APTD-156-
OPTIMUM WORKING FLUIDS
FOR AUTOMOTIVE
RANKINE ENGINES
VOLUME I - EXECUTIVE SUMMARY
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Mobile Source Air Pollution Control
Advanced Automotive Power Systems Development Division
Ann Arbor, Michigan 48105
-------
APTD-1563
Prepared By
D. R. Miller, H. R. Null, Q. E. Thompson
Monsanto Research Corporation
800 North Lindbergh Blvd.
St. Louis, Missouri 63166
Contract No. 68-04-0030
EPA Project Officer:
K. F. Barber
Prepared For
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Mobile Source Air Pollution Control
Advanced Automotive Power Systems Development Division
Ann Arbor, Michigan 48105
June 1973
-------
The APTD (Air Pollution Technical Data) series of reports is issued by
the Office of Air Quality Planning and Standards, Office of Air and
Water Programs, Environmental Protection Agency, to report technical
data of interest to a limited number of readers. Copies of APTD reports
are available free of charge to Federal employees, current contractors
and grantees, arid non-profit organizations - as supplies permit - from
the Air Pollution technical Information Center, Environmental Protection
Agency, Research Triangle Park, North Carolina 27711 or may be obtained,
for a nominal cost, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22151.
This report was furnished to the U.S. Environmental Protection Agency
by Monsanto Research Corporation in fulfillment of Contract No. 68-04-0030
and has been reviewed and approved for publication by the Environmental
Protection Agency. Approval does not signify that the contents necessarily
reflect the views and policies of the agency. The material presented in
this report may be based on an extrapolation of the "State-of-the-art."
Each assumption must be carefully analyzed by the reader to assure that it
is acceptable for his purpose. Results and conclusions should be viewed
correspondingly. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
Publication No. APTD - 1563
ii
-------
ACKNOWLEDGEMENT
The authors acknowledge with gratitude the helpful counsel of
Messrs. K. Barber, S. Luchter and F. P. Hutchins, all of the
Environmental Protection Agency's Advanced Automotive Power Systems
Development Division. Many persons within Monsanto Research Corp-
oration and Monsanto Company contributed substantially to the
research program reported here. These included R. C. Binning,
J. A. Conover, L. L. Fellinger, R. L. Green, R. L. Koch, R. J.
Larson, G. J. Levinskas, R. A. Luebke, N. F. May, J. T. Miller,
L. Parts, A. C. Pauls, P. F. Pellegrin, J. V. Pustinger, W. R.
Richard, A. D. Snyder, W. N. Trump, W. M. Underwood, J. A. Webster,
E. P. Wheeler and F. J. Winslow. Their manifold inputs are
gratefully recognized, as also are those of Professor P. A. S.
Smith of the Department of Chemistry, University of Michigan, who
served as a consultant.
Also recognized are the contributions made by D. B. Wigmore and
R. E. Niggemann of Sundstrand Aviation, Division of Sundstrand
Corporation, both in the Research Program and in the Engine
Optimization Studies performed for this report.
iii
-------
CONTENTS
Volume I. Executive Summary Page
Introduction 1
Summary 2
Conclusions 27
Recommendations 30
Volume II. Technical Section
Volume III. Technical Section - Appendices
Volume IV. Engine Optimization
iv
-------
INTRODUCTION
This is the final report of a research program performed by
Monsanto Research Corporation for the Environmental Protection
Agency, Advanced Automotive Power Systems Development Division
(AAPSDD) under Contract 68-04-0030. The objective of the work
was to determine the best working fluids for Rankine cycle
automotive power plants.
The Rankine Engine has been identified by EPA as a potential
alternative to the internal combustion engine for the automotive
application in the event the internal combustion engine cannot
satisfy the 19.75/76 Federal Emission Standards. Development work
has been proceeding on the Rankine engine, as well as other alter-
native power plants, under the direction of AAPSDD since initiation
of the Clean Air Act Amendments of 1970.
In a Rankine cycle power plant, the working fluid, which is
repeatedly vaporized, expanded and recondensed, plays a key role
in determining the competitiveness of this system. The thermo-
dyriamic, chemical and physical properties of the fluid have a
strong influence on the power plant efficiency, materials of con-
struction, component weight and size, the degree of hazard, if
any, and to some extent the power plant cost. Therefore, to make
certain that the Rankine engine could be evaluated at its greatest
potential, this study program was initiated to search out the best
fluids currently available for the automotive Rankine engine appli-
cation.
To accomplish this objective, specific guidelines were established
for the fluid screening process. These guidelines, contained within
the contract work statement, covered all significant fluid char-
acteristics: availability, compatibility, thermal stability, cycle
efficiency, cold start ability, applicability in principal engine
types, owner cost, toxicity, flammability and environmental hazard.
Quantitative targets set for these characteristics were intended to
approximate the upper limit of what could be achieved with presently
known materials and current technology. The results indeed reveal
that the criteria were well chosen, as only two possible fluids from
over 100 candidates survived the screening process, and these did
not totally satisfy all of .the stated requirements. In reviewing
the results it should be kept in mind that less stringent or
different selection criteria could very well have led to the
selection of different candidates.
The volume of reportable material resulting from this program has
dictated the physical subdivision of the detailed technical matter
into three separate volumes (II through IV) with this Volume I
serving as an executive introduction and summary.
-------
SUMMARY
A translation of system and working fluid contractual goals into
fluid screening criteria was the first step toward meeting con-
tract objectives. Contractual goals are listed in Table 1, while
the derived screening criteria are given in Table 2. A detailed
discussion of the derivation of the criteria is given in Volume II.
It may be noted that the screening criteria are mainly of the go/
no-gO) variety useful in the early elimination of unsuitable
materials. Additional requirements on cost, hazards, compatibility
and system operability were treated on an ad hoc basis involving
only a few of the more .promising candidate fluids.
On the basis of predicted thermal stability and melting/boiling
points, 112 candidate fluid constituents were identified and,
where possible, procured. Included in the candidate list
(Table 3) for comparison are several organic fluids previously
used or suggested as Rankine working fluids. Except for water,
all identified candidates are organic chemicals. Although
inorganic candidates were sought and a list of 32 possibilities
was similarly compiled, none was judged to be compatible with
low-cost metals at the 712°F minimum peak fluid temperature
needed to meet cycle efficiency requirements.
Binary and ternary mixtures of several of the 112 constituent
candidates were also identified with the expectation that such
mixtures offer wider latitude in matching diverse fluid re-
quirements .
Identified fluids were next subjected to screening tests under
the first five criteria. Criteria 1 through 3 required mainly
literature work. A special steel ampoule thermal exposure test
was employed to screen fluids for 720°F thermal stability under
Criterion 4. Criterion 5, designed to distinguish between fluids
requiring "reasonable" and "too-much" regeneration, drew on
published thermophysical properties or, when these data were not
available, on recognized property prediction procedures to com-
pute the I-factors.
Candidate screening results under Criteria 1 through. 5 are
summarized in Table 3- Here it is seen that the majority of
initial candidates failed to survive 200 hours at 712°P in the
Criterion 4 steel ampoule test. The few which did are listed in
Table 4 with their corresponding capsule survival times, I-factors
(at normal boiling points), and Criterion 3 (normal boiling point)
results. Brackets in the I-Factor column delineate the fluids
satisfactory per se (innermost bracket) from those suitable only
as components of a mixed (or complex) working fluid (outermost
bracket) and from those satisfactory as neither, all based upon
the tolerable regenerator size, Criterion 5-
-------
Table 1. PROPULSION SYSTEM AND WORKING FLUID GOALS
Propulsion System Characteristic
Type Expander
Approx. Net2 Full Power
Max. Weight
Max. Volume
Regenerator Size
Min. WF6 Condensing Temp.
Max. Low Ambient Temp.
Min. High Ambient Temp.
Max. Starting Time (idle at -20°F.
Max. Starting Time (65% power) @ 60°F,
Average Annual Usage
Life Expectancy
Working Fluid Characteristic
Presently available?
Min. Carnot Cycle Efficiency
Min. Ideal Cycle Efficiency
Max. Cost5 to Owner, 5 yr. period
Compatible with Low Cost Materials?
Compatible Lubricant Needed?
Health Hazard
Fire and Explosion Hazard
Environmental Hazard
Goal1
1-stage recip. and
1-stage turbine
1453 HP
1600 Ib
35 cu ft
Nil or small
220°F
-40°F
125°P
25 sec
*J5 sec
350 hr
3500 hr
Yes
Goal1
30%
$100
Yes
Yes - recip.
Maybe - turbine
Nil
Nil
Nil
NOTES:
1. Prime sources: 1971 Vehicle Design Goals (EPA) as amplified in
consultation with EPA
2. Gross expander shaft less feed pump shaft
3- Sundstrand calculation based on performance requirements of
1971 Vehicle Design Goals (EPA)
4. Prime source: contract work statement
5- Initial factory fill plus materials only for any subsequent
refills
6. "Working Fluid"
-------
Table. 2. WORKING FLUID SCREENING CRITERIA
1. Fluid component(s) known and available at reasonable cost.
2. Flow point1 below:
a) 20°C (68°F) for mixture component
b) -29°C (-20°F) for final working fluid
3. Normal boiling point 65 to 120°C (150 to 250°F)
4. Stable in mild steel ampoule test for at least 200 hours
at 720°F.
5. I-Factor2 (at normal dew point) between:
a) 0.65 arid °° for mixture component
b) 0.75 and 1.5 for final working fluid
6. Reference ideal cycle efficiency at least 30% (UAk < »)
7. Equivalent real cycle efficiency - report only - (UAk =
125 Btu/HP-hr-F)
8. Isentropic expansion enthalpy drop no greater than 200 Btu
Ib (turbine expander only, equivalent real cycle)
9. Isentropic expansion density ratio no greater than 25
(reciprocating expander only, equivalent real cycle).
NOTES:
1. Lowest temperature at which fluid will freely flow (melting
po.int in case of pure compound).
2. Parameter related to slope of saturated vapor curve on T-S
diagram (1 = vertical)
-------
Table 3. SUMMARY OF SCREENING TEST RESULTS, CRITERIA 1 to 5
«
Result,
Defined Organic Component Criterion No.
FSN Name I 1 3 1 5
1 benzotrifluoride 0 0 0 X X
2 o-bromobenzotrifluoride 3
3 m-bromobenzotrifluoride 0 0 X X
4 p-bromobenzotrifluoride 0 0 X X
5 2-bromo-l,4-difluorobenzene 00X6
6 o-bromofluorobenzene 0 0 X X
7 m-bromofluorobenzene 0 0 X X
8 bromopentafluorobenzene 0 0 X X X
9 p-bromofluorobenzene 0 0 X X
10 p-bromophenyl trifluoromethyl ether 0 0 X X
11 o-chlorobenzotrifluoride 0 0 X X
12 p-chlorobenzotrifluoride 0 0 X X
13 m-chlorobenzotrifluoride 0 0 X X X
14 o-chlorofluorobenzene 0 0 X X 0
15 m-chlorofluorobenzene 0 0 X X 0
16 p-chlorofluorobenzene 0 0 X X 0
17 chloropentafluorobenzene 0 0 0 X X
18 1,2-dichlorohexafluorocyclobutane 0 0 X X X
19 2,3-dichlorooctafluorobutane 18
20 1,2-dichlorooctafluorocyclohexene-1 20
21 1,2-dichlorotetrafluorocyclobutene-l 0 0 0 X
22 o-difluorobenzene 00000
23 p-difluorobenzene 00000
24 m-difluorobenzene 00000
25 2,4-difluorotoluene 0 0 0 " 0 X
26 2,5-difluorotoluene 0 0 0 0 X
27 2,6-dimethylfluorobenzene 0 0 X 0 X
28 2,3-dimethylfluorobenzene 0 0 X 0 X
29 3,4-dimethyIfluorobenzene X
30 l,3-di(trifluoromethyl)benzene 0 0 0 0 X
31 l,4-di(trifluoromethyl)benzene 0 0 0 X X
0 means pass
X means fail
42 means failed by implication by FSN-42
-------
Table 3. SUMMARY OF SCREENING TEST RESULTS. CRITERIA 1 to 5 (Cont'd.)
Defined Organic Component
FSN Name
32 o-fluoroanlsole
33 m-fluoroanlsole
34 p-fluoroanisole
35 fluorobenzene
36 p-fluorobenzotrifluoride
37 m-fluorobenzotrifluoride
38 o-fluorobenzotrifluoride,
39 fluoroether No. 3
40 2-fluoro-3-methylpyridine
41 p-fluorophenyl trlfluoromethyl ether
42 2-fluoropyridine
43 o-fluorotoluene
44 p-f luorotoluene
4 5 m-f luorotoluene
46 hexafluorobenzene
47 methylpentafluorobenzene
48 octafluorotoluene
49 pentafluorobenzene
50 pentafluoropropanol-1
51 perfluoro(methylcyclohexane)
52 perfluoroalkane-70
53 perfluorotributylamine
54 perfluoro-n-hexane
55 perfluoro-2-butyltetrahydrofuran
56 perfluoro(dimethylcyclohexanes)
57 phenyltrlfluoromethyl ether
58 phenyltrlfluoromethyl sulflde
59 1,2,4,5-tetrafluorobenzene
60 1,2,3,5-tetrafluorobenzene
61 1,2,3,4-tetrafluorobenzene
62 2,3,5j6-tetrafluorotoluene
Result,
Criterion No.
I 2 1 4 5_
0 0 X X X
0 0 X X X
0 0 X X
00000
37
0 0 0 X X
0
X
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
X
0
0
0
0
0
0
0
0
0
X
42
X
X
X
X
0
X
0
0
X
0
X
X
X
X
X
0
X
X
0
X
0 0 0 X X
0 0 X X X
0 0
0
0
0
X
0
0 0
0 0
0 0 OX X
0 0 X X
00000
0000
0 X X
0 means pass
X means fail
42 means failed by implication by FSN-42
-------
Table 3. SUMMARY OF SCREENING TEST RESULTS. CRITERIA 1 to 5 (Cont'd.)
Defined Organic Component
FSN Name
63 trifluoroacetophenone
64 1,2,4-trifluorobenzene
65 1,3>5-trifluorobenzene
66 2,2,2-trifluoroethanol
67 tris(trifluoromethyl)-s-triazine
68 resorcinol bis(trifluoromethyl) ether
69 3-chloropyridine
70 3-methylpyridine
71 pyrazine
72 pyrlmldine
73 toluene
7*1 methanol
75 1-propanol
76 benzene
77 monochlorobenzene
78 pyridine
79 tetrachloroethylene
80 trichloroethylene
81 trichloroethane
82 2,2,3,3-tetrafluoropropanol
83 perfluorosuccinic anhydride
84 perfluoroglutaric anhydride
85 undecafluoropropoxy-2-propanol-l
86 2,4,6-trifluoropyrimidine
87 hexafluoroacetone trihydrate
88 hexamethyldisilazane
89 perfluoroether P-ID (Allied)
90 water
91 acetic acid
92 morpholine
93 ethanol
0 means pass
X means fail
2 means failed by implication by FSN-42
Result,
Criterion No.
1
0
0
0
0
0
X
0
0
0
0
0
0
0
0
0
0
0
0
0
X
0
X
0
2
0
0
0
0
0
0
0
X
X
0
0
0
0
0
0
0
0
0
0
0
3
X
0
0
0
0
X
X
0
0
0
0
0
X
0
X
0
0
0
0
4_
X
0
0
X
0
X
0
0
X
X
0
X
0
X
X
80
X
X
X
5.
0
0
X
0
0
0
0
0
0
0
0
0
0
0
X
0
Corrosive
0
0
0
0
0
0
0
0
0
0
0
0
X
X
0
0
X
0
X
0
0
X
X
X
X
X
0
X
0
-------
Table 3. SUMMARY OP SCREENING TEST RESULTS. CRITERIA 1 to 5 (Cont'd.)
Defined Organic Component
FSN Name
94 neopentanol
95 ethylene glycol monomethyl ether
96 anisole
97 thiazole
98 N-methylmorphollne
99 thlophene
100 1,1,1,2,1,1,1-heptafluoro-2,3,3-tri-
chlorobutane
101 perfluoroheptene-1
102 2,3,5,6-tetrafluorophenol
103 pentafluorophenol
101 N-nonafluorobutyloctafluoromorphollne
105 2-methylpyridine
106 1-methylpyrldine
107 pentafluoropyridlne
108 2-chlorothiophene
109 2,6-dimethylpyridine
110 2-methylpyrazine
111 hexamethyldisiloxane
112 trlmethyl borate
Result,
Criterion No.
0
0 0
0 0
I 5_
0 X 0 X X
0 0 X X 0
0 0 X X
X
0 0 0 X X
000X0
0 0 0 X
0 0 0 X
0 X X
0 X X
0 0 0 X
0 0 X 0 X
00X00
0 0
0 0
X X
X 0 0
0 0 X X
0 0 0 0 X
0 0 0 X
Defined Mixtures
MFSN Components
1 FSN-73 and PSN-78
2 FSN-71 and PSN-90
3 PSN-90 and FSN-93
1 FSN-78 and FSN-91
5 FSN-75, FSN-76 and FSN-90
6 FSN-76, PSN-90 and FSN-93
7 FSN-73 and FSN-75
8 FSN-75 and FSN-76
000
0 0 0 X X
000X0
0 0 0 X X
0 0 0 X
0 0 0 X
0 0 0 X
0 0 0 X
0 means pass
X means fail
142 means failed by implication by FSN-12
-------
Table 3. SUMMARY OF SCREENING TEST RESULTS. CRITERIA 1 to 5 (Cont'd.)
Defined Mixtures
MFSN
9
10
11
12
13-
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
FSN-73
FSN-74
FSN-75
FSN-74,
FSN-73
FSN-71
FSN-78
FSN-23
FSN-30
FSN-61
FSN-23
FSN-66
FSN-70
FSN-71
FSN-90
FSN-90
FSN-90
FSN-90
FSN-46
FSN-46
FSN-70,
FSN-90,
FSN-70,
Components
and FSN-74
and
and
FSN-76
FSN-90
FSN-78 and FSN-90
and
and
and
and
and
and
and
and
and
and
and
and
and
and
and
and
FSN-76
FSN-90
FSN-90
FSN-78
FSN-78
FSN-78
FSN-96
FSN-90 ("Fluorinol")
FSN-90
FSN-90
FSN-105
FSN-106
FSN-109
FSN-110
FSN-107
FSN-49
FSN-90 and FSN-105
FSN-105 and FSN-106
FSN-90 and FSN-106
Result ,
Criterion
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
X
0
0
0
0
0
0
X
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
X
0
0
0
0
0
0
0
0
X
0
0
0
0
0
0
n
X
X
X
X
0
X
0
X
X
X
X
0
X
0
0
0
X
0
0
0
0
0
*
No.
5
0
0
0
X
X
0
0
0
0
0
0
0
0
0 means pass
X means fail
42 means failed by implication by FSN-42
-------
Table 4. THERMALLY STABLE COMPONENTS AND I-PACTORS
PSN Name
56 perfluoro(dimethylcyclohexanes)
89 perfluoroether P-ID (Allied)
111 hexamethyldisiloxane
67 trls(trifluoromethyl)-s-trlazlne
51 perfluoro(methylcyclohexane)
30 1,3-di(trifluoromethyl)benzene
28 2,3-dimethylfluorobenzene
27 2,6-dimethylfluorobenzene
48 octafluorotoluene
105 2-methylpyridlne
25 2,4-difluorotoluene
26 2,5-difluorotoluene
73 toluene
46 hexafluorobenzene
1|9 pentaf luorobenzene
59 lf2,1,5-tetrafluorobenzene
60 1,2,3,5-tetrafluorobenzene
107 pentafluoropyridine
109 2,6-dimethylpyridine
64 1,2,4-trifluorobenzene
65 1,3,5-trlfluorobenzene
35 fluorobenzene
22 o-difluorobenzene
23 p-difluorobenzene
24 m-difluorobenzene
70 3-methylpyrldlne
106 4-methylpyridine
78 pyridine
76 benzene
90 water
tt 0 = pass, X = fail
Stable
Hrs . at
720°P
>336
>336
312
>312
>336
>336
>288
>336
>336
>288
>336
>336
>336
>336
>336
>336
>336
>336
>288
>336
>336
>336
>336
>336
>336
288
240
>336
>336
NBPt
I- Result^
Factor Grit. 3
0.27 0
0.28 X
0.35 0
0.36 0
0.40 0
0.42 0
0.50 X
0.50 X
0.55 0
0.56 X
0.60 0
0.60
0.66
0
0
0.68 0
0.71 o
0.74
0.75
0.75
0.75
0.76
0.76
0.77
0.78
0.79
0.81
0.82
0.82
0.87
0.89
2.81
0
0
0
X
0
0
0
0
0
0
X
X
0
0
0
10
-------
Chemicals in the lower part of Table 4, i.e. the primary auto-
motive candidates, all fall in several chemical categories with
the characteristics shown in Table 5.
Table 5- ADVANCED FLUID COMPONENT CLASSIFICATIONS
I. Aromatics (benzene, toluene, pyridine and
methylpyridines)
A. inexpensive (articles of present commerce)
B. flammable
C. water solubility
1. Benzene, toluene - insoluble
2. Pyridines - soluble in all proportions
D. recognized toxicity
II. Polyfluorinated aromatics (fluorinated benzene, toluene,
pyridine and methylpyridines)
A. inherently expensive, several in limited production
B. Flammability
1. fluorobenzene, difluorobenzenes, trifluoro-
benzenes - flammable in normal air (by flash
point test)
2. tetra-, penta-, hexa-fluorobenzenes, penta-
fluoropyridine, octafluorotoluene - non-
flammable in normal air (by flash point test)
C. water insoluble
D. toxicity not well established, thought to be of
same order as parent aromatics
III. Water
11
-------
Of the simple fluids within the innermost I-factor.bracket in
Table 4, not all meet the Criterion 2 requirement for final
working fluids (flow point below -29°C). Those which do are
listed in Table 6 with their melting points (same as flow point)
and an indication of flammability in normal air.
Table 6. SIMPLE FLUID CONDIDATES
Melting Flam.
FSN Ft. (°C) Test*
60 1,2,3,5-tetrafluorobenzene -48 0
107 pentafluoropyridine -44 0
64 1,2,4-trifluorobenzene -32 X
35 fluorobenzene -42 X
22 o-difluorobenzene -34 X
24 m-difluorobenzene -78 X
78 pyridine -42 X
•flashpoint in normal air; 0 = pass, X = fail
When a requirement for non-flammability is added to Criteria 1 to
5, then, the only simple candidates to emerge are:
l,2,3»5-tetrafluorobenzene, and pentafluoropyridine
While test amounts of these materials are available, their prices
are so high as to suggest that even high-volume production would
not lower them enough to meet owner cost requirements of $100/5 y
supply. This added impetus to a search for mixed (or complex)
fluids to take advantage not only of the wider choice of physical
parameters so afforded, but also of the possibility of lowering
the cost below that of the most expensive component. Water is an
especially attractive component of such a mixture since it can be
expected to reduce flammability and toxicity as well as cost.
Some 31 mixtures were defined at various stages of the search
(Table 3). Most of these were eliminated from consideration when
they failed the Criterion 4 test, or were found to contain com-
ponents which had failed. Two more (MFSN's 1 and 13) are quite
flammable and were eliminated on that basis. The remaining group
is constituted as in Table 7.
12
-------
Table 7. COMPLEX FLUID CANDIDATES
1. hexafluorobenzene with
a. pentafluorobenzene (MFSN-28)
b. pentafluoropyridine (MFSN-27)
2. water with
a. pyridine (MFSN-15)
b. 2-methylpyridine (MFSN-23)
c. 3-methylpyridine (MFSN-21)
d. 4-methylpyridine (MFSN-24)
e. binary combinations of b, c and d
(MFSN-29, 30, 3D
Each of the three methylpyridines violates either the I-factor
or boiling point criterion. Retention of these as candidate
components is justified primarily on the grounds that they are
the only water-soluble organics (other than pyridine itself)
found to have reasonable thermal stability.
Predictive computer routines were used to enable cycle evaluations
of the advanced candidates of Tables 5, 6 and 7. Their functions
are listed in Table 8.
These routines were used extensively to analyze the behavior of
advanced candidates in the "Reference Ideal Cycle", defined as
follows:
Reference Ideal Cycle - An ideal Rankine cycle (clockwise
rectangle on a pressure-entropy diagram) operating between
a top (bulk fluid) temperature of 712°F and a bottom tem-
perature of 220°F (saturated liquid at condenser outlet)
and with the peak fluid pressure chosen as the lesser of:
a. that just preventing expansion penetration into the
2-phase region;
b. 1000 psia
13
-------
Table 8. COMPUTER ROUTINES EMPLOYED
1. Estimation of pure compound
a. critical constants (P, T, V)
b. perfect vapor heat capacity coefficients
c. vapor pressures
d. liquid densities
e. enthalpies of vaporization
2. Calculation of liquid activity coefficients from vapor
pressure data on particular complex fluids.
3. Calculation of vapor phase properties by a special general-
ization of the Redlich-Kwong equation of state to cover com-
plex as well as simple vapor mixtures.
4. Prediction of principle PVT/thermodynamic functions of either
simple or complex fluids, using as input, data found in the
literature, or generated as in 1 to 3 above, and having as an
option (when treating simple fluids) the use of the Hirschfelder,
Buehler, McGee and Sutton formulation.
5- Calculation of specified ideal or real Rankine cycles (only
with the Hirschfelder et al. formulation).
6. Plotting of temperature-entropy diagrams for any fluid processed
in i».
-------
Reference ideal cycles calculated for many of the Table 5-7
candidates are compiled in Table 9. The first column is a test
of Criterion 6, which requires an ideal cycle efficiency of at
least 30% at any level of regenerator effectiveness. All straight
organic fluids pass, whereas the pyridine-water mixtures fail.
While water itself appears disqualified, it is capable of higher
efficiencies by raising its upper operating temperature.
Other entries in the table demonstrate compliance with certain
interim (non-criterion) goals; namely,
1. condenser pressure from 5 to 50 psia
2. ideal expansion density ratio less than about 40
(to serve in a single expansion reciprocator)
3. ideal expansion enthalpy drop less than about 200
Btu/pound (to serve in a single-stage impulse turbine)
4. I-factor in the approximate range of 0.65 to » com-
ponents) or 0.75 to 1.5 (final fluids)
The ideal Rankine cycle with a fully effective regenerator is
only of limited applicability in evaluating working fluids,
especially as concerns cycle efficiency. A 100% effective re-
generator is by definition infinite in size. Since the space
available for an automotive regenerator is severely restricted
in the engine compartment, and since the cycle efficiency is
highly sensitive to extent of regeneration (as well as to ex-
pander efficiency), a more realistic "Equivalent Real Cycle"
was defined as follows:
Equivalent Real Cycle - A real Rankine cycle which matches
the Reference Ideal Cycle at the two extreme corners (ex-
pander and pump inlets), has engine and pump efficiencies
of 75% each, has pressure losses in the regenerator, vapor
generator and condenser representative of those predicted
by system contractors (in January 1972), and has an adiabatic
pump and expander. The "size" of the regenerator may be
limited by restricting its q/At (log mean) ratio (or equiv-
alently its UAk product) to a "reasonable value".
Equivalent Real Cycle calculations employing many of the same
simple fluids recorded in Table 9 are compiled in Table 10. Com-
parison of the entries with the reciprocating/turbine suitability
Criteria 8 and 9 reveals all fluoroaromatics and benzene suitable
for either engine type and all aromatics but benzene suited only
for turbine use.
15
-------
Table 9. SUMMARY OF REFERENCE IDEAL CYCLE CALCULATIONS ON THERMALLY STABLE CANDIDATES
Aromatic3
Benzene
Toluene
Pyridine
3-Methylpyrldine
Fluoro Aromatlcs
Fluorobenzene
m-Dlfluorobenzene
p-Dlfluorobenzene
1.2,3,5-Tetrafluoro
benzene
1,2,4,5-Tetrafluoro
benzene
Pentafluorobenzene
Octafluorotoluene
Pentafluoropyrldine
Hexafluorobenzene
Water
Pyridine/Water Solutions
6056
30$
60%
30%
60%
Pyridine
2-Methylpyridine
3-Methylpyrldine
30%
60% f4-Methylpyridine
Ideal
Cycle
Efflc.E*
31.8
31.7
31.5
32.1
31-9
31.6
31-7
30.6
30.8
30.3
30.2
30.3
31.1
20.5
27.4
28.3
27.9
27-9
27-3
27.7
27.6
28.6
Condenser
Pressure
(psia)
29.4
12.3
10.5
4.2
25.9
28.5
23.6
28.2
22.8
26.6
11.9
28.2
30.5
17.3
21.3
19.6
22.8
21.6
19-3
15-9
13.9
14.3
Expansion
Density
Ratio
40.3
145-9
145.1
338.1
42.4
5.8
Expansion
Enthalpy
Drop
(Btu/lb)
81.3
79-3
93-9
97.3
13-0
226.0
I-Pactor
at
220°F
0.85
0-75
0.98
0.89
46.6
43-7
54.6
70.1
56.9
59-1
0.65
0.70
0.71
0.76
54.7
15-2
86.9
41.6
40.3
44.8
35.3
29-7
38.2
35.2
0.77
0.68
0.55
0.75
0.54
3.34
36.3
47.1
33.8
49.4
39.1
65-4
40.1
76.4
154.9
125.9
160.8
107.3
167-1
112.3
165-5
114.8
1.66
1.33
1.48
0.89
1.75
1.21
1.73
1.20
* Full regeneration on all simple fluids: full regeneration or regeneration
only to the dew line, whichever occurs first, on all solutions.
** Concentrations shown are mol percents.
16
-------
Table 10. SUMMARY OF EQUIVALENT REAL CYCLE CALCULATIONS ON
SIMPLE THERMALLY STABLE FLUIDS
Aromatics
benzene
toluene
pyrldine
3-methylpyridlne
Fluoro Aromatics
f luorobenzene
m-dif luorobenzene
p-d if luorobenzene
1 j 2, 3 > 5- tetraf luorobenzene
pentaf luoropyridine
hexaf luorobenzene
Notes: 1. Regenerator Q/AT(log
Real
Cycle
Effic.(
22.8
22.7
23-7
24.8
22.1
21.9
22.3
21.2
20.9
20.7
mean)
Expansion
Density
%1} Ratio2
15-1
46.4
13-5
83.6
17.0
16.2
19-5
15.7
15.4
15-2
= UAk = 125
Enthalpy
Drop,( Btu/lb)
57.4
56.5
67.3
70.2
49-5
40.2
41.8
30.4
27-0
24.8
Btu/HP-hr-°F
2. Isentropic, from engine inlet to exhaust enthalpy
3. Condenser pressures
Table 9
and 220
°F I-factors
same as in
17
-------
A searching examination of the complex fluid candidates listed in
Table 7 was carried out to further limit the number of final can-
didates. This included experimental studies of flammability,
toxicity and freeze (flow) point, in addition to the development
of criteria data discussed above.
The first choice was obvious. Of the nonflammable fluids, only
the highly fluorinated benzenes stand a reasonable chance of
approaching the cost goal in large-scale production. Hexa-
fluorobenzene is too high melting to be used alone, but mixtures
with tri-, tetra- and pentafluorobenzenes possess the requisite
low flow points. Such mixtures actually result from the fluor-
ination of benzene, so processing economies could be expected.
As a particular model representing this class, the first final
Rankine candidate fluid, designated RC-1, was fixed as follows:
RC-1
Concentration
Constituent Mol Wt Mol '% Wt %
pentafluorobenzene 168.1 60.0 57-5
hexafluorobenzene 186.1 40.0 42.5
The only obvious way to avoid using the expensive fluoro compounds
and still secure some measure of fire hazard protection is to
employ an organic-water solution. Only the pyridines, with their
water solubility, permit this approach. Mixtures involving two
or more of the three methylpyridines and pyridine itself, all con-
templated in this category, offer some advantage in processing
economics because of the simultaneous production of all four in
practice. In consideration of freeze (flow) point depression,
cycle utility, and flash/fire points, it appeared that a mixture
of 2-methylpyridine in water was the best binary choice, so as
the second final candidate the representative model defined as
follows was chosen:
RC-2
Concentration
Constituent Mol Wt Mol %
water 18.0 65.0
2-methylpyridine 93-1 35-0
18
-------
Thorough experimental and computational characterization of the
two final candidates followed their definition. Physical property
measurements included flow points, vapor pressures, liquid densities3
specific heats, viscosities and thermal conductivities. Selected
values appear in Table 11. Computations using these and pure com-
ponent literature data then provided such derived properties as
vapor viscosities and thermal conductivities, ideal vapor heat
capacity coefficients, and liquid activity coefficients.
Thermodynamic tables and T-S diagrams were next prepared, and the
two principal Rankine cycles were calculated, as summarized in
Table 12. It is significant to compare the cycle efficiencies of
the two fluids. While RC-2 fails the 30$ ideal cycle efficiency
stricture, it actually shows a slightly higher real cycle effi-
ciency than does RC-1. The main reason is that the fixed "size"
regenerator is more effective in recovering the lesser superheat
of the RC-1 fluid. This supports the original use of the I-factor
screening criterion.
Application-oriented test work was also undertaken to further
explore candidate utility. Included were stability-compatibility
tests in dynamic pumping loops and steel ampoules, animal acute
toxicity experiments, flammability tests, lubrication studies, and
manufacturing cost estimations. Highlights are listed in Table 13-
Fluid RC-2 was found to be incompatible with SAE 4130 steel boiler
tubing at 720°F. Failure of the steel to passivate was evidenced
by formation of hydrogen gas and magnetic iron oxide. An attempt
to prepassivate the steel with water/steam reduced, but did not
stop, the water-iron reaction.
An idea of the inhalation toxicity hazards of the two final can-
didates was garnered from comparisons of the rat LC50 results with
(a) expected inhalation doses in a hypothetical "worst case"
Rankine automobile accident involving driver entrapment and fluid
release, and (b) known inhalation accidents involving common vol-
atile fluids. The inhalation hazards of both candidates are con-
sidered comparable to those of gasoline. Products of flame de-
composition of each fluid were identified.
Attempts to find lubricants with all characteristics desired in
the reciprocating engine application were largely unsuccessful.
Desired characteristics were (a) stable, compatible, and a liquid
at 712°F; (b) incompletely soluble in the working fluid at ambient
temperature; and (c) solidification point (in suspension in working
fluid) below -20°F. All lubricant candidates known to be stable
above 700°F were completely miscible in RC-1 and most were highly
soluble in RC-2. A number of highly silylated benzenes were syn-
thesized which had minimal solubility in RC-2 and thermal stability
comparable to the RC-2. These compounds suffered from inherent
expense and too-high solidification points.
19
-------
Table 11. SELECTED PROPERTIES OF FINAL CANDIDATES
Final Candidate
Property
Average molecular weight
Flow point, °F
Normal boiling point, °F
Condensing pressure
@ 220°F, psia
Heat of vaporization
6 NBPt, Btu/lb
I-Factor at 220°F
Liquid density at
77°F, g/ml
Specific heat at
77°F, Btu/lb
RC-1
175.3
172
30
79-1
0.72
1.549
0.290
RC-2
200
21
378
1.38
0.985
0.668
20
-------
Table 12. FINAL CANDIDATE CYCLE CALCULATION SUMMARY
Final Candidate:
Cycle1 :
Cycle Efficiency
Per 100 Net Cycle HP
Fluid rate, Ib/hr
Eng. exhaust, cfm
Pump rate, gpm
Pump HP
Heat flows, kBtu/hr
Vapor gen.., kBtu/hr
Regen., kBtu/hr
Conden. , kBtu/hr
Engine, % Ef f ic . 2
Pressure ratio
Density ratio
Isentropic 3
Exhaust quality, %
Enthalpy drop, Btu/lb
Pump, % Effic.2
Pressure ratio
Regenerator, % Effect.
UAK, Btu/HP-hr-°F
Log mean temp, diff., °F
Notes: 1. "Reference ideal"
earlier
2. Input specified
3. From engine inlet
RC-1
Ideal Real
30.0 21.0
7470 11500
259 3^0
12.2 18.8
7.34 16.8 -
848 1220
391 692
409 95.0
100 75
36.6 29.8
44.6 37.9
44.6 15.4
176 190
36.6 25-8
100 75
36.6 42.1
95-3 .91.8
12.0 12. 52
RC-2
Ideal Real
28.8 22.2
1760 2520
233 310
3.93 5-66
2.43 5-04
883 1150
64.8 213
629 890
100 75
43.6- 35.4
35 -.6 32.9
35.6 12.1
108 118
149 106
100 75
43.6 50.1
100 98.4
12. 52
32.6 55-6 0 23.0
and "equivalent real" as defined
condition to exhaust enthalpy
21
-------
Table 13- PERFORMANCE PARAMETERS OF FINAL CANDIDATES
Candidate
Stability. Compatibility
1000-hour dynamic loop test
Failure mode
Toxic Hazards
Inhalation
4-hr rat LC50, ppm
30-min rat LC50, ppm
Oral, rat, LD50, mg/kg
Dermal, rabbit, LD50, mg/kg
Irritation score
Rabbit skin (8 max.)
Rabbit eye (110 max.)
Flammability
Flash point, modified COC, °F
Fire point, modified COC, °F
Auto ignition temp., °F
Hot compartment ignition
rating (benzene3!*!)
Cost (to owner, 5 yr supply), est.$
RC-1
Pass
RC-2
Fail
Decomposition
16,000
40,0002
12,000
>8,000
3
10
8,000*
-
8101
2501
5.6'
711
e
e
e
0
112
130
145
1,060
8
<100
Notes:
Experiments conducted with a 40/60 mol % mixture of
2-methylpyridine/water
Experiment conducted with a 50/50 mol % mixture of
penta-/hexafluorobenzene by Imperial Smelting Corp-
oration
22
-------
A low cost synthetic fluid - biphenyl/terphenyl eutectic - was
evaluated as a miscible lubricant in friction-wear experiments
involving cast iron rubbing against various typical mating
materials. This material has promise as a miscible lubricant in
either RC-1 or RC-2. No petroleum-based lubricants can be ex-
pected to survive the high temperatures demanded in this appli-
cation.
Fairly detailed economic studies were carried out to project the
ultimate cost to the driving public of the RC-1 fluid. Present
manufacturing techniques are not suited to high volume low-cost
production. A considerable technological investment would be
necessary to bring the cost to the level indicated in Table 13-
The RC-2 fluid is inherently cheaper and is currently abundant.
Ancillary experimental studies were carried out to probe the
ultimate thermal stability of "Fluorinol 85" with and without
added refrigeration oil. The upper temperature limit of these
fluids was placed near 650°F. Another study of a class of semi-
organic heat stable fluids called "carboranes" revealed a
flammability deficiency.
The two final candidate fluids were subjected to a Rankine Cycle
Engine Optimization Analysis where the size, weight, and efficiency
of an engine for each of these fluids was determined. Fluid RC-1
was used to optimize an engine with a turbine expander and fluid
RC-2 was used to optimize an engine with a reciprocating expander.
An optimum automotive engine was determined by using a criterion
that weighed engine size and weight against engine overall efficien
(miles per gallon) over a time-weighted average of five operating
load points representing country, suburban, and city driving con-
ditions .
The Sundstrand Rankine Cycle Design Optimization Program used is
a comprehensive analytical design procedure in which detailed
design point and off-design calculations are made, and components
meeting cycle requirements are calculated. In the design procedure
a design point cycle is first calculated, from which engine com-
ponents are designed. The performance of the system using these
components is then evaluated by means of off-design cycle calcula-
tions. The overall system and its performance is then evaluated
against an optimizing criterion, the results of which are used to
vary the inputs to the design point cycle analysis. A new design
point cycle is then calculated, components designed, and off-design
performance evaluated. This procedure is repeated until an optimum
system, based on the optimizing criterion, is found.
23
-------
The first step in defining an optimum engine is to define the load
points for which the engine is to be optimized, and to define the
parameters of the optimizing criterion. For the purposes of this
study only steady state load points were to be used, and the
optimizing criterion was a trade-off between system efficiency
and system weight. The optimizing criterion was based on the
fact that high-efficiency systems tend to be high-weight systems,
and reduced-weight systems result in lower efficiencies. As
system weight is directly related to system volume, system size
is also being optimized by this criteria.
The load points used in the analysis consisted of one design point
load and four off-design load conditions. These load points were
derived from the Federal driving cycle and EPA's prototype vehicle
performance specification.
The load points selected were:
Speed Power % Time
Design Point maximum power acceleration 70 mph 108 HP 0.5
Off-design 1 steady speed 60 mph 32 HP 16.5
Off-design 2 steady speed 30 mph 8.4 HP 33-0
Off-design 3 acceleration 25-1 mph 9-92 HP 28.0
Off-design 4 acceleration l.?4 mph 1.143 HP 22.0
The power is the power output from the transmission to the drive
shaft, the engine power required being determined by the mechanical
losses, accessory requirements and the cycle calculations. These
load points were also used to calculate the fuel load required for
a 200 mile range.
The optimizing criterion was constructed using miles-per-gallon as
representative of engine efficiency, weighed against system weight
in pounds. In the optimizing process a weighting factor was varied
to place more emphasis on system efficiency or system weight to
see how this affected the system configuration. The optimum con-
ditions found in each calculation are listed in Tables 14 and 15. "
The turbine engine was optimized as a supercritical engine assuming
a turbine inlet temperature of 712°F (required for a 200°F con- v
denser in order to meet the EPA specifications of 42% carnot cycle
efficiency), and a 425 psia turbine inlet pressure. This pressure
was selected to keep pump work to a minimum since the fluid prop-
erties did not indicate any significant benefit in increased
adiabatic head or reduced regenerator size by going to a higher
pressure. The turbine tip speed was limited to 1200 ft/sec to
allow economical materials, and the condenser assumed was the
AiResearch design for the EPA Thermo Electron engine.
24
-------
Table 14. PRINCIPAL RESULTS_OF__ENGINE OPTIMIZATION CALCULATIONS
Engine Type Working Fluid
Weighting Factor, Effic. (Vs Weight)
Overall Miles/Gallon
Weights, Pounds
Expander and Gear Box
Condenser and Pan
Regenerator
Burner and Vapor Generator
Transmission and Drive Train
All Other Components and Fuel
Total System
Volumes, Cubic Feet
Expander and Gear Box
Condenser and Fan
Regenerator
Burner and Vapor Generator
Transmission
Unoccupied Space
Total System
Turb.
RC-1
0.5
173
195
380
355
322
1571
Turb.
RC-1
0.97
178
172
234
368
355
319
1626
Recip.
RC-2
0.5
13.33 13.81 18.89*
225
142
36
281
355
322
1361
0.23
7-55
4.13
9.58
1.50
15.^1
38.3
0.28
7.71
5-01
8.83
1.50
15.57
38.9
2.25
5-56
0.82
9.79
1.50
13-3
33-2
Based on 3% of stroke clearance volume and square wave
operation of valves porting 20$ of piston area
25
-------
Table 15. OPTIMIZED DESIGN POINT PARAMETERS
Engine Type Working Fluid
Weighting Factor, Ef f Ic . (Vs Weight)
Condensing Temperature, °F
Regenerator Press. Drop/Press. In, %
Regenerator Effectiveness, %
Condenser Effectiveness, %
Condenser Air Power, kW
Expander Speed
Tip Speed, ft/sec
Pitch Diam. , in.
RPM (3 in. stroke)
Admission
Arc/Total Circum. , %
Intake Angle, Degrees
Flue Gas Press. Split (Vap. Gen/
Total) %
Turb.
RC-1
0.5
179
1.27
93
81
5-48
677
4.96
-
7.7-1
-
80.4
Turb.
RC-1
0.97
179
1.27
95
81
4.93
677
6.00
-
100.
-
89.4
Recip
RC-2
0.5
242
1.43
87
80
5.0
1900
54
72.6
26
-------
The optimization analysis for the turbine engine was performed
over six values of the optimization weighting factor (which can
vary from 0 to 1; the higher the value, the greater the emphasis
on high efficiency). The result of this analysis showed that there
was no appreciable variation in the weight of the engine or in
miles per gallon over the.range of the weighting factor. It was
also seen that there was little variation in the design point
condenser temperature, which varied from 179°P to 200°F. It
appeared from an analysis of the results that this situation was
the result of the restriction of condenser size to one fitting
the frontal region of a normal engine compartment. The size
restriction on the condenser so severely restricts the condenser
size for the power level required that essentially the same size
condenser results with all values of the optimization weighting
factor. As a result, the condenser influence on the rest of the
power system is such that there is very little variation in the
optimum engine weight and efficiency throughout the range of the
optimizing weighting factor.
The reciprocating engine was analyzed as a subcritical engine
using RC-2. The design assumed an expander inlet temperature of
712°F and a 700 psia inlet pressure for the same reasons as in
the turbine engine, and to keep cylinder bore diameter a reasonable^
size for an automotive engine. The reciprocator was a single
acting, single expansion engine whose speed was limited to 2000
rpm to keep valve train forces within reasonable limits. The
condenser core was that of the AiResearch design developed for
the Thermo Electron engine.
As can be seen by the miles per gallon, the reciprocating engine
shows very encouraging results. However, it must be taken into
consideration that the simplified reciprocator model used assumed
a three percent of stroke clearance volume and "square wave"
valve operation that had flow areas equal to one-fifth the piston
area. These rather optimistic assumptions result in low valve
losses and high volumetric efficiency which may be difficult to
attain in an actual engine. The result of these assumptions is a
high efficiency expander which results in the high miles per
gallon.figures. It should be borne in mind when comparing these
figures to that of the turbine, that the turbine expander model
used was a well developed turbine analysis that gave realistic
results based on current state-of-the-art technology.
CONCLUSIONS
1. No organic or aqueous organic fluids were found to completely
satisfy all the requirements established in the contract for
Rankine cycle automotive power plant working fluids.
27
-------
Fluid RC-1 comes nearest to satisfying the automotive organic
working fluid requirements and goals set by AAPSDD. It meets
all of the established performance criteria for both turbine
and reciprocating engines, including efficiencies, temperature,
pressure and density limitations, and regenerator size.
Additionally, it is considered fire-safe and unlikely to cause
serious injury on accidental human exposure (subject to
verification in extended toxicity testing recommended as a
follow on). The fluid is quite stable at temperatures to at
least 720°F in low-cost system materials, and it is expected
to last for at least the 5-year span of initial vehicle owner-
ship. It possesses a distinct advantage over water in not
freezing at temperatures down to -4o°F. A low-cost, soluble
lubricant is available if needed. The main contractual
deficiency of the fluid is its projected cost of no less than
$112/over its expected five year life, as compared to a $100
contract limit. To arrive at the $112 cost it was necessary
to assume (a) successful advancement of the state of fluor-
ination technology beyond that now practiced, and (b) the
existence of a 50 million pound per year benzene fluorination
plant. Current prices of the higher fluorinated benzenes are
many times the projected price.
Turning to cycle practicality, the comparatively high flow
rate (Table 12) of the RC-1 fluid per net cycle horsepower
(related to its higher molecular weight) may be considered a
disadvantage despite the lack of a contractual limit. All in
all, this fluid is worthy of further development.
Fluid RC-2 is a back-up candidate for automotive use, either
in a turbine or reciprocating engine. It satisfies all per-
formance criteria except the 30% minimum ideal cycle efficiency.
Even so, the fluid is capable of high efficiency usage in a
real cycle with a compact regenerator. Although the fluid is
flammable, it is less so than common fuels. In burning it
releases considerably less energy on account of its water con-
tent, and its products of combustion are not hazardous to
humans. The fluid itself is unlikely to cause serious injury.
It is marginally stable at ?20°F in steel. SAE 4130 steel is
not a compatible material for boiler tube construction with
RC-2. The fluid acts chemically to corrode this steel, forming
F6304 and hydrogen. A more suitable steel alloy needs to be
found before this fluid can be considered a contender with
RC-1. Assuming a satisfactory steel were found, the RC-2 fluid
would cost the vehicle owner about $100 in the first five years
of vehicle life. This presumes three system refills. Tradi-
tional lubricants cannot be used with this fluid above ?00°F,
but low cost synthetics offer promise when mutual solubility
is permissible. Synthetic silane lubricants may be considered
if immiscibility is mandatory, but further development is
needed. Fluidity above -40°F is another feature of importance.
The prime ingredient of the fluid is commercially available at
28
-------
low cost now, but in lesser quantity than would be commanded
by a million cars-per-year market. RC-2 is worthy of further
development only in the unlikely event that candidate RC-1
must be abandoned.
^. Many organic fluids possess the state and thermodynamic prop-
erties needed to attain engine performance goals.
5. Very few fluids possess the requisite thermochemical stability
in steel at 712°F (and higher). Those that do fall in these
categories:
Aromatics
Fluoroaromatics
Aromatic tertiary amines
Perfluoroaliphatics
Water
6. Flammability eliminates the aromatics from contention, and
places the aromatic tertiary amines in a contingency category.
Since the aromatic tertiary amines (or pyridines) are quite
flammable and water soluble, their safe usage in automobiles
is contingent upon a suitable reduction in flammability
effected by mixing with water.
7. An excessive demand for regenerator space to maintain tol-
erable cycle efficiency rules out the perfluoroaliphatics.
This is clearly shown by the I-factor criterion and the
supporting cycle calculations.
8. Additional restrictions on vapor pressure, melting point
(or "flow" point), cost, and toxicity further narrow the
choice to
mixed fluorobenzenes and
aqueous pyridines,
from which categories the final candidates RC-1 and 2 were
chosen.
9- The search rationale adopted at the outset was basically
sound, although some adjustments were necessary. Particu-
larly valuable features were:
a. an efficient thermal stability screening program
b. development and effective use of the I-factor con-
cept for thermodynamic screening - including con-
firmation via I-factor - efficiency correlations
c. maximal use of computation throughout, but especially
in these areas:
29
-------
prediction of thermodynamics of non-ideal/
ideal mixtures
I-factor and physical property predictions
ideal and real cycle calculations
d. the concept of progressive advancement of candidates,
exposing each to increasingly rigorous examination.
10. As shown by the optimization studies, both the RC-1 and RC-2
final candidate fluids are appropriate fluids for automotive-
size engines and are conducive to good efficiencies for
automotive application.
RECOMMENDATIONS
Relative to fluid RC-1, it is recommended that
1. it be considered an acceptable organic fluid for a
prototype engine development program.
2. static thermochemical stability tests be performed
to learn how far above 720°F the fluid might be
driven
3. providing prototype experience is favorable, the
fluid be subjected to a carefully planned set of
exposure tests in the areas of
chronic inhalation
chronic dermal contact
environmental fate
as necessary to qualify the fluid for widespread use
4. encouragement be given to prospective manufacturers to
develop the necessary low-cost processes for large scale
production.
Relative to fluid RC-2 and assuming abandonment of RC-1, it is
recommended that:
5. additional work be undertaken to find a suitable steel
and/or steel passivation technique to avoid the iron-
water reaction at high temperature
6. if appropriate, additional synthesis work be done to
finalize an immiscible silane lubricant
30
-------
7. a fire hazard simulation be performed - preferably in-
volving a mock-up burner-vapor generator and a simulated
high pressure fluid leak
8. the first three recommendations under RC-1 be followed
9- additional engine optimization studies be carried out
using computer models of components which more closely
represent that for automotive type advanced state-of-the-
art designs. These models would give a better descrip-
tion of component weight, size, effectiveness, and
efficiency.
10. further engine optimization studies be performed that
optimize more parameters than were done for this pro-
gram. A more detailed optimizing criteria would also
be beneficial in determining an optimum engine.
31
-------
TECHNICAL REPORT DAYA
(Please rrcJ IniLn.ctiuns o;i //(..• ri true beJ
1. REPORT NO.
APTD -
1563
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Optimum Working Fluids for Automotive Rankine Engines
:Volume I- Executive Summary
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHORlS)
D.R." Miller, H. R. Null, Q. E Thompson
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation, SOONorth Lindbergh Blvd
St. Louis, Missouri 63166 and Sundstrand Aviation
Rockford, Illinois
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68 - 04 - 0030
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Mobile Source Air Pollution Control
Advanced Automotive Power Systems Development Division
Ann Arbor, Michigan. *~48105
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The objective was to determine the best working fluids for Rankine cycle automotive
ppwer plants. This study program was initiated to search out the best fluids currently
available for the automotive Rankine engine application. Specific guidelines were
established for the fluid screening process. These guidelines covered all significant
fluid characteristics: availability, compatibility, thermal stability, cycle efficiency
cold start ability, applicability in principal engine types, owner cost, toxicity,
flammability and environmental hazard. This Volume i serves as an executive introductidjr
and summary. The best Rankine candidate fluid, designated RC-1, was fixed as follows.:
pentafluorobenzene 60 Mol%, hexaflurorbenzene 40 Mol%. The best binary choice, RC-2
was fixed as follows: water 65 Mol%, 2-methylpyridine 35Mol%. Fluid RC-1 was used to
optimize an engine with a turbine expander and fluid RC-2 was used to optimize an
engine with a reciprocating expander. The optimization studies indicate that both the
RC--1 and RC-2 final candidate fluids are appropriate fluids for automotive-size engine
and are conducive to good efficiencies for autonotive application. Additional conclu-
sions and recommendations are given. The results of the study are summarized and
presented in fifteen tables.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lOENTIFIERS/OPEN ENDED TERMS C. COSATI 1'iclU/GfoUp
Air Pollution
Rankine cycle
Engines
Design
Turbines
Reciprocating Engines
Working Fluid
Mixed Fluorobenzenes
Aqueous Pyridines
13 B
21 G
21 E
14 B
7 C
13. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF HAGtS
36
20. SECURITY CLASS (This page)
«-s Unclassified
22. PRICE
EPA Form 2220-1 ,-V-73)
32
------- |