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)
------- |