COMMITTEE ON THE CHALLENGES OF MODERN SOCIETY
III
Hi
Illl
 1
LA CONFERENCE
     COMITE SUR LES DEFIS DE LA SOCIETE MODERNS

-------
          TECHNICAL REPORT

               OF THE

    CONFERENCE ON LOW POLLUTION

     POWER SYSTEMS DEVELOPMENT
NATO Committee on the Challenges of
           Modern Society
     Eindhoven, The Netherlands
      February 23, 24, 25, 1971
  Environmental Protection Agency
        Rockville, Maryland

-------
                          TECHNICAL REPORT

                              OF THE

                   CONFERENCE ON LOW POLLUTION

                    POWER SYSTEMS DEVELOPMENT



                        TABLE OF CONTENTS

CHAPTER                                                       PAGE

I.            Introduction                                    1-1

II.           FEDERAL MOTOR VEHICLE EMISSION GOALS FOR       II-l
              CO, HC, AND NOx BASED ON DESIRED AIR
              QUALITY LEVELS
                   Ronald Engel, Environmental Protection
                       Agency, USA

III.          THE ADVANCE AUTOMOTIVE POWER SYSTEMS PROGRAM  III-l
                   John H. Brogan, Environmental
                       Protection Agency, USA

IV.           THE POTENTIAL OF THE GAS TURBINE VEHICLE       IV-1
              IN ALLEVIATING AIR POLLUTION
                   Edward S. Wright, United Aircraft, USA

V.            THE STIRLING-CYCLE ENGINE                       V-l
                   R.A.J.O. van Witterveen, N. V. Philips,
                       The Netherlands

VI.           RANKINE-CYCLE POWER SYSTEM WITH ORGANIC-       VI-1
              BASED FLUID AND RECIPROCATING EXPANDER FOR
              LOW EMISSION AUTOMOTIVE PROPULSION
                   Dean T. Morgan, Thermo Electron
                       Corporation, USA

VII.          STIRLING ENGINE ACTIVITIES AT UNITED          VII-1
              STIRLING (SWEDEN)
                   Lars G. Ortegren, United Stirling,
                       Sweden

-------
 CHAPTER
                                                              PAGE
VIII.
IX.
X.
XI.
XII.
XIII.
XIV.
XV.
XVI.
NITROGEN OXIDE FORMATION IN THE COMBUSTION   VIII-1
CHAMBER OF THE INTERNAL COMBUSTION ENGINE
AND ITS SUPPRESSION BY MEASURES FROM
COMBUSTION TECHNOLOGY
     K. H. Newmann, Volkswagen, Federal
         Republic of Germany

A EUROPEAN CONTRIBUTION TO LOWER VEHICLE       IX-1
EXHAUST EMISSIONS
     Diarmuid Downs, Ricardo
         Consulting Engineers, United Kingdom

LOW EMISSIONS FROM CONTROLLED COMBUSTION FOR    X-l
AUTOMOTIVE RANKINE CYCLE ENGINES
     W. A. Compton, Solar Division of
         International Harvester Company, USA

HYBRID HEAT ENGINE/ELECTRIC SYSTEMS STUDY      XI-1
     Joseph Meltzer, The Aerospace
         Corporation, USA

ADVANCED TECHNIQUES IN ELECTRICAL VEHICLES    XII-1
     Bohers/Ducrot, Citroen Automobile
         Company, France

RESEARCH AND DEVELOPMENT ON A LITHIUM-       XIII-1
SULFUR BATTERY
     Elton J. Cairns, Argonne National
         Laboratories, USA

RESEARCH AND DEVELOPMENT PLAN OF ELECTRIC     XIV-1
CAR
     Shizume, Japanese Automotive
         Manufacturers' Association,  Japan

STUDIES BY FIAT ON THE ELECTRICALLY-DRIVEN     XV-1
AUTOMOBILE
     G. Brusaglino, Fiat, Italy

ELECTRICAL VEHICLES WITH FUEL CELLS:          XVI-1
WHY AND HOW?
     J. Beslier,  Society of Peugeot
         Automobiles, France

-------
                            ACKNOWLEDGMENTS









     The United States, as pilot country on the Air  Pollution Study




of the NATO Committee on the Challenges of Modern Society acknowledges




the assistance and hospitality of the Government of  The Netherlands




and The Philips Company who co-hosted the Conference.

-------
                              1-1




                           Introduction







The NATO/CCMS Conference  on Low Pollution Power Systems Development




was held in Eindhoven, The Netherlands,  on February 23-25,  1971.  The




Government of The Netherlands and Philips Company co-hosted the




Conference.






The total attendance for  the meeting was just under one hundred government




and industrial representatives from twelve countries, as well as non-NATO




international organizations.  The following countries were  represented:




The Netherlands, Belgium, Canada, Denmark, France, Germany, Great Britain,




Italy, Turkey, Sweden, Japan, and the  United States.  The Organization




for Economic Cooperation  and Development (OECD) and the European Economic




Community had observers in attendance.






The Conference was divided into two sessions with the first day reserved




for government representative presentations and discussions regarding




country programs and policies relative to the automotive pollution




problem.  The second and  third days were devoted to technical presentations




on research and development projects by both government and industry




attendees.






The "Summary of Proceedings of the Conference on Low Pollution Power Systems




Development"* was published by the U.  S. Environmental Protection Agency




(EPA)  in March 1971.  The Summary of Proceedings document gives an overview




of the Conference and includes brief summaries of each presentation made




at the Conference by government and industry representatives.






*This  document is available from the Environmental Protection Agency.

-------
                              1-2
This document includes the texts and figures of the technical presen-




tations that dealt with research efforts in the low pollution power




systems field.  Additionally,  the one technical paper dealing with the




rationale for the U. S. standards was also  included because of the




extensive amount of information on the relationship between air quality




and automobile emissions contained in the paper.   The U.  S. companies




who participated in the conference  are contractors to the Environmental




Protection Agency for low pollution power systems  development.  Other




discussion papers, which were  general in nature relating  to governmental




concern with the automobile pollution problems,  were not  included;




however, copies of these papers may be obtained from the  participants.




The summaries are included in  the Summary of Proceedings  document.

-------
                              II-l
                         Chapter II
FEDERAL MOTOR VEHICLE EMISSION GOALS FOR CO, HC, AND NO
                                                       A

            BASED ON DESIRED AIR QUALITY LEVELS*
                             by
               D. S. Earth, J. C. Romanovsky,
             E. A. Schuck, and N. P. Cernansky
             Environmental Protection Agency,
                            USA
            Presented at Eindhoven Conference by

                      Dr. Ronald Engel
 Assistant Director, Bureau of Criteria and Standards, APCO
             Environmental Protection Agency,
            Research Triangle Park, N.C., USA
*"Journal of the Air Pollution Control Association," Vol. 20,
 No. 8, August, 1970, pp. 519-523. Reprinted by permission of
 the Air Pollution Control Association,  Pittsburgh, Pennsylvania
 USA.

-------
                                    II-2
 Federal Regulations
                                 Background
     The Federal motor vehicle air pollution  control  program  has  its  origin
in the Clean Air Act, as amended in 1965.1'2   This  Act  borrowed heavily from
the motor vehicle air pollution control  program  already in  effect in
California '  and reflected testimony from  the automotive  industry that
similar controls were feasible for application nationwide.  The automotive
industry favored national  standards to forestall  a  proliferation  of regulatory
legislation at the state level.  National standards,  applicable to the 1968
                                              5
model year, were promulgated on March 31, 1966.   More  stringent  and  more
equitable national  standards,  applicable to the 1970  model  year,  were
promulgated on June 4, 1968.    In addition, on  February  10, 1970, Secrel
Finch published an  advance notice of proposed rulemaking indicating the
Department's
model years.'
Department's  intent to  adopt  more  stringent  standards  for  the  1973  and  1975
            7
                            Air Quality Criteria
     Under the 1967 amendment to the Clean Air Act,8 the Department of
Health, Education,  and Welfare is required to publish air quality criteria
Documents.  With criteria for photochemical  oxidants,  carbon monoxide,
hydrocarbons,   and nitrogen oxides either published or underway it now
becomes possible to consider future motor vehicle emission standards in

-------
terms of desired air quality consistent with information in the pertinent
air quality criteria documents.  The resumes of these documents describe
what in the Secretary's judgment are the concentrations and exposure time
which cause or contribute to, or are likely to cause or contribute to,
air pollution which endangers human health or welfare.
                      Desired Air Quality Goals
     Based on the criteria resumes and observed aerometric relationships,
Table I has been constructed to show the maximum values which would seem
to be consistent with health-related criteria.  Minimum safety margin
considerations have been incorporated into these levels.
Calculating Motor Vehicle Emission Goals
     Using the information in Table I, simple roll-back techniques, like
those used by California, could be used to calculate needed emission
reductions.  These techniques, however, involve a number of assumptions
which may not be entirely valid.  For example, inherent in the techniques
employed is the assumption that the increase in atmospheric concentrations
of primary pollutants will be directly proportional  to the growth in
emissions of the contaminant.  Also, the roll-back techniques employed
by California assumed a linear correlation to exist between the severity
of the manifestation, specifically oxidant index, and reactive hydro-
carbon emissions to the atmosphere.  The complexities of the photochemical
alterations involving reactive hydrocarbons in the atmosphere are too
great to permit such an assumption, a^ priori, because at a minimum the
role of the oxides of nitrogen must be taken into consideration.
     Clearly what is needed is to fully relate all  these effects  to
emissions and to predict future events in growth and required control

-------
                                    II-4
                                                              9-11
     Table I.  Desired Air Quality Goals for Health Protection


	Contaminant	       Concentration,  g/m^         Average, hr

Carbon Monoxide                <_ 10,000 (9 ppm)                 8

Photochemical Oxidants         i 125 (0.06 ppm)                 1

Nitrogen Dioxide3              4190 (0.10 ppm)                 1

A
 Preliminary estimated value  pending review of technical  report on direct

 health effects of nitrogen dioxide  and publication  of Air Quality

 Criteria  Document.

-------
                               II-5
using a comprehensive simulation model complete with modules reflecting
all input variables, meteorological variables including air transport
parameters, and most importantly-where oxidants, eye irritants, and
aerosols are concerned-the chemical kinetics describing the multiplicity
of the reactions which occur.  Factual data on existing air quality are
also required in order to validate and adjust the model as necessary.
Several studies are currently investigating the generation of these
simulation modelsJ2,13  Unfortunately, none of these models is yet
ready for general application.
         Relating Air Quality Goals for Nitrogen Oxides and
              Nonmethane Hydrocarbons to Oxidant Goals
     In the absence of a validated model which incorporates atmospheric
chemical mechanisms, a more restricted diffusion model may be employed
in which empirical relationships between primary atmospheric pollutants
and secondary reaction products are known.  Such relationships have
been established between photochemical oxidants and nonmethane hydro-
carbons .11  Analogous relationships have also been developed for the
oxides of nitrogen separately as well as in combination with nonmethane
hydrocarbons.  One important premise, based on experiences reflecting
the cause and effects associations observed to relate Los Angeles with
Pasadena, is that hydrocarbon emissions between the hours of 6 and 9 a.m.
result in peak oxidant concentrations 2 to 4 hours later.H  This is
particularly true on the days of greatest interest-days which because
of favorable meteorology favor maximum production of oxidant.
     Figure 1 includes an analysis of 3 years of data which shows that
on those several days a year when meteorological conditions were most
conducive to the formation of photochemical oxidants, nonmethane

-------
                                             II-6
   0.30
   0.25
 E
 Q.
 Q.
§ 0.20
O
LLJ
LU
O
X
   0.15
   0.10
   0.05
                                                          LOS ANGELES
                                          LOS ANGELES^X-TDENVER

                                   WASHINGTON*^
                                              " * LOS ANGELES

                                    \ A PHILADELPHIA
                                   LOS ANGELES
   PHILADELPHIA-


      PHILADELPHIA
     WASHINGTON /

[_ WASHINGTOIU/pH,LADELpH|A

 WASHINGTON
             7
        WASHINGTON, ^ ^ A AA A
               A

               t
               AA
        ** ^ /^ i~A '
  A


A


• A       A       Am


     A



             ^    \  .

   A    *•
« A  A  A  A  A  A    AA
       0          0.5          1.0          1.5          2.0          2.5

            6-9 A.M. AVERAGE NONMETHANE HC CONCENT RATION, ppm C


        Figure 1.  Maximum daily oxidant as a function of early morning non-

                 methane hydrocarbons; 1966-1968 CAMP stations; May through
                 October 1967 for Los Angeles.11

-------
                                II-7
hydrocarbon concentrations of 200 jjg/m3 (0.3 ppm C) for the 3-hour
period from 6 to 9 a.m. might produce an average 1-hour photochemical
oxidant peak concentration of up to 200 pg/m3 (0.10 ppm) 2 to 4 hours
later.I1  The hydrocarbon measurements were confined to 200 ;jg/m3, or
above, because of instrumentation limitations.  However, if the functional
relationships between the hydrocarbon and photochemical oxidant measure-
ments were extended to include the levels presented in Table HI as the
highest value consistent with health-related criteria, the corresponding
hydrocarbon concentration would be approximately 125 ng/m3 (0.10 ppm C).
     Figure 2 presents similar data relating 6 to 9 a.m. average nitrogen
oxides concentrations to maximum daily 1-hour average oxidant concentra-
tions appearing 2 to 4 hours later.  The envelope enclosing the data
presumably reflects those several days a year when meteorological
conditions were most conducive to the formation of photochemical
oxidants.  On such occasions, if the functional relationships between
the oxides of nitrogen and photochemical oxidant measurements were
extended to include the level of oxidants presented in Table 1, the
corresponding nitrogen oxides concentrations would be approximately
49 jjg/m3 as N02 (0.026 ppm).
     Figure 3 is a three-dimensional presentation of the same data
shown in Figures 1 and 2J1  The isopleths traced through the different
oxidant levels reflect the fact that the ratio of nonmethane hydro-
carbon to nitrogen oxides is important, along with the absolute level
of the separate independent variables, in determining the maximum
level of oxidant produced on days of favorable meteorology.  Presenta-
tion of the data in this form is useful  for qualitatively considering
the trade-offs between unilateral and joint control of hydrocarbons

-------
                             Table III.  Relationship of Air Quality Goals

                                    to Motor Vehicle Emission Goals



Emission
Carbon Monoxide
Nitrogen Oxides3


Type of
Effect
Direct
Direct


Health-Related
Air Quality Goals
10 mg/m3 (9 ppm)
190 ug/m3 (0.1 ppm)

Health-Related
Averaging Time,
hr
8
1
Emission
Goals from all
Vehicles to Achieve
Desired Air Quality,
g/mi
6.16
0.38

Control
Requirement for
All Vehicles, %
92.5
93.6
  (N02)            (NOe)       (N02)
Hydrocarbons
Indirect   125 yg/m  of oxidant
            (9.06 ppm)
1
0.15
99.0
aThe NOX emission goal is identical to the N02 emission goal since in Los Angeles, the city at maximum risk,

 all NO is converted to N02-
                                                                                                             03

-------
                     II-
  0.30
 D.
 Q.
   ).20
LU
cc
13
O
X
  0.10
o
5
3
                 1
          W - WASHINGTON
          P-PHILADELPHIA
          D- DENVER
          • - 0.3 TO 0.9 ppm C
          O-l.OTOS.OppmC
            • W
   -«-<•»• *I*    *O »»OO   O  •   OO
        •  P   «O         o  r> O
 Q •••• •Q«V»* ••Q* • •  •    (J  «U»
•           ^ O     O
•o  ••• ••••• o»»»   0*0 oo      o
                                                               00
                                                            o
                                                            o
                                    I
                                             I
                                        I
I
      0                 0.10                0.20                0.30

                6-9 A.M. AVERAGE TOTAL I\IOX CONCENTRATION, ppm


     Figure 2.   Maximum daily oxidant as a function of early morning nitrogen
               oxides; 1966-1968 CAMP stations.

-------
                          11-10
   1.2
I
S1-0
z 0.8
o
UJ

< 0.6
o
z
o
2
UJ
                                                        OXIDANT, ppm
                                                         •  0.07
                                                         A  0.10
                                                         •  0.12
                                                            0.14
                                                            0.16
   0.2
o>
                   PREDICTED LIMITS FOR 0.10
                         ppm OXIDANT


                  ' PREDICTED LIMITS FOR 0.05 ppm OXIDANT

                  	I	I	
                                                              I
     0         0.10                   0.20                   0.30
                  6-9 A.M. AVERAGE IMOX CONCENTRATION, ppm

  Figure 3.  Upper limit on maximum daily 1-hour average oxidant as a
            function of early morning nonmethane hydrocarbons and nitrogen
            oxides; June, July, August; Philadelphia, Washington, Denver;
            1966 through 1968.11

-------
                               11-11
and oxides of nitrogen.  It must be emphasized that Figure 3 at this
time is based on relatively few points.  Thus the curves can be better
defined as further data become available.  In view of this, the drawing
of quantitative conclusions from this graph is not wholly justified.
     Qualitative examination of Figure 3 shows that, depending on the
specific existing concentrations of oxidants, nitrogen oxides,  and
nonmethane hydrocarbons, the optimum control strategy for the most rapid
reduction of oxidants may be unilateral control of hydrocarbons,
unilateral control of nitrogen oxides, or joint control of both
nitrogen oxides and hydrocarbons.  It must be pointed out, however,
that all possible oxidant control strategy options are not available
to us.   Recall that the postulated health-related desired air quality
for N(>2 is 190 yg/m3 (0.10 ppm) for a 1-hour average (Table I).  Thus
no oxidant control strategy is acceptable which will not simultaneously
lower the N0£ values to the desired air quality.  Furthermore, our
estimate of available nitrogen oxides control technology is that there
is at this time no proven or postulated device capable of lowering
nitrogen oxides emissions from internal combustion engine mobile sources
sufficiently to achieve air quality concentrations of =0.026 ppm, which
we have shown would be required to control oxidant air quality concen-
trations to  125^ig/m3 (0.06 ppm) by unilateral nitrogen oxides control.
     We conclude that the most rational control strategy at this time
is to control nitrogen oxides to achieve the desired health-related air
quality for N0£ and to rely on control of nonmethane hydrocarbons to
achieve the desired oxidant air quality goal.  Subsequent calculations
will be based on this approach.

-------
                               11-12
     Again,  caution must be exercised  with  respect  to  the  permanency
of the recommended control  approach.   In  particular it must be  kept
in mind that all  factors are not  known relative  to  atmospheric  inter-
actions.   Because of this lack  of information  it is most difficult to
develop a comprehensive systems approach  to this problem.   A good
example of the need for caution was  discovered when the changes in the
Los Angeles  atmosphere between  1962  and 1967 were investigated.14
During this  5-year period,  in the May  through  October  months, the
6 to 9 a.m.  hydrocarbons decreased by  4 percent  and the nitrogen
oxides increased by 25 percent. 'Additionally  the maximum daily
1-hour average oxidant decreased  by  11  percent.   Focusing  attention
on only the  decrease in maximum oxidant level, however, does not
describe all the changes.  Thus this maximum oxidant decrease was
associated with an increase in  the number of days on which the  maximum
oxidant exceeded 0.1 ppm.  The  result  is  that  a  trade-off has been made,
i.e., an 11  percent decrease in the  maximum for  a 10 percent increase
in dosage.  This latter increase  in  dosage, and  to  a large extent the
decrease in maximum, may be related  to the  altered  ratio of HC  to NOX.
Thus control approaches which suggest  allowing increases in ambient
levels of nitrogen oxides are definitely  subject to question.  As a
result of this and other examples it is apparent that  we must be always
alert to the atmospheric stiuation on  a year-to-year basis.  Furthermore,
we must be prepared to modify any adopted control approach as new infor-
mation becomes available.
     In addition to these examples of  lack  of  information, there are
other contributing factors  which  may demand changing any projected
emission values at some time in the  future. During the next five

-------
                                11-13
years, for example, we will re-examine the Photochemical Oxidant and
Hydrocarbon Criteria documents.  In the intervening years it is  quite
possible that more precise measuring methods may be developed.  Thus
the oxidant criteria may become ozone criteria and the present
non-methane hydrocarbon measurement may be stated in terms of reactive
hydrocarbons.  Such changes will, of course, affect the emission goals
as calculated in this presentation.
     Summary of Calculations Relative to Motor Vehicle Emission
             Goals Designed to Produce a Health-Related
                        Acceptable Air Quality
     Calculation of the motor vehicle emission goals for CO, HC, and
NOX requires application of two equations.  The first of these equations
permits a calculation of the fractional reduction in ambient concentra-
tions necessary to achieve a specific health-related air quality.^
The equation is stated in terms of present air quality, desired air
quality, background concentrations, and the projected growth in
emissions.  Mathematically stated, it has the form:
             R = (GF)   (PAQ) - (DAQ)              m
                 TGF)   (PAQ) - (B)                u;
where R is the calculated fractional reduction required; GF is the
emission growth factor; PAQ is the present maximum air quality; DAQ
is the desired air quality; and B is the background concentration.
The second equation employs the result from Equation 1 to restate
the air quality reduction in terms of mobile emission rates.  Mathe-
matically expressed, this second equation has the form:
           (PER)   (1-R)  =  DER                   (2)
where PER is the percent emission rate giving rise to the present air
quality (PAQ); the expression 1-R is the statement of how much of the

-------
emission can be allowed without exceeding the desired air quality.
The multiplication, i.e., (PER) (1-R) thus yields the desired emission
rate  (DER) consistent with the desired air quality (DAQ).
      In order that the calculated emission goals be protective of
health, certain principles must be adopted relative to the values
inserted in Equations 1 and 2.  Thus, present air quality in Equation
1 must be representative of the worst situation observed in the entire
United States.   For HC and NOX values to be associated with health
effects of oxidants and eye irritants, the worst situation presently
occurs in Los Angeles.16  (The 6 to 9 a.m. maximum average concentrations
6bserved during the-high oxidant potential-season^were 5.3 ppm for non-
methane HC and 0.62 ppm for NO^.)             Los Angeles also had the
highest yearly 1-hour average N02 concentrations (0.69 ppm); thus, when
the health effects of N02 are being considered^air quality levels in
Los Angeles are used.  With respect to CO health effects, however,
the city of Chicago,  with a maximum 8-hour average of 44 ppm, represents
the .worst known case  in the United States.16  In terms of growth factors,
we must again assume  the maximum predicted growth in order to provide
adequate health protection.  For a time period from 1967 to ten years
beyond the year of application of these motor vehicle emission goals, the
maximum predicted growth factor for mobile emissions must be used.  Again,
to afford the maximum health protection, the background concentrations
used in Equation 1 must be the maximum supported by scientific inquiry.
For these purposes a  value of 1 ppm  was used for CO,10 0.1 ppm for HC,11
and 0.004 ppm for NOX.  The lowest supportable values for the desired
air quality,  as derived from the appropriate criteria documents were
used in Equation 1 and have been presented in Table I.

-------
                                11-15
     For the present emission rate (PER), we have related derived
emission rates to a 1967 rate of 82.6 grams per miles (g/mi) for CO,
14.84 g/mi for HC, and 5.93 g/mi for NOX.17  The HC and NOX rates are
representative of the 1967 vehicle mix in Los Angeles, while the CO
rate is representative of the 1967 vehicle mix in Chicago.
     The one factor which this treatment does not speak to is the
possible effect of the failure of vehicles to meet the stated emission
goals when they become standards.  Thus, surveillance data from vehicles
presently in consumer use show current Federal standards are exceeded on
the average by 13 percent in terms of CO emissions and 25 percent in terms
of HC emissions.!8  This, however, is a difficult factor to take into
account because it may change in the future.  If we ignore such past
failures, however, we are in essence announcing that before these emission
goals are established as standards we will have devised methods that will
force automobile manufacturers to provide control devices which will enable
vehicles in consumer use to meet required emission standards.  The pro-
jected emission limitations derived here will assume that devices and
control systems in light duty vehicles in use continue to exceed the
standards by the stated percentages.
     To proceed to the calculation of specific mobile emission goals
needed to achieve desired air quality goals, it now becomes necessary
to designate a target year for implementation of the required mobile
emission restrictions.  Selection of a target year should be based
upon the evaluation of many important factors such as:  the need to
meet the desired air quality goals at the earliest possible time;
the state-of-the-art of control  technology; and future advances in
our understanding of atmospheric photochemical processes.  It is

-------
                                11-16
 beyond the scope of the paper and the competencies of the authors
 to  evaluate all of the above factors and their interactions in detail.
 Thus, to avoid becoming embroiled in high-level policy questions which
 still remain to be resolved, we have arbitrarily selected 1980 as the
 target year on which to base our calculations.  In the event the decision
 is  made to select an earlier year for implementation it will be an easy
 matter to adjust the calculations accordingly.  Of course, the earlier
 the calculated mobile emission restrictions can be put into effect, the
 earlier we will be able to achieve desired air quality goals and the less
 stringent will be the necessary mobile emission limits.  In accordance
 with the above decision we used a 1967-1990 mobile emission growth
 factor.17'19
     In reviewing the required information items it will be noted that
 the  percentage contribution of current mobile sources to existing air
 quality, the projected mobile growth rate, and the background concentra-
 tions of each pollutant are the least well-defined.  Thus, background
 concentrations of HC are not delineated with any degree of accuracy.
 The same can be said about growth rates of mobile emissions.  Further-
 more, the percentage contribution of mobile emissions to total  emissions
 must at this time be based on logic rather than known fact.  In the case
 of Los  Angeles there seems to be general agreement that in the downtown
 area the smog-season emissions of CO,  HC, and NOX are mainly from mobile
 sources.   Consequently, the calculated emission goals were based on
 the latter assumption.   A similar assumption was made relative to the
Chicago data,  i.e.,  that the bulk of the ambient CO concentrations
steam from the mobile source.  If this assumption is incorrect, it
follows  that those stationary sources  which contribute significantly
to the  downtown area concentrations of CO, HC, and NOX must be

-------
                                11-17
controlled to the same degree as motor vehicles.  To the extent that
these sources are uncontrollable the restriction of the motor vehicle
would have to be increased.
     Table ilcontains a summary of the numerical values used in cal-
culating 1980 mobile emission goals.  Figures 4, 5, and 6 express the
results of the calculations for desired mobile emission rates for CO,
HC, and NOX, respectively, as a function of the desired air quality for
each pollutant without taking into account the potential deterioration
of the control measures for CO and HC.  In these figures a range'of
values above and below the Table 1 values of desired air quality are
shown since the exact health-related values are still  under study and
may be changed in the future.  Table  m contains  the  1980 motor vehicle
emission goals for CO, NOX, and HC required to achieve the desired
air quality after taking presently observed CO and HC control measure
deterioration into account.

-------
Table II.  Numerical Values of Parameters used in Calculating



             Mobile Emission Goals9"11'16'17'19
Parameter
1967 maximum air quality values
reflated to direct health effects
1967 maximum air quality of
oxidant precursors, (6 to 9 a.m.
average)
1967 average emission rates for
all motor vehicles
Maximum background concentration
Maximum growth factor of mobile
emissions
Air Quality Goals
CO
51 mg/m3 for 8
hr, Chicago
™
82.6 g/mi
1 mg/m3
(1 ppm)
2.18
10 mg/m3 for 8
hr (9 ppm)
HC
-
3.5 mg/m3 as
CH4 (5.3 ppm C),
Los Angeles
14.84 g/mi
0.1 mg/m3 as CH4
(0.1 ppm)
2.18
125 yg/m3 of
oxidant (0.06
ppm), 1-hour
average
NOX
-
1170 yg/m3 as
N02 (0.62 ppm),
Los Angeles
5.93 g/mi
8 yg/m3 as N02
(0.004 ppm)
2.18
125 yg/m3 of
oxidant (0.06
ppm), 1-hour
average
N02
1300 yg/m3 for 1 hr,
(0.69 ppm), Los Angeles
—
-
8 yg/m3 as N02
(0.004 ppm)
2.18 I
a
190 yg/m3 for 1 hr
(0.1 ppm)

-------
                                   11-19
   11
                       10
   10
CO
 E
 o>
 E

d  9
o
LU
LU
                          E
                          a.
                          a.
                      8  8
                         LLI

O
X
00
                         O
                         X
                         CO
                            I
I
      (4          5678!

                    TOTAL MOBILE CO EMISSION RATE, g/mi
                                                      •
      Figure 4.  Relationship of 8-hour average CO values to mobile emission
                rates.

-------
                              11-20
                                                                 0.070
        0.05       0.10       0.15       0.20       0.25
             TOTAL MOBILE HC EMISSION RATE, g/mi
                                                                 0.045
0.30
Figure 5.   Relationship of HC mobile emission rate to 1-hour average
          oxidant.

-------
                              11-21
       0.2        0.4        0.6        0.8         1.0
            TOTAL MOBILE IMOX EMISSION RATE, g/mi

Figure 6.   Relationship of 1-hour ambient IM02 levels to mobile IMOx
          emission rates.
1.2

-------
                                      11-22
III.  Summary
     The historical background of the development of Federal  mobile emissions
standards has been presented.   Based on issued and planned Criteria documents,
desired air quality goals for  health protection have been set at the following
levels:
       CO                      4 10,000 yg/m3 (9 ppm)  - 8-hour average
       Photochemical  oxidants   4125 yg/m3 (0.06 ppm)  - 1-hour average
       N0£                     4190 yg/m3 (0.10 ppm)  - 1-hour average
The complexities inherent in the control  of photochemical  oxidants have been
discussed and a control  strategy involving joint control  of hydrocarbons and
nitrogen dioxides adopted.   Using a  modified roll-back calculation with a 1967
baseline, the following  1980 total motor vehicle emission goals by which to
achieve desired air quality goals have been derived:
       CO              6.16 g/mi
       Hydrocarbons    0.14 g/mi
       NOX             0.40 g/mi
The calculation of future motor vehicle emission goals should be a continuous
process with new data being used as  they become available.

-------
                                     11-23

                                 REFERENCES

1.  Clean Air Act, Public Law 88-206, 88th Congress, 1st Session, December 17,
                                           3-1-2-
    1963.  In:  U.S. Statutes at Large, 77_:932-401  (1964J.


2.  Motor Vehicle Air Pollution Control Act, Public Law 89-272, 89th Congress,

    1st Session, October 20, 1965.  In:  U.S. Statutes at Large, 79^:992-1001

    (1966).


3.  "Technical Report of California Standards for Ambient Air Quality and

    Motor Vehicle Exhaust," California Dept. of Public Health, Berkeley,

    California, (1960).


4.  "Technical Report of California Standards for Ambient Air Quality and

    Motor VEhicle Exhaust.  Supplement Number 1, Crankcase Emission Standard,"

    State of California, Department of Public Health, Berkeley, California

    (Aug. 1961).


5.  "Control of Air Pollution from New Motor Vehicles and New Motor VEhicle

    Engines," Federal Register (Washington), Part II, 31_(61): 5170-5238

    (March 30, 1966).


6.  "Control of Air Pollution from New Motor Vehicles and New Motor Vehicle

    Engines," Federal Register (Washington). Part II, 33_( 108):8303-8324

    (June 4, 1968).


7.  "Control of Air Pollution from New Motor Vehicles and New Motor Vehicle

    Engines," Federal Register (Washington), Part II, 3j[(28):2791 (February

    10, 1970).


8.  Air Quality Act of 1967, Public Law 90-148, 90th Congress, 1st Session,

    November 21, 1967.  In:  U.S. Statutes at Large, 81_:485-507 (1968).

-------
                                       11-24
 9.  National Air Pollution Control  Administration,   Air QuaUty Criteria for
     Photochemical Oxidants. U.S. DHEW, PHS,  EHS,  Washington,  D.  C.  (March 1970),
 10.  National Air Pollution Control  Administration,  Air Quality Criteria for
     Carbon Monoxide, U.S. DHEW, PHS,  EHS,  Washington,  D.  C.  (March  1970).
 11.  National Air Pollution Control  Administration,  Air Quality Criteria for
     Hydrocarbons, U.S. DHEW, PHS, EHS, Washington,  D.  C.  (March  1970).
 12.  Wayne, L., e_t a1_., "Modeling Photochemical  Smog on a  Computer for
     Decision-Making," Paper presented at the APCA annual  meeting, St.  Louis,
     Mo. (Jwne 14-18, 1970).
 13.  Eschenroedor, A. Q. and J.  R. Martinez,  "Mathematical  Modeling  of  Photo-
     chemical Smog," General Research  Corp.,  Santa Barbara, California  (December
     1969).
 14.  Los Angeles County Air Pollution  Control  District  Data 1962  and 1967,
     Los Angeles, California
 15.  Larsen, R. I.,  "A New Mathematical  Model  of Air Pollutant Concentration
     Averaging Time  and Frequency,"  J.  Air  Pollution Control Assoc., Ijhl
     (January 1969).
16.  Larsen, R. I.,  "Relating Air Pollutant Effects  to  Concentration and
     Control," J.  Air Pollution  Control  Assoc.,  20:4 (April 1970).
17.  Kramer, R. L. and N.  P. Cernansky,  "Motor Vehicle  Emission Rates,"  U.S.
     DHEW,  PHS, EHS,  National  Air Pollution Control  Administration.   Durham,
     North  Carolina  (internal  document).
18.  Hocker, A.  J.,  "Exhaust emissions  from Privately Owned 1966-1969 California
     Automobiles:  A  Statistical  Evaluation of Surveillance Data," California
     Air Resources. Laboratory, Los Angeles, Calif.,  Supplement to Progress
     Report Number 17  (February 6, 1970).

-------
                                      11-25




19.  Landsberg, H. H., L.  F.  Frischman,  and J.  L.  Fisher, Resources in



     America's Future, Patterns  of Requirements  and Availabilities 1960-2000,



     John Hopkins Press, Baltimore (1963).

-------
                            III-l
                         Chapter III
                   THE ADVANCED AUTOMOTIVE

                    POWER SYSTEMS PROGRAM



                               by
                       John J. Brogan
Director, Div. of Motor Vehicle Research and Development, APCO
               Environmental Protection Agency
                   Ann Arbor, Michigan, USA

-------
                                Ill-2
 Introduction
 The Advanced Automotive Power Systems Program is an outgrowth
 of concern by our nation, led by President Nixon, that improvement
 of air quality in the United States is markedly dependent on the
 elimination of the automobile as a significant source of air
 pollution.  In the United States, the internal combustion engine
 in our automobiles contributes about half of the total pollutants
 from all sources to our air environment.  Figure 1 illustrates
 the automobile's share of the blame in the United States.  Based
 on 1970 estimates, nearly 70% of the carbon monoxide from all
 man-made sources comes from the automobile, also 40% of the
 hydrocarbons and about 40% of the nitrogen oxides appear from
 this source.  Less than 1% of the particulates and less than
 0.10% of the sulfur oxide emissions nationwide are due to this
 source.

 One result of the nationwide concern over the automobile has
 been the tightening of Federal exhaust emission standards that
 must be met on new automobiles.  This tightening of the standards
 is apparent by inspection of Figure 2.  This Figure shows the
 Federal exhaust emission standards through 1973 for the three
 major pollutants and estimated values for 1975/76.  The final
 values for these latter years have not been finalized, as yet.
 By model year 1975, a 90% reduction in hydrocarbons and carbon
 monoxide from the 1971 levels are shown, and, by 1976, the
 nitrogen oxides levels will be equivalent to a 90% reduction from
 the uncontrolled levels measured for the 1971 model year vehicles.
 There is considerable doubt that the conventional spark-ignition
 gasoline-fueled reciprocating internal combustion engine can be
 'cleaned up' adequately to meet all of these forthcoming standards.
As a result, alternate power systems which are inherently clean
when compared with the conventional engine must be considered as
potential replacements.

-------
                           III-3
MILLION
  TONS
    13O
   NATIONWIDE  EMISSIONS  ESTIMATES
                    •1970-
no

100

9O



7O

6O



4O

3O

2O

1O
                                             ALL SOURCES
                                                 •
                                               MOTOR
                                              VEHICLES
POLLUTANT- CO
                  HC
                     NO
PARTICULATES   SO
                          Figure 1.
FEDERAL  MOTOR VEHICLE EMISSION STANDARDS
                  grams/mile
        1971    1972    1973    1975    1976
  HC
  CO
  NOL
4.6
47.0
—
3.4
39.0
-
3.4
39.0
3.0
0.46
4.7
3.0
0.46
4.7
0.4-0.6
                          Figure 2.

-------
                                III-4
 Bases for the Program
 A panel of scientists originating in the Executive Office of the
 President was formed in 1969 to evaluate industry efforts in our
 country to develop alternate power systems.  The conclusion
 reached by the Panel was that industry was not making a serious
 effort to make such developments.  Lacking sufficient industry
 interest, a Federal research and development program was
 recommended.  This program was announced by President Nixon on
 February 10, 1970 (see Figure 3) and the program was underway
 by July.  In addition, in his February message the President
 announced another Federal program to stimulate industry by
 providing financial incentive to groups who independently develop
 their own alternate power systems.   This latter program is called
 the Clean Car Incentive Program.  Thus, the two parallel approaches
 to develop alternate power systems  appear as in Figure 4.  The
 total package consisting of the Research and Development Program
 and the Incentive Program is called the Advanced Automotive
 Power Systems (AAPS) Program.

What are the goals of the Advanced  Automotive Power Systems
 Program?   The program is intended  to provide insurance to our
nation that if a practical and virtually pollution-free power
 system could be developed, this development would culminate in
 a demonstrated system by 1975.   Achievement of this goal can be
 accomplished by demonstration of a  complete power system either
developed directly under Federal government monies or, by similar
type demonstrations whose development is sponsored by industry,
perhaps as a result of stimulation  brought about by the existence
of the Research and Development Program or of the Incentive
Program,  or both.

In addition,  the AAPS Program will  provide important information
to the Administrator of the Environmental Protection Agency on

-------
                   III-5
EXCERPT ROM MESSAGE ON ENVIRONMENT
              FEBRUARY 10,1970
      u... with the goal of producing an
      unconventionally powered, virtually pollution
      free automobile within five yeans."
                 — PRESIDENT NIXON
                  Figure 3.
    DEVELOPMENT OF ALTERNATE
     POWER SYSTEMS APPROACH
TWO PARALLEL PATHS
     FEDERALLY  SPONSORED R/D
     INCENTIVE PROGRAM FOR PRIVATE DEVELOPMENT
                  Figure 4.

-------
                                 III-6
 the  technical feasibility of meeting the stringent 1975/76
 emission  standards with use of alternate power systems.   Such
 information aids in judgments made by the Administrator  on
 possible  waiver requests to postpone enforcement of the
 standards.

 Research  and Development Program
 First,  the research and development part of the Program  will
 be discussed.  What are candidates for development in this
 Program?  There are five types of power systems that were
 initially part of the Program when it began in July 1970.
 These include the gas turbine, two kinds of hybrids—heat engine/
 electrics and heat engine/flywheel systems—Rankine cycle power-
 plants  and the all-electric system.  There are two additional
 systems,  the stratified charge engine and the Diesel engine which
 are  serious contenders for candidacy.  The particular versions of
 these last two systems under consideration are reasonably well-
 developed and are being considered as candidates because of
 their nearer term potential for meeting the 1975/76 emission
 standards when mass produced.  Thus, it is these two engines
 which we  expect will provide important information on whether
 alternate systems can meet the standards and this information
 will be available very early in the Program.

 Figure  5  lists all of the candidate systems and the stratified
 charge  and Diesel.   In general, the candidates were selected
 because the technology to improve, to improvise, and to  further
 develop each, exists in abundance in the United States.   In
 addition, independent research sponsored by private industry in
 the United States had been underway for some time on most systems
 selected.  Thus, the interest to further develop the systems was
 there.   Additional candidate systems may be brought into the
 Program.  We are continually reassessing new developments on other
 systems including developments made within our country and
worldwide.

-------
                                 III-7
It is pointed out that the Stirling cycle powerplant has been
seriously considered for candidacy since the Program work began.
The considerable strides made in the Netherlands and Sweden in
making the Stirling cycle powerplant a competitive system have
been followed closely by our staff and our Technical Advisory
Committee.  However, we are not ready as yet to bring this system
into the Research and Development Program.

The all-electric is shown apart from the other candidates in
Figure 5.  We do not see the complete development of the all-
electric system early enough to meet the 1975 date because the
type of batteries with adequate power and energy densities
needed to compete in performance and cost with the conventional
engine are presently in the basic research stage.  The battery
systems of special  interest herein are the high temperature
alkali-metal battery systems such as Lithium-Sulfur and Sodium-
Sulfur.  We recognize that there are several types of batteries
in use today for all-electric automobiles and small vans.
However,  in order to meet the driving requirements of the
American public, the development of completely new and more
powerful batteries  is needed.  As a result of these considerations
the all-electric car is not a candidate to meet the 1975 goal,
rather we see its development by 1978.  So, one may ask why is
this system in the  Program?  It  is in the Program because
eventually these new and better battery systems must be developed.
The cost of development is very high as is the risk.  For
industry at the present time, the costs and risks appear to out-
weigh the ultimate  advantage of having these batteries available.
By supporting the research and early development with government
monies we see a point being reached where industry will pick up
the further development of these batteries.

On what basis will  any of these propulsion systems be brought
through the development phases into complete system hardware?

-------
CANDIDATE  POWER  SYSTEMS
  HEAT  ENGINE/ELECTRIC HYBRIDS

  2nd GENERATION  ICE's-STRATIFIED CHARGE  ENGINE
                      DIESEL ENGINE

  RANKINE CYCLE

  HEAT  ENGINE/FLYWHEEL HYBRIDS
  ALL  ELECTRIC

                   Figure 5.

-------
                                III-9
The criteria which we consider are summarized in Figure 6.
Performance requirements are defined for the system and for
individual components.  Based on the designs formulated and the
tests conducted, we must have confidence that these requirements
can be met.  The characteristics of the largest selling model
type automobile in the United States for year 1970 form the
basis for engine weight and volume constraints, road performance
requirements and fuel economies.  Of course the exhaust emissions
must meet or do better than the emissions standards of 1975/76.
The safety characteristics of the automobile with the new engine
are another important consideration.  For example, it is de-
sirable that the working fluid in the Rankine cycle engine be
non-toxic and non-flammable.  The flywheel design must be such
that any structural failure of the flywheel itself not lead to
catastrophic failure outside of the confines of its casing.
The socio-economic impact noted on the slide relates to the
criteria that engine exhaust odor and noise must not be
ojectionable according to contemporary standards and that the
system cost to build in-quantity production be competitive with
the conventional engine with its exhaust treatment devices
installed on the latter.  National impact is listed last, yet,
it ranks as number one on anyone's list of criteria.  As one
example of how we evaluate the impact, let us consider impact
of the all-electric automobile.

Figure 7 shows the projected power generating capacity in the
United States beyond year 2000.  In some of the most densely
populated areas of our country, such as New York City, we find
that the demands for electricity, at times, are not being met
by the supply.  This is a city without any appreciable electric
car popultion.  On a nationwide basis, if all the vehicles on
the road today were all-electrics, the total demand for
electricity would exceed the existing capacity to supply by

-------
                   111-10
POWER SYSTEM  ACCEPTANCE CRITERIA


  • SPECIFIED SYSTEM  AND COMPONENT
   PERFORMANCE REQUIREMENTS

  • VERY LOW EXHAUST EMISSIONS

  • SAFETY

  • SOCIO-ECONOMIC  REQUIREMENTS
   NATIONAL IMPACT
                 Figure 6,
     PROJECTED POWER GENERATING
     CAPACITY  IN THE UNITED STATES
         O 2
         ill *-
         >105
         £ 8
         O 6
          10.4;
                   TOTAL POWER
                   CAPACITY-
-NATURAL GAS
-OIL
LHYDROELECTRIC
           1960 1970 1980 1990 2000 2010 2020

                    YEARS

                Figure 7.

-------
                                III-ll

about 40%.  In addition, the technology to control sulfur
oxides from fossil fuel burning power stations is still being
developed.  Further, the nuclear source curve shown in Figure 7
is very optimistic and since drawing this Figure, one year ago,
our experience has been that as time goes on this cross-over
point will move farther to the right.  Therefore, if recharge-
able alkali-metal batteries were fully developed right now and
all-electric automobiles produced in-quantity we would
aggravate our existing electrical power capacity problem and,
in addition, temporarily transfer the major sources of pollution
from a mobile to a stationary source.  So then these rre some
of the factors that we consider in judging whether a candidate
powerplant should go beyond the paper study phase and into
hardware, and, of equal importance, these factors influence the
timing of certain research work and the pace of development of
some candidates.

Individual Candidates and Their Status
Now let's take a closer look at the candidates and their status.
Each candidate was at a different state of development when the
Program began, and depending on our success in achieving
technological breakthroughs, each may enter the system hardware
phase at  different times during the next five years.  The
first 18-month phase of the Program is intended to be a period
of evaluation wherein the systems and their components are to
be designed, the critical components bench tested and decisions
made on whether to proceed to first generation hardware.

Rankine Cycle Engine.  Consider the Rankine cycle engine as
illustrated in Figure 8.  In this engine, there is an external
combustor and an enclosed working fluid which is heated,
expanded  to do work, then condensed to a liquid, with the fluid
being continuously recycled.  There are three types of Rankine

-------
                     111-12
     BASIC RANKINE CYCLE  ENGINE
                    Figure 8.
                                     WATER FLUID
                             S- ENTROPY
CANDIDATE RANKINE CYCLE POWER PLANTS
   ORGAN 1C WORKING FLUID- RECIPROCATING EXPANDER

   WATER BASE WORKING FLUID- RECIPROCATING OR A
                             ROTARY EXPANDER

   •ORGANIC WORKING FLUID-TURBINE EXPANDER
                     Figure 9.

-------
                                 111-13
systems which are presently in the design and component test
phase as shown in Figure 9.  The work on the organic working
fluid reciprocating expander system is summarized in the
technical paper presented in this meeting by a representative
from Thenno-Electron Corporation where this work is being
conducted in our country.  Work on the other versions of
Rankine cycle systems noted on the Figure are just beginning.
As we see them, the problem areas associated with a practical
design of the Rankine cycle appear on the next slide.  The
problems appear mainly in the inefficiency of components and
complexity of the control system and, of course, with its
exhaust emissions.  A representative of Solar Division of
International Harvester Company in San Diego, California, one
of three firms conducting government sponsored research and
development on improved combustors will summarize the work of
one of the high efficiency combustor systems at this meeting.
The next Figure summarizes a schedule of the planned work on
the Rankine cycle.  Complete system development for the three
types mentioned earlier with parallel research on the components
is shown.  We point out that the three systems contractors are
designing complete systems including all components.  An integral
part of this effort is to define the design requirements that
each component must meet.  For example, the temperature of the
condensate, maximum and minimum flow rates and heat release
rates are part of the condenser design requirements.  The
systems contractor and the backup component research contractor
are designing a condenser to meet the identical requirements.
The condenser research contractor will design and verify in
test his condenser designs for each of three systems.  The
first generation system hardware for any of the three systems
may have components made by one or more of the component research
contractors, depending on whose component design is the most
efficient with cost to produce being one of several important

-------
                 111-14
        RANKINE CYCLE ENGINE
                PROBLEMS
'CONDENSER SIZE WT.

'BOILER SIZE WT.

'CONTROL COMPLEXITY

'ENGINE EFFICIENCY

•MINIMIZE EMISSIONS
•FREEZING (water)

•LUBRICATION

'FEEDPUMP

•SEALS (non-water)
•VALVING DESIGN
                Figure 10.

RANKINE CYCLE ENGINE  PROGRAM

SYSTEM
DEVELOPMEIS
COMPONEN1
RESEARCH
1969
197O
System De
IT


Con
Con
Co IT



1971
sign Co


1st (
1972
mpone

1973
nt Test
1974
-3 Sys
1975
terns
generation Hardware


2nd G
rol System
denser
ibustio




n Rese





sn. Har
0—
arch (3
dware
	 o
1


                Figure 11.

-------
                                 111-15

elements applied in that judgment.  Lastly, you can see from
Figure 11 that delivery of the first prototype engines could be
as early as 1973.

Gas Turbine.  More work has been conducted by industry in our
country on the gas turbine than on any other candidate with an
unconventional engine for the automobile application.  For the
gas turbine we are focusing our efforts on solving some of the
problems which have plagued past attempts to get the gas turbine
on the road.  Some of these problems are noted in Figure 12.
These problems include the need for:  developing manufacturing
techniques for mass producing turbine inexpensively, for
improving the part load, fuel economy, combustion research to
improve combustor design and increasing system reliability.
Actually, if we  operated our automobile in the same way that
commercial jet aircraft are operated, namely, running most of
the time at constant speed, many of these problems would
disappear.  This gas turbine program is oriented toward solving
problems, and if successful, is intended to stimulate industry
to apply the results of this research to their own turbine
designs.  As we  see it, the demonstration of the turbine in the
automobile probably will be performed by industry.  It is because
industry is so close to practical hardware on this system that
our approach here emphasizes problem solving rather than a
government sponsored demonstration.

One of the big unknowns in getting the turbine on the road is
knowledge of the cost to mass produce this system; therefore,
early recognition of what are the high cost components of the
system is needed so that fruitful manufacturing research can be
properly directed.  A representative of Pratt & Whitney Aircraft
from the United  States will present findings on this subject in
this meeting.  That firm is on the team of contractors in this
program.

-------
                             111-16

Hybrids.  Now, for the hybrids,  the heat engine/electric and the
heat engine/flywheel.   The heat  engine/electric hybrid consists
of a small size engine such as in a Volkswagen, and an array of
inexpensive lead-acid batteries.   The manner in which these two
power sources are arranged is illustrated in Figure 13.   The
same series versus parallel configurations can be applied where
the flywheel replaces the battery system.  The system is designed
to extract power from the heat engine alone, or from both
sources, heat engine and battery, at the same time.  We find
that the parallel hookup is best where the conventional internal
combustion engine is used as the heat engine and a series
configuration appears best where a small gas turbine is used.
In either configuration the system operates by running the heat
engine at a constant speed—say  equivalent to 40 mph road speed,
and, where vehicle accelerations are required, the additional
power comes from the battery system.  Above a 40 mph road speed
the engine speed slowly converges on the new steady-state
level demanded by the driver with additional power required
during slack time provided directly from the battery system.
There are several potential advantages which this hybrid concept
offers.  One is that the engine  speed range is relatively small
with the attendant ease of control of exhaust emissions under
such a condition.  Another advantage is that the very high road
performance for a standard size  American automobile, of
approximately 4,000 Ib. weight,  can be demonstrated using a
relatively small and inexpensive heat engine.  A basic problem
with this type system appears in its relative complexity, higher
cost and the larger system volume required with use of two power
sources.  Development of improved lead-acid batteries to
accommodate the rapid charge-discharge characteristics needed
for this mode of operation is now underway in the Research and
Development Program.

In our analysis of this type hybrid many types of heat engines
have been considered by our contractors.  Two types were noted

-------
                     111-17
           OASTURBINEENCjINES

 MAJOR PROBLEM AREAS

       • MANUFACTURING COST
       •FUEL ECONOMY
       • EMISSIONS-NOXREDUCTION FOR 1980 GOALS
       •RELIABILITY
                   Figure 12.
            HYBRID  SCHEMATIC
SERIES HYBRID
DRIVES WITH
BATTERY ENERGY
STORAGE
 ELECTRIC
'GENERATOR
CONTROLS - ^

  1
                        BATTERY

PARALLEL HYBRID
DRIVES WITH
BATTERY ENERGY
STORAGE
     II
    ELECTRIC
 MOTOR/GENERATOR

   Figure 13.

-------
                             111-18

earlier.  Two firms in our country are conducting the bulk of
our work on this hybrid.   One firm is the Aerospace Corporation.
A member of the Aerospace Corporation team will report on their
work at this meeting.

As stated earlier, the heat engine/flywheel system would function
in a similar manner to the heat engine/electric with the battery
replaced by a mechanical storage device, namely, the spinning
flywheel.  We have progressed on this system from parametric
analysis of many practical flywheel materials such as shown in
Figure 14 and numerous practical configurations as shown in
Figure 15 to the design  aid fabrication of specific flywheels
for cars.

Today we are at the point where two flywheels are now being
tested to verify predicted energy and power densities.  One of
the flywheels is fiberglas and the other fabricated of 4340
steel.  Work on the configuration of the key component of this
system, the transmission, will begin shortly.  To give you a
feel for some characteristics of the flywheel:  for a full
size family car, the flywheel is made of 4340 steel in a
constant stress configuration weighing 42 Ibs. and operating
at a maximum speed of  24,000 revolutions per minute.

All-Electric Car.   Alkali-metal battery research and development
for the automobile application has been underway for more than
a year at Argonne National Laboratories.  Their work was reported
on in this meeting. We anticipate that the  'proof-of-principle'
for the high temperature lithium-sulfur system will be demonstrated
on single and multiple cells within the next 18 months.  This
'proof-of-principle' refers to the experimental demonstration
that the energy density and power density desired for a full-
sized battery can be obtained on an elemental cell basis.  Once
this proof-of-principle has been achieved, the battery work will
move into a development phase, first with a goal of a 2 kw battery,
then a 5 kw battery and then a 20 kw battery system.

-------
FLYWHEEL MATERIALS
MATERIAL
I8N 1-400
(MARAGING STEEL)
18N1-300
(MARAGING STEEL)
4340 STEEL
1040 STEEL
1020 STEEL
CAST IRON
2021 -T8!
(ALUMINUM)
2024 -T851
(ALUMINUM)
6AL-4V
(TITANIUM)
V GLASS
S-IOI4 GLASS
DENSITY
M -
L8S/IN3
0.289
0.289
0.283
0.283
0.283n
0.280
0.103
0.100
0.160
0.092
0.067
POISSON'S
RATIO
M
0.26
0.30
0.32
0.30
0.30
0.30
0.33
0.33
0.32
0.20
0.20
ULT
TENSILE
fou)KSI
409
307
260
87
68
55
62
66
150
130
250
YIELD
TENSILE
(Fty)KSI
400
300
217
58
43
37
52
58
140
-
-
REC. WORKING
STRESS
(
-------
              FLYWHEEL  GEOMETRIES
             PIERCED
              DISC
        RIM
       SOLID
       DISC
      CONICAL
      CONSTANT
       STRESS
      LOG OR
        BAR
        M
        I
   SHAPE

                      r
                                r
   STRESS
    MAP
       RIMSTRESS-v
  VOLUMETRIC
  EFFICIENCY
(FW$MIN.WT. HOUSING)
30.4%
20%
30.4%
62%
80%
6%
                          Figure 15.

-------
                            Ill-21

Stratified Charge Engine.  As mentioned earlier, the two
candidates which we are seriously considering for entry into the
Program are the stratified charge engine and the Diesel.  The
stratified charge engine is a spark-ignited gasoline fueled
internal combustion engine with many hardware characteristics
of the conventional engine.  Differences appear mainly in the
combustion chamber design, use of fuel injection, and in the
resulting combustion process.  In one version of this engine
the fuel is injected into the cylinder as the spark appears and
therefore the fuel burns as it enters.  The fuel burns initially
in an over-rich condition with fuel injectors designed to permit
fluid swirl at the top of the cylinder.  The burning expands
into the lower portions of the cylinder and, overall, the
resultant combustion products are similar to those obtained using
a very lean mixture.  Lean burning is very desirable from the
emissions viewpoint.  The initial work on this engine was
sponsored by the U.S. Army Tank-Automotive Command in the state
of Michigan.  The measured exhaust emission levels for the
stratified charge engine installed in jeeps and employing a
catalytic muffler are below the standards for hydrocarbons and
carbon monoxide set for 1975 but further work must be conducted
to reduce the nitrogen oxide emissions.  Several generations
of development have been funded and if this system is brought into
the Program, the work will emphasize achievement of reductions in
nitrogen oxide emissions and then fleet testing of this well-
developed system.

Diesel Engine.  The Diesel engine is not commonly used in American
made automobiles, mainly because it is heavy and costs more to
manufacture compared with the conventional Otto cycle engine.  If
the Diesel is brought into the Program, emphasis will be placed
on furthering the development of a low compression ratio Diesel
with a high swirl and prechamber design.  Exhaust emission levels
for hydrocarbons and carbon monoxide which are lower than the
1975 standards have been shown for this type engine without

-------
                         Ill-22

resorting to a catalyst.   The measured nitrogen oxide levels
on the first generation engine are relatively low but are within
reaching distance of the 1976 standard.  Work on this type Diesel
will concentrate on nitrogen oxides reduction, and on performance
durability and driveability testing.

Federal Clean Car Incentive Program
The goals of the Research and Development Program, the near
candidates and our approach to further development of each have
been discussed.  As stated earlier, there is another important
program underway in our nation to develop a virtually pollution
free automobile using an alternate power source.  This work
consists of efforts on the part of industry itself with
financial incentive provided by our government to inspire
independent work.  This describes the Federal Clean Car Incentive
Program.  The intent of the Program is to stimulate the private
sector to produce a virtually pollution-free automobile.  The
program provides a market for large and small auto and non-auto
manufacturers who often possess new and unique approaches toward
engine designs.  However, in the past, they have lacked incentive
to further independent development.

The Program description is summarized in Figure 16"  After
successfully passing stringent emissions and performance testing,
first on a leased Prototype car, then on 10 purchased copies of
the prototype for Demonstration, the successful engine system
will be further tested after procurement of 300 vehicles.  If the
low emissions levels are maintained and road performance satis-
factory, the car then receives certification as a low-pollution
vehicle, and this certification is significant.  The certified
vehicles which make their way through the Program will be
favored by purchase in quantity for Government fleet use.
Premiums as high as 200% over the basic price normally charged
to the government are permitted for the successful developer.

-------
  ELEMENTS OF THE  FEDERAL INCENTIVE PLAN
 PROTOTYPE
    PHASE
-J.EASE 1
  *
  /

Meet Prototype Specs.
DEMONSTRATION
   PHASE
                      -7EJT
FLEET TEST
  PHASE
Government Procure-
ment in Quantity
      i
   Legislation

 Low Emission
Vehicle Certification I
                                    •TES7\
                            .PURCHASE
               PURCHASE

               10 VEHICLES.
                         Figure 16.

-------
                              111-24

The legislation to accommodate such purchases appears in the
Amendments to the Clean Air Act, recently passed by our Congress
and signed by our President.  We point out that in this Program,
the developer retains all patent rights and that the total cost
to manufacture, including tooling costs plus a reasonable
profit, is provided to the developer.

The Incentive Program is interrelated with the Research and
Development Program in that, where desirable, a partially
successful candidate in the Incentive Program could receive
research funding to further improve the engine system and this
funding could come from the Research and Development Program.

Lastly, we mention that the Incentive Program is expected to
provide a valuable source of information from actual vehicles
from which to judge the capability of the industry to meet
1975/76 emission standards.

This Clean Car Incentive Program has just begun, with approximately
20 proposals from industry to enter the prototype phase received
recently,  and more to come.  While proposal evaluation is not
as yet complete, as we see it now, delivery of the first proto-
types of low emission cars into this Program will be made before
August of this year.   Figure 17 summarizes Program status.

The Incentive Program has been planned at $20 million over a
three year period,  The rate of use and extent of use of these
funds is dependent on the rate at which selected candidates
proceed through the Program.  At each test stage, any given
candidate can be eliminated.  The period of major costs should
be in 1972 and 1973 since a number of candidates would be
entering,  or, in the demonstration test stage in 1972 and the
fleet test of one or more candidates should begin in 1973.

-------
      PROGRAM   STATU S
 MARCH  1971  CONTRACTS (approx. 15)  FOR


        • Brayton Cycle  Gas Turbine

        -Rankine  Cycle

        •Heat Engine/Electric  Hybrid

        - Electric


* PROTOTYPE  DEMOSTRATBONS  IN  1971

        WITH GAS  TURBINES


                  Figure 17.
H
H
M
I

Ln

-------
                              111-26

World-wide Participation in Both Programs
Both Programs are open to firms residing outside of the
United States.  Most contractual work in the Research and
Development Program is awarded based on competitive proposals by
industry in response to Scopes of Work which are well publicized
and then sent to firms requesting them.  Multiple awards or
contracts are common to the Research and Development Program.
For example, we now have three different firms under contract
and working on three different approaches to develop high
efficiency combustors for the Rankine cycle system.  There will
be four parallel approaches considered for the gas turbine
combustor design.  As a step toward improving lines of communi-
cation between ourselves in the Program and industry in your
country, we intend to define mechanisms to provide the Scopes
of Work for new contracts to you and other interested industrial
firms.  In addition, the Scopes of Work for all on-going
contracts, and final reports from completed contracts will be
made available to all interested parties.

We welcome your participation in both Programs.  Any firm or
government representative who wishes information on participation
will be provided written material in this meeting.

Summary
I will now summarize this talk.  We have defined the Research and
Development Program and the Incentive Program.  Together, they
form the Advanced Automotive Power Systems Program.  Their goals
are similar—to produce a virtually pollution free automobile—
their methods differ however.  We feel confident that ultimate
goal will be achieved within the 5-year time frame.  If the
conventional internal  (Dmbustion engine can be modified adequately
by industry, so much for the better.  But, it is these two
Programs which will provide the insurance to our nation in the
event that they cannot make it.

-------
                              111-27

Lastly, we urge that industry in your countries compete in
the Programs.  We will do all that is reasonable to encourage
this activity.  In this way all of us will benefit, here and
at home.

-------
                  IV-1
               Chapter IV
THE POTENTIAL OF THE GAS-TURBINE VEHICLE

      IN ALLEVIATING AIR POLLUTION
                   by
            Edward S. Wright
     Chief, Ground Systems Analysis
         Research Laboratories
            United Aircraft
    East Hartford, Connecticut, USA

-------
                                    IV-2


Introduction

Emissions from motor vehicles account for approximately 60 percent of

the annual total air pollutants emitted in this country (see reference

1); the three major pollutants emitted by automobiles—unburned hydro-

carbons (UHC), carbon monoxide (CO), and nitrogen oxides (NOX)—comprise

approximately 63, 93, and 46 percent, respectively, of the total of

each of these pollutants.  Consequently, there has been a great deal of

interest in ways and means of reducing the emissions of Otto cycle

internal combustion gasoline engines, and of finding inherently cleaner-

burning alternatives to these engines.


It is generally believed that if the reciprocating internal combustion

engine cannot be cleaned up sufficiently in a reasonably economic manner,

the gas turbine is the most likely alternative.  Its inherent emission

characteristics are generally considered to be satisfactory (presently

or potentially), and the primary issue is whether it is technically and

economically feasible.  Based on several well-advertised programs of

the major automobile manufacturers in the United States and abroad, it

appears that the technical problems are surmountable, but there is, as

yet, considerable doubt concerning the economics of automobile gas

turbines.


The purpose of this paper is: (a) to summarize the emission characteristics

of automobile gas turbines, (b)  to examine the potential of these engines

for satisfying automobile requirements, and (c) to assess their ability


(1) A study sponsored by the National Air Pollution Control Administration is
    being  conducted currently by the United Aircraft Research Laboratories,
    with the purpose of examining the probable manufacturing costs of several
    candidate automotive gas turbine engines, including those with simple
    and regenerated cycles, and with free and single-shaft turbines.  The
    results, when published!, should provide an interesting data source in an
    area where the available literature is scarce.

-------
                                    IV-3
to compete with cleaned-up reciprocating  internal combustion engines




by identifying their relevant technical  (performance) and economic




(manufacturing-cost and fuel-consumption) characteristics.  The paper




is intended to provid background information  for a panel discussion at




the 1970 ASME Winter Annual Meeting,,







Emissions




Uncontrolled Emissions of Otto Cycle-Engined  Automobiles--An automobile




utilizing an Otto cycle (gasoline fuel,  reciprocating) engine with no




emission control devices has three  sources of emission--the exhaust,




the crankcase, and the fuel supply  system (fuel tank and carburetor).




Virtually all of the CO, NO , and lead,  and about 60 percent of the UHC

                           X



come from the exhaust of the vehicle; about 30 percent of the UHC comes




from the crankcase; and the remaining 10  percent of the UHC comes from




the fuel supply system (see reference 2).  In terms of quantity of emissions,




the pollutants in the exhaust, based upon a weighted average of engine




sizes, are 0.50 Ib/hr UHC, 3.5 Ib/hr CO,  and  0.17 Ib/hr NO  (see reference
                                                          X



3).  Variations of + 20 percent or  more  are typical.  The total emissions




from an uncontrolled Otto cycle-engined  automobile have been estimated




as follows (see reference 1): unburned hydrocarbons--520 Ib/yr; carbon




monoxide--1,700 Ib/hr; and nitrogen oxides--90 Ib/yr.







Emission Standards—The first nation-wide standards for automotive emissions




were issued in 1966, and were to take effect  on all 1968 model year cars




and light trucks (see reference 4).  These standards are given in Table  1

-------
                                   IV-4
in terms of concentration, and pertain to engines with 140 cubic inches

or large displacement.  The standards for 1970 model year cars, also

shown in the following, tighten the 1968 standards by about 30 percent.

An important change in the standards is a change from concentration

(ppm) to a mass-per-mile basis.  It will be noted that these standards

do not contain restrictions on NO  or evaporations.
                              Table 1
      Crankcase

      Exhaust
        Unburned Hydrocarbon
        Carbon Monoxide
                                     1968 Model Year
          Zero
                        1970 Model Year
             Zero
        275 ppm           2.2 g/mi
    1.5% volume fraction   23 g/mi'
Proposed future emission standards, in g/mi., are presented in Table 2,

based on data in reference 5.

                    Table 2.  Emission Standards

                                 	 Model Year
     Pollutant

   Unburned Hydrocarbon
   Carbon Monoxide
   Nitrogen Oxides
   Particulates

     Total
1971
2.7*
23
-_**
— **
1975
0.5
11
0.9
0.1
1980
0.25
4.7
0.4
0.03
21.8
12.5
5.4
    ^Includes evaporative losses from fuel supply system.
   **No defined standard; estimated output for uncontrolled emission
     is 5.8 g/mi of nitrogen oxides and 0.3 g/mi of particulates.
The proposed 1971 model year standards include an evaporative loss, and

the proposed 1975 model year standards include NOX and particulate require-

ments.  The 1975 standards represent an approximate 80 percent reduction in

-------
                                    IV-5
UHC, and a 50 percent reduction in CO from the 1971  standards, and an



estimated 85 percent reduction in NO  from the uncontrolled automobile.
                                    x


The proposed 1980 standards are those put forth by the President's



Environmental Advisory Board and represent reductions by about 50



percent in the 1975 standards.





Several methods of adapting the gasoline engine to meet these standards



have been proposed, and major efforts are being undertaken by the



automotive industry to meet them.  However, the effectiveness of some



of the control devices has been questioned (see reference 6), and,



consequently, alternative propulsion systems have been proposed.  There



is considerable evidence to suggest that the gas turbine has the potential



of low emissions without control systems.





Emissions from Gas Turbine Engines—A gas turbine does not have a component



that serves a function similar to the crankcase of the Otto-cycle engine.



Lubrication, where necessary, is usually by a method which could be



referred to as dry sump, with no chance of the fuel-air mixture venting



to the atmosphere from the oil tank.





The fuel supply system of a gas turbine consists of a fuel tank, injection



pump, and injectors; hence, there are no float chambers or other reservoirs



to allow evaporative losses such as those from the carburetor of automo^



tive reciprocating engines.  Since a turbine would probably operate on



lower volatility fuels, such as kerosene or diesel oil mixed with low-



octane gasoline, the evaporative losses from the fuel tank are more easily



minimized.

-------
                                    IV-6
Since neither the fuel system, the lubrication system, nor the fuel,




itself, (in terms of lead or sulfur content) is likely to be a source




of pollution from the vehicular gas turbine, it is evident that the




combustion process is the only significant area of concern.  As described




in the following, the continuous combustion process at overall lean




fuel-air ratios leads to inherently low emissions levels of UHC, NOX,




and CO for gas turbine engines.






The concentrations of exhaust pollutants in gas turbine cannot be compared




readily with those of the Otto-cycle engine because of the disparity of




fuel-air ratios.  Therefore, a method of comparison using a parameter




called the emission index (see reference 7) was devised.  The emissions




index (El) is, essentially, a measure of the pounds of pollutants




emitted per pound of fuel burned; it can be calculated by using the concen-




tration of pollutant and the fuel-air ratio of the engine.






The overal1 lean fuel-air ratios and the absence of quenching on cool




walls results in complete combustion and, consequently, low levels of UHC




and CO emission.  The UHC's are of two major types; the first is formed by




those hydrocarbons that pass through the combustion chamber unchanged;




the second consists of those which are produced in the low-temperature




region of the combustion process.  The formation of both types, as well




as the formation of CO, takes place in the outer zones of the combustion




chamber,  where the air injected to cool the liner walls has lowered the




local temperature.  The emissions index is highest at idle for UHC and CO.

-------
                                      IV-7
As the engine power output increases  (i.e., the equivalence ratio




increases), combustion improves, the  emissions index  for UHC and CO




becomes negligible, and the absolute  emission levels  decrease  signifi-




cantly—despite the increase in fuel  consumption due  to higher power



output.






As in the Otto cycle engine, the formation  of NO   is  probably  caused by
                                                x
the oxidation of the nitrogen in air  (0 + N_), and measurements to date




have indicated that it is a strong function  of temperature and a weak




function of pressure, at least for practical  gas turbine cycle pressures.




These measurements show NO  levels distinctly lower than those expected
                          X



for equilibrium conditions (see reference 7), in contrast with the NOV
                                                                     .A.



levels of Otto cycle engines.  Furthermore,  the NOX is essentially frozen




at the combustor exit, and concentrations remain the same during the




expansion process„  Among the reasons postulated for this low NOX concentra-




tion are the rapid quenching of the combustion products with the excess




air introduced into the combustor for cooling and the locally rich burning




near the fuel droplets in the primary combustion zone.  Since temperatures




at the combustor exit are a function  of both  combustor inlet temperature




and the equivalence ratio, it has been assumed commonly that improvements




in cycle efficiency leading to higher combustor inlet temperatures (as a




result of higher pressure ratios, or  of regeneration) or higher turbine




inlet temperatures will automatically lead to higher N0x emission levels,




because of the associated higher combustor temperatures.  This assump-




tion is not necessarily valid„  Analytical and experimental research




conducted at the United Aircraft Research Laboratories has indicated that

-------
                                    IV-8
the bulk of the NO  formation occurs due to the strong recirculation
                  X

patterns characteristic of the primary combustion zones, since  in  straight-

through flow, the exposure of the mixture and products to high  temperature

in the flame front is not of sufficient duration to allow significant NOX

formation (see reference 8).  Thus, the prospect exists that further

research, which yields knowledge of how variations in primary combustion

zone geometry, residence time, combustion temperature, and fuel-air ratio

affect overall NO  production, may indicate how combustor designs  must
                 X

be modified in order to reduce the already low NO  emissions level of
                                                 X

vehicular gas turbines even further.


Comparisons of Automobile Engine Types—Table 3 presents comparisons of

the emissions index (ratio of pollutant weight to fuel weight x 10-^) for

several automobile reciprocating engines and types of gas turbines, reported

from various sources and compiled in reference 7.  Although these  data

are not completely compatible and despite their wide variation, they can be

used as a basis of some general conclusions concerning the emission of CO

from a typical gas turbine.  Using the El as a basis, the emission of CO

from a typical gas turbine is from 1 to 10 percent of an uncontrolled Otto-

cycle engine, the emission of UHC is 1 to 20 percent, and that  of  NO
                                                                    X
from 30 to 80 percent.

               Table 3. Comparison of Emission Indices

     Engine/Engine Pollutant              CO             UHC*       NO

     Reciprocating (30 mph)              241             6.1       16.3
     Reciprocating (cold start)           513            35         14.5
     Reciprocating                       407            35         13
     Regenerative Gas Turbine (30 mph)     5.3           0.3       13.5
     Gas  Turbine (cold start)             50             1.1        9.8
     Aircraft Turbojet                     3.3           0.4        5.5
     Aircraft Turbojet                    19             25

   ^Expressed as hexane.

-------
                                    IV-9
The procedure of totaling emission indices, used  in  this paper,  is  adopted


in order to provide gross comparisons between  the two  types  of engines.


It is felt that this simplified approach has validity,  since the emission


standards of Table 2 are based on reciprocating engine outputs,  rather


than specifically on precise health-based  criteria for particular pollu-


tants.  On this El basis, the gas turbine  ranges  from  1 to 11 percent


of that of the untreated Otto-cycle engine.




Korth (see reference 9) and others have measured  emissions from  regenerated


automobile gas turbines operating in typical duty cycles.  It is very


difficult to instrument mobile vehicles properly  for accurate measure-


ments of the extremely low concentrations  of pollutants typical  of  gas


turbine engines, and wide variations have  been reported among several


makes of engines.  Therefore, extrapolations of the  test-stand measure-


ments of an aircraft gas turbine to the automobile standard  duty cycle


are of interest.  Although such an engine  would not  be applied to an


automobile, its emissions characteristics  might closely parallel those


of a simple-cycle automobile engine of acceptable efficiency.  For  such


an engine applied to an automobile type duty cycle,  an average total El


of 10.0 may be assumed which, at an assumed fuel  consumption of  270 g/mi^i1


yields a total emission (UHC + CO + NOX) or 2.7 g/mi.   Thus,  the extrapo-


lated emissions level is approximately 50  percent of the proposed 1980


standard for the treated gasoline internal combustion  engine, even  though


its NO  levels exceed the standard for that particular pollutant.
      x
(1)  About 11 miles per gallon.

-------
                                    IV-10
It must be emphasized that this emission level is predicted  for  gas




turbine combustion technology of the 1960's, with no treatment,  and with




no emissions criteria considered in the combustor design.  Furthermore,




the ability of treated gasoline automobiles to continue to meet




emissions criteria over the lifetime of the vehicle is open  to serious ques-




tion, even if costly maintenance and inspection procedures are adopted




and enforced.  In the absence of perfect maintenance, if a 1980  Otto




cycle engine reverted to emissions of the level of the 1975  standard,




it would cause as much pollution as 4.6 untreated gas turbines (as




described in the foregoing); if it is reverted to 1971 standards, it




would cause as much pollution as 11.4 gas turbines; and if it reverted




to its natural untreated operating state, it would cause as  much pollution




as 35 gas turbines.






The occurrence of the poor mixture/excess blowby condition commonly




observed on today's automobiles on the road would lead to even higher




levels of pollution.  For the gas turbine,  poor adjustment of fuel to




the combustor is unlikely to cause pollution problems., If the mixture




becomes fuel-rich, the engine simply puts out more power; if the car




were operating at maximum throttle, and the owner ignored overheat




warnings,  the engine might fail ultimately, but the pollution control of




the gas turbine is fail-safe.






Alternative Power Plants—Other than the gas turbine, electric,  steam,




and Stirling engines are the most frequently mentioned of low-pollution




alternative power plants to the gasoline reciprocating engine for

-------
                                     IV-11
vehicular propulsion.  Well-reported efforts have been directed by




various organizations toward investigating steam and electric alterna-




tives to the gasoline engine.  These efforts nearly always confirm the




difficulties inherent with these approaches, lending support to the




strong economic pressures to continue the usage of gasoline engines for




automobiles, even though control of their pollution will involve sub-




stantial and costly modification.  At present, the major automobile




manufacturers are publicly committed to the reduction of emissions from




gasoline engines as their preferred approach to the pollution problem,




but have identified the gas turbine as their first alternative.






From an emissions standpoint, gas turbine emissions are likely to be




iroughly equivalent to those emitted by a central power station producing




an equivalent amount of power for electrically powered vehicles.







Other Vehicular Applications--In terms of gross magnitude of emission of




pollutants, the automobile far exceeds all other vehicles combined, both




because of Otto cycle engine characteristics and sheer numbers„  Despite




the much lower levels of UHC, CO, and M)  emissions contributed by diesel-
                                        X



powered trucks, buses, trains, and off-highway equipment of various types,




special localized situations regarding pollution may favor the gas turbine




for these applications as well.  For example, the urban diesel bus, while




theoretically a low emitter of pollutants, is severely criticized for




introducing noise and smoke to sensitive areas, and gas turbines may alleviate




these problems as well.

-------
                                     IV-12
Conclusions Concerning Emissions—The technical literature documents

the fact that the gas turbine, at least in most of its forms, does  indeed

have  significantly lower emissions than the conventional Otto cycle engine

and can meet proposed future standards for UHC and CO without modification

(see  references 7 and 9) .  The ability of the turbine to meet future N0x

standards is not as clear, although it is hoped that the N0x emission will

be amenable to solution when sufficient research and development are per-

formed.  Nonetheless, even without NOX reduction, total untreated emissions

(UHC, CO, NOX) of the gas turbine are likely to be only 50 percent  of those

of the treated Otto cycle engine which meets proposed 1980 standards, and

less  than 3 percent of the output of an untreated Otto cycle engine.  If

all Otto cycle engines were replaced by gas turbines, the motor vehicle

contribution to pollution referred to on page 1 would probably be reduced

from  60 to 2 percent.  In other words, it is likely that vehicular-caused

air pollution would be eliminated as a cause for serious environmental

concern.


Other Pertinent Characteristics

Engine Power Output—Figure 1 shows torque/speed characteristics of typical

reciprocating engines and five gas turbines of various configurations, i.e.,

single-shaft, differential with 10 percent compressor speed, single-stage

free-turbine, and multi-stage free-turbine configurations.  Clearly, the

vide variations in the characteristics imply that when discussing "typical"

gas turbines, it is essential to specify the configuration.  The configura-

tion most commonly considered for vehicular applications^ has been  the single'
(1) The Chrysler, Rover- and General Motors automobile engines, and the GM,
    Ford, and Leyland truck engines have used the free-turbine configuration.

-------
                                      IV-13
stage free  turbine, since the  rising torque curve below rated speed

yields positive stability when the output shaft  speed  is in a fixed ratio

to the wheel speed.  (As wheel speed drops due to increased resistance,

the increased torque of the  engine tends to restore  the system to its

original  speed.)
      3
      o
2.8

2.4

2.0
      ?*-°  1.6
            1.2

	100% CONSTANT HORSEPOWER
	 MULTISTAGE FREE TURBINE
	SINGLE STAGE FREE TURBINE

	DIFFERENTIAL TURBINE (WITH 10%
       COMPRESSOR OVERSPEED)

	 DIFFERENTIAL TURBINE (FOR
       CONSTANT COMPRESSOR SPEED)
—,	RECIPROCATING
_._.	SINGLE SHAFT TURBINE
                 0.2  0.4  0.6 0.8  1.0  1.2
               OUTPUT SPEED  RAT!O N/NO


      Figure 1.  Engine torque speed  characteristics.
The acceleration lag of gas  turbine engines is due  primarily to the

fact that  compression ratios  are  more constant in reciprocating engines

(see reference 10).   Nevertheless,  many techniques  are  available to reduce

the lag to acceptable limits.   These include the momentary removal of load

at some point  in the power transmission train, reduction of engine shaft

mass inertia,  introduction of variable geometry, or allowing momentary over-

fueling (overheat)  in a fashion analogous to the accelerator pump of an

Otto cycle engine.

-------
                                    IV-14
Because the gas turbine compresses far more air than is actually consumed




in the combustion process, it is more sensitive to ambient conditions




than its reciprocating counterparts: its power output and fuel consumption




are both affected more adversely by high temperature and altitude, and




the engine must be sized to provide adequate performance at reasonably-




adverse conditions.  This fact implies that cars used extensively at




high altitudes and temperatures should be equipped with higher powered




engine options and also that certain potential "optional extras," such




as water injection, might increase the flexibility of the engine„






Maintenance, Reliability, and Life—Experience with aircraft and industrial




gas turbines has indicated that, because of their basic simplicity,




these engines can operate for thousands of hours with only minor routine




maintenance.  The same can be expected for automobile gas turbines,




provided their unique requirements are adequately satisfied in the design.




One very significant requirement is the ability to withstand the thermal




stresses associated with the frequent changes in power and, where regener-




ated engines are required, the addition of cores and seals may complicate




the maintenance requirements.






Transmission Considerations--The engine torque/speed characteristics shown




in Figure 1 are unsatisfactory for vehicle propulsion without torque




multiplication.  Examination of a spectrum of transmission configurations is




needed to match the wide variety of torque/speed curves characteristic  of




various configurations of gas turbines to various vehicle requirements.

-------
                                    IV-15
Ideally an infinitely variable  (IV) transmission would be desirable




for use with gas turbines.  In  effect, the IV transmission decouples




vehicle wheel speed from engine speed and allows the engine to deliver




full power at any wheel speed  (subject to traction limitations and trans-




mission efficiency), so that the torque speed characteristic of the




engine is irrelevant to the output of the transmission.  The lead candi-




dates for vehicular gas turbine IV transmissions are electrical, hydro-




static, and mechanical.






The electrical system would consist of a generator (or alternator) driven




by the gas turbine, an electric motor (or motors) driving the axle or




wheels, and a control system,,   Although the electrical system would yield




almost ideal power application, its manufacturing cost for automobiles




is likely to be high.  Hydrostatic transmissions, consisting of displace-




ment hydrostatic pumps and motors, would require the development of




successful turbine shaft speed  pumps and might have noise and safety




problems.  Mechanical infinitely variable transmissions rely on rolling




contact along a variable radius, and generally suffer from Hertzian stress




problems at the contact surfaces.  Therefore, they tend to result in




comparatively large and heavy  installations.  Nevertheless, they merit




serious consideration for automobile gas turbines.






The electrical transmissions described in the foregoing allow consideration




of hybrid automobiles and buses (see reference 11) in which power is drawn




not only from the prime mover,  but also from energy storage devices, such




as batteries of flywheels.  These hybrid approaches permit shutdown of the

-------
                                   IV-16
prime mover (and thus elimination of all emissions) during  operation  in

emissions-sensitive areas, and also permit charging the energy  system

at the most efficient and/or lowest emissions operating point of the

engine, thus reducing overall emissions.


Transmissions with a finite number of gear ratios are attractive because

of their low cost and high efficiency.  For a given engine  application,

the higher the basic engine torque at reduced speed, the fewer  the gear

ratios needed; thus, a single-shaft engine would probably require an

impractical number of gear ratios.  Hydraulic torque converters are used

in automobiles, where they are generally teamed with automatically shifting

two- or three-speed geared transmissions; they are inexpensive, conven-

ient, and adequately efficient.  For a single-shaft gas turbine, such

a transmission with a minimum of 5 or 6 gear ratios (or a multiple-stage

torque converter) may furnish acceptable performance at reasonable cost.


Vehicle Performance--The prime mover and transmission must  be selected to

meet some mission envelope of vehicle performance (desired  accelerations,

velocities, and elapsed times for a range of payloads).  Standard computa-

tional procedures (see reference 12) relate propulsion power requirements,

via transmission efficiencies, to road, grade, and air resistance; the

acceleration performance depends, additionally, on the mass and rotational

inertia of the vehicle.  With appropriate allowance for installation  and

accessory losses, these procedures serve to establish the engine power

requirementso   Installed horsepower, as presently defined for automobiles,


(1)  For a free-turbine engine, the hydraulic torque converter is probably
    redundant.

-------
                                    IV-17
can be misleading,  since  a typical  automobile equipped with a nominal

250-hp engine will  require about 30 hp to cruise continuously at 65

mph.  (A truck at maximum gross  weight equipped with a 250-hp engine,

on the other hand,  is  likely  to  require its full power to cruise at

that speed.)  A  typical automobile  engine duty cycle, adapted from

information supplied by Rover (see  reference 12), is reproduced in

Figure 2.
                       0-15  15-30  30-60  60-90 90-120120-150
                               POWER LEVEL (hp)
          Figure  2.   Typical automobile engine duty cycle.
The weight of the engine  is  important  for  establishing automobile perfor-

mance.  Present engines and  radiators  average  16.5  percent of total car

weight (see reference 11), or approximately  660  pounds for the nominal

4,000-pound automobile.   If  the  corresponding  gas turbine weighed 160

pounds, the 500-pound saving would  allow significant  reductions in power

output for a given performance level,  since  peak power requirements for

the automobile are largely determined  by the acceleration desired, and

acceleration power is directly proportional  to mass.   For example, the

-------
                                    IV-18
 power  required at the road solely to accelerate  a  3,500-pound  automobile




 at  5 mph per second is obviously 12..5 percent  less than  that  required




 for a  4,000-pound automobile  (106.3 hp versus  121.5 hp at  50  mph).




 Applying standard road and aerodynamic resistance  calculations to  typical




 automobiles, summing power requirements, and assuming equal  transmission




 efficiencies, leads to the conclusion that equivalent performance  at  50




 mph could probably be achieved with 11.8 percent less engine  output




 power  in the lighter vehicle  (135 hp versus 153 hp) .







 Fuel Consumption





 Historically, regenerators have been regarded  universally  as  essential




 for practical gas turbines in automobile applications, because simple-




 cycle  engines have been considered unsuitable  for  this application on the




 basis  of poor full-load thermal efficiency and, even worse, part-load




 efficiency.  Unfortunately, the addition of the regenerator considerably




 diminishes the appeal of the gas turbine engine in other respects, since




 it: (a) adds complexity, cost (manufacturing,  maintenance  and  development),




weight, and volume;  (b)  reduces specific power output and  reliability;




 and (c) complicates  control and response.







 Simple-cycle engine  efficiency (and specific power output) can be described




 conveniently by the  use of en ly three parameters,  i.e., pressure ratio,




 turbine inlet temperature, and flow path efficiency, where the latter




 represents both component efficiencies and pressure  losses and is a




measure of the aerodynamic sophistication of the engine  (see reference 14) .




Figure 3 shows the relationship between full-load  sfc and  these parameters

-------
                                      IV-19
 for a fixed  flow path  efficiency of 0.69 (see reference 15).  As can be


 seen, a wide range  of  combinations  of pressure ratios and turbine inlet


 temperatures will yield  sfc's  below 0.50,  which has been the goal for


 regenerated  engines for  automobiles.

                   0.80
                      1500    1700    1900    2100
                        TURBINE INLET TEMPERATURE,°F


        Figure 3.  Fuel consumption relationship  for  simple-
        cycle engines with constant flow-path efficiency

 The development  of small,  simple, advanced-technology components, with


 high efficiencies and high pressure ratios (see reference 16), permits


 reconsideration  of the simple-cycle gas turbine as an automobile power


 plant.  Although the fuel  consumption of this engine will be higher than


 regenerative designs based on  the same component  technology,1 and possibly


 slightly higher  than that  of competitive gasoline reciprocating engines,


 this engine may  be competitive  for  several reasons:  namely (a)  automo-


 bile customers are not necessarily  deterred by higher fuel consumptions


 (power mileages),  provided the  other  features (including  price)  appeal to


 them; (b)  the amount  of fuel burned by the gasoline  internal  combustion
(1)  Both simple- and regenerative-cycle engine concepts based on small,
    simple,  advanced-technology components are under study at United Air-

    craft Research Laboratories as part of the aforementioned manufacturing
    cost study.                                                            5

-------
                                    IV-20
engine, as well as the cost of the fuel may Increase as a result  of




pollution control measures; (c) the gas turbine can burn lower-cost




fuels; and (d) the possibility exists of governmental fuel  taxation




policy to favor the intorduction of low-pollution vehicles.






Manufacturing Cost—Although definitive production cost estimates




are highly proprietary at this time, General Motors has released data




indicating an opinion that the cost of automobile gas turbine engines




will be approximately three times that of an untreated reciprocating




engine, and from 1% to 2 times that of an eventual, pollution-treated




reciprocating engine as shown in Figure 4 (see reference 17).  These




costs appear too high for serious competition with reciprocating engines,




and, therefore, methods of reducing this level of manufacturing cost would




appear mandatory in order for the gas turbine to replace reciprocating




engines in quantities sufficiently large to have a significant effect on




air pollution levels.






With regard to manufacturing cost, Eckert predicts that a proper gas turbine




design adopted to modern manufacturing techniques should result in a cost




division of 85 percent for materials and 15 percent for labor (see




reference 10).  Therefore, reduction of material costs are  of primary




interest in reducing manufacturing cost.  Since superalloy  material costs




predominate in the raw material cost of today's gas turbine  engine, one




approach to reducing the manufacturing cost is to substitute ceramic materials




wherever possible for refractory  superalloys  in order to take advantage




of raw material costs of  5 to 40  cents per pound as opposed to  $1.50  to

-------
                                    IV-21
to $6.00 per pound; however,  the lack of tensile strength in ceramics




in an impediment to this  approach (see reference 18).
—
'1 80M960-NO CONTROLS
\ 1
«
E
0
5,60
UJ
Q
* 40
O
z
O
20
Z
O
ta
1 0
\
\

\
-1
M966-EXHAUST
_|
11970-IMPROVED EXHAUST
- '1970-CALIF. STD
51 IKLIN v»,
- 197X-EXHAUST REACTORS G£,\ J"" *!!1E'
,1111111111/11111" SltAM, tit.
1 | | .//////
                          RELATIVE POWERPLANT COST
           Figure  4.   Carbon monoxide reduction economics,
Another approach is to increase  the  specific  output  of  the engine by




increasing pressure ratio,  turbine inlet  temperatures,  and flow path




efficiencies, thereby reducing the size and material content of cost-




critical parts.






As stated in a previous footnote, a  study is  under way  concerning the




probable manufacturing cost of high  specific  output  engines.   Projections




will be made concerning probable manufacturing  costs of gas  turbine  engines




for automobiles in units of 100,000  and 1,000,000 annually,  based on




material content and related costs.

-------
                                    IV-22
Conclusions




Properly designed gas turbine engines for vehicular applications  offer




significant potential for alleviation of reciprocating engine-caused




air pollution.  From the emissions standpoint, this potential  is  repre-




sented by a predicted total emission level only 2 to 4 percent  of that




characteristic of untreated gasoline reciprocating engines and  only 15




to 30 percent of that of reciprocating engines meeting proposed 1975




emissions standards.







Therefore, widespread adoption of vehicular gas turbine engines has the




potential of practically eliminating vehicular air pollution as a




subject of serious environmental concern, provided that the performance,




efficiency, and manufacturing cost of these gas turbine engines are




competitive with the reciprocating engines they would replace.







Acknowledgment







The assistance of Drs.  F. L. Robson and C. T. Bowman of the United Aircraft




Research Laboratories in the preparation of the emissions section is




gratefully acknowledged.

-------
                                    IV-23
                             References
 1.  Morse,  R. S., "The Automobile and Air Pollution," A Program for
    Progress USGPO,  December 1967.

 2.  Middleton, J. T., "Future Air Quality Standards and Motor Vehicle
    Emissions Restrictions," National Conference on Air Pollution,
    Paper A-2, December 13, 1966.

 3.  Maga, J. A., and Kinosian, J. R., "Motor Vehicle Emission Standards--
    Present and Future," SAE Paper 660104, January 1966.

 4.  Anon.,  "Control  of Air Pollution from New Motor Vehicles and New Motor
    Vehicle Engines," The Federal Register, Vol. 31, No. 61, March 30, 1966.

 5.  Anon.,  "Air and  Water News," December 22, 1969 and March 9, 1970.

 6.  Brubacker, J., and Grant, E. P., "Do Exhaust Controls Really Work?",
    Second Report, SAE Paper 670689, 1967.

 7.  Sawyer, R. F., Teixiera, D. P., and Starkman, E. S., "Air Pollution
    Characteristics  of Gas Turbine Engines," Journal of Engineering for
    Power,  Transactions of the ASME, Vol. 91, Series A, No. 4, October
    1969, pp. 290-296.

 8.  Marteney, P. J., "Analytical Study of the Kinetics of Formation of
    Nitrogen Oxides  in Hydrocarbon Air Combustion," Combustion Science
    and Technology,  Vol. 1, No. 6.

 9.  Korth,  M. W., and Rose, A. H., Jr», "Emissions from a Gas Turbine
    Automobile," SAE Paper 680402.

10.  Eckert, B., "Has the Automobile Gas Turbine a Change?", A.T.2, Vol. 69,
    No. 9,  Sept.

11.  Hoffman, G. A.,  "Hybrid Power Systems for Vehicles," Symposium on Power
    Systems for Electric Vehicles, U. S. Department of Health, Education
    and Welfare, National Center for Air Pollution Control, 1967.

12.  Anon.,  "Truck Ability Prediction Procedure," SAE Handbook, Supplement
    82, Society of Automotive Engineers, May 1968.

13.  Penny,  N., "Rover Case History of Small Gas Turbines," SAE 634A,
    January 1963.

14.  Wood, H. J., "A  Polytropic Technique for Gas Turbine Performance
    Prediction and Evaluation," SAE 660161, January 1966.

-------
                                     IV-24
15. Kahle, G. W., and Wright, E. S., "Thermal Efficiency Versus Engine
    Price—Optimizing for Industrial Vehicles," ASME Paper No. 67-6T-41,
    March 1967.

16. Kenny, D. P., "A Novel Low Cost Diffuser for High Performance Centrifugal
    Compressors," ASME P,aper No. 68-GT-38, March 1968.

17. Anon., "G. M. Progress of Power," G.  M. Report, May 1969.

18. McLean,  A. F.,  "The Application of Ceramics to the Small Gas Turbine,"
    ASME Paper No.  70-GT-105, May 25, 1970.

-------
           V-l
        Chapter V
THE STIRLING-CYCLE ENGINE
            by
R. A. J. 0. van Witteveen
      Stirling Group
      N. V. Philips
Eindhoven, The Netherlands

-------
                                   V-2


 Air  Pollution  is a very serious issue which  is now  causing  considerable

 concern  throughout the world.  The convening  of  this meeting  demon-

 strates  this concern and we sincerely trust  that it will  not  only high-

 light  the problem but will also contribute to the solution  of air pollution,


 It is  hoped that the intensive Stirling engine research programme in  the

 Netherlands offers some positive contribution to the alleviation  of the

 air  pollution  problem and in particular, the automotive pollution.


 Based  on our experience with this engine, its characteristics,  the

 progressed state of the art and further prospects,  we are convinced that

 the  Stirling engine should be regarded as one of the promising  candidates

 for  alternative vehicle propulsion.  It represents  a very clean engine

 in the heavy duty vehicle application (on and off highway), and even  for

 light duty vehicles.


 The  aim  of this presentation is to provide an impression  of:

   (a) the nature of the work we have been doing and the  resultant
       expertise which has been achieved
   (b) the progress made
   (c) the future potential of the Stirling engine  and the envisaged
       applications
   (d) the formulated plans for the future


 It is impossible to adequately portray all the above facets within the

 time scale allocated for this presentation and for  this reason  tomorrow

 afternoon has been set aside to enable members of the conference  to

visit the facilities at Philips.

-------
                                  V-3





The initial development of the Stirling engine carried out over many




years was not specifically directed towards vehicle use.  However, this




period was valuable in providing a sound basis for the later research




and development phase which commenced during  the early 60's.







Fig. 1 shows a 1-98 engine undergoing test.   Engines are  identified




by the first digit indicating the number of cylinders and the remaining




digits the piston displacement.  About 30  of  the 1-98 engines have




been built used principally as test beds for  various experiments.  The




original rating was 10 HP but later versions  were rated at 20 HP.




Maximum speed is 3,000 RPM.







Fig. 2 illustrates another engine concept  namely an opposed piston type




which can be used for ultra-silent underfloor boat propulsion.  A 97%




efficient wormgear reduction and coupling  is  built into the crankcase.




Its rating is 100 - 200 HP at 3,000 RPM.







During 1968, it was decided to design a four  cylinder in-line engine which




could be used for vehicle as well as stationary applications, with a




rating or 100 - 200 HP.  Another criteria  was that it should be capable




of being mounted vertically or horizontally.  Fig. 3 shows this engine.







The total experience on modern Stirling engines during the past ten years




has been achieved on approximately 50 engines of seven different types.




Powers range from a few tens of watts to 400  HP and the total running hours




amount to 25,000.  Component testing has also been a feature of the




development programme and close to 1,000,000  hours have been logged over




the ten year period.  Engines have been endurance tested  for 5,000 hours

-------
               V-4
Figure 1.  1-98 Engine during test,

-------
Figure 2.  Four cylinder opposed piston engine.

-------
                  y-6
Figure 3.  Philips 4-235 engine.

-------
                                   V-7





at full load and vital running components for  over  10,000 hours.






We have a total of 30 dynamometer  test  stands.  Fig. 4  shows  three  of




them which are used to test the 4-235 engine.






During the past three years, a policy has been  formulated which high-




ligh£s the development of engines  for vehicle  applications as being of




prime importance for these vehicle applications.






Emissions




Fig. 5 illustrates the external combustion system which is the basis




of the modern Stirling engine.  The power piston and the displacer




are enclosed by the heater head, a composite unit comprising  the




burner, heater tubes, preheater and regenerator.  The system has some




excellent features.






The combustion is continuous and takes  place in a space all surrounded by




hot walls of 700°C (1,300°F), so there  is complete absence of quenching




effects, and combustion will be most complete.  This is also accomplished




by a very good mixing in the burner itself.  Furthermore, the amount of




excees air can be chosen freely, such that the design has a considerable




measure of flexibility.






Due to the pre-heating of the incoming  air the flame temperature is




rather high; nevertheless, the amount of NOX formed is very low.  This




is because local hot spots are avoided, and because the residence time in




the burner is very short (5 to 10 milliseconds).

-------
                                                                                                   1
                                                                                                  00
Figure 4.  View of test cells with 4-235 engines installed,

-------
                               V-9
Figure 5.  Heater head with piston and displacer.
              STIRLING   EXHAUST  EMISSION  (GVM)
HC
CO
NO-
     30 V. REC.
              1968-71  1972-75
              CYCLE   CYCLE
.02

1.0

.8
.16
.03

1.4

1.0
.20
                   1970-71
               STANDARDS
 2.2
23.0
                        1975-76      APSV
                   STANDARDS   PROGRAM
  .46         .14

 4.8        620

.4-.6         .40
      Figure 6.  Table of emissions,

-------
                                   V-10




Fig. 6  shows the results obtained on exhaust measurements.   The




emissions on a GVM basis are calculated for the  old  and  the  new cycle test.




The  standards and also the proposed standards are  shown.   It can be




seen that Cx, Hy and CO are extremely low.  The  NOX  figures  are valid




for a burner which is not specifically optimized for  low NO  ,   New




designs which use recirculation show measured values  which lead to




numbers as low as 0.2 GVM.  Recirculation offers no particular  problem




because it only marginally effects engine output and  efficiency.






It is also interesting to consider the comparison with the 13 mode




Californian emission cycle for heavy duty vehicles which is  shown in




Fig. 7.






Other developments envisaged are for so-called suppressed flame temperature




burners which also result in extremely low NO  emissions.
                                             X





Noise and Vibrations




The comparison between a Diesel and Stirling engine of the same power




is shown in Fig.  8 and amounts to about 20 dB.  The noise level from  a




Stirling engine is almost inaudible at a distance of  50 metres.






Controls




Fuel control is by means of a thermostat system.  Power/torque  is




controlled by means of working gas pressure.  A full  load change  in




either direction can be made within 0.3 seconds.  A cold start  takes  less



than 20 seconds.

-------
                                 V-ll
        HEAVY DUTY  VEHICLES  (> 6000  IBS)   Gm/BHR hr
                  Stirling



        HC   .04

        NOX 2.0 (0.7)rec

        CO                  2.2
2.1 (0.74)rec
               1973 _ 74     1975
                Calif.    Federal
12.5       5


40       25
Figure 7.  Comparison with 13 mode California Emission  Cycle.
        dB
        uo
        130
        120
        110
        wo
        90
        80
        70

                                                          I
                                          structure-borne noise
                                              refit
                                         sound pressure level
                                              ref:J0002ubor
            4   6  8 100     2
  4  6  8 WOO   2     468 10000
           FREQUENCY.cps
                Figure 8.  Noise  level comparison.

-------
                                   V-12






Braking




By varying the gas pressure in the working and buffer space it is




possible to obtain a braking effect up to 80% of full torque.







Overload




Engines are rated for continuous loads;  short time overloading is




possible.







Multi-fuel




The single cylinder demonstration engine in Fig. 9 illustrates the




various fuels which can be used with a Stirling engine.







B.H.P./Efficiency/Torque




These characteristics are shown in Fig.  10 and 11 for various values




of working gas pressure.  Note the advantageous torque increase with




decreasing speed compared with the Diesel engine.







The decision to concentrate on vehicle applications automatically poses




the question as to what precise vehicle applications are envisaged.




There is a distinction between professional or heavy duty vehicle propulsion




and light duty vehicle propulsion (passenger cars).  Different require-




ments as to power-weight ratio and initial costs in relation to fuel




economy and maintenance and durability will hold.  Stirling engines are




(will be) suited for both categories.







Heavy Duty Vehicle Application




The 4-235 engine shown in Fig. 12 will be mounted in a bus and is mounted




horizontally.  It is initially rated at 100 HP with a capability of being




increased to 200 HP.  The weight is 750 K.g.  The illustration shows the

-------
Figure 9.   Multi-fuel demonstration engine,

-------
                                       V-14
 OVERALL
EFFICIENCY
       n
                  500
WOO
1500          2000

ENGINE SPEED n.rpm
                                            t.BHP
                         Figure 10.  Efficiency  curves.
    dynamometer
   f torque
   Jatm
         250     SOO    79O
        engine speed n.rpm
1000     1290    1500    1790    2OOO    2890    2100
                             Fieure 11.  Toraue

-------
                                                                                  <
                                                                                  I
Figure 12.  Philips 4-235 engine with gearbox.

-------
                                   V-16





engine assembly with a Voith-DIWA type 502-3 gearbox.   Ten  of these




engines will be built for this and other applications.







Fig. 13 shows the bus with engine installed in the rear and  also  showing




the cooling system which necessitates an additional radiator to cater




for the increased heat load present in the Stirling cycle.   Fig.  14




shows the actual bus.






It is hoped that a considerable amount of experience will be obtained




from the bus project which will provide a sound basis for future




development.







Light Duty Vehicle Application




The passenger car is undoubtably the application which  commands most




attention and one of the main characteristics required  is a  low weight




power ratio.  A different and attractive concept appears to  be possible




which would bring the ratio down to as low as 2 - 3 pounds per HP.







The double acting principle is shown in Fig. 15.  The arrangement is




that the hot space (expansion space) of one cylinder is  connected to the




cold space (compression space) of the next cylinder with a heater,  regener-




ator and cooler between.  The double acting principle can be achieved by




more than one type of drive.  One of the most promising  of designs  is




the swashplate engine shown in Fig. 16.  Besides a volume and weight




advantage there is also a gain in simplicity of design  and cost reduction.






Heat Pipe System




A particularly attractive heat source for the Stirling  engine is  obtained

-------
                            V-17
Figure 13.  Installation of bus engine and cooling system.

-------
                             V-18

Figure 14.  Bus powered by Philips 4-235  Stirling Engine.

-------
                  V-19
                                             XPANSION SPACE
                                                  HEATER
                                                  REGENERATOR
                                                  COOLER
                                       W—^COMPRESSION SPACE
Figure 15.  Double acting Stirling cycle.

-------
                                                                                                               Crosshead



                                                                     Cihnder       Compression       Piston Rod    /  Slider Bearing

                                                                                      Space
Thrust

Bearing
Burner-Air Inlet
                                                                                                    Oil Pumps


                                                                                             Cooler Tubes
                                                                               Regenerator
                                                             Preheater


                                                      Connecting Ducts

                                                                                                                                                               i
                                                                                                                                                              M
                                                                                                                                                              o
               Exhaust Outlet
                                           Figure 16.   Swashplate  engine.

-------
                                    V-21

by the application of heat pipes.   A heat pipe is a closed volume in

which is incorporated a medium which can absorb and reject heat at a

very high rate; such a medium is sodium.  By a process of evaporation

and condensing, the external heat system of the Stirling engine is

vastly improved.  Fig. 17 shows a heat pipe system incorporated in a

swash plate engine.  Two major advantages emanate from such a design.

   (a) External and internal heat transfer can be considered
       independently and the Stirling cycle can be optimized
       to a still higher specific output.  It is possible to use
       higher speeds because heater flow losses are reduced.
   (b) It is hoped that the combination with the combustor
       can be made such that the combustion process and heat
       transfer are partly simultaneous.  The maximum flame
       temperature can be reduced to 1,600°C where hardly any
       NO  formation occurs.
         X


The development phase for such a promising combustion design has

already commenced.  Engines utilizing the heat pipe system represent

engines which are commonly referred to as being of the "second generation."


Reviewing the proceeding engine types one endeavors to be realistic.

The hardware experience gained has  been considerable; soon we will

acquire vehicle experience.  However, the engine development stage is

only just beyond the laboratory concept.  A considerable development

task particularly in the field of cost reduction and engineering lies

ahead before an economical acceptable engine is produced.  A considerable

effort is also being made in the field of new fabrication methods.  This

effort in advancing Stirling engine technology is almost entirely

carried out by Philips and its licensees in Germany and Sweden.  Other

license agreements are being negotiated.

-------
                            y-22
Figure 17.  Swashplate engine with heat pipe system,

-------
                                   V-23






In Eindhoven the Stirling effort is divided into two main groups:




the Research Group Stirling Engines and the Product Group Stirling




Engines.  The Research Group is primarily engaged in basic Stirling




engine research while the Product Group is engaged in engine develop-




ment for both vehicle and stationary applications.

-------
                            VI-1
                        Chapter VI
  RANKINE-CYCLE POWER SYSTEM WITH ORGANIC-BASED FLUID AND

RECIPROCATING EXPANDER FOR LOW EMISSION AUTOMOTIVE PROPULSION
                             by
                       Dean T.  Morgan
                   Special Products Staff
                Thermo Electron Corporation
                Waltham, Massachusetts, USA

-------
                               VI-2
1.  BACKGROUND ON WORK LEADING TO AUTOMOTIVE SYSTEM
    DESIGN

     Thermo Electron Corporation has been involved in Rankine-cycle
system development since 1963; this effort has been concentrated on
completely self-contained portable powerplants in the horsepower
range of fraction to ~ 200 hp.  During this period the systemhas
evolved from one of very limited practical application to one capable
of competing with internal combustion engines  in a number of com-
mercial  applications.

     The Initial work on  Rankine-cycle systems at TECO from 1963
to 1967 was concentrated on steam as a working fluid.  It was realized
that steamhad some severe limitations for a low-cost,  reliable system
capable of use at low ambient temperatures,  but the development of
alternative working fluids had not advanced sufficiently for practical
consideration of their use.   Since 1964,  many new working fluid
candidates have been proposed,  and sufficient thermodynamic and
physical property data have  been developed for evaluating their
potential as working fluids for commercially oriented applications.
After extensive analysis of all of these fluids,  thiophene was selected
in 1967 as the current state-of-the-art working fluid which best ful-
filled the system characteristics required for commercial  Rankine-
cycle systems.  Development of components and systems using this
fluid continued until September  1970,  culminating in operation of a
complete, self-contained 5 hp package.

     The primary limitation  for the use of thiophene in systems to be
operated by the general public is its high flammability and  toxicity...
In September 1969,  TECO began laboratory testing  of trifluoroethanol,

-------
                              VI-3
either  pure or mixed with water, as a working fluid.  The mixture of

                                                       *
85 mol% trifluoroethanol, 15 mol% water (Fluorinol-85)  was  selected



as the  best composition.   Fluorinol-85 has the r mo dynamic and physical


properties very similar to thiophene and, in addition, is completely


acceptable from a flammability and toxicity viewpoint for systems to


be operated by the  general public.  Since September 1970, testing of a


complete,  self-contained 5 hp powerplant (see Figure 1) with Fluorinol-



85 working fluid has been carried out with satisfactory results.   Both


thiophene and Fluorinol-85 have been operated in the same 5 hp package


(after  careful flushing and cleaning when changing fluids) with prac-



tically identical power and efficiency.




       Fluorinol-85, with its  completely acceptable safety characteris-


tics, represents  a  significant  advancement in Rankine-cycle technology


for automotive propulsion systems. Work on use of the  TECO Rankine-


cycle system as an automotive propulsion powerplant was initiated in


June 1969 under contract to the Division of Motor Vehicle Research


and Development, Air Pollution  Control  Office,  Environmental Pro-


tection Agency.   During the first year of the program,  a conceptual


design of the complete system was developed.  This conceptual design,


described in Sections 2 and 3 of  this report,  indicated a strong potential


of the TECO approach for a practical and competitive low-emission



automotive propulsion system.   The program is now in the second


year and involves conversion of  the conceptual design to a more


detailed, optimized design and experimental development on several



components,  specifically:
 Halocarbon Products Corporation, Hackensack, N. J.

-------
1 .    3   lew  Gen.eira.toir  Set  with  Side   Panels  Removed .

-------
                              VI-5
    a.  Analysis and bench-testing of full-size expander intake



        and exhaust valving approaches.





    b.  Detailed design, fabrication, and loop testing of a full-size



        feedpump.





    c.  Fabrication of a boiler third stage section and measurement



        of pressure drops and heat transfer rates.





    d.  Full-size rotary  shaft seal testing.





    In the remainder of this report,  a summary description of the



component designs, system performance and characteristics, and



system packaging in an intermediate size American car are presented,



based on the APCO-sponsored studies.  The component designs are



based on thiophene, since the details with Fluorinal-85 have not been



completed.  The differences are generally small; where differences



occur with the Fluorinol-85, they are pointed out.





    The system component sizes are  based on  the peak power require-



ments for an intermediate size American car to give a 0-60 mph



acceleration time of ~ 15 seconds or better and a top speed of 90-95 mph



Using these criteria,  the system peak shaft horsepower  is 100 shp. The



reference car size for the APCO alternative powerplant studies was



recently increased to a full-size American car. This  modification



will require an increase in the system power output of approximately



20% to meet the performance goals.






2.  COMPONENT DESCRIPTIONS




2. 1 System Working  Fluid and Cycle  Characteristics




    The working fluid to be used in the system is Fluorinol-85, a:



mixture of trifluorethanol and water containing  85 mol%  trifluoroethanol

-------
                             VI-6
and  15 mol% water.  As discussed above,  sufficient experience



has  been obtained with this fluid so  it can be considered a state-of-



the-art working fluid.  In  Table 1, a summary is given comparing



important cycle parameters for different trifluoroethanol-water



mixtures; the equivalent calculation is also presented for thiophene,



since this working fluid was used as the basis for the  component



designs given in this report.  Based on these numbers, Fluorinol-85



was  selected as the  optimum composition,  and is being used in testing



the 5 hp system.   This composition also  has the minimum freezing



point of -82 °F; in Figure 2,  the effect of water content on the freezing



point is illustrated.   The boiler outlet temperature and pressure are



selected as  550°F and 700 psia, respectively, to maximize efficiency



and  minimize the engine displacement for a given power output.  This



cycle is illustrated on the  pressure-enthalpy diagram of Figure 3.  It



should be noted that trifluoroethanol appears to be more stable ther-



mally than thiophene,  so that it maybe possible to  increase the boiler



outlet temperature to 600 °F during the program with a resulting im-



provement in cycle efficiency.  While Fluorinol-85 appears the opti-



mum composition at the present, future testing may indicate an alter-



nate  composition is optimum.




     In Table 2,  the overall design point cycle characteristics  for the



system are  presented.  The design point is based on a peak power



requirement of 100 horsepower at an engine speed of 2400 rpm.  For



comparison, the design point characteristics are presented in Table 3



for thiophene, the working fluid used in the  conceptual design study.



The  component designs presented in Section 2 as well  as the overall



system performance predictions of Section 3 are based on thiophene,

-------
                                 VI-7

                                TABLE 1

       CALCULATED CYCLE PARAMETERS FOR SEVERAL FLUORINOLS
                     AND COMPARISON WITH THIOPHENE

                     Release Pressure         =  75 psia
                     Condensing Temperature  =  200 °F
                     Subcooling             T  =A20°F
            Expander  Thermal Efficiency, rj.
                                           ExPth
            Expander  Mechanical Efficiency, n
                                            Exp,
                                                M
            Regenerator Effectiveness,  r)
            Pump Overall Efficiency, rj
                                        Reg
                                        OA
0. 8

0. 9

0. 9

0. 7
Fluid
Thiophene
Fluorinol
100
Fluorinol
85

Fluorinol
61

Fluorinol
51

Boiler
Outlet
Temperature,
°F
550
550
550
600
550
600
550
600
Boiler
Outlet
Pressure,
psia
500
400
50C
600
400
500
600
700
700
500
600
700
800
700
500
600
700
800
800
)7
Cycle,
%
1!'.. 7
K-.2
1!>. '
Ih, 0
If-. 8
If,. 2
:,(.. '••
It'.. 6
: ; ! , F>
ir . <'
I'''. !-'
Ib. v
If.. P.
1 7 „ 0
15. r
If. 7
1!'. 7
I1 . >>
1".0
QR/QB*
0.21
0.31
o.::9
0. 26
0. 24
0. 20
0. J9
0, 175
0. 2,2
0. M-
0 . Ji 3
0. 1 .1
0. 10
0. 16
0. 10
0. 087
0. 075
0. 061
0. 11
Expander CID,
in3/np
at 2000 rpm
1. 70
1. 84
1. 71
1, 58
1. 52
1.40
1.37
1,34
1. 23
1.56
1. 50
1.42
1.38
1.42
1.47
1.40
1.31
1.26
1.25
WE, Shaft
less
Wp, Shaft
Btu/lb
32.9
26, 5
23. 6
30.4 i
34. 8
36.3
37. 0
37.2 !
43. 1
44.4 |
4"-. 8
45.9
45.6
50, 2 ;
51.2 !
52. 9 i
52. 9
53. 3
58. 7
I
Regenerator heat transfer rate/Boiler heat transfer rate.

-------
                              VI-8
O
o

42
.£
o
0.
N
0)
4>
0
-10
-20
•30
•40
•50
•60
7O

\






























s
\













X—
f













"• -.






























^s,















-------
                      VI-9
                           BO   200   ZEO   40   240    ISO   BOO    SZO   MO   39Q   S*
•O   SO
Figure 3.  Pressure-Enthalpy Diagram and Cycle  Conditions
           for Fluorinol-85.

-------
                            VI-10
                          TABLE  2
       DESIGN POINT SPECIFICATIONS, FLUORINOL-85
Working Fluid

Boiler Outlet Temperature
Boiler Outlet Pressure
Boiler Heat Transfer Rate
Boiler Efficiency (HHV)

Expander Design Point Intake Ratio
Expander Displacement
Expander Speed
Expander Piston Speed
Exp.  Horsepower less Feedpump Power
Expander Thermal Efficiency
Expander Mechanical Efficiency
Expander Overall Efficiency

Regenerator Effectiveness
Regenerator Heat Transfer  Rate

Condensing Temperature
Condensing Pressure
Subcooled Liquid Temperature
Condenser Heat Transfer Rate

Working Fluid Mass  Flow Rate
Organic Volumetric  Flow Rate
Feedpump Overall Efficiency
Feedpump Power

Cycle Efficiency
Overall Efficiency
Fluorinol-85

550°F
700 psia
1, 600, 000 Btu/hr
82.5%

0. 137
107 in3
2400 rpm
1000 ft/min
103.2 hp
84. 6%
91.5%
77.5%

90%
276, 000 Btu/hr

217°F
43 psia
197°F
1, 312, 000 Btu/hr

7330
11.70 gallons/minute
70. 0%
6.4 hp

16.4%
13.5%

-------
                            VI-11
                          TABLE 3
               DESIGN POINT SPECIFICATIONS
                THIOPHENE WORKING FLUID
Working Fluid

Boiler Outlet Temperature
Boiler Outlet Pressure
Boiler Heat Transfer Rate
Boiler Efficiency (HHV)

Expander Design Point Intake Ratio
Expander Displacement
Expander Speed
Expander Piston Speed
Expander HP less Feedpump Power
Expander Thermal Efficiency
-Expander Mechanical  Efficiency
Expander Overall Efficiency
Expander IMEP
Regenerator Effectiveness
Regenerator Heat Transfer Rate
Condensing Temperature
Condensing Pressure
Subcooled Liquid Temperature
Condenser Heat Transfer Rate

Organic Mass  Flow Rate
Organic Volumetric Flow Rate
Feedpump Overall Efficiency
Feedpump Power

Cycle Efficiency
Overall Efficiency
Thiophene

550°F
500 psia
J.58 x 106  Btu/hr
82. 5%

0. 137
184 in3
2000 rpm
1000 ft/min
K*7 2 hp
8-, 6%
c l , 5%
7'\ 5%
127,4 p s i

90.0%
0. 249 x 106 Btu/hr

216. 2°F
i:.r. 0 psia
 %. 2°F
*. 25 x 106  Btu/hr

7377 pounds/hr
' "• I gallons/min
5><->. 7%
5.25 hp

\(,. 7%

-------
                              VI-12
since the details with Fluorinol-85 have not been completed.  The
differences are generally small; where differences occur with the
Fluorinol-85, they are pointed out.
2. 2  Expander Design
     Establishing the dimensions of a new expander design depends,
among other things, on the  prediction of  the indicated and mechanical
efficiencies of the expander at the design condition.  Consequently, a
detailed analysis was carried out with thiophene to determine these
efficiencies as functions of  piston speed and load,  using measured
efficiencies obtained with the 5 hp expander on test at Thermo Electron
as a check.  The performance with Fluorinol-85 should be almost
identical to the thiophene results.  As a result of the analysis, it is pos-
sible to plot expander  efficiency versus piston speed at various
loads.  One such plot is shown in Figure 4 for an indicated mean
effective pressure  (IMEP) of  125 psi.   The rapid drop in efficiency
with piston speeds  above 1000 ft/min occurs due to inlet valve losses.
     From this analysis, a piston speed of  1000 ft/min was selected
for the design condition of 103 bhp at a vehicle speed of 95 mph.  The
reduction in expander  size which could be realized by selecting a
higher piston  speed would probably be more than lost in boiler and
condenser size increases due to  lower overall cycle efficiency.
     The IMEP and BMEP are determined by the cycle design condition,
and the  BMEP and  the piston  speed determine the piston area required
to develop the desired horsepower.  A 90 °V of four cylinders was
selected as being reasonably  compact without either an excessive
number of moving parts or  excessive torque variation.  The V design

-------
                             VI-13
  90
  80
o:
£70
o
  60
  50
  40
                            £)ual Inlet Valve
             Single Inlet Valve
Release Pressure - 75psia
             Pi - SOOpsia
             Ti = 55O°F
         200     600    .1000    1400     1800    2200
                        Speed, ft/min
             Figure 4.  Overall Expander Efficiency Variation
                      with Piston Speed

-------
                             VI-14
results  in a short engine for a given number of cylinders and facilitates
packaging of the system.  With four cylinders,  the resulting bore with
Fluorinol-85 is 3.68 inches.  The mean piston speed (1000 ft/min at
design)  and the engine speed are related by the expression

                           S = 2  LN
where S = mean piston speed,  L = stroke,  and N = rpm.  The selected
design point speed of 2400 rpm, based on a reasonable bo re-to-stroke
ratio  (1.47) and on valve train dynamics, results in  a stroke of  2. 5
inches.   The basic expander dimensions and specifications are given
in Table 4,  with the values previously used for thiophene given for
comparison; cross-sectional views of the expander'with dimensions
given for thiophene are presented in Figures 5  and 6 with hydraulic
expander valving, slipping-clutch transmission, and feedpump
incorporated.   The overall expander dimensions will be reduced
somewhat with Fluorinol-85.
    All materials are identical to those now used in  automotive
internal combustion engines.   The Rankine-cycle expander differs
from  the current automotive internal combustion engine in two very
important aspects of its design:  the inlet valving and the bearing
design.
2.2.1  Variable Cut-Off Inlet Valving
    The importance of having  a valving  system with  variable cut-off
is established in  Section 3.  Apart from the difficulty in varying the
cut-off, the valving problem is considerably more severe than in
internal combustion engines.   To avoid excessive losses,  the high

-------
                           VI-15
                          TABLE 4
       EXPANDER DIMENSIONS AND SPECIFICATIONS
  Working Fluid
     Thiophene
   Fluorinol-85
Configuration
Bore
Stroke
Displacement
BMP
(feedpump work deducted)
IMEP at Design
BMEP at Design
Four Cylinders, 90 °V
4. 42 inches
3. 0 inches
184 in3
103 at 2000 rpm

127 psi
117 psi
Four Cylinders  90 °V
3.68 inches
2. 50 inches
107
103 at 2400 rpm

175 psi
160 psi

-------
                                  VI-16
0
-J
3
                Pigure 5.    V-4 Expander with Hydraulically
                             Actuated Valves, Front View.

-------
                                                                                                  H
                                                                                                  I
©'
    Figure"6.   V-4 Expander with Hydraulically Actuated Valves, Side View.

-------
                              VI-18
density of the vapor at expander inlet conditions necessitates an inlet


valve comparable in diameter and lift (and therefore mass) to the


intake valve of an internal combustion engine.  However, the valve


event is much shorter in the Rankine  expander than in the internal


combustion engine.  The design point intake ratio of 13. 7%  corres-


ponds to a  maximum valve event of 60°, whereas in internal com-


bustion engines the inlet valve  event  is on the order of 240° or more.


In a cam-operated system at a given  speed,  the acceleration and2



therefore,  the stress level are  proportional to the lift  divided by


the square of the valve  event; the cam stresses are much higher


at a given expander speed for the vapor expander.




     One way of overcoming the problem is shown in Figure 7. In


this system two concentric inlet valves in series are driven by two


separate camshafts.  Cam number 1,  driving inlet valve 1,  has fixed


timing with respect to the  crankshaft.  Cam number 2  has variable
                                                            i

timing with respect to the  crankshaft, and the total valve event'  is


determined by the overlap of the two  valves.   In this way, relatively


long cam events can be used,  giving  reasonable sized camshafts.


Figure 7 also shows that the mean valve opening area can be higher


with this approach than with a single  valve,  which should compensate


for the lower flow coefficient of the two valve system.   A modifica-


tion of this concept is being set up for bench testing at TECO.




    Other  approaches to variable cut-off inlet valving  are shown in


Figures 8 and 9-   These are both hydraulic devices.   A directly-


actuated hydraulic system is shown in Figure 8: A cam-operated


plunger pump operates  a hydraulic column which acts on a stepped


piston on the inlet valve stem.   The pump plunger  is constructed with

-------
                VI-19
               7TZJ.C.
           CRANK ANGLE
Figure 1.   Two Inlet Valves in Series.

-------
                  VI-20
                                                        LOW PRESSURE
Figure  8.     Directly Actuated Hydraulic Valve.

-------
LOW
PRESSURE
                                                                                                       360"
                                                                                            OF"
                                  Figure 9.     pilot Operated Hydraulic Valve.

-------
                             VI-22
a helical undercut so that its angular position in its bore determines



its effective stroke, thus varying inlet valve duration.  This system



is quite similar to diesel engine injection systems but has the, disad-



vantage of a relatively large power requirement.




     Another hydraulic scheme is shown schematically in Figure 9.



In this system,  a pump supplies high pressure oil at  700  1000 psi



to a  rotary valve (shown as two valves for simplicity in  Figure 9).



The  rotary valve supplies the high pressure oil to alternate sides of



a piston connected to the inlet valve.  Cut-off adjustment is obtained



by moving the rotary valve  axially in its housing.  Thermo Electron



Corporation has contracted with the British Internal Combustion



Engine Research Institute for fabrication and bench testing of a full-



size valving system based on this concept.  A subcontract to American



Bosch Corporation has also been arranged under Phase  II  of the APCO



program at TECO for bench testing of a similar concept which uses a



solenoid-operated pilot valve in place of a rotary valve.  Both of these



approaches have a low power requirement of about one horsepower for



a 4 cylinder expander.





     It is thus  expected that bench test results from three different



approaches will be available for selection of the expander  intake



valving system to be used in the system.   Phase II of the APCO



program also  involves bench-testing of a full-size exhaust valve



scaled up from the exhaust  valve used on the TECO 5 hp expander.




2. 2.  2  Bearing Design and Selection




     The expander bearing design is strongly'influenced by the  type



of transmission used*  If the expander is coupled directly  to the

-------
                             VI-23
driveshaft, as in many early steam cars,  journal bearings relying
on hydrodynamic lubrication cannot be used; roller or ball bearings
must be used, because of the high bearing loads which could occur
at essentially zero rpm.  On the other hand, if a conventional torque
converter were used,  bearing sizes, at least on the crankshaft,  can
be comparable to an internal combustion engine of the same bore,
since the peak cylinder pressures are roughly the  same and the
inertia loading is lower on the  Rankine expander (because of the
1000 ft/min limit on piston speed).  The wrist pin  bearing is more
heavily loaded than in the conventional four-stroke  internal combustion
engine bebause the load on the pin never reverses;  in this respect,
it is much like the wrist pin in a two-stroke engine.  Any single
speed transmission can load the  expander bearings fairly heavily
when the expander is idling at 360 rpm and the  clutch is engaged.
At these conditions, with a maximum intake ratio of 80%,  the bearing
loading is such that bearing sizes associated with a two-stroke diesel
of the same bore would be barely adequate for the  Rankine expander.
A transmission with two forward speeds as well as a  reduction in
the maximum intake ratio,  as planned for the system, will alleviate
this situation.
     It is expected that, with development,  conventional journal
bearings can be  used in the expander if  the transmission used
permits the expander to idle at zero vehicle speed.  However, to
minimize the development problems and the consequences of
momentary lubrication failure on the system, needle  bearings
will be used throughout the expander for the first prototypes.
Needle bearings have been used in the 5 hp expanders tested
at Thermo  Electron Corporation with excellent reliability.

-------
                             VI-24
2. 3  Feedpump Design
     A full-size feedpump sized for the thiophene is under loop-test
at TECO as part of the Phase II of the APCO program.  This pump
was designed and constructed at TECO.  The primary factors con-
sidered in the selection and design of the vapor-generator ieedpump
are that the pump must be positive displacement because of the high
discharge pressure.   The lubricity of either Fluorinol-85 or  thiophene
is relatively poor and its liquid viscosity low.  The pumping rate must
be variable from basically  zero to 15 gpm for thiophene over a 800 -
2000 rpm range,  and the pump  must operate with low net positive
suction head (NPSH) without cavitation, since the NPSH is provided
primarily by subcooling of  the liquid coming from the  condenser.
     The feedpump selected, illustrated in Figure  10,  is a 5-cylinder
piston pump driven by a wobble plate; its characteristics are summar-
ized in Table 5.  The  selection of a piston pump was based on testing
of several types of positive displacement pumps, including gear and
vane-type pumps, at Thermo Electron.  In general,  high leakage
rates (low volumetric efficiency) and high wear rates have been
encountered for all pump types  other than piston, which have given
completely satisfactory performance.  With the piston pump all
bearing surfaces can be oil-lubricated.
     The variable pump rate is obtained by incorporation of variable
displacement in the feedpump, permitting the pumping rate to be con-
trolled at the desired  rate regardless of feedpump speed.   The method
used to obtain variable displacement is to vary the effective displace-
ment of the feedpump by bypassing all or part of the pump output to
the  condenser.  The effective displacement is obtained by moving the

-------
         BY-PASS
                                                                          H
                                                                          I
                                                                          N5
                INTAKE
Figure 10.  Feedpump Cross Section.

-------
             VI-26




          TABLE  5





FEEDPUMP CHARACTERISTICS
Working Fluid
Pumping Rate
Number of Cylinders
Volumetric Efficiency
Overall Efficiency
Range for Maximum Pumping
Rate
Total Displacement
Bore
Stroke
Materials of Construction
Housing
Pistons and Valves
Bearings
Thiophene
15 gpm
5
90%
80%
800-2000 rpm
5. 51 in3
1. 875 in
0. 4 in

Cast Iron
Hardened Steel
Needle
Fluorinol-85
11.7 gpm
5
90%
80%
960 -2400 rpm
3,58 in3
1.51 in
0. 4 in

Cast Iron
Hardened Steel
Needle

-------
                              VI-27
entire cylinder block to regulate the axial position at which the bypass



port is uncovered during the discharge stroke of each of the five pistons



The bypass porting is configured so that pumping to the high pressure



exhaust occurs always  from bottom dead center; this procedure mini-



mizes the  discharge pressure transients and noise level of the pump.



The bypass flow is also never pumped to the high pressure,  thereby



minimizing the feedpump parasitic  load on the system.  The control



lever for positioning the cylinder block uses a bellows rather than



a sliding seal to maintain the hermetic nature of the system.





     Spring-loaded poppet suction and discharge valves are used.



The suction valve is constructed in the cylinder and is made as



large as possible  to minimize the pressure loss through the valve



and the tendency for cavitation.,  The smaller discharge valve is



located in  the cylinder  head.  Common suction and discharge plenums



for all five cylinders are incorporated in the housing  castings. The



use of five cylinders  was based on reducing pressure transients due



to the flow variation  from the piston pump.  These transients must



be maintained  sufficiently small on the suction side of the pump that



the liquid pressure never falls below the vapor pressure of the sub-



cooled liquid.  A computer analysis of the pressure transient behavior



indicated that a five-cylinder pump would be required to prevent



cavitation with 20 °F subcooling at the pump suction.





     The pump  drive could be either crank or wobble-plate.  The



wobble-plate drive was selected because of  its compactness and



easier packaging with the expander,  its lower weight and vibration,



its quieter  operation at higher speeds, and its more convenient



geometry for variable displacement incorporation.

-------
                               VI-28
2. 4  Burner-Boiler Design





       The burner-boiler design has not been worked out in detail for



Fluorinol-85.  In Figures 11,  12 and 13 the boiler tube bundle design



and the burner cross sections are  presented, based on detailed com-



puter analysis for the thiophene  fluid.  The design requirements are



summarized in Table 6.  Some modification will be necessary for



Fluorinol-85,  due to the larger superheat requirement for Fluorinol-



85 relative to thiophene.   The changes will be restricted to modifica-



tions of the tube  bundle in each stage,  with no change in the overall



burner-boiler  envelope or in the flow paths in the boiler.  The factors



considered in arriving at the boiler design,  in addition to heat transfer



performance,  were low materials  cost, low volume,  easy construction,



and low combustion side pressure  drop.  Maintaining a low combustion



side pressure  drop is of extreme importance in meeting the develop-



ment goals of fast  startup and low  parasitic power loss.  For startup,



the complete combustion  system must  be electrically driven by the



battery.   To achieve fast startup,  it is essential that the burner be



operated at its maximum rate.   The total peak power requirement



must therefore be maintained at  as low a level as possible, consistent



with the packaging  envelope  and heat transfer rating required.  The



current design at full burning rate  requires about 1  1/4 hp total to



operate the combustion air blower  and fuel nozzle compressor.  This



power level is practical considering the battery, alternator and motor



sizes required.





       With reference to Figure 11, the  combustion gases, at a tempera-



ture of ~  3300°F, flow from the  combustion chamber into the center of



the tube bundle and radially  outward through the tube bundle.   The flow

-------
                       VI-29
                                                           23 IN
Figure 11. Cross Section Through Burner-Boiler,  Short Axis.

-------
                          VI-30
L	
   Figure 12.    Top View of Boiler Tua3 Bundle.

-------
                          VI-31
Figure 13.  Cross Sections Through Automotive-Size Burner.

-------
                        VI-32

                      TABLE 6

      BURNER-BOILER DESIGN REQUIREMENTS
Reference Cycle Boiler Heat                  /
    Transfer Rate                   1.60 x 10  Btu/hr

Maximum Boiler Heat  Transfer
    Rate                            1. 70 x 106 Btu/hr

Burner Design Maximum Heat                 ,
    Release Rate (HHV)              2.06 x 10  Btu/hr

Boiler Design Efficiency (HHV)       82.5%

Turndown Ratio                      15/1

-------
                              VI-33
path of the organic through the tube bundle is illustrated in Figure 14.
The organic first flows through stage 3,  from which the combustion
gases are exhausted; this provides the lowest organic temperatures
in the boiler at the combustion gas outlet and a high boiler efficiency
without air preheat.  It is important that an extremely compact and
efficient heat transfer surface be used in this stage to maximize the
boiler efficiency with acceptable pressure  drop on the combustion
side.  The organic next flows through the inner stage (through which
the combustion gases first flow),  with a resultant high heat transfer
coefficient.   Because of the high gas temperature and extended surface
on the combustion side,  coupled with the high heat transfer coefficient
on the organic side,  a very high heat transfer rate can be obtained in
the first stage.  The organic next flows through the superheater coil
or stage 2.   This stage is a bare tube coil,  since the  controlling
thermal  resistance is on the organic side and an  extended heat transfer
surface is not required on the tube.
     The characteristics of the three boiler stages are given in Table
7; Figure 15 presents the calculated design point temperature and
pressure profiles  through the boiler.   In the last or third stage,  a
matrix made of  steel and copper balls  brazed together and to the tube
is used.  This type of extended  surface provides  a very high heat trans-
fer rate per unit volume and is  amenable to mass production techniques.
Thermo  Electron Corporation is currently making heat transfer  and
pressure drop measurements on a full-size section of the boiler third
stage as  part of Phase II of the  APCO program.   Other types of
extended surface for the boiler  third stage will also be evaluated ex-
perimentally.

-------
                   VI-34
        ORGANIC FLOW TO ENGINE
ORGANIC FLOW
   TO BOILER "

   STAGE NO.  3
                                COMBUSTION
                                    GAS
                                   FLOW
      Figure 14. Organic Flow Path through Boiler Tube Bundle

-------
                                      TABLE  7
                          BOILER DESIGN SPECIFICATIONS

Stage
No.

1
2
3


Total
^
Woo f
Transfer
Rate
Btu/hr
1. 05 x fo6
0. 33 x 106
0.36 x 106

Combustion Gas Temp.
°F
entering

3330
1670
] 114

leaving

1670
1114
467

1 1
]. 74 x 10°

—
—
i
Tubing
Length
ft.

17
35
27


79

Pressure Drop
Combustion
Side in w. c.

0. 17
0. 33
2. 26


2. 76

Organic Side
psi

—
—
__


46

  be Spe c if i cations
Inner Tube   ID
             OD

Outer Tube  ID
             OD
0.930"
1. 000''

1. 125"
1.315"
F_in_ Specifications

Fins/Inch      10
Fin Thickness  0. 012"
Fin Material   Copper
Fin Height     0. 356"
                                                                                                    <
                                                                                                    I—I
                                                                                                    I
                                                                                                    Ln
Matrix Specifications

Ball Size            3/32"
Ball Material  50% Carbon Steel, 50% Copper
Matrix Thickness   0.5"
Matrix Height  0.970"
   (between tubes)

-------
                                VI-36
    650
    600 -
    550 ^
    5OO -
i- CL
3  -
Q. \fi
E £
o> a.
450 	
    400 -
     350
    300 -
                                            Tg, = 3330

                                               1895.9

                                            Tg3= 1190.0

                                            Tg,, = 490.24
                     20     30    40    50

                        Tubing Lengthfrom Inlet. Feet
                                              60     70
80
      Figure 15.    Design Point Boiler Temperature  Profile

-------
                              VI-37
    An intermediate heat transfer fluid (water) is used,  with double



tube construction in the boiler tubes,  to positively prohibit hot spots



on the organic side of the boiler.   Stages 1 and 2 are connected to the



same water reservoir,  which represents the high pressure side of



the boiler.  The water side of stage 3 is separate and represents the



low pressure water side of the boiler.





     The  combustion chamber design,  illustrated in Figure 13,  is



based on a volumetric burning rate of 2. 8 x 10  Btu/hr/ft  and  is



scaled from the burner used in the 5 hp system currently under test.



The burner is constructed integral with the boiler tube bundle,  as



illustrated in Figure 11.  To  reduce the pressure drop within the



diametrical clearance available,  two identical burners operating



in parallel are used rather than one longer burner with the same



combustion chamber diameter.  The pressure drop at maximum



firing rate for the two-burner setup is 1. 5" w. c.





     While no pollution measurements are available on the full-scale



burner design illustrated,  Thermo Electron Corporation has completed



measurements on a smaller scale burner  (140,000 Btu/hr) with per-



formance characteristics similar to the burner illustrated in the



design.   This burner  is being used on a 5 hp system now on test at



Thermo Electron Corporation.  Figures 16 and 17 present the steady



state emission levels from this burner as a function of excess air



for burning rates of 105,000 Btu/hr and 50,000 Btu/hr,  respectively.



It is apparent that the emission levels are extremely low.  To indicate



the transient performance of the  burner, the burner was oscillated



between 50, 000 and 105, 000 Btu/hr burning rates with constant fuel-to-



air ratios maintained; CO and unburned hydrocarbon  emission levels

-------
                      VI-38


a.
a.
C/5
z
o
(/)
UJ


200-
180-
160-
140-
120-
100-
80 -
60 -
40 -
20 -
STEADY
Q=50,OOC
FUEL JP-<






*^ 	 *- — ^^i^
• • • •
°~' I I ! I I I
0 10 20 30 40 50 60
                                                          NO
                                                          CO
                                                          CH
              EXCESS AIR (%)
Figure 16.  Effect of Excess Air on Emissions,  50,000 Btu/hr

-------
200-
180-
160-
140-
120-
100-
 80-
 60-
 40-
 20-
  0-
I
0
                                  STEADY STATE DATA
                                  Q=105,000 BTU/HR
                                  FUEL JP-4
                                 'X NO
            i
           10
 I
20
      30    40
EXCESS AIR (%)
 I
50
 i
60
                                                          OJ
                                                          VD
Figure 17. Effect of Excess Air on Emissions, 105,000 Btu/hr

-------
                           VI-40
were monitored continuously while a bag sample was  collected through-



out the run for NO measurement at the end.   The results are indicated



in Table 8.  As indicated in Section 3, use of these emission concentra-



tions with the system performance gives gm/mile emission levels  sig-



nificantly less  than projected limits under the recently enacted Clean



Air Legislation in the U. S.




2. 5  Condenser Design




    The condenser fan power represents the largest parasitic load of



the system.  A very important goal in the design  of the component  is



to minimize the required fan power.  The factors considered in the



condenser design, as well as its integration and packaging with the



system,  were:




    a.   An efficient (low f/j ratio) extended surface should be used



         on the air  side of the exchanger.





    b.   The organic side should be circuited to provide high vapor



         velocities leading to high organic side heat transfer



         coefficients.





    c.   The frontal area of the condenser (as well as the air free



         flow area)  should be made as large  as possible.




    d.   Highly efficient condenser fans should be used.





    e.   The system packaging and vehicle modifications should



         be designed to  minimize grille  (intake) and engine com-



         partment (exhaust) air pressure losses.

-------
                 VI-41

              TABLE 8

     TRANSIENT EMISSION DATA
FIRING
RATE
(BTU/HR)
105,000
50, 000
50, 000
50,000
50, 000
105, 000
105, 000
105,000
50,000
50,000
50,000
50, 000
105, 000
105,000
105, 000
50, 000
50, 000
50, 000

EXCESS
AIR
(%)
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%

CH
X
(PPM)
6
-
5
-
4. 5
401
7
5
5
4
4. 5
-
151
-
6
-
4. 5
4

CO
(PPM)
60
-
30
25
70
1801
90
-
80
25
20
-
1000+1
-
75
75
30
15

NO
X
(PPM)
-
-
-
-
-
-
-
-
-
-
-
-
-
-

-
-
422
392
ELAPSED
TIME
(MIN)
0
. ?
1. 0
1. 5
2. 0
2. 5
3. 0
3. 5
4. 0
4. 5
5. 0
5. 5
6. 0
6. 5
7. 0
7 5
8. 0
9. 0

NOTES:
        i
        2
Short duration peak ^,10 sec.
Exhaust gas sample collected during
the 9 minute run.   Then two samples
were drawn and NO analysis performed.

-------
                            VI-42
    f.    The condenser fan control should be optimized to minimize



         the fan power requirement by reducing the fan power when



         not required.  The condenser fans are directly driven by



         the propulsion engine.




    The condenser core of the conceptual design is illustrated in



Figure  18 with the design point characteristics, as obtained with



the condenser computer program, given in Table  9.




    The condenser core is similar to a Ford radiator with louvered



fins, except that the flattened tubes have  a heavier wall (0.030" vs



0.005")  and are constructed in one integral piece with partitions used



to provide  the desired vapor-side flow path.  Copper fins with 2. 5 mil



thickness are used,  and the tubing is made of carbon steel  rather



than brass  as in the Ford radiator.  The  frontal area used  in the design



represents the maximum practical area for an intermediate size



American car with some rework of the front end frame and grille.




    It is expected that improvements can be made to the condenser



design illustrated here.  Use  of a strip fin rather than the louvered



fin results  in an improved f/j ratio and will reduce the fan power



required for  a given condenser physical size.  If the fluid compati-



bility properties are  satisfactory, use of an all-aluminum condenser



will significantly reduce the condenser weight and will permit easier



construction.




2. 6 Regenerator Design





    The regenerator design conditions are illustrated in Figure 19



and the  regenerator design in Figure 20.   While the design is based



on thiophene,  the requirements  are very  similar to those for

-------
                                                                                            3 0"
                                               — 501L	
H
I	==
                                                                                              ZO. 6'
                             Figure 18. Condenser Design.

-------
                            VI-44
                         TABLE 9

      CONDENSER DESIGN POINT CHARACTERISTICS


Heat Rejection Rate
    (20 °F Superheat Entering                         ,
     20 °F Subcooling Leaving)               1.25 x 10  Btu/hr

Length                                      50 in.

Height                                      20. 6 in.

Depth                                       3.0 in.

Frontal Area                                7. 15 ft

Condenser Inlet Pressure                    25 psia

Design Ambient Air Temperature            95 °F

Air Pressure  Drop                          3.45 in.  we

Ideal Fan Power                            7. 91 hp

Air Flow  Rate                               63, 000 Ib/hr

-------
           Vapor

    25  psia, 348°F
(550-A P) psia, 285°F
1
                                        Q = 249,000 Btu/hr
                           (25-A P) psia, 230°F
            Liquid
                           550 psia, 199°F
H
I
                                                    Ln
                    Liquid Flow Rate = Vapor Flow Rate = 7377 Ibs/hr
                     Figure 19.  Regenerator Design Point Requirements.

-------
-775-
                                                                                                                   <
                                                                                                                   H
                                                                                                                    I
                                 Figure 20.  Regenerator Design.

-------
                           VI-47
Fluorinol-85 and only minor design changes are expected.   The design


is based on obtaining a  compact regenerator with geometry suitable for


packaging directly above the expander in the expander compartment


and with low pressure drop on both the liquid and vapor sides.   On the


vapor side, a brazed ball matrix extended surface with 1/16" ball


diameter is used.  The exchanger  is divided into four parallel liquid


circuits; in each circuit,  the vapor passes through four separate


stages,  permitting  the exchanger to approach a pure counterflow


exchanger.


                                                                  *
     The regenerator is also designed to function as an oil  separator.


The  low vapor velocities coupled with the ball  matrix should provide


very efficient separation of the lubricating oil  droplets from the ex-


haust vapor from the expander.  A drain line returns the collected oil


back to  the expander crankcase.



2. 7  Automatic Transmission



     An  automatic transmission will be used which permits the  main


propulsion expander to  idle at  zero vehicle speed and thereby drive


the  system and vehicle  accessories.   The use  of a transmission also


permits  matching very  closely the  driving characteristics  of current


automobiles powered by the I/C engine with three-speed automatic


transmission.  The  Dana  Corporation,  Toledo, Ohio,  as part of Phase


II of the APCO program,  is designing a two-speed slipping clutch


transmission based  on the type of transmission they develop and


produce for off-the-road vehicles.
 Patent Application Pending

-------
                              VI-48
2. 8  Controls





        The startup and control system is designed to provide completely



automatic  startup and operation of the system.  The  functions required



of the  driver are identical to those now required in current,  conventional



cars with automatic transmission.  The  controls are assembled from con-



ventional components with the exception  of the burner fuel-air control,



which  is currently under development at TECO.





        The primary control problem in the system is control of the burning



rate and of the pumping  rate to the monotube vapor generator to main-



tain boiler outlet pressure and the temperature within specified limits



over any type of transient encountered by the  system.  The control



system is based on operating the boiler as close to quasi-steady state



as possible over all transients;  both burning  rate and pumping rate are



maintained as closely as possible to the  values  corresponding to the



instantaneous vapor flow rate from the regenerator.





        A  schematic layout of the principal control components is presented



in Figure 21.  The feedpump control, used to control the pumping rate to



maintain boiler  outlet pressure, is  simplified by the  fact that the organic



flow rate at any expander rpm is approximately linear with intake ratio.



The feedpump and expander valving are thus directly operated as a unit



by the  accelerator pedal; a vernier  control working from the  boiler outlet



pressure is used to reduce deviations from the  design point and to eliminate



any unbalance in the system automatically.  A mechanical governor is used



to limit the maximum intake ratio as a function of rpm and to govern



expander speed  at idle.





       In order to maintain the burning rate at  a value corresponding to



the organic flow rate into the boiler,  the burner control uses an orifice



in the organic line to sense almost instantaneously any changes in the

-------
           Accumulator  for
            Constant  Fuel
              Pressure
Pressure By
 pass Valve

-------
                             VI-50
organic flow rate.  This signal,  along with a similar signal from an
orifice in the fuel line, is  used with a diaphragm controller to provide
the  proper  fuel and air flow rates.  If necessary, the fuel-to-air
ratio can easily be varied  as a function of turndown to minimize
pollutant emissions at any burning rate.  A vernier control operating
from the boiler outlet temperature  is used to reduce deviations from
the design point value; at low power levels,  when the organic flow is
low, this temperature control becomes the primary control on the
burner.  The fuel control valve cross  section is presented in Figure 22
and the air control valve cross section is presented in Figure 23.

3.  SYSTEM PERFORMANCE AND PACKAGING
3. 1  Performance
     The important decisions •which have a strong influence on the
overall system performance and cost relative to the internal com-
bustion engine are the method of driving accessories at  zero vehicle
speed and  the choice  between constant intake ratio  expander valving
with throttle valve control and variable intake ratio expander valving.
With respect to the first decision,  two alternatives are feasible. An
auxiliary5  constant speed expander can be used to drive all accessories,
permitting the expander to be  coupled directly  to the driveshaft (through
a simple geat transmission for forward, reverse,  neutral, park control).
A more complex transmission can also be used, permitting the main
propulsion expander to idle at zero vehicl/fe speed so that all accessories
canbe drivendirectly by the  main propulsion expander; this transmission
is still simpler than that required for the I/C engine-driven system,

-------
      '-"-"••~\   \    \--~~
<3o;  iff,     ,25.  \    -J5j
 \ i29'. \ V2Z)   A  \ 27, \ fcl
  \-T\T    \   \  \ \  f
fir*)
     ,-««
\  (^V. ^)
\  \^-\    ^  \ ^
                                                                                             <
                                                                                             M
                                                                                             t
       Figure 22.  Fuel Control Valve Assembly,

-------
                                                                      ft*
                                                                                                      I
                                                                                                     Ul
                                                                                                     K>
                                              - Up*.Kea?n
Figure 23.  Combustion Air Control.

-------
                             VI-53
       It has been the conclusion of this study that the use of the main


propulsion expander to drive all accessories is preferable.   Even though


the accessories must be larger when driven by the variable-speed main


propulsion expander, the accessory designs are identical and the same


number of parts must be processed; very little cost differential exists


between the  different sized components.  Any cost reductions due to


smaller accessory components are more than counterbalanced by the


requirement for an additional expander of 1 5  20 hp to handle  short term


accessory peak loads, with governor-throttle valve control.  The system


with  all accessories driven by the main propulsion expander,  therefore,


seems preferable in terms of simplicity, cost,  and packaging; this


system has been selected as the optimum approach.



       With respect to expander valving, a detailed analysis with


computer modeling of the expander and boiler performance for all


operating conditions has been carried out, comparing a system with


constant intake ratio expander valving and throttle valve control and


a system with variable intake ratio expander valving.   The results


are summarized in the performance maps presented in Figures 24


and 25 for constant IR and variable IR (IR    = 0. 29)  systems,
                                        max

respectively.  Comparing these two performance maps for the same


boiler and expander sizes, the following conclusions can be made:



       a.   The system with variable IR valving has a  peak efficiency


           of 18. 5% versus  15. 0%  for the system with constant IR


           valving.   This 20-25% improvement in efficiency (or mpg)


           occurs over a large power-speed region including the region

-------
                                                                                          Ln
0
 0
200    400     600   800    1000    1200    1400   1600   I8OO   2OOO  2200
                             Engine RPM
             Figure 24.  Performance Map with 184 CID Expander and
                         Maximum Intake Ratio of 0. 29.

-------
                                       VI-55
o>
*
o
I
no


100




80




60




40




20




 0
                     Full Power, Variable IP
                         of O 8 Max
                                                                             14%
    0     200    400    600    800
                                      1000   1200   1400
                                      Engine RPM
1600     1800   2000   2200
                  Figure 25.   Performance Map with  184  CID Expander

                               and Constant Intake Ratio of 0. 137.

-------
                            VI-56
         of 20 - 40 hp and 600 - 1200 rpm,  where the system would


         operate most of the time.



    b.   The peak power for the equivalent-sized expander and boiler


         is much greater over most of the expander speed range for


         variable IR valving than for constant IR valving.  Calculations


         of 0 - 60 mph wide-open-throttle  times indicate a 50%  increase


         for constant IR valving  relative to variable IR valving.



For these two reasons,  there exists an extremely strong incentive for


incorporation of variable intake  ratio  valving in the expander; this


type of valving is used in the reference expander design presented


earlier.



    In Figure 25, the dotted line'to the left of the  plot represents the


peak system power if a maximum intake ratio of 0.8 (the maximum


practical) rather than 0.29 were used.   (IR)      = 0.29 was selected
                                          max

as optimum for  the folio-wing reasons;



    a.   As evident from Figure 25,  very little performance  is lost


         by decreasing (IR)     from 0. 8 to  0. 29.
                          max


    b.   The required feedpump displacement is decreased by a


         factor of 2. 7.



    c.   The low vehicle speed condenser rejection rate is reduced,


         permitting greater utilization of  ram air.



    d.   The wide-open-throttle system efficiency  at expander  speeds


         to 800 rpm is increased.



    With reference to Figure 25, the  part-load performance  of the


system is an extremely  important consideration.  Thus,  while  the


design point efficiency,  defined by the peak power requirements

-------
                             VI-57
for acceleration, is 13 7%,  the efficiency increases under part-load


conditions where an automobile normally operates   The increase


with the variable IR valving  occurs because part-load  operation is


obtained by reducing the IR below the design point value of 0,137,


providing a more efficient expansion in the expander,  reduction of the


condenser pressure under part-load operation also occurs, again


leading to a more efficient cycle.  The region of high efficiency


(> 17%) is broad, that is, a high efficiency is obtained over a broad


range of engine power and speed.  Thus,  while the peak thermal


efficiency of the Rankine-cycle system (18.5%) is  much less than


that of the I/C automotive engine (~30%), the average efficiencies


for typical consumer driving cycles are  much closer,   The Ford


Motor Company has calculated, using the performance map of Figure


 25, with (IR)     = 0. 8,  the fuel economies for different driving
             max

conditions of the Rankine-cycle system with single-speed slipping


clutch transmission installed in an intermediate size American car,


and compared them directly with the fuel economy calculated for the


same body with 302-2V engine and three-speed automatic transmission.


The results are summarized in Table 10 for both steady speed and


dynamic driving cycle operation.   For the customer average driving


cycle,  the mpg for the 302 CID I/C engine is 15.7,  versus  12,7 mpg


for the Rankine-cycle system.  For steady speed,  zero grade operation,


the Rankine-cycle engine has better fuel economy than the  I/C engine


up to 50 mph and poorer fuel economy above 50 mph.  While calculations


have not yet been made,  it is expected that use of a two-speed trans-


mission as planned will improve the city fuel economy significantly,


resulting  in an improved  customer average.  This change,  coupled


with a potential for increasing the  peak cycle temperature to 600 °F,

-------
                                                  TABLE 10

                                       VEHICLE ECONOMY  PROJECTIONS
                                       THERMO ELECTRON CORPORATION
                                           RANKINE-CYCLE ENGINE
VEHICLE:  Intermediate size American car,  Wheelbase -  116 in.
TIRES:     7.75 x 14, Rolling Radius -  1. 08  ft. ,  Rev/Mile -  778
(IR)     = 0.
   max

Ford Production Engines
(1969 Model)
302-2V 3-Speed Automatic
250- IV 3-Speed Automatic
Rankine- Cycle Engines
182-CID Clutch Drive
Single Speed
182-CID Torque Converter,
1. 88 Ratio Speed- Up Gear
182-CID Torque Converter
2. 77 Ratio Speed-Up Gear
Idling
Speed
rpm

500
600

300
300
300
Fuel Flow
Ibs/hr

3. 75
2. 68

2. 00
2. 00
2. 00
City
mpg

13. 3
12. 9

9.6
10. 1
10. 3
Fuel Economy Computer Program PB1213
Suburban
mpg

18.0
19.6

15. 8
15.0
15.2
Customer
Average
mpg

15. 7
16.3

12. 7
12. 5
12. 8
Steady Speed mph
30
mpg

27.3
27.0

33. 1
32.4
32. 3
40
mpg

23. 3
25. 4

25. 1
24. 6
24. 8
50
mpg

20.4
22.9

20.4
19. 7
19. 5
60
mpg

18.0
20. 1

16. 7
16. 0
15. 7
70
mpg

16 2
I 7. 6

13. 4
13.0
12. 9
30-70 mph
mpg

21. 0
22. 6

21.7
21.2
21.0
                                                                                                                    I
                                                                                                                   Ui
                                                                                                                   CXI

-------
                           VI-59
could ultimately result in better fuel economy than the 1969 302-2V I/C


engine.  It should also be noted that the fuel economy of the I/C engines


decreases year by year as additional pollution controls are required.



    The acceleration performance of the vehicle with a given power-


plant is strongly dependent on the transmission used.



    Performance characteristics  of the Rankine-cycle system  in an


intermediate size American car are presented in .Figures 26 through


29 and Tables 11  and 12 for a torque converter transmission and


single-speed clutch transmission with IR     = 0. 8 and for single
                                       max

and two-speed clutch transmissions with IR    = 0. 29.   These cal-
                                          max

culations were prepared by the Ford Motor Company and the Dana


Corporation, using existing computer programs, for the  Rankine -


cycle system with different types of transmissions and for the Ford


production 302-2V engine with three speed transmission.  The basic


input to these calculations was the performance maps presented in


Figures 24 and 25.



    The most important conclusions from these calculations are:



    a.  The Rankine-cycle system with 184 CID expander (thiophene


        working  fluid) should be capable of providing 0-60  mph


        acceleration times of less than 15.0 seconds, taken as the


        criterion for  acceptable performance.  The  acceleration


        performance  is  fairly dependent on the type  of transmission


        used.  A maximum level  grade vehicle speed of  90  100 mph


        should be attainable,  irrespective of type of transmission


        used.

-------
                                VI-60
   1800
   1600
   1400
   1200
en
.0
LJ

0)
o
D
   1000
   800
   600
   400
   200
     0
TWO SPEED CLUTCH

   2.5/1  Ratio to ISOOEngineRPM Downshift
     I/ I  After Downshift
   2.0/1 Ratio to 1800 Engine RPM Downshift

     I/I After Downshift
        Single Speed Clutch
                1
       0      20




        Figure 26.
   40     60      80      100

         Vehicle Speed, MPH
120
 Comparison of Tractive Effort for Single and Two

 Speed Clutch Transmission.

-------
                            VI-61
   2200
   2000
    1800
    1600
    1400
UJ
_>
O
O
    1200
    1000
    800
    600
    400
    200
                   TWO SPEED CLUTCH
                   2. 5/1  Ratio to 1800 Engine rpm
                   downshift, 1/1 after downshift.
                                      ]    I
                ,. Intermediate size American car,  302-2V
                  Engine,  Three Speed Transmission
          Two Speed
          Clutch, 2. 5/1
          Ratio to 1800 Engine
       —rpm downshift, 1/1
          after downshift.
                                            3420 Eng. RPM
                                      I
               20     4O      60     80
                            Vehicle Speed, MPH
                                            100
120
140
Figure 27-  Comparisons of Tractive Effort for an Intermediate Size
           American car,  powered by 302-2V Engine and by Rankine
           Cycle System with 134 CID Expander,  IR    = 0.29
                                                  max       '
           with Dana Two Speed Clutch Transmission.

-------
                              VT-6?.
CO
,0
H
u
nJ
*H
H
    2200
    2000
    1800
    1600
    1400
    1200
    1000
     800
     600
    400
    200
                  Intermediate Size American Car, 302-2V
                  Engine,  Three Speed Transmission
.0/1 Ratio to 1800 Engine Rpm
 1/1 After Downshift
                20      40      60      80      100

                         Vehicle Speed, Mph
                         120
140
  Figure 28.  Comparison of Tractive Effort for an Intermediate Size
              American Car powered by 302-2V Engine and by Rankine
              Cycle system with 184 CID Expander,  IR    = 0.29, wit.i
              Dana two-speed clutch transmission.

-------
                              VI-63
  2200
  2000
   1800
   1600
v>
O
5t
UJ
0)
o
O
   1400
 1200
 1000
   800
   600
   400
   200
                  -Intermediate Size American Car, 302-2V—
                  Engine,  Three Soeed Transmission
^Maximum Torque Capacityof System
 Rankine Cycle, 184 CID, IRm(^0.8
 Single Speed Clutch Transmission
       Rankine Cycle, l84CID,IRma=0.8
       Toraue Converter Transmission
            D Converter)
             Low Speed Torque
             Used in Performance
             Calculations
                                          3420 Eng.RPM


                                        2000 Eng. RPM
                                  I
                                    I
     0
Figure 29.
              20     40     60      80     100
                            Vehicle Speed.MPH
                                  120
140
              Comparison of Tractive Effort for  an Intermeliate Size
              American Car Powered by 302-2V  Engine and by Rankine-
              Cycle System  with 184 CID Expander,  IRmax = 0. 8,  with
              Single Speed Clutch Transmission or with Torque Converter
              Transmission.

-------
                                                                  TABLE 11
                                                    VEHICLE PERFORMANCE PROJECTIONS

                                                      THERMO ELECTRON CORPORATION

                                                           RANKINE CYCLE ENGINE
VEHICLE: Intermediate Size American Car, Wheelbase - 116 in.

TIRES:     7.75 x 14,  Rolling Radius  -  1.08 ft. ,  Rev/Mile -  778
(IR)    =0.
   max

Ford Production Engines
302-2V 3-Speed Automatic
11-1/4 Dia. Converter
250- IV 3 -Speed Automatic
11-1/4 Dia. Converter
Rankine Cycle Engines
182-CID Clutch Drive, Single Speed
200- CID Clutch Drive, Single Speed
220 -CID Clutch Drive, Single Speed
240- CID Clutch Drive, Single Speed
182-CID 1. 88 Ratio Speed-Up Gear
12-5/16 Dia. Converter
182-CID 2. 77 Ratio Speed-Up Gear
12-5/16 Dia. Converter
Transmission
Gear Ratio
2.46, 1.46, 1.00
2.46, 1.46, 1.00

None
None
None
None
None
None
Axle
Ratio
2. 79
2.79
1.64
1.64
1.64
1.64
3.06
4. 50
N/V
36. 2
36.2
21. 3
21. 3
21. 3
21. 3
21. 3
21. 3-
Car
Weight
Ibs.
3539
3522
3539
3539
3539
3539
3539
3539
Performance - Computer Program PB1 1 1 1
0 -4 Sec.
ft.
89. 9
78. 2
70. 0
77. 2
86.0
94.4
74. 2
79. 8
0 - 10 Sec.
ft.
469.2
405. 7
407. 9
442.0
480.4
516. 7
412. 2
414. 6
0-60 mph
sec.
11. 9
15.6
14.2
12. 7
11.4
10. 3
14.4
14. 3
0 - 1/4 Mile
sec.
18. 8
20. 6
20. 3
19. 5
18. 7
18. 0
20. 1
20. 1
Passing at
50 mph
sec.
9.72
11. 59
10.30
9. 74
9.22
8. 79
10. 97
10. 85
                                                                                                                                                      <
                                                                                                                                                      H
                                                                                                                                                      I

                                                                                                                                                      -P-

-------
                            VI-65





                           TABLE 12


            VEHICLE PERFORMANCE PROJECTIONS,
IR    = 0.29
  max
54 ft # Subtracted from Full Torque Curve at All Speeds for Accessories
Transmission
Single Speed
Two Speed
Two Speed
Two Speed
Two Speed
Two Speed
Gear
Ratio
1/1
2.5/1
1/1
2.5/1
1/1
z.oo/:
1/1
1. 75/1
1/1
1.50/1
1/1
Engine Downshift
Speed
-
1800
1800
1800
1800
1800
0-60 mph
Acceleration Time
Seconds
19.6
14.4
14.5
14.6
14.9
15.5
Gradability
17.0%
48. 9%
42.5%
37.2%
31.9%
26. 7%

-------
                            VI-66
    b.   The Rankine-cycle system with two-speed clutch transmission

         and IR    = 0. 29 provides a close approximation to the tractive
               max
         effort delivered by the 302-2V internal combustion engine with

         three-speed automatic transmission.

    c.   The Rankine-cycle system with two-speed clutch transmission

         and IR    = 0. 29 provides a gradability of 49% with 54 ft-lbs
               max
         of torque subtracted for driving accessories.

3. 2  System Weight and Estimated Cost

    In Table 13,  the weight of the Rankine-cycle system reference

design based on thiophene (957 Ibs  total) is compared with thg.t of

the 302 CID engine with three-speed transmission (806 Ibs total).  Sub-

stantial weight reductions may be expected in the Rankine-cycle system

with development.  For example,  use of an all-aluminum condenser

would reduce the weight of this  component by approximately 50 Ibs.

Use of Fluorinal-85 will also permit reducing the expander disp'lace-
                 3         3
ment from 184 in to  107 in , with a resulting reduction in the ex-

pander weight.

    Using the conceptual design drawings,  cost estimators from the
Ford Motor Company have prepared a preliminary large-volume

production manufacturing cost estimate for the system.  The  cost

was quoted relative to the 1970  302-2V with the result:

          Cost,  Rankine  Cycle System _     +0.31
               Cost,  I/C System             -0.15

The system manufacturing cost estimate will be updated as the devel-

opment continues.

-------
                         VI-67
                      TABLE  13

       TABULATION OF  TOTAL SYSTEM WEIGHT
AND COMPARISON WITH  302-2V INTERNAL COMBUSTION
         SYSTEM WITH 3-SPEED TRANSMISSION


Expander Assembly
Feedpump
Expander Subsystem
Transmission
Burner- Boiler
Regenerator
Condenser
Radiator
Controls, Fans, Accessory
Drives and other Miscellan-
eous Components
Total
Rankine
Reference
Design
220
45
265 Ibs
135 Ibs
273 Ibs
54 Ibs
115 Ibs


75 Ibs
957 Ibs
302 in3 V-8
with 3 -speed
Automatic

479 Ibs
159 Ibs



54 Ibs

114 Ibs
806 Ibs

-------
                            VI-68
3. 3  System Packaging

     For the conceptual design,  the system, was packaged in the engine

compartment of an intermediate size American car; a full-size,  com-

plete mockup of the system was constructed in the engine compartment

of this car with excellent results.   In Figure 30, a photograph of the

mockup is presented and in Figure 31,  sketches illustrating the expander

transmission, burner-boiler, regenerator,  and condenser  locations

in the system are illustrated.  The only change required in the engine

compartment in packaging the system was modification of the frame

and fender panels at the very front of the car to facilitate placement

of the condenser.

3. 4  Emission Projections for Rankine-Cycle Automotive Propulsion
     System

     Using the burner emission levels obtained from the burner devel-

oped at Thermo Electron Corporation for a 5 hp Rankine-cycle cur-

rently on test,  projections have been made of the emission level  of

unburned hydrocarbons, CO, and NO from a Rankine-cycle automotive

propulsion system using 10 mpg fuel economy; the results  are pre-

sented in Table  14, and compared with projected Federal standards

for 1975 and  1980;* presented also are results for an uncontrolled

I/C engine and the IIEC targets  for emission control for the I/C
engine.   The Rankine-cycle system emission levels are lower than

the projected 1980 standards by a factor of 5 for unburned  hydro-

carbons, by a factor of 13 for carbon monoxide emission,   and by
a factor of 1.6 for NO  emission.
 The projected Federal standards in Table 14  are those existing
 prior to the passage of the Clean Air Legislation in the U. S.

-------
                                                                                                  <
                                                                                                  I—I
                                                                                                   I
Figure 30.  Photograph of Rankine-Cycle System Mockup In Intermediate

            Size American Car Engine Compartment.

-------
                              VI-70
                            r-.-l.--L

\
\

,/
'

\ VJ 1'
x i-4^--4
i
''^- -
n ^

•













n
	 i



                                           \.   I
                                       n    l   L
                                            V	J	
                                 FtCOPUMP
Figure 31.   Position of Major Components for 100 hp Rankine-Cycle

             Powerplant in an Intermediate Size American Car Engine

             Comoartment.

-------
                                   TABLE 14


EMISSION COMPARISON OF RANKINE CYCLE WITH INTERNAL COMBUSTION ENGINE
System
Emission HC
t
| Level CO
I gms/mi NO
Projected
1975/1976
Standards
0. 46
4. 7
0. 4
Uncontrolled
I. C.
Engine
5. 5
24
5 5
IIEC Targets
for
I, C. Engine
0. 82
* 7. 1
0.68
TECO Projections
for Rankine- Cycle
Propulsion System
0. 05
0, 35
0 25
                                                                                      H
                                                                                      I

-------
                           Vl-72
3. 5  Major Conclusions




    a.   The Rankine-cycle system offers a very strong potential for



         minimum emissions of particulates, unburned hydrocarbons,



         nitric oxide, and carbon monoxide.




    b.   The use of Fluorinol-85 as a working fluid, with a moderate



         maximum cycle  temperature of 550 °F,  permits a significant



         cost reduction relative to the equivalent steam system. This



         reduction may permit the Rankine-cycle system to be com-



         petitive costwise with the equivalent I/C system,  particularly



         since the stricter emission level standards will require sig-



         nificant cost increases in the I/C system.




    c.   An organic-based working fluid, Rankine-cycle system



         approximately equivalent  in performance to a 302 CID



         I/C engine with 3-speed automatic transmission can be



         packaged in current automotive engine compartments



         with only minor  modifications required in  the sheet



         metal and frame.





    d.   The system development and design is sufficiently advanced



         to insure a high  probability of meeting the design goals and



         performance estimates presented in this report.

-------
                        VII-1
                     Chapter VII
STIRLING ENGINE ACTIVITIES AT UNITED STIRLING (SWEDEN)
                          by
                   Lars G. Ortegren
                 Vehicle Applications
       K. B. United Stirling (Sweden) AB & CO.
                     Malmo, Sweden

-------
                                 VII-2



Introduction


The development of Stirling engines in Sweden is taken care  of  by


KB United Stirling (Sweden) AB and Company.  The main object  of


the company is the development and adaptation for production  of


Stirling engines.  At present, about 70 engineers are engaged in the


Swedish Stirling engine project.



General information on the Stirling engine principles, its properties


and potentials is given in a brochure, copies of which are available


at the conference.  The most important characteristics of this  engine


are its very clean exhausts and silent operation.



A major part of United Stirling's resources is being used for the


development of a 200 hp four cylinder in-line engine, suitable  for a


number of applications, but especially for operation in areas where the


environmental properties are of great importance.  Generally speaking,


this engine will have outer dimensions allowing the exchange of most


existing diesel engines in city buses, mining vehicles, boats and


stationary installations.



The 4-615 Engine Program


The 200 hp engine to be produced by United Stirling is named 4-615 (four

                 O
cylinders, 615 cmj of swept volume per cylinder).  A development program


is in progress, starting with prototype 4-615A engines running  in labora-


tories during 1971.  The manufacturing and testing of a number  of


gradually more advanced prototype engines is scheduled for the  following

-------
                                  VII-3
years.  Series deliveries are planned  to  commence  in 1976,  preceded




by earlier pilot production.







A very important background for  this development  is  an extensive  use




of value analysis technique in order to approach  a production version




with minimum outer dimensions and manufacturing cost.







The consecutive versions will be gradually  more simplified  and adapted




to series production methods.  Further increased  realiability and ease




of maintenance are other major objects of work.  The engine principle,




with heat generation and working cycle separated from each  other,




facilitates independent parallel development  of components,  e.g.,




burners, preheaters, control components,  etc.







The 4-615 engine is a four cylinder, displacer type  engine,  with  rhombic




drive mechanism.  Nominal power  output is 147 kW  (200  hp) at 2,400  rpm




at the main output shaft.  When  choosing  design parameters,  efficiency has




been given a high priority, which,  for the  rather  conservative values of




heater tube temperature and mean cycle pressure, has resulted in  a  rather




slow running engine with a relatively  large swept  volume.   Thus the crank-




shafts turn at 1,550 rpm at nominal power output,  and  a  speed increaser is




used to adapt the output shaft speed to existing automotive  transmission.







One important result of value analysis work applied  to the  4-615  engine is




a considerably reduced size, as compared  to the present  state of  art.




The estimated dry weight of this engine is  900 -1,000  kg  (2,000 - 2,200




pounds^ .

-------
                                  VII-4
A brief preliminary specification of this engine is as follows:

   Rated gross power               200 hp at 2,400 rpm
   Maximum gross torque            80 kpm (580 Ibft) at 700 rpm
   Maximum brake efficiency in
     vehicle application           37%
   Installation dimensions
     Length                        1,200 mm (48 inches)
     Width                         500 mm (22 inches)
     Height                        1,100 mm (44 inches)
   Fuels                           Diesel oil, kerosene
   Optional fuels                  LPG, LNG
Traction Application Projects

The installation and testing of engines in vehicles has always been

considered a very important part of engine development.  This is

particularly true in the case of Stirling engine development, because

of the inherent relations between engine performance (power, efficiency)

and properties of the external cooling system.


City bus propulsion is considered one of the most attractive fields of

application for United Stirling's engine type 4-615.  To make possible a

near-future commercial realization of Stirling engine powered buses, a two-

phase vehicle development project has been started.


For basic test and evaluation, a Philips type 4-235 engine will, as Phase I,

be installed in a test vehicle during 1971.  In this phase, theoretical

calculations and considerations concerning vehicle performance will be

verified.  A practically useful theory for matching engines, transmissions

and cooling systems will be established.  Performance is scheduled to be

demonstrated during 1972.

-------
                                 VII-5
Phase II, which is planned to be completed during 1973, consists of the




development of a city bus, using a 4-615 Stirling engine prototype.  In




this phase, bus and propulsion systems will be mutually matched in order




to achieve a near-optimum solution with regard to driving characteristics,




space utilization, system noise level, passenger compartment heating,




etc.  The development will be performed in close cooperation with a bus




manufacturer.  Experience feedback to production engine design is a




very important part of this phase.







As a result of this two-phase development project, production of city




buses with exceptionally good environmental properties will be made




possible.  Besides the very low exhaust emission level described earlier




in this paper, a remarkably low noise level will be the result.




Referring to the standardized measuring distance of 7.5 metres (25 feet)




from vehicle center line, a noise level below 70 dB(A) can be expected




under most urban operation conditions.

-------
                                    VIII-1
                                 Chapter VIII
NITROGEN  OXIDE  FORMATION IN THE CO~MBUSTION CHAMBER OF  THE  INTERNAL COMBUSTION

      ENGINE AND ITS SUPPRESSION BY MEASURES FROM COMBUSTION TECHNOLOGY
                                      by
                               W. E. Earnhardt
                            Research Department 2
                              Volkswagenwerk AG
                                Wolfsburg, FRG
                     Presented at Eindhoven Conference by
                                K. H. Newmann
                          Research Development Group
                              Volkswagenwerk AG
                                Wolfsburg, FRG
                        Translated for EPA by SCITRAN
                       (Scientific Translation Service)
                        Santa Barbara, California, USA

-------
                              VIII-2
                             Summary

Production and decomposition of nitric oxide in the internal
combustion engine was investigated,  with consideration of the
kinetics of the elementary chemical reactions.  A simple
mathematical model of nitric oxide formation in the engine
was developed.  It consists essentially of an appropriate
reaction mechanism for the combustion process in the engine
and a thermodynamic analysis of the combustion process.
Conclusions obtained from this model concept were then tested
on a one-cylinder engine using such measures of combustion
technology as stratified charge operation to influence the
combustion process so that nitric oxide formation is
inhibited because of too low temperature,  in spite of the
presence of oxygen.  In this way the combustion is controlled
so that it is still possible to oxidize carbon monoxide and
hydrocarbons almost completely.

-------
                              VIII-3
                             Contents
1.    Introduction

2.    Model of Nitric Oxide Formation in the Engine
2.1  Reaction Mechanisms for Nitric Oxide Formation
2.2  Model of Flame Propagation and Thermodynamic Analysis
     of the Combustion Process
2.3  Calculation of the Kinetics of Nitric Oxide Formation

3.    Combustion Technology Measures for Suppression of Nitric
     Oxide

4.    Summary

5.    Bibliography

-------
                             VIII-4
1.   Introduction

     Investigation of non-equilibrium processes of the
processes occurring in the combustion chamber of the internal
combustion engine has for some time been of major scientific
interest.  In view of the drastic measures against air pollution
by automobile exhaust which have been required by legislators,
it is pressingly necessary to find ways to decrease the
injurious materials carbon monoxide (CO),  nitric oxide (NO) and
unburned hydrocarbons (CH) contained in the exhaust gas, in
order to attain more complete combustion and thus a cleaner
exhaust gas.

     In the following report,  the thermal formation of nitric
oxide within an internal combustion engine is studied theoretically
because of its particular importance for environmental protection.
Thus a model of automobile nitric oxide formation was developed,
and its transferability to a real engine was studied.  On the
basis of the conclusions obtained from that,  it was attempted
to control the combustion process through measures from
combustion technology so that the NO concentration would remain
as small as possible,  and also that the oxidation processes
necessary for a reduction of the HC and CO would be  completed.

     We shall not consider here the potential for NO reduction
by exhaust gas recirculation or by use of so-called  "diluents"
such as C02? t^O, He, Ar, No, etc;  by use of reduction catalysts;
by choice of unconventional fuels, or by addition of certain
additives to promote combustion.  In this respect,   see
G. H. Meguerian [11].

-------
                             VIII-5

2.   Model of nitric oxide formation in the engine

2.1  Reaction mechanisms for nitric oxide formation

     Thermal nitric oxide formation can be represented by the
gross reaction

                         N  + 0  z=± 2NO
Such a reaction consists of a number of elementary reactions
which,  considered microscopically,  describe the collisions
between molecules,  radicals,  or atoms,  through which new
elements are formed.  The elementary reactions make up the
chemical transformations actually going on,  while the gross
reaction reflects only the total result of these elementary
reactions,  without considering the intermediate products.

     On the basis of a literature study,  the thermal nitric
oxide formation in combustion processes can be characterized,
for example,  by the reaction mechanisms shown in Figure  1.
Reaction mechanism  I  was stated recently by H. K. Newhall and
S. N. Shahed [l],  and mechanism II by L. S. Caretto and  co-
workers [2].  Equations (2) to (5) of the first mechanism in
Figure 1,  to be sure,  are contained in the reaction scheme
suggested by K. Vetter [3] as early as 1945,  while Equations
(3) to (8) of the second mechanism had been stated by H.  K.
Newhall and E. S. Starkman [4j as well as by G. A. Lavoie and
co-workers [5].  Since both calculation and experiment have
shown [6, 7]  that the nitrogen monoxide, NO, is the only
nitrogen oxide of significance in the engine,  those elementary
reactions in these mechanisms which contain nitrogen dioxide,
NCL, or nitrous oxide, NoO, as participants can be neglected,
without causing any large error.

     W. Bernhardt [8] showed recently that with respect to
nitric oxide formation in the engine,  the reactions

-------
                              VIII-6

                      N + O + M^NO    +M

                and   OH  +  N  F^  NO    +  H

are of secondary importance.   See also [9].  Thus two  equations
remain  in each of the reaction mechanisms shown in Figure  1.
These are well known under the name of the Zeldovich mechanisms  [10]

     If one compares the velocity constants in the Zeldovich
mechanism,  using the kinetic reaction data of K. L. Wary and
I. D. Teare  [12],   it appears that the reaction

                          N2  + 0 ?± NO + W

is clearly the rate-determining reaction for thermal nitric
oxide formation because of its high activation energy.  See
also Figure 4.

-------
                     Zeldovich mechanisms
N
^ 0 + M
02+N
N2+0
I02 + 0
^
-
-
NO
NO
NO
NO
+ M
.0
+ N
•>• 0?
fy
\
N2 +
02 +
OH +
N2 +
N7 +
0
N
N
02
OH
^
^
N
N
N
*
N-
0 +
0 *
0 +
? 0 +
>0 +
N
0
H
0
  N20+0
N2+ 0+M
NO  +0+M
NO  * NO
N20  +M
  N2 0  +0
 N+0  + M
A/2 + 0  v- M
NO  +NO
NO  +M
                                                                         Research  2
                                                                         BE-70-01
                                                                                       M
                                                                                       I
The triple collision partner M can be any particle from the reaction
volume;  it acts  only to remove energy.
   Figure 1.  Reaction mechanisms for the formation of nitric oxide in  engines.

-------
                             VIII-8

2.2  Model of flame propagation and thermodynamic analysis of
     the combustion process

     In the model of flame propagation used in this work it is
assumed initially that during the combustion in the combustion
chamber there are areas with quite different temperatures.  The
region already passed through by the flame front contains a burned
mixture of high temperature.   The region not yet reached by the
flame contains fresh gas of relatively low temperature.  It is
assumed,  however,  that in both regions it is possible to have
intensive mixing and thus a more rapid decay of the temperature
differences,  and that the energy exchange from the unburned
region to the burned region occurs so rapidly that the fresh
gas temperature suddenly rises to the higher temperature of
the exhaust gas.

     So far it has not been possible to establish generally
valid analytical relations for the flame advance within an
internal combustion engine.  It is assumed,  however,  that
the model concept used here sufficiently describes the processes,
especially of the actual pressure curve while the combustion
process goes on.  Along with the pressure,  which is taken as
constant everywhere in the combustion chamber,  the volume  V
is given as a function of time,  from the well-known kinematic
equation for piston motion.  According to this model concept,
the total volume is made up of the volume of the fresh gas and
the volume of the burned gas.

The temperature in the unburned region is calculated assuming
an isentropic change of state for an ideal gas.  This is a crude
approximation,  because the prerequisite for application of
equilibrium thermodynamics is not given.  Because of the lesser
effect of temperature in the unburned gas on the following
calculations,  we can proceed in that way here.

-------
                             VIII-9

     The  temperature  in the burned region can be determined
by the  time  course  of conversion of the chemical energy into the
internal  energy of  the combustion gases.  For this purpose, of
course, a thermodynamic analysis of the processes in the engine
must  previously have  been performed.  In this analysis it must
be considered that  the working fluid changes during the process
because of the chemical reaction.  Figure 2 shows the result of
such  an analysis for  a VW Type 3 engine (3010 rpm, air ratio
 A =   0.93,   full load operation).

     Extensive descriptions of the procedure for determining
the energy transformations in an internal combustion engine
can be found in W.  Hinze [13]and W. Bernhardt [14].  For that
reason we will not  discuss it extensively here.  Figure 3 shows
the time  course of  the temperature in the unburned and burned
regions of the combustion chamber for the combustion process
being studied.      The figure also shows the mean temperature
in the working gas,  as would be given from an adiabatic mixing
process.   In the temperature determination it was assumed that
the state behavior  of the reacting substances is given by the
equation  of  state for ideal gases.

-------
  1TJ
  H-
  OP
  c
  H
  0)
o> m
3 3
 X
  n
p t^
  P
H-, 3
C Oq
3 CD
O
r+ H-
H. 3
O
3 S3
3" CD
CD i-i
  3
>-t P
CD I—1
P
n n
rt O
H' 9
O CT
3 C
3 o
CD 3
CD
C
CD
7.5

 J

1.25



1.0



0.75



05



0.25



0
                                                              internal  energy  of the combustion

                                                                                           gases
                                                                                          CD
                                                                                          3
                                                                                          M-H
             0
                                                                            reaction  time
         research 2
                                                                                                         BE-70-03

-------
 OQ
  c
CD

OJ
H H
CD H-
OQ 3
h" CD
O
3 n
C/1 O
  C
O i-j
i-h W
  CD
rt
^ O
CD Hi

n rt
o tr
3 CD

C rt
C/l CD
rt 3

O CD
O rt-
tr1 C
P H
3 CD
CT
n> H'
  H'
  I-H
  H-i
  CD
  H
  CD
  3
  H-
         3000-

          K

         2600-
      CD
      tn
      3


      2  2200-
      

                                                                                           region
                                                                                                                     1—I
                                                                                                                     H
                                                                                                                     I
                       1,6
3,2
4,8     .6,4      8,0      9,6
                                                                           11,2
                                                                                  12,8     74,4      76,0  ms  17,6
                                                                     reaction  time
         research   2
                                                                                                    BE-70 -02

-------
                             VIII-12

2.3  Calculation of the kinetics of nitric oxide formation

     The model of flame propagation which has been described was
used to calculate the kinetics of nitric oxide formation.  The
temperatures and volumes or densities calculated by means of this
model are particularly needed.  For the Zeldovich mechanism,
the time change of the NO concentration can be written in the
form presented in Figure 4 for isochoric processes, in which the
density remains constant.

     Problems arise in the application of this differential
equation,  however,  because of the determination of the time
course of the concentrations of the participants in the reaction,
especially of the 0 atoms.  Starting from the known initial
concentrations of Oo and N~ immediately before the beginning
of the combustion process,  the concentrations of Oo, 0  ^ and N
are calculated approximately for the temperatures in the different
regions of the combustion chamber by use of the equilibrium
constant,  considering the dissociation and the fuel conversion
which has already occurred.  With the known concentrations
 "0^ ' ^2 '  0 and "N    »  known density, and known temperature,
the differential equation in Figure 4 can be solved by means of
the method of Runge-Kutta.  Thus the NO formation can be followed
analytically in the unburned as well as in the burned region of
the combustion chamber.

     As an example,  Figure 5 shows the curve for the NO concen-
trations in the two regions existing, according to the model
concept,  with a VW Type 3 engine at  1000 rpm,  ignition advanced
to 12° before TDC,  and an air ratio of    % =  0.93.  The
nitric  oxide  formation  is  shown  graphically,  starting  at the time of
ignition ( t = 0 ), during the combustion and expansion  processes.
                                                       o
The nitric oxide concentration is shown in moles per cm.

     As was to be expected from studies carried out previously
by means of equilibrium theory,  the nitric oxide formation  in
the fresh gas is negligible in comparison to that in the burned

-------
Oq
 c
 CD

 4^.
 n
 (a
 I—"
 n
 H'
 O
 O
 l-h
 O
 in
 O

 l-h
 O
 l-i
 3
 PJ
 rt
 H'
 O
OO
1
06
O
0
O'
O'
0'
00
O'
CD
O



























/
/



/
7



X
/





^—





























7.6 3.2 4.8 6,4 8.0 9.6 77.2 72.8 74.4 1t
reaction time ^











$.0 77.6
ms
^uP BE-70-OU
research 2

-------
                      time change of the NO concentration for the Zeldovich mechanism:
OQ
C
H
CD
H
H'
3
CD

O
O
e
l-t
O
Hi
2
O
n
o
o
CD
fo
H-
H'
O
                 d(5
                      NO _
in which O is the density,Ojis the number of moles of component  i per

mass,  t  is the time, and  k.  is the velocity constant of reaction j:








1
2
3
A
reaction

N2 + 0 — NO + N
NO + N — N2 + 0
02 + N — NO * 0
NO + 0 — 02 * N
velocity constant CfTT^

' mol • sec
7- 10 13 exp C-75000/ R T]
1,55- ]013
15,3 • 109 T exp (-7080 /
3,2 • 109 T exp [-39100/

R-T]
R-T]






T in K , R in ccU/(mol- Kj
research 2

BE-71-01

-------
                             VIII-15

region because of the lower temperatures in the fresh gas
region (  < 1000 °K).   The nitric oxide concentration in the fresh
gas is so slight that it is barely detectable in Figure 5.  It
is a characteristic of the thermal nitric oxide formation in an
Otto engine that it becomes perceptible only quite late,
after about a third of the total combustion period has passed.
[This applies only for the average nitric oxide concentration,
referred to the total combustion volume,  which we are considering
here.]  The maximum of the (average) NO concentration occurs only
when two thirds of the combustion period has passed.   After the
maximum has been reached,  there is hardly any detectable change
in the nitric oxide concentration during the expansion process.
This result is in distinct agreement with the experimental studies
of H. K.  Newhall and E. S. Starkman [4, 16],  who recorded the
monochromatic emission of NO by means of a monochromator directly
during the expansion,  and with the reaction kinetic calculations
of H. K.  Newhall [17].  On comparison of the engine nitric oxide
concentrations in the exhaust gas, as predicted by this, model of
engine nitric oxide formation,  with the values measured for an
actual engine,  it can be seen that this model is valid only
for nearly stoichiometric mexture ratios and for the region  X > 1.
The reason is that in the rich region ( X < 1 )  the effect of
the fuel is not considered.  This has also not been done in
other models just recently proposed [18, 19].

     According to G.  H. Meguerian [11],  with a fuel excess in
the Zeldovich mechanism, the oxygen and nitrogen radicals react
with the unburned or partially burned fuel molecules.  This
leads to lower nitric oxide concentrations,  and the NO values for
15% fuel excess are about 30% below the values obtained for the
stoichiometric mixture.  See also [19J.  In the thin region
(  X  >  1 )  the predicted NO values agree well with those measured
in the engine exhaust,  because the flame velocity decreases with
increasing air ratio,  and this effect is considered in the
flame propagation model.

-------
                             VIII-16

3.   Combustion Technology Measures for Suppression of Nitric Oxide

     In the calculation of the kinetics of nitric oxide formation
it appears that temperatures above 2600 °K are necessary for
formation of NO in the burned region of the mixture.  See Figures
3 and 5.  By contrast,  the CH and CO oxidation processes go on
even at considerably lower temperatures.  If it were possible to
control the combustion process in the engine so that the temperature
of the combustion gas would remain below about 2600 °K,  the
nitric oxide formation would be inhibited.  This would certainly
lead to lower NO concentrations in the exhaust gas.  Another
important conclusion from the model of engine nitric oxide
formation is that NO is formed only quite late in large concen-
trations in the burned mixture.  Since the flame front of the
burned region has already passed,  we speak here of "post-flame
reactions".

     On the basis of knowledge obtained from the model concept
of NO formation,  an engine combustion process was tested with
a CLR  one-cylinder engine  in research at the Volkswagenwerke AG.  ^
The process appeared to be able to decrease simultaneously the
CH and CO concentrations as well as the NO concentration.  This
amounts essentially to use of a pre-combustion chamber,  in which
a nearly stoichiometric or even a rich mixture is first ignited
and burned.  The design of the combustion chamber is such that,
after passage through the pre-combustion chamber, the flame front
advances into the main combustion chamber.  There it meets the
load-controlled main mixture at relatively low temperatures.
(The main combustion chamber can even contain completely pure air.)
In this way the hot combustion gases are rapidly lowered in
temperature,  so that in spite of excess oxygen only small amounts
of NO can be formed.  The temperature which results is still high
enough,  though,  to allow rapid oxidation of carbon monoxide and
hydrocarbons.  A pre-combustion chamber arrangement which is
similar in principle was suggested recently by Newhall [20].

       These  investigations  were  performed by  Graduate  Engineer
  I.  Geiger.

-------
                             VIII-17
 Figure  6.
              Combustion  chamber with pre-combustion chamber,
           y**\/s\jp*!*jf*vu^
           '
                        „ ___ , ____„_. ^
                     "      '  '   ''    '
                                        	
                                 .
Figure 7.
           Combustion in a combustion chamber  with a pre-
           c ombus t ion chamber .
Figure 8.   Combustion in a  combustion chamber with a pre-
           combustion chamber.

-------
                             VIII-18

     Figure 6 shows the arrangement of the pre -combust ion chamber
and the main combustion chamber with a VW Type 3 cylinder head.
In these studies,  the ratio of pre -combust ion chamber volume to
main combustion chamber volume was between 1:10 and 3:10.  With
this pre-combustion chamber arrangement,  very slight exhaust
gas emissions were measured with a CLR one-cylinder engine
at 2000 rpm under full load operation with optimum ignition
timing.  The emissions were as low as  40 ppm NO, 40 ppm CH
(measured as hexane) and 0.1% CO by volume at air ratios of
  X   =  2  to  X  = 3.

     In conclusion it should be mentioned that the combustion pro-
cesses in the combustion chamber could be filmed by use of a
high-speed camera (6,000 pictures /second) so as to  test
whether the combustion technology measures applied  satisfied
the information obtained from the model of engine NO formation.
Figure 7 shows 64 pictures from a single combustion process.
They were made at 2000 rpm,  ignition at 30° before TDC,  effective
                                2
mean pressure   P    = 2.5 kp/cm ,  an air ratio of  A = 2.5 in
the main combustion chamber and an over-all air ratio  ^n~G = 1.6,
                                                        6c;>
using the high-speed camera.  In this case we measured 170 ppm NO.
This is more than an order of magnitude less than in conventional
combustion chambers.

     Figure 8 shows an enlargement of one photograph from
Figure 7.
4.  Summary

     The nitric oxide concentration in the exhaust gas can be
predicted for stoichiometric and fuel-poor mixtures with the
simplified model of engine nitric oxide formation described here.
It is shown that, along with the velocity constants of the reaction

                         N + 0 i=i NO +  N

-------
                             VIII-19

the temperature  in  the  region passed by the flame front is the
principal quantity  affecting  the  prediction.

    Based on conclusions  from the  model,   and with basic
changes in the previous combustion  processes,   such as stratified
charge operation,   exhaust gas emission measurements on a CLR
one-cylinder engine with a modified VW Type 3  cylinder head
showed NO at 5 to 10% and  CH  at 10  to 25% of the levels observed
with conventional combustion  processes.  Pre-combustion chamber
arrangements,  combined with  stratified charge operation could
prove to be potential solutions for significant reduction of
the injurious materials in automobile exhaust  gas.  In spite of
the good results obtained  so  far,   it is doubtful whether this
method alone will be  suitable to attain the drastic reduction
of emissions required by American legislators  after 1975,
with retention of good  specific power and economy.

-------
                              VIII-20
                            REFERENCES


1.   Newhall,  H.  K.  and S.  M.  Shahed.   Kinetics of Nitric Oxide
      Formation in High Pressure Flames.   Paper presented at
      Thirteenth Symposium (International)  on Combustion, Salt Lake
      City,  Utah,  August 1970.

2.   Caretto,  L.  S., L. J.  Muzio, R.  F.  Sawyer and E.  S.  Starkman.
      The Role of Kinetics in Engine  Emission of Nitric  Oxide.
      Paper  presented at the  AICHE-Meeting,  Denver,  Colorado,
      August  1970.

3.   Vetter,  K.  Kinetics of Thermal  Decomposition and Formation of
      Nitric  Oxide.  Z. f. Elektrochemie,  Vol. 53,  1949, pp. 369-80.

4.   Newhall,  H.  K.  and E.  S.  Starkman.   Direct Spectroscopic Determina
      tion of Nitric Oxide in Reciprocating  Engine  Cylinders.   SAE-
      Paper  No.  670, 122,  1967.

5.   Lavoie,  G. A.,  J. B. Heywood and  J. C.  Keck.   Experimental and
      Theoretical Study of Nitric Oxide Formation in  Internal
      Combustion Engines.   Combustion Sci.  § Technol., Vol.  1, 1970,
      pp. 316-26 .

6.   Wimmer,  D. B.  and L. A. MacReynolds.   Nitrogen  Oxides and  Engine
      Combustion.   SAE Trans.,  Vol.  70, 1962.

7.   Campau,  R. M.  and J. C. Neerman.   Continuous  Mass Spectrometric
      Determinations of Nitric Oxide  in Automobile  Exhaust.   SAE-
      Trans., Paper 660 116,  Vol. 75, 1967.

8.   Bernhardt, W.   Studies of the Nonequilibrium Processes of  the
      Reactions  Occurring in the Combustion  Chamber of the Internal
      Combustion^Engine.  Lecture at  the  VDI Thermodynamics  Col-
      loquium, Wiirzburg, 5 October 1970.

9.   Campbell, I. M.  Chemical Mechanismus  Relevant  to the Production
      and Emission of Nitric Oxide and Carbon Monoxide from Com-
      bustion Engines.  Lecture 7 in:  A short Course on Fundamen-
      tals of Engine Exhaust Pollution, University of Leeds,
      September  1970.

10.  Zeldovich, Ya.  B.  The Oxidation of Nitrogen in Combustion
      Explosions,  Acta Physocochimica URSS,  Vol.  21,  1946, pp. 577-
      628.

11.  Meguerian, G.  H.  Nitrogen Oxide-Formation, Suppression, and
      Catalytic Reduction.  Paper PD 23.   Presented at the World Oil
      Congress in Moscow,  June 1971.

12.  Wary, K.  C,  and J. D.  Teare.  Shock-Tube Study of the Kinetics of
      Nitric Oxide at High Temperatures.   J. Phys.  Chem., Vol. 36,
      1962,  pp.  2582-96.

-------
                             VIII-21
13.  Hinze. W.  Procedures  for Thermodynamic Evaluation  o£ Test
      Stand Results  for  Internal  Combustion Engines  Studied.
      Dissertation,  Dresden Technical University,  1956.

14.  Bernhardt, W.  Thermodynamic  Evaluation of Test  Stand Results
      to Determine Laws  of Energy Conversion  in  the  Engine.  VW
      Report T 327,  11 June, 1969 (unpublished).

15.  Bernhardt, W.  Formation of Nitric Oxide  in  Internal Combustion
      Engines.  Brief Report from Research 2, 18 December 1969
       (unpublished) .

16.  — .  Control  of  Oxides of Nitrogen.   Fifth Status Report of VW
      of America  to  California Air Resources  Board,  Los Angeles,
      VW Report V/70,  (unpublished).

17.  Starkman, E.  S.  Formation of Exhaust Emission in the Combustion
      Chamber.  XIII.  Congress of FISITA, Bruxelles, Belgium,
      Paper No. 15. 3. D. , June 1970.

    Newhall, H. K.   Kinetics of Engine -Generated Nitrogen Oxides and
      Carbon Monoxide.   Twelfth Symposium  (International) on Combus-
      tion, The Combustion Institute, Pittsburgh,  Pennsylvania,
      1969, pp. 603-13.

19.  Heywood, J. B.,  S. M.  Mathews and B. Aven.   Predictions of
      Nitric Oxide Concentrations in a Spark-Ignition Engine
      Compared with  Exhaust Measurements.  SAE -Paper 71001, -Vol. 11,
      15 January  1971, Detroit, Mich.

20.  Muzio, L. J., E. S.  Starkman  and L. S. Caretto.  The Effect of
      Temperature Variations in the Engine Combustion Chamber on
      Formation and  Emission of Nitrogen Oxides.   SAE -Paper 710158,
      Detroit, Michigan, Vol. 11,  15 January  1971.

21.  Newhall, H. K. and I.  A. El-Messiri.  A Combustion  Chamber
      Concept for Control  of Engine Exhaust Air  Pollutants Emissions
      Combustion  and Flame, Vol.  14, 1970, pp. 155-58.
18

-------
                            IX-1
                        Chapter IX
A EUROPEAN CONTRIBUTION TO LOWER VEHICLE EXHAUST EMISSIONS
                            by
                      Diarmuid Downs
                    Managing Director,
                   Ricardo and Company
                     Shoreham-by-Sea,
                   Sx., United Kingdom

-------
                            IX-2
     I do not propose to discuss whether the present limits and




proposed future limits for road vehicle exhaust emissions are




sensible economically or well founded scientifically.    I will




only remark in passing that the NQx regulations will be the most




difficult to meet and will have the greatest effect on the economics




of operation of motor vehicles.   It behoves us to be sure,




therefore, before imposing them, that they are really essential




medically and environmentally, and that we have our priorities




right in relation to other atmospheric pollutants.




     In this Note, I have deliberately confined myself to two




areas of interest and concern to the automobile engineer




          1.   The small European automobile and its problems




in meeting future US legislative requirements.




          2.   The contribution which combustion chamber design




can make to reducing exhaust emissions from the diesel engine.

-------
                             IX-3
The European Automobile and the American Market.




     As Consulting Engineers engaged in design, development and




research on internal combustion engines, we have been very active




in assisting European car manufacturers to meet the U.S. Federal




and Californian State exhaust emission limits.   Our experience is




probably unique in the range of vehicle type we have studied and




the variety of approach we have used to meet individual requirements.




     To meet the U.S. Regulations up to and including those imposed




for the 1971 model year, we have used in the main what has come




to be known as the Cleaner Air Package (CAP) approach, involving




a tightening of production tolerances and detailed adjustments to




the carburettors and ignition settings over the load and speed




range, combined with the use of such devices as intake air heaters




and manifold air depression limiters.   The art is to reach the




required exhaust emission levels without unacceptable loss of




driveability.   This is harder to achieve with the small European




car than with the larger and generally more powerful American car.




For .this reason, a number of European car manufacturers have adopted




petrol injection in place of carburation, which, because of the




more precise metering and delivery of the fuel over the operating




range including the transients, enables the present exhaust emission




limits to be attained without sacrifice of driveability;  in fact,




in most cases, with driveability enhanced.   For the 1972 model




year, it seems probable that further refinement of the CAP approach,




particularly if associated with petrol injection, will enable the




lower limits associated with the change of test procedure, to be



attained.   In some cases, however, it may be necessary to use




manifold air oxidation, a device which has been used intermittently

-------
                           IX-A
since the early days of pollution control,  but which,  because of




the cost of the air pump and the problems posed by its installation




and drive, has generally been abandoned wherever possible.




     We have been working for some time with our eye on 1975 and




Table I summarizes the present position.  The first line gives the




limits which we thought we were going to have to meet prior to the



signing of the Nixon/Muskie bill last December.    The second line




gives our present target, necessarily somewhat vague in regard to




NOx.   Some typical results obtained with small European cars fitted




with catalytic afterburners are shown in the second half of the



Table.   It can be seen that, ignoring NOX for the moment,  the




pre-Muskie limits for HC and CO have almost been reached with




copper/chromium catalyst and could probably be still further



reduced with more development work.   With the platinum catalyst,




even the limits required by the Nixon/fauskie bill can be met in




regard to HC and CO.   When the attempt is made to reach the NO




limits as well, by the use of Exhaust Gas Recirculation (EGR)




combined with a platinum catalytic afterburner, the NO  is certainly
                                                      X



reduced, but the HC and CO are then too high.   These results were




obtained with an unleaded fuel and it should be emphasized that




the performance of these catalytic systems in  respect of endurance




has not yet been fully assessed.



     It would seem that it is going to be very difficult, if not




impossible, to reach the figures required for  1975 by the use of




EGR, without unacceptable loss of driveability.   For this reason,




we believe that a double catalytic system holds out the best hope




for a European car, with a first stage reducing the NOX and a




second stage oxidising the HC and  CO.   We have no results to




report on such a system as yet.

-------
                           IX-5
                          TABLE 1
               AUTOMOBILE EXHAUST EMISSIONS
               (1972 FEDERAL TEST PROCEDURE)
                                         HC        CO      NOX

1975 U.S.  FEDERAL LIMITS                 ~        ~

   1.  Pre Nixon/fouskie                  0.5      11.0      0.9

   2.  Post Nixon/kuskie                 0.^6      4.8    0.4-0.6




RESULTS WITH SMALL EUROPEAN CARS

   1.  Cu/Cr Oxidation Catalyst          0.6       8.5      5-0

   2.  Pt Oxidation Catalyst             0.3       3-0      5-0

   3.  Pt Catalyst + EGR                 1.7      10.0      1.0

-------
                            IX-6
     Although the US legislative requirements between 1972 and 1975



are not entirely clear,  it would appear that we shall have to meet




low limits for NOX in California in 1974,    As the figure would not




appear to be nearly as low as that proposed for 1975, however, it




could probably be met by the use of a small amount of EGR combined




with a catalytic afterburner, without too much loss of driveability.




The Federal limits for 1975 could probably be met in regard to CO




and HC by an oxidation catalyst system.   It is going to be very




much'more difficult to meet the 1975 limits for NOX, and we in




Europe would certainly hope that the application of these limits




would be postponed until 1976, as is provided for in the Nixon/Muskie




bill.



Exhaust Emissions from the Diesel Engine.




     Attempts to lower the exhaust emissions from the spark-ignition




engine by changes in combustion chamber design have been disappointingly




ineffective.   The same is not true of the diesel engine where the




swirl chamber system, long recognized as giving a cleaner exhaust




in regard to smoke than the direct injection system, is now shown




to be superior also in regard to gaseous exhaust emissions, particularly




nitrogen oxide.   This is illustrated by comparative tests carried




out on a single-cylinder engine of 1600cc (96 cu.in.) capacity




fitted with  a)  a Ricardo Comet V swirl chamber combustion system




and  b)  a Direct Injection combustion system.   Fig. 1 shows




comparative figures for carbon monoxide emissions over the load and




speed range of this particular engine.   The CO emissions are very




low in both cases, in comparison with the figures which would be




obtained from a spark-ignition engine, but, even so, those from




the swirl chamber system are lower than those from  the direct

-------
                             IX-7
EMISSIONS  OF  CARBON MONOXIDE  FROM  TWO
I2O X  I4O mm.  DIESEL ENGINES AT  OPTIMUM
INJECTION  TIM1MG5.
^  COMET ~JT COMBUSTION  CHAMBER
B.  DIRECT  INJECTION   COMBUSTION CHAMBER
                                                       DRG. No D. 21653
                                                       DATE:-  4- 11-70
                                                                   Plfi. I
 600   ;  8QO    :  looo
1600   !   I80O     200O  : ,

-------
injection system.   The unburned hydrocarbon figures shown in Fig.2,




although they are low in comparison with those from an untreated




gasoline engine, give no cause for complacency when we remember the




improvements which have been made in the latter unit in recent years.




Here again the swirl chamber engine gives lower hydrocarbon emissions




than the direct injection unit.   Fig. 3 shows the nitrogen oxide




figures and here it can be seen that the emissions from the swirl




chamber unit are less than half those from the direct injection unit.




     An additional advantage of the swirl chamber system is that the




injection timing may be retarded from the optimum with a big reduction




in NOX, and incidentally a reduction in noise, but with only a small




change in the smoke-limited power, as shown on Fig.  4-      A




corresponding change in the injection timing of the direct injection




engine, although it also reduces the nitrogen oxide concentration




considerably, at the same time reduces the smoke-limited power




appreciably.   There are very good technical reasons for the better




performance of the swirl chamber system in regard to nitrogen oxide




production.    The high rates of fuel/air mixing and therefore of




combustion of this system enables retarded injection timings to be




used, vrlth a lowering of nitric oxide concentration as noted above,




without the severe effect on smoke-limited power experienced with




the slower burning Direct Injection unit.    Also, with the swirl




chamber system, all the fuel is initally injected into only half




the air and only afterwards is this rich fuel/air mixture mixed




with the rest of the air in the cylinder and burning completed.




This results in a lower temperature/volume/time integral for the




the swirl chamber in comparison with the direct injection system




and this is an important parameter due to the strong temperature

-------
                              TX-Q
   EMISSIONS  OF  UNBURNED HYDROCARBONS  FROM
   TWO  I2O X I4O  mm.  DIESEL-  ENGINES AT OPTIMUM
               INJECTION TIMINGS
DRG. No. D. 21654
DATE:-  4 -II- 70

           FIG.2
_A^  COMET T COMBUSTION  CHAMBER
_B_  DIRECT INJECTION  COMBUSTION CHAMBER

 EMISSIONS  QUOTED  IN ppm. CARBON
                                               oo I   ieoo     2000

-------
                            IX-10
 EMISSIONS OF" OXIDES OF NITROGEN FROM TWO
 I2O X I4O mm. DIESEL ENGINES AT  OPTIMUM
              INJECTION TIMINGS
_A_ COMET TT COMBUSTION  CHAMBER
_B_ DIRECT INJECTION COMBUST/ON  CHAMBER
 EMISSIONS  QUOTED IN ppm. NO
DRG  No.  D 21655
DATE-   4-11-70
           FIG 3
                       ;: izoo  :  : 1400 ;
                            RPM

-------
                         TX-ll.
EMISSIONS OF OXIDES  OF NITROGEN FgQM
TWO  120 x I4O mm.  DIESEL  ENGINES WITH
INJECTION TIMING RETARDED  BV A-° FROM OPTIMUM

k_ COMET T COMBUSTION  CHAMBER
B  DIRECT  INJECTION COMBUSTION CHAMBER
EMISSIONS  QUOTED  IN  ppm NO
DRG.  No. D. 21656
DATE -  4- II - 70

          FIG 4

-------
                            IX-12
dependence of nitric oxide formation.




     In the United States, the State of California Air Resources




Board has proposed regulations for exhaust emissions from heavy




duty diesel engines, as set out in Table 2.    The second half of




the Table gives results derived from the single-cylinder engine




tests, just described.   It can be seen that CO is no problem, but




that the standard Direct Injection engine is marginal for 1973 and




that even the swirl chamber combustion system will not meet the




1975 limits in standard form.   It should be mentioned that these




single-cylinder engine test results are generally in accord with




those obtained from a wide variety of multi-cylinder units.




     With retardation of injection timing and possibly irith other




minor modifications to the engine, it should be possible to get down




to 5 gms/hp/h  with the swirl-chamber engine, but it is going to




be very difficult to do it with the Direct Injection system.




     Using EGR with the swirl chamber engine should enable figures




of 3 gms/hp/h to be achieved, and somewhat similar or slightly




higher values should be obtained with water injection.




     Neither EGR nor water injection are very attractive solutions,




however, because of their possible effect on engine durability in




addition to their influence on performance.   A catalytic method




of dealing with NOX, as is proposed for the spark-ignition engine,




would appear to be denied us on the diesel, as no one has yet




developed a catalytic method of eliminating nitrogen oxide which




will work in the oxidising atmosphere almost always present in the




diesel engine exhaust.    To this  extent nitric oxide poses a much

-------
                             IX-13
                DIESEL ENGINE EXHAUST EMISSIONS
                          (gma/hp.h.)
                                              CO      HC + N02
CALIFORNIA PROPOSALS                          ~      	

     1973                                     40         16

     1975                                     25          5


EUROPEAN DIESEL ENGINES

  1. Direct Injection                         2.6        13-4-

  2. Ricardo Comet Swirl Chamber              3-0         7.3

  3. Ricardo Comet 4° Retard                  5-5         5-9

  4. Ricardo Comet 4° Retard
       20% EGR                                8.2         3-0

  5. Ricardo Comet Optimum Timing
       2:1 Water/Fuel Injection               4.0         3-4

-------
                            IX-14
more difficult problem for the diesel engineer than for the gasoline



engineer and brings me back to the question I asked right at the



beginning.    Are we really sure that it is necessary to achieve such



low limits of nitric oxide in the engine exhaust?

-------
                       X-l
                    Chapter X
     LOW EMISSIONS FROM CONTROLLED COMBUSTION

       FOR AUTOMOTIVE RANKINE CYCLE ENGINES
                        by
   W. A. Compton, J. R. Shekleton, T. E. Duffy,
                 and R. T. LeCren
Solar Division of International Harvester Company
            San Diego, California, USA
       Presented at Einchoven Conference by
                  W. A. Compton
           Assistant Director-Research
Solar Division of International Harvester Company
            San Diego, California, USA

-------
                                     X-2
                                   ABSTRACT
        Rankine cycle engines have a high potential of meeting the emission levels
established by the 1970 amendment to the Federal Clean Air Act for the 1975-76 auto-
mobile. This paper discusses a Solar research and development program sponsored
by EPA/APCO which demonstrates a full scale 2 million BTU per hour working model
of a Rankine cycle engine combustor and controls which can surpass the emission goals
established.

        Special features of the combustor  are the unique methods of precisely control-
ling both the fuel and air to provide optimum flame  performance at any engine power
level.  This paper discusses the special requirements of the Rankine cycle engine and
shows why the very wide range of fuel flow required necessitates use of special tech-
niques  in fuel atomization,  fuel and air control,  and aerodynamics.  Sufficient discus-
sion is included to show the design methods that are necessary,  in general, to achieve
low emissions in continuous flow combustion systems.  Emphasis is placed on the
importance of interfacing a combustion system with other engine parts if a successful
low emission, wide turndown ratio combustor working model is  to be achieved. Suf-
ficient  discussion on combustion kinetics is included to advise on approaches necessary
to minimize NO formation in external combustion systems while maintaining high
efficiency and low CO and unburned hydrocarbons.

-------
                                        X-3
I.   INTRODUCTION

        The Rankine cycle engine has a high potential of meeting the established 1975-
76 emission levels when installed in a family car,  thus eliminating an atmosphere of
undesirable fumes now commonly contributed to the internal combustion engine.  The
major portion  of such an effort must,  however, be devoted to a system to develop a
full-scale (2 million BTU per hour) prototype combustor system to demonstrate that
the desired low emission levels can be met with the proper operating performance.

        Solar has addressed itself to these problems by drawing on  its many years of
gas turbine experience, studies of combustion kinetics, and applying a novel fuel
atomization method. In addition, a precise air-fuel  ratio control concept is necessary.
The solution lay in continuously modulating the fuel flow to match engine power demands
and to similarly control the air-flow to the combustor so that optimum combustion
could be obtained under all power level conditions.  Such a method of control lay outside
the state-of-the-art of present day methods and special techniques had to be applied to
the solution of this problem.

        Throughout the design, automotive features have been emphasized.  Low
emission, compactness, high  response, low cost potential, and high efficiency have
been major considerations in the design selection process.  The system described in
this report includes all necessary controls to supply and regulate both fuel and air.
High response rates are obtained while continuously  maintaining an optimum air-fuel
ratio with the lowest emissions.  An axial blower is  used to supply the required air to
the combustor.  Air-fuel ratio and fuel rates are controlled by a single power demand
lever.  The effects of air temperature, pressure leakage, flow, speed,  and fuel back
pressure are automatically compensated by simple mechanical control systems.

H.   1975-76 EMISSION LEVELS  FOR MOBILE ENGINES

        The goals set for this program were those for a six-passenger automobile
vehicle with a maximum test weight of 4600 pounds,  tested for emissions in accordance
with the procedure outlined in the November 10, 1970 Federal Register (Ref. 10). Values
are:

           • Hydrocarbons*       1.65 mg/gram  fuel   (0.46 gram/mile)
           • Carbon Monoxide   16.25  "   "     "    (4.7    "     "  )
           • Oxides of Nitrogen** 1.38  "   "     "    (0.4    "     "  )
           • Particulates         0.1   "   "     "    (0.03  "     "  )

            *  Total  hydrocarbons (using 1972 measurements procedures) plus
                total aldehydes.  Aldehydes to be  0.16 mg/gram fuel maximum.
            ** Computed as  NO2-

-------
                                        X-4
        The task of meeting the goals are graphically displayed in Figure 1, comparing
typical emission levels for light duty vehicles of 1960 and 1970 to those published as
requirements for 1975-76 in the 1970  amendment to the  Federal Clean Air Act.

III.   PERFORMANCE GOALS

        The basic performance goal of the demonstration system  is to develop a full
scale (2 x 10^ BTU/hr) prototype combustion system which could be  capable of achiev-
ing the emission levels when such a system is integrated into Rankine  cycle engines
installed in an automobile.  In order to have the widest possible technical significance
( to a potential family of yet undefined engines),  extreme boundary limits were placed
upon the performance goals.  These limits appear in the power range, transient
response, parasitic power,  and volume goals.  The importance of transient response
was emphasized because previously reported combustor and engine tests with external
combustors  indicated that on-off cycling and rapid transients were a major source of
emissions.  The following performance goals were used in synthesizing the combustion
system design:

          •  Heat release:  2 x 106 to 2 x 104 BTU/hr in a maximum  of 1.33
             cubic feet combustor volume.  The unit shall be able to control  the
             time average heat release at any point from 2 x 10^ to 2 x 10
             BTU/HR with low emissions.  This represents a 100 to 1 ratio of
             heat release between maximum  and minimum.

          •  Fuels:   Diesel No.  1,  Jet A, or Kerosene

          •  Rapid transients  in firing rates  without severe degradation of
             emission performance.  A goal  of 50 percent power change per
             second or 1 to 100 percent power change in 2 seconds has been
             established to allow interface with fast  response flash vapor
             generators.

          •  Minimum volume consistent with automotive research goals.
             Weight factors were considered, but prototype construction
             practices were used in the early research  model.

          •  Parasitic power losses of less  than 2 horsepower without vaporizer.

          •  Air density variations  caused by altitude ranges between  zero and
             5000 feet,  and temperature variations due  to variations from 0 to
             130° F will be accounted  for in the design.

-------
                                     X-5
                                NO2-mg/g
1960
ITYP)
                                    1
                   1970
                   (TYP)
                                    -25-
                                    — J
                                      0-
                                      3-
                                         1971
                                    -10-
41.S
             1*4
                 HC-
                                           20     40     «0     10
nng/g
                                                CO-mg/g
                                               I960
                                               (TYP)
                                                                          290
       FIGURE 1.   COMPARING TYPICAL LIGHT DUTY VEHICLE EMISSION
                   LEVELS TO 1975-76 LEVELS
         • Rapid startup to full power in less than three seconds.

         • High Reliability and low cost for automotive applications shall be
           inherent in the design approaches.


IV.  COMBUSTION SYSTEM DESIGN


4.1  REACTION KINETICS


       Three rules must be obeyed to burn liquid hydrocarbons quickly and efficiently
(Ref. 1, 2, 3,  and 4):
       1)  The fuel must be rapidly evaporated

-------
                                        X-6
        2)   The fuel and air must be rapidly mixed.

        3)   The rate of chemical reaction must be maximized

        The rules are satisfied when:

        _              a)  Fuel droplet size is small
        Evaporation
                        b)  Droplet to air relative velocity is high

                        c)  Air is injected at high velocity through a large number
        Mixing            of holes into a small combustor
                        d)  A large number of fuel injection points are used
        Reaction        e)  The air-fuel ratio is stoichiometric

In practice, unless heat losses are involved, it is necessary to add additional air in
order to avoid chemical dissociation losses.

        With variations,  these rules are the basis for the design of most combustors
used in cars, gas turbines,  power plants, and home heating systems.  Highly efficient
combustion can be obtained, free of emissions of smoke, carbon monoxide or fuel.
Unfortunately, this design method can result in high emissions of nitric oxide (NO).

        The rate of formation of NO has been found accurately represented by
          dNO            14  -67,000/T
         ——= 6.62 x 10   e
          dt
                                               T
1/2
                (Ref.  5)
where        NO, N£,  and 02 are concentrations of nitrogen, oxygen, and nitric
             oxide in mole fractions

             t is time in seconds

             T is temperature,  ° K

             P is pressure in atmospheres.

-------
                                        X-7

Time can be assumed infinite after one second and the resultant equilibrium values
of nitric oxide are shown in Figure 2.  Peak NO emissions occur at slightly greater
than stoichiometric air and fall rapidly when excess fuel is present.  The greater the
time spent at top temperature at any air-fuel, the greater is the amount of  NO formed
as shown in Figure 3.

        In a  reducing atomsphere, the lack of excess oxygen permits the use of
catalysts to break  down any nitric  oxide that is formed and so in an automobile, pre-
sents additional flexibility in control (Ref.  6).  In combustion systems using excess
oxygen,  catalysts are not practical and the only method of control is to limit the initial
NO formation.  The rule governing such a combustor design is:

        1)    Reaction time must be small in those parts of the combustor where
             air fuel is near stoichiometric.

             Therefore,  combining the rules for efficient and fast  combustion with
             minimum NO formation,  we would have a combustor of small  size,
             having a high pressure loss, a multiplicity of fuel injection points and
             an air-fuel distribution as follows:

             a)  A well mixed primary fuel-air zone having excess fuel, of
                 sufficient volume to permit maximum reaction of fuel

             b)  A secondary fuel-air zone where excess air is rapidly and
                uniformly added  and  having no more volume than is necessary
                to assure completion of the combustion reaction
             c)  A tertiary fuel-air zone (when needed) where any required
                excess air is added and where no reaction occurs.

Such  rules would apply in gas turbines, Rankine cycle engines, and any other combus-
tion system where air-fuels much  greater than stoichiometric are  involved.  In
Rankine cycle external combustors operating near stoichiometric,  some additional
control is permitted because heat losses to the boiler walls can bring down flame tem-
peratures substantially.  Often, in such applications,  inlet air preheat is used which
has the reverse effect.  In gas turbines, because wide operating conditions result in
large changes of air-fuel, there is only one unique point (usually a high power condition)
where the rules can be ideally maintained.  A penalty in increased emissions at other
operating conditions is ordinarily accepted and this is the cause of high hydrocarbon
emissions near airports because of ground taxi or start up.  Power plants in autos do
not have any unique operating point and typically have to start frequently and undergo
rapid power level changes.  In a gasoline engine,  it is therefore necessary to precisely
control the air-fuel ratio by means of a carburetor.  It is also considered essential in
Rankine engines used in an auto that the airflow into the combustor be controlled so
that the  rules for minimum emission are held at any fuel flow.

-------
       10
FIGURE 2.
                         10
                             AIR/FUEL RATIO
EQUILIBRIUM CONCENTRATIONS BY VOLUME OF CARBON
MONOXIDE AND NITRIC OXIDE AS A FUNCTION OF AIR/
FUEL RATIO

-------
                                   X-9
  10.000 i-
   i.ooo U
  o
                            10   12   14    16   18   20    22   24
    100 U
FIGURE 3.    EMISSION OF NITRIC OXIDE AS A FUNCTION OF AIR FUEL
              RATIO AND TIME

-------
                                      X-10
4. 2  AIR AND FUEL CONTROL SYSTEM DESIGN APPROACH

         The control  system was designed to regulate the flow of fuel and air to set
power outputs as a function of an input command from vaporizer control.  A fuel ratio
range  of 100 to 1 has been established as a design goal for this demonstration system.
Since present state-of-the-art controls have difficulties providing turndown ratios of
more than 15 to 1,  this particular requirement has been  recognized as a critical design
area requiring unique solutions. In order to optimize the low emission characteristics
of the  combustor, it is essential that throughout the entire power range (1  to 100%),  an
optimum air-fuel ratio be maintained.  An optimum envelope in  which the air-fuel
ratio should be centered across  the entire power range is discussed in detail in Section
5 and forms the basis for the control.

         By directly mechanically coupling a fuel metering valve and the large diam-
eter air metering plate (Fig. 4), the fuel flow area and the air flow area can be kept
in an exact ratio correspondence. Once the area ratios have been fixed, it is only
necessary to regulate pressure  drops  in order to maintain the desired weight flow
ratios.  Figure 5 shows the overall mechanical arrangement of the demonstration com-
bustor system.

         For low pressure rise, air compression can be neglected and the  weight
flow relationships can be written as:
                                                                         (1)
             Wf = Kf Af j4Pf                                              (2)

where        W  = mass air flow into combustor
              a
            A   = air metering valve area

             P  = upstream pressure (blower discharge)

             P  = downstream pressure
              LJ

             T  = air temperature at metering valve

             W  = fuel  mass flow

           AP  = pressure drop across fuel valve.

-------
                BYPASS VALVE
          *t.o
    FUEL METfRMG
    VALVE  1.0 to IM LB/HH
Af REGULATOR
L_.
FUEL
PUMP

•Ul UK 1
:~i i~
Q]

                                                  30 GPH
                                                          LV^I
       ^g  P    ^msu^

             ,1 O    _-<-	A
                            OK IF ICE
                              An
          -^ Of
p\ '.\\vsn:
L-. . r-
Iw.V.VX.' v
                 AT,

           P /T  COMPENSATION

          	VALVE

        - kfcV"*""^-
3EALED CATrTY
            FIGURE 4.   CONTROL SYSTEM SCHEMATIC

-------
FIGURE 5.  DEMONSTRATION SYSTEM

-------
                                      X-13
If the inlet pressure (altitude) and temperature are constant, a simple relationship
can be written for the air fuel ratio:
                                                                          (3)
        To ensure the air-fuel ratio remains a function of flow area, only the pres-
sure drop across the fuel valve will be controlled proportional to the pressure drop
across the air metering ports.   A APf regulator valve performs this function by a
force balance across diaphragms.  Operation of this component is illustrated in Fig-
ure 4.  PI pressure is connected (through a density compensator) to the bottom side
of diaphragm A]_.  Downstream pressure (P2) is connected to  the opposite side of the
diaphragm.  Fuel system pressure is regulated by a flapper valve that recirculates
excess fuel back to the tank.  A small diaphragm on the fuel side balances the pressure
across the air metering ports A^ against the fuel pressure.   If the air pressure c rop
increases (causing Wa to increase), the force across the system  becomes unbalanced
and the fuel pressure  regulator moves up; reducing the by-pass flow, and thus increas-
ing the fuel pressure and its mass flow across the fuel metering valve.  Since the  force
balance must be maintained across the APf regulator, we have:

         E  Forces = Ag (APf - PJ =\(PI~ Pg)

             and where AP is large compared to P
                         i                      £
                        t
                where P  = P (times) ambient air temperature and pressure
                               compensation factor
                 A!    -
we have     AP  = ~—(P -P)                                           (4)
              t  A2    1   2

Thus equation (3) can be rewritten as
                          ' P2) A2
                          _ 4  _ z
                                   = constant
        Blower speed changes due to voltage or load variations, blower efficiency
reductions due to fouling or wear, and by-pass valve leakage variations are auto-
matically compensated by maintaining the  AP as a function of the air valves
pressure drop.

-------
                                       X-14
        It can be seen from equation (1) that the air flow is a function of
Thus, an altitude and ambient air temperature correction is applied by adding a
TI compensator (Fig.  4).  Blower discharge pressure (P^) is admitted to the AP regu-
lator by a contoured needle valve.  This valve is positioned by a force balance across
a diaphragm having P^, T^ on one side and cavity sealed with air at standard conditions
Thus, as the air temperature increases,  the sealed air will expand and move the valve
upwards.  This action will increase the pressure drop across the contour valve  and
thus reduce PI to a corrected value of PI on the bottom side diaphragm A.  As PI
decreases, the fuel pressure will drop, producing the desired fuel flow reduction as
the ambient air temperature  increases.

        Air to the combustor is metered across a series of twelve ports cut into two
plates (Fig. 6).  An  input lever  rotates one plate with respect to the stationary backup
plate.  The metering port areas are caused to open or close as the input power lever
rotates matched ports in each plate.  By  maintaining a known relationship to power
lever position and metering port area, the port contour can be arranged to provide
near linear control as a function of power lever position and voltage with the AP regu-
lator compensating for differences in the range of 1 to 2.5 inches of
         The capability of the fuel regulator to control fuel pressure levels propor-
tional to differential metering valve pressures of 1 to 2. 5 inches of H2O is vital to
obtain a wide flow range control with high response that maintains an accurate air fuel
ratio from 1 to 100 percent power demands.  Initial analysis indicated that the use of
motor voltage (speed) regulation could not meet the range accurately or the stringent
transients improved by the Federal driving standards. Inertial lags of the motor would
keep it out of synchronization with the fuel flow for an unacceptable percentage of the
standard driving cycle.  By maintaining the fuel pressure proportional to the AP across
the air metering plates, the demonstration  system allows limited voltage control (to
reduce parasitic losses and noise at normal driving speeds) with precise air-fuel ratio
control  under transient conditions.   If the by-pass valve is maintained closed for the
power range from 100 to 70 percent, the air flow into the  combustor can be accurately
controlled by proportionally lowering the voltage to the blower motor while the regu-
lator compensates for the reduced pressure differential across the air metering valve
at lower speeds.  Since a mechanical linkage maintains a  known ratio of air valve area
to fuel valve area,  the regulator will automatically maintain the air fuel ratio as the
power is reduced.  A limitation on  this sytem is the lower differential pressure signal
that can maintain accurate fuel pressure regulation.   Analysis and test has indicated
that 1. 0-inch of H2O is a reasonable lower  limit. Using this value, the part load

-------
                                       X-15
                 MrmiNG onmcra
                  TTP. (11) PLCS,
                                                             TAN MOTOR
                                                            COOLING FINB
        FIGURE 6.    SWmL ELIMINATION AND PRESSURE EQUALIZATION
                      BAFFLE ARRANGEMENT
power demands of the combustor system can be established.  Although the combustor
requirements are  moderate (1.25 HP for the optimum configuration and 2. 3 HP for
the demonstration system), addition of a typical boiler configuration can make  parasitic
power losses at part loads unacceptable. If a boiler of the type described in Reference
(11) was used, an  additional 2.4 horsepower would be required.  Total parasitic power
levels would then be as high as 3. 65 horsepower.  If the system were to  require this
high power level from full power down to idle condition, a highly undesirable condition
would exist.  Lowering the input voltage across a voltage range that allows accurate
AP compensation by the regulator will eliminate the motor inertial speed lags since
the air fuel  ratio changes are continually maintained by the regulator.   Power reduc-
tion by operating at 1. 0 inch of ^O can be  estimated by simple calculations based on
fan laws for a series wound motor.  For voltage changes of approximately 2 to 1, the
following relationships  give reasonably accurate results.

        For a given voltage change V  to V

            flow Q:         Q/QO = V/Vo
                             1  A    L   £
                                           -/
            pressure p     Pi/p2 = (Vi/V2)

            fan BHP is:     BHP  /BHP0 = (V /V

-------
                                       X-16
         By maintaining the bypass valve closed and reducing the voltage to the motor,
the flow  through the metering valve will drop as a function of both the motor voltage
and area change of the metering valve.  Pressure will drop as a function of the square
of flow.  By using 1. 0-inch as the lower limit of adequate control,  the bypass valve
can be maintained in a closed position until the  voltage is reduced to  1/2 or 0. 7077-^
Since power is approximately a cubic function,  we will obtain (0.707)3 (BHP), or a
65 percent reduction in power.  With the optimum system and a relatively high air
side boiler pressure drop, a parasitic power loss reduction of approximately 2.4 horse-
power.   Total power with this high pressure drop boiler  would be approximately 1.25
horsepower at vehicle power demands below 35 percent.   One difficulty with this
approach is that it is more difficult to make the system linear.  However, since this
is not a driver input command linearity, it is of secondary importance if the correct
air-fuel ratios can be maintained.

         Fuel must be accurately mete red from 109 to 1. 0 pounds per hour.   At low
fuel rates (1. 0 pounds per hour), the flow  is approximately two drops per second.
Standard valves do not have sufficient linear range to accommodate these severe
requirements. Additionally,  the valve should not be sensitive to temperature induced
fuel viscosity changes.   A new approach to this problem  has been taken by the applica-
tion of a dual  slotted shear valve (Fig.  7) consisting of two flat (ground and lapped)
plates with matched contour slots 90 degrees to each other. At the intersection of the
two slots, a square orifice is formed whose area is a function of the  relative position
of the top movable plate.   The square shape (and thus the discharge coefficient) can
be maintained constant throughout the entire 100 to 1 area ratio.  Since the plates are
in contact, fuel will flow only through the slot in each plate and not  between the plates,
thereby reducing the clearance leakage path to the microfinish of the contacting sur-
faces.  Figure 8 shows the assembled fuel valve of the demonstration system.  A drain
groove is provided between the upstream pressure and the metered outlet fuel passage,
thereby reducing the potential leakage pressure to the level of the frictional flow drop
to the spin cup.  Since this is normally less than 0.5 psid, resulting leakage of metered
fuel into the drain system will probably be negligible.

         The size of the orifice slots used is a trade-off between four factors.

           •  Fabrication capabilities - requires large dimensions
           •  Contamination - requires large dimensions

           •  Backpressure sensitivity - requires high pressure and thus,  small sizes
           •  Temperature sensitivity - it  is desired that changes in fuel temperature
             have little effect of the coefficient of discharge.  This factor requires
             a high Reynolds number and thereby small orifice dimensions.

-------
                                     X-17
                                               10 : 1 SLOT
                                                                 DRAIN TO TANK
MOVABLE PLATE —
FLOW AREA SHADED
-ROM Fl'El , ( \
REGULATOR (J \






( ri

LJ 4
DENED AND LAPPED __/
VR PLATE SURFACES
h'r
1
V
SPRING
/ POWER LEVER
/ POSITION INPUT

10 PSIG u \j, ^ y^^\
-^^- ' ' ' '^ ' ', ', \\ \
<• - f -^B^l ^V ^^-^ /
J t *. ». ^ J X,_ j*
I /f^»^» ^**"™ "
^ 1
DHA1,N
0 PSIG ^J n«^- TO SPIN CUP
     FIGURE 7.   FUEL METERING VALVE CONCEPT FOR 100:1 TURNDOWN
4.3  FUEL INJECTION, ATOMIZATION AND IGNITION

        The requirements of the fuel injector are:

        1)   Small droplet size

        2)   Reliable and low cost

        3)   Multiple injection points

        4)   Precise spray angle.

In detail,  these requirements entail:

        1)   Small droplet size must be obtained with kerosene at any fuel flow between
            1.0 and 109 pounds per hour with fuel viscosity, varying with ambient
            temperature, ranging from less than 1 to over 16  centistokes.

-------
                            X-1S

        FIGURE 8.  FUEL METERING VALVE ASSEMBLY
2)   The injector must be of proven design, unsophisticated,  not requiring
    a high fuel pressure pump or other complex or costly auxiliaries.
    Fuel orifices must be large to prevent contamination.
3)   The object of multiple injection is to provide the maximum interface
    of fuel and air and hence provide fastest and most uniform mixing so
    important for optimum air-fuel ratio and fastest reaction rate.  Pack-
    aging in an auto requires,  at this stage, maximum design flexibility.
    If the combustor were mounted horizontally,  being at least 10 inches
    in diameter, the manifold  head effect would require, for good fuel
    distribution, a fuel pressure of over 10, 000 psi,  and is impractical.
    The mixing must therefore be mainly achieved by good design of air
    injection and providing the maximum surface of fuel spray to the air.

4)   Precise spray angle is necessary to control fuel air mixing and also
    to obtain ignition by assuring that the spark is  adjacent to but not
    smothered by the fuel droplets.  Ignition is easiest at low velocities
    and hence should be done at 1.0 pound per hour.   (An ignition failure
    at 109 pounds per hour would be a dangerous fire  risk and serious
    polluter.)

-------
                                      X-19
        There are four principal classes of fuel injectors (1) fuel pressure, (2)  air
assisted, (3) rotary,  and (4) vaporizer (Ref.  ). The first three operate on the same
principle; i.e.,  the fuel is presented to the air at a high relative velocity which shatters
the fuel into small droplets. Vaporizers rely on heat input, usually from the flame,  to
vaporize the fuel. A design study, adumbrated below, indicates the rotary injector to
be the best choice.

        1)   Pressure injectors cannot atomize fuel flows as low as 1 pound per hour
            except when nonviscous fuels are used.  Orifice size would be minute and
            fuel pressures above 250,000 psi are needed.  Spray angle is extremely
            sensitive to viscosity.

        2)   Air assist injectors use high velocity air to atomize the fuel.  There
            are two types, one using high air volume and low pressure (8-inch water)
            (Ref.  2).  The other using high air pressure (several psi) and low volume.
            Some  of these injectors use such high  air pressures that sonic velocities
            are reached (Ref.  7).  The high air volume system is not practical, be-
            cause at low fuel flows, the air flow required would be  far too much for
            combustion.   The  high pressure system injects air at such high velocities
            (400 ft/sec) that at low fuel  flows, where combustor air velocities are of the
            order of 1 ft/sec,  a flame could not be stabilized.  In addition, the spray
            angle  is sensitive  to viscosity,  especially at low fuel flows and, coupled
            with the high air velocities,  makes reliable ignition doubtful.  Unless high
            fuel pressure is used at high fuel flows (with resultant small orifice sizes)
            air flow has to be  high and substantial power is required. Both types of
            injectors need power absorbing and costly auxiliaries and involve the  use
            of small air and fuel passages.

        3)   A rotary atomizer does not  need auxiliaries as it can be mounted  directly
            onto the fan required to deliver combustion air.  No fuel pressure or  fuel
            orifices are needed as the fuel  is passed through a large tube (3/16-in.
            diameter) that passes through the center of the fan motor shaft, and trick-
            led onto the cup surfaces.  The fuel adheres to the cup  wall and is ejected
            from the outer lip of the cup at a velocity of about 40 feet/second.  The
            power required to drive the cup is negligible.  Fuel spray angle is extremely
            precise and independent of fuel flow or viscosity (no ignition failures have
            occurred in tests to date).

        4)   A vaporizer requires an auxiliary source of heat for ignition and  initial
            flame propagation.  The response is slow  and the heat losses involved at
            1 pound per hour would  prevent combustion unless auxiliary heating were
            continuously supplied.  The  large variation in heat release makes it

-------
                                       X-20
             most difficult to prevent overtemperature of the vaporizer at high flow
             rates while maintaining sufficient heat at low flows.  It is sensitive to
             fuel volatility, hence requires a more precisely refined fuel type than
             the rotary injector and flame performance will be significantly influenced
             by ambient day temperature (in a cold start it will smoke).   These objec-
             tions apart,  a vaporizer is a desirable system.  This is because the
             absence of fuel droplets avoids evaporation delay and allows a more
             uniform mixing of fuel and air.

4. 4  COMBUSTOR DESIGN

        The combustor must be small  in size in order to  fit in an auto and also to
limit combustion  time.  The use of high combustion velocities requires high air injec-
tion velocities for proper mixing and hence high pressure  loss.  A limit of auxiliary
fan power prevented  a pressure loss of more  than 8 inches of water being available.
This dictated the  use of a relatively large combustor (Fig. 5 and 9).   The combustor
was a scaled up version of a similar rotating  cup combustor used in a Solar 10 KW
gas turbine under development.  Construction techniques are typically as used in gas
turbines,  except that the best available oxidation resistant material was used (Hastelloy
X). This  permitted  operation of highest possible wall temperatures and minimized
film cooling.  Film cooling,  by its very function, implies  low rates of mixing  (Ref.  9)
and breaks a fundamental rule in mixing of fuel and air in  that it must be good.

        Air was  arranged to enter the  combustor through various ports, the air-fuel
ratio calculated,  and computer runs made of the resultant emissions. Figures 10 and
11 are typical results.  Combustion time  at full fuel flow is nominally 0. 01 second,
and this increases to 1.0 second at minimum fuel flow.  The resultant increase in NO
emissions at low  flows is clearly seen in  Figure 11.

V.  DEVELOPMENT TESTING

5.1  COMBUSTOR TESTING

        Initially, the fan was not used  and air was supplied by a remotely located air
compressor that supplied air via a plenum to the combustor.  This guaranteed uniform
air distribution.  For convenience of test, the cup was driven by a small electric motor
independent of the fan (Fig. 12).

        After development of the combustor and cup, it was possible to maintain an
efficient flame over  a range of fuel flows  from over 109  pounds per hour  to less than
0. 3 pounds per hour, provided that the  air-fuel ratio was kept within reasonable
limits  (Fig. 13 and 14).  A range of air-fuels were tested  and Figures 15  and 16 show
the emissions of CO  and NO at the optimum air-fuel together with the results of a 10

-------
                                      X-21
                                            EXHAUST STACK & LOCATION
                                                  FOR BOILER
        FIGURE 9.  SIDE VIEW OF ROTATING CUP COMBUSTOR ASSEMBLY


percent error in air-fuel.  Figure 17 shows the optimum air-fuel required for mini-
mum emissions.  During the testing, hydrocarbon emissions were monitored, at
optimum air-fuels, below background levels.  This is not surprising in view of the
close proximity of the test facility to the San Diego airport (200 yards).  The results
are noteworthy in that:

        1)   The NO emissions fall off sharply at low fuel flow and the CO emissions
            increase.  This is due to the high heat losses from the flame at these
            conditions.

        2)   The large  increase in emissions in the mid-fuel flow range when any
            deviation from optimum air-fuel occurs.  This was found caused by
            poor fuel atomization  that occurred in this region and was subsequently
            eliminated by design changes.

        3)   The large  variations in optimum air-fuel as a function of fuel flow.  At
            high fuel flows, this is partly attributable  to the poor fuel atomization
            which varied  as a function of fuel flow and due to the changes in fuel-air

-------
                       X-22
  FIGURE 10.  AIR/FUEL RATIO ALONG COMBUSTOR FROM
             COMPUTER ANALYSIS
                        DISTANCE - INCHES
FIGURE 11.  NO» CO CONTENT FROM COMPUTER ANALYSIS

-------
                           X-23
ROTATING CUP
                                 ROTATING
                                CUP, MOTOR
  FIGURE 12.   ROTATING CUP AND MOTOR ASSEMBLY
  FIGURE 13.  FLAME AT 109 POUNDS/HOUR FUEL FLOW

-------
                                       X-24
             FIGURE 14.   FLAME AT 0.25 POUND/HOUR FUEL FLOW
             mixing that inevitably occur.  At low fuel flows, the most efficient
             flame resulted when the primary zone was leaned out with excess
             air.  This shortened  the flame, reduced heat losses, and hence pro-
             vided more efficiency.

        4)   The increase in CO as a function of fuel flow.  A not unexpected result
             in view of the time dependence of the reaction.   By extension of the
             secondary combustor zone, the CO emissions can  be reduced and, pro-
             vided good fuel-air mixing has been obtained, NO emissions theoretically
             would not increase significantly.

        Tests were done with a water cooled heat exchanger  mounted on the rear of
the combustor and designed to simulate the vaporizer that, in the final design, must
be used.  NO emissions were significantly reduced,  CO increased slightly, but hydro-
carbon emissions increased an order of magnitude.  It is concluded that heat losses
to the vaporizer must therefore be minimized by shielding it from the flame.

        Tests were then done using the fan and  control system  previously shown in
Figure 4.   The initial combustion characteristics were totally different and emissions
unacceptably high.  Two main features were  noted.  At high air flows, circumferential
air maldistributions were so high that raw fuel escaped from  the exhaust.  At low air
flows a pronounced swirl occurred  in the flame.  Air distribution circumferentially
was good but axially considerably different from the rig tests.  Aerodynamic analysis

-------
                                     X-25
    u
    U
      17
       16
       15
       14
       13
       12
       11
       10
    I
FIGURE 15.
                      16.25 - PROPOSED LIMITS 1975-76
                                Wa nnn = 2835 LB/HR AIR
                                  100%
              10    20
                        30   40
                                  50
                          AIR FLOW,
                          60
                         Wa
                       Wa"
                                            70   80
                                            ±10% FROM OPTIMUM
                                                 AIR-FUEL
                                                           OPTIMUM AIR-FUEL
                                                           j
                                          90   100
                         100%
EMISSIONS OF CARBON MONOXIDE AS A FUNCTION OF COMBUSTOR
AIR FLOW,  AT OPTIMUM Am FUEL  FOR MINIMUM EMISSIONS AND
ALSO WITH A ±10% DEVIATION OF AIR FUEL FROM OPTIMUM

-------
                                      JX-26
      u
      D
      fo
      o
      O1.8
        1.6
        1.4
      X
      O 1.2
      O
      s
        0.6
        0.4
                                           PROPOSED LIMITS OF EMISSION
                                             ± 10% FROM OPTIMUM
                                                  AIR-FUEL

                                                  OPTIMUM
                                                  AER-FUEL
                                    Wa   „  2835 LB/HR OF AIR
                                      jLOu /o
                                                              j
                10
                     20
                          30
                               40
                                     50
                                AIR FLOW,
                            60   70
                              Wa
                            Wftioo%
                                                    80
                                                         90   100
FIGURE 16.
EMISSIONS OF NITRIC OXIDE AS A FUNCTION OF COMBUSTOR AIR
FLOW,  AT OPTIMUM AIR FUEL FOR MINIMUM EMISSIONS AND
ALSO WITH A ±10% DEVIATION OF Am FUEL FROM OPTIMUM
indicated these two problems were caused by two design features in the fan design.
The fan used was a highly loaded solid vortex design typical of those used in aircraft
where minimum size and weight is needed (Ref.  8).  The velocities in it were high
and to supply air to the combustor, substantial diffusion was required. Diffusion is
unstable and results in circumferential maldistributions.  At low air  flows,  the pres-
ence of  slight swirl from the fan results in the establishment of a free vortex in the
combustor of relatively high velocity (the fan runs continuously at full speed).  Because
the combustor pressure  drop is low (0. 0008-inch water at 1. 0 pounds per hour of fuel)
this has a significant effect on the discharge coefficients of the air metering holes with
consequent axial maldistributions. Aerodynamic problems of this sort are the bane
of combustion design.  In gas turbines, not only is efficiency reduced but serious hot
spots are  caused.  The much wider range of air flows of the Rankine  combustor makes

-------
                                      X-27
                                         50
                             60
70
80
90   100
                                    Wa
    FIGURE 17.
                Waioo%
AIR-FUEL RATIO FOR MINIMUM EMISSIONS AS A FUNCTION
OF COMBUSTOR Affi FLOW
the problem of precise air-fuel control more critical and the tests indicate it to be a
leading problem if emissions are to be kept low and vaporizer life to be long.

        The expedient of introducing a large plenum between fan and combustor
(Fig. 18) provided a temporary solution and rig and fan tests were found to aerody-
namically duplicate each other.  This was confirmed by exhaust temperature traverses
of both rig and fan air supplies and which were in close agreement (Fig.  19).  The
profile of temperature is indicative of inadequate mixing.  Apart from detracting
from vaporizer reliability, it is bound to raise emissions because optimum air-fuels
are not maintained.

-------
                                    X-28
 FIGURE 18.  FAN AND COMBUSTOR ASSEMBLY UTILIZING LONG MIXING DUCT
       5. Or
  4.0


a
X
u
5 3.0
CO

s
     BJ
     U
     J


     8
       2.0
        1.0
       0.0
                       FAN AIR SUPPLY
                                      RIG AIR SUPPLY
           1400    1600     1800     2000    2200     2400     2600     2800

                            BOILER INLET TEMPERATURE, °F

FIGURE 19.  REPEATABILITY OF RADIAL PROFILE OF TEMPERATURE INTO

            BOILER USING BOTH FAN AND RIG AIR SUPPLIES

-------
                                      X-29
        As the plenum involved a duct of 36 inches in length, it is not acceptable for
use in an auto.  The solution lies in a fan designed specifically for the combustor, of
larger diameter and lower velocity.  The air control system would retain the present
features and it is doubtful whether any other air control system,  such as a butterfly
valve, would be satisfactory because of the asymmetry of air that would result.

        After further development, endurance tests indicated the most significant
problem to be carbon build  up  in the combustor walls. Temperature and emission
measurements, as well as visual flame observation indicate that improvements in
air-fuel mixing are needed; and it is expected that further development can provide
substantial performance improvements.

5.2  CONTROL SYSTEM DEVELOPMENT TESTS

        Control system development tests have been completed on all major compo-
nents.  Final fuel metering valve performance tests have demonstrated the gain to be
61 pounds per hour per inch stroke with a linearity of ±2.5 percent across a range of
3 to 115 pounds per hour  (Fig. 20).  Good repeatability and flow control to 0. 5 pounds
per hour indicates the valve has a dynamic range greater than 200 to 1 (Fig.  21).  Fuel
pressure regulator valve (Fig. 22) development tests have extended its range  and
improved accuracy.  An important goal of extending the control range to 1. 0 inch of
H20 input actuator pressure differential has been achieved.  Blower motor power reduc-
tions of as much as 65 percent can now be compensated with the fuel regulator.  Correct
fuel flow can be maintained within ±4 percent with actuator input differential pressures
ranging from 1. 0 to 2. 5 inches of H2O (Fig. 23).

        Air metering valve problems caused by high dynamic head, swirl,  and turning
losses by use of an off-the-shelf blower, required high pressure loss baffles and flow
straighteners to be installed.  As a consequence, the blower motor was required to be
operated approximately 6 volts above  its normal input.  Additionally, combustor tests
with the air valve showed that  potential aerodynamic problems existed that would require
considerable development effort to minimize volume.  Analysis of these  problems
indicated that the best solution would be to specially design a matched fan-motor com-
bination to the functional  and geometric requirements of the combustor.  It was estab-
lished that the demonstration system should utilize a simple dump plenum to make the
fan more aerodynamically compatible with the air valve and combustor.  Thus, a 36-
inch extension between the fan and air valve was incorporated into the demonstration
valve.  Figure 24 shows typical performance on bench tests of the air metering valve.

-------
                                                                       SLOPE - 61 PPH/INCH
                                                                                            PRESSURE DIFFERENTIAL: 10 PSID
                                                                                            FUEL: JP-5
                                                                                            TEMPERATURE: 76'F
                                                                                           O DATA RECORDED 12/11/70
                                                                                            • DATA RECORDED 12/15/70
0.1     U.2
                       0.4      0.5
                                              0.7      0. s
                                                                     1.0     1.1      1.2     i.'J
                                                       FUEL VALVE POSITION (INCHES)

        FIGURE 20.     FUEL  METERING VALVE  FINAL  WEIGHT FLOW CALIBRATION
U)
o

-------
                                      2-31
         10.0
         a.o
          8.0
          7.0
        8
        I
        OS
        H
        O,
6.0
        H

          3.0
          2.0
          1.0
                      SLOPE - 64.2 PPH/IN.-
                                           SLOPE • 61 PPH/IN.
                                                PRESSURE DIFFERENTIAL:  10 PStt)
                                                FUEL: JP-5
                                                TEMPERATURE: 76 *F
                                         O DATA RECORDED 12/11/70
                                         • DATA RECORDED 12/15/70
                      LINEARIZATION REFERENCE POINT (1.0 PPH AT 0.011 INCHES)
             0      0.02    0.04     0.06    0.08    0.1     0.12     0.14    0.16
                               FUEL VALVE POSITION (INCHES)

FIGURE 21.  FUEL METERING VALVE FROM 0.5 TO 10 POUNDS PER HOUR
                (FINAL WEIGHT FLOW  CALIBRATION)

-------
                                      X-32
                    fyssssaas^^                            -
      FIGURE 22.   FUEL REGULATOR ASSEMBLED WITH Pl - P2 ACTUATOR


5. 3  EMISSION MONITORING

        All emission data presented was taken with Beckman Model 315A,  CO, CO2,
and NO analyzers along with Beckman Model 402 Hydrocarbon Analyzer.  The emis-
sion monitoring system is designed in accordance with existing vehicular specifications
as defined by Federal Standards, the Automobile Manufacturing Association and the
State of California.  It was selected for use in this program because its wide spread
acceptance permits direct comparison of results with other investigators' work.  The
CO, CO2, and NO analyzer uses NDIR cells selected for the ranges required for the
program.   Table I lists the ranges,  repeatability and interference factors that are
characteristic of this measurement system.   In order to achieve a 0-150  ppm range
for NO, an exceptionally long (41 inches)  NDIR cell is incorporated for this important
parameter.  Figure 25  shows the control  panel and strip chart data readout employed
with these instruments. Sampling was accomplished with a 1/8-inch diameter cooled
probe for CO, CO2, and NO  to prevent reaction in the sampling line.  Hydrocarbons
were sampled with a line maintained  at 350° F by means of electric heaters.

VL  CONCLUSIONS

        The 1975-76 emission levels for automobiles are feasible when using a Rankine
engine combustor.  Package  size and power requirements of the combustor need not be
excessive,  the components used are not complex, are capable of low cost mass produc-
tion and,  as involve proven concepts, high reliability is attainable.  The fuels used
need no critical refining capabilities, and are much safer and less expensive than
gasoline.  They can be  instantly burned in the coldest weather and do not  have any

-------
                                                                       • 11X1% FLOW
                                                                       A  *C* FLOW
                                                                       •   It FLOW
      1.1
                      1.4       l.j       1.6       1."       i.S       Lit       2.0
                          CONT&UL ACTUATOR INPUT PRESSUREt>lFFER£NTUL (INCHES OF WATEB)
                                                                               2.1
                                                                                               2.3
           U)
1.5
FIGURE 23.  REGULATOR CALIBRATION WITH 0.008 INCH FLAT RUBBER DIAPHRAGM

-------
         100
            —   I
      B
         70
         r,o
         40	
                       CALCULATED
                       ACTUAL        I /
                            TEST WITH 2.25 WIDE
                            BY-PASS VALVE
                            (BAND ON ORIFICE
                            PLATE SIDE)
                                      j     i
         Ki
          0 ^-	1
            0   10    20
                        7% FtOW@ 7. 5% PLP
                                           «Me>KE VELOCITY
5% FLOW @ 5.9% PLP  :CALIBRATIONS
2%FLOW (§!3.8% PLP  |     |      j          I
             r>0    (JO    70    80   90   100
                           30    40
                            POWER LEVEL POSITION (PLP) %
FIGURE 24.   FLOW CONTROL PERFORMANCE OF AIR METERING VALVE

                                    TABLE I
       INSTRUMENTATION ACCURACY USED FOR  DATA ANALYSIS
Constituent
Gas
CO
C°2
NO

H C
x y

Ranges
0-1000 ppm by Vol.
0-2.5% by Vol.
0-250 ppm by Vol.
0-16% by Vol.
0-5% by Vol.
0-1000 ppm by Vol.
0-150 ppm by Vol.
0-5 ppm
16 ranges through
0-250,000 ppm
Repeatability
% of full Scale
1.0
1.0
1.0
1.0
1.0
±1.0
±2.0
<±5

Interference
Interfering
Gas
co2
H20
	
CO
co2


Mole
Percent
5.0
3.3
	
5.0
5.0
—

Extraneous
Response
3.4 ppm
1. 1 ppm
	
1 ppm
3 ppm
---

Calibration
Gas Accuracy
±r,%
±5%
±5%

±5%


-------
                                      X-35
   FIGURE 25.   TOTAL VIEW OF BECKMAN GAS ANALYZER INCLUDING THE
                 FLAME  IONIZATION DETECTOR
warm-up emissions as occurs on current engines.  They need no special additives for
combustion control, nor will the current automotive evaporative control systems be
necessary.

        Development  is necessary, especially in regard to integration with other
engine components because they can affect both the shape of final performance of the
combustor and particularly, the critical aerodynamic  interface necessary for low
emissions.  The most critical area is in air-fuel  control. To achieve the best possible
performance (and  the results,  while below current goals, could be substantially  im-
proved) requires aerothermodynamic sophistications more typical of advanced gas
turbine combustors than conventional heating systems.

        Further development of the combustor components would yield substantial
benefits in both size and power requirements (Fig. 26).  However, if an economic
vaporizer of reasonable  size was used (Ref.  11),  the horsepower requirements of the
combustor would increase from 1. 25 to  3. 65.  By developing  the inherent variable fan
speed capability of the fuel air control system, the parasitic losses at part loads could
be maintained at approximately 1 horsepower. Typically, maximum power is  required
only transiently and normal power output is 35 percent or less than maximum.  This
suggests that the design  should be optimized for the normal engine load demands of 35
percent power or less.

-------
        If the motor speed was reduced to supply the air required for 35 percent
power, the total power required of combustor and vaporizer would be reduced from
3.65 to 1.25 horsepower, and full fan power would only be required  in occasional
engine power demands above 35 percent where fan speed would be increased.  Such
a design concept provides considerable potential for combustor improvements as it
permits the use of higher combustor pressure losses and a smaller  combustor.








-4 	 — 31-




N
/
1.16 FTJ
COMBUSTOR
PLUS CASE






Tc
'r



••


^n r~"
UP j
OTOR| !
-1 i
12.7f, DIA
(J. M FT3
AIR VALVE
|_ 	 	 	 ,




...
FAN
MOTOR
FAN





'

1





7 1

^ r





nw

-3

                                                   PANCAKE
                                                   DC MOTOR
                                                  FUEL INLETI
                                                               FAN RLADE

                                                                FAN STATOII

                                                                 CONTROL SECTION
                 TOTAL VOLUME 1 89 FT"
                 DEMONSTRATION SYSTEM
                                                             OPTIMUM SYSTEM

3
Total Volume (ft )
Length (Inch)
Diameter (Inch)
3
Combustor Volume (ft )
Combustor Diameter (Inch)
Horsepower
Combustor Loss (Inch Water)
Diffuser Loss (Inch Water)
Metering Loss (Inch Water)
Overall Pressure Loss (Inch Water)
Fan Efficiency (%)
Motor Efficiency (%)
Overall Efficiency (%)
Motor Speed (rpm)
Demonstration
System
Design
1.89
31.0
13.00
0.687
11.00
2.30
8.00
3.00
2.00
13.00
75.0
75.0
56.2
14000
Optimum
System
Design
1.09
14.25
13.00
0.687
12.00
1.25
6.0
0.5
1.5
8.0
85.0
75.0
68.0
7000
Spec
1.33



2.00








   Note:  Ignition, Fuel and Air Regulators not included.

         FIGURE 26.  TWO FAN SKETCHES - PRESENT AND OPTIMUM

-------
                                    X-37
VII.  REFERENCES

  1.    Lefebvre,  A. H., "Theoretical Aspects of Gas Turbine Combustion
       Performance".  Co A Report Aero No. 163, The College of Aeronautics,
       Cranfield,  England (1966).

  2.    Lefebvre,  A. H. and Miller,  D.,  "The Development of An Air Blast
       Atomizer for Gas Turbine Application".  Co A Report Aero No. 193,
       The College  of Aeronautics, Cranfield,  England (1966).

  3.    Spalding, D. P.,  "Performance Criteria of Gas Turbine Combustion
       Chambers".  Bunhill Publications, Ltd., London, England (1956).
  4.    Hottel, H.  C., Williams, G.  C, and Miles, G. A.,  "Mixedness in the
       Well Stirred Reactor".  Eleventh Symposium on Combustion,  1967,
       L.  C.  Card 55-9170.

  5.    Caretto, L.  S., Sawyer, R. F. and Starkman, E. S., "The Formation of
       Nitric Oxide in Combustion Processes".  Central  States Section/Combustion
       Institute (1968).

  6.    Yolles, S., Wise, H.  and Berriman, L.  P., "Study of Catalytic Control
       of Exhaust Emissions for Auto Cycle Engines".  Stanford Research
       Institute (1970).
  7.    Hawthorne, W.  R. and Olson,  W.  T., "Design and Performance of Gas
       Turbine Power Plants".  Princeton University Press, 1960,  L. C.
       Card 58-5027.
  8.    De Kovats, A.  and Desmur, G., "Pumps, Fans and Compressors".
       Translated by R.  S. Eaton, M. A.   Blackie and Sons, Ltd.,  London,
       England (1958).
  9.    Clarke,  J. S. and Jackson, S. R., "General Considerations in the Design
       of Combustion  Chambers for Aircraft and Industrial Gas Turbines".
       Joseph Lucas,  Ltd.,  Burnley, England (1962).

 10.    Federal Register, Vol. 35, No. 136, July 15, 1970 and No.  219, Nov. 10,  1970.
 11.    Strack, W. C.,  "Condensers and Boilers for Steam Powered Cars;
       A Parametric Analysis of Their Size, Weight and  Required Fan Power".
       NASA  TND-5813,  Lewis Research  Center,  Cleveland, Ohio (1970).

-------
                     XI-1
                  Chapter XI
  HYBRID HEAT ENGINE/ELECTRIC SYSTEMS STUDY
                      by
                Joseph Meltzer
Director of Pollution and Resources Programs,
          The Aerospace Corporation
         El Segundo, California, USA

-------
                                    XI-2
      This briefing summarizes the results of a comprehensive broadbased
study aimed at determining the feasibility of using a hybrid heat engine/
electric propulsion system as  a means of reducing exhaust emissions from
street-operated vehicles.  In this hybrid concept,  the source of power is a
combination of heat engine and batteries (in essence, the heat engine
supplies  steady state power and the batteries  supply transient power
demands). The study examined—for several  classes of vehicles--many
types of heat engines, batteries, and other major  components,  as well as
several design configurations.  Following a review of the associated tech-
nologies,  hybrid performance,  exhaust emissions, and major component
requirements were determined. Based on these results, recommendations
can be formulated to  ensure the development of critical powertrain  com-
ponents for an early demonstration of prototype  vehicles.
      In the propulsion of the  hybrid heat engine/electric vehicle, the
ultimate  source of all energy to be expended is the heat engine.  The key to
success in reducing exhaust emissions is good part-load and  full  load
efficiency of  powertrain components, and the  ability to restrict operational
requirements £or the heat  engine to those of supplying road load power and
(in conjunction with a generator) recharging advanced high power/high
energy batteries that supply acceleration power.   With this idea in mind,
the study was tailored to examine six classes  of vehicles:  the 4000-lb family
car, 1700-lb commuter car, low-  and  high-speed  postal/delivery van, and
low- and high-speed intracity bus.  For  each  class of vehicle, five  engines
were included in the powertrain:  spark ignition,  compression ignition,  gas
turbine,  Rankine cycle, and Stirling cycle.  Lead-acid,  nickel-cadmium,
and nickel-zinc batteries were studied for adequacy in supplying acceleration
power to each vehicle.   Also,  a wide range of ac and dc  motors,  generators,
and power conditioning and control systems were  evaluated for performance
efficiency, weight,  simplicity, and cost.

-------
                               XI-3
     Throughout the study, the following ground rules prevailed:

     •     The hybrid vehicle should match the conventional
           automotive vehicle in acceleration,  speed,  gradeability,
           curb weight,  and powertrain weight.

     •     External recharge of the battery should not be required.
           This requirement was simulated in computations by
           requiring that the heat engine-driven generator recharge
           the battery to the original state-of-charge prior to the
           end of a selected emission driving cycle.
     •     The battery is to discharge only when the vehicle is
           undergoing acceleration, not on a smooth grade  or at
           cruise conditions.

     •     The heat engine is to supply steady road load power
           and is not required to undergo  rapid acceleration.

     •     Only design concepts compatible with near term
           (1972-1975) prototype vehicle development are to be
           considered.

     The following set of charts summarizes the  study content,  parameters

examined,  and the results.  Some of the results are highlighted in the text

below for the family and  commuter car.

     •     Only the spark ignition internal combustion engine and
           the gas  turbine engines can be  practically packaged
           into the hybrid heat engine/electric vehicle with the
           performance specified in the study.

     •     All hybrids examined showed marked emission reduc-
           tions  over  current conventional vehicles.

     •     If currently available technology--not including catalytic
           converters--is used, no version of the family car could
           meet  1975  emission standards  (HC = 0.46 grams/mile,
           CO =  4.79  grams/mile and  NO- =0.4 grams/mile).
     •     If advanced technology--including the catalytic converter
           for the internal combustion engine--is used,  all versions
           but the diesel could meet 1975  standards (except for  the
           minor NO2 excess for the spark ignition family car
           version).
     •     Commuter  car emissions are less than one-half of those
           for the family car and with  advanced technology  easily
           meet the 1975 standards.   (The commuter car weighs
           only  1700 Ib and has reduced acceleration and maximum
           cruise speed capabilities.)

-------
                       XI-4
Emissions are approximately 10 and 15 percent lower
for the parallel powertrain configuration as compared
to the series configuration in the family and commuter
cars, respectively.

Study results are based on hot  start data.  Incorporation
of cold start effects would still allow the advanced tech-
nology versions of some hybrid vehicles to meet
1975 standards.

Regenerative braking has essentially no effect on
emissions.
Motor and engine part-load  characteristics and motor
efficiency are extremely important in determining
emissions.

Improved lead-acid batteries could be used in near
term hybrids.  Increased energy density and power
density capabilities are needed.  Battery lifetime and
improved charge acceptance are the most critical areas
requiring improvement.   Nickel-zinc looks promising
for the post-1975 period.

Battery charge acceptance characteristics play an
extremely important role in determining  resultant
vehicle exhaust emissions.

Hybrid fuel consumption is about the same as for the
current conventional car.

For the hybrid using a spark ignition engine, high-
production costs would range from 1.4-1. 6 times
todays car.

-------
                      XI-5
                    PURPOSE  OF STUDY

 TO DETERMINE :
   • FEASIBILITY OF HEAT ENGINE/ELECTRIC HYBRID AUTOMOTIVE VEHICLES
   • POTENTIAL REDUCTION IN AUTOMOTIVE EXHAUST EMISSIONS
   • MOST PROMISING DESIGN CONCEPT IN EACH  VEHICLE CLASS:
      • FULL-SIZE FAMILY CAR
      • SMALL COMMUTER CAR
      • DELIVERY  AND POSTAL VAN
      • CITY BUS
 TO RECOMMEND '
   • TECHNICAL DEVELOPMENT PLAN FOR CRITICAL COMPONENTS
   • TECHNICAL DEVELOPMENT PLAN TO ASSURE PRODUCTION VEHICLE
     IN  1975 -1980 PERIOD
      • SCHEDULE
      • RESOURCES  ALLOCATION
      • MILESTONES (e.g. TEST BEDS, PROTOTYPES    )
                 THE HYBRID  CONCEPT

•  POWER FOR PROPULSION IS SUPPLIED BY TWO SOURCES : HEAT
   ENGINE  AND  BATTERIES

•  POWER FOR ACCELERATION IS SUPPLIED BY THE BATTERIES -
   CRUISE  POWER IS SUPPLIED BY THE HEAT ENGINE

•  HEAT ENGINE SIZE CAN  BE REDUCED

•  HEAT ENGINE OPERATES OVER  RESTRICTED RPM RANGE

•  ENGINE DESIGN AND OPERATION  CAN BE OPTIMIZED TO REDUCE
   EXHAUST EMISSIONS

•  ALLOWS INTERMEDIATE  STEP BETWEEN CURRENT INTERNAL
   COMBUSTION AND PRACTICAL ALL-ELECTRIC VEHICLE OF  1985-1990

-------
                           XI-6
ADVANTAGES  OF OPERATING  HEAT ENGINE IN HYBRID VEHICLE

     •  RAPID  ENGINE ACCELERATION REQUIREMENT  IS REMOVED
        • EXPECT NO STUMBLE  FROM LEAN ENGINE OPERATION
        • EXPECT IMPROVED  EXHAUST EMISSIONS

      • RPM AND LOAD RANGE ARE RESTRICTED
        • EXPECT IMPROVED COMBUSTION AND FUEL  CONSUMPTION
        • EXPECT IMPROVED EXHAUST EMISSIONS
        • REDUCE DESIGN  REQUIREMENTS FOR CATALYTIC CONVERTER
               SCHEMATIC  OF HYBRID CARS
   SERIES CONFIGURATION
   PARALLEL CONFIGURATION
HEAT
ENGINE

L
GENERATOR




CONTROL
SYSTEM
t
BATTERIES



GEARING




MOTOR
	 , WHFFl 
-------
                             XI-7
                     VEHICLE  COMPONENT  ARRAY
     ENGINES
       • I.C. SPARK
       • DIESEL
       • GAS TURBINE
       • RANKINE CYCLE
       • STIRLING CYCLE
     BATTERIES
       • LEAD-ACID
       • NICKEL-CADMIUM
       • NICKEL-ZINC
     MOTORS
       • A.C. INDUCTION
       • D.C. SHUNT WOUND/
        EXTERNALLY EXCITED
       • D.C. COMPOUND
       • D.C. SERIES
       • D.C.BRUSHLESS
                  GENERATORS
                   • DC
                   • A.C. (ALTERNATOR)
                  POWER CONDITIONING AND CONTROL
                   • SILICON CONTROLLED RECTIFIERS
                   • INVERTERS
                   • SOLID STATE INTEGRATED CIRCUITS
                   • CYCLOCONVERTER
                   • RELAYS/SWITCHES
                   • RESISTORS/INDUCTORS
                   HYBRID VEHICLE SPECIFICATIONS
FAMILY
CAR
V MAX (miles/hr) 80
V GRADE AT GRADE
(miles/hr, AT %) 40 AT 12
RANGE (miles) 200
LOADED WEiGHT(lb) 4,000
COMMUTER
CAR
70
33 AT 12
50
1,700
INTRA - CITY
BUS
40
6 AT 20
200
30,000
DELIVERY /POSTAL
VAN
40
8 AT 20
60
7,000
ASSIGNED POWER
TRAIN WEIGHT (Ib)    1,500
ASSIGNED POWER
TRAIN VOLUME (ft
28
           600
16
            6,000
175
                I  ,700
                                          42
ACCELERATION
  EQUAL TO CONTEMPORARY AUTOMOTIVE VEHICLE

-------
                                  XI-8
         DRIVING  CYCLES  FOR EMISSION COMPARISONS
              60
           •&
FAMILY AND   E
COMMUTER    £

           9s
CAR
                                        7.5 miles
INTRA-CITY
BUS
DELIVERY
VAN
              60
           •s.

              »
           £
                       200      400      600     800

                                        TIME , sec




                         0 1  miles


                            1    |    |
                0   10   20  30   40   50

                        TIME, sec
              60
                               0 2 miles
                       20
                               40      60

                                 TIME, sec
                                               80       100
                                                      1000     1200     1400
          DESIGN  DRIVING CYCLE,  4000-lb FAMILY  CAR
               MAXIMUM ACCELERATION

                       - HIGH SPEED CRUISE
                                  - HIGH SPEED CRUISE
                                   FOR RANGE
                                                        12% GRADE
                                              10,330 10,340

                                         TIME, »c
                                                              11,061  11,075

-------
                    XI-9
      POWER REQUIREMENTS  FOR 4000-Ib FAMILY
             CAR AT MAXIMUM ACCELERATION
            ( AT THE POWER CONVERTER OUTPUT )
                         POWER TO ACHIEVE
                         GRAOEABILITY
                         REQUIREMENT
                                      ROILING RESISTANCE PLUS
                                      AERODYNAMIC DRAG
                     40     50    60
                      VELOCITY, mph
SERIES  CONFIGURATION - VARIATION  OF HEAT ENGINE
           POWER  WITH  VEHICLE SPEED
                        LEVEL ROAD
                                POWER
                               DELIVERED
   EXCESS TO BATTERY
   AND/OR ENERGY
   DUMP CIRCUIT-,
   MINIMUM
   ALLOWABLE
   POWER-
POWER REQUIRED FOR
STEADY ROAD LOAD
                   VEHICLE SPEED, mph

-------
                      XI-10
ELECTRICAL CONTROL SCHEMATIC, SERIES CONFIGURATION


HEAT f
ENGINE V
T
I
i
I
L. 	





\_ ALTERNATOR^.
1 RECTIFIER
1 t
i i i
1 ' If
1 L-
| j





BATTERY
VOLTAGE | i




. . J. _,. .
1 "
1
i

- CONTROL SYSTEM
POWER
* LEVEL



T
I FOOT
i PEDAL




* 1
-, T 1 1
If ' '
	 _ J |
VOLTAGE J
, 	 1
RPM IQlIf




V WHEELS
L/






	 MECHANICAL POWER
                                    ^^- ELECTRICAL POWER
                                    	SENSING OR CONTROL
       WEIGHT COMPARISON FOR ELECTRIC GENERATORS
     10-
            i     i  T  i      i     i   i  r
    I10
      I0l	L
                              DC-
                      10              I02
                    CONTINUOUS RATED POWER, kW

-------
                                 XI-11
             WEIGHT  COMPARISON FOR  ELECTRIC MOTORS
          1400

          1200

          1000

         . 800

         J
           600

           400

           200
DC MOTOR + SCR CONTROLLER
            DC MOTOR
                       AC MOTOR + INVERTER
                       +  COOLING SYSTEM
                            FAMILY CAR
                         I  A       I
              0          50         100        150        200
                              CONTINUOUS RATED POWER, hp
                                                250
               HYBRID VEHICLE  PERFORMANCE  EVALUATION
                         DIGITAL COMPUTER  PROGRAM
         CASE INPUT
HFAT ENGINES EMISSIONS
VEHICLE WEIGHT
VEHICLE VELOCITY  HISTORY
BATTERY CAPACITY
BATTERY RECHARGE EFFICIENCY
REGENERATIVE BRAKING EFFICIENCY
TRANSMISSION EFFICIENCY
GEAR TRAIN  EFFICIENCY
GEAR RATIO
MINIMUM GENERATOR CURRENT
              ACCOUNTABILITY
          HEAT ENGINE POWER
          AUXILIARY POWER
          GENERATOR EFFICIENCY
          SYSTEM VOLTAGE
          MOTOR EFFICIENCY
          MOTOR SPEED
          MOTOR TORQUE
          VEHICLE ACCELERATION
          VEHICLE VELOCITY
          TRACTIVE EFFORT
          STEADY ROAD HORSEPOWER
          VEHICLE  KINETIC ENERGY
          ROAD GRADE
          AERODYNAMIC DRAG
          TIRE PRESSURE
          TIRE ROLLING RADIUS
          BATTERY CHARACTERISTICS
          (INCLUDING  MAX 8 MIN
          ALLOWABLE VOLTAGE)
        OUTPUT
VEHICLE EMISSIONS
BATTERY STATE-OF-CHARGE
BATTERY  CURRENT
BATTERY  VOLTAGE
GENERATOR CURRENT
MOTOR CURRENT
ENERGY TO ROAD
ENERGY TO BATTERY
ENERGY DISSIPATED
TIME
DISTANCE  TRAVELLED

-------
                            XI-12
          BATTERY DISCHARGE CHARACTERISTICS
            DURING DHEW URBAN DRIVING CYCLE
                      4000-lb FAMILY CAR
                   GENERATOR OUTPUT =38Amp
                    38 AH LEAD-ACID BATTERY
  IUO
»*  99
V  98
o
   97
   96
                                  I
           200
400
  600     800
ELAPSED TIME, sec
                                         1000
                              1200
1400
          LEAD-ACID BATTERY DEVELOPMENT GOALS
                    4000-lb FAMILY CAR
DHEW EMISSION
DRIVING CYCLE
38 AH BATTERY OPERATING
FOR 1370 sec
7.5 VEHICLE MILES WITH
73 BATTERY CHARGE /
DISCHARGE CYCLES AND
1.34 AH DEPTH OF
DISCHARGE
DESIGN DRIVING CYCLE
38 AH BATTERY DELIVERED
462 Amp


1
                                      HYBRID VEHICLE-FAMILY
                                      CAR DEVELOPMENT GOALS
                                     38 AH BATTERY AT LESS THAN
                                     5% DEPTH OF DISCHARGE TO
                                     DELIVER UP TO 500 Amp

                                     5000 hr OF OPERATION
                                     AND 100,000 VEHICLE
                                     MILES  WITH 975,000
                                     CHARGE / DISCHARGE CYCLES

-------
                           XI-13
   COMPARISON  OF ENERGY/POWER  DENSITY CHARACTERISTICS
         OF LEAD-ACID AND NICKEL-ZINC  BATTERIES
                        WITH DESIGN GOALS
POWER DENSITY,Watt/lb
                   300
                   200
                    100
                            ADVANCED LEAD-ACID
                 BATTERY DESIGN GOALS
                  4000-lb FAMILY CAR
                   1500-Ib AVAIL ABLE
                  POWERTRAIN WEIGHT
                  395-lb BATTERIES
                  10% DEPTH OF DISCHANGE
                                  SU LEAD-ACID
                                               NICKEL-ZINC
                                  10          20
                               ENERGY DENSITY, Walt-hr/lb
                                         30
             CYCLE  LIFE OF LEAD-ACID BATTERIES
         I06
         10-
         10'
                    I    I    I    r
                     GOULD-NATIONAL
                  \  BIPOLAR (1968)
                    \
                      \
                        \
              EAGLE-PICHER
              MILK TRUCK
              SIMULATION
                               N ^-STATE-OF-ART
                                  \
                                           HYBRID
                                           GOALS  —
   SLI
REQUIREMENT
(SAE  TEST)
                        PRESENT
                        CAPABILITY ESB
ESB
PROJECTION
                   20      40      60       80
                     DEPTH OF DISCHARGE, percent
                                   100

-------
                               XI-14
           CHARACTERISTICS  OF SECONDARY BATTERIES
       CHARACTERISTIC                      BATTERY TYPE

RELATIVE COST*
DEMONSTRATED CAPABILITY**
  POWER DENSITY,  W/lb
  ENERGY  DENSITY, W-hr/lb
  CYCLE LIFE AT DEPTH OF DISCHARGE  137,400 at 6.7%  34,000 at 25%  190 at 100%
       * BASED ON ACTIVE MATERIAL COST
      **POWER AND ENERGY DENSITIES NOT DEMONSTRATED SIMULTANEOUSLY
LEAD-ACID
1.0
328
23.3
NICKEL-CADMIUM
II. 1
450
18
NICKEL-ZINC
2.5
180
22
           COMPARISON OF HEAT ENGINE CHARACTERISTICS
                           (89HP ENGINE)

GASOLINE
DIESEL
GAS TURBINE
RANKINE CYCLE
STIRLING CYCLE
WEIGHT
NUMBER
320
756
294
802
1090
VOLUME
FT3
11.4
17.3
9.1
12.7
21.6
SFC
*IBHP-HR
.50
40
.57
.85
42

-------
                          XI-15
        EFFECT OF AVAILABLE  POWERTRAIN  WEIGHT  ON
                    BATTERY  DESIGN  GOALS
                        4000-lb FAMILY  CAR
1000

800

600

400

200
     PEAK DEMAND   93.2  kW
                 DIESEL
.STIRLING
  1000    1400     1800     3200

      AVAILABLE POWERTRAIN WEIGHT, Ib
                   PEAK DEMAND = 0.41 kW-hr  AND
                   BATTERY DISCHARGED  10% OF CAPACITY
£  100


I  80

£  60


£  40

§  20
                                                      DIESEL
                                    STIRLING
               §   0-
               8   1000
           1400     1800    2200

       ft/AILABLE  POWERTRAIN WEIGHT, Ib
           EFFECT OF BATTERY RECHARGE EFFICIENCY
                   ON BATTERY DESIGN GOALS
                         4000-lb FAMILY CAR
              FINAL STATE-OF-CHARGE   INITIAL STATE-OF-CHARGE
               AT IJB =0.7, PEAK ENERGY DENSITY = 10.4 W-hr/lb
                 1.15


                 1.10


 NORMALIZED       I 05
   BATTERY
 PEAK ENERGY
   DENSITY        1-00
  (W-hr/lb)/
(W-hr/lb)vro%  0.95


                0.90


                0.85
       50
60  / 70       80       90

"RECHARGE EFFICIENCY (T/B),%

-------
                       XI-16
      INSTALLED BATTERY REQUIREMENTS AND
        PROJECTED BATTERY CAPABILITIES
                                       T
                             VEHICLE
                             WEIGHT = 4000 Ib
                             NICKEL-ZINC
                             BATTERY
                              INSTALLED
                              BATTERY REQUIREMENTS -,
                           FAMILY CAR
                           S. I. ENGINE
                           SERIES CONFIGURATION
         0       10      20     30      40      50
            MAXIMUM INSTALLED ENERGY DENSITY, W-hr/lb
TYPICAL SPARK IGNITION ENGINE EXHAUST EMISSIONS
             vs AIR/FUEL RATIO (GASOLINE)
     _    CURRENT
        SPARK IGNITION
       ENGINE A/F RANGE
                    14      16
                    AIR/FUEL RATIO
                                                  22

-------
                    XI-17
  SPARK  IGNITION ENGINES (GASOLINE) HYDROCARBON
       EMISSION, STEADY STATE DESIGN LOAD
       I     I   I
                   I
                                         I     I   I   I
                           PROJECTED TECHNOLOGY-LEAN
                                        A/F = I9
                                        A/F=I5 =
                         PROJECTED TECHNOLOGY
                        A/F= 22, CATALYST, RECIRCULATION
                             i   i  i
                  10                I02
                  DESIGN BRAKE HORSEPOWER
10'
  HEAT ENGINE EXHAUST EMISSIONS, HYDROCARBON -
STEADY STATE DESIGN LOAD, LARGE ENGINES (>50hp)
I.U


0.8

fo.6
O-
CD
1 —
§.
^0.4

0.2
Q
i i i i i
i— I i 	 1 STATE-OF-THE

—






—
-
—
i







2!
u_
—

-------
                            XI-18
HYBRID  (HEAT ENGINE/ELECTRIC) VEHICLE  EMISSIONS SUMMARY
            4000-Ib FAMILY  CAR/DHEW  CYCLE*
                 CARBON MONOXIDE EMISSIONS
   8.0 r
   5.0
 „ 4.0
 8
   2.0
    1.0
>
-

-
-
-


£
1 — .

*AIR CONDITIONER NOT ON. HEAT ENGINE OPERATED
CONTINUOUSLY AT 18.4 Bhp (20.8% OF DESIGN
POWER LEVEL OF 88.8 Bhp )
\- 1975 STANDARD
Q CURRENT TECHNOLOGY
Sk
-


-------
                            XI-19
HYBRID  (HEAT ENGINE/ELECTRIC) VEHICLE EMISSIONS SUMMARY
            4000-lb FAMILY CAR/DHEW CYCLE*
                   HYDROCARBON EMISSIONS
     1.15
     0.8
  2  0.6
  e
     0.4
     0.2
7
I,
                           I	1 CURRENT
                           1	' TECHNOLOGY

                                PROJECTED
                                TECHNOLOGY
       * AIR CONDITIONER NOT ON. HEAT ENGINE OPERATED
        CONTINUOUSLY AT 18.4 Bhp(20.8% OF DESIGN
        POWER LEVEL OF 88.8 Bhp )
                                            1975 STANDARD
Ik
                                            rh
             S.I. ENGINE
        DIESEL   GAS TURBINE   RANKINE
                   STIRLING
              VEHICLE EMISSION COMPARISON
      CONVENTIONAL OPERATION vs HYBRID OPERATION
                  SPARK-IGNITION ENGINE
50
40
30
EMISSION LEVEL,
grams/mile
20
10
1970
CONVENTIONAL
s S.I. ENGINE
- 8
8
7— MOHF HHF
1 	 ' CYCL
* . •
CONVENTIONAL
S.I. ENGINE
(A/F=I5-I6) + RECIRC.
CONVENTIONAL
S 1. ENGINE LEAN
(VARIABLE A/F) OPERATION,
' 3 ' NO RECIRC. miurFn
o IA/F-IQ1 ADVANCED
n ~ ' .''3| - TECHNOLOGY
° PLUS
o^o ^ ^A/F= 22+ CAT.+ RECIRC.,
W
_§
CONVENTIONAL HYBRID VEHICLE /DHEW CYCLE
VEHICLE (4000-lb FAMILY CAR)

-------
                         XI-20
COMPARATIVE EMISSION LEVELS OF  THE FAMILY AND COMMUTER CAR
120
110
100


80
0
5 70
z
2 60
CO
£ 50
&40
i —
cc
if 20
10
—



__
u
~ •
•
1
	 •
~L
r
P

^
^
^
%
//
'/A
I

1975 STANDARD
—

1975 STANDARDS
fa HC -





r
%
%

///
I
'//
'//
//
'//
CO -
N02-
046 gm/mi
4.7 gm/mi
0.4 gm/mi


	

ADVANCED TECHNOLOGY ~~
WRALLEL CONFIGURATION

t
S.I. GAS TURBINE S.I

Sj
i :
-
GAS TURBINE
«^^y
                 FAMILY CAR
                                     COMMUTER CAR
 EFFECT OF BATTERY RECHARGE EFFICIENCY ON N02 EMISSIONS

      FAMILY CAR-SERIES CONFIGURATION-PROJECTED TECHNOLOGY
                                                STIRLING
                                                RANKINE
                                                1C. ENGINE
                                                S.I. ENGINE
              50      60     70     80

                   RECHARGE EFFICIENCY!^),'
90
      100

-------
                              JXI-21
            EFFECT OF VEHICLE WEIGHT ON N02  EMISSIONS
    FAMILY CAR/DHEW CYCLE-PARALLEL CONFIGURATION-CURRENT TECHNOLOGY
          1.31	r
          1.2
§1
Q "-- in
UJ —. '-U
      2*
          0.9
                                                    -GAS TURBINE
                                                     I.C. ENGINE
                                                     RANKINE
                                                     STIRLING
                                                     'DIESEL
         0.8
          4000    4100   4200    4300    4400    4500    4600    4700    4800
                                 VEHICLE WEIGHT, Ib
EFFECT OF BATTERY CAPACITY AND TYPE ON HC, CO, AND N02  EMISSIONS
             FAMILY CAR/DHEW CYCLE-SERIES CONFIGURATION
e. u

ATIO,
5 STANDARDS
(T>
D EMISSION F
/(gm/milelij-
D —
D ro
kl y
^ -^
Z §. 04



°c
1 1 1 1 1
-
r \^
- A^_

1975 STANDARDS
HC -0.46gm/mile
_ CO -4.7 gm/mile

-
1 , 1
10 20 30
1 ' ' 1 ' 1 '
S.I. ENGINE/PROJECTED TECHNOLOGY
r BASELINE BATTERY CAPACITY USED
FOR FAMILY CAR EMISSION
CALCULATIONS
I
IIC


i 	 Ni-Zn BATTERY
0 Ni-Cd BATTERY


Y 1 , ,1,1,
40 50 60 70 8C
                            BATTERY CAPACITY, Amp-hr

-------
                              XI-22
   EFFECT OF DRIVE  MOTOR EFFICIENCY ON N02 EMISSIONS
FAMILY CAR/DHEW CYCLE-PARALLEL CONFIGURATION-CURRENT TECHNOLOGY
 ,„-??
       1.40
       120
 ¥  I.  1.10
    "  1.00
      0.90
      0.80
            DIESEL
                  /STIRLING - DIESEL
                   RANKINE
                   I.C. ENGINE
             GAS TURBINE
         40    50    60    70    80
                 MOTOR  EFFICIENCY (ijj,
                                     90
                                           100
            HYBRID VEHICLE  FUEL CONSUMPTION AND
                 PROJECTED PRODUCTION   COSTS
                          FUEL  CONSUMPTION
               VEHICLE

            COMMUTER CAR
            FAMILY CAR

            LOW SPEED VAN
            HIGH SPEED VAN

            LOW SPEED BUS
            HIGH SPEED BUS
SERIES CONFIG.
   (mi /go I)
     26
     I I

    3.75
     4
    1.25
    I 50
PARALLEL  CONFIG
   (ml/pal)
     30.5
     12.5
         HIGH-PRODUCTION COSTS COMPARED TO CONVENTIONAL CAR
                   VEHICLE
            CURRENT CONVENTIONAL CAR

            HYBRID CAR
              SPARK IGNITION
              DIESEL
              GAS TURBINE
              RANKINE
              STIRLING
                 RELATIVE COSTS
                       I
                    1.4 -1.6
                    1.5 -1.7
                      -1.6
                       2+
                      2.25+

-------
                           XI-23
                            SUMMARY

•  HYBRID HEAT ENGINE/ELECTRIC FAMILY  CAR

    •  SHOWS MARKED REDUCTION IN  EMISSIONS OVER TODAY'S CONVENTIONAL
      HEAT ENGINE-DRIVEN AUTOMOBILE

    •  FOR NEAR-TERM DEVELOPMENT, THE I.C. SPARK IGNITION ENGINES
      OFFER MEANS FOR MEETING 1975 STANDARDS  BY PERMITTING
      LEAN OPERATION

    •  FOR THE FUTURE, THE GAS TURBINE CAN EXCEED 1975  STANDARDS
      FOR ALL EMISSIONS.

    •  FOR THE FUTURE,  RANKINE CYCLE  AND STIRLING CYCLE CAN FAR
      EXCEED 1975 STANDARDS, BUT WEIGHT AND VOLUME CONSIDERATIONS
      MAY LIMIT THEIR USE.  EXTENSIVE DEVELOPMENT ACTIVITY IS REQUIRED

    •  SIGNIFICANT IMPROVEMENTS IN  ELECTRIC MOTOR PERFORMANCE
      APPEAR  TO BE READILY ACHIEVABLE

    •  LEAD-ACID BATTERY TECHNOLOGY IS  AVAILABLE (BUT NOT BEING
      PRODUCED)  WHICH CAN SATISFY HYBRID NEEDS FOR POWER AND ENERGY
      DENSITY REQUIREMENTS BUT MARKED IMPROVEMENT IS NEEDED IN CYCLE
      LIFE AND IN  CHARGE ACCEPTANCE CHARACTERISTICS

•  HYBRID HEAT ENGINE/ELECTRIC FAMILY CAR
    •  PARALLEL vs SERIES POWERTRAIN CONFIGURATION
       • PARALLEL HAS  10-15% LOWER EMISSION
    •  THE PARALLEL CONFIGURATION IS A SLIGHTLY MORE COMPLEX SYSTEM
      WITH LESS DESIGN  FLEXIBILITY
    •  BATTERY RECHARGE EFFICIENCY
       •  SMALL EFFECT ON EMISSIONS SHOWN OVER  A REALISTIC RANGE
          OF EFFICIENCY
    •  REGENERATIVE BREAKING EFECTS
       •  SLIGHTLY DECREASES DEPTH  OF DISCHARGE
       •  NEGLIGIBLE EFFECTS ON VEHICLE  EMISSIONS
    •  TOTAL VEHICLE WEIGHT
       •  BATTERY DESIGN GOALS ARE  VERY SENSITIVE TO POWER SYSTEM
          WEIGHT ALLOCATION
    •  COLD START EFFECTS
       •  FOR HC AND  CO INCREASE  FROM  15 TO 50% ( /VERAGES ABOUT 30%)
       • FOR NOX SLIGHT DECREASES

•  HYBRID HEAT ENGINE/ELECTRIC COMMUTER CAR

   • LOW-WEIGHT  REDUCED-PERFORMANCE  CAR CAN  REDUCE EMISSIONS
     BY A FACTOR  OF 2.5 COMPARED TO THE HYBRID  FAMILY CAR

•  HYBRID HEAT ENGINE /ELECTRIC BUS AND VAN

   •  DIESEL AND TURBINE ENGINES LOOK  ATTRACTIVE FOR BUS AND
     VAN APPLICATIONS WITH MINIMUM DEVELOPMENT EXPENDITURES
   •  POTENTIAL IMPROVEMENTS IN HYBRID BUSES AND VANS  COULD
     LEAD TO SIGNIFICANT REDUCTION IN  N02 AND CO
   •  VEHICLE EMISSIONS DATA ARE REQUIRED OVER REALISTIC
     DRIVING CYCLES ON CURRENT BUSES AND VANS BEFORE
     COMPARISON OF THE HYBRID MODE CAN BE ASSESSED
   •  BATTERY POWER DENSITY AND ENERGY DENSITY SHOULD  EASILY
     BE MET FOR THE HYBRID BUS APPLICATION

-------
                   XII-1
                Chapter XII
 ADVANCED TECHNIQUES IN ELECTRICAL VEHICLES
               Bohers/Ducrot
Engineer, Research and Development Division
Citroen Automobile Company/Electricity Section
              Nanterre, France
       Translated for EPA by SCITRAN
      (Scientific Translation Service)
      Santa Barbara, California, USA

-------
                              XII-2
     Ladies and gentlemen, my name is Pol Ducrot.   I am an
engineer with the Research and Development Division of the
Citroen Automobile Company — Electricity Section.

     Let me, in turn, thank the Philips Company for inviting me
here.  Thanks to them, I shall report to you on the advanced
studies we are carrying out in the field of electric vehicles.

     In many fields, from the old drive to the recent SM, Citroen
has been in the vanguard— for  example,  in  the  area  of vehicle-
ground communications or oleopneumatic suspension for vehicles
of a more modest type.

     Thus, it is not surprising that the Company is engaged in
significant research in the area of electrical drive particularly,

     One of the systems we are studying will be described here.

     This study is drawn from a research program carried out in
close cooperation with the Compagnie Prancaise de Raffinage and
the Battelle Institute of Geneva.

     As we all know, an electric drive system must include a
source of electrical energy, together with a unit  to convert 'it
into mechanical energy in a form adapted to the flexibility
requirements which characterize drive.

-------
                              XII-3
     Unfortunately, nobody has yet found an ideal energy source:
quiet,  nonpolluting, inexpensive, economical to run,  very light,
and...  so simple to design in sketches of future vehicles.

     One must look for compromises and optimizations  to reduce
the often unpleasant disparity between the dream and  reality.

     This is the purpose of the first part of the presentation,
dealing with a cell supplied with air and liquid or gaseous
hydrocarbons.

                  Optimization of the Fuel Cell

     This is a solid electrolyte cell, the principle  of which
is well-known.  We know that the working temperature  may be
selected between 800 and 1000° C, which enables us to avoid  usinj
precious catalysts like platinum and to use such inexpensive
fuels as hydrocarbons, since conversion becomes quite simple.

     The specific performance obtained in this way is quite
interesting.  The inconvenient aspects of this high temperature
should also be mentioned.

     The choice of materials is more limited, and the technology
is more difficult.  The electrolyte thickness should  be small,
both to reduce the internal voltage drop in the electrolyte  and
to minimize the mass to be raised to a high temperature, and
thus the energy required for the proper temperature.

     In parallel with the laboratory tests, an optimization
calculation by computer was carried out beginning with the
different mathematical models to take into account the effects
of the  numerous parameters of the cell, which include, for
instance :

-------
                               XII-4
     — arrangement of the converter (exterior or integrated)
     — cell dimensions
     — maximal and minimal electrolyte temperatures
     — activation polarization
     — average partial pressure of oxygen in the cathode
        compartment
     — blast-engine output
     — heat flux released by parasitic combustion

                             Photo 1

     Each element has a current-voltage characteristic depending
on its location in the cell.

     Here are the typical characteristics of an element:

Rp .  JE designates the voltage drop in the electrolyte related to
its ionic conductibility.

              AEA designates  the activation polarizations
              AEG designates  the concentration polarizations

                             Photo 2
     The design shows here the cell with an external converter to
be used in models 1 and 2.

                             Photo 3

     Here in Model I, the current drainage occurs perpendicularly
to the active faces.   We shall term it "cells with transverse
current".

-------
                               XII-5
                             Photo
     In Model II,  current drainage occurs parallel to the active
faces and the elements are arranged in series by a suitable
incorporated arrangement.  We shall call it:   "cell with internal
series arrangement".

                             Photo 5.

     Here we see the  beginning of Fortran writing of the program
relative to the latter model supplied by computer.

                             Photo 6
     Here,  by way of example,  is one of the curve networks
supplied by calculation of Model I with a transverse  current.
It gives the net output of the cell core as a function of the
power-volume ratio for 3 values of the electrolyte thickness.
These curves correspond to a cell center which is completely
heat-insulated.   The preheating of the air is done without  loss
by heat  recovery.

                             Photo 7

     Here is the same network  for Model II with internal series
arrangement.

                             Photo 8

     Here is a universal optimization curve.   The curve in  a heavy
line  is  the envelope of the different special curves  represented
by the dotted lines.

-------
                               XII-6
     This curve shows the compromise between the power-volume
ratio and the net output of the cell core,  taking into account
the blast-engine.

                     Electrical Drive System

     After this first series of photos,  we  shall look at  a
special system of electrical drive by an asynchronous, three-
phase motor with casing, supplied by a continuous current source
and the result of bench tests carried out in 1968,  beginning with
a rectangular or sinusoidal modulation drive.

     There is no need to emphasize the strength and low price
which characterize the asynchronous motor with casing which can
be produced in a large range of velocities  and outputs.

     On the other hand, its adaption to  drive  starting with a
fuel cell requires an "undulator" designed  especially for that
use .

                             Photo 9

     This is a schematic diagram of the  universal control.   We
can single out:

     —— the undulator bar control.
     —• the mechanisms for guiding the bar  control.   They will
        receive, regardless of the type  of  control  used,  orders
        for voltage V and frequency fs,  and will transform these
        signals into properly implemented signals in time to
        obtain the desired voltage amplitude and frequency in the
        stator.  These signals will drive the  bar control circuits

-------
                              XII-7
    — operational circuits for supplying V and. fs references.
       These circuits send out the two parameters:  the flux and
       current for the motor.

    We then have to select an undulator pattern suited for this
application.  Two things are certain:

    — It must have "two" return diodes, sometimes called "free
       wheel diodes."
    — It cannot be of the "series" type where the commutation
       capacity is placed in series and which does not have the
       desired operational flexibility.

    Even limiting ourselves to extinction by condenser discharge,
several extinction circuits can be imagined.

                           Photo 10
    Since some order was needed, they were classified according
to the following criteria:

    — the extinction class, termed A or B depending on whether
       the commutation inductance is or is not traversed by the
       motor current.   This makes the extinction circuit either
       heavily or slightly dependent on the main current.
    — extinction systems, which may be either direct or al-
       ternating.

    The table you see  illustrates this classification.

    At the top, the continuous direct current extinction systems
    No.  1:  universal.  All thyristors are cut off simultaneously
    No.  2 and 3'   individual or semi-individual.

-------
                               XII-8
     At the bottom:  "alternating current" extinction systems.
     No. 4 and 5:  complementary extinction systems — very
                   economical — where lighting one group causes
                   the extinction of the other group.
     No. 6 and 7:  individual or semi-individual extinction
                   systems, which are much more complex.

                   Semi-individual alternating current extinction
                   of class B was chosen for our tests, because
                   it had the required flexibility for sinusoidal
                   modulation allowing smaller harmonic content
                   in the current wave.

                            Photo 11
     This picture is taken from our sinusoidal control patent
and shows the control ensemble schematically.

     To simplify the design, the thyristor protection circuits
in dV and dl ,       ,  ,      ,
   —     — have not been shown.
   dt     dt

     We see three control groups U, V and W, with for each:

     — Tl and T2 main thyristors
     — free wheel diodes Dl and D2
     — extinction thyristors SI and S2
     and an inductance/extinction capacity circuit.

                            Photo 12

     Here are the output voltages of the undulator with rectangular
modulation.  The form factor T/TM is 50% here.

-------
                               XII-9


    UR - Us shows the  composite  voltage  between phases.

                            Photo 13

    This shows the  law governing the  change  of the  cut-off
frequency as a function of  output frequency.

    The cut-off  frequency  is  a whole  multiple  of 6  fs.   When
starting, fs = 10 Hz, K = 16,  fM  = 960 Hz.

    We see that  K successively takes  on  the  values  16,  8J  4,  2,
1,  so that fM never  exceeds  2  Khz.

                            Photo 14
    Here we see the theoretical  behavior  of the  current  in  a
main thyristor with sinusoidal  modulations  at cos    =  1

                           Photo 15
    This photo is also taken  from the  same  Citroen  patent.  To
obtain the correct cut—off  sequence,  a  special  sinusoidal
modulation system was perfected.

    Here is the principle  employed:

    Above, diagram A shows  a  sinusoidal  voltage wave  1 with
frequency fo.

    At Instants tl, t2, t3, following  periodically  at intervals
of T  = 1/fe, we measure the instantaneous value of  the sinusoidal
voltage with respect to the  reference level  3,  a voltage shown by
iengths al, a2 , a3 . . .

-------
                              XII-10
     Thus, a series of asymmetrical triangular signals  is
generated as shown in diagram B.

     We see that their steep fronts bl,  b2,  b3 occur at instants
tl, t2, t3-

     — their heights are proportional to the instantaneous  values
        al, a.2, a3 of sinusoidal  voltage 1 measured at  these
        sampling times tl, t2, t3-
     — the oblique fronts cl, c2,  c3 all have the same given
        slope p.
     — in diagram C the triangular signals  are transformed  into
        rectangular impulses, each  with a width equal to the
        duration between the rigid  front of  the corresponding
        triangular signal and the moment when the oblique  front
        reaches the reference level visible  in diagram  B.

     Thus, impulse II has a width equal to the period dl.

     So, we see that the impulses I are synchronous with the
sampling with frequency fe and are  modulated in width in accor-
dance with the instantaneous value  of sinusoidal voltage 1 at
those times, since the gradient provides a voltage-time linear
conversion.

     If the reference level is identified with the zero level of
the triangular signals, the width d of the impulses is  directly
proportional to the instantaneous value of sinusoidal voltage
measured with respect to level 3.

     These impulses I are used to control turning on and off the
thyristors inserted in the group  commuting an undulator phase..

-------
                              XII-11
     One of these thyristors is  kept  on as  long as  impulse  I
lasts,  and it is turned off by forced extinction during the
interval T  - d.
          e

     Clearly, we need only sample the same  sinusoidal  voltage  1
at other times offset by 2£ to obtain the impulses  needed to
control two other groups.

     A  closer examination of the system would show  that a second
order harmonic would appear with an amplitude proportional  to
frequency fo; however, a simple  and effective correction was
introduced by modulating the reference level.

                            Photo 16

     Here we see a simplified diagram of the sampling  and
amplitude-time conversion apparatus.

     — For each impulse f, the sampling condenser C  is charged
        very quickly by the push-button p.
     — It is discharged permanently  through the constant current
        source lo.
     —- The comparator enables us to  obtain a logic signal
        defining the commutation sequence for an undulator  phase.

     One should note that  this signal is the image  of  the phase
output  voltage.

                            Photo 17

     Here is the drive curve which shows the behavior  of the motor
coupling as a function of velocity.

-------
                              XII-12
     For this motor,  a 3.6 Mkgs starting coupling was used for
a nominal output of 10 KW with a maximum velocity of 12,000 rpm.

                            Photo 18

     Here is the output measured for rectangular and sinusoidal
control.

     We see that in the average velocity range the rectangular
command system yields a greater output for the laws of cut-off
frequency used.  This can be explained by the unequal losses in
the extinction circuits.

     We have also noted increased fatigue in the sinusoidal
control extinction circuits —- including the extinction con-
densers — when we attempt to obtain a starting coupling.

     For the sake of completeness,  I should point out the  slightly
greater complexity of the sinusoidal control electronics, with
little effect on price.

     On the other hand, I should stress the coupling modulations
derived from pulsating parasitic couplings with rectangular
modulation.

     In starting, their relative amplitude is on the order of
20%3  and the frequency is 60 Hz.

     Consequently, we must anticipate a transmission to avoid
mechanical fatigue, not to cause resonances and to retain  the
qualities of comfort  and silence so highly valued in electric
vehicles.

-------
                              XII-13
    I shall  conclude this  presentation  with  one  small  detail:

    The continuous  current  source  used  In  these  tests  was  not
from a fuel cell.

    Thank you, Mr.  President,  I  am prepared  to answer  questions,
to the extent that they  are  not too compromising  — I think you
will understand.
     Company address:

          Societe  des  Automobiles  Citroen
          qua!  A.  Citroen
          Paris  XV,  France

     Research  Division:

          Pol  Ducrot
          Service  Recherches
          Societe  des  Automobiles  Citroen
          1 rue  P. Millet
          92 Nanterre, Prance

-------
                                P lane h e
 Fig. 16. Electiic
characteristics of an
       element.      ,-
                                         conversion  '
                                         compartment
                                               anode
Figure ^.  Principle of
:  the heart of a cell  n:
with external conversion
                                                     joxes
                                                            conducting band
                                                                       fuel distribution
                                                                           channels
                                                                                     inter-plaaue
                                                                                       connectors
                                                                              membrane
                                active
                                surface
                                 porous
                                support

                                 connectors
Fig.  1.  anode box model 1
                                                                                             braces
                                           fuel  distribution
                                          ^channels   [
                                              connecting  band
                                             [connected with
                                                the metal parts
                                                —^ of the box

                                              connectors  for
                                              internal series
                                              connection
                                                                                                   electrically insulated
                                                                                                   from porous support
           inter-plaque connectors
                        **
Fig. 10.  anode box, Module II
      Ph.'I
                                     ph.a
                 -^                   Ph,3
                 'Appendix 5 FORTRAN listing of program II

     Legend  for  Ph.  6  and  Ph.  7

1 — activation polarization: 0.10 V
2 — electrolyte temperature: 800 °C
3 — cathode thickness  (In.O ): 50u
                         Z, J
                                                 Ph.
 cell with transferse current
      exterior converter
             cell with internal series
                     connection
                  exterior  converter!
                yield
                                                                            calculated points^, ^7
                                                                            closest to optimum JF(!)J
                                                                            maximum current    *~—-
                                                                            through electrolyte '^.
                                                                            combustion yield  at   -":
                                                                            the  center of core
                                                                                                            power
           Ph. &
                    Ph ,   7
                            Ph.  <3
                                                                                                            volume ratio

-------
                          ,  7        ',
                        ondulator   V
         supply control^.,  j   motor
                                          In
 traction_
 braking:
conversion
                       =p
                    transducers
                                 o o
                                 3 -H
                                 G 4->
                                 •H O
                                      O JJ
                                          class A
total
                                           semi-
                                          indiv.
                                                  In class B
                                                      no known example  FIG.7
                                                      no known example
                                       -Q complementary  diagonal
                                                      impossible
                Ph.  6
Fig. 2.3  Change in cutoff
frequency f  as a function of
      output frequency  fq.
               •Values  of k  16,  8,  4,  2,  1
                                                                                 Ph.1.
                                                                            Ph.
                             -J
                Ph. -13
                                1968
                           -{current
                             : voltage
                           -^- couple

                              -^velocity
                                                                               output
                                                                               voltages
                                                                               at the
                                                                               ondulator
                                                                 for rectangular
                                                    modulation with Fm =  12 Fs and T/T =
                                                  TT"'  n'Lr,^ n  voltage delivered by  50%
                                                  U , U_ and U                     '
                                                   R   s      r  ondulator phases'
                                                  U^, Ugcomposite  voltage"  x"l

                                                  U neutral motor potential
                                                   K
                                                  U -U .U    neutral-phase voltage
                                                — R  n  phR	   ;
                                    Fig. 1.7. theoretical
                                     curve  for current in a main,
                                     thyristor for sinusoidal
                                     modulation  for a charge
                                    with
                                                                                           FIG. t
                                                                              Ph.  -t 5"
                                                            couple-yield (m kg) /(%)[:-|:pr July 1968L
                                                                ^pittBi
                                         Ph.
           Fig.  1.2.   Traction curve
   numerical values;
                                                                                                                    trigger
                                                                                     Icrn
                                                                               constant
                                                                                current
                                                                               source
                                                                  Fig. 12.  Simplified
                                                                  diagram of a sampling
                                                                  and amplitude         |
                                                                  time conversion device
                                           •^j-, •- i--fyield_with
                                           'ield-half c-^~^e jsinusoidal  ph A 6>
                                                •. :.•;".''~T "• r-|Ll_: modulation
                                                   /;iFig.  3.1.  total yields
                                                                          ..TilL; !X; r.:_itotal chargeJIii-'
                                                                           li-l-i i-!-'. ~> ^-|:-;-:-i-f-i-f-H-rl couple delopped
                                                                          —r T  ] i   . : 1^^, -  - \-   ,    I    c       r *•
                                                                                                                                      fxl
                                                                                                                                      H
                                                                                                                                      M
                                                                                                                                      I
                                                                                                                                      M
                                                                                                                                      Ln
                  ph  -ir

-------
                       XIII-1
                    Chapter XIII
RESEARCH AND DEVELOPMENT ON A LITHIUM-SULFUR BATTERY
                         by
                  Elton J. Cairns,
      Section Head, Energy Conversion Section,
            Argonne National Laboratories
               Argonne, Illinois, USA

-------
                                   XIII-2
Introduction




The Air Pollution Control Office of the Environmental Protection  Agency




(EPA) has initiated a broad program for the development of  low-emissions




vehicles.  A number of alternatives to the internal combustion engine




can be considered to be candidates for use as power plants  in such




vehicles.  Some alternatives are listed in Slide 1 (copies  of the




slides are attached).  The Argonne National Laboratory, under the




sponsorship of EPA, is pursuing the development of lithium/sulfur




batteries for all-electric vehicles.  Some of the goals that ANL has set




for itself (in keeping with the EPA goals) are shown in Slide 2.  The goal




of 220 W-hr/kg cannot be met by any conventional battery, hence the interest




in the lithium/sulfur system, which uses a liquid lithium anode, a liquid




sulfur cathode, and a molten salt electrolyte containing lithium halides,




and operates at 375°C.







Though this program is still in the laboratory stage, some  interesting




results have been obtained which are worthy of review.  The manner in




which the lithium/sulfur cell operates during discharge is  indicated in




Slide 3.   It is particularly important to provide for the removal of




the product la^S from the reaction site, and to supply more sulfur and




electrons for reaction, without losing any material to the  surroundings„




A major portion of our effort is centered around this process.







Experimental Results




An example of a small lithium/sulfur laboratory cell is shown in Slide 4.




Voltage-current density and voltage-capacity density curves for such a

-------
                        XIII-3




SOME ALTERNATIVES TO THE INTERNAL COMBUSTION ENGINE


     1.  BRAYTON CYCLE   (GAS TURBINE)

     2.  RANKINE CYCLE   (STEAM ENGINE)

     3.  STIRLING CYCLE

     4.  HYBIRD:  HEAT ENGINE PLUS BATTERIES

     5.  ALL-ELECTRIC:  SECONDARY BATTERIES
                    EPA PROGRAM


         HIGH SPECIFIC ENERGY Li/S BATTERY
             FOR ELECTRIC AUTOMOBILES


           GOALS:       220 W-hr/kg

                        220 W/kg

                        1000 CYCLES

                        $10/kWhr


          LOW COST, HIGH SPECIFIC ENERGY

-------
ANODE-FELTMETAL
  CONTAINING Li
            CATHODE-FELTMETAL
               CONTAINING S

ELECTROLYTE: e.g. LiBr-RbSr
                   CELL REACTIONS
         CELL:  Li/Lit ELECTROLYTE )/S (+ Li)
         ANODE: Li0-*- Li+4- e~
         CATHODE: 2Li+ +2e" + S°-»Li2$
         OVERALL: 2Li°+s°->-Li2s

-------
                 XIII-5
                             ANODE LEADS
           CATHODE LEAD
                                  Nb PLUNGER
                               Cu GASKET
                              Nb CATHODE HOUSING
                                  CATHODE CURRENT
                                     COLLECTOR
                                     CONTAINING
                                      SULFUR
                                   S.S. ANODE CUP
                                  ALUMINA CRUCIBLE
ELECTROLYTE
  LiBr-RbBr
 ANODE FELTMETAL
CONTAINING LITHIUM

-------
                                   XIII-6
cell are shown in Slides 5 and 6.  The capacity density of Slide 6 is


too small for the goals of Slide 2.  Therefore, efforts have been made to


improve the capacity density by modifying the structure of the cathode


current collector, as shown by some examples in Slide 7.  The voltage-


capacity density curves for constant-current operation of lithium/


sulfur cells with a comb, laminated, and enclosed laminated structures


are shown in Slides 8, 9, and 10, respectively.  The laminated cathode


has yielded the highest capacity densities,  and the enclosed laminated


structure has shown the best cycle life.  Another long-lived cathode is


the so-called reservoir/cathode, which is comprised of a disk cathode


with a space above it filled with sulfur, and an electrolyte-wetted layer


of porous material below the disk current collector, to prevent the escape


of sulfur.



Some of the important performance parameters for various cathodes are


summarized in the table of Slide 11.  The 220 W-hr/kg goal of Slide 2

                                o
corresponds to about 1.2 A-hr/cm , or 0.4 A-hr/gm.  The cycle lives of


the enclosed laminated and reservoir cathodes are within a factor of 2-3


of the goal.  It can be seen that some further improvements in the cathodes


are necessary before serious scale-up and engineering work can be carried


out.



In addition to the experiments with laboratory cells, several other areas


of investigation are being pursued, as indicated in Slide 12.  The phase


equilibrium investigations have as their objective the identification of


electrolytes and additives to sulfur which minimize the solubility of sulfur-

-------
CURRENT COLLECTORS
STAINLESS STEEL FELT
  POROSITY PORE SIZE
 O   80%     240/A
 D   80%      29 M
 A   40%      l9i
2.5
    Li/LiF-LiCI- Lil/Li in S
ANODE AREA = 2.6cm2
CATHODE AREA =0.7 cm2
INTERELECTRODE DISTANCE = 0.3cm
CELL TEMPERATURE » 375°C
SHORT-TIME DATA
                                                    A
                                                        D
                   21012345
                    CURRENT DENSITY, A/cm2
                              8
                             0

-------
  3.0
                                Li/Li Br- Rb Br/Li in S
  2.5
LJ
CD
<
h-
o 1.5
        \
                          CATHODE AREA
                          ANODE  AREA
                          INTERELECTRODE
                                DISTANCE
                                  = 0.7 cm2
                                  = 2.6cm2

                                  =  I .Ocm
LU
O
.0
  0.5
CURRENT COLLECTOR
POROSITY
PORE  SIZE
TEMPERATURE
CURRENT DENSITY
 S.S FELT
= 80%
= 29/im
= 375°C
= 0.33 A/cm2
   0
     0
           0.05       O.I         0.15        0.20
               CAPACITY DENSITY, A-hr/cm2
                                               0.25

-------
   DISK
      COMB
LAMINATED
ENCLOSED LAMINATES
     CATHODE  CURRENT COLLECTOR STRUCTURES

-------
3.0
      0.26 A/cm2
      RECHARGE \ 2
                               T
                 Li/ Li Br- RbBr/Li in S
                   0.53 A/cm2 RECHARGE
0.5
0.0
   0
0.53 A/cm2
DISCHARGE

   i	
ANODE  AREA  2.92 cm2
CATHODE AREA  1.89 cm2
INTERELECTRODE DISTANCE  I cm
TEMPERATURE 395°C
CATHODE CURRENT COLLECTOR
  GRAPHITE, 1.4/APORE SIZE, 63% POROSITY
THEORETICAL CAPACITY 1.45 A-hr/cm2
	I	I	
         O.I
     0.2
0.3
0.4
                     CAPACITY  DENSITY, A-hr/cm'
   0
                10                  20
     PERCENT OF THEORETICAL CAPACITY  DENSITY
                                                                     I
                                                                     M
                                                                     O
                                                                  ON
                                       30

-------
3.0
2.5
     0.45 A/cm2 RECHARGE I
                 Li/LiBr -RbBr/Li in S
                    ANODE AREA 2.7cm2
                    CATHODE AREA 0.96 cm2
                                   NTERELECTRODE  DISTANCE
                                  TEMPERATURE 390°C
                                  THEORETICAL  CAPACITY
                                     DENSITY  2.06  A-hr/cm2
   0.72 A/cm2
  RECHARGE  2
       04 A/cm2  DISCHARGE 3
          0.31 A/cm
     DISCHARGE
                                                              0.52 A/cm2
                                                             DISCHARGE I
        CATHODE CURRENT COLLECTOR
           4 SULFUR ELEMENTS 1.6mm THICK
             80%  POROSITY  30yu- PORE SIZE
           5 ELECTROLYTE  ELEMENTS 0.45mm  THICK
             83%  POROSITY  25/z PORE SIZE
                                                   I
   0
0.
0.2          0.3          0.4
 CAPACITY DENSITY, A-hr/cm2
0.5
0.6
                       I
                                           I
   0
             10                      20
      PERCENT OF THEORETICAL  CAPACITY  DENSITY

-------
UJ
o
>
o
              I     I    I
                                                                0.18 A/cm CHARGE - 34
                    0.2 A/cm" CHARGE -17
                                                              0.2 A/cm' CHARGE-2
                                                      0.2 A/cm2 DISCHARGE-2
    .0
   0.5f
    0
Li/LiCI-Lil-KI /Li in S

  ANODE AREA   2.5 cm
  CATHODE AREA  2.53 A/cm
  INTERELECTRODE DISTANCE I cm
  TEMPERATURE  380°C
  CATHODE CURRENT  COLLECTOR
     3 GRAPHITE ELEMENTS
     63 %  POROSITY  1.4^. PORE SIZE
     MOLYBDENUM FOAM CASING
  THEORETICAL CAPACITY DENSITY 0335 A-hr/cm2
0.2  A/cm DISCHARGE-34
                                                     0.2 A/cm  DISCHARGE- 17
     0
               0.05                  O.I

                     CAPACITY DENSITY,  A-hr/cm;
         0.15
     0
         10              20             30             40
             PERCENT OF THEORETICAL CAPACITY DENSITY
                 50

-------
                              XIII-13
CATHODE
  TYPICAL PERFORMANCE CHARACTERISTICS

    FOR VARIOUS CATHODE STRUCTURES


W/cm2x A-hr/cm2**  A-hr/cm3 A-hr/gm  CYCLES  LIFETIME
                                                hr
DISK
COMB
LAMINATED
ENCLOSED
LAMINATED
RESERVOIR

3
3.5
4
4
2
n TM^MOTT1
0.2
0.4
0.5
0.2
0.36*
V 1 s*m TMTT
0.6
0.3
0.5
0.15
0.15
?-DTTT ffTDnn
0.21
0.15
0.17
0.05
0.05
17 n T C T A Ml
<5
<2
<10
419
>400
"'T?
<10
<5
<20
588
>500

**ONE-HOUR RATE
                  OTHER AREAS OF INVESTIGATION


              1.  PHASE EQUILIBRIA:  Li2S-S-LiX

              2.  INTERFACIAL PHENOMENA

              3.  CATHODE MATERIALS STUDIES

              4.  SOLID ELECTROLYTES

              5.  MATERIALS EVALUATION

-------
                                  XIII-14










bearing species in the electrolyte.  A minimum solubility is desired in




order to minimize the rate of loss of sulfur from the cathode.  Some




typical results are shown in the form of a pseudo-ternary phase diagram




in Slide 13.  Here, the extent of the phase marked LS is to be minimized.







Interfacial phenomena are also important in retaining sulfur.  The cathode




current collector should be well-wetted by sulfur, but not by electrolyte.




The wetting properties of various solids by sulfur and by electrolytes




are being studied.  Cathode materials (i.e., sulfur, plus various additives




such as phosphorus) which have high electronic conductivities, low




viscosities, low vapor pressures, and other desirable properties are being




sought.  Investigations of the suitability of solid lithium-ion conductors




for use as electrolytes has recently begun.  Materials stability continues




to be an important area of investigation.   An up-to-date summary of the




corrosion rates of a number of candidate materials of construction in




lithium-sulfur mixtures and in pure lithium is given by Slides 14 and 15,




respectively.







Electric Vehicle Performance Calculations




In order to evaluate the potential performance of an all-electric vehicle




powered by a lithium/sulfur battery, some computer calculations have been




performed for an electric automobile having the characteristics shown in




Slide 16.  The power requirements for this vehicle were calculated from




the equations given in Slide 17, using the values of the constants




shown in Slide 18,  The driving profiles assumed for the purposes of the

-------
       A
       D
       O
       0
THREE  PHASES (quench)
TWO  PHASES (quench)
PHASE  BOUNDARY
DTA
Li2S-S-(LiBr-RbBr)
 PSEUDO TERNARY
  SYSTEM 360 °C
LiBr-RbBr

-------
                 XIII-16
  CORROSION  BY 20 %  Li~S MIXTURE 375°C
INCONELJ702)
2RK65 SS
ZIRCALOY-2
347 SS
HASTELLOY-X
ALUMINUM
205 SS
                          AVE. RATE, I00-300hr
                          MAX. RATE, 620 hr

                          AVE. RATE, 620 hr
            012345
                CORROSION RATE.rnm/yr
6

-------
                 XIII-17
CORROSION  BY  MOLTEN  LITHIUM  AT 375° C
BeO°
Th02
AIN
BN
LiAI02
MgOb
BeOC
BeOd
(
! 1 1 1 ill! ; } \ \
3
^ TEST DURATION 1000-1200 hr
P
z

'//A

///////\

/////////A

//////////.

'////////////////////\ \/ / / / //
a. HOT-PRESSED, HIGH-PURITY
b. SINGLE CRYSTAL
c. RECRYSTALLIZED GRADE
d. COMMERCIAL GRADE
lilt till n 1 1
D 0.5 1.0 12
CORROSION RATE, mm/yr
1 1 «\
               20             40
            CORROSION  RATE, mils/yr
480

-------
                          XIII-18
            ELECTRIC VEHICLE CHARACTERISTICS
CURB WEIGHT



PAYLOAD WEIGHT



BATTERY WEIGHT



   SPECIFIC ENERGY (4hr RATE)



   SPECIFIC POWER  (Ihr RATE)





ACCESSORY POWER



AIR CONDITIONING



POWER STEERING



TOTAL
                     1588 kg



                      227 kg



                      397 kg


                      220 W-hr/kg


                      220 W/kg




                      230 W


                     4400 W



                     1900 W


                     6530 W
(3500 lb)



(500 lb)



(875 lb)


(100 W-hr/lb)



(100 W/lb)
              POWER REQUIREMENTS EQUATIONS




                 "P      "P
                                 1.1 Ra)
Pr = V(Rr + R^ + Rg





              V2
           Rr = Tc x W
           Rw = fA CD Af



           R  = W sin e
            o

           R  = W dV
            a   g ar

-------
                       XIII-19
                    VALUES OF CONSTANTS
ROLLING RESISTANCE COEFF.  (T )           0.0175



AIR DRAG COEFFICIENT  (CD)                0.35



FRONTAL AREA  (A£)                        2 . 32m2



TRANSMISSION EFFICIENCY  (E  -E )          0.82
                         ^ m  ej


TOTAL ACCEL. /LINEAR ACCEL,               1.1



                                                  "6
AIR DENSITY/ gc
                                         1.25 x 10

-------
                                   XIII-20
calculations are presented in Slide 19.  For each profile the range was




calculated for no accessory power for 230 W (lights, heater, etc.) and




for 6530 W (air conditioning and power steering added).







The laboratory results were put into the electric vehicle calculations




in the form of E = f (q, i) equations, where E is cell voltage, q is




capacity density, and i is current density.  Slide 20 shows a set of E - q




plots for various i, drawn by the computer.  These equations, together




with those of Slide 17 and the driving profiles, were combined with a




cell and battery design similar to those shown on Slides 21 and 22 to




yield the results shown in Slide 23.  The ranges shown are less than




the EPA goal of 322 km (200 mi); hence, there is a need for improvement




in the capacity density of the laboratory cells and/or the design of the




battery.  These results do show, however, that electric vehicles based on




lithium/sulfur batteries can be expected to have ranges of 150-250 km, if




25% of the curb weight of the vehicle is alotted for batteries.






Acknowledgement




This paper represents a summary of the work of the many people listed in




Slide 24.  They are the ones to whom credit should go for the accomplish-




ments described above.  I also wish to thank the Air Pollution Control




Office of the Environmental Protection Agency for support of this program.

-------
0
    URBAN  DRIVING PROFILE
     SUBURBAN DRIVING PROFILE
                                                                  i
                                                                  NO
              CROSS-COUNTRY DRIVING PROFILE
O\J
60
40
20
0




—
i i i i i i i i i i i , ,
300    600    900    1,200    1,500


                  TIME, SEC
1,100

-------
                    Li / LiBr -RbBr/ Li in S

                            375°C

                        EMPIRICAL  FIT
                                                             CURRENT DENSITY

                                                                  A/cm2


                                                                A   0.095


                                                                A   0.094

                                                                V   0.091


                                                                •   0.049
0.40 —
0.20
    0  0.01   0.02 0.03 0.04  0.05  0.06 0.07  0.08  0.09  0.10


                  CAPACITY DENSITY,  A-hr/cm2
i
M
N>

-------
SEAL CLAMP
         &•:••••-; SULFUR , X-.'. ,•%
            ELECTROLYTE
                                                               H
                                                               I
                                                               NJ
1;33cm
 ELECTRICAL INSULATOR
  (HIGH TEMR  PLASTIC)
           LITHIUM/SULFUR CELL CONCEPTUAL DESIGN

-------
             XIII-24
56 cm
  THERMAL
 INSULATION
                              METAL
                             CONTAINER
                        POWER  TERMINALS
95 cm
67 CELLS
PER STACK
                            COPPER BAR
      LITHIUM / SULFUR  BATTERY
         CONCEPTUAL DESIGN

-------
                     XIII-25
  300
  2501-
E 200
LU
O

<  150

LL)
_l
U

£  100
   50
    0
           X
                       ACCESSORY LOAD
                         BO 6530 W
                             230 W
                         mm o w

                                           150
                                           100
                                           50
         URBAN
        DRIVING
        PROFILE
        32 Km/hr
        (20mph)
                  SUBURBAN
                   DRIVING
                   PROFILE
                   67 Km/hr
                   (41 mph )
CROSS-COUNTRY
    DRIVING
    PROFILE
   106 Km/hr
    (66 mph )
                                           0

-------
                     XIII-26
                  EPA PROGRAM
R. K. STEUNENBERG




J. P. ACKERMAN




B. A. FEAY




M. L. KYLE




H. SHIMOTAKE




R. RUBISCHKO    (GOULD)




D. M. GRUEN     (CHM)




A. J. ZIELEN    (CHM)




T. W. LATIMER   (MSD)




J. N. MLJM)Y     (MSD)




D. E. WALKER    (EBR-II)
R, M, YONCO
J. R. PAVLIK




F. J. MARTINO

-------
                    XIV-1
                 Chapter XIV
RESEARCH AND DEVELOPMENT PLAN OF ELECTRIC CAR
                      by
                 Mr. Shizume
Japanese Automotive Manufacturers' Association
                    Japan

-------
                                  XIV-2


1.  Fundamental Plan

   (a)  We intend to develop a "new type electric car" for the purpose

   of use in the town from 1971 to 1975.  New type electric cars include

   the following variations:

       (1) Small scale passenger car.
       (2) Small scale cargo car.
       (3) Medium scale passenger car.
       (4) Medium scale cargo car.
       (5) Bus


   The development cost is estimated about thirteen million dollars.

   We develop it by concentrating the R & D abilities of national

   organizations, private automobile companies,  universities and so on.

   Our main purpose to develop it is to protect  the environment (air

   pollution caused by exhaust gas or noise caused by engines).


   (b) We will develop it according to the following procedure:

       (1) We will develop the experimental electric car by pursuing the

       optimal structure and performance to accomplish this purpose.  We

       will make the primary experimental electric car and test its

       performance and safety, etc. for the first three years.  Then we

       will make the complete experimental electric car and not only

       test its performance and safety but also judge it from the view

       of technique, social needs or economy for the following two

       years.  The performance data shown later is the performance of

       primary electric cars which we will develop for the first three

       years.


       (2) We will research and develop several components used to the

       experimental electric car such as Pb-PbC^ battery, new type

-------
                                  XIV-3










       battery, new type motor (thyristor motor, wheel motor, etc.)




       controller, plastic body and so on.  We will use this result




       to develop the experimental electric car.







       (3) On the other hand we will study the utilization system of




       electric cars from the view of city traffic system, standardiza-




       tion, optimal total energy system and so on in order to use it




       well and promote the popularity.







Append ix




National Research and Development Program




   (1) Object




   From the standpoint of national interest, many fields of industrial




   technology need urgent research and development which requires a




   great deal of expense and a long-term period for their fulfillment.




   It is demanded of the Government to promote such important research




   and development positively by making plans, bearing research expenses




   and by organizing research abilities.







   In answer to these acute needs the National Research and Development




   Program System was established in 1966.







   (2) Method of Selection




   Each project is selected from the projects proposed by governmental




   organizations, national laboratories or private companies under the




   following criteria.

-------
                               XIV-4
     (a) The project is very important and immediately needed  to
     achieve leveling up of industrial construction, effective
     exploitation of natural resources or preventing public nuisance.
     (b) Its techniques are advancing and have effective  spin-offs.
     (c) Large amount of financial resources over a long  period are
     needed to research and develop the techniques and, these, in
     addition, are subject to heavy risks.
     (d) The target of R & D of the techniques must be set and also
     the technical method to achieve it must be forecasted.
     (e) The R & D of the techniques requires the concentration of the
     R & D abilities of national organizations, private companies,
     universities and so on.
Now the National R & D Program System (NRDPS--known as Ogata-project--

in Japanese) includes the R & D of "Magneto-hydrodynamics generator,"

"Large-scale digital computer," "Desulfurization process," "New

process for olefin production," "Sea water desalting and by-product

recover," and "Remotely controlled undersea oil drilling rig."

From April 1971, "Electric car," "Jet engine used to civil aircraft,"

and "Pattern recognition system" will be starting to develop.


(3) Method of Management—A special organization for managing the

System has been established in The Agency of Industrial Science and

Technology (AIST).   The organization manages each of six projects.

The following five headings has the responsibility of the management.

    (a) Selection of R & D project.
    (b) Planning the actual programme.
    (c) Selection of sub-contractors.
    (d) Operation of the System.
    (e) Management of the achievements.

Another organization, such as the advisory committee or evaluating

committee, assists the organization in managing the System.

-------
2.  Research and Development Plan  of  Electric  Car

Establishment of Judging
Standard 6t judgement
of Electric Car
i
Development of Experimental
Electric Car
Development of Several
Conpoents
Decelopment of Charging
System
Study of Utilization System
Development Cost
(Million Dollars)
1971
Study of .
Standard
raent of
Equipmen
Design 1 	 >

Design 1 	 *

Fundamental

| System Stud


1.25
1972 1973 1974 />, 1975
Tudijing Establishment Develoocent of
Develop- — * of Judging — ^ Test 	 > Testing Equip- — » Final 1
testing Bshod ~/^ ment Test |
: Study & Esta- A 'V
blishment of
Judging Standard

Trial Production T Production
of Primarv Exoeri- s[ _, . ~l -. r .,,,,, ,
. . _,J . ' ^ Design > of Final Expei'l-
mental Electric e- ^ mcntal Elcctric
Car ^ * **
A. v>Qr
1 t*—

Trial Pi'uducLlou ^Design S ^ ^ 	

. J

Studyl (Application Study] *-•


y |— ^Application Study] 'Trial Production}

*\ *, i
^[System SLuuy| 	 >-


3.64 3.97 2.56 2.0
Remarks
Nations. l Laboratory
Car Maker
Private Company
& National
Laboratory
Private Company
National Organizatior
13.42
(TOTAL)
                                                                                                                                                    I
                                                                                                                                                    Ln

-------
3.  Specification & Performance of the Experimental Electric Car


            (The End of 1973)

Passenger + Payload (kg.)
Total Weight (kg.)
Maxmum Speed (km/h)
Range (km)
Acceleration
Ability (0 - 30 km/h)
(Sec)
Climbing Ability
(Speed of 6 degree slope)
(km/h)
Cargo Car
Small Scale
2 + 200
1,100
70
130 - 150
5
40
Medium Scale
2 + 1,000
3,500
70
180 - 200
5
40
Passenger Car 1 Bus
Small Scale
4 or 2 + 100
1,000
80
130 - 150
4
40
Medium Scale
5 or 3 + 300
2,000
80
180 - 200
3
40
Large Scale
60 - 80 persons
15,000
60
230 - 250
8
40
                                                                                                                                          I
                                                                                                                                         ON

-------
                        XV-1
                     Chapter XV
STUDIES BY FIAT ON THE ELECTRICALLY-DRIVEN AUTOMOBILE
                         by
                   G. Brusaglino
       Chief, Electrical Research Department
           Fiat Research and Development
                    Turin, Italy

-------
                              XV-2
     Electrically-driven automobiles have a long history at
Fiat.  As early as the last world war,  some electric cars were
built on chassis and bodies of that  period.

     Within the framework of research aimed at  reducing air
polution, in 1961 Fiat began new studies  on the possibility of
using electrical drive in the light  of current  technological
possibilities.

     For the first experiments,  we used vehicles provided with
a clutch and gearbox, namely:  a Fiat 200 van and a Fiat 1100
automobile.

     In orienting these studies, we  took into consideration the
various possible drive systems with  an eye to defining the best
applications for use in automobiles, especially urban areas.  We
found that such systems must satisfy the following conditions:

     1.  High overall output of the  whole drive system, i.e., of
the motor and its control system under all vehicle operating
conditions,  especially in stop-and-go city driving.

-------
                               XV-3
     2.  Capability of quick response in traffic.  Thus, the
following were deemed necessary:

     — fast acceleration;
     — automatic braking effect when accelerator is released
        (similar to effect in thermal motor) with possibility of
        later energy recovery.
     — possibility of velocity control and coupling with no
        discontinuity or with a discontinuity contained within
        limits judged acceptable for thermal motor vehicles.

     3.  Availability of high specific outputs even at small rpm
rates to allow good motor operation.

     This characteristic enables us to bring into play certain
laws of motion using appropriate control to reduce acceleration
time, taking into account passenger comfort.

     These conditions must be accompanied by feasibility of each
element in the system, low price, and inexpensive maintenance.

     In the drive system of the vehicle mentioned above,we  used  a
compound excitation motor where one of the coils  can be supplied
independently.  The armature has no dissipation element:  velocity
regulation depends only on variation of the current in the  in-
dependently-supplied coil.

     This system results in variable motor velocities in a  con-
tinuous fashion over a range of values which,  through special
design  features, has proven sufficiently broad.   Moreover,  this
system  has, naturally, the possibility of dynamic braking by
energy  recovery with a braking effect similar  to  that of a  piston
engine.  In fact, under over-excitation conditions with respect

-------
                               XV-4
to the velocity imposed by the car,  the motor shifts to operating
as a generator with no discontinuity.

     The energy expended for regulation is practically negligible
and the output of the drive system,  identified with that of the
motor, assumes values close to 95% under normal operating condi-
tions .

     The output in the transitory regime may  reach the same
values by similarly setting the de-excitation gradient in
correlation with velocity.

     In this system, the motor velocity must  not  drop below a
certain regime corresponding to maximum excitation;  consequently,
the installation must be completed with a gearbox and transmission.
Nonetheless, the system has the advantages of durability,  economy
and high overall output, which were  the bases of  the research
orientation.

     We next attempted to simplify the operation of experimental
cars by eliminating the transmission and gearbox.

     For that purpose, we added a nondissipating  regulation system
to the armature circuit to permit the  motor regime to vary from
zero velocity to the velocity when regulation of  the independent
field begins.

     A system of this type was used  on a car  derived from the
Fiat 850 model.

     The characteristics of the car  are as follows:

     — Car                       Fiat 850 (with  structural
                                  modifications)

-------
                               XV-5
     — Total unloaded weight      1025 kg
     — Batteries                  lead, special type with
                                   plastic vat
     — Battery weight             320 kg
     — Battery voltage            96 V
     — Motor                      Fiat, 6 poles, compound
                                   excitation
     — Nominal output peak        21 CV
     — Maximum peak output        45 CV
     — Maximum velocity           72 km/h
     — Acceleration               from 0 to 50 km/h in 8  seconds
     — Range at 60 km/h on
        level road                 65 km

     In order to ascertain more realistically the performance
of the regulation system developed for a small light car es-
pecially suited for use in town,  a Fiat 500 car was  transformed  to
electrical drive.

     To simulate the performance  which could be supplied by these
future, high energy density,  light-weight batteries, the car  was
fitted with a limited number  of lead batteries.

     The characteristics of the car are as follows:

     — Car                        Fiat 500 (with structural
                                   modifications)
     — Total unloaded weight      730 kg
     — Battery weight             160 kg
     — Battery voltage            96 V
     — Motor                      Fiat, 6 poles, compound
                                   excitation
     — Nominal output             21 CV

-------
                               XV-6
     — Maximum peak output        45 CV
     — Maximum velocity           80 km/h

     The operational characteristics of this  car,  aside from its
range which was not significant in this study,  were judged satis-
factory.  The car was easy to drive in city traffic,  due to the
outstanding pick-up.

     We should point out that these cars were produced especially
for the purpose of testing certain drive systems.   These systems,
moreover, allow us simply to increase the performance to the
point of making them similar to those of traditional  vehicles  for
use not only in the city.

     We are at present producing a vehicle fitted  with a drive
system with an asynchronous motor fed by batteries through a
static converter (produced in collaboration with Philips Co. of
Milan).

     This experience was regarded as a prelude  to  the use of such
a system in buses.

     In particular, we plan the production of hybrid  buses with
batteries recharged by an electric generating group,  which could
be brought about by a gas turbine.

     The drive systems tested so far use motors of almost entirely
conventional types, even if they are designed with particular
features to make best use of certain characteristics.

     The relative control modes consequently  conform  to the need
to adapt the characteristics of the motor to  the requirements  of
driving the vehicle — elements which can not always  be reconciled,

-------
                               XV-7
     At this point, it was regarded as opportune to approach the
problem of a drive system as a whole anew and to revise the
motor design by bringing it closer to the requirements of a
vehicle through more rational regulation, i.e., considering the
structure of the motor as intimately connected to the system
of control.

     Current Fiat research programs on electrical vehicles are
based on these principles.

-------
                      XVI-1
                   Chapter XVI
ELECTRICAL VEHICLES WITH FUEL CELLS:  WHY AND HOW?
                        by
                    J. Beslier
      Chief of Electrical Equipment Research
            Peugeot Automobile Company
                  Paris, France

-------
                             XVI-2
          In order not to impair road traffic, the electric
     vehicle should have a performance close to that of other
     vehicles on the road.  It must have sufficient speed,
     good acceleration ability, and a normal action radius.
     But strengthening the car structure to ensure safety
     in case of impact makes the chassis heavier and will
     make the construction of small vehicles more and more
     difficult.   All this leads to a vehicle having medium-
     sized dimensions, equipped with a powerful generator,
     with large energy capacity.  Only the fuel cell meets
     the requirements of the problem.

          The associated Companies PEUGEOT-ALSTHOM have signed
     an important contract for research with ESSO-USA, which
     should lead to a methanol-air cell that would be suit-
     able for the contemplated vehicle.
     In an electric vehicle, the energy consumed by the motor
comes from a generator which is to the electric motor what the
boiler is to the steam engine.   To answer the problem posed,
which Is the reduction of pollution,  one needs a nonpolluting
generator and a silent motor transmission unit.

-------
                              XVI-3
                      What Vehicle to Make?

     Before examining each of these points, we must define the
main characteristics of the desired vehicle in order to select
the best adapted elements.  In fact, we may attempt to produce a
vehicle answering a very special list of specifications (delivery
vehicle, taxi, vehicle for city use, small public transporta-
tion vehicle) or, on the other hand, we may try to convert the
standard vehicles presently used which are included in the various
possible categories.  We shall not deal here with heavy weights
and other large specialized vehicles which are not customarily
produced by the Peugeot Company.

     Very diverse considerations will influence the choice.   They
deal with industrial requirements, technical capacities, market
demand and the norms imposed by regulations.

                     Effects of Regulations

     As far as the latter are concerned, it is undeniable that
they play a large role in the work of the Research Section.
Whether it is a question of passive or active safety,  nuisance
caused by the vehicle or its operational features, all these
problems have caused significant chassis changes both with regard
to structure and equipment.

     Improved safety in case of collision involves strengthening
the structure.  To withstand front, back, and side collisions,
while providing survival space, there must be more and more  space
for the structural elements which are indispensable for protection
of the main body.  The latter will tend to increase to permit the
installation of protective devices for the passengers.   Another

-------
                               XVI-4
consequence of this strengthening is a substantial increase in
chassis weight.

     All this leads us to believe that it will become increasingly
difficult to reduce vehicle dimensions.

                          Market Demand

     It is undeniable that the vehicle best corresponding to the
demands of European consumers is the average-sized vehicle.
Very small cars have had only ephemeral or limited success.  They
remain relatively expensive, especially when the price is compared
with their very limited space, performance; and comfort capacities.
By a very small vehicle, we mean vehicles designed to carry two
persons, with a minimum of exterior accoutrements.

     The average-sized vehicle owes its success to the fact that,
for a price many can afford, It offers a good compromise allowing
practical and comfortable use in diverse business or family
circumstances.

     This success does not seem to be diminishing,  and it leads
us to think quite naturally that a new vehicle will have a greater
chance of market penetration if it corresponds to the needs of the
consumer —that is, if it is in the average-sized bracket.

                     In dus t r ial_ Re q_u ir e men t s

     As far as  industrial installations are concerned, it is
apparent that the replacement of the internal combustion motor
by an electric  generator-motor unit will involve very complex
and costly reconversion problems.

-------
                              XVI-5
     For these reasons, it appears inevitable that the classical
vehicle will continue to be produced in parallel.  The new
investments required will be very high^ and traditional equipment
will have to be written off.

     The last factor coming into play will be purely technical in
nature:  what motor, and especially what generator, will be used?

                            The Motor

     Many studies have been made around the world on developing
electrical motors suited for automobile propulsion.  This is a
very special problem, because the conditions of use are quite
different from those for industrial motors and locomotive engines,
For the latter, the weight and dimensions are not as important
as in the automobile, since longevity must be considerable:
2000,000 km represents the yearly use of a locomotive.  On the
other hand, the automobile motor must be able to be mass-produced
and must be light and compact.  Finally, it must be suitable for
the various operating regimes encountered by a vehicle used in
different circumstances and climates, by demanding or careless
customers.

     Motors can be classed into two large categories, according
to whether they operate on direct or alternating current.  The
development of semiconductors enables us to improve flexibility
and transmission output and to contemplate the production of DC
motors with no collector or supplying AC motors from a DC source.
Finally, it is possible (and may be of safety interest) to use
the motor for braking.  Finally, let us note that the motor out-
put can be divided to activate two motorized axles or four
motorized wheels.

-------
                              XVI-6
     The subject of motors is quite vast.  Much research remains
to be done before we can draw up a complete, technical, and
economical balance sheet for the unit comprised of the motor and
the electronic power circuits.

                          The Generator

     The sources of electrical energy,  or generators,  fall into
two categories:  storage batteries or secondary generators which
only restore the energy which they store in electrochemical form
and cells or primary generators, which  produce electrical energy
directly from fuel.

     Storage batteries are well known,  especially the  lead storage
battery whose low mass-energy can be improved only slightly.   Many
other types of storage batteries have been studied or  are being
tested.  It appears that real progress  can be made,  but at the
cost of hitherto undetermined mechanisms.  Improvements are also
possible in the area of recharging, the duration of which is  a
significant handicap.  The latter is still encumbered  by the
inevitable need for a complicated infrastructure connecting the
users with a distribution network, using a counter for tabulating
the energy absorbed.

     Cells are also the object of many  studies, and various fuels
have been investigated.  Aside from dangerous, difficult to handle
products, or costly products, there are a number of possibilities.
The most interesting may be methanol combined with atmospheric
air:  it is a relatively easily obtained fuel, with an
acceptable cost, which can be distributed by existing  service
stations with no new infrastructure. However, if this product
can be used commercially with no major  difficulties, its use to
produce electrical energy proves difficult:  it is a stable body

-------
                              XVI-7
which liberates ions reluctantly even in the presence of noble
catalysts.

     The latter must be disregarded due to their rarity.  The
search for an effective catalyst which is cheap, easily prepared
and long-lasting is a big problem.  Nor must one underestimate
the difficulties relative to technology of the cell itself and
its equipment.

     Work is progressing, however, and various laboratories have
announced positive results with small cells which are still bulky
and expensive.  However, their great merit lies in showing that
this path is open.

                     Vehicle Characteristics
     The preceding conditions enable us to define the principal
characteristics of the vehicle.

     We have seen that, because of safety requirements,  it will
be very difficult if not impossible to make a very small vehicle.
We have also seen that this vehicle will be relatively heavy.

     To guarantee adequate commercial success, it will have to
have a performance approximating that of comparable traditional
vehicles, or at least, not much lower if the users are to be
satisfied with it.  This point is likewise important for traffic
flow since, as we know, traffic flow is better, the more homo-
geneous it is as far as acceleration and velocity capabilities are
concerned.

     We can see that the vehicle with a classical lead storage
battery is far from meeting these conditions because of its low

-------
                              XVI-8
capabilities.  The considerable mass of the required batteries
limits the acceleration, and the low mass-energy makes it
necessary to limit the velocity.  The mass of the batteries
itself requires a heavier structure.

     The problem of the small vehicle with a storage-battery is
well-known by Peugeot, who produced approximately ^400 during the
years of extreme restriction, 19^1-^3-   It was the VLV,  a light,
two-seater, town vehicle with lead batteries and, thus,  of very
limited performance like all vehicles of the type.  With this
little car, approximately 70 km could be covered with a  peak
velocity of 36 km/h.   Its total weight  was 365 kg, of which 160
kg was batteries.  Restored and modernized,  it is certain that
its performance could be increased somewhat.  Velocity could be
raised to 55 to 60 km/h perhaps, but only by retaining the old
type of chassis which does not conform to safety norms.
                       VLV PEUGEOT
     The new batteries being tested will certainly  be  much better
than the present ones, but,  short  of significant  progress, their
weight will still limit the  possibilities of the  vehicles  using
t hem.

-------
                              XVI-9
     To obtain the minimal performance defined above,  it  is
necessary to have a generator providing at least 450  wh/kg.   Let
us recall that lead storage batteries have difficulty  in  reaching
40 wh/kg.  Only a fuel cell appears capable of giving an  adequate
energy density to ensure velocity, acceleration and action
radius.

     One should note that the weight of a storage battery is
clearly proportional to the stored energy, meaning that the
weight increases rapidly with the required action radius. In the
case of a fuel cell, this is not the case because the weight  of
the cell is a function of the power it supplies, and the  action
radius is a function of the capacity of the fuel tank. With
sufficiently powerful fuels, as in the case of methanol,  the
consumption in kg per kW/h is low relative to the overall weight
of the vehicle.

     Since the fuel cell is the only solution leading to  the
production of a vehicle answering the problem posed, we had  to
find this cell.  The studies carried out in this field by the
Alsthom Company interested Peugeot, since they were performed
in accordance with a method taking into consideration technological
problems from the outset.  It is not a question of only  performing
laboratory tests which cannot be transposed to industrial uses.
The development in parallel of two techniques is doubtless more
difficult, and perhaps slower, but only on the surface because
it eliminates the final necessity of spending a great deal of
time exploring paths which may not be practical.  The two companies
collaborated to study a vehicle with a cell. Alsthom, brought  to
Peugeot its experience in the field of electrical motors  and
control circuits.  This Association has been working since the
end of 1967, and Interesting results have been obtained.

-------
                              XVI-10
     Above an experimental vehicle is shown.   It  is a small car
capable of transporting 11 passengers plus the driver with 175 kg
of luggage.  The velocity is 95 km/h.
     The vehicle is derived from a type J? Peugeot,  and is to be
used for electrical transmission tests.

-------
                              XVI-11
     It is equipped currently as a hybrid with a controlled
ignition leading to a constant velocity of the electric generator
working in parallel with the batteries.  The generator assembly is
placed in the rear of the vehicle, and it is anticipated that it
can easily be replaced by a fuel cell.

     Even in its present form, the performance of the vehicle is
comparable to that of traditional vehicles of its category.

     In the special field of cells, the level of development
obtained with hydrazine cells has interested ESSO-USA, who has
just signed an important research contract with the Alsthom-
Peugeot Association to develop the methoanol-air cell.

     The French Government has recognized the value of the program
and is aiding the Alsthom-Peugeot Association by loans supplied
through the General Delegation for Scientific and Technical
Research.

     We have seen the reasons leading Peugeot to develop research
on an average-sized electrical vehicle.  This type of vehicle
satisfies the requirements of reducing nuisances due to atmos-
pheric and noise pollution, and at the same time answers a
natural consumer demand.

     For  commercial  success,  it  is also  important that  the
proposed  vehicle need not require too  many modifications
in the driving habits of the users.  They should be able to
switch from one vehicle to the other without disorientat ion or
any special constraints.  The problem of recharging storage
batteries, which requires time and special fixed installations,
would always be an unpleasant constraint.

-------
                               XVI-12
     In conclusion, we feel that we should seek to produce a
vehicle which, for the user, will differ very  little  from
vehicles currently in use.  The electric vehicle with a fuel
cell is a solution, but it is still far in the future.

-------