United States
            Environmental Protection
            Agency
                 Office of Mobile Source Air Pollution Control
                 Emission Control Technology Division
                 2565 Plymouth Road
                 Ann Arbor, Michigan 48105
EPA 460/3-88-002
   July 1988
f/EPA
            Air
Near-Cylinder Dissociation
of Methanol for
Automotive Application

-------
                                        EPA 460/3-88-002
     Near-Cylinder Dissociation  of
Methanol for Automotive Application
                       by

           Engine and Vehicle Research Division
              Southwest Research Institute
                 6220 Culebra Road
               San Antonio, Texas 78284

               Contract No. 68-03-1984

           EPA Project Officer: Robert I. Bruetsch


                    Prepared for

         ENVIRONMENTAL PROTECTION AGENCY
        Office of Mobile Source Air Pollution Control
           Emission Control Technology Division
                 2565 Plymouth Road
              Ann Arbor, Michigan 48105
                    July 1988

-------
     This  report  is  issued  by  the Environmental  Protection  Agency  to  report
technical  data  of  interest to  a limited  number of  readers.    Copies are available
free  of  charge  to  Federal  employees,  current   contractors  and  grantees,  and
nonprofit   organizations   -   in limited   quantities      from  the  Library  Services
Office,  Environmental  Protection  Agency,   2565  Plymouth  Road,  Ann  Arbor,
Michigan  48105.
     This  report  was  furnished  to  the  Environmental   Protection  Agency  by
Southwest  Research Institute,  6220 Culebra Road, San  Antonio, Texas,  in  fulfill-
ment of Contract  No. 68-03-1984.   The contents of  this report are  reproduced
herein  as  received from  Southwest  Research Institute.   The  opinions,  findings,
and  conclusions  expressed  are those  of  the  author  and  not necessarily  those  of
the  Environmental  Protection  Agency.  Mention  of  company or product  names  is
not  to  be  considered  as  an  endorsement  by  the  Environmental  Protection
Agency.
                        Publication No. EPA-460/3-88-002
                                          11

-------
                                  FOREWORD
     This  project  was  initiated   by  the   Control  Technology  Assessment  and
Characterization Branch  of  the Environmental Protection  Agency  in  September,
1984.   The  effort on  which  this   report is  based was  performed  by  the  Engine
and  Vehicle   Research  Division of Southwest  Research Institute  and  was  com-
pleted in July, 1988.

     This  project  was  identified   within  SwRI  as  03-8352,  and  the  Project
Managers,  in chronological  order,  were  Mr.  R.   G. Phatak,  Dr.  Robert  Thring
(acting),  Mr.  Gerald  Driscoll,  and  Mr.  Gary  Stecklein.   This  final  report  was
compiled  by  Mr.   Roy Hull  and  Mr.  Charles  Urban  and was  reviewed by  Mr.
Charles  Urban and Mr.  Gary Stecklein.  Mr. Charles  Urban also  provided  tech-
nical  review  of  the heat  exchanger subsequent  to  evaluation  and  recommended
design   improvements.    The  Project  Officers at  the  Environmental  Protection
Agency   over the  course of  this  project  were  Mr.  Robert  Wagner,   Dr.  Karl
Hellman (acting), and Mr.  Robert Bruetsch.
                                        in

-------
                                    ABSTRACT
     This   report   describes   the   design  analyses  and   experimental  evaluation
toward  developing  a system  of dissociating  methanol  for  use in  a spark-ignited
passenger  car  engine.    Maximum  dissociation  of methanol  fuel  attainable  using
exhaust  gas heat  was  the  basic   goal  of  this  project.    The heating  value  of
dissociated  methanol  is  about  20  percent higher  than  that  for  liquid  methanol.
Therefore,  significant  energy  savings  appear  to be  attainable  if  a  large  percent
of the  engine  fuel needs  can be dissociated  by  use  of  the otherwise wasted heat
of the  exhaust  gas.   A  heat  exchanger  consisting of  a methanol  evaporator,  an
injection  system  to  provide  fuel   to  the evaporator,  and  a  dissociation  catalytic
reactor  was designed,  fabricated and  tested.   The dissociation  efficiency  of  this
initial  reactor  was  not as  good as expected.   The  relative  poor performance  of
the  dissociation  reactor   apparently   was  associated  with   fuel  injector  nozzle
heating,  blockage  of passageways   in  the dissociation  reactor,  and  an  ineffective
dissociation  catalyst.    The   heat   exchanger  was  redesigned  to   incorporate  im-
provements, and four heat exchangers were fabricated using the final design.
                                         IV

-------
                            TABLE OF CONTENTS

                                                                          Page

FOREWORD	iii

ABSTRACT	iv

LIST OF FIGURES	vi

LIST OF TABLES	vi

SUMMARY	vii

I.     INTRODUCTION	1

         A.  Project Objectives	1
         B. Background Information	1
         C. Scope of Work	3
         D.  Methanol Dissociation Heat Exchanger.	      .3
         E. Methanol Fuel	4

II.    HEAT EXCHANGER   INITIAL DESIGN	5

         A.  Fuel Injection System	5
         B. Evaporator and Reactor	9
         C. Catalyst	15

III.   EVALUATION OF HEAT EXCHANGER	      .18

         A.  Test Engine	      ...  18
         B. Heat Exchanger Test  Results	     	   21
         C. Analysis of the Catalyst	23

IV.   VALVE TRAIN AND CONTROL  SYSTEM CONSIDERATIONS	27

         A. Valve Train Considerations	27
         B. Control System Considerations	     	   27

 V.   HEAT EXCHANGER   FINAL DESIGN	29

VI.   CONCLUSIONS AND RECOMMENDATIONS	31

REFERENCES

APPENDICES

-------
                               LIST OF FIGURES


Figure                                                                      Page

  1     Spray Pattern of Methanol Injector	8

  2     Injector Modifications to Obtain Spray Pattern	8

  3     Heat Exchanger Outlet Shell and Exhaust Manifold	12

  4     Heat Exchanger Shown with Reactor at Left and
        Evaporator on the Right	12

  5     Junction of Evaporator and Reactor with Evaporator
        on the Right	13

  6     Top View of Evaporator Cover	13

  7     Reactor End View showing Dissociated Product Outlet
        and Exhaust Gas Inlet	14

  8     Test Engine Setup to Evaluate Heat Exchanger	19

  9     Heat Exchanger Performance Versus Methanol Flowrate	24

 10     Cylinder Fuel Needs Versus Methanol Flowrate	25



                               LIST OF TABLES

Table                                                                       page

  1     Design Specifications for SwRI Methanol
        Dissociation Heat Exchanger	11

  2     Initial SwRI Test Data on NAPS-Z Engine as Received	20

  3     Summary of Data from Engine Test on  NAPS-Z Engine
        as Received	       .20

  4     Data from Evaluation of Methanol Dissociation Heat Exchanger	21

  5     Methanol Mass Flowrate Versus Percent Dissociation	22
                                        VI

-------
                                    SUMMARY
      A  program was  conducted  by SwRI to  analyze, design, and  test a  methanol
dissociation  heat  exchanger  for  automotive  applications.    The  major  objective
for  methanol  dissociation  is  to  provide   improved  engine  operating  efficiency
compared  to  conventional  operation  with   liquid methanol.    In-cylinder  dissoci-
ation  of methanol  was  initially  desired  but was shown early in  the  program to
be  unachievable.    An  alternative  approach  was  taken  to  locate  the  heat  ex-
changer  near  the  exhaust  port  and  to  inject   the  products of dissociation  into
the  combustion chamber without  prior  mixing  with combustion  air.    Maximum
dissociation of methanol from  available  exhaust heat was a primary goal.

      A  gain  in  engine  thermal  efficiency  is  possible  in  theory,  because  the
exhaust  heat  can  be  used   to   dissociate  methanol into  hydrogen  and  carbon
monoxide  according  to  the  reaction  CH3  OH	-*~2H2  +  CO.   Through  this
dissociation  process,  the  energy  content  (lower  heating  value) of  liquid  methanol
is increased  about  20  percent.   Mathematical and  design  analyses were  made of
catalytic  reactor  and  evaporator  sections  for   use   in a   dissociation  heat  ex-
changer.    These  analyses   were  based   on  using   a  Nissan  NAPS-Z  two-liter
engine.   A  counterflow design   was  chosen,  whereby  the  exhaust gases  flowed
first  over  the catalytic reactor  and  then   over  the   evaporator.  This  counterflow
design  was  selected  over  crossflow   or  parallel-flow systems  as  being  the  most
efficient for the dissociation process.

      Based  on the analyses,  a heat  exchanger approximately 52 cm  (20.5  inches)
long  by 9.5  cm  (3.75)  inches  outside diameter,  including  reactor  and evaporator,
was  designed  and  fabricated.    The  Johnson Matthey  Catalytic Systems Division
was   selected  to  provide   the  catalysts  for  the dissociation  reactor.    For  the
methanol dissociation  catalyst,  a  proprietary base  metal  catalyst   (Type H)  was
used.    For  the  exhaust-side catalyst, precious  metals  (platinum  and  palladium)
were used.  The fuel system for injecting liquid methanol into the evaporator
consisted of  a  modified  Stanadyne  DB2 distributor-type  pump and  an  injector
designed and fabricated by SwRI.

      Performance  evaluations of  the  prototype  dissociation  heat  exchanger  were
conducted  at  selected engine speed  and  load conditions  using  the exhaust from
one  cylinder  of  the  NAPS-Z  engine.   This engine was  utilized  as  an  exhaust
generator,  and  the  products  of dissociation  were not  used as fuel  for the  engine
in  these evaluations.    Dissociation  rates  in the  reactor  ranged  from   1   to  96
percent  of  the  total  methanol  injected  into the evaporator;  the   higher  dissoci-
ation   rates  being  associated  with  low  methanol  injection  rates.    This  repre-
sented from  0.6 to  9.8  percent  of  the  engine  cylinder  fuel   needs  at the  test
points  evaluated. The  evaporator  was adequately  sized to  vaporize  100  percent
of the engine  cylinder fuel needs at 3000 rpm,  but  only  50 percent at 1500 rpm.
The  vaporization rate  was greater  at 3000 rpm because  of  the   higher  exhaust
temperature and flowrate at that engine speed.

     Design improvements  were  made to the dissociation  heat  exchanger.   These
included a  relocation  of  the  methanol injector,  a  modification  to  the  flow areas
in the  reactor,  and   the  addition  of fins  to   the  evaporator.    Four  heat  ex-
changers were  fabricated to  the   final design specifications, and provided  to  the
EPA.

                                        vii

-------
                                I.  INTRODUCTION

      This  section   discusses  the  objectives,  background,   and  scope  of  work
associated with development of a methanol dissociation heat exchanger.

A.  Project Objectives

      The  original goal  was to provide  reactors  inside  the  cylinders  of a metha-
nol  engine  (in  a pre-chamber) to  vaporize  and  dissociate a significant portion of
the  total engine  fuel  needs.   The dissociated  methanol  was  to  provide improved
operating  efficiency  compared  to conventional  operation  with  liquid  methanol.
At  the  outset  of  the  program,  the  in-cylinder  dissociation  was  determined  not
possible  given  the  limited  volume  (hence limited  catalyst area), high temperature
(potential  catalyst  damage),  and   presence  of  oxygen   and  combustion  products
(catalyst  oxidation  or  poisoning)  that are  typically  found  in  combustion  cham-
bers.

      The project efforts were then directed toward the following objectives:

      0     Design,  fabricate  and  test  a  heat  exchanger   for  "near-
           cylinder"  dissociation  of methanol.    This  was  to  utilize   a
           heat  exchanger   for  each   cylinder  and discharge  the   dis-
           sociation products injected directly into  the cylinder.

      0     Design and  fabricate  engine  modifications  to accommodate
           direct  injection   of the  heat  exchanger  products  into  the
           cylinder.

      0     Procure  or  fabricate a  fuel  injection  system  to  inject  and
           meter  fuel into the heat exchanger.

      0     Based  on  the   analysis  of  test  results,   modify  the   heat
           exchanger design to  optimize its performance.

           Fabricate  four   new   heat  exchangers  to   the  optimized
           design.

Design  efforts  in several  areas  were  more  extensive  than originally  planned,  and
major problems  were  experienced with  the engine  to  be  used   in testing  the
methanol dissociation heat  exchanger.    In order  to  concentrate  remaining  efforts
toward   development  of  the  dissociation heat   exchanger,  program  requirements
were subsequently  redirected  to  omit the  modifications to  the  engine  (controls
and  intake  valve)  and  the actual operation  of  the  engine  on  the   products  of
dissociation.

B.  Background Information

      It  is  possible  to  enhance  the energy  content of  liquid methanol  by endo-
thermically   reacting  it  over  a  catalyst   to  form  gaseous  hydrogen   and  carbon
monoxide.   Dissociated  methanol  is  methanol  decomposed  by  one  of  two pro-
cesses.    The first  approach  separates neat  methanol  into  hydrogen  and  carbon
monoxide primarily in a reaction as follows:

-------
                              CH3 OH—*-2H2 + CO

The second  approach  uses  an  equimolar  solution  of methanol and  water  and is
known   as  steam   reformation.    This   method   can   be  demonstrated   by  the
following reaction:

                          CH3 OH + H20—»-3H2 + C02

Even  though  steam  reformation   provides  more  hydrogen,   dissociation  of  neat
methanol  provides  for  greater  improvement  in  thermal  efficiency  than  that
provided by steam reformation.

     There  are  several  reasons  for  the   interest  in  using   dissociated  methanol
versus neat  methanol  in internal  combustion  engines.   The  primary  reason is  to
reduce  specific fuel consumption.   This can  be   accomplished because the lower
heating  value  (LHV)  of the  dissociated products (H2  and CO)  is about  20  per-
cent  higher  than  that  of  liquid   methanol and 13  percent   higher  than  that  of
vaporized  methanol.   In  addition,  the  engine  can  operate  at very  low  equiva-
lence  ratios  with  the  hydrogen-rich  dissociated methanol  as  compared  to liquid
methanol.    A second reason  for  using  dissociated  methanol  is  that  harmful
emissions  are  reduced.   Hydrocarbon and aldehyde  emissions are very  low  with
dissociated methanol versus liquid methanol.   Oxides  of  nitrogen are  also  lower
with  dissociated methanol   when  operating  at  low equivalence ratios  in  recipro-
cating  engines.

     The  following comments  on  emissions are based on ideal mixtures of disso-
ciated  or  reformed products of  methanol.   Actual emissions  levels would  proba-
bly  be  between  those  produced on liquid  methanol  and those produced  on these
ideal mixtures.

     Hydrocarbons  -  Most  of  the  literature  suggests  that hydrocarbon  (HC)
emissions  should  be  very   low with  either  reformed  or  dissociated  methanol  as
compared  to  liquid  methanol.   This  is because there are no hydrocarbon  com-
pounds  in  either  reformed  (3H2  + CO2)  or  dissociated  (2H2  + CO)  methanol.
Engines  operating  on  mixtures  of H2 and CO  or  H2  and  CO2 generate  hydro-
carbons  only  from the lubricating oil  that  enters  into  the   combustion  chamber
past the rings  or  valve  stem  seals.    Thus,  hydrocarbon emissions  when  using
dissociated  methanol   are   a   function  of  engine   condition,  design  and  load.
Hydrocarbon  emissions from an  engine  fueled  partially  with  dissociated  methanol
should at  least  be  less than the HC emissions  from  the same engine  fueled  with
liquid methanol.t1)*

     Oxides  of  Nitrogen  -   NOX  emissions  are  low   at very  lean  equivalence
ratios  (low load)   when  operating  on  dissociated methanol.  (1>2)   No  data were
found  in  the  literature  for  engine  operation  with  dissociated  (or  reformed)
methanol at  high  loads.   Operation  on dissociated methanol  at  high  loads, with
equivalence ratios  of  about  1.0, should produce somewhat higher NOX  emissions
compared to  liquid methanol  because  of the  higher  flame temperature  of the H2
+ CO mixture. (2)
      Superscript  numbers  in  parentheses  designate   references  at   the  end  of
this report.

-------
     Carbon  Monoxide  -   At  very   low equivalence  ratios,  CO  emissions  with
dissociated  methanol  tend  to   be  somewhat  higher  than  with  liquid  methanol,
because  of  passthrough  of  unburned  CO.    At  higher  equivalence  ratios,  CO
emissions are typically  less with dissociated methanol. (2'

     Aldehydes -  Aldehyde emissions are  not  readily  formed  from  the primary
methanol dissociation   products:  hydrogen and  carbon  monoxide.   Because  disso-
ciation  to  H2  and  CO  is not  complete  and  other products  of  dissociation  have
not been adequately defined, however,  effect  of  dissociation  on aldehydes  is not
currently known.

C.  Scope of Work

     This  program was  directed  toward dissociation  of  methanol in  the  com-
bustion  chamber  or in  a  reactor located  close  to  the combustion  chamber,  so
the  products  of  dissociation could  be  directly  introduced  into  the  combustion
chamber.  Three  technical  tasks were specified and included the  development  of
design  specifications,   procurement  and/or  fabrication  of  hardware  and   engine
components,  and   engine  testing  followed  by  system   design  improvements.    A
fourth task delineated the program reporting requirements.

D.  Methanol Dissociation Heat Exchanger

     The term "heat  exchanger",  as  applied in this report,  refers  to  the  assem-
bly into which liquid  methanol is  injected  and out of  which dissociated products
(2H2 + CO)  are discharged.   The  heat  exchanger  is made up of two  components.
These   are   the  "evaporator"  into  which  the  liquid  methanol  is injected  and
vaporized,  and the  "reactor" where   the  vaporized methanol  is  superheated  and
dissociated into hydrogen  and carbon monoxide.

     Heat exchanger design objectives  were to:

     0     Make maximum  use  of exhaust  heat down  to  about 400°F
           for  dissociation  in the  reactor  section  and  use  heat  below
           400°F for vaporization in the evaporator section.

     °     Design   for  maximized  dissociation   of  methanol  under all
          engine operating conditions.

          Accommodate  transient  engine operating conditions through
          the  use  of  an auxiliary fuel  system.   This  auxiliary  system
          would fuel  the  engine  during start-up  and would  provide
          supplementary  fuel when the amount of dissociated fuel was
          insufficient for the power requirements.

     Heat exchanger design considerations  included the following:

          The  reactor design  analysis  was  to  be  directed  toward plate/fin  and
          concentric tube  designs  with  a  counterflow configuration  to provide
          maximum heat  transfer area  for  a given   mechanical  volume.    The
          design was to be  based on analyses  of steady-state  conditions.

-------
     "     The  possibility  of  using fins  on  both  sides  (exhaust  and  methanol)
           was to be investigated.

     0     Depositing  catalyst  materials on both  sides  of  the  reactor  was  to be
           investigated.    Catalyst   material  on  the  exhaust  side  could  increase
           reactor  temperature through  oxidation  of  any  HC and  CO  present  in
           the exhaust.

     0     Vaporization  of  methanol  in  the  evaporator  was  to occur  under all
           operating  conditions utilizing  the  residual  exhaust  heat  energy   after
           the exhaust gases have passed through the reactor.

     0     Pressure  drop  across  the  reactor  and  transient  response  of the  heat
           exchanger  under  a range  of  engine  operating  conditions  were  to be
           calculated.

E.  Methanol Fuel

     To meet future energy needs, a synthetic  fuel derived  from domestic   coal,
natural gas, and  biomass would  be especially attractive  if the use  of such fuel
also led  to increased  engine efficiency  and  improved  exhaust emissions.    One
favorable  candidate   that  has  emerged  from  extensive   engine   testing   in  the
U.S.A.  (3)  and  Germany  (4)   is methanol.   Methanol can  be made  from synthesis
gas  using  readily-available   catalysts  and  reactor  designs.    Based  on  thermal
cycles  (5'6),  methanol appears to  be  the most  cost-effective liquid fuel that  can
be  made from  coal.   Projections (7)  indicate  that  methanol  could  become econo-
mically competitive with gasoline.

     The  production of  methanol  from synthesis  gas  is  an  exothermic  process
(21.68  kCal/mole)  by the reaction  CO +  2H2	^CH3OH,  which is carried out in
a  catalyst  bed  at  relatively  low temperatures  and  high pressures.   Use of  waste
heat  from  a combustion engine  to dissociate methanol  back  into carbon mono-
xide  and  hydrogen,  an endothermic  process,  provides  an opportunity  to  recover
some of the energy lost during the coal to methanol conversion.

     Although  methanol  has  several  good  combustion  qualities,  namely,   high
octane  number,  high  flame   speed,  and  broad  ignition  limits, it  also has   some
less  desirable  properties such as  high  heat of  vaporization,  a tendency  to  pre-
ignite, and low  heat  of combustion.   Problems  with neat methanol  include  cold-
start  ignition,  cylinder  wear, and  vehicle  range.    To alleviate difficulties  with
cold-start  ignition,  the  volatility  of methanol  is  increased  by  adding  components
such  as  gasoline  or dimethyl ether.   A  system  using  exhaust heat  to dissociate
methanol   would  not  enhance  cold-start  ignition   in  an  engine,  but it   could
improve  thermal  efficiency  and reduce  exhaust emissions.   The   main advantage
of methanol dissociation is that it  provides  a  method of  utilizing  waste heat and
thus  the   potential  for  reducing  the  specific  fuel  consumption  of  the  spark-
ignition engine. (8~u)

-------
                   II. HEAT EXCHANGER - INITIAL DESIGN

      The  term "heat exchanger", as used  in  this  project, includes the evaporator
into  which  liquid  methanol is  injected,  and  the  reactor where  the  methanol  is
dissociated into hydrogen and carbon monoxide.

A. Fuel Injection System

      The  fuel injection  system,  consisting of  a  pump and a suitable  injector for
each  evaporator  chamber,  must  be  capable  of  delivering the  correct  amount  of
fuel  in a  spray  pattern   that  evenly  covers  the  walls of  the  evaporator,  to
assure  rapid  vaporization.    The  pump  can  be  a multi-cylinder,  distributor-type
pump capable of  supplying  fuel  to  each individual  injector.   Compatibility  with
methanol fuel is  a  requirement  for both  the pump  and  the  injector  as  well  as
any  connecting lines, fittings  and gaskets.   Pump and nozzle  wear is a  concern
if the  pump and  nozzles are  designed  for diesel fuel but have  to handle metha-
nol.   Materials changes are required to avoid wear and corrosion.

      To provide  the greatest  possible  vaporization, there  is a  need to  spray the
walls of the  evaporator  with a forceful impingement  (to  minimize  boundary  layer
effects)  of small  droplets  over the entire  surface  of  the  evaporator (to minimize
evaporator volume).

Development  Effort -  An  initial  survey  of  SwRI  engineers  having  experience
with  methanol-fueled  engines and  fuel  injection  systems  in   general  failed  to
disclose  any  suitable pumps  or  injectors  that  would  meet  the requirements for
the  fuel  system.    Commercial suppliers  of  pumps  and  injectors   were not  sur-
veyed at that time.  SwRI experience had  shown  that  use of  neat  methanol  in
standard diesel injectors and  pumps results in early failures  of these  components
due  to  lack of lubrication.   Therefore, an in-house  effort was  made  to  design  a
pump and injector system based on the SwRI experience.

      This  fuel injection  system  design  effort  was directed  towards   providing  a
system that would:

           Deliver  from 0.2 to 11 kg/hr (0.5  to 24.5  Ibm/hr) of  metha-
           nol per cylinder.

      0     Deliver fuel evenly from each injector.

      °     Atomize the fuel sufficiently to promote rapid vaporization.

     0     Provide  injectors  to operate at internal  evaporator  temper-
           atures  of  up  to  126°C (259T) and  external exhaust  temper-
           atures   up  to  177°C  (350T),  and  include  a  check   valve
           feature  to minimize vapor lock.

     °     Provide  pump and  injectors  to  generate  and  withstand  fuel
          pressures up to 2100 kPa  (300  psig).  This  value  is the  sum
          of  the  maximum reactor  operating  pressure  of  700 kPa  (100
          psig) and the  maximum expected  pressure  drop  across the
          fuel nozzle  required  to  achieve  good atomization  of  1400
          kPa (200 psig).

-------
     0     Assure that the  injector  produces a  methanol  spray  pattern
           that  maximizes  coating  the  heated  inside   walls   of   the
           evaporator.

     0     Provide  pump  and  injectors  to  perform  reliably and with
           sufficient  durability without the  inherent  lubricity provided
           by petroleum fuels.

     Initial  effort  resulted  in  a  workable  system  that  produced  good  spray
atomization and  distribution  at  high flowrates,  as observed  within a  clear  plastic
tube using  Stoddard solvent.   Testing of  the  SwRI  designed pump  and  injectors
is  described  in Appendix  A.  The pump  had two separate pistons,  each driven on
the opposite  side of  the  same  eccentric.    The  outlet from  each piston was  sent
to two  of the four nozzles.   Bench  tests  were  performed  using  the  pump  (driven
by  an  electric  motor) connected to  four  nozzles.    Each of  the  nozzles  was
installed in  a clear  plastic  tube  with  an  inside  diameter  of 7.9  cm  (to  simulate
the  3.1   inches  diameter  of  the evaporator).     Each  tube  had  provisions  for
collecting  and  draining  the  sprayed  solvent,  so  that  the quantity  of  solvent
sprayed  from  each  nozzle could  be  determined  over a  test  period  and compared
to each  other.   This bench  testing showed that the  initial design had two  major
shortcomings: (1) lack  of  balanced  fuel  delivery  between   nozzles  at  all  flow-
rates, and (2) poor  atomization  quality at  low pump speeds (low  flowrates).  In
order to  overcome   these shortcomings,  a  second generation pump  was  designed
using the  nozzles  and  pump eccentric housing  from the  first  generation design.
The second  generation  pump  design  (shown in Appendix  A) uses  a single  piston
that  operates  at constant  speed  to  provide  good   atomization  at  all  flowrates.
Changes in flowrate  were accomplished by adjusting the  pressure regulator.

      Subsequent  to the in-house  pump and injector  design  efforts,  a search  was
conducted for commercially available  injection  systems  that are  compatible with
methanol.   Three  manufacturers  were  contacted:  Stanadyne,  American Bosch, and
Robert  Bosch.   Each was asked  if one of their  production  fuel injection  systems
(pump  and   nozzle)  or   a  modified  production  system  would  operate  on  neat
methanol.    Stanadyne responded with  assistance  applicable   to  the  specific pro-
gram needs,  since  they  had  experience  in operating  their distributor-type  pumps
with  methanol.   There  is an  inherent  advantage in using  a  distributor pump over
an  in-line pump in that  no  other  source of lubrication  is  required.   An in-line
pump  typically uses engine oil  for lubrication  in  the cam area,  but  a  distributor
pump  uses only  the fuel  it  is supplied with  to  lubricate the internal  pump  parts.
Thus,  if a  distributor  pump can be  successfully  used  with methanol,  contami-
nation of  the engine oil  by methanol  and  contamination of  the  fuel  with engine
oil  can  be avoided.  Engine oil  in  the methanol could have a  deleterious effect
on  the  dissociation  catalyst.    Internal  changes  to  the  distributor  pump  were
necessary,  however,  to  avoid  wear and corrosion when  using methanol  instead of
diesel fuel.

      At this point  in the program there  were  three  possible directions  to pursue
to obtain a suitable fuel injection system. They were to:

           Use internal design effort  on  the  second generation  pump.
           This  concept  was  designed   to  provide  a   balanced  fuel
           delivery  between    each  of the  nozzles   and  to  give good
           atomization.    It  was   not  a  proven  concept,   so  there  was
           risk that it would not achieve these goals.

-------
            Use  a  completely  commercial  based (Stanadyne)  system  for the
            pump  and  nozzles.     This  could  be  either   an  off-the-shelf
            unmodified  system  that  has  demonstrated  reasonable  life   using
            neat  methanol,  or a modified  commercial system  that  has demon-
            strated  its  durability.    Some  changes  to  the commercial  nozzles
            would probably be required to obtain the  desired spray pattern.

      0     Use  a  combination of  commercial  pump  and  the existing
            nozzles.   This approach  could  be of  value if  the existing
            nozzles  worked well with the commercial  pump.

 Each  of these  approaches  would  require  some  effort  to  adapt the  pump to  the
 engine and to the control system.   The  third approach (i.e.,  the  combination  of
 a commercial  Stanadyne  DB2  pump  and  existing SwRI  designed  injectors)  was
 chosen at that time.

      There was some  difficulty  in  assessing the  actual durability  of standard or
 modified  commercial  pumps  operating on  methanol.    Though there  were  some
 published   data  discussing  hardware  C12-14),  that  typically   was  not the  primary
 purpose for  the publication.   Published  data did not  involve  extensive  operating
 hours   and  discuss  only  limited  experience.    Discussions   with   manufacturers,
 likewise, indicate that their experience was  limited.

      The  SwRI designed  injector  performed satisfactory  in  evaluations at  room
 temperature  but  did  not  perform  satisfactorily   at  elevated   operating  tempera-
 ture,   In   addition,  it  was  considered  desirable to use injectors  that  were more
 readily available.   The  final system  utilized a commercially  modified  fuel injec-
 tion  pump  and modified  commercially available  injectors;  these  units  are  des-
 cribed as follows:

 Pump  - The Stanadyne  DB2 distributor type pump is  used  commercially  on small,
 high-speed diesel  engines.   To enable the  pump  to  safely  operate on methanol,
 the following pump modifications were made by Stanadyne:

      0     The  housing  was  nickel  plated  to  retard  corrosion  of  the
           aluminum by methanol.

      0     Ceramic rollers were used in place of steel.

           Tool  steel  was  used  instead  of  sintered  steel  for transfer
           pump liners and rotor.

     °     The  governor  section   and  the  valve  cage  were  nickel
           plated.

Injectors  -   The  Stanadyne  Model  14 (or  CVA  Model  5760  403)  poppet nozzle
injector  was  modified  to  provide  the wide spray  pattern  desired  for  the  heat
exchanger  application.    The  spray  pattern  obtained  is  shown in  Figure  1,   and
the  modifications  to  the  Stanadyne  Model  14  injector  necessary  to achieve  this
pattern is shown  in Figure 2.

-------
                           5"
                                   \
                                       \
                                          \
 L

                                                        MODIFIED  STflNRDYNE
                                                         MODEL  14  INJECTOR
         FIGURE 1. SPRAY PATTERN OF METHANOL INJECTOR,
    SEE BODY
 MODIFICRTION OETRIL
     RDD PIECE TO VRLVE
   SNR6E  OVER SURFRCE 'B*
RND HHCHINE TO SPECIFIED 5HHPE
                                                           BLEND PIECE TO VRLVE SEHT.
                        BODY
                        8.178 REFERENCE - RCTURL
                        DIRMETER DETERMINED BY
               16        42 INCLUSIVE RNGLE TO EDGE
                        OF VflLVE SEBT CONTRCT HREH
       NOTE

       USE STRNRDYNE MODEL 14
       OR CRV MODEL 5768 483
       POPPET NOZZLE INJECTOR
                                                  VRLVE
                                                   CUT OFF BODY PORTION HS SHOHN
                                 INCLU
                                                      •VRLVE SERT CONTRCT RRER

                                              BODY  MODIFICflTION DETRIL
FIGURE 2.  INJECTOR MODIFICATIONS TO OBTAIN SPRAY PATTERN.

-------
B.  Evaporator and Reactor

      The  initial  intent  of  this  program  was  to  have  the  dissociation  of  the
liquid methanol  occur  in  the  combustion chamber.   In  the report Design  Specifi-
cations for  In-Cylinder Dissociation  of Methanol  (Appendix  B),  SwRI  concluded
that  there   would  be  several  disadvantages  to  locating  the  catalyst within  the
combustion  chamber.   As stated  in  the  design specifications,  the major limita-
tion  was  that  the  evaporator  and  reactor  (catalyst)  volumes  required  would  be
greatly in   excess  of  space  available in  the  combustion  chamber.    Additional
stated disadvantages of in-cylinder dissociation were:

           Contamination  of  the   catalyst   with  combustion   products,
           rendering it useless in a short period of time.

      0    Heating of  the  catalyst  would  primarily occur  during  com-
           bustion, rather  than from  waste  heat  in  the  exhaust,  and
           that results in lower cycle efficiency.

           The    necessity   of  supplying   methanol,    and   therefore
           dissociation,   during  the   intake  cycle,  resulting  in   low
           volumetric efficiency of  the engine.

      0    Premature  combustion of  dissociated  methanol,  resulting  in
           loss  of engine  efficiency  and  damage to  the  engine's struc-
           ture.

      These  limitations  were  eliminated  by  (1) separating the  catalyst  chamber
from  the engine's combustion  space  and (2)  locating  the catalyst chamber  in  the
exhaust stream  so  that  waste  heat  from the  exhaust gases  can  readily  be  used.
However, such  design  would  still  be  a  departure  from previous  work  of others
in  that the  dissociated gases would  be inducted directly into  the cylinder rather
than  through the  intake  manifold  with combustion air.   Products  of dissociation
were  to  be  introduced  into the cylinder through  a separate induction valve.   To
conserve heat, the heat exchanger  was to  be located as  close as  possible  to  the
exhaust port.   Maximum  dissociation of methanol was  the goal,  and the  guide-
lines to be followed in designing the heat exchanger were:

      0     Use  a  counterflow  heat  exchanger,  finned   on both  sides
           with  either a plate-fin or concentric  tube configuration.

      0     Provide sufficient  area  to  handle  maximum  heat flux  from
          the exhaust above 205°C (400°F.)

     °    Catalyze both sides of the heat exchanger.

     0    Investigate,  by  analysis,  the  transient behavior  of  the  heat
          exchanger.

     0    Construct  the exchanger  so  that low  temperature exhaust is
          used  for  vaporization, with the higher  temperature  exhaust
          being used for dissociation of the vaporized methanol.

-------
     °     Design  the  heat  exchanger  for  steady state  conditions,  and
           provide  separate  fueling  provisions  for  start-up  and  other
           transient conditions.

     Design analyses  of  the  heat  exchanger  were  performed  by  SwRI and   a
consultant.  Analysis   of   the  evaporator  by   the   consultant  is   contained  in
Appendix  C.   Initial heat exchanger analysis  by  SwRI  (Appendix  D)  covered both
the  evaporator  and the  reactor.    After reviewing  the  analyses  and the  related
literature,  an  evaporator  design  was  selected,   and  a  decision  was reached  to
perform a series  of tests on a single  unit  on the  engine to determine the  actual
performance of  the  evaporator  and  the  reactor.   Design  specifications  for the
heat exchanger that was  fabricated  for testing  are  listed in  Table  1.   The heat
exchanger  was  fabricated  by  SwRI  using  catalyst substrates  provided by  Johnson
Matthey.    Figures  3-7  are  photographs  of  the  heat  exchanger  showing  the
assembled  heat exchanger and its component  parts.    These  figures  are  described
in the  following paragraphs:

      Figure 3 -       The  heat exchanger  outer  shell  is shown  attached  to the
                      exhaust  manifold.   Ports  in   the  outer  shell  near the  ex-
                      haust  manifold  were  used  to  measure  exhaust  gas  tempera-
                      ture and pressure as  it enters the  heat  exchanger.    Other
                      ports  in the outer  shell were  used to  measure  exhaust gas
                      temperatures  and  pressures  at  locations  between  the  reac-
                      tor  and  evaporator and at  the exit  of  the evaporator.  The
                      tube extending from the right  side of  the outer shell  is for
                      discharging exhaust gases.

      Figure 4 -       The   heat  exchanger   is  shown   with   the   reactor   (where
                      dissociation  takes  place)  at   the   left   and  the  evaporator
                      (where methanol vaporization takes place)  to the right.

      Figure 5         A  closeup  of the  junction  of  the  evaporator  and  reactor is
                      shown  with  the  evaporator  to the  right.   The  single  ex-
                      haust  annulus around  the evaporator is  visible  as  are  three
                      of  the  four tubes  that  port  the   vaporized  methanol  from
                      the  evaporator  to the  reactor.    The  outer  shells  of  both
                      the   evaporator  and   reactor   were  made  from 0.127  mm
                      (0.005-inch)  thick  stainless  steel  and served  only  to  retain
                      the  metal substrate during fabrication.

      Figure 6 -       This  view  shows  the top  of  the  evaporator   cover.   The
                      nozzle is in  the  center,  with  its  fuel line extending  upward
                      and its spring tension adjustment on top.

      Figure 7         The  reactor  exhaust  inlet  and  dissociated  methanol  outlet
                      are  shown.   The  manifold connects  the  two methanol annuli
                      and  has a flared connection that  permits attachment  to the
                      outlet  port  welded  to the  outer  shell.     The  metal  catalyst
                      substrate can be seen in the three exhaust  annuli.
                                         10

-------
                    TABLE 1.  DESIGN SPECIFICATIONS FOR

              SwRI METHANOL DISSOCIATION HEAT EXCHANGER
Dimensions
              Heat Exchanger Housing:
             Inside Housing: Reactor -
                         Evaporator

        Connection Between Evaporator
                         and Reactor:
 58  cm  (23 inches) long by 9.5 cm (3.75
 inches) in diameter

 20.4 cm (8.05 inches) long
 10.2 cm (4.0  inches) long
3.8 cm (1.5 inches)
Remainder  of the length inside  housing is  for the fuel injection  nozzle, exhaust
discharge, and a manifold for the dissociated products.
Design Flowrates
                         Exhaust Side:     64  kg/hr  (141  Ibm/hr)  with less  than
                                          25 mm (1.0 in.) Hg. pressure drop

                            Methanol:     0.2  to  11  kg/hr  (0.5  to  25  Ibm/hr)
                                          injector rating
Materials

Housing and all structural parts are 304 L stainless steel.

Reactor Catalyst Details

         Dissociation Catalyst Material:
               Apparent Surface Yield:

                 Actual Surface Yield:


                Specific Surface Area:

                        Bulk Density:

                  Washcoat Thickness:

            Exhaust Catalyst Material:
Johnson  Matthey Type H base metal (2
annuli)

3.9 kg/hr-m2 (0.8 lbm/hr-ft2)

          kg/hr-m2  (3.7xlQ-4  lbm/hr-
ft2)

125 m2/g

0.5 g/cm3

0.05  mm

Johnson  Matthey  precious  metal;  9  Pt:
1 Rh (3 annuli)
                                       11

-------
  FIGURE 3.    HEAT EXCHANGER OUTER SHELL AND EXHAUST MANIFOLD.
FIGURE 4.    HEAT EXCHANGER SHOWN WITH REACTOR AT LEFT  AND
            EVAPORATOR ON THE RIGHT.
                              12

-------
FIGURE 5.    JUNCTIONS OF EVAPORATOR AND REACTOR WITH
            EVAPORATOR ON THE RIGHT.
      FIGURE 6.    TOP VIEW OF EVAPORATOR COVER,
                         13

-------
FIGURE 7.    REACTOR  END VIEW  SHOWING  DISSOCIATED PRODUCT
            OUTLET AND EXHAUST GAS INLET.
                               14

-------
 C.  Catalyst

      The design criteria specified for the reactor catalysts were:

      0    Provide  adequate   heat  at  the  required   temperature   to  the  catalyst
           active sites.

      0    Provide sufficient catalyst surface area.

      0    Have  low  pressure  drop,  allowing  the  product  gases  to  charge  the
           engine cylinder in the time  available.

      0    Provide  thermal   control  to   maintain  the  catalyst   in  its   optimum
           temperature  range, or  at  a  minimum,  provide safeguards  from  over-
           heating  the  catalyst  material,  which would result in  permanent  dam-
           age.

      0    Resist the potentially corrosive gases.

      0    Resist the  vibration  levels of  an automotive  engine operating  at  high
           speed.

      0    Provide  low  thermal  inertia,   thus enabling   the  reactor  output  to
           follow changing engine speed-load conditions.

      Design  concepts  included  (1)  the  use  of   the exhaust  gas  as  the  total
 source  of heat  and  (2)  the  use  of  the   exhaust  gas and  engine  coolant.   The
 first   concept  would   be  applied   by  locating  the  evaporator  and  the  catalytic
 reactor directly  in  the  exhaust  stream.   The  second  approach  would use  engine
 coolant  to  provide  heat  to  the evaporator,  and  the  catalytic  reactor would  be
 located in the exhaust stream.   The advantages of  using the engine  coolant  are
 that the  vaporizer  design could be compact due  to  the   liquid-to-liquid heat  ex-
 change  and   that  some  of  the  coolant  heat,  otherwise   rejected,   would   be
 recovered.    Disadvantages  of  using  the  engine  coolant  are  a  more  complex
 system,  relative   to  using exhaust  heat  only,  and  a  longer time  for  warm-up,
 because  the  coolant and  engine  block  must achieve  operating temperature  before
 vaporization  could  occur.   The  time delay  in producing quantities of  vaporized
 methanol,  which would then  be fed to  the  reactor,  could  result  in  overheating
 of  the reactor due  to lack  of  methanol  vapor.   It  would  thus  be  necessary  to
 add a  device  to divert  exhaust from   the  reactor to  prevent overheating  during
 warm-up.   Because of  the long  warm-up  period and  added complexity,  the  use  of
 coolant for vaporizing  methanol was  rejected  in favor of using exhaust gas only.

     The  catalytic  reactor design  was  proposed  by  SwRI  to  be  of  cross-flow
 tube  design  with  methanol  flowing  through  tubes  coated  on the  inside  with  a
 catalyst.    This  design  was  later  replaced by a  counterflow  design,  in  which
 coaxial tubes  coated  with catalytic material would handle exhaust gas  flowing  in
 one  direction and methanol  or methanol vapor  flowing  in the other.   The coun-
 terflow design  was considered to  make  more  efficient use  of the  exhaust heat,
offsetting  the   added  difficulty   involved  in  fabrication.     This   counterflow
arrangement  has  been  shown  in   previous figures.    To maximize  heat  to  the
reactor,  catalyst  material  was also  used on  the exhaust  side  to  increase  reactor
temperature  through  oxidation  of  any CO and  HC  remaining   in  the  exhaust
gases.

                                        15

-------
      Three  catalyst  manufacturers  were  contacted  for  assistance  in  the  analysis,
design,  and  fabrication  of  the  catalytic   reactors.    These   manufacturers  were
Johnson Matthey (JM), Conoco, Inc., and W.R. Grace & Co.

      Conoco supplied copies of patents  and  a report on their previous  catalytic
dissociation  efforts.    Catalyst  yields  were   not   given   in  the  report  provided.
The  patents  described the  catalyst  composition,  for  each of  the  patent  appli-
cations,  the  catalysts  were  applied  to  pellets  in  a  fixed  bed.   Based  on  the
geometric  surface area  of  the  pellets  used   in  their  supported  bed  simulation
unit,  a  representative from Conoco  calculated  yields between  0.5 to  2.9  kg/hr-
m2 (0.1 to  0.6  lbm/hr-ft2) for  idle  to  full  load, respectively^15)   The product
yield  values were  based  on  the throughput  of methanol at 85%  dissociation.   The
size  for  a  single reactor  for   the  engine  would  be  approximately  8x18x38  cm
(3x7x15 inches).   Conoco's  choice of  catalyst  was based on  overall  yield,  yield
at  low  temperatures, long-term  stability,   thermal stability,   and  physical  inte-
grity.   It  was  mentioned  that  none  of  the better  catalysts used  at Conoco  were
substantially better (e.g.,  2  times)  than  the others.   Conoco  was  not  selected  to
supply   a   catalyst,   because  their  pellet-type  catalyst   applied   to  a   crossflow
reactor would result in an unuseably large heat exchanger.

      W.R.  Grace  declined  their  involvement  in   supplying  catalyst  materials  at
that time.

      Johnson  Matthey proposed  to  design, build  and  catalyze a reactor.    The
proposed reactor was a  single  methanol  annulus   design  similar to  configuration
number  6  in  Figure  3B  of Appendix D.    Their  preference for  a single annulus
design  was  to reduce cost and complexity.    With connections, the length  of the
reactor unit  would  be less than  30  cm  (12   inches).    It  would  be  capable  of
operating at  1035 kPa (150 psig) and  would  be  catalyzed  on  both  the  methanol
and  exhaust  sides.    Subsequently,  specifications   for  a double-annulus  design
were  provided  to JM for review.   They  replied  that they   could  produce   that
design.   Therefore, a decision  was made to have   JM fabricate  the double annu-
lus design.    SwRI would  design  and  build  a  separate  evaporator to  mate  with
the reactor.    SwRI  proposed  to assemble  and  test one   of  these  heat  exchanger
units  on one  cylinder of  the  NAPS-Z engine to  obtain  experimental  data.    The
output  of  the reactor for  this  first  test  would not be injected  into  the engine,
but would  be analyzed  to determine  the  product  species, the  mass  flowrates of
vaporized  and  dissociated  products,  and  operating temperatures  and  pressures.
Copies of correspondence with Johnson Matthey  are contained in Appendix E.
                                        16

-------
      Johnson  Matthey  provided  the   following  specifications  on  their  proposed
catalysts:

      Dissociation Catalyst Material:             Johnson Matthey Type H Base Metal
           Apparent Surface Yield:                     3.9 kg/hr-m2 (0.8 lb/hr-ft2)
           Actual Surface Yield:             1.8  x 10'3 kg/hr-m2 (3.7xlO-4 lb/hr-ft2)

      Specific Surface Area:                                               125 m2/g

      Bulk Density:                                                        0.5 g/cm3

      Washcoat Thickness:                                                  0.05  mm

      Exhaust Catalyst:                          Precious Metal Proprietary (9Pt:lRh)

      Reactor Material:                                               Stainless Steel

      Thermal  Conductivity:                         K = 24 W/m-k (14 Btu/hr-ft-°F)
JM  provided  the  catalyzed  metal  foil  substrates  and  assembled  the  tubes  and
substrates  into  the  reactor  configuration.   The  tubes making  up  the concentric
catalyst  dividers  were  fabricated  by  SwRI and  supplied  to  JM.    The  partially
assembled  reactor  was  delivered  to  SwRI,  where it  was  visually inspected, prior
to  final assembly,  to  assure  all  passages  in  the methanol and exhaust  catalysts
were open.

     After  assembly of  the  evaporator  to  the  reactor,  thermocouples,  pressure
taps,  and  the injector  nozzle were  installed.   The  assembly  did  not  leak  when
pressurized  to 69.0 kPa (100 psi)  while submerged  in water.  This  prototype heat
exchanger  (evaporator/reactor)  was  then  evaluated   on  a  test  engine, and  the
results   of  those  tests  and  of  a  post-test  analysis  are  covered  in   the following
section.
                                        17

-------
                   III.  EVALUATION OF HEAT EXCHANGER


     A test  plan, contained in Appendix  F,  was prepared  following  an EPA/SwRI
meeting  in  February,  1986.   The purpose of  the  planned  testing  was to  deter-
mine  the performance  of the  heat  exchanger,  compare  the  experimental  results
with  the desired and calculated  values,  provide  data  to  optimize  the  heat ex-
changer  design,  and  balance  the  individual  reactor  and evaporator  designs.   The
test  engine   setup  is shown  in   Figure  8.   For  these  evaluations,  the  product
output from the  dissociation  heat exchanger  was sampled and evaluated  but was
not injected into the engine intake.

A.  Test  Engine

     A  Nissan NAPS-Z  two  liter engine  was  selected for  the  evaluation of the
prototype methanol  dissociation  heat exchanger and this  engine has two  spark
plugs  per  cylinder.   Operating  the  engine  on  only one  spark  plug  provides  a
convenient location  in which  to  install  a  valve  for  induction of  the  products of
dissociation.    The  combustion  chamber,  a  crossflow  design,  is  also   well  suited
for these test purposes.

     A  cylinder head was sectioned to determine  if a  new cylinder  head casting
would be  required   or  if existing heads  could be  altered for the  test  program.
After  layouts  of  heat exchangers  were  made, based  on using  a new  head design
and on  using  existing  heads,  a  decision  was  made  to   use the  existing  head
design.   Appendix  G  contains  design  options  for  the  heat  exchanger  gas ad-
mission valve and head modifications.

     A  NAPS-Z engine,  received  from  EPA  to  use  in  the evaluation  of the
dissociation  heat exchanger, was  installed  in  a  test  cell and  run  on  methanol at
several points  matching  those in EPA  test  data  provided  to  SwRI.   Results of
this initial  testing  are  shown in  Table  2.   All  fuel  consumption  data  taken at
SwRI  exceeded that  taken  at the EPA.   The  specific fuel mass flowrate  trans-
ducer  used   was  a  Micromotion  unit  calibrated  for  methanol.   Its  calibration, as
well as  that of  the dynamometer, were  checked and reconfirmed when  this  fuel
consumption  difference  became apparent.    A laminar-flow  element was  installed
on  the engine induction  air  system  to  permit  measurement of air  mass  flowrate
and calculation of equivalence ratios.

     The engine  was then operated  at  three  test  points, data  from which are
summarized  in  Table 3.    The engine  started and  operated smoothly  and  steadily
during this  entire  test sequence.   The engine was then  shut  down  to review the
data.   When preliminary testing  was resumed,  engine  failure occurred.   Repairs
included honing the  cylinder bores,  installing  new  low-compression  pistons  and
rings,  and installing  new connecting rods.   The damaged pistons were inspected,
photographed and sent to the EPA.

     The cause  of  the  engine failure  was  not  conclusively  determined.   Damage
such  as  this is  usually caused  by high  temperatures that result  from excessively
lean operating  and/or pre-ignition.   However,  test  data  shown  in  Table  3  indi-
cate that the  engine  was  running  richer  than stoichiometric.   Subsequently,  it
was determined  that  the  fuel-rich  operation  apparently  resulted  from  misin-
stallation  of  an  air-fuel  ratio  controller  provided   with  the engine.  Following
repairs,to  return    the   engine   to   standard   overhaul   condition,   break-in

                                       18

-------
         Standard Exhaust
         Manifold
                                        One Stanadyne Pencil Nozzle
                                                  Methanol  to Evaporator
To  Laboratory
Exhaust  System
                                             -Evaporator
                                             /Reactor
                         Reactor
                    Products  Out  to
                   Gas  Chromatography
                   For  Analysis
                 Engine Driven
              Stanadyne DB2  Pump
r                                                                Methanol  thru  Standard  Engine
                                                                Injection System
              Engine Induction Air
NAPS-Z Engine
                                  «—Three  Stanadyne  Nozzles
                                     Discharging into Reservoir
                                            Data Acquisition
       No.

        1

        2

        3

        4

        5

        6

        7

        8

        9

       10

FIGURE 8.
                               Location
                                                                          Parameter
                 Methanol into the evaporator

                 Methanol between the evaporator and reactor

                 Dissociated products out of reactor

                 Exhaust out of engine

                 Exhaust between reactor and evaporator

                 Exhaust after evaporator

                 Engine crankshaft

                 Engine intake

                 Methanol injected into manifold or ports

                 Engine coolant

                    TEST ENGINE  SETUP TO  EVALUATE
        Temperature, pressure
          Mass flowrate
        Temperature

        Temperature, pressure
          Product analysis
        Temperature, pressure

        Temperature, pressure

        Temperature, pressure

        Mass airflow

        Mass flowrate

        Temperature

        Temperature

   HEAT  EXCHANGER
                                            19

-------
Table 2.  Initial SwRI test Data on NAPS-Z Engine As Received
Spark
Engine Advance
Speed (rpm) °BTDC %Loa
1500 27.0 30
1500
1500
3000
3000
3000
3000
4500
4500
4500

18.0
12.0
39.0
39.0
32.0
22.0
45.0
24.0
24.0
Table 3.
70
100
30
50
70
100
30
70
100
Fuel Consumption
Tnrnue (N-m^ (ke/hr)
d EPA
45
96
141
44
75
105
150
42
96
140
Summary of Data from
Engine Spark
Speed Advanced
(rum) "BTDC






4500
3000
1500
Power Output
fkW)
EPA SwRI
64.9 63.9
33.0 32.8
7.1 7.0
23.5
32
27
SwRI
45
96
140
45
76
104
156
42
95
141
Engine Test
EPA
4.7
7.6
11.0
9.2
12.4
15.9
22.3
14.2
24.3
32.2
on NAPS-Z
Fuel
Consumption
Toraue (N-m) fke/hr)
EPA £
140
105
45
BSFC
fke/kW-hr)
EPA
0.496
0.482
0.662
SwRI
0.604
0.512
0.814
JwRI EPA
136 32.2
104 15.9
45 4.7
A/F
EPA
6.9
8.9
9.4
SwRI
38.6
16.8
5.7
Ratio Eqi
SwRI
5.8
6.2
4.9
SwRI % Difference
5.8
8.2
12.4
11.4
15.5
19.1
24.2
19.7
28.7
37.8
23
8
13
25
25
20
9
39
18
18
Engine as Received
Air
Consumption
fke/hr)
EPA SwRI
224 222
142 105
44 28
livalence 0 Ratio
EPA SwRI
0.93 1.25
0.72 1.03
0.70 1.32






                           20

-------
running  was performed.   The  engine  continued  to  run fuel-rich, with equivalence
ratios  ranging  from  1.5  at  low  loads  and  speeds   to  1.15  at  high loads  and
speeds.   Following  the  break-in running, the engine  was prepared  for  measuring
heat exchanger parameters and dissociation product makeup.

B.  Heat  Exchanger Test Results

      The  methanol dissociation  heat  exchanger  was  evaluated,  using  the  exhaust
from  one cylinder, over  a  range  of  engine speeds  and power  outputs.   Results
for  several test  runs  are  summarized  in  Table  4 and 5.   Appendix H contains  a
copy  of   the  report  submitted  to  EPA  during   September,  1987  covering  the
engine evaluation of the methanol  dissociation  heat exchanger.

      Table 4.  Data from Evaluation of Methanol  Dissociation Heat Exchanger*


                                 RUN 4   RUN 3   RUN 9     RUN 5     RUNS
ENGINE  OPERATION
  Speed, rpm                    1500     1500     1500       1500       3000
  Torque, ft-lb                  30       30       30         30         40
  Methanol/Cylinder, Ib/hr       11.7      12.7      10.8b       12.7        28.2
  Exhaust/Cylinder, Ib/hr         813      791      840C        773        1148C
  Exhaust Temperature, °F

DISSOCIATOR
  Insulatedd                      NO      NO      YES        NO        YES
  Methanol Flow, Ib/hr           0.50      1.00      1.10        2.00        6.30

Exhaust Temperature,  °F:
  Before Reactor                 790      707      877°        658        1167°
  After Reactor                  652      419      721         171b       863
  After Evaporator               422      210      244         162        691

Methanol Temperature, °F
  In Evaporator                  392      198        352       162        345
  After Reactor                  518b      666        788       576        1140
aData presented in order based on methanol flowrate into the dissociation.
bData appears to be somewhat low.
°Apparent minor discrepancy in these data based on location in system.
dAround the  outside of the heat exchanger unit.
  Based  on  the  temperature data  given  in Table  4,  the  evaporator  was  appar-
ently  unable to  vaporize  the  total fuel needs of  an engine  cylinder  at  the  1500
rpm operating  condition.   This is concluded  from  the  relatively low  temperature
(i.e., less  than 200°F)  in the  evaporator in  Runs  3 and  5.    Insulating  the  heat
exchanger  assembly  resulted  in  some  improvement  in  vaporizing  potential,  as
illustrated  by  the  increased  temperature  in  the  evaporator  in  Run  9.   At  the
3000 rpm  engine operating  condition, the total methanol fuel needs of an  engine
cylinder could have been vaporized  by  the evaporator.
                                       21

-------
            Table 5.  Methanol Mass Flowrate Versus Percent Dissociation

                  Methanol           % of Engine
Run Mass Flowrate Cylinder
No. fke/hr) Needs
6 0.01 1
4 0.2 19
7 0.4 29
3 0.4 31
9 0.5 35
5 0.9 63
8 2.9 89
Dissociated Products Engine
Run as of Speed Torque
No. % Cylinder Needs from) CN-m)
6 0.6 1500 41
4 4.2 1500 41
7 2.0 1500 41
3 1.6 1500 41
9 1.8 1500 41
5 0.6 1500 41
8 9.8 3000 54
Percent
Dissociation
96
22
7
5
5
1
11
Heat Exchanger
Condition
Insulated
Uninsulated
Uninsulated/pressurized
to 242 kPa
Uninsulated
Insulated
Uninsulated
Insulated
      Conditions  in  the  heat  exchanger  were  stable  at  low  to moderate  engine
speeds  and  loads,   enabling  determination  of  methanol  dissociation  rates.    At
high  engine speeds  and  with high  loads  at lower  engine speeds,  the conditions
in the  heat exchanger  were  very unstable,  and  meaningful  methanol  dissociation
rates could not be determined.

      The  mechanical  design  of  the  heat  exchanger  was  generally  satisfactory.
An  exception  was  the  injector  location in  the  exhaust  stream;  high  injector
temperatures at  high speeds  and  loads were encountered.   The  lack of  adequate
injector  cooling  was a  major factor  in  the  inability  to  operate  the  heat  ex-
changer  at  high  engine  speeds and loads because  of  the severe deterioration  of
injector spray characteristics.
                                        22

-------
      For  the  evaluations  shown  in Table  5,  the dissociation rates  in the  reactor
 ranged from 1  to  96 percent  of the total  methanol  injected into the evaporator;
 the  higher  dissociation  rates  were  associated  with  low  methanol injection  rates.
 These  dissociation   rates  represented  from  0.6   to   9.8  percent  of  the  engine
 cylinder  fuel  needs  at  the  test  points  evaluated.    Heat  exchanger performance
 versus  methanol  flowrate  is shown  in  Figure 9  and  the percent of cylinder fuel
 needs versus methanol flow rate  is shown in Figure  10.

      The  evaporator was  adequately sized to vaporize 100  percent of the  engine
 cylinder fuel needs at  3000  rpm,  but  only about  50 percent at  1500 rpm.   The
 vaporization rate was greater  at  3000 rpm  because  of  the  higher  exhaust  tem-
 perature and flowrate at that engine speed.

 C.  Analysis of the  Catalyst

      Following  analysis  of  the  heat  exchanger  test  results and the  conclusion
 that  satisfactory  heat exchanger  performance was  not  achieved,  the reactor was
 shipped to  Johnson Matthey  for  a  post-test   analysis  of  the  reactor  catalyst.
 Specifically, JM  was asked  to  determine  the  catalyst  active  surface  area  and  the
 condition  of  the  catalyst  and  to  conduct  other  analyses,  as  appropriate,  to
 determine   why   the  dissociation  yields  were  so  low.   Based  on  the   catalyst
 design  criteria,   (e.g.,  optimum  dissociation  temperature  of  370°C)  dissociation
 yields should have been much higher.

      The  methanol  dissociation  and  the engine  exhaust  catalysts  were  checked
 by  the  JM Analytical  Lab  for  effective surface  area  (B.E.T.).   These  catalysts
 were  also   scanned  by  x-ray  for the presence  of catalyst  poisons  and  the coke
 deposit  at  the  reactor  outlet  was  analyzed  by x-ray  fluorescence  (XRF)  for
 contaminants.   Preliminary results  of  the analyses  by  JM  are  given  in Appendix
 E.

      The  effective  surface  area  measured  for  the  dissociation  catalyst  was  less
 than  one  meter  squared  per  gram (m2/g).   The  corresponding result  for  the
 engine  exhaust  catalyst was  just  slightly  over one  m2/g.   Elements  found on  the
 dissociation  catalyst were  Al,  Cr,  Fe, and  Zn,  and  elements  found  in  the coke
 deposited  on the dissociation  catalyst were  Ca,  Cr,  Fe, Pb, and Zn.    Elements
 found on  the  exhaust catalyst  were Al,  Ca,  Cr,  Cu,  Fe,  Pt, and  Zn.  The high
 temperature  resistant stainless  steel  metal support  is  an alloy  containing  Al,  Cr,
 and Fe which could  be picked up by the x-ray.

      The  expected   effective  surface area  for  a  metal supported  catalyst of  the
 type  used  would be  somewhere  on the  order of  8 to  13  m2/g,  so  the  effective
 surface area  of  both  catalysts  was an  order of  magnitude  lower than  expected.
 Usual  reasons  for   loss  of   surface area  are poor  washcoat  adhesion,   thermal
 sintering,  and  masking/poisoning.    A scan  with  a  20-power lens did  not reveal
 bare  patches of  metal  which  would  indicate  washcoat adhesion problems.   The
 600°C (1100°F)  temperatures  noted  by  SwRI are  not as  high  as  in automotive
 applications  and  would  not  seem to be  a  prime candidate  for  low  surface  area
 numbers.    Welding  the  catalyst  reactor  could  be  a  source  of  high  temperatures
(and  metal poisons),  and  masking  or  poisoning effects are  a  definite possibility.
The  coke  deposit scanned  by XRF  showed the presence  of  Ca  and Pb  as well  as
the   expected Cr,   Fe,  and  Zn,  but  the amount  of  each   of  these elements  is
unknown.    Each could  be  a  poison in large  amounts on the  catalyst   surface.

                                        23

-------
        96% @ 0.02 LBM/HR
o
cr

% DISSOCIATION
% CYLINDER FUEL
REQMTS
POSSIBLE WITH
DISSOCIATION
PRODUCTS
1500 RPM
& 30 LB-FT
A
O



3000 RPM
& 40 LB-FT
O



                         Percent dissociation as a function of methanol
                         mass flow rate
                                                                     O
                         Percentage of cylinder fuel need provided by
                         dissociated methanol
                                    I
                                I
              1.0
2.0        3.0        4.0         5.0

     METHANOL MASS FLOW RATE (LBM/HR)
                                                                  6.0
7.0
      FIGURE 9.     HEAT EXCHANGER PERFORMANCE VERSUS METHANOL MASS
                    FLOWRATE.
                                           24

-------
   100
    75
C/D
Q
LJJ
    50
LU
Q
>
o
    25
                                                                             X
                                                                               X
                                                                  X
                                                               X
                               X
                                                            X
                                                      X
                                                 I
                                            I
                1.0
2.0         3.0         4.0        5.0         6.0


     METHANOL MASS FLOW RATE (LBM/HR)
7.0
          FIGURE 10.    CYLINDER FUEL NEEDS VERSUS METHANOL MASS
                         FLOWRATE.
                                           25

-------
The  x-ray  scan  of  the catalyst  is  not  sensitive to  the  lighter  elements,  so a
masking  by  carbon  compounds would  not  be  detected.    The  very  fact that  a
coke  deposit was  produced within  the reactor  may  indicate  occurrence  of  some
other undesired catalytic reactions.

      A  check  of  the heat  exchanger  design  was  made,  and  generally  it appears
that  appropriate  design  criteria were  used  in  all  but a couple  of  areas.    The
primary design  flaw observed  was  the location  of  the  methanol  injector in  the
exhaust stream without provisions for cooling.

      In some areas,  potential  for  design  improvements were  noted.   One design
improvement  involved   the  exhaust  flow  through  the  reactor  section  of   the
methanol dissociation heat  exchanger.   With  reference  to the view  of  the  reac-
tor given  in  previous  Figure  7,  the  center  section  of  the  exhaust  catalyst  was
designed  to  carry  the  same  exhaust  heat  per  circumferential  unit  as  the  two
outer  sections  of  the exhaust  catalyst.   Since  the  center  sections  provides   heat
to  methanol  catalysts  on both  radial  sides,  it should  be  designed to carry  more
exhaust  heat.   Incorporating this  design  improvement  involves  only  minor  rede-
sign.    Another  design  improvement  would  be  to  insulate  around the outer   dia-
meter  of   the  reactor   with a material  such  as  3M Interam  heat  expandable
ceramic  insulation.    Due  to the limitations  on  the  outside  diameter  of the   heat
exchanger,  however,  incorporating  such  internal  insulation  would  require exten-
sive redesign of the reactor.

      Based  on  the  overall design   criteria,  the  heat  exchanger   should  have
provided  relatively  good  dissociation  of  methanol at  several  of  the  conditions
over  which  it was  operated.  Some of the  findings in  the  inspection  of the  heat
exchanger  assembly  by  SwRI  and  of  the  catalysts  by  Johnson  Matthey  probably
explain  the  low  dissociation   efficiency.     The  primary finding  was   that  the
methanol catalyst had an  effective area  an  order of magnitude less  than antici-
pated.   No  specific  cause  was  found  for  this  major reduction  in effective  area.
Other  findings  of  potential  significance  were   that  the  actual  cross  sectional
flow  area of  the  center section of  the  exhaust  catalyst  was less than half  the
designed  area  and  that  significant  carbonaceous  deposits  were  present  on   the
methanol  catalyst.   A number  of  the  outer  flow  tubes in  the  methanol  catalysts
were  blocked by plugs of the  carbonaceous material.

      Fabrication  of  subsequent  units  should  include   additional  inspection   to
assure  meeting  design  specifications.    With  the  catalyst,  this  should  include
determinations  of  the  effective surface  area and  the  dissociation  efficiency  of
the  actual   batch  of  catalyst  substrate  to  be   used  in  the reactor  assemblies.
During  assembly  of  the  system,  checks  should  be made  to  assure  all  important
physical dimensions and criteria meet design specifications.  Also, the type of
catalyst used in the  the prototype heat  exchanger should  be reconsidered  as  to
whether it remains as the  primary choice.
                                        26

-------
         IV.  VALVE TRAIN AND CONTROL SYSTEM CONSIDERATIONS
      Design  considerations  toward  a  valve  train  and  control  system  for  the
 methanol dissociation heat exchanger are  discussed in this section.

 A.   Valve Train Considerations

      An initial objective  of the  program  was  to  inject  the  products  of dissoci-
 ation  directly  into the  combustion  chamber  without first mixing  with  the  com-
 bustion  air  charge.   To  achieve that  objective,  a  dissociated products  inlet  valve
 is  required  which  operates  independently  of  the  engine   intake  and  exhaust
 valves.   The selected  engine (Nissan NAPS-Z)  has  two  spark  plugs per  cylinder.
 One  of  the spark  plug  holes  can be  used  for the  dissociated  products   inlet
 valve, and the  engine can  be  run  using  the  remaining  spark  plug per  cylinder.
 The  turbulence  and  mixing  that will occur should  assure  that  one spark  plug  per
 cylinder will be sufficient.

      Several  methods  to actuate the valves  for  admitting dissociated  products  to
 the  cylinders  were  considered.   These included  electronically-actuated assemblies
 that  could  be  designed  to perform  both  gas admission   and  flowrate  control
 functions.   While there  appears to  be  some  advantages  to this method, it is  the
 most  expensive  approach  and  also  the  most  risky.   The expense  is  due to  the
 low volume production  of suitable  solenoids  and the  need  for a separate control
 system to  open  and  close  the  valves  at  the  proper  times.    The  risk  is  high,
 because  the concept is  not  proven  and it is not known  whether  development  of
 working hardware could be achieved.

      Mechanical  valve actuation methods  (overhead  cam and  overhead  valve  with
 rocker  arm)  were  reviewed to  determine  which  was  the  most  suitable.    The
 specific  requirements  of the  dissociated  product  valve  timing  necessitate careful
 attention  to valve dynamics.   With the  short  duration  (73°  of crank  angle) and
 high  maximum  engine speed  (6000 rpm),  the  valve will  have to  be  accelerated
 quickly  to  achieve  its full open position.   The  high accelerations  will  yield  high
 forces, and  such forces  are  best handled  by a  valve train of  minimum mass and
 proven high  speed  potential.   The  short  duration  requires a cam  profile  without
 convex surfaces or  a  cam  with an all-around  convex  surface  of  large  diameter
 (approximately  15-cm  base  circle).   An  alternative to  either  of  these  would  be
 to  use a cam  with  a convex  contour  and a small lift,  combined  with  a rocker
 arm,  to  increase  the  valve lift  to  the  required  amount.    This  approach,  however,
 increases  the  total  valve  train  mass   and  makes  packaging  impractical.    After
 considering  these  various  approaches,  direct  actuation  using  a   small  diameter
 cam  was  chosen.   The  layout  for the  heat exchanger was enlarged  to  include
 the valve  gear  and a  computer  program  was  written to calculate  the  cam profile
 and   the  valve  accelerations.    The dissociated products valve  train  design  is
 contained in Appendices I through K.

 B.  Control System Considerations

      Because  the  dissociation  of  methanol  will  not   occur  at  engine  start-up
(cold  exhaust),  and  may not  be  adequate  to  fuel the  engine  at  some operating
conditions,  a dual system for fueling  the  engine will   be required.   One of  the
systems is  for  port-injection fueling, and  the  other  is  for  injection of methanol


                                        27

-------
into  the  catalytic reactor.    Separate  control  systems  could be  required  for the
two systems. Parameters to be sensed for control purposes include:

       0    Evaporator temperature
           Catalytic reactor temperature
           Engine speed
           Engine load (throttle position or manifold vacuum)
       0    Exhaust oxygen content
       0    Exhaust temperature
           Hydrogen in reactor

      At  this  time,  no  practical  method  is  known  for  continuously  measuring
hydrogen concentration.

Design  Concept  - A  conceptual design  of  a control system  was developed follo-
wing  the  testing of  the  heat  exchanger.  Operation  of  the  heat  exchanger  made
some   of  the  desired  characteristics  of  a  control  system  apparent.    Control  of
the engine  is  thought best  done  with  two  separate  control  loops.    One  loop
would control  the heat exchanger  and maintain  it  at  some  condition,  such  as  a
predetermined   evaporator  temperature   that  was  above  the  equilibrium  point.
The other control system would control the engine fuel  system.    Flow from the
heat  exchangers  would be restricted  or  stopped  for  low  fuel  demand  conditions.
Otherwise  all  available reactor  contents  would   be discharged into  the  combus-
tion  chamber  each  time  the  valve  is  opened   and  the  manifold  fuel  injection
system would make up  for  the balance of fuel needs.

Reactor  Variability  - It  became   apparent, when  operating  the   heat  exchanger
that it is difficult  to control  the evaporator  temperature.    Also, exhaust   tem-
perature varied  substantially  from  run  to  run  and  drifted  even  when  the  heat
exchanger  was  operating  under what  should  have  been steady-state   conditions.
It   appears  unlikely   that   multiple  heat  exchangers  (such  as   one  for   each
cylinder)  can  be maintained  at  similar  conditions.    If  each  heat  exchanger
operates at different  conditions, rates of dissociation will differ.   With  different
rates  of  dissociation,  the  equivalence ratio  in  each cylinder  will  differ and  it
will be  difficult to  take  full  advantage of  the extreme lean  operation  possible
with  hydrogen-rich  mixtures.     A  single  heat   exchanger for the engine  would
eliminate  such variation  and be more space  efficient than  multiple  units,  but  a
single  unit would likely  present major  challenges  in  the  design  of the evapo-
rator and the dissociation reactor.
                                       Z8

-------
                     V.  HEAT EXCHANGER - FINAL DESIGN
      A  technical  design review was  conducted  and  the  results  were  transmitted
to  the  EPA  for review.   A  copy  of  the  design  review  and  proposed  improve-
ments  is  given  in  Appendix  L,   and  the  findings  are  briefly  summarized  as
follows:

Dissociation  Catalyst  -  The   base  metal   catalyst  used  in  the  initial  heat  ex-
changer  did  not provide  satisfactory  dissociation.    From  a brief  review of  the
published  literature and  discussions with individuals  at JM,  EPA,  and  SwRI, it
appears  that  a  fully  appropriate   dissociation  catalyst  composition  has  not  yet
been identified.   No catalyst  composition  has  been  found that  will provide  the
essential   characteristics   of   high   dissociation   efficiency   at  low  temperature,
durability   at   somewhat  higher temperatures,  and   negligible  coking   character-
istics.   The  primary  requirement  for  developing  a  methanol  dissociation   heat
exchanger  is the identification of a suitable dissociation catalyst composition.

      Until sufficient  operating  history  has  been  developed on a  specific  catalyst
composition,  it  is  recommended that a sample  from each batch be analyzed  to
assure it meets design specifications.    At  minimum,  such  analyses  should  include
determination  of effective   surface  area and  bench  determination  of  dissociation
efficiency.

Evaporator -   The   present  evaporator  involves  a  closed   cylinder  wrapped  with
corrugated  metal catalyst  substrate.    Because the  engine  exhaust  gases  first  pass
through  the   exhaust  catalyst  in  the  dissociation-reactor,   which  should  oxidize
essentially   all available  HC and CO,  the catalyst  wrapped  around  the  evaporator
appears  to   serve  no   useful   function.    Additionally,  the   corrugated   catalyst
substrate  effectively  isolates  most   of  the  exhaust  flow  away from  the  evapo-
rator.    By  removing   the  catalyst  substrate from  around  the  evaporator  and
installing  longitudinal fins   onto the  outer surface,  heat   transfer  from  the  ex-
haust  to  the  evaporator  should  increase   significantly.    Another  potential  im-
provement  is  to have the  cone  in  the  evaporator  open  to  the  exhaust,  rather
than closed as  in the present design.

      A  commercially available  injector,  that can be   utilized with  only  minor
modification,   is  considered  to  be satisfactory  and  about  as good  as   can   be
attained  without major  redesign  effort.    An  essential   design  modification  is
relocation of the methanol injector out of  the  exhaust stream.

Dissociation  Reactor  -  One   design  improvement  involves  the   exhaust  flow
through  the  reactor  section  of the  methanol dissociation  heat exchanger.   With
reference  to  the original   design  of  the  reactor,  the  center  section  of  the  ex-
haust catalyst  was  designed  to carry  the   same  exhaust heat  per  circumferential
unit  as  the  two outer  sections  of the exhaust  catalyst.    Since  the  center  sec-
tions  provides  heat  to  methanol  catalysts  on   both  radial  sides,  it  should   be
designed to  carry  more exhaust heat.   Another recommended improvement  is  to
increase  the  methanol dissociation  annuli  from  two  to  three.   A  third  improve-
ment would  be  to  increase  the  void   space at  the  entrance and  exit  of  the
dissociation annuli  and  to  provide  four exit ports  from  the  annuli,  rather  than
the two as  in the initial design.
                                         29

-------
Expected  Result  of  Design  Improvement    The  mechanical  design  improvements
described  should  improve the  operation  and  dissociation  efficiency  of  the  heat
exchanger.  These  improvements, however,  are not expected  to  provide anywhere
near the  order  of magnitude  improvement  desired.   It appears that selection  of
the  catalyst   composition  is  the   most  important  criteria   toward   meeting  the
desired dissociation efficiency.
                                     30

-------
                  VI.  CONCLUSIONS AND RECOMMENDATIONS
      Tests  of  the heat  exchanger  showed  the  dissociation  rates  to  be  1  to  96
percent   of   the  total  methanol   injected  into  the  evaporator.     These   rates,
however, only   represented  1  to  10  percent   of the  cylinder  fuel needs  at  the
engine  test  points evaluated.    The  catalytic  reactor  size  or the  effectiveness of
the  catalyst  (or  a  combination  of  size  and   effectiveness)  will  have  to  be  in-
creased  by roughly an  order of magnitude to  achieve the dissociation needed to
meet  20-25  percent  of  the   engine  fuel  needs  at   the  lower  range  of  power
operating conditions.

      The  evaporator  was  adequate  to  vaporize  100   percent  of the  engine  fuel
requirement  at  3000  rpm and  50 percent at  1500 rpm.    Exhaust  temperatures at
the  evaporator   were  higher  than  they  would  have  been  if  the dissociation  rate
in  the  catalytic  reactor  had   been  higher.    A more  effective  reactor   would
absorb  more  of  the  exhaust heat,   and  the  evaporation  rate  would  be  reduced,
possibly  to  the  extent  that  the  evaporator  size or  effectiveness   would   have to
be increased.

      Injector overheating occurred  as a consequence  of it being mounted  in  the
exhaust  stream.   When  injector temperatures   exceeded 205°C  (401°F),  the  injec-
tor  nozzle  spray  pattern  deteriorated.    The   lack  of  adequate injector  cooling
was   a  major  factor   in  the  inability   to  operate  the  heat   exchanger   at  high
engine  speeds  and loads.  One  flowrate  of  methanol  was evaluated  at  a  higher
engine  speed and load,  54 N-m (40 Ib-ft) torque  at  3000 rpm, and it  resulted in
the  highest  rate  of   dissociation   (11  percent)  and  the  highest  percentage  of
cylinder fuel requirement (10 percent).

      The  catalyst  used  in  this   reactor  to   dissociate   the   methanol  provided
unsatisfactory   dissociation  yields   at  all  engine  power   levels evaluated.     The
power  levels  evaluated  are  considered reasonably  typical   of  engine operation  for
light-duty vehicles.    Although  exhaust  flow  through  the  reactor  annuli  was not
optimum,  it  appears   that  a  catalyst  with  higher  dissociation effectiveness  at
lower temperatures is needed to enable a reasonable reactor size and mass.

      Test results  indicate that  if  a  separate  heat  exchanger is  used   at  each
cylinder,  variations  in  dissociation   rates  from  cylinder-to-cylinder  will occur and
equivalence  ratios  will  vary.    Complex  controls will likely  be  required  to take
advantage  of the extreme  lean  operation  possible  with  hydrogen fuel.   A  single
catalytic  reactor  for  the  engine   would  eliminate cylinder-to-cylinder  variations
and  be  more  space efficient.   Evaporator  and dissociation  reactor design  for a
single unit, however, would likely become more  complex.

      Heat exchanger  performance   was  shown  to be inadequate  because  of  the
poor  performance of  the  catalytic  reactor  and  the injector.    Cause for  the poor
performance  of the injector  has  been  determined.   Specific  cause for  the poor
performance   of  the  catalytic  reactor,  however,  can not be  specifically  deter-
mined  from  the  available data.    It is  recommended  that  subsequent  methanol
dissociation   heat  exchanger  development include  bench  testing  in   the  catalyst
selection  process  and  baseline  bench  testing  of  the  batch of  catalyzed  substrate
actually used.
                                        31

-------
                                 REFERENCES


 1.    I. Yamaguchi,  et  al, "Development Research on  Dissociated Methanol Fueled
      Spark Ignition Engine, " SAE Paper  852201.

 2.    D.  McCall,  et  al,  "Performance  and Emissions  Characteristics  of a  Spark
      Ignition  Engine  Fueled  with  Dissociated  and Steam  Reformed  Methanol,"
      Department of Mechanical Engineering, Texas A  and M University.

 3.    Most,   W.J.  and   Longwell,  J.P.,  "Single-Cylinder   Engine  Evaluation  of
      Methanol-Improved  Energy Economy  and Reduced NOX,"  SAE  Paper 750119,
      February 1975.

 4.    Bernhardt,  W.E.,  "Engine  Performance  and Exhaust  Emission  Characteristics
      From  a Methanol-Fueled  Automobile,"  General Motors  Symposium  -  Future
      Automotive Fuels;  October  1975.

 5.    Leonard,   J.P.   (Chem  Systems),  "Prospects  for  Coal-Based   Transportation
      Fuels Fading Fast," Oil and Gas Journal, 30 May  1983, pp.  93-97.

 6.    Cox,  T.P.  and  Rykowski,  R.A.,  (EPA), "Methanol:  Its  Production, Use  and
      Implementation," ASME Paper 82-DPG-25.

 7.    "Large  Coal-to-Methanol  Plant Design,"  The  Oil  and Gas  Journal,  p.  102,
      March 27,  1978.

 8.    R.  Toepel,  et al,  "Development  of Detroit  Diesel  Allison  6V-92TA  Methanol
      Fueled Coach Engine" SAE Paper 831744.

 9.    A.  Konig,  et al,  "Engine  Operation on  Partially  Dissociated  Methanol,"  SAE
      Paper 850573.

 10.   R.  Bechtold, "The  Theoretical Limits  and Practical  Considerations  of  De-
      composed  Methanol  as a  Light Duty Vehicle  Fuel,"  VI  International  Sym-
      posium on  Alcohol Fuels Technology, May 21-25, 1984, Volume I.

 11.   J. Finegold, "Dissociated  Methanol  Vehicle  Test  Results,"  VI  International
      Symposium on Alcohol Fuels Technology, May 21-25, 1984, Volume I.

 12.   Bundel,  H.,  "Implementation Experiences  with MWM  Pilot-Injection Diesel
      Cycle Engines Burning  Alcohol as  Main  Fuel,'  MWM  Diesel Ltd., Brazil, pp.
      1-56.

13.  Neitz,  A.,  et al,  "Results of  Further  Development  in the M.A.N.  Methanol
     Engine,' M.A.N., p. 1-303.

14.  Pischinger,   F.,  et  al,  "Operations  and  Exhaust  Emissions Behavior of  the
     Direct Injection Alcohol Diesel Engine," p. 1-319.

15.  Telephone  communications  with Dr.  Frank Burke,  Coal  Research  Division,
     Conoco, Inc.
                                       32

-------
                           APPENDICES
A -  TESTING OF SwRI PUMPS AND INJECTORS
B -  DESIGN SPECIFICATIONS - IN-CYLINDER DISSOCIATION OF METHANOL
C -  EVAPORATOR DESIGN ANALYSES - PERFORMED BY A CONSULTANT
D   HEAT EXCHANGER DESIGN ANALYSES - PERFORMED BY SwRI
E -  CORRESPONDENCE WITH CATALYST MANUFACTURER
F -  TEST PLAN FOR EVALUATING HEAT EXCHANGER
G -  CYLINDER HEAT AND CATALYTIC REACTOR DESIGN OPTIONS
H -  EVALUATION OF METHANOL DISSOCIATION HEAT EXCHANGER
I   ENGINE MODIFICATIONS FOR A DISSOCIATED PRODUCT VALVE
J -  VALVE ASSEMBLY CALCULATIONS
K -  CAM PROFILE PROGRAM AND OUTPUT
L -  METHANOL DISSOCIATION HEAT EXCHANGER-DESIGN REVIEW

-------
          APPENDIX A
 TESTING OF SwRI DESIGNED AND
FABRICATED PUMP AND INJECTORS

-------
 TESTING OF THE SWRI DESIGNED AND FABRICATED PUMP AND INJECTORS
A.   Fuel Injection System

     Details  of the  fuel  injection system  are shown  in  Figures 1  and  2.   The
design is based on previous work at  SwRI  that  includes  a data base  of results.
Liquid methanol will  be supplied by  the  engine-driven fuel pump  to the electric
motor-driven  fuel  pump shown  in Figure  2.   This  pump  consists  of an electric
motor  driving  two  separate,  but  identical,  positive  displacement  single  piston
pumps via  an  eccentric.   The  eccentric  has a  ball  bearing mounted  around  it.
The  outer bearing shell  contacts  the  plungers which  eliminates  sliding  motion  at
their  bases.   Liquid  methanol  enters  the ends  of each  pump  through  an  inlet
check  valve  as  the  plunger  retracts  under  spring  force.    The  inlet check  valve
is  moved  from its open  to its closed position  by friction between the  plunger
mounted O-ring and the internal check valve diameter.  The trapped methanol is
expelled  through  a second  check valve to  the  pump  outlet.   The  plunger  rides
on a  teflon lip  seal  in  the  methanol end  of the bore.  An oil bath separates the
plunger  from  the  bore at the eccentric end.   Methanol from each piston  pump is
directed  to  two  evaporator  nozzles.    Each  piston  has  been  sized  to  provide
methanol to  two  nozzles over  the entire  range of  engine  requirements  (assuming
all engine fuel demands are met through  this pump).

      A  typical  nozzle  cross-section  is shown  in  Figure   1.    One  of  these  is
installed   in  each  evaporator.     Methanol  enters  the  nozzle  through   the   inlet
port,  goes through the fuel  channels,  and fills  the  void  on both sides of  the
plunger  piston.    As  each  pressure  pulse  is  provided by  the  pump,  the  valve
unseats  and  ejects  a quantity  of  liquid methanol  into the  evaporator.   Because
the pump is a positive  displacement  type,   the  metering  of  methanol  into  the
evaporator is  a direct  function of  the  pump  speed.    The  control system  will
provide  a  signal  to  the  electric  motor  based  on  its  measured   speed.    The
adjusting screw on the nozzle  is provided so that  fuel flow rate from individual
nozzles can be adjusted to the same value.

B.    Test Set-up

      A  series  of bench  tests   was  performed using  the  pump  (driven by  an
electric  motor)   connected  to  the  four   nozzles.     Each  of   the  nozzles  was
installed   in   a  clear  plastic  tube  with   a  three-inch  inside  diameter  (actual
diameter to be  3.108  inches).  Each  tube  had provisions for holding and  draining
the  sprayed  methanol  so  that  the   quantity of   methanol  sprayed  from   each
nozzle could  be  determined over the test  period  and compared  to each other.
Stoddard solvent was used in place of methanol for testing.

      The  system  was  evaluated using the  following   test  matrix  showing  nozzle
number as  a function of nozzle position and  test  number.
                                         A-2

-------
A
1
2
3
2
1
B
2
1
1
4
2
C
3
4
4
1
3
D
4
3
2
3
4
                   Test
                  Number      	Nozzle Position
                     1
                     2
                     3
                     4
                     5
     Each  test  included  operation  at 600,  950,  1200  (except  test No.  4),  and
2400 pump  rpm.   Nozzle  volume output  and individual pump  piston output were
measured  as  a  function of  time at each  of  these  speeds.   Test  number 5  is a
repeat  of test  number  1,  and  was  done to  check repeatability  of  the  system.
The  left pump  piston output was  always  connected  to  nozzle positions A  and C,
while the right pump  piston output  was  always  connected  to  nozzle  positions  B
and  D.   Pump  and nozzle  volume flowrates  were plotted  for each  of  the  tests as
shown in Figures 3 through 12.

     The  nozzles were  adjusted  to  produce a balanced  output at 2400 pump  rpm
as set up in test numbers 1 and 5.   Test number 2 interchanged  the  two nozzles
that  were connected  to  each pump  output so that  each pump  was delivering to
the  same  two  nozzles,  but in different positions.    Test  numbers  4  and  5
exchanged  nozzles  so  that  each pump  delivered  fuel  to  one  nozzle  that  previ-
ously received flow  from the other pump.

     Several observations are  apparent from the plots:

     1.    For test numbers 1, 2 and  5  -

     0     Volume  flowrate  while   matched  (cylinder  to  cylinder)  at one pump
          speed is not matched at other pump speeds.

     0     Nozzle  flowrate  versus  pump  speed  is  slightly  non-linear  (per  nozzle
          flowrate  plots).

          Pump output (sum of both applicable nozzles)  is not a  linear  function
          of pump speed (per pump  flowrate plots).

     2.    For test numbers 3 and 4 -

     0     Nozzle outputs are not  balanced.
     0     Pump output is  not a linear function of pump speed.

     From these observations  we  can  conclude that:

     0     The  system  design  (pump  and nozzles)  does  not  provide  for equal
          output from each nozzle  at  all  pump  speeds after the  nozzles  are
          balanced  at one pump speed.

     0     Pump output would not  appear  to  be a linear function of pump speed.
          Part of  this may  be  due  to  measurement  methods,  however,  because
          nozzle output increases  almost linearly with pump speed.
                                         A-3

-------
           Pump  piston  output  characteristics   differ   even  though  they   are
           dimensionally  the same.    This  is based  on  results of  tests  1 and  2
           which  show  that   interchanging  nozzle  positions  so  that  they  still
           receive  fuel  from  the  same  pump  piston  does not  appreciably change
           the   nozzle   flowrate  characteristics  versus  pump   speed.     Changing
           nozzle  position   (tests  3  and  4)  so  that  two  of  the  four  nozzles
           receive  fuel  from  the other  pump  piston changes  the  relative  flow-
           rates between nozzle pairs.

     In  addition  to  these conclusions,  it  was  apparent  during  the  tests  that
atomization was poorer at low  fuel  flowrates  than at  high flowrates.   Good
atomization quality  depends on both  a  high  pressure  drop across the  nozzle  and
on  a high  frequency  of fluid  pressure  pulses.   Thus,  the scheme  to  adjust  fuel
flowrate   by changing  pump   speed  will  not  provide  for  optimum  atomization
quality over the range of required flowrates.

C.  Second Generation  Pump Design

     The  first  generation design  had   two  major   shortcomings:   (1)  lack  of
balanced  fuel  delivery between nozzles  at  all flowrates,  and  (2)  poor  atomization
quality  at low pump  speeds  (flowrates).   In  order  to  overcome these  a second
generation pump was  designed  as  shown  in Figure  13.    It uses  the nozzles  and
pump eccentric housing from the first generation design.

     The  second generation design  has a  single  piston  that  operates  at  constant
speed  to  provide  good  atomization at  all  flowrates.   Changes  in  flowrate   are
accomplished  by  adjusting the  pressure  regulator.    Low flowrates  are accom-
plished  by  reducing the spring  tension in  the  regulator  piston  which allows it to
displace  as the  pump  cycles  and  accommodates  the pump  output.   At higher
flowrates,  the  spring   tension   on the  regulator  piston  is increased  so that  part
or  all  of  the  pump output is  discharged  through  the  nozzles.   Discharge through
the  nozzles occurs  when  the   pressure required  to further displace  the  regulator
piston  exceeds  that  required  to   force  fuel  through   the  nozzle.    Maximum
flowrate   occurs  when  the  spring  tension  in the  regulator  piston  is  sufficiently
high to preclude any piston  displacement before the nozzles  discharge fuel.
                                           A-4

-------
Adjusting Screw
                                                  Valve Stem (Steel)
                                                  Fuel Channels
                                                 Valve Spring

                                                 Clearance Between
                                                 Plunger and Drive


                                                 «	Methanol Inlet
                                                 Adjusting Spring
              FIGURE 1.  SwRI-DESIGNED NOZZLE
                                     A-5

-------
     Clearance
   Check Valve Outlet (Viton)

   Check Valve (Inlet) (Delrin)

    Methanol Outlet
     to Nozzles

   O-ring (for friction)

    Lip seal (Teflon)


Drain
   Housing (Aluminum)

   Seal (Viton)
     Electric
Pump
Symmetrical
About Center!
Line
    Crankcase
     (Aluminum)
     Bearings
      (3 typ)
                  Electric
                   Motor
                   Drive
                        FIGURE 2.  METHANOL FUEL PUMP
                                           A-6

-------
 c
"E
LU
I—
<
o
_i
LL

LU
^
ID
_J
O
250


225


200


175


150


125


100


 75


 50


 25
-iI    i   |	1

    NOZZLE FLOWRATE

           vs

       PUMPSPEED

 1-2-86 DATA (TEST NO.l)
                                                     T
                T~
                         NOZZLE NO. 1A

                         NOZZLE NO. 2B

                         NOZZLE NO. 3C
                         NOZZLE NO. 4D
      0 I  I
            I   i   I
      j	L
j	I	i	I	L
         600   800   1000   1200  1400   1600   1800  2000  2200  2400


                           PUMP  SPEED  (rpm)
                    FIGURE 3.   NOZZLE FLOWRATE VS. PUMP SPEED

-------
     200
   (D
     190
     180
     170
00
O
     160
     150 -
  O
  >
   140
          T	1	r
                         T	1	r
                                   RIGHT PUMP (B AND D)
                                          LEFT PUMP (A AND C)
                             1-
        PUMP FLOWRATE
              V8
          PUMP SPEED
        2-86 (TEST NO. 1)
I  .   I
       600  800  1000 1200 1400  1600 1800 2000 2200 2400
                    PUMP SPEED (rpm)
                    FIGURE 4.   PUMP FLOWRATE VS. PUMP SPEED

-------
I
VO
        250
        225
     ^  200


    "E   175
    LLJ
    h-
    <
O
__J
LL

LU
     150


     125


     100


     75


O   50


     25


      0
          -J   I—T
                     I
                             NOZZLE FLOWRATE

                                    vs

                                PUMPSPEED

                           1-2-86 DATA (TEST NO.2)
                                                            T
T
                                                       NOZZLE NO. 1A

                                                       NOZZLE NO. 2B

                                                       NOZZLE NO. 3C
                                                       NOZZLE NO. 4D
                                                       	I	I	J	J_J
         600   800   1000   1200   1400  1600  1800  2000  2200  2400


                            PUMP SPEED (rpm)
                          FIGURE 5.   NOZZLE FLOWRATE VS. PUMP SPEED

-------
 0)

 "o  200
K)
    190
    180
   170
UJ

< 160
150
    140
    130
    120
 o
 LiJ

                             RIGHT PUMP (B AND 0)
                         < _
                       PUMP FLOWRATE
                             V8
                         PUMP SPEED
                       2-86 (TEST NO. 2)
                                  LER PUMP (A AND C)
       600  800  1000  1200  1400 1600 1800 2000 2200 2400
                    PUMP  SPEED (rpm)
                 FIGURE 6.   PUMP FLOWRATE VS. PUMP SPEED

-------
 c
 E
 E
<
o
_i
LL
LU
o
    400
    350
300
    250
    200
150
     100
     50
       0
                      NOZZLE FLOWRATE
                              MS
                          PUMPSPEED
                    1-2-86 DATA (TEST NO.3)
NOZZLE NO. 1A
NOZZLE NO. 2B
NOZZLE NO. 3C
NOZZLE NO. 40
         600    800   1000   1200   1400   1600   1800  2000  2200  2400
                      FIGURE 7.   NOZZLE FLOWRATE VS. PUMP SPEED

-------
200
                                RIGHT PUMP (B AND D)
                                  LEFT PUMP (A AND C)
                                    PUMP FLOWRATE
                                          vs
                                      PUMP SPEED
                                  1-2-86  (TEST  NO. 3)
=> 130 -
	I	
                       	I	
 110
    600   800  1000  1200   1400  1600  1800  2000  2200  2400
                     PUMP SPEED (rpm)
                FIGURE 8.    PUMP FLOWRATE VS. PUMP SPEED

-------
 c
 E
LLJ
t—
<
o
_j
LL
UJ
O
250

225

200

 175

 150

 125

 100

 75

 50

 25
   0
          T
    NOZZLE FLOWRATE
           vs
       PUMPSPEED
 1-2-86 DATA (TEST NO.4)
                                  NOZZLE NO. 1A
                                  NOZZLE NO. 2B
                                  NOZZLE NO. 30
                                  NOZZLE NO. 4D
         600
700
                         800
900
1000
1100
1200
                       FIGURE 9.   NOZZLE FLOWRATE VS. PUMP SPEED

-------
 CD
 ^
 O
    210
   200 -
^E
 |  190
LU
(—
<
    180
 O  170 -

 LL

 LU  160
 O  150
        1    r
       600
                700
                                      RIGHT PUMP (B AND D)
                                     LEFT PUMP (A AND C)
                                         PUMP FLOWRATE
                                               VS

                                           PUMP SPEED
                                       1-2-86 (TEST  NO. 4)
800      900     1000

PUMP SPEED  (rpm)
1100
1200
                   FIGURE 10.  PUMP FLOWRATE VS. PUMP SPEED

-------
c
E

E

LU
h-
<
o
_j
LL

LJJ
O
250


225


200


 175


 150


 125


 100


 75


 50


 25


   0
-|  i    |   i	1   i   |   i   |—


         NOZZLE FLOWRATE

                vs

            PUMPSPEED

       1-2-86 DATA (TEST NO.5)
                            NOZZLE NO. 1A

                            NOZZLE NO. 2B

                            NOZZLE NO. 30
                            NOZZLE NO. 4D
J	L
                      1
J	L
            1
1
j	L
1
         600   800   1000   1200  1400  1600  1800   2000  2200  2400


                           PUMP  SPEED (rpm)

                      FIGURE 11.  NOZZLE FLOWRATE VS. PUMP SPEED

-------
 CD
    Z IU
   200
    190
   180
cr  170
   160 -
   150 -
E


LLJ
O
LL

LU
O
   130 Li
           1  T
                                    RIGHT PUMP (B AND D)
                                     LEFT PUMP (A AND C)
                            PUMP  FLOWRATE
                                  vs
                              PUMP SPEED
                          1-2-86  (TEST NO.  5)
       600   800  1000  1200  1400  1600   1800  2000  2200  2400
                        PUMP  SPEED  (rom)
                  FIGURE 12.  PUMP FLOWRATE VS. PUMP SPEED

-------
     (4) PLflCES.
                                           TVP (4) PLACES
                                           ADJUST FLOW 7O
                                      O3-32S2-30I-Z (TO
                                             CONCEPT
            GEMERAT/OM  Ftl£L
                                26401 '8352-403
SOUTHWEST RESEARCH INSTITUTE
        IAN AMfONIO. THAI
FIGURE 13. SECOND GENERATION SwRI-DESIGNED FUEL INJECTION PUMP
                             A-17

-------
              APPENDIX B




        DESIGN SPECIFICATIONS




IN-CYLINDER DISSOCIATION OF METHANOL

-------
          DESIGN SPECIFICATIONS

"IN-CYLINDER DISSOCIATION OF METHANOL"
          SwRI Project No. 03-8352

         EPA Project No. 68-03-1984
               Submitted by

           Ramkrishna G. Phatak
      Engine Research and Development
  Department of Engine and Vehicle Research
         Southwest Research Institute
             November 30, 1984
                     B-2

-------
INTRODUCTION

      The objective of engine design specifications  is to provide a guideline for the
project work.  Consistent  with  the technical objective  of the program "In-Cylinder
Dissociation of Methanol," the various alternatives considered and the design specifica-
tions presented in this document are aimed at developing a methanol-operated engine
with  fuel economy,  driveability and exhaust emissions at  least equal to the automotive
gasoline engine selected for the experiments.  The method of  achieving  this goal are
discussed in the appropriate section of this document.

COMBUSTION CYCLES, LOCATION OF CATALYST AND FUEL CONSIDERATIONS

      Automobile  engines  of  the four-cycle  type operate either  on gasoline  or diesel
fuel.   The combustion cycle of the gasoline engine is the Otto cycle, in which  the
charge  induced during the  suction stroke is  a homogeneous mixture of fuel (gasoline)
and air. The diesel  engine  operates on the diesel cycle, or constant pressure  cycle, in
which the charge induced into the engine's cylinder consists of air.  Fuel is injected into
the engine's  cylinder at an appropriate time in the cycle.  In  Otto cycle engines,  the
combustion of the fuel/air  mixture is initiated by high voltage  electric sparks through
the spark plugs.  The combustion of diesel fuel in a diesel cycle engine is accomplished
through self-ignition.

      Because of the  self-ignition properties of the fuel,  diesel engines do not require
any ignition  assistance.  However, to initiate  this process,  the fuel is required to be
admitted to  the engine  when  the  combustion air temperature  and pressure  are
sufficiently high  to  cause  the  fuel to self-ignite. In diesel  engines, the required  high
pressures and temperatures are  achieved through the use of a high compression ratio.
Also, the self-ignition property of the  fuel enables the engine to operate over a wide
fuel/air ratio.

      Change in  load  is achieved by altering the amount of fuel admitted into  the
combustion space of  the engine.  The air  supplied  to the engine  is controlled  only
through the engine's design, and no provisions are required to alter the air flow to the
engine.   Use of higher compression ratios, and the engine's ability to operate on wide
fuel/air ratios, makes the diesel cycle engine thermodynamically  more efficient  than
                                             B-3

-------
the Otto cycle gasoline engine.  The compression ratio of Otto cycle gasoline engines is
usually  limited to 10:1  due to fear of preignition/detonation of the  fuel/air mixture.
Although combustion of fuel-air mixtures much leaner than chemically-correct (stoichi-
ometric) is theoretically possible,  burning in the internal combustion engine  poses
several  problems.  In conventional Otto cycle automobile engines,  the fuel-air  mixture
is  supplied at or very near  stoichiometric, and  load  is  controlled by regulating the
amount  of  air and fuel induced  into the engine's cylinder. The amount of air is varied
by opening or closing a butterfly valve  in  the intake  manifold/carburetor.  Such  an
obstruction to the airflow reduces the volumetric efficiency of the  engine.  The low
compression  ratio,  inability  to  burn a lean  fuel-air  mixture  and loss of  volumetric
efficiency make the Otto cycle  engine thermodynamically less efficient than the diesel
cycle engine.

      The concept of in-cyUnder  dissociation of methanol can be adapted to either
diesel or Otto cycle engines; however, there are certain advantages and disadvantages
associated  with  each approach.   Before the applicability and suitability of the cycles
are considered, it is essential to  consider the possible  locations of the catalyst  in  an
engine.   In the specific  context of  this particular project,  it is possible to locate the
catalyst chamber within  the combustion space of the engine. Alternatively, a catalyst
chamber can be  designed as part of  the cylinder head.   The  first approach  has the
following disadvantages:

      1)    contamination of  the catalyst  with  combustion products, rendering it useless
           in a short period of time,
      2)    exposure  of  the catalyst to high combustion heat, causing a heat loss  that
           results in lower cycle efficiency,
      3)    the necessity of supplying  methanol, and therefore dissociation, during the
           intake cycle, resulting  in low volumetric efficiency of the engine,
      4)     premature combustion  of dissociated methanol,  resulting  in loss of engine
           efficiency and damage  to the engine's structure,
      5)     inability  to  use  exhaust heat (waste heat) for dissociation of methanol,
           thereby  being unable  to improve  the thermodynamic  efficiency  of the
          engine.
                                         B-4

-------
     These disadvantages can be eliminated if:


     1.    the catalyst chamber is isolated from the engine's combustion space,

     2.    the catalyst chamber location is designed to use the  waste heat from the
           engines exhaust gases.


     Since differences in the combustion cycle are  due to  the characteristics of the

fuel, it is essential that the hydrogen (H) and carbon monoxide (CO) are characterized
as  fuel  for  an  internal  combustion  engine.   The  primary products  of  dissociated

methanol are 66 percent  hydrogen, and  33  percent carbon monoxide.  Both gases are
suitable  for  internal combustion  engines.   Table 1  shows  the typical properties of
methanol, hydrogen,  and carbon monoxide.


                            Table 1.  Properties of Fuels
Heating Value
      BTU/lb

Ignition Temperature
      oF

Theoretical air requirements
      Ib/lb

Ignition Limit, air-fuel ratio

      Lower
      Upper
                                   Methanol     Hydrogen
                                    (MeOH)         (H)
8471
 842
 6.4
 5.5
 26
51600
 1040
  34
  4
 77
Carbon Monoxide
      (CO)

      4321
      1121
       2.5
      12.5
       75
     For combustion cycles and  operating efficiency, ignition temperature  and lower
ignition limits are important.  Both gases have high ignition temperatures and a very
wide ignition limit, which is highly desirable for fuel economy and power.


     Another important consideration is  the  flame  speed.  Both gases have very high
flame speeds.  Use of hydrogen  fuel in spark-ignition  internal combustion  engines is
associated  with  several problems,  such  as  flashback,  preignition,  severe cylinder
                                           B-5

-------
pressure oscillation, and rough running due to a high rate of pressure rise.  However, the
presence oi CO in hydrogen suppresses all the above without affecting flame speed.

Combustion of Dissociated Methanol in a Diesel Cycle

     Combustion of dissociated methanol in a diesel cycle  is possible. Because of high
ignition temperatures, however,  the compression  ratio of such an engine must  be  in
excess of  80:1.  It would not be possible to  convert an existing  gasoline automotive
engine to  run at such a high compression ratio.  Such an engine, if developed, would be
very bulky and would  result in a loss of overall fuel economy.  Administering the  gases
to the cylinder would also be a problem.  The pressure required to inject the  gas into
the combustion chamber  would be more than 10,000 psi. Dissociation of methanol in a
catalyst chamber results  in pressure rise. The pressure depends on the catalyst  used,
the temperature, and the volume of the chamber.  Preliminary calculations  indicate
that for  a one  liter  engine under  the  best  possible conditions,  the pressure in the
chamber will be approximately 150 psi.  This being far less than the required injection
pressure,  additional power will have to be expended to increase the pressure of the
gases.  The additional work will obviously result in a loss in fuel economy.

     The  only advantage of using the diesel cycle would be an excellent control on the
combustion by regulating the  rate of fuel (gas) injection. However, the anticipated poor
fuel economy and higher NOX emissions make this cycle unsuitable.

Combustion of Dissociated Methanol in Otto Cycle

     Conventional Otto cycle, in which a homogeneous fuel-air mixture is induced into
the engine's cylinder,  is certainly not suitable  for this project. The fuel in the  catalyst
chamber at about 150 psi can be admitted to the engine during the induction stroke with
the help of a third valve in the cylinder head.  However,  introduction of  gas  into the
cylinder during the induction stroke  will  reduce the engine's power output considerably
due to a loss in the volumetric efficiency.  The fuel is required  to  be  admitted only
after  the exhaust  valve is closed; however,  this would not be  a problem.  The
combustion of fuel in such a cycle will be initiated through a spark at an appropriate
timing.  The  high flame  velocity means that the engine can run  on  a highly retarded
                                          B-6

-------
timing (an advantage to reduce the NOX emissions) and timing changes for various loads
and speeds would be minimal.  Engine load control will be  accomplished through  the
metering  of  fuel and/or  air  induced into the engine.  As  stated earlier,  the major
disadvantage of the Otto cycle as described above is reduced power output.

Combustion of Dissociated Methanol in a Modified Otto Cycle

      Engine  power can be  considerably increased if  the fuel is supplied after or just
before the intake valve is closed during the compression stroke.  The cylinder pressure
at the time of intake valve closing  is estimated to be equal  to or  slightly  less than
atmospheric pressure.  Therefore, introducing the dissociated methanol fuel (already at
approximately 150 psi) is not only easier but has other beneficial effects. It is expected
that the effects  of introducing the high pressure gas  will be similar  to a supercharged
engine. Increased power and efficient combustion can be therefore expected from  the
engine. Furthermore, by  shaping the  port and suitably locating the fuel gas inlet valve
in the cylinder head,  air swirl can be generated. Air swirl in conjunction  with the high
flame speed is expected to improve the combustion efficiency.

      It is well known that air standard efficiency of an engine increases with  increases
in the engine's compression ratio. Considering the high-octane rating of both the gases,
it appears possible to increase the compression ratio of the engine.  The actual increase
will be decided upon the structural strength of the engine selected for the experiment.

      Initiation of combustion will be through  a spark plug.  As  discussed earlier,  the
engine can be operated with a highly retarded ignition timing. In addition, the required
changes in the timing due to changes in speed and load will be minimum.  Load control
can be through regulation of both air and fuel; however, it is advantageous to operate
the engine at open throttle to maximize power output.

      Regardless  of the combustion cycle selected,  the engine will operate in different
modes during  starting/warm-up,  and  when  the engine is warmed up.  During  the cold
starting and  warming up  period, there would  be insufficient heat available  to cause
dissociation of methanol.   The  fuel, therefore,  will  have  to be admitted  with  the
induction air.  The fuel can  be supplied either through a carburetor, a throttle-body fuel
                                         B-7

-------
injection system, or a port injection system. Once there is sufficient heat available for
dissociation of methanol, the mode of fuel should be gradually changed from the liquid
fuel in the induction manifold to dissociated methanol in the cylinder.

      Based on the above descriptions, the following design specifications are suggested
for the combustion cycle:

      1.    A  modified  Otto cycle engine operating in  dual-mode; that is, a standard
           Otto cycle during starting and warming up and a modified Otto cycle during
           operation.
      2.    During starting,  liquid  methanol fuel will be supplied in the manifold or port
           injection.
      3.    During  operation, dissociated  methanol will  be  supplied to  the  engine
           cylinder through a third valve.
      4.    Dissociated  methanol  will be supplied  to the engine cylinder  after or  just
           before the intake valve is closed.
      5.    Methanol  will be  injected  into the  dissociation  chamber  soon after  the
           chamber is emptied into the engine cylinder.
      6.    The charge will be ignited with the help of a spark plug.
      7.    Open  throttle (or no throttle) operation is desirable during the modified Otto
           cycle operation.

DEVELOPMENT OF CATALYSTS AND CATALYST CHAMBER SPECIFICATIONS

      The selection of catalyst materials, catalyst configuration, and the  pre-chamber
dimensions  will  be  made  following  a bench  type experiment  to screen different
materials and obtain other  design data.  Performance  data on methanol dissociation
catalysts  for automotive use are  reasonably well-known for steady-state atmospheric
conditions.  This  data are not readily  available for  transients and high pressures in  the
range  of  2 to 10 atmospheres.   Such data are  essential to  the design of  the pre-
chamber-type dissociator proposed in  this program.

      Figure 1 shows the design of the reaction chamber for the test  rig. The design is
an approximation of the chamber  to be used on  the engine.  The  catalyst chamber is
                                          B-8

-------
•GAS SAMPLING VALVE
     SCREEN
GAS COLLECTION
     AREA
 CATALYST
   BED
   AREA
                                                                                METHANOL INJECTOR
VAPORIZATION AREA


ELECTRICAL HEATER
                                                                                GAS SAMPLING VALVE
                                                                           SCREEN
                      FIGURE 1.  SINGLE PASS CATALYST TEST CHAMBER

-------
annular  in  cross section  with an  electrical  heater element  through  the  center,
representing the  heat from  the exhaust flow. The total volume of the chamber will  be
100 cm3.  A fuel injector  will spray methanol onto a set of copper fins.  This is  to
provide space and time for vaporization to take  place.  Methanol vapors will then travel
through the catalyst bed.  Following timing cycles as  in the engine, a blowdown  valve
will open, simulating the release of dissociation  products into the combustion chamber.

      The experimental chamber will be instrumented for pressure rise, temperatures  at
several locations, and  gas  chromatography will  be used  to  analyze the  products  of
dissociation.  The gas analysis  will include the determination of CO, H£, CJ-ty,  H2O,
(CH3)2O, and unreacted  methanol.  Mass  balances will be made in order to access
possible  formation  of  solid carbon in the  catalyst bed.  By  use of a high speed gas
sampling  valve,  the gas analysis history can be  attained  at  various  segments of the
pressure rise following the injection of methanol.

      This  experiment will  provide information  on the  transport  properties  of the
catalyst  matrix, namely, the effects of conduction and heat capacity.  These properties
play a role during the thermal recharge cycie and  the dissociation cycles.  The overall
reaction rate depends  on  both the  chemical rate and the heat and mass transfer to the
catalyst.  At sufficiently high temperatures  (300°C), the chemical reaction rate is  quite
fast and should not be rate-limiting. The transport steps are:

      1.    Mass transfer of reactants to the catalyst and the desorption of products.
      2.    Heat transfer to the catalyst bed.

      In a closely packed  bed, the  gases are in  intimate contact with the catalyst and
diffusion lengths  are short.  However, if desorption of products is  slow,  the reaction
rate  will  be  limited by  the  number of available sites.   If  the rate is limited  by
desorption of  products, the  effect  can be ascertained by varying the temperature and
the catalyst particle size.  Increasing the  available surface area of the catalyst will
increase  the  rate  if  mass transport and desorption of  products  is  rate-limiting.
Dissociation rates will be determined from pressure rise data.
                                           B-10

-------
     The effects of bulk mass transfer could be examined by altering the geometry of
the reaction chamber.  Such a modification  may take  the  form as shown in Figure 2.
Here an internal  cylinder has been added to the design, creating a longer flow path for
the reactants, hence a higher flow velocity through  the catalyst  bed.

     Heat transfer  to the  catalyst bed  is an important aspect of  the dissociation
process.    Both vaporization  and dissociation  of  methanol  are very  endothermic
processes,  which require the  maintenance  of  a high  temperature (300°C).   Several
high-response thermocouples will be installed in  the reaction chamber shown in Figure 2
to assess temperature gradients and rates of heat recharge.

     Finally, the data from these tests,  will be  used  to predict  rates of dissociation in
real engine tests.  The simulated catalyst chamber  test data will also be of value in
measuring catalyst performance and longevity.

DESIGN SPECIFICATIONS - ENGINE AND
COMBUSTION CHAMBER CONSIDERATIONS

     Engine and the combustion chamber configurations must be selected to match  the
objective of the project and the above selected combustion cycles. Also important in
the selection  of  the engine configuration  is the  development of hardware.  In this
particular project, due consideration must be given to  the design  and development of
the cylinder head.

     Since  the  objective of the project  is  to study  the in-cylinder dissociation  of
methanol concept for an automobile engine, the selected engine  must be suitable  for
automotive applications. A  four-cycle engine is best  suited for  the experiments.  From
a dissociation point of view,  an air-cooled engine may be preferred over a water-cooled
engine; however,  air-cooled gasoline engines are no  longer used on automobiles.  It is
preferred that we select a late model automobile engine of the state-of-the-art design.
On  a four-cycle  engine running at 6000 rpm,  the injection event  repeats  every 20
milliseconds; therefore,  regardless of how the fuel  is admitted, the minimum  time
available for dissociation of  methanol is at least 20 milliseconds.  Since this available
time is more than that required, engine speed is not likely to be a limiting factor.  All
                                        B-ll

-------
          •GAS SAMPLING VALVE
w
i
           CATALYST

              BED
                                                                                         METHANOL INJECTOR
                                                                                       ELECTRICAL HEATER
                                                                                       GAS SAMPLING VALVE
                                                                                       SCREEN
                                FIGURE 2.  DOUBLE PASS CATALYST TEST CHAMBER

-------
late  model  automotive engines are of the overhead valve type, a superior  design as
compared with the side valve engines.  In this project, too, an overhead valve engine is
highly  desirable.   A  conservative engine design  (structural)  is  more suitable.   As
discussed in the earlier sections, some thermal efficiency gains are likely if the engine's
compression ratio is  increased.   Such an increase is possible  only  if  the engine's
structure can take the additional load due to  higher working pressures.  This particular
point will be considered in the engine selection process.

     The engine  design  configuration  should  be  suitable  for the  proposed design
modifications.  As discussed earlier in this document, placing the  catalyst chamber in
the combustion space has many  disadvantages  and would  require  extensive modifica-
tions to the engine. The  most promising alternative is the use of heat from the exhaust
gases,  and possibly from the  cooling water  jacket.   In order  to use heat  from the
exhaust gases, it is desirable that  the catalyst  chamber be located in the  exhaust port
of the  cylinder head. For the combustion cycle selected, and for the highest volumetric
efficiency, it is essential that the  dissociated methanoi be supplied to the  cylinder just
before or soon after the intake valve is closed.  This requires that there is a fuel supply
valve in  addition to the  intake and exhaust valves.  Figures 3 and 4 show a conceptual
layout for the  cylinder head design.  From this layout  it is clear that we need ample
space  in the exhaust port to accommodate the catalytic chamber.   It  is  therefore
required that the  engine's cylinder head design  configuration has a cross-flow head, or
that it be possible to  convert  the  head to a cross-flow type.  For efficient combustion
and  high BMEP,  several  automobile  engines  use  hemispherical  heads and four valves
driven through double overhead cams. Although the hemispherical combustion chamber
and  four valves  are advantageous, there are very difficult  hardware  problems.   The
hemispherical combustion chamber heads with four valves usually do not have  the space
to accommodate a third valve  and  the large space required for  the catalytic  chamber.
Furthermore, machining the valve gear surface  at complex angles (hemispherical heads)
is much  more difficult  than valve gear surfaces  parallel or at  right angles  to the
cylinder  axis.   It  is felt that the type  of fuel, together with the  intake swirl,  will
produce  results  comparable  to  a hemispherical  combustion  chamber with  a  much
simpler "bathtub" shaped  combustion chamber.  From a hardware modifications point of
''lew, an overhead earn is prefer^d.    The engine will most likely need an additional
camshaft to drive the third  valve.   In order to  keep the hardware development costs to
                                           B-13

-------
                                                                                AHfiUS T K»i f £.     I
W
 I
                                                                      fxtauir VQIT
                                                                                           Dfl&OCJJbrfD sttfriv&JOt.

                                                                                           FKQOVtrt rp crtr-spfif
                                                                                    TOLKftANCC*
                                                                               QT* Mtoo
                                                                                                                    SOUTHWCST KSIAKH MSTITUTI
                                                                                                                            HfAD  LflYOUT
                                                                                                                                            of
                                                                                                                       COO« ID«P«T MO

                                                                                                                       26401
                                FIGURE 3.  CYLINDER HEAD LAYOUT - DOUBLE PASS CATALYST CHAMBER

-------
              ^Jtffl rOtj THJfO YAt. ¥£,
w
 I
                                                                                         fSXJUST C*»
1

i



\
7
\
2>JSS
ro c
1
                                                                      £X##US7 POf7
                                                                                                            SPACf *0^ CArAtyg T
            ~r*T^r> wot.r£.
                                                                        TOLCN NCH
                                                                                    C?T2O*J_ 
-------
a minimum, we are proposing a four-cylinder in-line engine.  Also, it is desirable
that we select an over-square engine.
     Based on the above descriptions, the following  set of design specifications are
recommended:

     1.    A four-cylinder, spark-ignition automobile engine
     2.    Engine speed up to 6000 rpm
     3.    Larger bore engine; that is, stroke/bore ratio as small as possible
     4.    Two overhead valves, one cam engine
     5.    Cross-flow type cylinder head design
     6.    A conservatively designed engine

Based  on  the above  requirements,  two  engines  are considered  suitable  for  the
experiments. The engine specifications are shown in Table 2.
Make & Model
Bore x Stroke
    (mm)
Displacement
    (liters)
Cam
Maximum Power
Maximum Torque
Fuel System
                          Table 2.  Engine Specifications
                                       1
Datsun 200 SX, 1982
     84.5 x 88

        2.2

     Single OH
102 Bhp (9 5200 rpm
129 Ib ft @ 2800 rpm
 Bosch L-3etronics
  U.S. VW Rabbit
    79.5 x 86.4

        1.6

    Single OH
74 Bhp (§ 5000 rpm
90 Ib ft @ 3000 rpm
 Bosch K-3etronics
                                           B-16

-------
     The following modifications to the engine are recommended:
     1.    Fit  a heat exchanger  (catalytic  chamber) into the exhaust port of each
           cylinder.
     2.    Provide the engine with a third valve  to admit the fuel into the engine.
     3.    Provide valve gear for operating the third valve.
     4.    Design a new exhaust manifold.
     5.    Design/modify any other component as a result of the above modifications.

DESIGN SPECIFICATIONS FOR CATALYTIC CHAMBER

     The program  objective is to use the  waste heat (heat lost to cooling water and
heat lost to exhaust gases).  Although part  of the cooling water's heat can  be used  for
evaporation of  methanol, the temperature of cooling water is far less when compared to
the required temperature for dissociation of  methanol.  The primary  heat source  for
methanol dissociation has to be the hot exhaust gases.  Maximum heat transfer can be
achieved if the heat exchanger is placed in the exhaust port and as close as possible  to
the exhaust valve opening.  The size and surface area of the heat  exchanger required
cannot be finalized at this time.  An experiment is being planned to determine the size
of the heat  exchanger.   Two designs, shown in Figures 3 and 4,  are possible.  The
material for  the catalyst  chamber will be either copper  or stainless steel.   The space
provided in the catalyst  chamber will be filled with the  catalyst material.  Liquid
methanol will be supplied to the catalyst chamber through a fuel  injector. In operation,
the temperature  and pressure in  the  catalyst chamber  will be monitored.  Detailed
design  specifications of the catalyst chamber can be  determined at  the completion of
the experiments.

DESIGN SPECIFICATIONS - GOALS FOR DRIVEABILITY,
FUEL ECONOMY, AND EXHAUST EMISSIONS

     The project's goal will  be  to  achieve driveability, fuel  economy, and exhaust
emissions equal to or better than the engine selected for these experiments.
                                           B-17

-------
     Driveability  measurements/assessments are  usually  subjective  in  nature.   For
these discussions we are  referring to  the  report  on CRC Project  No. CM-118-84.
Appendix A shows the detailed procedure for evaluating driveability.

     Individual raters are  provided with a procedure  to  start and idle the engine and
drive the car(s) at specified speeds and distances. The driveability assessment is based
on the  severity of  the  malfunctions, such  as  (1) hesitation,  (2) stumble,  (3) surge,
(4) stall, and (5) backfire.   The  malfunctions are rated  subjectively.  For example, a
trace of malfunction and heavy malfunction are rated as 1 and 4, respectively.

     In the current project, testing of an automobile to CRC procedure for driveability
rating is not planned. However, the malfunction factors suggested by CRC will be used
to compare engine performance.  In our experiments, the selected engines (baseline and
after modifications to run on dissociated methanol) with all its controls will be suitably
instrumented to measure hesitation, stumble, and  surge.  For  example, the degree of
time delay between  opening  the throttle and  an  increase in  engine speed  could be
measured and compared to quantify hesitation.  Likewise, surge characteristics of an
engine can be assessed by monitoring the dynamometer output.  For the engine running
on dissociated methanol, acceptable driveability will be achieved by suitably modifying
the fuel control system.  As discussed elsewhere in this document, the  engine will likely
have the control systems,  one for liquid  injection  of methanol  in the ports, and the
other for liquid injection of methanol  in the catalyst  chamber.   It  is  anticipated that
both systems will have to be tuned to achieve acceptable driveability.

     The project's goal is to achieve fuel economy with  dissociated methanol equal to
or better  than the  selected engine. The fuel economy testing and evaluation will be
based on dynamometer testing of  the engine. The engine(s) will  be dynamometer tested
for predetermined speeds and  loads and the fuel consumption  will  be measured.   The
first step in determining the speed and loads is to select a driving cycle.  Using the in-
house computer program and the selected engine/vehicle  data, the operating speed and
engine  load will be calculated for the driving cycle.  The fuel consumption  of the engine
operated on both  gasoline and dissociated methanol will be compared to assess the  fuel
economy gained/lost.  It is anticipated that improved fuel  economy  will  be  achieved
through the use of  dissociated  methanol.   Fuel economy  goals will be  achieved by
                                         B-18

-------
maximizing the use of waste heat  and optimizing the engine to run on lean fuel-air
mixture. Slight supercharging of the engine and air swirl will be attempted to improve
the combustion efficiency of the engine.

     Testing of the automobile with the dissociated methanol engine is  not planned in
this project.  It is therefore not possible to relate the engine's emissions to 1975 Federal
Test Procedure.   Our emission goal, therefore,  is to  achieve  identical steady-state
exhaust emissions from the gasoline and dissociated methanol versions of the engine.  In
both cases the engine  will be  tested on  a  dynamometer for  pre-determined speed-load
conditions.  It is possible to use the same speed-load  point for both fuel economy and
emissions testing.   The  emissions measurements  will  include  total HC,  CO, Nox> and
aldehydes.  The measurements will be made after the  exhaust  catalyst.   The reduction
in emissions will be achieved through  combustion modifications.  For  example, our
attempt will be to  run  the engine at or below an equivalence ratio of .5 with as retarded
an ignition timing as possible.

DESIGN SPECIFICATIONS - AUTOMATIC CONTROLS

     The dissociated  methanol operated  engine will be equipped with two control
systems: one  for  port injection of  liquid methanol  and the other  for injection of
methanol in the catalyst chamber.  It is suggested that for liquid  methanol injection in
the ports, we replace the electronic control system on  the engine with  a new adjustable
system.   We  are  proposing two control  systems, one  each for liquid  methanol and
dissociated  methanol.    We are suggesting an intermittent fuel  injection  system(s)
similar  to  Robert  Bosch L-Jetronics   or LH-Jetronics  system.  Intermittent  fuel
injection is considered to be  advantageous for  maximum  dissociation  of methanol.
During  start and warm-up the port  injection of  liquid methanol would be operational.
The temperature in the  catalyst chamber will be measured and an appropriate signal
will be used to activate the  methanol dissociation system.  The in-port injection system
will either be cut off or kept  in operation; however, at a  much reduced  fuel  injection
rate.  At low idling and very low loads,  when the  exhaust temperature  is expected to be
low, the fuel supplied  to the engine  will be suitably shared  by the two systems.  The
engine will be fitted with both these systems at the beginning of  engine testing.  Initial
engine tests will be run without  the electronic  controls.  At  each  speed/load point, the
                                          B-19

-------
engine performance will  be optimized  by  optimizing the spark-timing for the lowest
possible equivalence ratio fuel-air mixture.  Also, warm-up  mixture and cold starting/
warm-up spark timing  requirements of the  engine will be evaluated and optimized using
the external controls.  Also,  engine tests  will be performed to optimize  the  ratio  of
liquid injected methanol in the port and dissociated methanol system at  several speed-
load points.   The above  optimization tests  will provide  a  guideline to program the
microprocessor in the automatic control systems.

      The control system will be  used primarily for  controlling spark timing, fuel-air
ratio and, if required,  exhaust gas recirculation. It is anticipated that NOX emissions  of
the dissociated methanol  engine  will  be lower  than the gasoline engine  due to the
engine's  ability to  run on low equivalence ratio fuel-air mixture.  Also, the retarded
ignition   timing  will  help  reduce the production of NOX.   An  EGR system will be
introduced only if the lean mixture and retarded ignition  timing fail to  bring the NOX
equal to the gasoline engine.  We  are planning to use  the fuel quantity supply signal  to
control the flow of exhaust gas to the engine.

      Our preliminary  calculations indicate that rate of heat  transfer to  the  catalyst
chamber will  not require any control.  Therefore, no attempt  will be  made to control
the flow of exhaust gases over the catalyst  chamber.

      An alternative control system for  this, engine would consist of essentially a speed
governor, controlling the  RPM through the  quantity of fuel injected.  The system would
be the electronic version  of the diesel's engine-type governor.   Depending  on  the load,
the engine would operate unthrottled (middle to maximum load) or throttled  (light load).
The  throttle plate will be required on  this engine, unlike the  diesel, because the charge
is homogeneous.  The leanest equivalence ratio obtainable  is  estimated to be 0.3 as
demonstrated by the Solar Energy Research  Institute  (SERI).  The diesel, on the other
hand, can idle at  an overall equivalence ratio  of 0.01  due to charge stratification.  Some
charge stratification is anticipated in this design; however, to what degree is difficult to
predict.

      Figure  5 is a block  diagram of  the control  system.    A proportional-integral-
derivative  (PID)  analog controller (or the  digital equivalent),  would be  used.   A  PID
controller uses both  the integral and the derivative of the error (difference between the
                                        B-20

-------
w
PROPORTIONAL-
  INTERGRAL-
  DERIVATIVE
 CONTROLLER
     (PID)
                    RPM
                 SET POINT
                                             ELECTRONIC
                                            PROPORTIONING
                                               CIRCUIT
                                                                     PULSE OUTPUT TO INJECTORS
                                                        MANIFOLD   INTAKE PORT
                                                        PRESSURE    INJECTOR,

                                                       £
                                               THROTTLE
                                                 PLATE
                                              CONTROLLER
                                                             THROTTLE PLATE
                                                                 SERVO
                                                     RPM
                                                    INPUT
                                                                                                •METHANOL
                                                                                                 INJECTOR
  CATALYST
TEMPERATURE
  SENSOR
                                                                                       L—ODKt C
                                                                           RPM PICKUP
                                     FIGURE 5.  SIMPLIFIED BLOCK DIAGRAM OF
                                          ALTERNATIVE CONTROL SYSTEM

-------
actual RPM and the setpoint) as well as the error itself to determine the control output.
It would for example provide enrichment of the mixture during an acceleration due to
the increase in the derivative of the error.

     The output of the PID would then be divided between the intake port injector and
the catalyst chamber  depending  on  factors  such as  the catalyst  temperature,  the
quantity of fuel required, and the rate of change of fuel delivery.

     The  temperature sensor in the  catalyst bed would divert  methanol to  the port
injectors if  the temperature was too low.  This condition would occur during cold start
or high load conditions  when there would not be sufficient heat to dissociate the fuel.

      During conditions of acceleration and of increasing  load but constant RPM, the
heat supplied to the catalyst will  lag the heat required for dissociation.  By diverting
methanol to the port injectors during these transient conditions, the thermal lag of the
catalyst bed can be avoided.

      While  examining  possible  approaches in developing this control scenario, it is not
yet clear whether analog is more appropriate than digital circuitry for this  project.

DESIGN SPECIFICATIONS - TEST PLAN

     Engine testing will  be done in three phases. During the first phase, called  the
baseline test, the  engine  will be  tested on an  engine  dynamometer to establish fuel
economy, exhaust emissions and driveability.  For  fuel economy and emissions measure-
ments, the engine will be tested at the following speed-load conditions:

                     Speed (rpm)        Load BMEP (psi)
                         Idle                 29.5
                         1750                10,18,35
                         1950                  39
                         2150                 11,43
                         2350                  48
                         3000      Corresponding to  max power
                         5000      Corresponding to  max power
                                         B-22

-------
     The speeds and loads in the first 6 rows are based on a vehicle fuel economy and

performance simulation computer program.  The input for the program  was based on an

average of 97 models of 1983 U.S. built cars (Table 3). Table 4 shows the output of the

computer simulation program.  Depending upon the engine and the car  selected, the

speed/load numbers in the first five rows will change slightly.  The speeds selected  in

the last two rows will depend upon the speeds corresponding to the max torque and max

power of the selected engine.


                    Table 3.  Input Data for Computer Program
                 1983 U. S. Production Cars -   97 Models

                        Average Statistics:
                             Frontal Area:
                                  Weight:

                            Displacement:
                                    Bore:
                                  Stroke:

                      Number of Cylinders:
                               Axle Ratio:
                Wheel Revolutions per Mile:
                              Wheel Base:
                           Fuel/Air Ratio:
                                  Timing:
                       Compression Ratio:
                        Engine Speed, rpm:

          Four-speed Standard Transmission

                         Computer Results
                                     MPG
                                EPA City:
                            EPA Highway:
                                   BMEP:
                                   FMEP:
                                     BMP:
                                    BSFC:
 20.6 sq. ft.
 2900 Ibs (inclusive of driver and pay-
 load)
 183.7 in3
 3.633 in
 3.306

 5.361
 2.9965
 780 rpm (assume GR 78-14)
 105 in.
 16 (14.93 stoichiometric, assumed)
 MBT (assumed)
 8.575
 2142 @ 55 mph
                                                AF = 26 ft2     EFF. AF = 20.6 ft2
  23.4 @ 55 mph
       18.6
  25.8 @ 50 mph
49.6 psi @ 55 mph
20.3 psi @ 55 mph
  24.6 @ 55 mph
.5458 Ibs @ 55 mph
25.6
19.0
27.4
43.2
20.6
21.5
.5900
      Engine rpm and load corresponding to about  20  mph will  be  used for evaluating
                                         B-23

-------
    Table f.   SwRI Vehicle Fuel Economy and Performance Simulation
HP Req't node: RL
Engine Type  SJ
Air/Fuc?l Ratio :  \t>. 0
Pressures: l)Exh (psia):  15.0
Temperatures (F): DAir:   70
Stoich Air/Fuel Ratio.  14 93
Vehicle Eff Frontal Area (ft2>
Nornal Lpadori Weight  (lb>:   320(1.
Orivetrain Type: ST
N/V Ratio (Top Gear)    38.95
Piston Displacement (i.n3>   103.70
Bore x Stroke  (in):  3.633 x 3.306
Compression Ratio:    8.6
Volumetric Efficiency:   85
Fuel LHV (BTIJ/lb):   19020.
Fuel Density:  5.86
Electrir Cooling Fan:  NO      Air Cnnd;
Ign Timing, (fron hBT  -t-RTD, -ADV): 0
Gearbox Ratios:     3.000    2.200     i
                                                 2)Pnro (in-Hg)•
                                                 2>CooldnT:   ISO.
                                                                   29  92
                                              20 . 60
                                                     L/R
                                           Rated rpn   4000
                                                                  1 .70
                              NO

                              500
                                                            1 . 000
                                                                       MA
                           Specifications:
            Max BHP 6> rp«                   120  P 4(1 DO.
            Redline                          4000  rpn
            Adher.ion Linit                   1387.0 Ibf
            Track                            57.7 in
            Uheelbase                        105.3 in
            H(?i(;ht                           30.0 in
            Ter,t Weight                      3200.  Ibn
            Weight Dist                       'JO
            Axle Ratio                       2 997
            «ph/iOOnrpn                      2E.67
            Gear Ratios
                 Gear 1                      3.000
                 Gear ?                      2.200
                 Gear 3                      1.500
                 Gear 4                      i.000
            Top Speeds in Gears:
                 Gear 1                      34.? nph
                 Gear ?                      46.7 nph
                 Gear 3                      613.5 nph
                 Gear 4                     102.7 nph
            Max Speed
            Redline
            Tine tu Speed
                  nph
                 li- 90
                 (i- fln
                 n- 7ii
                 0- t,Q
                 0- 50
                 0- 40
                 0- 30
                 0- 20
Fuel Conr.unp tion .

 Idle fy  3.83 )bn/hr
 MPG= 23.5 H 20.nph
 MPG= 20.8 P ?5.nph
 MPG= 27.7 (? 30 nph
 HPG= 32.5 ft 35. nph
 MPG= 26. 3 (? 45. nph
 MPG= ?7 4 e 50 nph
 rtPG= 25.6 e 55.nph
 MHC,'= 24 0 (•! 60 nph
 HMF.P (Ace* i)=  42.
 BMEP (Ace* 2)=  59.
 EPA City EstinatP:
 EPA  Hwy Estinafe:
 City/Hwp Conposite:
                             Perf orn.incp :
                     SPT
                      29 58
                      21 .37
                      14 .78
                      1 0 . 87<
                       7.96
                       5.43
                       3.46
                       2 . 2ft
                                113 69 nph
                                102.7(1 nph
       b p'

       Bl*
      Man Vac = 21.6  in-Hg
  don Vac= 19.4 in-Hg
  n.in Vae= IB.8 in-Hn
  Man Vac = 18.1 in-Hn
  Man Vac= 15.5 in-Hn
  Man V.ac= 14. U in-Hg
  Man Vac= 14.0 in-Hg
  Man Vac= 13 i in-Hn
  Man Vac= 12.2 in-Hg
;i    Man Vac= 13.0  in-Hg
;i    Man Vac=  9.4  in-Hg
 0  MPC   (L.A4)
        Steady-State  MPG
                                     RPM=1714
                                     RPM=2142.
                                     RPM--1753.
RP11= 1753.
RPM=1947
RPM-214,?.
RPH=2337
    ""'- 16
(2)

(3)
(4)
(4)
(4)
(4)
(4)
0
                                                         HMFP-
                                                         BHUf-
                                                         BrtFP-
  A/F= 16.0
     RPM- 1753.
       '1= 214J
10
10
17
28
35
3V
43
47

-------
driveability.   For better comparison  between  the gasoline engine and  the dissociated
methanol engine, we will try to quantify the malfunctions.

      In  the second  phase  of testing  the engine will be dynamometer tested with the
dissociated  methanol  as the fuel.  These  tests will  be run  with the  fuel  injection
equipment hardware but without the electronic controls.  For these tests the electronic
controls will be replaced with the manual  controls.  The objective of these tests  is to
verify the suitability of the catalyst chamber design for engine operation. The engine
will be tested for the following speed-load conditions.

              Speed (rpm)                 Loads (F = Full Load)
                  1000                    F, 0.8F, 0.6F, 0.4F, OF
                  2250                    F, 0.8F, 0.6F, 0.4F, OF
                  3000                    F, 0.8F, 0.6F, 0.4F, OF
                 WOO*                    F, 0.8F, 0.6F, 0.4F, OF
     (*or rated speed of the engine)

      Alternatively,  the speed-load points  used in the first set of  tests can be used in
place of the above speed-load points.  At each speed-load point maximum fuel  economy
will  be  achieved  by  varying ignition timing and  optimizing the use  of dissociated
methanol as  engine  fuel.   The  optimized  ignition   timing  and  the  fuel  delivery
characteristics of both  fuel injection systems for each  speed-load point will provide
input data for programming the electronic controller.

      Prior to the third phase of testing  the electronic controllers in the fuel  injection
systems  will be programmed. In this phase the engine will be tested for fuel  economy
emissions and driveability  with the electronic controllers.  If the test  results  are found
to be short of the goals, then further  modifications will be considered for the catalyst
chamber and control systems.  For example, if the NOX emissions are higher  than the
goal,  then an EGR system and a related control will be added to the engine.
                                           B-25

-------
        APPENDIX C






EVAPORATOR DESIGN ANALYSES




PERFORMED BY A CONSULTANT

-------
              EVAPORATOR DESIGN FOR SOUTHWEST RESEARCH  INSITIUTE
     Use Con-figuration 4 in Table 3c.  Design conditions are: Methanol




saturation exhaust flow rates are 132 Ibm/hr-cyl at 4500 rpm, gas pressure




drop = 2.5 psi.  Desire a set of hand calculations that outlines the design




analysis and the






DESIGN CONCEPT






     A diesel spray nozzle will spray methanol droplets on the inner surface




of the 3.353 in diameter evaporator tube.  Heat transfer from the hot exhaust




gases will vaporize the methanol.   We will assume that the spray nozzle will




fully wet the entire surface of the evaporator tube.   The evaporator tube will




be inclined at 45 degrees.






POSSIBLE EVAPORATOR DESIGN CONCEPTS






     There are two basic design concepts that may be  considered:




     1.   A shel1-and-tube boiler,  having gas passing  inside a bundle of small




     diameter horizontal tubes,  with nucleate boiling of methanol on the




     outside of the tubes.






     2.  Evaporation of a thin methanol  film that is gravity drained from the




     inside surface of a vertical  (or inclined)  tube.   Exhaust gas flows




     parallel  to the  outer  surface  of the tube.






     SwRI has specified the second  type  of evaporator.  There are two possible




methods  of  feeding  the mehtanol  to  the  evaporator:




     1.   Feed  the liquid at the  top  of  the evaporator  tube.






     2.   Spray  the  liquid,  as droplets,  over the entire length of the
                                          C-2

-------
     evaporator tube.




   SwRI has specified that the second  method  is  desired.   The  methanol will




enter at ambient temperature, approximately 90 F.  Hence,  it must be heated  to




its vaporization temperature  (259 F).








DESIGN CONSIDERATIONS






     Because the methanol will -form a  film on the inner tube surface, it may




ruptura upon heating.  Should film rupture occur, the dry  heat transfer




surface will be ineffective for evaporation.In order to assure surface




wetting, use of capillary grooves or a porous surface coating would help to




maintain surface wetting.  When all of the methanol is evaporated, the tube




wall will become dry.






     It is unlikely that a "threaded"  (axial or circumferential grooves)




surface would provide an increase of surface area.  It is  probable that the




tips of the v-threads would be unwetted, because of film rupture.   The contact




angle of the methanol should be small  (15-20 deg).  Hence, it is unlikely that




such grooves would provide any significant film surface area increase.   For an




axial grooved surface, the forces acting on the film are gravity and surface




tension forces (acting at the contact angle).






     It is possible to solve the momentum and energy equations for a smooth




inner tube surface and determine the liquid film thickness, assuming no film




rupture.   Since the film is laminar, the heat transfer coefficient is equal to




the liquid conductivity  (k)  divided by the film thickness  (B).   The problem is




complicated by two factors:   1.   The wall temperature varies along its length,




and 2.   Uniform spray distribution along the tube length is more complicated




than liquid supply at the top  of  the tube.






     If  an  axial  grooved  inner tube surface were used, the solution would be
                                       C-3

-------
 furthur  complicated.  However,  I think the problem  can  be  solved.






 PRELIMINARY DESIGN






      A hand calculation has been completed, which outlines  the  methodology  of




 the  evaporator design, assuming the methanol evaporation coe-fiicient  is




 known.   This calculation is attached.  It is noted  that the  thermal properties




 of methanol used are somewhat different than included in the SwRI  report.   My




 properties were taken from Perrys Chemical Engineer's Handbook, 6th ed.,




 page  3-204.






      The analysis assumes that axial threads are on the inner tube surface.




 Two  thread heights are considered - 0.015 and 0.030 in.   The analysis assumes




 that  the thread grooves are filled to their tips with methanol.  Thus, the




 film  thickness (B) is the groove depth.  It is assumed that the thread tips




 are  not  wetted.  The evaporation coefficient is calculated as h =  k/B.  It  is




 also  assumed that the active inner surface area is the inner tube




 circumference.  Methanol  has a quite low thermal conductivity,  approximately




 0.095 Btu/hr-ft -F.  The calculated evaporation coefficients are 33 and 74




 Btu/hr-ft -F for B = 0.030 and 0.015 in,  respectively.  Approximately 807. of




 the total thermal resistance is on the methanol side for B = 0.015 in.






     The analysis calculates the UA required to:  1.  Raise the methanol to




 saturation temperature,  and 2.  To evaporate the methanol.   I believe the




 analysis method is correct  for liquid supplied  at the top  of the tube.




 However, I am not so sure about the situation for a spray  of methanol over  the




 entire tube length.   For  this case,  liquid at 90 F is sprayed on the film




 surface.   The mechanism  by  which the methanol is raised  to saturation




 temperature and evaporated  is unclear.






     Using  a  methanol  latent heat  of 405  Btu/lb (rather  than 500 Btu/lb as




shown in  the  SwRI  report),  the calculated  tube  lengths are 2.92 and 1.63 ft





                                        C-4

-------
for 8 = 0.03  and  0.015,  respectively.   It  is  suggested  that  it  may  be




difficult for a single spray nozzle  to  uniformly  wet  the  entire  length  ?{  A




tube 2-3 ft long.   Hence,  the possibility  of  liquid  supply at the top of the




tube may be somewhat reasonable.






CONCLUSIONS






     The analysis has shown that the dominant thermal resistance is on  the




methanol side.  For the  chosen design concept, the required  heat-exchanger




length will decrease as  the liquid film thickness is  made smaller.  A somewhat




complex analysis  would be  required to estimate the film thickness for a tube




having axial  thread grooves; however, it is probable  that such and analysis




can be done.  Consideration should be given to final  selection of the method




of liquid supply  to the  inner tube surface, e.g., whether spray distribution




over the entire tube surface can be  achieved.






RECOMMENDATIONS






     It is recommended that some furthur analysis be  performed to estimate  the




probable methanol liquid film thickness, and the distribution of the heat




transfer coefficient over  the tube length.






     The simplest concept  for analysis would be a smooth inner tube surface.




It is possible that this analysis can be extended to  treat the axial thread




groove geometry.






     Should SwRI desire the Consultant to undertake the above analysis,  I




would estimate that 5-8 days would be required for the plain  tube case.   An




additional  10  days may be required to extend the analysis to  the axial  thread




geometry.   The Consultant would  hire the services of  a Graduate Student  to  do




detailed  calculations.   The cost  of the Graduate Student would be in the $800-




1,000 range.   Some computer time  would be involved,  perhaps '1300.  Actually,
                                          C-5

-------
this problem would be a good one for an M.S.  Thesis.   This option may be of


interest,  if your  time schedule  permits.   An  alternate,  or parallel  approach


would involve experimental  investigation  to determine  the methanol  film


coefficient.
                              Submitted  by
                                             7
                                           Ralph  L.  Webb


                                           July  17,  1985
                                         C-6

-------
                        DR. RALPH L. WEBB
                      512 W. FOSTER AVENUE
                     STATE COLLEGE, PA 16801

                                                   August 9, 1985

Mr. William E. Likos
Fuels and Lubricants Research Division
Southwest Research Institute
Post Office Drawer 28510
San Antonio, TX 78284

Dear Bill:

    This letter is in response to our telephone conversation on
about July 31.  You requested that I provide additional guidance
on how one may predict the methanol film film thickness in the
evaporator, and thereby calculate the methanol heat transfer
coefficient.  I have spent approximately 12 hours on this
question, and believe that I know how to perform the necessary
calculations.  However, it would be necessary to derive the
applicable differential equations and perform a numerical
solution.  The problem of thin film evaporation of a uniform
spray onto a grooved tube is not one, whose solution can be
looked up in a book.  Although I believe it should be possible to
analytically solve this problem, some time would be necessary to
do the job.

     The purpose of this letter is to describe the fluid flow and
heat transfer phenomena in the evaporator tube, and to describe
how one may proceed to solve the problem.

     If the methanol were added at the top end of a plain
evaporator tube, the solution can be taken from Nusselts theory
of film condensation.  However, the evaporator design you
envision introduces three additional complications.  These are:

     1.  Introduction of the methanol via a uniform spray over the
         full length of the tube.
     2.  Introduction of a subcooled, rather than saturated fluid.
     3.  The use of axial grooves on the inner surface of the tube,

     I will comment on the complications introduced by each of
of the above factors, and indicate how one may approach the
problem analysis.

     Consider, first, a vertical, plain evaporator tube.  If
saturated methanol were introduced at the top of the tube,
Nusselt's theory of film condensation is directly applicable to
the "reverse" problem of film evaporation.  If the entering fluid
                                   c-7

-------
is subcooled, it is a fairly straight forward matter to calculate
the tube length required to raise the subcooled fluid to its
saturation temperature.

     Introduction of the methanol by a uniform spray over the
full length of the tube substantially changes the velocity and
thickness of the liquid film.  Attachment A outlines how one may
proceed to derive and solve the differential equations for the
film thickness.  I think that the resulting film thickness will
be smaller than that of a film introduced at the top of the
tube.  Contrary to the Nusselt theory, it appears that the
differential equations must be numerically solved.

     The use of an axially grooved evaporator tube introduces
surface tension forces that act to slow the drainage of the
film.  Attachment B shows how one would calculate the static
height of the liquid film inside a grooved evaporator tube.  Note
that there is no flow of the film for this analysis.  However,
the same surface tension forces would exist for the dynamic
condition, when the film is gravity drained from the tube.
Attachment C shows how one would approach this problem for the
liquid introduced at the top of the tube.  As shown by the
analysis, it is necessary to write a force balance, which
accounts for the pressure, gravity and surface tension forces on
the liquid film.

     Now, if one combines all of the above three special features
(uniform spray over the tube length, grooved tube and entering
subcooled liquid) the differential equations become more
complex.  However, I believe it is possible to formulate the
necessary momentum and energy balances, which define the
differential equations that must be solved.  Another factor
involved, that I have not discussed, is the inclination angle of
the tube.  However, one may easily account for this by including
the component of the gravity force in the flow direction.

     I think that the Nusselt solution for a plain tube, with all
liquid introduced at the top would yield a conservative answer,
relative to that for the uniform liquid spray. Furthur, I feel
that the heat transfer coefficient for a plain tube would be
higher than for a grooved tube.  The main function of the grooved
tube is to prevent surface dewetting, which is quite likely to
occur for a plain tube.  Attachment D contains calculations for
the evaporation coefficient in a plain tube inclined at 45
degrees, having all liquid introduced at-the top of the tube.
Introduction of the liquid by a uniform spray over the full tube
length will probably result in a higher evaporation coefficient,
and the use of axial grooves will probably act to decrease the
coefficient.
                                c-8

-------
     The above information is all I can accomplish in "eight hour
analysis," which you specified.  In regard to future evaluations
for the evaporation coefficient, I will suggest two possibilities
for your consideration:

     1.  Forget any furthur analysis, and experimentally determine
         the evaporator performance.
     2.  Attempt to solve the governing differential equations of
         the problem.

     I would be pleased to attack either of the above approaches
for you.  Such work would probably best be performed under a
research subcontract with Penn State.  I believe it should be
possible to set up an experimental apparatus to measure the
evaporation coefficient.  The analytical approach would require a
matter of weeks to perform.

     I hope this information is of value in identifying the
probable value of the evaporation coefficient, and approaches to
its more accurate determination.

                         With my best regards.
                         Ralph L. Webb
                                  c-9

-------
           APPENDIX D






HEAT EXCHANGER DESIGN ANALYSES




       PERFORMED BY SWRI

-------
        HEAT EXCHANGER DESIGN FOR DISSOCIATING METHANOL
                           WITH EXHAUST HEAT

1.0  INTRODUCTION

     This report covers  the  design specifications of the individual cylinder heat
exchangers for vaporizing  and dissociating  methanol  using exhaust  heat.   The
following paragraphs discuss the specifics of the design analysis, the assumptions
and methodologies used,  and the results and recommendations.  Included are a
sketch of the design, and  its details.

2.0  ASSUMPTIONS AND LIMITS (Used in the calculations)

     2.1  Catalyst

          2.1.1   Recommendations by Johnson Matthey (3  M)

                  Catalyst formulation  and required  area  were discussed with
several people at 3 M. The required  area to dissociate a pound of methanol per
hour varied  from 0.165 to 0.98  square feet.  The catalyst material for these
values is copper oxide and zinc oxide.

                  Catalysts  containing  copper  were  not recommended  due to
poor high  temperature resistance.  3  M  recommended catalysts using platinum,
chromium, or palladium.  Yields equal to or greater  than the reported values were
predicted by 3 M.

          2.1.2   Catalyst Material (Dissociation Side)

                  No  specific catalyst  material was  used in  the calculations.
Plots of reactor length versus catalyst area were made to finalize the reactor size
once a value of yield was  selected.

          2.1.3   Catalyst Material (Exhaust Side)

                  None was assumed. However, addition  of an oxidizing catalyst
on the exhaust  side would increase the total heat flux to the  methanol side  and
improve the heat exchanger effectiveness.

          2.1.*   Heat Exchanger Surface Area Required for Catalyst

                  This is considered  as  an  independent variable in the reactor
design  and is discussed in  the analysis section.

          2.1.5   Maximum Allowable Catalyst Temperature

                  No  means  are provided  to protect  the  catalyst  from  the
maximum exhaust temperature.

          2.1.6   Minimum Catalyst Temperature to Achieve Significant Yields

                  These  are summarized in Table 3.
                                         D-2

-------
     2.2   Reactor

           2.2.1   Coefficients of Thermal Conductivity

                  17-7-PH Stainless steel  = 22.0 BTU/(hr)(ft2)(op/ft)

           2.2.2   Specific Heats (BTU/lbm - OR)

                  Air =  0.24
                  Exhaust =  0.27
                  Methanol (vapor)  = 0.41

           2.2.3   Heat of Vaporization of Methanol

                  500 BTU/lbm

           2.2.4   Heat of Dissociation of Methanol

                  1500 BTU/lbm

           2.2.5   Structural Descriptions of 17-7-PH Stainless Steel

                  Ultimate Strength - 87 ksi
                  Yield Strength  -  39 ksi
                  Endurance Limit  -  31 ksi

           2.2.6   Minimum Fin Spacing

                  The fins on both the exhaust  and methanol sides can be no
closer  together than 0.05 inches (20 fins per inch) to assure that the wash coat
follows the fin contours and does not bridge them.

           2.2.7   Pressure Drop

                  The maximum allowable pressure  drop on the exhaust side of
the reactor and evaporator is 2.5 psig.

           2.2.8   Methanol and Exhaust Flowrate Limits (Ibm/hr - cylinder)

                  Extrapolated from EPA NAPS-Z Engine Data

                	Engine Condition	        Methanol       Exhaust
                Idle (500 rpm,*= 0.7)                      0.7            8.0
                Full Power (6000 rpm & 100% load,
               $ =0.93                                  24.5          190.0

3.0  ANALYSIS

     3.1    Required Reactor Pressures and Volume

     (1)    Minimum Desirable Valve Size  - The minimum preferable size for the
           gas inlet valve should be such that its mach index number does not
                                         D-3

-------
          exceed 0.6.  Using the following relationship (Reference  1) this valve
          size was determined:

                          z =      Ap x S	
                                 Ai x Ci x a

where:         z  =      Valve mach  index number = 0.6 maximum desirable
              Ap  =      Piston area  = 8.8 in2 (NAPS-Z engine)
              Ai  =      Inlet valve area, which calculates to 0.44 in2
               S  =      Mean piston speed = 1553 ft/min  for NAPS-Z engine
                         at 5500 rpm
               a  =      Sonic  velocity  of  inlet  gas =  1318  ft/sec  under
                         expected inlet conditions
              Ci  =      Mean valve  flow  coefficient =  0.65  typically for a
                         well-designed poppet valve with  its lift equivalent to
                         25% of its diameter

          The Ai  =  0.44 in2 yields a minimum desirable valve diameter of 0.75
          inches.

     (2)  Maximum  Allowable Valve  Size - The NAPS-Z engine existing spark
          plug holes are approximately 0.5  inches in diameter.  The sectioned
          engine head  reveals that  one of the spark plug holes in each  cylinder
          could be enlarged to 0.625 inch in diameter without compromising the
          structural  integrity of the head.

     (3)  Critical  Flow Through  the  Maximum  Allowable  Valve Size -  The
          0.625-inch diameter  valve  is smaller  than  the 0.75  inch minimum
          desirable size. It will yield a higher mach index at maximum inlet gas
          flowrates.  The calculated mach indexes at 100% engine load are:

                         Engine Speed             Mach
                            (rpm)                 Index
                             5500                  0.85
                             4500                  0.70

     (4)  Valve Timing -  The opening and closing locations  of the methanol
          inlet valve are limited by the following engine parameter values as a
          function of crank angle:

          •   Maximum spark advance
          •   Induction air intake valve closing
          •   Cylinder compression pressure

          The maximum spark advanced shown in the Reference 3 test data was
          45° BTDC.  The  point of  induction air  inlet valve closing  was not
          documented  but  is assumed at  a common value of  45°  ABDC.   The
          cylinder compression of induction air was assumed to be isentropic and
          raises the  cylinder pressure to 72 psia with the crank at 45° BTDC.
          Given these constraints the valve  opening was made at 45° ABDC and
          its closing at or before 45°  BTDC.  This allowed 90° of crank angle for
          the introduction  of  dissociated  products  into  the  cylinder.   The
          corresponding time  at this crank angle  for several engine  speeds  is
          summarized below:
                                         D-4

-------
                                       Maximum                 Effective*
           Engine Speed             Valve Opening            Valve Opening
              (rpm)                    Time (ms)                 Time (ms)
               500                       30.0                      15.0
               1500                      10.0                      5.0
               4500                       3.3                       1.7
               6000                       2.5                       1.3
* 50% of maximum to account for reduced valve area during opening and closing.

      (5)   Required Reactor Pressures to Provide Adequate Product Flowrates

           The reactor pressure must be maintained between 100 to  150 psig so
           that the required product mass can be delivered to the cylinder during
           the brief valve  opening  time.   Flowrates  were  calculated  using
           (Reference 1):
                              m = AZ x aoi  x  poix $1

      where:   m  =       Mass flowrate of product
              A2 =       Valve port nominal area
              a01 =       Sonic velocity at valve opening
              P01 =       Density of dissociated products in reactor
               $1 =       Compressible flow function based on empirical data
                          (Reference 1)

           A summary of product flows per cylinder per cycle at 100 and 150 psig
           reactor pressures  compared  to  the  required  amount of  product to
           maintain 100% power is summarized as follows:

                              Amount of Products
                             Entering per Cylinder        Amount of Products
           Engine Speed      	(Ibs x 10~3)	      Required per  Cylinder
              (rpm)         150 psig       100 psig          (Ibs x 1Q~3)
               500             4.3             2.6
               1500            1.5             0.9                0.14
               4500            0.5             0.3                0.14
               6000            0.4             0.2                0.14

           Mass flowrates in each  case are below calculated critical values.

      (6)   Minimum Reactor Volume to Maintain Required Pressures

           Assuming  the isothermal expansion (worst case) of the  reactor  gases
           through the  opened valve,   a  minimum  of 6  inches^ reactor plus
           evaporator volume is required to maintain 100 psig minimum  pressure
           when the starting pressure is 150 psig and maximum product  delivery
           is required.

      3.2   Heat Balance

           In  the  proposed reactor  system liquid  methanol is  injected into an
evaporator, where it vaporizes. The saturated methanol vapor is  then superheated
in the evaporator and enters the reactor where it dissociates into CO and  \\2- The
exhaust gas from the engine provides the heat for all requirements:  vaporization,
superheat,  and dissociation.  The  exhaust gas flows counter to the direction of
                                       D-5

-------
methanol flow, so heat is first transferred from the exhaust gas to the  dissocia-
tion process and last to the vaporization process.

           A primary question is what percentage of the  total engine fuel flow
can be processed in  the manner described above, given the quantity and quality of
the heat available in the exhaust.  This discussion will concentrate on the engine
operating conditions of 4500 rpm, 100% load and 500 rpm, 30% load.

           Parameters of importance in the analysis were:

           Heat of  vaporization of methanol                        500 BTU/lb
           Heat of  dissociation of methanol                        1500 BTU/lb
           Specific heat of liquid methanol                      0.58 BTU/lb-op
           Specific heat of vaporized methanol                  0.44 BTU/lb-OF
           Exhaust  flow rate at 4500 rpm; 100% load                132  Ib/hr-cyl
           Fuel flow rate at 4500 rpm; 100% load                  20.5  Ib/hr-cyl
           Exhaust  flow rate at 1500 rpm; 30% load                  24  Ib/hr-cyl
           Fuel flow rate at 1500 rpm; 30% load                    2.6  Ib/hr-cyl
           Minimum temperature for the dissociation  reaction             400°F

           The exhaust temperatures measured by EPA at 4500 rpm, full load,
and 1500 rpm, 30% load were  1300°F and  851°F  respectively.  Our calculations
show  that, by insulating  the exhaust port,  these temperatures  can be raised to
1400°F and 911°F,  respectively.  In our analysis, we considered the effect of the
standard port (as used by  the EPA) and  an insulated port (giving the higher exhaust
temperatures). Insulation would be  accomplished, for instance, by a ceramic port
insert.

           In the proposed system,  the pressure within the evaporator and reactor
is uniform, and  this  pressure is  defined solely  by the  temperature of  the
evaporator, since  in the  evaporator  the  methanol  liquid  and vapor  exist in
equilibrium. A few  equilibrium states are:

           Pressure (psia)              Temperature (°F)
               150                          289
               100                          259
                29                           180
               14.7                          149

Thus, if the evaporator temperature can be maintained at 289°F, the fuel pressure
to the engine will be 150 psia.  Since it is necessary to inject the fuel gases during
the compression stroke, and since significant fuel pressure is required to do this in
the time available,  a high evaporator temperature  (producing a high fuel pressure)
is a necessity.  A heat balance  can  determine  the temperature of the evaporator,
and hence  the  pressure,  at least approximately.   A first approximation can be
readily obtained  by assuming an  infinite heat  exchanger  area,  so  that  the
complexities of varying  heat transfer coefficients  and fin  efficiencies can  be
ignored.  A more  accurate approximation takes into account  these  factors.  Both
approximations will be made in  this  summary.

      The quantity  of heat required for processing the methanol was calculated
and the effect upon the exhaust gas temperature was determined.   For different
temperature differences  between methanol and exhaust gas,  the maximum flow
                                      D-6

-------
rate of methanol was calculated.  Figure  1 shows  the  percent of the total  fuel
flow needed to run the engine at 4500 rpm,  fuel load that results in  complete
vaporization and dissociation by the available exhaust heat.  Two cases are shown.
The first is for an insulated exhaust port and an infinite heat exchanger where the
exhaust gas temperature is equal to the  liquid methanol  temperature at the
entrance to the vaporizer.   The second is for a 100°F difference  between the
exhaust gas and the liquid methanol at the vaporizer entrance, which  implies a
small heat exchanger.   A standard exhaust port  is  used.  Thus, the upper curve
shows  the maximum amount of methanol that can  be dissociated, and the lower
curve provides a conservative estimate.

           Figure  2 shows similar results for the engine  condition of 1500 rpm,
30% load, and   = 0.7.

           In  either case, running the fuel pressure at 165  psia (as  opposed to a
lower pressure) has a relatively small  penalty on the amount of fuel that can be
processed through  the  reactor  and  evaporator.  Note that increasing  the  fuel
throughput decreases the fuel pressure, but even  when the pressure is reduced to
very low values, the fuel throughput  has not reached 100%.  This is because there
is insufficient  heat in the exhaust for dissociation  and vaporization of 100% of the
methanol even at reduced temperatures corresponding to low pressures.

           Insulating the exhaust port will permit raising the fuel throughput by
about 7%  at 1500 rpm, 30%  load, and by about 8% at 4500 rpm, fuel load, for the
same evaporator design.

     3.3   Heat Exchanger Design For Heat Transfer (Step 1)

           3.3.1   Reactor

                  (1)     Reactor Design - The reactor was designed to use all
                         available exhaust heat from the engine down  to 432°F
                         (432°F  exhaust  gas  temperature  was calculated to
                         yield a 400°F  wall temperature).   This  temperature
                         was chosen as  the  lowest point that significant  yields
                         of  dissociated  methanol  could be produced from  the
                         best  catalysts.  The  cross-sectional  geometry  was
                         chosen  based  on  a   trade-off of  outside  diameter
                         (limited  by   engine  cylinder-to-cylinder  spacing),
                         number of  annulli  (a  low number desired to reduce
                         manifold    complexity),   available   metal   catalyst
                         substitute  material (from Behr)  that can be coated
                         (maximum  of 20 fins  per inch according to Johnson-
                         Matthey).

                  (2)     Reactor Heat  Transfer  - The length of the reactor
                         was  determined  from   the   chosen  cross-sectional
                         geometry  and  the  calculated  required heat  transfer
                         area  necessary to  cool  the  exhaust gases to 432°F.
                         The  maximum  mass   flow   rate   condition   was
                         determined to  be the  limiting case for heat transfer
                         area.  The  NTU effectiveness method, as outlined in
                         Reference  2, was used to perform  the  heat transfer
                         calculations.   The Reynolds number  was calculated
                                        D-7

-------
                         from the cross-sectional geometry  and exhaust mass
                         flow rate.  The heat transfer parameter and friction
                         factor were selected from plots for each configuration
                         as a function  of  Reynolds number.  Fin effectiveness
                         was  calculated  from  the  fin  geometry.    Heat
                         exchanger exhaust  side effectiveness  was chosen and
                         along  with the fluid  outlet temperature differences
                         used to  calculate   the  temperature  change  in the
                         exhaust  gas.   This temperature  difference and the
                         capacity  rate ratio  were used to  calculate  the  NTU.
                         The NTU was then used to determine the required heat
                         transfer  area  and,  for  a   given   cross-sectional
                         geometry, the reactor length.

                  (3)     Pressure  Drop - Pressure  drop through the reactor
                         exhaust  side  was calculated for  each configuration
                         using the formulas in Reference  2.

                  (4)     Summary of Reactor Designs - Figure 3 and Table 1
                         summarizes the various  reactor geometries  and their
                         properties.

           3.3.2   Evaporator

                  (1)     Evaporator Design  - The  analysis of the evaporator
                         was similar to that of the reactor in that it considered
                         the  temperature of the exhaust  leaving the reactor
                         and  the  minimum  temperature   (267°F   exhaust
                         temperature was calculated to  yield 259°F wall tem-
                         perature) at which methanol would vaporize under an
                         elevated reactor pressure (150 psig).

                  (2)     Evaporator Heat Transfer  -  The same procedures dis-
                         cussed  for  the  reactor  were used  to calculate the
                         evaporator required heat transfer area.

                  (3)     Evaporator Pressure Drop  - The pressure drop on the
                         exhaust side was calculated using the same methods as
                         that used for the reactor.

                  (4)     Summary of Evaporator Designs - Figure 3 and Table
                         2 summarize the various evaporator configurations and
                         their properties.

     3.*   Heat Exchanger Heat Transfer Optimization (Step 2)

           Design  of the reactor/evaporator  using  the  432°F  exhaust  outlet
temperature from  the  reactor resulted in a reactor that  could dissociate  more
methanol  than  the  evaporator could vaporize (on  a  heat transfer basis).   To
correct this a balanced  design was  sought by setting the yields of the evaporator
and  reactor equal.   The  parameters of one reactor configuration  and  one
                                         D-8

-------
evaporator configuration are summarized below:

               Product Yield  Catalyst   O.D.   Length   A P
 Component      (Ibm/hr)     Area (ft2)   (in)    (in)    (psig)   Effectiveness

Reactor      #6     17.9          2.57      4.0     1.8     0.15        0.85
Evaporator   #4     17.9                   4.0     4.8     0.37        0.97

These values are based on a reactor inlet temperature of 1400°F, a reactor outlet
and evaporator inlet temperature of 550°F and an evaporator outlet  temperature
of 267°F.  The 1400°F exhaust temperature  would  be possible by insulating the
exhaust port as discussed in paragraph 3.1.

     3.5   Heat Exchanger Design for Catalyst Area (Step 3)

           (1)     Catalyst Area -  The available catalyst area for each reactor
                  configuration  was calculated based on the heat exchanger area
                  and  compared to the area requirements  of  Johnson Matthey.
                  With  all reactor designs, the  required catalyst  area (per
                  Johnson Matthey) is greater  than the  required heat transfer
                  area.  The length of any design,  therefore, must be increased
                  to achieve  the necessary catalyst area.  This  is shown  on
                  Figure 4 which plots % dissociation versus  reactor  length for
                  both Johnson  Matthey and  Wei and Chen yield data. If  a more
                  appropriate catalyst  surface  yield value  than  that recom-
                  mended  by Johnson Matthey or that demonstrated by Wei and
                  Chen is  deemed appropriate,  then the required reactor length
                  can be determined.

           (2)     Discussion of  Catalysts and Surface Yields - A  summary  of
                  catalyst properties is shown  in  Table 3.   A  review  of  the
                  literature  suggests  that  higher  surface yields  than  those
                  achieved by  Johnson Matthey are possible  (Wei  and Chen
                  data).   If so,  the length  of the  reactor  could be reduced  as
                  discussed previously.  In any case, however, the reactor would
                  probably contain greater surface  area  than that  required for
                  heat transfer.

           (3)     Effect  of  Catalyst  Area  Requirements  on Pressure Drop -
                  Increasing  the reactor length to  provide  adequate catalyst
                  area will result in an increase in pressure drop. This is shown
                  in Figure 5, which plots catalyst surface  area versus pressure
                  drop for the various reactor configurations.  The  point repre-
                  senting the minimum  heat transfer  area is shown  for each
                  configuration.

     3.6   Fabrication and Assembly Details

           (1)     Separate reactor and evaporators are used for several reasons:

                  •       Design of the reactor  and evaporator can be optimized
                          independently.  This is particularly important for the
                          evaporator where the  uniformity of spray impingement
                                        D-9

-------
                          necessitates  a   different   configuration   for  the
                          methanol side from that used on the reactor.

                  •       Vaporization and  dissociation function separation will
                          minimize the possibility of reactor flooding.

                  •       Reactor and evaporator performance can be indepen-
                          dently measured as can the  state of the fluid in each
                          during testing.

           (2)     The reactor tubes and other structural parts are made of 0.030
                  inch or  thicker material  and the  fins of  0.00*  inch thick
                  material. Type 17-7 PH stainless steel or an equivalent will be
                  used for  this application.   This  material and  these   tubing
                  thicknesses will yield maximum tubing stresses of less than 15
                  ksi.  This type of  stainless  steel permits welding, maintains
                  high strength at high  temperatures and contains aluminum
                  which will provide better adherence of the alumina-based wash
                  coat.   Pieces will be  furnace brazed with a high temperature
                  brazing filler metal such as BNiCr and/or TIG welded.

           (3)     Figure 3a shows the reactor and evaporator mounted in-line in
                  one section of exhaust pipe. It may be necessary to mount the
                  reactor above  the  evaporator and connect them  with  a "U"
                  section of pipe to reduce the overall length of the assembly.
                  Such an arrangement would not change the basic design  or the
                  operating characteristics of the system shown in the figure.

     3.7   Transient Conditions

           Changes in engine fuel requirements will necessitate a change  in the
flow rate of dissociated products into  the engine.  Under steady-state conditions
dissociated product generation and mass flowrates will be a function of exhaust
temperatures and mass flowrates.  Because the reactor cannot supply 100%  of the
engine needs, direct  methanol  injection  into the intake  manifold will be used
during steady-state conditions to supplement the reactor output.  Mass flow rates
of direct methanol injection  will be  a  function of  reactor  output under these
conditions.  Under conditions where increased power is  required, direct methanol
injection will have to make up the difference in fuel demand during the period of
the transient and  until the reactor/evaporator  reaches steady-state  production.
Lag in the output of the reactor is a function of the total internal volume and the
mass and thermal conductivity of the heat transfer surfaces, and  will need to be
determined experimentally.  Under transient conditions,  the fuel control  system
using  inputs  from the engine, throttle  position  and  reactor temperatures  and
pressure  will be designed  to meet engine fuel  needs by correctly  proportioning
fuel delivery to both the intake manifold and the evaporator.

4.0  RECOMMENDATIONS

     The final size of the  reactor is determined by the catalyst yield predictions.
SwRI has the most confidence in the yield data provided by 3-M which was derived
from a coated tube configuration.  The combination of the reactor configuration
number 6 and evaporator  number 4,  appears to be  a reasonable design.  This
configuration, sized according  to the catalyst  activity  of  the J-M  tube data,
                                      D-10

-------
results in a reactor length of 13.6 inches and  an evaporator length of 4.8 inches.
The total pressure drop would be 1.5 psid.

     J M would  provide an  alumina washcoat to both sides of the reactor.  The
catalyst  on each  side would  be dependent upon the final recommendation of 3 M
or EPA.

     This design provides  a  factor of safety of 7.6 for the heat transfer into the
reactor.   No safety factors exist for the catalyst yield predictions  and  the
evaporator design.  These safety  factors could be provided at the expense  of a
higher pressure drop in the exhaust system.
                                       D-ll

-------
                             REFERENCES
1.    Taylor, C.F., "Internal Combustion Engine in Theory and Practice," Volumes
     1 and 2, 1980.

2.    Kays, W.M., and A. L. London, "Compact Heat Exchangers," 2nd Edition.

3.    EPA  test  data on  380  NAPS-ZMD engine - test numbers  HD812481  and
     HD812489
                                   D-12

-------
  120
EVAPORATOR TEMP (Tooil) ,.°F
        149 180
              259
289
Q 100
w
p
e
   80
   60
          EVAPORATOR AREA INFINITE, (Tgas - Tboil = 0)
                  (INSULATED EXH. PORT)
           LIMITED EVAPORATOR AREA
           (Tgas - Tboil = 100°F),(STANDARD EXH. PORT)
                              I
                 50          100          150

                         FUEL PRESSURE - PSIA
                                      200
           FIGURE 1 - FUEL THROUGHPUT AT 4500 RPM,
                        FULL LOAD, 0 = 1.0
                                   D-13

-------
   EVAPORATOR TEMP (Tboil) ..°F

149  180             259
                                            289
O 80
hJ
W
P
      \
H
8
   60
   40-
          fAPORATOR AREA INFINITE
             - Tboil = 0), (INSULATED EXH. PORT)
           EVAPORATOR AREA LIMITED, (Tgas - Tboil = 100°F)
                          (STANDARD EXH. PORT)
                50           100          150

                      FUEL PRESSURE - PSIA
                                             200
            FIGURE 2 - FUEL THROUGHPUT AT 1500 RPM,
                         30% LOAD,  = 0.7
                                D-14

-------
    INTAKE VALVE
EXHAUST VALVE
                                                                               DISSOCIATED   A
                                                                              PRODUCTS OUT  \
                                                         DISSOCIATED CH3OH
                                                        PRODUCTS TO ENGINE
                                                                            REACTOR    SECTION R2-R2

                                                                            R1
                      WATER
                      JACKET
NAPS-Z ENGINE CYLINDER
        HEAD
                          COMBUSTION CHAMBER
                                                                                       EVAPORATOR
                      METHANOL
                                                        METHANOL SIDE
                                                     FIN CONFIGURATION TO
                                                       BE FINALIZED LATER
                                                                                                  LIQUID CH3OH
                                                                                                  FROM PUMP
      SECTION R1-R1                     SECTION E-E
       NOTE: DIMENSIONS DEFINED AND
             DISCUSSED ON FOLLOWING SHEETS
                     FIGURE 3A -  REACTOR ASSEMBLY FOR EVAPORIZATION AND DISSOCIATION OF METHANOL

-------
                                           Diameters (inches)
                                                                            Dr
Configuration   Type    PID2P^p^'p^     pft -    D7'f    pR
      1          A    1.000  1.100  1.140   1.336
      2          B    0.800  0.996  1.036   1.436   1.476  1.672
      3          B    0.800  0.996  1.036   1.436   1.476  1.672  1.712   2.112   2.152
      4          B    0.800  1.300  1.340   1.740   1.780  2.280
                                                                                  Fin Description
5

6
A

B

B
                      3.544  3.744  3.804  4.000

                      3.088  3.283  3.344  3.744   3.804  4.000
                             --  Strip fin Figure 10-56 Kayes
                                 and London, 3rd Edition,
                             --  .098" height, 20 per inch,
                                 0.125 inch length in direc-
                          2.348  tion of flow
                             --  11 plain fin/inch: Figure
                                 10-26 Kayes and London,
                                 2nd Edition
                             --  Figure  10-56 Kayes and
                                 London, 3rd Edition
                             --  Figure  10-56, Kayes and
                                 London, 3rd Edition
                2.372  2.568   2.628  3.028  3.088  3.284  3.344   3.744  3.804  4.000  Figure 10-56, Kayes and
                                                                                        London, 3rd Edition
                 Type A
                                             TypeB
   Dimension
       Dl
       D2
       D3
       D4
          Description
     Inner closed cylinder
     Reaction side annulus
     Wall
     Fin - exhaust annulus
                                                Dimension
                                                    Dl
                                                    D2
                                                    D3
                                                    D4
      Description
        FfN
      8EHR SUBSTRATE
                                              D9
                                              DID
Inner closed cylinder
Fin
Wall
Reaction side annulus
Wall
Fin
Wall
Reaction side annulus
Wall
Fin
            FIGURE 38.  CROSSECTIONAL GEOMETRIC DATA OF REACTOR CONFIGURATIONS

-------
Configuration
     1
     2
     3
     4
 Type
   A
   A
   A
   A
Dimension
   Dl
   D2
   D3
2.064
1.858
3.744
3.538
D?
2.1040
1.898
3.804
3.598

D3
2.300
2.300
4.000
4.000
Description
Fin Description
Figure 10-56 Kayes and London, 3rd Edition
Figure 10-60 Kayes and London, 2nd Edition
Figure 10-56 Kayes and London, 3rd Edition
Figure 10-60 Kayes and London 2nd Edition

                                         Inside diameter of evaporation chamber
                                         Wall
                                         Fin
              FIGURE 3C.  CROSSECTIONAL GEOMETRIC DATA OF EVAPORATION CONFIGURATION

-------
                                                  MAXIMUM POSSIBLE
                                                  DUE TO AVAILABLE
                                                  HEAT ABOVE 259 °F
JM SURFACE YIELD
  0
   25
    50          75

% DISSOCIATED
FIGURE 4 - REACTOR LENGTH (4 IN. DIA.) vs % DISSOCIATION

                AT 4500 RPM, * = % LOAD
                      Texh =1400°F
                           D-18

-------
               EXCESS HEAT TRANSFER AREA BEYOND
                     DESIGN POINT, TYPICAL
                                REACTOR NUMBER - SEE FIGURE 3
           5          10          15          20

            CATALYST SURFACE AREA (FT2)
FIGURE 5. PRESSURE DROP VS. CATALYST SURF ACT AREA
                          D-19

-------
                                                  Table 1.  Reactor Configurations
N3
o

Configuration
1
2
3
4
5
6
7

Exhaust
Annulli
1
2
3
2
1
2
3

Reaction
Annulli
1
1
2
1
1
1
2

Length
(in)
8.90
6.09
4.262
23.70
5.442
3.36
1.68
Outer
Diameter
(in)
1.336
1.672
2.348
2.28
4.000
4.000
4.000
Pressure
Loss
(psid)
14.9
2.6
0.632
0.220
0.950
0.277
0.040
Volume
Catalyst
(in*)
0.716
4.70
8.43
22.9
5.44
7.48
6.75
Catalyst
Area
(ft2)
0.45
3.0
5.35
14.5
3.45
4.79
4.28
                                                 Table 2.  Evaporator Configurations

Configuration
1
2
3
4

Exhaust
Annulli
1
1
1
1

Evaporation
Annulli
1
1
1
1

Length
,0n)
3.810
6.200
4.078
4.085
Outer
Diameter
(in)
2.300
2.300
4.000
4.000

Pressure Loss
(psid)
2.93
1.15
0.816
0.316

-------
                                                 Table 3 - Summary of Catalyst Properties
a
i
(S3
                Catalyst
JM
CuO ZnO pellets           698
JM
1/4" tube coated with
CuO ZnO                   698
Co Sch of Mines
Pd/La pellets               572
       Surface
        Yield
       lb/hr-ft2
                                            0.165
   Surface Area
Required to Produce
 19.1 Ib/hr CO + 2H
                              116
                                             0.98
                             19.5
                                             0.15
                               119
                                                                                           Comments
                          o Laboratory pellet data is similar in yield
                            rates
                          o Space velocities (    1/hr) are lower than
                            what has been used in automotive reactors
                          Based on an actual coated tube, which we
                          can duplicate. JM suggest that it may have
                          been heat limited, implying that higher yields
                          may be possible by improving the heat flux.
                          Higher temp, resistance of Cu is poor, other
                          catalyst materials (Pt, Cr2O3, Pd)  can be
                          used with similar or improved yields.
         Conoco

         Wei and Chen
         Methane formation from
         CO + H2
         Raney  Nickel
         Mass diffusion of CO
         into methanol limit
         on rate
560
           4.2
         4.5
           .16
         119
                          Awaiting report from EPA
                                                                           An example of possible rates for the decom-
                                                                           position reaction if the heat and mass trans-
                                                                           fer limits are higher than those obtained by
                                                                           pellet-filled tubes
                                                                           Similarity between mass diffusion rate and
                                                                           pellet data rates in which the velocities were
                                                                           low suggest that diffusion may be limiting

-------
       Addendum to Appendix D
Letter  to  EPA  Containing  Follow-up
Discussions  and Clarifications/Revisions
to Heat Exhanger Design Analysis
                    D-22

-------
SOUTHWEST  RESEARCH  INSTITUTE

POST OFFICE DRAWER 28510 • 6220 CULEBRA ROAD • SAN ANTONIO. TEXAS, USA 78284

ENGINES, EMISSIONS AND VEHICLE RESEARCH DIVISION
                                      June 25,  1985
 Mr. Carl Hellman
 United States Environmental Protection Agency
 Motor Vehicle Testing Laboratory
 2565 Plymouth Road - ECTD
 Ann Arbor, MI 48103

 Subject:   Follow-up Information on Monthly Progress Report No. 8, Dated May
           31, 1985

 Dear Mr. Hellman:

      The  following paragraphs contain our responses to the questions you  raised
 during  our June  17  telephone call.  Where applicable, page numbers refer to the
 sequential order  of pages in  the technical  discussion  attached to our Monthly
 Progress Report  No. 8 (MPR //8).

 Parameter Values on Page 5

      The values for fuel and exhaust mass flowrates used in paragraph 3.2 were
 those assigned originally to this analysis. The values used in the  final calculations
 (as summarized  in  MPR  #8) are  shown below along with the appropriate values
 from the EPA test results (Test No. HD-812489) on the NAPS-Z engine.

                       Mass Flow        SwRI           EPA
                         Rate        Value Used     Test Data
  Engine Condition     Parameter     (Ibm/hr-cyl)   (Ibm/hr-cyl)   % Difference
 4500 rpm, 100% load      Fuel           17.8           17.7           0.6
 4500 rpm, 100% load   Exhaust         142.8          141.3          1.1
 1500 rpm, 30% load        Fuel            2.6           2.6            0.0
 1500 rpm, 30% load     Exhaust         25.8           26.7           3.4

 Most of the  differences in values used versus the  values from the EPA data are
 small enough not to  appreciably affect the results of the heat balance analysis.

 Values of Dissociated Products Mass Flow Versus Choking Flow on Page 4

      Table 1 summarizes the  values of maximum possible product flow into the
 cylinder and required  product  flow for 100% cylinder  requirements.  The  table
 includes values  of  maximum   possible product  flow   per cylinder for reactor
 pressures at 100 and  150 psig  and product  temperatures at 400°F and 1200°F.
 These   temperature  values represent  the  boundaries   for the  product as it  is
 discharged  into  the  cylinder.    The  number  after  the  virgule in  each case
 represents the choking flow at  that condition.  The maximum possible amount of
 product injection is the lower of the two values. This is  because  further decreases
 in downstream pressure (cylinder pressure) such as that which occurs at product
                                        D-23

-------
valve start to open (14.2 psia) and full valve open  positions (23.9 psia), will not
cause an increase (or decrease) in the choking flow rate.

     The phenomenon of choking flow and the  values  of mass flow  rates at a
mach number of 1 is discussed in some thermodynamic texts.  The particular one
used  was written  by G.  J.  Van  Wylen,  entitled  "Thermodynamics" and  was
published in  1963 by John Wiley and Sons.

     Also, the equation for calculating mass flow rate, given the proposed value
size, is not sensitive to the value of absolute pressure.

     It can be concluded, by looking at Table 1, that  the  maximum  possible
product flow in most cases is at or  near choking flow.  (It is this at both reactor
pressures because of the  large pressure  differential.)   This  poses  no problem,
however, because the required amount of products  for  100% fuel needs  is well
below  these  maximum  possible  values as  shown in  the far right  column.  In
addition, the  section on heat balance (MPR #8),  calculates that there is not
sufficient exhaust heat to  vaporize and dissociate 100% of the engine fuel needs,
so somewhat less that the required amount of fuel will be injected each cycle (not
considering the increased heat of combustion of the dissociated products relative
to liquid methanol).  Even if the dissociated product valve duration is reduced to
73°  from 90°  Crank angle (19%  reduction in duration),  the maximum  possible
amount of product flow will still exceed the  amount required.

Sensitivity of the Heat of Dissociation to Pressure

     The heat of dissociation is independent of pressure for  an ideal  gas and
nearly  so for other reactions.  Pages 70 and 77 in "Thermodynamics for Chemists,"
by Samuel Glasstone,  D. Van  Nostrand Company,  discusses this. The  actual value
of the heat of dissociation and, for  that matter, the heat of vaporization vary
depending on  the reference,  but tend to cluster about certain values as shown in
Table 2.
                                          D-24

-------
Connector Between Reactor Housing and Cylinder Head

      This can be reduced to virtually zero length which will shorten the overall
size of the  heat exchanger and reduce heat looses.  We will look at exhaust gas
distribution  as it exits  the  port to assure that  the  transition to  the  reactor
provides uniform flow through  the annuluses.  Insulation can also be applied to the
outside of the connector (if a connector  is  required) and the  heat exchanger
housing to minimize heat loss.

Value for Thermal Conductivity

      The value  for thermal conductivity used in the calculations in MPR #8 was
22 Btu/(hr)(ft2)(°F/ft).  Discussions with Professor Webb indicate that this was in
error and a  value of  8 to 10 Btu/(hr)(ft2)(°F/ft)  would be more appropriate.  This
change has  no  effect  on the  final  reactor design because  the  catalyst  area
requirements exceed those for heat  transfer.   The evaporator lengths  shown on
Table 2 of MPR  #8  will be increased by about 31% to the values below:

          Configuration                            New Length (in.)
                 1                                         5.0
                 2                                         8.1
                 3                                         5.3
      We will, however, investigate other materials for the evaporator that  may
have  higher coefficients of thermal conductivity.  There is no need for material
compatibility between the evaporator and reactor.  In addition, the evaporator is
exposed  to lower exhaust temperatures and does not need  to  have  a  surface
compatible with  the  catalyst washcoat.  An evaporator material  with a higher
thermal conductivity  will  reduce the overall size of the heat exchanger.

Evaporator Flooding

      During our June 10 telephone conversation you mentioned your concern  over
reactor flooding with the inclination of the reactor and evaporator as shown on
Figure 3A in  MPR  #8.    We had  considered this possibility  in  configuring the
reactor/evaporator as shown.   We theorized that the lower evaporator would
prevent liquid fuel from entering  the  reactor and, if insufficient  vaporization did
occur then vapors would not be passed into the reactor which would  mean that the
exhaust gas temperatures would not decrease appreciably as it passed through the
reactor.  These higher temperatures would then be available to vaporize the liquid
methanol  in the evaporator.  This does not mean  that liquid  methanol could not
build  up  in the evaporator.  This is a possibility,  particularly as the  evaporator
heat transfer effectiveness is lowered with the presence of the bulk liquid rather
than having it spread  out on the  evaporator surfaces.

      The   disadvantage  of  having the  reactor/evaporator  inclined  downward
toward the engine is  that  liquid methanol could enter the  reactor.  Its presence
                                          D-25

-------
will force the reactor to operate at the saturation temperature of  the vaporized
methanol (289°F at  150  psig reactor pressure)  which  is below the  minimum
temperature  for significant  dissociation.   While vaporization may occur, the
engine will  receive  either  liquid  or vaporized methanol.  This may  be a  less
objectionable consequence  of  overfueling than puddling of fuel  in the evaporator.
In any case it may be necessary to include a safety valve in the line between the
evaporator and reactor or at the reactor outlet.
                                     Sincerely,
                                                  0.
                                     Gerald Driscoll
                                     Senior Research Engineer
                                     Vehicle Systems
                                     Department of Engine and Vehicle Research
/tme
Attachments
                                        D-26

-------
                       Table 1.  Maximum Possible and Required Product Flows Per Cylinder
                               Per Cycle at Various Engine and Reactor Conditions
Engine          Maximum Possible Amount of Products Entering/Cylinder/Cycle (Ibm x 10~3)       Required Product
 speed             Reactor Pressure at 150 psig            Reactor Pressure at 100 psig              Flow
 (rpm)             &00°F*               1200°F             400°F             1200°F            (Ibm x 1Q-3)
  500              4.7/4.4**            3.6/3.4             2.9/3.1            2.1/2.3

 1500               1.5/1.5              1.2/1.1             1.0/1.1            0.7/0.8                0.14

 4500               0.5/0.5              0.4/0.4             0.3/0.3            0.5/0.3                0.14

 6000               0.4/0.4              0.3/0.3             0.3/0.3            0.2/0.2                0.14


 *  Temperature of dissociated products.

 ** First value in each cell indicates product quantity based on pressure difference between reactor and cylinder at
    45° BTDC. Second value indicates choked or maximum amount of products.

 NOTE: Slight differences between values shown here and in our MPR //8 are due to the use of different values for the
        ratio of specific heats (k).  This table uses k =  1.4. The data in MPR //8 uses k = 1.3.

-------
               Table 2.  Values of the Heats of Vaporization and
                          Dissociation of Methanol
Value (Btu/lbm)

      545



      474


      482


      481
CH3OH HEAT OF VAPORIZATION

                        Reference
     Lalk. T.R., "Evaluation of Dissociated and Steam Reformed
     Methanol as an Automotive Engine Fuel," May 1984, NASA
     CR 168242, DOE/NBB-0064, May 1984, p. 48.

     Taylor, C.F., "The Internal Combustion Engine in Theory
     and Practice," Vol. I, p. 46.

     Baumeister and Marks, "Standard Handbook for Mechanical
     Engineers," pp. 4-14.

     Bechtold and Timbario, "The Theoretical Limits and
     Practical Considerations of Decomposed Methanol as a
     Light-Duty Vehicle Fuel," 1984 Alcohol Fuels Symposium,
     pp. 1-24.
Value (Btu/lbm)

     1500*

     1202


     1188
 CH3OH HEAT OF DISSOCIATION
(Not including heat of vaporization)

                        Reference
     Lalk, T.R., p. 48.

     Finegold, J.G., "Dissociated Methanol-Vehicle Results,"
     April 1984, SERI/TP-234-2245, p. 2.

     Bechtold, pp. 1-24.
 * This may include some specific heat used to raise the temperature of the vapors
   to the point where dissociation may occur.
                                        D-28

-------
                 APPENDIX E
CORRESPONDENCE WITH CATALYST MANUFACTURER

-------
   SOUTHWEST  RESEARCH  INSTITUTE

   POST OFFICE DRAWER 28510 • 6220 CULEBRA ROAD • SAN ANTONIO, TEXAS, USA 78284


   ENGINES. EMISSIONS AND VEHICLE RESEARCH DIVISION         AUgUSt 16, 1985
Dr. Alan D'Alessandro
Johnson-Matthey Inc.
436 Devon Park Drive
Wayne, PA 19087

Subject:   EPA Reactor Specifications for Dissociating Methanol

Dear Dr. D'Allesandro:

     I have attached  specification sheets that EPA recently sent to  us listing reactor
parameters and values that they would like used. These are the ones I mentioned in our
recent telephone  conversation.  Would you please  review these  and tell me if you can
comply with them, if you would recommend other valves  or materials, and if you would
like me to set up a conference call with EPA to discuss concerns that you have?

     As  defined in the attachments  I sent you previously, the EPA would like  to use
reactor  configuration number 7 rather  than number  6.  To  permit the reactor to fit
inside a tube with an 3.584 inch inside diameter, the exact dimensions of the  tubes and
the mesh and fin  material need to be changed  from the  values previously listed.   You
may want to adjust these  dimensions  to fit available  fin and mesh  thicknesses  and
tubing sizes.  The length will need to be calculated to achieve the desired catalyst EPA
recommends or what you feel is necessary.

                HEW DIAMETER DIMENSIONS  FOR CONFIGURATION NO. 7

                                              DIMENSIONS  (IN)
           DIAMETER                                            DIAMETRAL
         DESIGNATION                 DIAMETER                 DIFFERENCE

              D!                        1.942
                                                                  0.196
              D2                        2.138
                                                                  0.060
              D3                        2.198
                                                                  0.400
              D4                        2.598
                                                                  0.060
              D5                        2.658
                                                                  0.196
              D6                        2.854
                                                                  0.060
              D7                        2.914
                                                                  0.400
              D8                        3.314
                                                                  0.060
              D9                        3.374
                                                                  0.196

                                       3.570
                                             E-2

-------
     I will be on vacation from August 20 through September 3.  I will call you when I
return to  ask you for your  decision on building the reactor to configuration  7  and  to
discuss  any questions  you may have.   If  it  appears  that  we can go  ahead  with
fabrication, I would like to pursue this with a minimum of delay.

     If you have any questions while I am on vacation,  please feel free to contact Bill
Likos at (512) 684-5111, x-3135.

                                         Sincerely,
                                         Gerald D. Driscoll
                                         Senior Research Engineer
                                         Vehicle Systems
                                         Department of Engine and Vehicle Research
/ds

cc:  R. H. Thring
     B. Likos (02)
                                          E-3

-------
          INTAKE VALVE
                \
EXHAUST VALVE
                                              DISSOCIATED
                                                                                    PRODUCTS OUT
                                                               DISSOCIATED CH3OH
                                                              PRODUCTS TO ENGINE
                            WATER
                            JACKET
                                                                                  REACTOR    SECTION R2-R2

                                                                                  R1
      NAPS-Z ENGINE CYLINDER
              HEAD
                                COMBUSTION CHAMBER
                                                                                             EVAPORATOR
                             EXHAUST


                            METHANOL
                                                              METHANOL SIDE
                                                           FIN CONFIGURATION TO
                                                             BE FINALIZED LATER
                                                                                                       LIQUID CH3OH
                                                                                                        FROM PUMP
            SECTION R1-R1
     SECTION E-E
NOTE: DIMENSIONS DEFINED AND          REACTOR-EVAPORATOR ASSEMBLY
      DISCUSSED ON FOLLOWING SHEETS  FOR EVAPORATION AND DISSOCIATION
                                     OF METHANOL
                                     FIGURE 3A

-------
                                                       Diameters (inches)
M
I
      Configuration   Type    Dj
                            D/;
                                             D-,
                                                                             Fin Description
            1
            2
            3
A
B
B
B


A

B

B
                Dimension
                   03
                                                                    --  Strip fin Figure 10-56 Kayes
                                                                        and London, 3rd Edition,
                                              	098" height, 20 per inch,
                                                                        0.125 inch length in direc-
0.800  0.996   1.036   1.436   1.476  1.672  1.712   2.112   2.152   2.348  tion of flow
1.000  1.100   1.140   1.336
0.800  0.996   1.036   1.436  1.476  1.672
                            0.800  1.300   1.340  1.740  1.780  2.280


                            3.544  3.744   3.804  4.000

                            3.088  3.283   3.344  3.744  3.804  4.000
                                                                    --  11 plain fin/inch: Figure
                                                                        10-26 Kayes and London,
                                                                        2nd Edition
                                                                    --  Figure 10-56 Kayes and
                                                                        London, 3rd Edition
                                                                    --  Figure 10-56, Kayes and
                                                                        London, 3rd Edition
                            2.372   2.568   2.628  3.028  3.088  3.284   3.344   3.744  3.804  4.000  Figure 10-56, Kayes and
                                                                                                    London, 3rd Edition
                              Type A
                                              TypeB
           Description '
     Inner closed cylinder
     Reaction side annulus
     Wall
     Fin - exhaust annulus
                  - FIN
                  BEHR  SUBSTRATE
                           Dimension

                              D2
                              D3
                              D4

                              D6
                                                          D9
                                                          DID
                                             Description
                                       Inner closed cylinder
                                       Fin
                                       Wall
                                       Reaction side annulus
                                       Wall
                                       Fin
                                       Wall
                                       Reaction side annulus
                                       Wall
                                       Fin
                         FIGURE 3B.  CROSSECTIONAL GEOMETRIC DATA OF REACTOR CONFIGURATIONS

-------
Configuration
     1
     2
     3
     4
Type
 A
 A
 A
 A
                 Dimension
                    Dl
                    D2
                    D3
2.064
1.858
3.744
3.538

2

D,
.1040
1.898
3.804

3.598

2
2
D,
^^^^^~
.300
.300
4.000
4
.000

Figure
Figure
Figure
Figure

10-56
10-60
10-56
10-60
Fin
Kayes
Kayes
Kayes
Kayes
Description
and
and
and
and
London,
London,
London,
London
3rd
2nd
3rd
2nd
Edition
Edition
Edition
Edition
Description
                     Inside diameter of evaporation chamber
                     Wall
                     Fin
                                                                                 FIN
             FIGURE 3C.  CROSSECTIONAL GEOMETRIC DATA OF EVAPORATION CONFIGURATION

-------
SOUTHWEST  RESEARCH  INSTITUTE

POST OFFICE DRAWER 28510 • 6220 CULEBRA ROAD • SAN ANTONIO, TEXAS, USA 78284

ENGINES, EMISSIONS AND VEHICLE RESEARCH DIVISION          _
                                       June 18, 1985
 Dr. Alan D'Alessandro
 Johnson-Matthey Inc.
 436 Devon Park Drive
 Wayne, PA  19087

 Subject:    Catalyst Selection and Application to a Reactor for Dissociating
            Methanol

 Dear Dr. D'Alessandro:

      I am writing with regard to our recent telephone conversation to provide you
 with some information concerning the  reactor that we  are  designing,  and to ask
 for some information that we need to finalize this design.

      We are conducting a program for the EPA to evaluate direct injection of
 dissociated methanol (CO and H2) into  the combustion chamber of an engine. We
 need a catalyst manufacturer to provide an appropriate catalyst and to apply it to
 a reactor that we would supply.

      With regard to this we would  like to ask,  first of all, if Johnson-Matthey
 would be willing to assist us by recommending a catalyst and applying it to the fin
 tube surface of our  reactor.  The reactor will operate at 100 to 150 psig and we
 would  like  the  catalyst  to  survive maximum  exhaust temperatures of  about
 1300°F.  A catalyst  that would produce significant product yields at temperatures
 as low as 400°F is also desired.  The primary products of dissociation should be
 CO and  H2 under these  conditions.   The attachments show  the basic reactor
 configurations we are considering.

      If  you  can  assist us,  we  need  certain  information about  your  selected
 catalyst  to permit us to finalize the reactor design:

      1.    Catalyst yield at  the elevated pressures and at several temperatures.

      2.   Approximate degradation  of catalyst yields as a function of time and
            temperature.

      3.    Maximum  allowable  temperature  the  catalyst can be  exposed to and
           retain its function.

      4.    Minimum  temperature at which  significant (e.g. 50%) yields may be
           realized.
                                         E-7

-------
      5.    Description of product species at the elevated pressure.

      6.    Materials that  the  reactor  should  be  made  of  to  provide  good
           adherence for the wash coat.

      7.    Minimum  substrate  fin spacing that  will permit the  wash coat  to
           conform to the substrate and not bridge the fins.

      8.    Suggested prefabricated high  density  fin or substrate material  to
           provide maximum catalyst surface  area in a minimum volume.

      9.    Approximate cost and time required to perform the application.

      Your early response to our inquiry would be appreciated.  Please call me  at
512/684-5111, extension 2047.

                                      Sincerely,
                                     Gerald D. Driscoll
                                     Senior Research Engineer
                                     Department of Engine and Vehicle Research
/tme
Attachments
cc:   R. Thring
     W. Likos
                                      E-8

-------
                      APPENDIX F
      TEST PLAN FOR EVALUATING HEAT EXCHANGER
WITHOUT OPERATING ENGINE ON PRODUCTS OF DISSOCIATION

-------
          SUMMARY OF PREPARATIONS FOR HEAT EXCHANGER
                          LABORATORY TESTING
Purpose of Testing

     To measure  heat exchanger  performance  under actual engine  operating
conditions so  that a comparison  of  analytical and experimental  results  can^ be
made, the actual capabilities to dissociate methanol can be assessed and the final
heat exchanger design details can  be determined.

Preparations

1.0  Heat Exchanger

     1.1   Complete fabrication  of  heat exchanger shells and ship tp Johnson
           Matthey 0  M).

     1.2   Have J M apply substrate, catalyze and assemble tubes.

     1.3   Finish heat  exchanger  plugs, ports, ducts, and mounts

     1.*   Design  provisions for mounting nozzle  to heat exchanger after decision
           on nozzle is made.

2.0  Fuel Injection

     2.1   Need information from Stanadyne on the suitability of their  nozzles
           for this application and a design for the nozzle tip to spray methanol
           over the inside evaporator surface.

     2.2   If  Stanadyne  nozzles  appear  suitable they need to be  evaluated for
           their spray  pattern on  the test stand at SwRI.

     2.3   If  the Stanadyne  nozzles cannot  be  made  suitable then the existing
           SwRI design can be used.

3.0  Engine Modifications

     3.1   Prepare package of existing calculations and send to EPA.

     3.2   Design  modifications to adapt Stanadyne DB2 pump to NAPS-Z engine
           (need engine).

     3.3  Complete engine  design modifications with regard to valve train and
          cam drive and submit drawing package to EPA (need engine).

4.0  New Parts Required to Mate Engine and Heat Exchanger

     Design and fabricate  parts  to connect  the  engine and heat exchanger; the
heat exchanger and exhaust manifold; and the heat exchanger to various sensors.
                                       F-2

-------
5.0  Control System

     No control system design efforts  need to be accomplished other than to
provide for adjustment and locking of the Stanadyne pump rack.

6.0  Test Setup

     6.1   The  NAPS-Z engine would be  installed  in  a  standard  test cell and
           connected  to a dynamometer, coolant pipes,  and the cell exhaust
           system.

     6.2   The  engine  would operate  on its existing unmodified fuel injection
           system. It would require no modifications other than the  addition of
           the Stanadyne pump.

     6.3   The heat exchanger would be mounted in one housing that connected
           one cylinder exhaust port to the exhaust manifold.  The other exhaust
           ports  would be connected  by housings  that  did  not  contain  heat
           exchangers.  Exhaust  from  one  cylinder  would  flow through the
           reactor, then through the evaporator, counterflow to the methanol.

     6.*   Methanol would be pumped at high pressure by the  Stanadyne pump to
           the nozzles. One nozzle would be  mounted in the evaporator and spray
           methanol on its inside surface.  The  other nozzle  discharges would be
           routed back to the methanol fuel supply. Methanol  would be vaporized
           in  the evaporator  and  these  vapors  would flow into the reactor.
           Reactor  discharge  would   be   collected   for   analysis  via   gas
           chromotography.  The dissociated products would  not be injected into
           the engine.

7.0  Test Instrumentation

     The  engine  and  heat exchanger  would  be instrumented to monitor the
following parameters.  The values of these parameters would  be used  to measure
the performance of the heat exchanger (MPR  #10).

             Parameter                        Location
     Temperature                    Methanol into the evaporator
     Pressure
     Mass flowrate

     Temperature                    Methanol between the evaporator
                                     and reactor

     Temperature                    Dissociated products out of
     Pressure                        reactor
     Product analysis
                                       F-3

-------
      	Parameter	      	Location	

      Temperature                    Exhaust out of engine
      Pressure

      Temperature                    Exhaust between reactor and
      Pressure                       evaporator

      Temperature                    Exhaust after evaporator
      Pressure

      Engine speed                    Engine crankshaft

      Mass airflow                    Engine intake

      Mass flowrate                   Methanol injected into manifold
                                     or ports

      Temperature                    Engine coolant


8.0   Test Procedure

      The engine would be operated  over  a matrix of loads  and speeds.  These
would represent steady-state conditions.  The heat exchanger  would be subjected
to a matrix of pressures and flowrates at each engine test  point.  Measurements
from  the instrumentation listed in paragraph 7.0  would be obtained at each test
point.

9.0   Use of Results

      The test results will be used to define the following:

      9.1   Heat exchanger, evaporator, and reactor  exhaust side effectiveness in
           terms of methanol and exhaust temperatures at each test point.

     9.2   Efficiency  of  the dissociation  process in terms of methanol dissociated
           versus total methanol flow at each test point.

     9.3   Heat exchanger total, evaporator and reactor pressure drops at each
           test point.

     9.*   Comparison of  above  experimental results   with  calculated  design
           specifications.

     9.5  Qualitative  and quantitative  analysis  of all  product  constituents  to
          determine  if  compounds  or elements  other  than  hydrogen,  carbon
          monoxide, and vaporized methanol are present at each test point.

     9.6  Optimum Heat exchanger total  size  and relative sizes for the reactor
          and evaporator.
                                      F-4

-------
         APPENDIX G
CYLINDER HEAD AND CATALYTIC
   REACTOR DESIGN OPTIONS

-------
        CYLINDER HEAD AND CATALYTIC REACTOR DESIGN OPTIONS
     Four layout  drawings  were  made  which  show  two possible  arrangements,
one  which would  make  use of  the  existing  cylinder  head and one  which  would
use a new, modified cylinder head casting.

     Layout   drawing  number  03-8352-003   shows  the  overall assembly of  the
cylinder  head  with  the  catalytic  converters.   An  independent catalytic  converter
is  planned for  each cylinder.   In this  layout only the exhaust port  configuration
along  with the  converter  housing  and  the   converter are  shown.    A  casting is
being  prepared  for  the  converter  housing.    This  casting  will  have  provision  for
accommodating  either  two  or  four  converters.    The converters   will  be  bolted
onto the  housing.    Depending  upon  the design  selected (two  or  four  converters
per  housing), one   or  two  converter housings will be built.    The  advantages or
disadvantages  of  each  of  the  arrangements  are  discussed  later   in  this  report.
The housing(s)  will  be bolted to the cylinder head in the existing  holes for  the
exhaust  manifold.    Also,  the  hole  patterns on  the  converter  housing's  outlet
flange  are such that  the  exhaust manifold   can  be  used without any  modifica-
tions.

     The  layout  drawing  also  shows  the  initially  proposed  construction   of a
crossflow  catalytic  converter.   The converter  consists  of header tanks  at both
ends.   Stainless steel  tubes of  approximately 0.125  and  OD  x  0.015  inch  wall
thickness   would  form  the  body to   support  the  catalyst.    The  tubes  would
provide  large  surface  areas  for  catalytic  reaction and  heat  transfer  from  the
hot  exhaust  gases  to the  catalyst.   A methanol fuel injector  would be accommo-
dated  in  one of  the header  tanks.   The header  tank on the  other  side will be
used to  collect  the  dissociated  products.   This crossflow design was discarded in
favor  of  a counterflow  design  which is  more efficient  for  a  dissociation reactor
application.

     In  the   above  design   configurations,  no  cylinder  head   modifications  were
contemplated.    However,  adoption  of  this  design  means  some  heat  loss  will
occur  from  the  exhaust  gas  to  the  water  jacket.   In the  present cylinder head
design, a  portion  of the exhaust port  is surrounded  by  the  water  jacket.    The
heat loss  to  the cooling  water  can be  eliminated  if  the  cylinder  head  casting is
modified  as  shown  in  drawing  03-8352-005.   In  this design modification, a  cored
air space  is  added  between the  port  wall and  the water jacket.    Alternatively,
ceramic  lining  of   the port  surfaces  or  a port  liner made  up  of  an  insulating
material will  be  considered to reduce the heat loss.

     Layout  drawings  03-8352-004  and  -006 show the  gas  (dissociated  products
of  methanol)  admission valve.   In drawing 03-8352-004,  the  valve is designed to
fit  into  the  spark   plug  hole  adjacent to the exhaust  valve/port.    Although  some
enlargement  of  the  existing 14  mm  spark plug  hole seems  possible,  this  design
will restrict  the  gas  flow  area  to  slightly  more  than  0.1  square  inch.    It  is
estimated  that,  at  5200  rpm  engine speed,  the  dissociated products  will  have to
be  admitted  to the cylinder  in 1.6  milliseconds.   This  result was  based  on  the
assumption that the  gas  admission  valve will  be  opened  for  a  period of  50  crank
degrees  during   the compression  stroke  of  the  engine.    The  volume of  the
dissociated methanol produced in the  catalytic  converter  at  full  load-rated   speed
of  the NAPS-Z  engine is  estimated  to  be 1.6 cubic  inches.    In  reality,  however,
                                         G-2

-------
the  construction   features  of  the  converter  and  the  joining  passages  are  the
limiting  factors.   Based  on  current  design of  the  converter,  it  is  estimated that
the  volume  of  dissociation  products  to be  transferred  to  the  cylinder is  about
3.5 cubic  inches.   The calculations indicate  that the  velocity of  the gas  through
the valve  would  be in excess of  1000  ft/sec.   Such  a high velocity is  considered
to  be  undesirable;  therefore,  it  is  essential  that,   among  other  things,  the  gas
inlet  valve be designed as  large  as possible.    Depending  upon  the decision made,
this issue will be suitably addressed in the detailed design of the components.

      Layout  drawing  03-8352-006 shows  the  gas inlet valve in  a  modified (new
casting  and  machining)  cylinder  head.   By  providing  an  independent boss  in  the
casting the valve size is considerably enlarged.

      It  can  be  seen  that,  due  to   the   large  size  of  the converter,   which  is
necessary  for  heat  transfer  and  catalyst area  considerations,  it  has  not  been
possible  to  incorporate  the  converter  within  the  confines  of  the  cylinder  head.
however,  the  essence  of  in-cylinder   dissociation  has  been  retained,  since  the
dissociated  products  are  still  admitted  directly  to  the   cylinder,  without  being
first mixed with incoming air.

      Valve  operating  mechanism   is   another  important  factor  in  the  design  of
the  in-cylinder   methanol  dissociation   scheme.   Both modified   and unmodified
cylinder head  layouts show the valve  operating  mechanism.   A separate camshaft
driven  by  the  engine  is  being  considered.    The  cam  is  designed to   directly
operate  on   the   follower   attached  to  the   valve  stem.    No  difficulties  are
anticipated in the design  of the  cam/follower  assembly  and  the  camshaft  drive
arrangement.   However,  for  the  existing  unmodified  cylinder head,  lubrication of
the  valve  stem and  the  drive mechanism  is likely  to be  far  more complex than
the  one for  the  modified  cylinder head.   In  the  modified cylinder  head  design
the  camshaft  supports,  the cam,  the   follower and   valve  stem are  all  accommo-
dated  on  the  cylinder head.   The new design has  larger modified  rocker covers
to house the  gas  inlet valve  and  the  valve  train.    A modified rocker  cover will
be used to house the entire valve gear (intake, exhaust  and gas inlet valve).

      As  an   alternative   to   the   mechanically  operated   valve,   consideration   is
being  given  to  an  electrical  solenoid-operated valve  design.   The possibility of
electrical  valve  operation  depends  upon   the  existence  of  an  electrical  solenoid
with  large force  and  displacement and capable of operating at a  frequency  of  at
least 2600 cycles/minute.

      Either  of  the  two approaches  discussed  above  is  possible.    The  modified
cylinder head approach,  though  advantageous, is  considerably  more  complicated
since an  entirely  new  head must  be  developed.   The  use  of the  existing  cylinder
head  design has  some functional   disadvantages;  however,from  an  execution  point
of  view  it  is  a  much simpler  and  less   time-consuming approach.   As  already
mentioned,  after  considerable thought,  the unmodified cylinder head was  chosen
for future work.

      Reasons  for  this  recommendation  are  as  follows.     There  are  four  main
factors  involved:  (1)  the  location  of  the  converter;  (2)   the gas  velocity  in  the
transfer  valve;  (3) the  insulation  of   the  exhaust  port;  and (4)  the  valve  train.
As  far  as  the location of  the  converter   is  concerned,  due to the  heat  transfer
and   catalyst   area  considerations,  the size   of  the  converter  is  such   that  it
cannot  be fitted   within  the  confines  of   the  cylinder head (neither  the  existing
                                          G-3

-------
head  nor redesigned  head).   There  is  not  benefit,  therefore, in designing  a new
head  from  this point  of  view.   The gas velocity in  the  transfer valve  would  be
too high  with the standard cylinder head, using  the 14 mm spark  plug  hole,  but
enlarging  that  hole  will   reduce  the  gas velocity  to  a  reasonable level.    The
production  of a  new  cylinder head  would  enable  the  exhaust  port  to  be more
effectively  insulated,  but  adequate  insulation  with  the  standard  head  can   be
achieved  by  fitting  an exhaust  liner.    The  valve  train will be somewhat  more
difficult to  lubricate  with the unmodified cylinder head, but this problem  can  be
overcome.

      The time  required  to  design and  develop  the  new  cylinder  head is esti-
mated to be  about  24 weeks.   If  the existing  cylinder  head  is used,  the time
required to design  and build the converter housing,  converters, valve  and  valve
gear  is estimated  to be  10 weeks.   Most of the  above parts are common  to  the
existing cylinder head or a  new cylinder head design.
                                          G-4

-------

-------
O
                                                                         SOUTHWEST RESEARCH INSTITUTE
                                                                                     l»» *IITOIIIO,TI>U
                                                                                            Y&*-V£
                                                                        SIZE
                                                                            CODE IOCNT NO.
                                                                           264OI
                                                                                     OWAWINO NO.

-------

-------
o
I
00
                                                                                                      SOUTHWIST MSEAtCH INSTITUTf
                                                                                                             - Moturieo
                                                                                                                             ncf)o
                                                                                                        COM IO«N'T MO.

                                                                                                        26401
03-8352-006

-------
                 APPENDIX H
EVALUATION OF THE METHANOL DISSOCIATION
              HEAT EXCHANGER

       (Includes a Table of Additional Data)

-------
              SPECIAL TECHNICAL PROGRESS REPORT
EVALUATION OF THE METHANOL DISSOCIATION HEAT EXCHANGER
                         SwRI Project 03-8352
                       EPA Contract 68-03-1984
                             Prepared by

                           Gerald D. Driscoll
                   Vehicle Research and Development
                      Southwest Research Institute
                           6220 Culebra Road
                       San Antonio, Texas 78284
                             Submitted to

                          Mr. Robert Bruetsch
              United States Environmental Protection Agency
                    Motor Vehicle Testing Laboratory
                      2565 Plymouth Road - ECTD
                      Ann Arbor, Michigan  48103
                            September 1987
                                  H-2

-------
INTRODUCTION

       A  heat  exchanger  system  has  been  evaluated  to  determine  its  effective-
ness  in  dissociating  methanol  into hydrogen and  carbon  monoxide.   This  system
consists  of  evaporator  and reactor sections that  are  heated  by  engine  exhaust
gases.  Methanol  is   injected  into  the  evaporator  and,   after  being   vaporized,
passes over  a  dissociation  catalyst in  the reactor.   This  heat exchanger  system,
designed  for application  to only  one  engine cylinder,  is  10  cm  (four  inches)  in
diameter  by  50 cm  (20  inches)  long.   Background  information  on the  design  of
the  evaporator  and  the  reactor  is given  in the  section  "Heat Exchanger  Design
Considerations."

SUMMARY

     The  methanol  dissociation  heat   exchanger was  evaluated  over  a  range  of
engine speeds and  power outputs.   Conditions in  the  heat  exchanger  were stable
at  low to moderate  engine speeds and  loads,  enabling  determination  of methanol
dissociation  rates.   At high  engine speeds  and with high  loads  at  lower  engine
speeds, the  conditions  in  the  heat exchanger  were  very  unstable,  and meaningful
methanol dissociation  rates  could not  be determined.   The  lower engine  power
operating  conditions  are  considered  to  be  representative  of  typical  steady-state
engine operating conditions for  light-duty vehicles.

     The  mechanical design  of the heat  exchanger  was generally  satisfactory,  but
some  additional  improvements  in  injector  mounting   and  injector  cooling   are
needed.   The lack  of  adequate  injector cooling was a  major  factor in the  inabi-
lity to operate the heat exchanger at high engine speeds and loads.

      In  these evaluations,  the  dissociation rates in  the reactor  ranged from  1  to
96   percent   of  the   total  methanol  injected   into the  evaporator;   the   higher
dissociation  rates  being   associated  with  low  methanol  injection  rates.     This
represented  from  0.6   to  9.8  percent of  the  engine cylinder  fuel  needs   at  the
test  points  evaluated.    The  reactor  size,  or  the  effectiveness  of   the catalyst,
will have  to be  increased  by  a factor  of  2  to  4  to  achieve sufficient dissocia-
tion  to  meet 20  to  25  percent  of  the  total  fuel  needs of the engine  at   the
lower power operating conditions.

     The  evaporator  was  adequately sized to vaporize  100  percent  of the  engine
cylinder   fuel  needs  at  3000  rpm,  but  only  50  percent  at  1500  rpm.    The
vaporization   rate  was  greater   at  3000   rpm because   of  the  higher  exhaust
temperature and flowrate at  that engine speed.

TEST MATRIX

      The  engine, with  the  heat exchanger   mounted,  was  operated   at  engine
speeds that  ranged  from  1500 rpm to 4500 rpm  and  loads  that ranged from  30
Ibs-ft  to  110 Ibs-ft.   During  initial  testing,  however,  it  became apparent  that
the  evaporator   reached  excessive  temperatures when  operating  at  high  exhaust
temperatures  and  flowrates.   A  major factor was  overheating  of  the  methanol
injector,   resulting  in   injector  malfunction.    This  limited   testing   to  low  load
conditions at 1500 and 3000  rpm.

      Evaluations  conducted  involved  the  determination   of  dissociation,   with
several  methanol flowrates  through the  heat   exchanger  at two  engine  operating
                                        H-3

-------
conditions.   Also, evaluations  were  conducted  to  determine  the effect of  reactor
pressure and temperature on the rate of dissociation.

METHANOL DISSOCIATION IN HEAT EXCHANGER

      A summary  of  the test  results is given  in  Table  1.   Table  1  includes the
dissociation  rate  in  percent on  a mass basis  versus methanol  mass flowrate  into
the heat  exchanger.    The  amount of  dissociated  products as  a function  of  total
engine  cylinder  fuel  needs  at  these test  points varied  from  0.6  to  9.8  percent.
The  reactor  performed more  efficiently  at  the  higher engine  speed  and  load
conditions, because of the high exhaust temperatures and flow rate.

Effects of Fuel Rate

      A comparison of  dissociation rate, as  a  function of  methanol  mass  flowrate
into the  heat exchanger, is  included in Table  1 and shown in Figure 1.    There
is the  expected  inverse relationship  between the  percent of dissociation  and  fuel
rate.

      Figure  1  also  shows  the product of  the dissociation  rate  and  the   percent
of  engine cylinder fuel need.   This product represents  the  amount  of dissociated
products  that could  be  supplied  as  a  percent  of  the total  cylinder  fuel  needs  at
a  specific  engine operating  condition.   The   dissociation  rate as  a  function  of
cylinder requirements at the 1500 rpm test  point  appears to  reach a  maximum  of
4.2 percent  of  requirements at about  0.5 Ibs/hr methanol flow through  the  heat
exchanger.   Only  one flowrate  was  used at the higher engine speed  and load,and
it  resulted  in   a relatively  high   rate  of dissociation   (11.0  percent)  and  the
highest percentage  of  cylinder fuel  need  (9.8  percent).   Excessive  temperatures
in  the  heat  exchanger  at  higher engine speeds and loads  eliminated  the  possi-
bility  of  testing  at  those  conditions.   The  percent  of  cylinder fuel  needs, as  a
function of heat exchanger methanol mass flowrate, is shown in  Figure 2.

Effects of Internal Pressure

      Two comparative  evaluations  (Test runs  3  and  7) were performed  at the
same  test conditions,  except for  an elevated   evaporator and  reactor  pressure  in
Run  7.    Results of  these  two  tests  are   given  in Table  2.   There  was little
difference  in  the rates  of dissociation, between  these   two  tests.    This  result
was  to be  expected  because Johnson-Matthey  reported  that  the dissociation at  8
atmospheres  of  pressure would  be   about   93   percent   of  the  dissociation  at  1
atmosphere.   Although  the  literature  discusses the  tendency  of the  reaction  to
slow  down  and  reverse  at  higher  pressures,  change   in  pressure  should  have
negligible  effect over the range of  pressures expected in this application.
                                          H-4

-------
         96% @ 0.02 LBM/HR
   20
   15
o
cr
tu

% DISSOCIATION
% CYLINDER FUEL
REQMTS
POSSIBLE WITH
DISSOCIATION
PRODUCTS
1500 RPM
& 30 LB-FT
A
O



3000 RPM
& 40 LB-FT
O



                          Percent dissociation as a function of methanol
                          mass flow rate
                                                                       O
                          Percentage of cylinder fuel need provided by
                          dissociated methanol
               1.0
2.0        3.0        4.0         5.0

     METHANOL MASS FLOW RATE (LBM/HR)
                                                                   6.0
7.0
             FIGURE 1.  HEAT EXCHANGER PERFORMANCE VERSUS
                         METHANOL MASS FLOWRATE
                                       H-5

-------
O)
Q
LU

Z3
u_

QC
LU
Q
>
u
                1.0
2.0        3.0         4.0        5.0



     METHANOL MASS FLOW RATE (LBM/HR)
                                                                    6.0
7.0
        FIGURE 2.   CYLINDER FUEL NEEDS VERSUS METHANOL  MASS FLOWRATE
                                             H-6

-------
                           Table 1.  Methanol Mass Flowrate Versus Percent Dissociation
       Methanol    % of Engine
% Dissociation
Engine
Run*
No.
6
4
7
3
9
5
8
Mass Flowrate
(Ibm/hr)
0.02
0.5
0.95
1.0
1.1
2.0
6.3
Cylinder
Needs
0.6
19
29
31
35
63
89.4
Percent
Dissociation
96
22
7
5
5
1
11
X
% Cylinder Needs
0.6
4.2
2.0
1.6
1.8
0.6
9.8
Speed
(rpm)
1500
1500
1500
1500
1500
1500
3000
Torque
(Ib-ft)
30
30
30
30
30
30
40
Heat Exchanger
Condition
Insulated
Uninsulated
Uninsulated/pressurized to 35 psig
Uninsulated
Insulated
Uninsulated
Insulated
Analysis of heat exchanger products for run numbers 1 and 2 could not be made because test equipment was not set up to
handle the significant quantities of methanol in the sample.

-------
    Table 2. Effects of Evaporator Internal Pressure on Rates of Dissociation

Run Number                                   3          4

MeOH Mass Flowrate (Ibm/hr)                   1.0        0.95

Average Reactor Pressure (psig)                  0          33

Percent Dissociation                             5          7

Average Reactor Temperature (°F)*              432        236

Insulation                                 Uninsulated     Uninsulated

Effects on Temperature

    Two  comparative  evaluations  (Runs  3 and 9)  were  made  at  the  same  test
conditions,  except  that  the  heat exchanger  was wrapped  in  an  insulated blanket
for  Run  9.   Results  of  these  two  tests  are  summarized   in  Table  3.    The
dissociation  rates  were  the  same for  both  tests,   although the  average  evapora-
tor/reactor  temperature for  Run  9  was  138°F  higher  than  that  for  Run 3.
Johnson-Matthey  indicated that maximum dissociation occurred  with this catalyst
at 370°C  (698°F)  and  that  there were significant  yields  at 300°2C  (572°F).   For
Run  9,  the  average  temperature  between  inlet   and outlet  of  the reactor  was
300°C (570°F).   This  temperature  is  lower  than  that  for maximum dissociation,
but  it  is  at  the point  of  significant yields.   Operation  at an  average  reactor
temperature  of  222°C  (432°F)  in  Run  3   should have resulted  in  a  lower
dissociation  rate.

              Table  3.  Effects of Temperature on Rates of Dissociation

Run Number                                   3          9

MeOH Mass Flowrate (Ibm/hr)                   1.0        1.1

Average Reactor Pressure (psig)                  0          0

Percent Dissociation                             5          5

Average Reactor Temperature (°F)*              432        570

Insulation                                 Uninsulated     Insulated

Evaporator Temperature (MeOH)(°F)             198        352

Reactor Temperature (MeOH)(°F)                666        788

Exhaust Temperature Before Reactor (°F)         707        877

Exhaust Temperature Between Evaporator
   and Reactor (°F)                             419        721

Exhaust Temperature After Evaporator (°F)       210        244

 Average  reactor   temperature   is  average   of  internal   evaporator  and   reactor
outlet (MeOH) temperatures.
                                      H-8

-------
Effects of Changes in Engine Load and Speed

       Run 8  was at higher  engine  load and  speed than  the  other tests and  had
a  higher  methanol  flowrate  through  the  heat  exchanger.   The  results are sum-
marized in Table 1  and  Figure  1.    It  was  not  possible  to  lower  the  methanol
flowrates  in  this test  because  the  evaporator  temperature  would  have  become
excessive.   The dissociation rate  in Run  8 was  11  percent.   It should be  noted
that the  average reactor  temperature  for  this test  was  395°C  (742°F),  which  is
higher  than  for  any  other  run  and  was  close  to the  optimum  temperature of
370°C   (698°F).    The combination of  higher  reactor  temperature and higher  heat
flowrate  was  apparently  responsible  for the  dissociation  rate  being  higher  than
in  any of the  tests  at  the  engine speed  of  1500  rpm.   In  addition,  the  test at
3000  engine  rpm had  the highest portion of  dissociated products  as  a function
of total cylinder fuel needs.

RESULTS OF EVAPORATOR MAPPING

Approach

     The  capacity   of  the  evaporator  to  vaporize  methanol  as  a  function  of
flowrate  was  also  determined.    All  testing  of the  heat exchanger was  accom-
plished  with  the  evaporator at  a  temperature   higher   than  the  boiling  point
(equilibrium  temperature)  of  methanol.   To  determine the maximum capacity of
the  evaporator  to   completely  vaporize  the  methanol,  a  series  of  tests  were
conducted.    Evaporator   internal   temperature  as  a  function  of  flowrate   was
determined at  1500  engine  rpm  and  30  Ibs-ft  torque  at  ambient  and  high
pressures,  with  and  without  insulation.   Testing  was  also  done  at 2000  engine
rpm and  35  Ibs-ft.   Testing  at higher speeds and loads,  however,   was difficult
for two  reasons.  First, once  the temperature  was above  the equilibrium point it
was difficult  to achieve  stability;  the  internal   evaporator  temperature changed
at a rate  of  up  to 2°C per  second.   Second,  attempts  to  find  the minimum fuel
flowrate  that  still  maintained  the   evaporator   at  the   equilibrium  temperature
were   unsuccessful,  because  after  any excess fuel  vaporized  in  the  evaporator,
the temperature  would  rise  rapidly.   Control problems   were  even more  severe
when the heat exchanger was wrapped in an insulating blanket.

     Testing  the evaporator,  by  measuring its  internal  temperature and  plotting
it  as  a  function  of  fuel  mass  flowrate through  the  evaporator,  assumes that  the
evaporator  vaporizes  all  of  the  fuel  once  the  internal  temperature  exceeds  the
equilibrium  temperature.     This   seems  correct   in  theory,   but  has  not  been
verified.  An   attempt  toward verification  through  a  heat  balance  between  the
exhaust and methanol sides of the evaporator was unsuccessful.

Results

     Results  of  evaluations  without  insulation at low and  high  pressure  in  the
heat exchanger with  the  engine  operating  at  1500  rpm  and 30 Ibs-ft  of torque
are shown in  Figure  3.    Pressure  versus  boiling  point  for   methanol   is  also
included.   The   equilibrium  temperature   for  each  test  closely   matched  the
theoretical  value.  Results  of  evaluations  with insulation  are  shown  in  Figure 4.

     The  maximum  fuel  rate at  which complete  vaporization  can  be expected  is
1.17 Ibs/hour without insulation  and  1.7  Ibs/hour  with insulation  with the engine
operating at  1500 rpm  and 30  Ibs-ft  of  torque.   Insulation increases  the  capacity
of  the  evaporator  about  45  percent   at  this  engine  operating  condition.    The
                                         H-9

-------
   500 r—
                       SYMBOL
ENG. SPEED
  (RPM)
TORQUE
 (LB-FT)
EVAP. PRES.
  (PSIG)
COMMENTS
 ALL DATA
  TAKEN
  W/OUT
INSULATION
   400
cc
cr
LLJ
Q-
O  300
a:
O
Cu
CC
   200
                                                           METHANOL BOILING
                                                       TEMPERATURE VS PRESSURE
                                                        (PERRY'S CHEMICAL ENGR.
                                                        HNDBK, 6TH ED., PP. 3-204)
   100
                                I
                                                         50
                         I
                                     100
                                      150
                                                              PSIG
                               1.0                        2.0

                             METHANOL MASS FLOW RATE (LBM/HR)

            FIGURE 3.   EVAPORATOR TEMPERATURE VERSUS  METHANOL
                     MASS FLOWRATE WITHOUT INSULATION
                                                  3.0
                                        H-10

-------
    500 i—
                         SYMBOL

                            X
                            0
  ENG. SPEED   TORQUE   EVAP. PRES.

     (RPM)  _    (LB-FT)      (PSIG)    COMMENTS

     2000        35         0       ALL DATA
     1500        30         0      TAKEN WITH
                                   INSULATION
    400
GC
D
a:
LU
Q.
tr
O   300

oc
O
CL
    200
                   1.0
2.0
3.0
4.0
5.0
6.0
                              METHANOL MASS FLOW RATE (LBM/HR)


         FIGURE 4.  EVAPORATOR TEMPERATURE VERSUS METHANOL MASS

                          FLOWRATE WITH  INSULATION
                                          H-ll

-------
     evaporator capacity  at the  higher  engine speed  test point  was three  times  what
     it  was  at the  lower  speed, while  the engine  only consumed  about  35  percent
     more  fuel.   These  data  are summarized  in  Table  4.   The evaporator maximum
     capacity,  as  defined  by   the  maximum  flowrate  that   yields   an  evaporator
     temperature  slightly  over  equilibrium  temperature,   is  less   than  the  engine
     cylinder  needs at 1500 rpm  and 30  Ibs-  ft,  but it exceeds the cylinder needs  at
     2000 rpm and 35 Ibs-ft.

                            Table 4. Maximum Evaporator Flowrates

                                                         Evaporator             Maximum
   Engine                                 Evaporator       Maximum            Flowrate as a
Speed   Torque                            Pressure        Mass Flowrate   Portion of Cylinder
(rpm)   (Ib-ft             Insulation          (psig)           (Ibm/hr)        Needs (%)

1500      30              Uninsulated        0                 1.17                  37

1500      30              Uninsulated        46                1.22                  38

1500      30              Insulated           0                 1.70                  54

2000      35              Insulated           0                 5.05                 112


     EVAPORATOR SIZE

          The  maximum  allowable  evaporator  flowrates can  be  determined  at  any
     engine  test  point as  discussed  in  the previous  paragraph.   The  evaporator  size
     is  adequate  at  2000 engine  rpm  and  35  Ibs-ft  and at higher  loads  and  speeds.
     In Run  8, at 3000  rpm and 40 Ibs-ft, the evaporator temperature  was  well  above
     the  equilibrium  temperature,   and  the  methanol  flowrate  through   the  heat
     exchanger  was about  90  percent of the cylinder  fuel  needs.   If the  slope  of the
     evaporator temperature versus  fuel  flowrate  is  similar to that  at 2000 rpm  and
     35  lbs-ft,then  the  maximum  fuel  rate  at  the equilibrium  temperature  would  be
     about  110  percent  of  the  cylinder  requirements.    Again,  it is  assumed   that
     maximum flowrate can be defined as a function of evaporator temperature.

          It  should  be  noted that  all evaporator   testing  was  done  with  the current-
     sized  reactor.     The  current  reactor  does not  produce  high dissociation,  so  it
     does not use large  amounts of exhaust  heat.   Therefore,  more  exhaust  heat  is
     available  for vaporization  of  the  methanol.    A   more  effective  reactor  could
     require a larger evaporator.

          While  evaporator mapping  indicated  that  the evaporator  could not  handle
     100 percent  of  the  cylinder  fuel  needs at the  lower  power level, the evaporator
     could  easily  vaporize  the fuel  rates required   to  achieve the  optimum dissociation
     rates in terms of the  maximum percent of the cylinder fuel needs.
     REACTOR SIZE

     General

         From  the  data  summarized in Table  1,  it  is  apparent that  this  reactor  will
     not provide  high dissociation  rates  at low  engine loads and  speeds.    Also,  the
                                                H-12

-------
percent of  cylinder  fuel  needs that can  be met by  dissociation  products  will  be
fairly  low.   Based  on the  test data,  a reactor  two to  four  times the capacity  of
the  present  one  (8.05  inches long)  would  be  required  to  provide  dissociated
products accounting  for  20  to 25  percent  of  the  cylinder  fuel  needs at  low  to
moderate engine power levels.

Adding Heat to the Exhaust Side of the Reactor

    Improvements  in  overall   reactor  dissociation   effectiveness   at   low   engine
power  output,   without   increasing   the  size  or  effectiveness   of  its  catalyst,
requires  an  increase  in  exhaust  temperature.    There  is  little  temperature loss
between the  engine  exhaust  port  and  the  inlet  to  the  reactor  based  on test
data.   Insulation could decrease  the heat  loss, but  the  benefit would be minimal.
Exhaust  temperature  could  be  increased   by  burning  additional  fuel   in  the
exhaust side of  the  reactor.   The  exhaust  side  of the  reactor is coated  with  an
oxidizing   catalyst,   and   adding  fuel  would  increase   the  exhaust   temperature.
However,  the  gain  in  dissociation  and  engine  efficiency  would  not  offset  the
reduction in thermal efficiency  from  adding fuel into the exhaust.

ANALYSIS OF SAMPLES

      Samples  for analysis were  taken  from the  base  of the  heat  exchanger and
routed  through  a heated line  to a  series  of impingers.   The first impinger was
chilled  to   approximately  4°C  with an  ice-water  bath.   The second and third
impingers  were  cooled  to  about -50°C with  an  isopropyl  alcohol-dry ice bath.
The  temperature  of  the  effluent  leaving the  condensers was  sufficiently  low  to
assure  that  almost   all  of  the  water,  methanol, and dimethyl  ether  were con-
densed  out, and  the  essentially  dry  gas  was  collected  in  a Tedlar  bag.    This
sample  bag and the  liquid  samples were then  transported  to  the Gas  Chromato-
graph Laboratory for analysis.

      The  gas  chromatograph  (GC)  utilized a  15'  x 0.125"  S.S. column   packed
with  Parapak  Q  80  -  100  mesh,  and was equipped with  a Valco  six-port  gas
sampling   valve   and   a  thermoconductivity   detector.     Calibration  standards,
containing   measured   volumes  of  hydrogen,  nitrogen,   carbon monoxide,   carbon
dioxide,  methane  and  methanol,  were  prepared  using  a  Tracor-Atlas  10  liter gas-
blending cylinder.    These  standard blends  were injected into the GC; then  the
identified  peaks were  integrated,  and the  response  factors  for   each compound
were   calculated.    The   gas   samples  were  injected  and analyzed  in  the same
manner as  the standards.

      The  liquid  samples  were  analyzed  for  water,  DME  and  methanol.   The
volume of  the  water  was  determined  by  coulometry using a Photovolt  Titrator.
The volume of DME  was determined  by gas chromatography  using a  fused-silica,
open-tubular column.  The amount of  methanol was  determined  by  the difference
between these two compounds and  the total for all three  compounds.
The  results  of these  analyses  for  the tests  conducted  are  summarized  in  Table
5.   This  table also  includes  the  calculated value  for  the rate  of dissociation for
each test.
                                       H-13

-------
                          Table 5. Gas Chromatograph Analysis of Heat Exchanger Samples
I—>
-1^
 Date
 Time
 Test Code (Run Number)
 Hx Fuel Flow (Ibm/hr)
 Cylinder //I Exhaust Mass Flow (Ibm/hr)

 Air
 Amount Liquid Condensed (g)
 Volume Gas Collected (1)
 Sampling Time (minutes)
 H2/CO Mol Volume Ratio

 Gas Sample
 Air (N2/O2) Mol Volume Ratio
 Volume 96 MeOH
 Volume 96 H2
 Volume % CO
 Volume 96CH/,
 Volume 96 CO2

Condensate
 Weight % H2O
Weight % DME
Dissociation (%)
8/12/87
14:30
3
1.0
16.3
200
107.5
25
6.3
61.8/16.4
—
17
2.7
0.2
1.9
3.2
1.0
5
8/13/87
15:45
4
.5
17.11
93.4
142
30
3.2
44.7/11.6
1.9
30
9.5
0.3
2.0
3.9
3.1
22
8/17/87
11:30
5
2.0
17.36
215.5
13
13.7
3.5
57.4/14.6
4.3
16.6
4.7
0.4
2.0
.3
.4
1
8/25/87
16:18
6
0.02
17.26
o'
9
30
1.7
0.007/0.002
1.6
57.2
33.8
1.8
4.7
None
None
96
8/27/87
11:25
7
.95
17.90
71.3
4.3
30
3.7
3.6/0.2
17.5
54.8
14.9
2.2
6.8
2.8
3.5
7
9/4/87
12:15
8
6.3
46.3
42.2
14.0
15
2.3
3.9/0.2
8.1
60.0
25.7
1.1
1.1
0.98
0.4
11
9/4/87
14:05
9
1.1
17,20
200.0
25.0
30
2.1
5.1/1.1
1.7
61.0
28.7
0.7
1.3
0.83
N/A
5

-------
HEAT EXCHANGER

History

      The  basic  concept of a  counter-flow heat exchanger made up  of concentric
tubes  for  the reactor  was proposed  by  the  EPA.   The heat  exchanger includes
an  evaporator to vaporize  the  methanol  and a  reactor  to dissociate  the methanol
vapor.   SwRI   undertook  the   detailed  design  of  the  reactor  and  enlisted  the
assistance  of  Dr. R. Webb of  Pennsylvania State  University,  as recommended by
the  EPA,  to help   design  the   evaporator.    Fuel  is injected into the  evaporator
from  an  engine-driven,   specially  modified,   distributor-type  pump   built  by
Stanadyne.  Modifications  to  the  pump  included  nickel   plating the  interior and
using  ceramic rollers  in  the  precharge  section.   The  methanol  injection  nozzle,
custom designed and build by  SwRI,  differed from existing  production nozzles  in
that  it provided full  spray  impingement  on  the evaporator  walls  for  maximum
evaporation.    The   reactor substrate  and  the  methanol  and  exhaust  side  catalyst
coatings  were   recommended   and   provided   by   Johnson-Matthey.     Assembly
specifications  and   instrumentation to  monitor  the  heat  exchanger  performance
were defined by SwRI.   The  selection  of  the  catalysts  and  the basic  design and
dimensions were approved by  the EPA  in a July  31,  1985  letter  to  SwRI con-
cerning design specifications.

Current Design Consideration

      Several  design considerations  became  apparent during testing  of the  proto-
type heat exchanger. Selected considerations  are discussed as follows.

Methanol  Injector   -   The methanol  fuel  injector is  currently  located within the
exhaust stream  in  the  heat exchanger.    With  lower than  maximum  fuel flowrates
through  the  evaporator,  the  evaporator  temperature  could exceed 205°C (340°F).
When  the  internal  evaporator   temperature exceeded  205°C,  the  quality  of  the
spray   deteriorated.     The    primary   cause   of  deterioration  of   the   spray
characteris-  tics    was  a  reduction  in  the  injector  cracking  pressure  (cracking
pressure  was  monitored  continuously).    Reduction   in  cracking  pressure  was
apparently  caused   by  a   loss   in  spring  tension  at  the  elevated  temperatures.
Vapor  lock  can  occur  with  loss  in  cracking pressure.    The  seal in  the  top  of
the  injector  lost   its   elastic  properties  and  allowed  methanol to  leak  directly
into the exhaust stream.   The  methanol injector  should  be  relocated  outside  the
exhaust stream and incorporate provisions for additional cooling.

Evaporator   -    Significant temperature  instabilities  were noticed during  testing,
particularly  with  the  heat exchanger  insulated.    The   cause  appeared  to have
been the   thermal  lag  in  increasing and  decreasing  the  temperature  of  the  metal
in  the  heat  exchanger  as  conditions  changed.    Consideration needs  to  be  given
toward evaporator performance under transient engine operation.

       Evaporator  capacity,  defined  as  the  maximum   fuel  rate   that  can   be
injected  and  still  maintain the  internal  temperature  above  the  boiling point  of
methanol,  is  very  sensitive to  injector  spray  quality.    A  slight  drip  from  the
injector  can  reduce the  maximum  vaporization  rate  by 50  percent,  based  on
observations  during  testing.   The  degree  of  atomization  from  the  injector  is  a
function  of  the cracking pressure;  the  droplet  size  decreases  as  the  cracking
pressure is  increased.  Atomization  is  advantageous  when  injecting  fuel  into  a
combustion chamber  for  immediate  burning.   However,   in  this case  where it is
desirable  to   propel  the methanol  against  the  entire  surface of  the  wall  of  the
evaporator, the  optimum   droplet  size  is  that   which  allows  the  evaporator  to
                                          H-15

-------
operate at  maximum capacity under a  given exhaust  gas flowrate  and  tempera-
ture.    Moving  the  injector  outside  the  exhaust  stream,  where  it  would  be
exposed,  would  enable  adjustment  of  the cracking  pressure  while  observing the
evaporator  temperature.   This would  permit adjustment  of  the  cracking pressure
to achieve optimum droplet size.

     The  catalyst  used  in   this   reactor  to  dissociate  the  methanol  provided
unsatisfactory  yields  at low to  moderate  engine  power levels; these  power  levels
are considered  representative  of  typical  light-duty  engine use.   A  catalyst  with
higher  dissociation   effectiveness  at  lower  temperatures   is  needed  to  enable  a
reasonable reactor  size  and  mass.   The catalyst  furnished by  Johnson-Matthey  is
their  proprietary Type  H.    This  catalyst  has   a  maximum  dissociation  rate  at
370°C   (698°F),  with  significant  yields  at   300°C   (572°F).    In  the  evaluations
conducted,  the  average  methanol-side  reactor  temperatures  with  insulation   were
generally  at or  above  300°C.   At  lower loads  and  speeds,  such  as idle,  lower
temperatures  with  reduced   dissociation  can be  expected.    Higher  loads   and
speeds  will  generate temperatures  considerably   higher  than  the  370°C  optimum
temperature.

Exhaust  Back-Pressure   -    The  increase   in  exhaust back-pressure was  signi-
ficant,  being  about  1.6 inches of Hg  with the engine  operating  at  3000  rpm  and
40  Ibs-ft. This  increase  in  back-pressure was sufficient  to  increase the  exhaust
temperature at  the cylinder by about 80°C at this engine operating condition.

CONCLUSIONS  AND RECOMMENDATIONS

     Given  the  current  catalyst effectiveness,  it  appears  that  satisfactory   heat
exchanger performance  cannot be achieved with  a  design  that will fit  reasonably
under  the  hood of  a  vehicle.   Redesign  of the  existing  reactor  would require
that it be  increased in size  by a  factor of two  to  four  to  provide  minimum
satisfactory  dissociation  rates  at low engine  power levels.   Such  an increase  in
size would  result  in  increased  thermal  lag  and  could  require  an increase  in
evaporator  size  to  offset  lower  exhaust  temperatures exiting  from  the  larger
reactor.

     Given  the  low  dissociation  effectiveness at low  to  moderate  engine speeds
and loads,   it   is   doubtful   the  current  heat   exchanger   design   would  prove
satisfactory  under  transient  operation in a  vehicle.    Most  advantageous  would  be
a  dissociation  catalyst  that  is more  effective over  a  wider  temperature  range.
There is, however, no known source for such a catalyst.
                                           H-16

-------
           ADDITIONAL DA*m FKM EVACUATION OF MEIHANOL DISSOCIATION HEKT EXCHANGER DISSOdATOEP1
PARAMETER

Date
Time
 RUN 6
RUN 4
RUN 7
RUN 3
                                                                          RUN 9
RUN 5
                              RUN 8
08/25/87  08/13/87  08/27/87  08/12/87  09/04/87  08/17/87  09/04/87
1618      1545      1125      1430      1405      1130      1215
1500
30
12.8
56.3
20
17.3
88 lc
1500
30
11.7
56.7
20
17.1
813
1500
32
13.2
58.4
20
17.9
822
1500
30
12.7
52. 6b
20
16. 3b
791
1500
30
10. 8b
58.0
22
17.2
840C
1500
30
12.7
56.7
20
17.4
773
3000
40
28.2
157.0
28
46.3
1148C
ENGINE OPERATION

  Speed, rpm
  Torque, ft-lb
  Methanol Consumption, Ib/hr
  Intake Air, Ib/hr
  Spark Advance, °BTC
  Exhaust/Cylinder, Ib/hr
  Exhaust Temperature,  °F

AMBIENT

  Temperature,  "F                96         92        89        97        87        90        88
  Barometric Pressure,  in.Hg     29.1      29.0       29.3      29.0      29.2      29.2      29.2

DISSOdATOR

  Insulatedd                     YES        NO        NO        NO        YES       NO        YES
  Methanol Flow, Ib/hr           0.02      0.50      0.95       1.00      1.10     2.00       6.30
  Nozzle Pressure, psig         225       135       110        90        45       150        175

Exhaust Pressures, in.  Hg:
  Before Reactor                 0.3        0.1       0.3       0.0       0.2       0.1       1.1
  After Reactor                 0.2        0.0       0.2       0.0       0.1       0.0       0.5
  After Evaporator               0.0        0.0       0.0       0.0       0.0       0.0       0.0

Exhaust Temperature,  °F:
  Before Reactor
  After Reactor
  After Evaporator

Backpressure on Evaporator, psig 0.0

Methanol Temperature,  °F:
  In Evaporator
  After Reactor

 aData presented in order based on methanol flowrate into the dissociation.
 bData appears to be somewhat low.
 GApparent minor discrepancy in these data based on location in system.
 dAround the outside of the heat exchanger unit.
905°
908C
732
0.0
790
652
422
0.0
763
—
165
33
707
419
210
0.0
877C
721
244
0.0
658
17 lb
162
0.0
1167°
863
691
0.0
                                          392
                                          518k
                    176
                   198
                   666
                   352
                   788
                                                  162
                                                  576
                             345
                             1140

-------
        APPENDIX I
ENGINE MODIFICATIONS FOR A
DISSOCIATED PRODUCT VALVE

-------
      SUMMARY OF ENGINE MODIFICATIONS FOR OPERATION AND CONTROL
                         OF DISSOCIATED PRODUCT VALVE

I.    General

      The dissociated product valve for each cylinder will be actuated by  a  separate
additional camshaft that  will be  driven via a short chain from an additional  sprocket
attached to the existing camshaft sprocket.  The drive will be purely mechanical.  This
simplicity has the advantages of  providing reliable and repeatable valve operation.  It
has the  disadvantage of not allowing variation in valve opening duration which could be
used  to control  the flow of dissociated products  to  the  cylinder.   Approaches  using
solenoid operated valves  were considered, but  they were expensive, were still in  the
development stage  and their  full-time  control  would have added  substantially to  the
demands of the control system.

2.    Dissociated Product Inlet Valve Assembly

      In order to save project time and money and to reduce development risks, we
chose an existing valve assembly  for this purpose.  The valve is designed as  an exhaust
valve for a Honda motorcycle engine.  Given its design and proven street  performance
we feel that it will perform adequately as a  dissociated products  inlet  valve.   The
motorcycle exhaust valve is exposed to high accelerations given  the engine redline of
approximately  10,000 rpm, and  high temperature gas flows  over 1500°F-  The  valve
face  will be exposed  to  somewhat  lower combustion  chamber  temperatures in  the
NAPS-Z engine operating on methanol than those of the gasoline  fueled Honda engine.
Also, the dissociated product maximum temperature of about 600°F means that  the
dissociated product inlet  valve will be exposed  to lower gas flow  temperatures than it
would on the  engine  it is designed for.

      The stock  NAPS-Z  engine head is aluminum and it has steel valve seats pressed
into it to provide adequate durability. Our design provides for a "caged" valve assembly
similar  in design concept  to that  used on large  bore low speed diesel  engines and to the
third  valve  in  Mitsubishi  MCA jet engines.   Diesel engines  contain  this  feature to
minimize  the time  required to  perform valve  overhaul or replacement.   This  design
feature  was included in our design to provide a material suitable for hardening at  the
valve seat, to provide a  guide and mount  for  the valve  and  valve spring, and to port
dissociated products from the reactor outlet to the cylinder.  Providing  a  valve seat
that  adequately  resists  wear is  important  particularly with  hydrogen  and carbon
monoxide as  the primary  products being transmitted through this valve. Dry gases like
natural  gas  or propane,  and we suspect, hydrogen  and  carbon  monoxide, cause an
increase in valve wear over that experienced with unleaded gasoline.

3.    Pushrod

      A  short pushrod  is provided between the  overhead cam  and the  valve stem.  Its
primary purpose is to provide a mechanical link  between the cam surface and the  valve
stem.   If the valve chosen had a longer stem  it would not be necessary to have  the
pushrod. In fact, an extension could be friction welded to the valve stem to provide the
additional length needed.   However, the pushrod performs two other functions  which
make this concept  desirable.  Most importantly,  it accommodates side loads  induced
from  the  cam.   This arrangement minimizes  bending loads on  the  valve stem  and
friction  between the stem and the valve guide.  It also permits the cam follower to be
integrated  with  the pushrod so that  the valve  will accurately follow  the cam profile
(which  includes  concave  surfaces) and  so that  an adequately  wide  surface  can be
                                          1-2

-------
provided to reduce wear to acceptable levels.  Calculations covering contact loads are
discussed in paragraph 10.

*.   Overhead Cam

     The cam profile was  defined using a computer program written especially for that
purpose. A number of  general purpose  cam profile and  valve train analysis programs
are available  in  the  department, but they are not suitable for  the unique  cam  shape
required for this application.  The concave surfaces of the cam are  necessary given the
short duration of  valve lift, the small diameter cam  defined by the engine configura-
tion,  and  the  desire  to  keep  the valve  train  mass to  a minimum.   Calculations
summarizing the  cam profile design and the computer listing and output are attached in
Appendix IVB.

5.   Dissociated Product Valve Cam Chain Drive

     The NAPS-Z engine  manual that we purchased and the documentation you sent us
are adequate to define  the cam drive concept.  We do have a head that provides some
dimensions, but practical determination of details requires an engine complete with all
valve train components. A general concept using a chain drive between the existing and
new camshafts has been devised.  The  specifics  will be worked out after the engine is
received.

6.   Dissociated Product Valve Temperature and Pressure Characteristics

     The temperature environment for the valve is less stringent than that  which it is
designed for (exhaust).  The back of the valve will be exposed to the reactor pressure of
up to 100 psig, but it will be prevented from opening by sufficient preload on the valve
spring.  We had considered some pressure balanced valve  designs to  remove variation in
reactor pressure  as a consideration in  determining valve  spring rate, but the added
complexity, particularly with regard to sealing and lubrication of the balancing piston,
made this concept impractical.


7.   Valve Assembly Details

     Appendix J  summarizes the calculations for these.    Some  specific  values are
summarized below.

                     Parameter                                  Value
     Required valve lift                                        0.156 in.
     Design Cam Speed (@ 4500 engine rpm)                     2250 rpm

     [Minimum spring force required                               46 Ibf
     to seat valve against maximum
     (150 psig) reactor force

     Combined spring rate (as measured)                        330 Ibf/in.
     Combined maximum allowable spring compression            0.50 in.
                                                          (0.557  in. for outer)
                                                          (0.510  in. for inner)

     Spring force available at zero lift                          113.5 Ibf
                                                           ( 46 Ibf required)
                                             1-3

-------
     	Parameter	          	Value	
     Available spring force at maximum                         165 Ibf
     valve lift

     Valve assembly dynamic weight                           0.194 Ibm

     Maximum allowable cam generated                     328,639 in/sec^
     deceleration (spring controlled)

8.   Cam Profile

     The equations and the computer program used for developing the cam  profile are
listed in Appendix K along with the cam radius versus degrees.  A summary of critical
values  are as follows:

     Ramp-height = 0.034 in.
     Maximum acceleration between stations 2-3 = 217,309 in/sec^
     Maximum acceleration between stations 3-4 = 138,569 in/sec^
     Maximum valve displacement  = 0.177 in.
     Maximum valve displacement needs  to be between 0.156 inches minimum lift and
     0.190 inches maximum lift  (sum of ramp and required lift).

9.   Follower Radius

     The cam  profile listed in  Appendix  K   was drawn approximately 20 times larger
over the approximately 45° one-half lift  duration to provide a visual check that the
calculated  cam profile  was acceptable.  This large  scale drawing of the cam profile
also provided a useful tool for determining the maximum allowable  follower radius of
0.06 inches.

10.  Contact (Hertz) Stresses

     The contact stresses between the cam  surface and  the curved follower on the
pushrod were checked to see if  these were excessive. Maximum Hertz  stresses which
are a  result of both static  (spring)  and  dynamic  forces  were calculated  as follows
(Reference 2, pg. 87):

                                                   Maximum Stress
               Cam Width (in.)                           (psi)
                     0.5                                  231
                    0.625                                207
                    0.75                                 189

     Hertz stresses are  not a   problem  so cam and follower  design  will  primarily
consider wear as the  limiting constant.

11.  Follower Jumping Speed

     The speed at which the follower will no longer maintain contact with  the cam
surface is a function of the cam profile, camshaft  speed and natural frequency of the
valve gear.  Using the methods outlined in  Reference 1, Volume II, page 541, the natural
frequency of the  valve train was found to be  excited at 3868 crankshaft rpm. This is
well within the operating speed  range  of the engine.  A change to correct  this will be
made (one man-day effort).
                                           1-4

-------
12.  Valve Guide Material Considerations

     The preferred material for  the valve guide is cast iron.  It is usually chosen for
production applications because of its low cost and resistance  to wear.  In this case a
high carbon low-alloy tool  steel will be chosen because of ease  in machining, medium
strength and wear resistance.

13.  Valve Guide and Cam Follower Lubrication

     A thin  film of  oil must  be  established between the cam follower and the cam
surface and some oil must be  available  at the valve guides and pushrod guides.  Given
the conceptual design shown on our layout drawings it will not be a problem in  providing
oil to the pushrod guide.  Oil supplied to the cam and follower will drain down over the
pushrod guide and then to the  valve guide.  In conventional  engine design some  care  is
taken to assure that  the valve  guide is not covered with oil but does receive lubrication
from oil splashed in the rocker cover volume.  The reason  for avoiding an accumulation
of oil at  the  top of  intake valve guides is that the differential pressure  between the
rocker box and the intake port tends to  drive oil into the port and then to the cylinder
thus causing excessive  oil consumption and exhaust emissions.  This particular applica-
tion has a pressure differential between the dissociated product port and the rocker, but
the pressure on the valve side will most often be greater than that in the valve cover.
Nonetheless, oil  needs  to be directed at the dissociated  product valve  guide and not
allowed to submerge it.  An oil drain back to the engine  sump will be provided at the
base of the dissociated product valve assembly.

1*.  Valve Seat Material

     The dissociated product  valve seat will be incorporated in the  sleeve that is part
of the  valve assembly.  This assembly will be made up  outside the engine then inserted
in the  head and fastened in  place.  It is a "caged" valve concept, much like those used
on larger engines.  The sleeve material will be a  medium carbon, medium alloy, heat
treatable steel such  as SAE 3140 or 4140.  It will be  hardened after initial machining
and before final grinding to 40-46  Rc.

     The basic attributes are that the seat must resist pound-in, hot hardness, abrasion,
corrosion,  and creep  at high temperature.    Because of  assembly considerations, the
head material (aluminum alloy)  was never considered for a valve seat.  It  would be
poorly  suited for this purpose.  This particular application may be more severe  given the
essentially dry gas composition (CO and H2) from the  reactor.  However,  the less than
100% reactor  efficiency  will mean  that some vaporized methanol will be passing and
this will be to the benefit of the valve seat by reducing the tendency to wear.

     Thermal  expansion  and  thermal conductivity of the valve seat  are  of  some
concern also.   The application of similar alloy steels in valve seats in aluminum-based
heads is fairly common.  A nominal press fit of 0.003 in./in. diameter is suggested  in
SAE Recommended  Practice  3610b for cast iron or aluminum  heads.   Because the
dissociated product valve seat will  be  exposed  to  lower gas  temperatures than the
exhaust valve  seat and because  coolant passages  are about the same distance from
each, this basic approach  is considered satisfactory.

15.  References

15.1 Taylor, C. F., "The Internal Combustion Engine in Theory and Practice,"  Volumes I
     and II, The  MIT  Press, Revised Edition, 1985.                                   1
15.2 Shigley,   J.  E.  and Mitchell,  L.  D.,  "Mechanical  Engineering  Design,"  Fourth
     Edition,  1983.

                                           1-5

-------
         APPENDIX J




VALVE ASSEMBLY CALCULATIONS

-------
                     SUMMARY OF CONDITIONS FOR CAM  DESIGN
     Valve Lift
port -
          -  7TDL
     L  = D/4
                     opening
                                                   n
                                             D
                                D
                        0.75
                       0.625
                                        0.1875
                                        0.156*
     Cam Speed

     (= Vi engine speed)
                                             rpm
Eng Speed
Cam Speed
Cam Speed
6000
3000
5000 4500
2500 2250
rad/sec
^262 ^236
37.5*2* TT
      Minimum Spring Force Required (to seat valve with reactor pressure)
                      Valve Diameter (in.)
100 psig
150 psig
                        0.75
                       M Ibf
                         66
                                         0.625
                                         31  Ibf
                                          46
*.   Spring Rate

     (150 + 180 Ibf/in)  =  330 Ibf/in

5.   Allowable Compression

     Large:  0.557 inch
     Small:  0.510 inch
     Allow.:  0.5 inch

6.   Required  Lift

     L = 0.156 in. (see paragraph 1)

7.   Force Available with Zero Lift

     0.5 - 0.156  = 0.344  inch

     .344m(180 Ibf/in + 150 Ibf/in)  =  113.5 Ibf

     113.5 Ibf  > 66 Ibf  required minimum
                                            J-2

-------
8.   Maximum Force at Maximum Lift

     0.5 in. (180 Ibf/in. 4-150 Ibf/in.) =  165 Ibf

9.   Valve Assembly Weight

                                          grs             With fe Spring

           Valve                           26.0                26.0
           Two springs                    54.5                27.3
           Keeper and washer              9.9                  9.9
           Pushrod           (assume)      25                 25.0
                                                             88.2 grs

           88.2 grs x  2.205 Ibm =  0.194 Ibm
                      1000 grs

10.  Maximum Allowable Cam Deceleration

     F = ma, a =F =  386.4  in.x 165 Ibf
                m   0.194 lb-sec^

                =  328,639 in/sec2
                                           J-3

-------
           APPENDIX K




CAM PROFILE PROGRAM AND OUTPUT

-------
            CAM PROFILE FOR DISSOCIATED PRODUCT INLET VALVE
Cam profile symmetrical
   about this line

         0°
                       0.6  inch
                      m
                                    DEGREES
                                      0
                                      i
 4
 5
 6
 7
 C;
 9
 10
 1 1
 12
 13
 14
 15
 16
 17
 IS
 13
 28
 21
                                     24
                                     25
                                     26
                                     27
                                     2'!"'
                                     29
                                     30
                                     31
34
35
                                      7
                                     38
                                     33
                                     40
                                     41
                                     42
                                     43
                                     44
(INCHES)
RFID I US
  7 7 6 9 7
  7 7 4 Pi g
  7784
  76593
  7 6 9 8 "'
  75495
  74935
  74106
  73393
  72448
  71526
  76546
  6353
  68526
  67578
  6671 1
  65946
  65295
  6476
  64338
  64912
                         6 J -/V y
                         632
                         63622
                         62344
                         62667
  62483
  6231 1
  621 3 3
  61956
  61778
  616
  61422
  61244
  61067
  60889
  6071 1
  60533
  60356
  66178
  6
  6
  6
  f,
                                 K-2

-------
10  ' EPflCHf'1
20  * CflLCULRTE  CRM  PROFILE
30  ! G. ORISCOLL, JULY  1'3S5
40 K= 9983  !  DISP UNDER flCCEL
50 L=.156  !  IN
60 W=75*PI  '  RRD/SEC  -'3000 RPM>
r*^ -•"- H ^v — M
63
64
65
66
  70
  30
  90
 100
 110

 120
 130
 140
 150
 155

 170
 130

 190
 200
 210
 220
 230

 240
 250

 260
 270
 280
 290
 300
 310
 320
 330
 349
 350
 360
 379
 389
 399
 489
 419
 420
 430
 449
 450
469
470
480
 R7=0
 V5=9
 V6=0
 V7=190
 S6=0
 S5=0
 Tl=0
 D=l.'13500 !  TIME INCREMENT
 B=.6  !  BRSE  CIRCLE RflDIUS
 INTEGER I,J
 SHORT  fl<1009),V<1009>,3^1009
 >,H<1800>
 PRINT  "K=";K
   PRINT "L="J
              "IN-
   PRINT "W=";W;"RflO/SEC"
   PRINT
   PRINT
   RRD
           IN/SEC  RRMP  SPEED
           TO
       MB=";B;"IN RflDIUS"
       "D = ".;D; "SEC* INCREMENT1
W9=24  !
FOR J=l
! J=5
GCLEflR
IF J=2 THEN 299
IF J=3 THEN 359
IF j=4 THEN 419
SCflLE 9,  02,-599999,599999
 J=l
XflXIS 0,  002,0,.02
YflXIS  01,100999,-590900,590
999
MOVE  014,490900
LflBEL "flCCEL"
GOTO 469
SCflLE 9,  02,-250,259 !  J=2
XflXIS 0,.092,9,.02
YflXIS .01,59,-259,259
MOVE  014,299
LftBEL "VELOCITY"
GOTO 469
SCflLE 9,.02,0,.2  !  J=3
XflXIS 0,.892,9,.92
YflXIS .01,.92,9,.2
MOVE  914,.18
LflBEL "DISPL"
GOTO 468
SCflLE -1.2,1.2,-.9,.9 !  J=4
XflXIS 0,  2,-l,1
YflXIS 0,.2,-!,1
MOVE .5,.8
LflBEL "PROFILE"
MOVE 0,0
FOR T=9 TO   0134  STEP D
I=T-'-D
                                                              DWELL
                                                              RflMP
                                                            fl
                                                            V
490 T1=.77166706*PIXW
500 T2=7*PI-''8'"M
510 T3=15*P
515 T4=PI/W
520 T5=17*P
530 T6=9*PI,'3-'l-4
540 T7=0
550 T8=0
560 IF T=V
799 S=9
719 S,
    H*SIN
789 GOTO 1829
799 fl=8
899 R=fl
819 V1=V
820 V=V9
839 V=V
849 S1=S
859 S=V9*D+S1
869 S,
    H*SINCT*W>
939 S3=MflXCSl,S>
949 GOTO 1828
959 fll=fl ! fll=PREVIOUS  fl
968 fl= *W* < T-T2 > >
979 flx'2+Vl
1985 I  V=-<16.6751*K*L*W*COS<16.
                                                            fl
                                                            V
                                   K-3

-------
1887
1918
1829
1925

1838
1835
1940
1850
I960
1878
1980
1898
1188
1110

1128
1139
1135
1136
1140
1158
     V5=MflX'rV5, V)
     'MKI>=V
     S1=S
     !  S=K*L*SIN<16
                       T,fl
                       T,V

                       H*COS
     '  PRINT T.:S
     H=S+B
     H=H
     IF J=l THEM DRRW
     IF J=2 THEN DRflW
     IF J=3 THEN DRflW
     IF J=4 THEN DRflW
     ,H*SIN
     GOTO 1328
     fl=8
     S2=S
     V2=V
     fll=fl
     !  fl=-<69.5149*:»
1155 fl=-
 1157
 1160
 1178
 1180
 1190
 1195

 1208
 1218
 1220
 1239
 1240
 1245

 1258
 1260
 1270
 T288
 1290
 1300
 1318
 1329
 1330

 1340
 1358
 1368
 1378

 1380
 1390
 1480
1410
1428
1438
1448
1458
1468
1470
1430
     N<:T4xe:T6-T2>*W*»
     !  PRINT T;ft
     fl=fl
     fl7=MIN*L*SIN<4
     -T2»+S2
     S6=MflX=S
     !  PRINT  T;S
     H=S+B
     IF  J=l  THEN DRflW T,fl
     IF  J=2  THEN DRflW T,V
     IF  J=3  THEN DRflW T,S
     IF  J=4  THEN DRflW H*COSCT*U)
     ,H*SIN
     GOTO  1828
     ft=8
     fll=fl
     fl=278. 8596*K*L*WA2*SIN<: 16.6
     FKI>=ft
    S1=S
    S=S
     ! PRINT fl;V;S
    H=S+B
    H=H
    IF J=l THEN  DRflW  T,fl
1498 IF J=2 THEN DRflW  T,V
1500 IF J=3 THEN DRftW  T,S
1518 IF J=4 THEN DRflW  H*CQS
     ,H*SIN
1520 GOTO 1828
1530 fl=8
1548 fll=fl
1550 fl=0
1560 R=fl
1570 V1=V
1580 V=-V9
1598 V=V
1688 S1=S
1610 S=V*D+S1
1620 S=S
1638 H=S+B
1648 H=H
1658
1668
1678
1688
1698
1700
1710
1728
1738
1748
1758
1768
1778
1738
1798
1388
1818
! PRINT FhV
IF J=2 THEN
IF J=3 THEN
IF J=4 THEN
GOTO 1828
fl=0
fl=fl
v=8
V< I )=V
S=8
S< I >=S
H=S+B
H< I )=H
IF J=l THEN
IF J=2 THEN
IF J=3 THEN
IF J=4 THEM
;S
DRflW
DRflW
DRflW









DRflW
DRflW
DRflW
DRflW

T f V
T,S
H*COS









T,fl
T-. V
T,S
H*COS
I860 IF J=2 THEN DRflW T,-V
1878 IF J=3 THEN DRflW T..S
1888 IF J=4 THEN DRflW H(I>*COS"RflDIUS"
       FOR 1=188 TO 120 STEP -1
       PRINT 188-1,HCI)
       NEXT I
     NEXT I
     IF J<4 THEN 1948
     FOR Z=9 TO 2x75 STEP D
     DRflH B*COS,B*SIN = -
           11MflXV = "
           "«INV="
           "MflXS
-------
              APPENDIX L






METHANOL DISSOCIATION HEAT EXCHANGER




             DESIGN REVIEW

-------
SOUTHWEST   RESEARCH  INSTITUTE

POST OFFICE DRAWER 28510 • 6220 CULEBRA ROAD • SAN ANTONIO, TEXAS, USA 78284
                                     February 12, 1988

ENGINE AND VEHICLE RESEARCH DIVISION


Mr. Robert Bruetsch
United States Environmental Protection Agency
Motor Vehicle Testing Laboratory
2565 Plymouth Road  - ECTD
Ann Arbor, MI  48103

Subject:        Methanol  Dissociation  Heat  Exchanger  -   Design  Review   and
                Proposed  Improvements,   EPA   Contract  No.  68-03-1984,  SwRI
                Project No. 03-8352.

      The technical  design review  findings  and  proposed improvements  presented
are  primarily  based on observations of  the  preparer  since  he  became  involved
with  this  project  in  December of  last  year.   In  keeping  within project  con-
straints,  the  design  review  has been  limited.   The  proposed improvements  are
considered  to  be  incremental   and   should  not   be  misconstrued  as changes   that
will   assure  satisfactory  methanol  dissociation  performance.    Areas  of  technical
concern or improvement are discussed  as follows:

DISSOCIATION CATALYST  - The  base  metal catalyst  used in  the  initial  heat
exchanger  did  not  provide  satisfactory  dissociation.   A copy of  the  preliminary
results  of  the  catalyst  analyses  by  JM  is  included as  Attachment  A.   From  a
brief review  of the published  literature  and discussions  with individuals  at  JM,
EPA, and SwRI, it appears  that a  fully appropriate dissociation  catalyst  compo-
sition  has  not  yet  been  identified.   The writer  has  been  unable   to  find  a
catalyst  composition  that  will  provide  the   essential  characteristics   of   high
dissociation   efficiency   at   low   temperature,   durability  at  somewhat   higher
temperatures,   and  negligible   coking characteristics.    The   primary  requirement
for  developing  a methanol  dissociation  heat exchanger  is  the identification  of  a
suitable dissociation catalyst composition.

VERIFICATION OF CATALYST EFFICIENCY - Until  sufficient  operating history
has  been developed  on a specific  catalyst composition, it  is  recommended  that  a
sample  from  each batch  provided  be analyzed  to  assure it  meets  design  specifi-
cations.   At minimum, such  analyses  should  include  determination  of  effective
surface  area and bench determination of dissociation efficiency.

CATALYST  RECOMMENDATION -  The representative  from Johnson Matthey,  like
the   writer,  feels   that   reconsideration  of   catalyst   composition   and  bench
evaluation  of  samples  of the  actual batch  to  be  used  are   essential.   From  the
first  contact  in  December  of  last  year,  the   JM  representative  has  emphasized
their  current  overload  with   high  priority  programs.   Johnson   Matthey   will  be
unable  to  schedule  review,   formulation,  fabrication,   and  bench  evaluations  of
methanol  dissociation   catalysts  for several  months.    Total   time   to  obtain
catalysts  substrates   is  estimated  to be  six  months.    If   a  currently  applied
catalyst  composition, such  as   platinum,  is  used  and  no  bench evaluations  are
required, it  was indicated that such substrates  could likely be provided  within  a
couple  of months.   As previously  discussed, however,  the writer  has been  unable
to  find  any  reasonable  basis  on  which  to  recommend  any  specific   catalyst
composition.
                                           L-2

-------
EVAPORATOR  -  The  present  evaporator  involves  a closed  cylinder  wrapped  with
corrugated  metal  catalyst  substrate.    Because  the  engine  exhaust  gases  first
passes  through  the  exhaust  catalyst  in   the  dissociation reactor,  which  should
oxidize  essentially  all  available  HC  and   CO,  the  catalyst  wrapped  around  the
evaporator  appears  to  serve  no  useful   function.    Additionally,  the  corrugated
catalyst  substrate  effectively isolates  most  of  the  exhaust flow  away  from  the
evaporator.    By  removing  the  catalyst substrate  from  around  the  evaporator  and
installing  longitudinal  fins  onto  the  outer  surface,   heat   transfer  from   the
exhaust   to   the   evaporator   should  increase  significantly.     Another   potential
improvement  is  to have  the cone  in the  evaporator  open  to  the  exhaust,  rather
than closed  as in the present design.

DISSOCIATION REACTOR -  One  design  improvement involves  the  exhaust  flow
through  the  reactor  section  of the  methanol  dissociation  heat  exchanger.    With
reference  to  the  original  design  of the  reactor   shown  in  Attachment  B,   the
center section  of the  exhaust  catalyst  was  designed to  carry  the  same exhaust
heat  per circumferential  unit  as the two  outer  sections  of  the exhaust  catalyst.
Since  the  center  sections   provides  heat   to  methanol  catalysts  on  both  radial
sides,  it should  be  designed  to  carry  more  exhaust  heat.    Incorporating  this
design   improvement  involves   only  minor  redesign.     Another   recommended
improvement  is  to  increase  the  methanol  dissociation  annuli  from  two  to three
as  illustrated  in  the  revised  design  shown  in  Attachment  B.    A  third  improve-
ment  would  be  to  increase  the  void  space  at  the  entrance  and   exit  of   the
dissociation annuli  and  to  provide  four  exit ports from the  annuli, rather  than
the two as in the initial design.

INJECTOR DESIGN  AND  LOCATION -  The  commercially available injector  that
can  be  utilized  with  only  minor modification is  illustrated in Attachments C  and
D.   This configuration is considered to  be  satisfactory and  about as good as  can
be  attained  without  major  redesign  effort.   An  essential design  modification  is
relocation  of  the  methanol  injector out  of the  exhaust  stream.    Relocation  is
shown conceptually in Attachment E.

INTERNAL INSULATION  -  Inclusion of  internal insulation,  such as 3M  Interam,
around  the  outer diameter of  the  dissociation  reactor  would  provide  for   good
conservation of  heat,  and  it would also  provide  for  a  much  more  professional
appearance  than  does  insulation  installed  around  the  overall  heat  exchanger
assembly.   The  disadvantage  of  internal  insulation is  the  resultant  17  percent
loss  of  potentially  useable  cylindrical area.  That  loss in area can  be  made up
by  addition of  the  smaller  annulus  in  the  center  of  the  dissociation  reactor  as
illustrated   by  the  revised  design   shown  in  Attachment  B.    Adding  internal
insulation  in  the area  around  the  evaporator will   result  in  the necessity to  also
reduce the diameter of the evaporator,  as illustrated in Attachment D.

      A  possible  solution  to  the  current  size  limitation  of  3.75  inches overall
outside diameter  on  the  heat  exchanger  would be to  make  the  heat exchangers
elliptical.     Without  extensive  design   and  fabrication   efforts,   however,  an
elliptical  shape   would  be  impractical.   Fabrication  and  assembly  of  the current
                                        L-3

-------
cylindrical  design  are extremely difficult,  and an  elliptical design would  increase
the fabrication difficulty severalfold.

HEAT EXCHANGER ASSEMBLY  -  It  is proposed  to  fabricate  two separate two-
unit assemblies.   Attachment  of each  two-unit  assembly  to  the  Nissan NAPS-Z
engine  will  make   use  of  existing  exhaust  manifold   attachment  bolt  holes.
Attachment  of  four  separated  individual  units  to  the  engine   appears   to  be
impractical,   and  fabrication   of   a  single  four-unit  assembly   is  considered
significantly more difficult.

EXPECTED  RESULT OF  DESIGN  IMPROVEMENT  -   The  mechanical   design
improvements  described  should improve  the operation  and dissociation  efficiency
of  the  heat  exchanger.    These  improvements,  however, are  not  expected  to
provide  anywhere  near the order  of magnitude improvement  desired.   It appears
that selection of  the catalyst composition  is the  most  important  criteria  toward
meeting the desired dissociation efficiency.

Prepared by:                              Submitted by:
Charles M. Urban                         Gary L. Stecklein
Staff Engineer                            Director
Engine Systems Research & Development    Vehicle Systems Research Department

/ew

Attachments
                                        L-4

-------
                              ATTACHMENT  A

                ANALYSIS  OF CATALYST BY JOHNSON MATTHEY


FAX TO:       SOUTHWEST RESEARCH DESTITUTE
FRCM:         R.MCDCWELL    
DATE:         4 JAN 88       K*7  "^
SUBJECT:      PRELIMINARY PESUIITS - METHANDL REACTOR


ATTN:         MR. C. URBAN             FAX #  (512) 684-7523

Both the MeCH dissociation and the engine exhaust catalyst have been
checksd by our Analytical Lab for B.E.T.  surface area.  These catalysts
were also scanned by xray for the presence  of  catalyst poisons.  A coke
deposit discovered at the product cutlet  (Figure 9 iii the SwKE report)  was
analyzed by XRF for contaminants.
The surface area (S.A. )  measured for the rHaiaofiirfcinn catalyst was less
than one mater squared per gram (m2/g) .   The corresponding result for the
engine exhaust catalyst was 1.2 m2/g. Elements  found on the dissociation
catalyst were Al, Or, Fe, and Zn.   Elements found on the exhaust catalyst
were Al, ca, Cr, Cu, Fe, Ft, and Zn.  Elements found in the coke deposit
were Ca, Cr, Fa, H>, and Zn.  The high temperature resistant stainless
steel metal support is an alloy containing Al, Cr, and Fe which could be
picked up by the xray.

The expected B.E.T. surface area for a metal supported catalyst would be
soMWhere on the order of 8 to 13 m2/g,  so the low result on both
catalysts is surprising  Usual reasons for loss  of S.A. are poor washcoat
adhesion, thermal sintering, or masking/poisoning. A scan with a 20x lens
did not reveal bare patches of metal which would indicate washcoat
adhesion problems.  The 600 degree C. temperatures noted in the SWRI
report are not as high as we see in automotive applications and would not
seem to be a prime candidate for low S.A. numbers.  Welding the catalyst
reactor may be a source of high temperatures ( and metal poisons ) .
Masking or poisoning effects are a definite possibility.  The coke deposit
scanned by XRF shows the presence of ca and Ifc as well as the expected Cr,
Fe, and Zn.  However, the amount of each of these elements is unknown.
Each could be a poison in large amounts on the catalyst surface.  The xray
scan of the catalyst is not sensitive to the lighter elements, so a
masking agent of carbon compounds would not be detected.  The very fact
that a coke deposit was p»'-**iraH at the exit port of the reactor may
indicate some other undesired catalytic reactions going on, perhaps caused
by these other
A meeting to Ai*e*ta* the problem and your potential need for four
additional prototypes will be held early this week.  Dr. Paul Ellgen or
Jay Hoffman will phone you to let you knew where we stand.

                      Johnson Matthey
                   CATALYTIC SYSTEMS DIVISION
                     436 DEVON PARK DRIVE. WAYNE, PA
                                       L-5

-------
                                ATTACHMENT B

                        METHANOL DISSOCIATION REACTOR
         EXHAUST
         CATALYST
                           ORIGINAL DESIGN
METHANOL
CATALYST
INSULATION
(OPTIONAL)
                            1IETHANOL
                            CATALYST
                           EXHAUST
                           CATALYST
                              REVISED DESIGN

                                   L-6

-------
                                        METHANOL  FUEL  INJECTOR
                             Differential Area
                                        VWAAAA/VS/	1
                                     Max Valve
          Pintle
                  Valve Seat
      Lift (Preset)
                                            Nozzle Valve Closed
   SEE BODY
MODIFICRTION DETHIL
     HDD PIECE 70 VflLVE
    SHRGE OVER SURFHCE 'B'
flND  MHCH1NE TO SHRPE R5 INDICRTED
                                 SURFHCE  'B
                            BLEND PIECE TO VflLVE SEflT
                            PIECE SHOULD NOT OVERLHP SEflT CONTHCT flREH

                        BODY
                                                         i?0 REFERENCE - RCTURL DIHMETER SHOULD BE n RESULT
                                                       OF «2 INCLUSIVE HNGLE  TO EDGE OF VHLVE SEHT CONTflCT HREH
                                                             VRLVE
                                                                        CUT OFF BODY  PORTION flS SHOHN
                                                                           GRIND TO UNIFORM RRDIUS  TRNGENT
                                                                           TO VflLVE SEflT ON INSIDE
                                                                           flND TO 112' LINE ON OUTSIDE
                                                                           HITHOUT CUTTING VRLVE SEflT
                         STANADYNE MODEL  14 POPPET
                         NOZZLE  INJECTOR
                                             VHLVE SEHT  CONTHCT HREH

                                     BODY MODIFICflTION  DETRIL
                                                      L-7

-------
          ATTACHMENT D




METHANOL INJECTOR SPRAY PATTERN



[



1 f


2"
i
f


















'






i
1
1
i
/
I >
1 /
'' '

/
I >
i /
, /
/
. i
1 /
1 /
i /

1'
/
j '
I /

I/
1
i\
I 1
| 2
WITH
. 1 „



—
—
—

i i
Lj
A.
t V
/\
\
\
1 \
1 \



















.5" ID

MODIFIED
STANADYNE
/ MODEL 14
^ IMECT°S


i
,
I
\ |
\ '
\ .
\
\
1 X
\
\ '
\
\
\ l
\
\ 1
\ '
1

1 1
\
\ I
\
I t
\(
ii
\
\
l\
,\
i
_J
OPTIONAL ' '
INSULATION
-« 	 3
' ID
-W.


t f


1.7"
i






'


4.4"











^











'







                 L-8

-------
      RELOCATION  OF METHANOL  INJECTOR
                         EXHAUST
                         OUTLET
         EVAPORATOR
         netocoupLE
                                    LOUD rCTHANCL HET
                5
                   ORIGINAL DESIGN
EXHAUST

OUTLET
                     UOUDfCTHANO. (LET
                     EVAPORATOR .
                     nemocoudE
HOOFED STANAOrM! M
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-460/3-88-OOZ
2.
4. TITLE AND SUBTITLE
NEAR-CYLINDER DISSOCIATION OF METHANOL
FOR AUTOMOTIVE APPLICATION
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS


12. SPONSORING AGENCY NAME AND ADDRESS


3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
July 1988
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-03-1984
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report describes the design analyses and experimental evaluation
toward developing a system of dissociating methanol for use in a spark-ignited
passenger car engine. Maximum dissociation of methanol fuel attainable using
exhaust gas heat was the basic goal of this project. The heating value of
dissociated methanol is about 20 percent higher than that for liquid methanol.
Therefore, significant energy savings appear to be attainable if a large percent
of the engine fuel needs can be dissociated by use of the otherwise wasted heat
of the exhaust gas. A heat exchanger consisting of a methanol evaporator, an
injection system to provide fuel to the evaporator, and a dissociation catalytic
reactor was designed, fabricated and tested. The dissociation efficiency of this
initial reactor was not as good as expected. The relative poor performance of
the dissociation reactor apparently was associated with fuel injector nozzle
heating, blockage of passageways in the dissociation reactor, and an ineffective
dissociation catalyst. The heat exchanger was redesigned to incorporate im-
provements, and four heat exchangers were fabricated using the final design.
17.
a. DESCRIPTORS
KEY WORDS AND DOCUMENT ANALYSIS
b.lDENTIFIERS/OP6N ENDED TERMS
Methanol Methanol Fueled Vehicles
Methanol Exhaust Emissions Light-Duty Vehicles
Motor Vehicles
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified

c. COSATI Field/Gioup

21. NO. OF PAGES
182
22. PRICE
EPA Fprm 2220-1 (9-73)

-------