&EPA
United States
Environmental Protection
Agency
Motor Vehicle Emission Lab
2565 Plymouth Road
Ann Arbor, Michigan 48105
EPA-460/3-81-035
December 1981
Air
Sulfate Control Technology
Vehicle Testing
-------
SULFATE CONTROL TECHNOLOGY VEHICLE TESTING
by
Dennis F. Hess
Mary H. Keirns
Kenneth C. Bachman
Exxon Research and Engineering Company
Products Research Division
P.O. Box 51
Linden, New Jersey 07036
EPA Contract No. 68-03-2342
Project Officer
Robert Wagner
Environmental Protection
Office of Air, Noise, and Radiation
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
2565 Plymouth Road
Ann Arbor, Michigan 48105
This report is the Final Report and is submitted in fulfillment
of Contract No. 68-03-2342 under the sponsorship of the U.S.
Environmental Protection Agency.
December 1, 1981
-------
DISCLAIMER
This report has been reviewed by the Office of Mobile Source
Air Pollution Control, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
-------
TABLE OF CONTENTS
Page No.
1. INTRODUCTION 1
2. OVERVIEW 4
3. CONCLUSIONS 12
4. RECOMMENDATIONS 14
5. LOW EXCESS AIR VEHICLE - DESIGN #1 15
5.1 Design and Fabrication 15
5.2 Test Results 27
6. THREE-WAY CATALYST VEHICLE - DESIGN #1 33
6.1 Design and Fabrication 33
6.2 Test Results 36
7. THREE-WAY CATALYST VEHICLE - DESIGN #2 44
-.7.1 Design and Fabrication 44
7.2 Test Results 44
8. LOW EXCESS AIR VEHICLE - DESIGN #2 60
8.1 Design and Fabrication v 60
8.2 Test Results 66
9. REFERENCES 78
-------
LIST OF TABLES
Table No.
2-1
5-1
5-2
5-3
5-4
5-5
6-1
6-2
6-3
7-1
7-2
7-3
7-4
8-1
8-2
8-3
8-4
Title
Summary of Vehicles Tested and Contract
Tasks Completed
Effect of Main-Jet Size on A/F Ratio
Catalyst Specifications
EGR Distribution (at 48 kph: 30 mph)
Low Excess Air Vehicle Exhaust Emission
Levels
Exxon Research Carburetor Temperature
Instability
Three-Way Catalyst Volvo Exhaust Emissions -
Preliminary Test
Three-Way Catalyst Volvo Exhaust Emissions -
0 km. Test
Summary of Kilometer Accumulation Data for
Three-Way Catalyst Equipped Volvo
Three-Way Catalyst Equipped Volvo - Design
#2 - Preliminary (0 km.) Exhaust Emissions
Maintenance Performed to Three-Way Catalyst
Volvo - Design #2
Summary of Kilometer Accumulation Data for
Three-Way Catalyst Volvo - Design #2
FTP CO Emissions: Three-Way Catalyst
Vehicle - Design #2
Reference Voltage Potentiometer Setting vs.
Air-Fuel Ratio
Air-Fuel Ratio vs. Vehicle Speed
FTP Exhaust Emissions
FTP Exhaust Emissions - Recalibrated Cold
Page No.
11
16
25
28
30
31
36
39
41
44
49
50
57
66
66
67
68
Start Circuit
8-5 FTP Exhaust Emissions - Englehard PTX-514 68
Monolith
-------
LIST OF TABLES (CQNT'D)
Table No. Title Page
8-6
8-7
8-8
8-9
FTP Exhaust Emissions - Engelhard PTX-514
Monolith - Increased EGR Flowrate with
Spark Port EGR
FTP Exhaust Emissions
Monolith
FTP Exhaust Emissions
Durability Interval
FTP Exhaust Emissions
- Englehard PTX-516
- After 2000 km
- Recalibrated
69
70
70
75
Carburetor
-------
LIST OF FIGURES
Figure No.
5-1
5-2
5-3
5-4
5-5
5-6
5-7
5-8
5-9
6-1
6-2
7-1
7-2
8-1
8-2
8-3
8-4
8-5
8-6
8-7
8-8
Title
SEFE System Pictorial Diagram
Edelbrock Aluminum Manifold - Before
Modification
Intake Manifold Exhaust Crossover and
Plenum Chamber
SEFE Auxiliary Exhaust Pipe
Edelbrock Aluminum Manifold - After SEFE
Modification
SEFE Heated Plenum - Connection to Auxiliary
Exhaust Passage
Exhaust Gas Flow Through SEFE Modified Intake
Manifold
EGR Entrance Port Location
Modified EGR Inlet
Bosch Feedback Fuel Injection System
Pictorial Diagram
"Action Circle" of Feedback Controlled
Fuel Injection System's Logic
Volvo Lambda-Sond Feedback Control Fuel
Injection System
Regulated Emissions vs. Accumulated Kilometers -
Three-Way Catalyst Vehicle Design No. 2
Hoi ley Feedback Carburetor System Pictorial
Diagram
Hoi ley Feedback Carburetor System Operational
Logic
Hoi ley Feedback Carburetor and Vacuum Control
Regulator Valve
Oxygen Sensor Voltage Response Curve
Original Carburetor Airbox Curve
Defective Carburetor Airbox Curve
Recalibrated Carburetor Airbox Curve
CO Emission Stability at Various Vehicle
Page No.
17
19
20
21
22
23
24
26
29
35
37
45
58
61
62
63
65
72
73
74
76
Speeds
-------
LIST OF ABBREVIATIONS AND SYMBOLS
Abbreviation/Symbol
FTP
SET
HWFET
A/F Ratio
SEFE
EGR
ECU
TWC
EFE
VCRV
Description Page No.
1975 Federal Test Procedure 3 Bag
CVS Cold-Hot
Sulfate Emission Test
Highway Fuel Economy Test
Air to Fuel Ratio: Mixture Ratio
in LBS of Air/LB of Fuel
Super Early Fuel Evaporation System
Exhaust Gas Recycle
Electronic Control Unit
Three-Way Catalyst
Early Fuel Evaporation
Vacuum Control Regulator Valve
2
2
2
4
4
4
6
6
16
64
-------
SECTION 1
INTRODUCTION
Research described in this report was conducted under EPA
Contract No. 68-02-2342, entitled "Sulfate Control Technology Vehicle
Testing". The Contract work period spanned from December 1975 to
January 1978. The advent of oxidation catalyst emission control systems
on 1975 vehicles to control carbon monoxide (CO) and hydrocarbon (HC)
emissions raised concern over the formation and emission of sulfates
(aerosol sulfuric acid particulates - reference 1). During combustion
in the engine most of the gasoline sulfur is oxidized to sulfur dioxide
(S02). In a non-catalyst equipped vehicle the S02 passes through the
exhaust system and into the atmosphere undergoing little further oxidation.
However, in an oxidation catalyst equipped vehicle the S02 is further
oxidized to $03 over the catalyst along with CO and HC. The resultant
sulfur trioxide ($03) under exhaust conditions combines with water vapor
present from the combustion process to form sulfate (S04=) which hydrates
to form sulfuric acid. Research programs to determine automotive sulfate
emission rates, atmospheric dispersion of sulfates, and human response to
sulfates were initiated when public health concerns over sul fate exposure
were raised.
When this work was initiated, laboratory studies (2) had found
that the dominant factor affecting the rate of sulfate formation (S02
oxidation) over oxidation catalysts was the level of oxygen concentration,
or, oxygen partial pressure in the feed gas. As the level of oxygen in
the catalyst's feed gas was decreased CO and HC oxidation rates fell off,
but not as severely as the rate of S02 oxidation. At low oxygen partial
pressures a "break-point" was found where there was only a small debit in
CO and HC conversion efficiencies; yet S02 oxidation, and therefore
sulfate formation, was almost completely suppressed. Thus, by proper
choice of the oxygen partial pressure over an oxidation catalyst, sulfate
production could be controlled to low levels without severely affecting
CO or HC oxidation.
In 1975 commercial automotive catalysts were oxidation catalysts,
which oxidized CO and HC, while leaving NOX unaffected. Prototype
"three-way" catalyst preparations that could simultaneously oxidize and
reduce all three (CO, HC and NOX) automotive gaseous emissions were just
becoming available for test work. To operate in this manner, however,
the catalyst required that the feed gas proportions of CO, HC, NOX, and
oxygen be very close to a stoichiometric A/F mixture. Because this
narrow operating window results in a minimum of excess oxygen over the
catalyst, three-way catalysts were also expected to have low sulfate
emission levels.
In this contract the design, hardware development, emissions
characterization, and long-term durability testing of low sulfate
emissions control systems were the research goals. Contract work was
divided into three separate tasks:
-------
- 2 -
Task I - The design of two catalyst systems which minimized sulfate
emissions while maintaining low levels of CO, HC, and NOX emissions.
The first system was to be an oxidation catalyst system making use
of the low oxygen partial pressure concept to limit sulfate emissions;
while the second was to be a three-way catalyst design for simultaneous
control of CO, HC, and NOX emissions. Systems were to be designed
for 80,000 kilometer emission targets of 0.25 g/km HC, 2.1 g/km CO,
and 1.2 g/km NOX on the 1975 CVS COLD-HOT START TEST (FTP). To
insure achievement of the 80,000 km emissions limits, 0 km target
emission levels were to be 0.12 g/km HC, 1.0 g/km, and 0.6 g/km NOX
(1.2 g/km NOX if control is by exhaust gas recycle), as measured on
the FTP. The sulfate emission target was 6.21 mg/km at 96 km/hr
steady state operation. Steady state conditions were chosen for
the sulfate testing, rather than the FTP, to minimize the catalyst's
storage/release effect on sulfate. This target level corresponds
roughly to the sulfate emissions of a non-catalyst equipped vehicle
operating under the same cruise conditions (3).
Task 2 - The modification of two vehicles to incorporate the com-
ponents designed in Task 1. At least one of the vehicles was to be
of intermediate size with a V-8 engine. Once assembled, the vehicles
emission control systems were to be optimized, so that sulfate and
gaseous emissions would meet the specified target levels.
Task 3 - Measure the sulfate, sulfur dioxide, and gaseous emissions
of the vehicles at specified intervals over a total accumulation of
80,000 kilometers. Both vehicles would be tested at accumulation
points 0 km, 2,000 km, 6,000 km, 30,000 km, 60,000, and 80,000 km.
Emissions were to be measured during the test sequence of:
FTP (Federal Test Procedure)
20 Minute Idle Period
2-SET'S (Sulfate Emission Test)
HWFET (Highway Fuel Economy Test)
2-SET'S
2 Hours - 96 km/hr Cruise
The time variation of sulfate emissions, deterioration of catalyst
CO, HC, and NOX conversion efficiencies, and emission control
system mechanical hardware durability were among the parameters of
importance that would be determined from kilometer accumulation.
An eight month time period was allotted for performance of
these contract tasks. As will be discussed in more detail in the
report, hardware failures and design inadequacies in the original
vehicles chosen necessitated a contract modification (68-03-2342
Modification No. 1) at the expiration of the contract time period. All
work on the two original vehicles/emission control systems were stopped,
two new vehicles/emission control systems were chosen (Task 1), and
their assembly and testing (Tasks 2 and 3) were to be completed within
an eight month contract extension. In addition, the modification added
hydrogen cyanide (HCN) and ammonia (NHs) to the set of emissions to be
-------
- 3 -
measured during the testing sequence. Hydrogen cyanide and ammonia
emissions, like sulfates, are unregulated emissions. The sampling
and analytical techniques for determining the tailpipe levels of these
unregulated pollutants are discussed in detail in References (4) and
(5). Characterization of vehicle HCN and NH3 emission magnitudes and
rates, although not of prime concern to the contract goals, was added so
that preliminary data could be quickly accumulated by the EPA.
-------
- 4 -
SECTION 2
OVERVIEW
The purpose of this section is to summarize in concise form
the work performed and events that occurred during the contract time
period, as well as to establish the "contract tasks" that were completed.
A detailed treatment of the information contained in this overview is
presented in individual sections that comprise the main body of the
report.
Two catalyst system designs were originally chosen to fulfill
the requirements of the contract. They were: (1) a "low excess air"
vehicle equipped with V-8 engine and oxidation catalyst and (2) a
three-way catalyst equipped vehicle with a closed-loop electronically
controlled fuel metering system with an oxygen sensor. The low excess
air vehicle, by proper control of engine air-fuel ratio (A/F), would
maintain the oxidation catalyst at the low oxygen partial pressures
necessary for the minimization of sulfate formation. Thus, the vehicle
would use the oxidation catalyst for control of CO and HC emissions.
By controlling the excess air level over the catalyst, sulfate emissions
would be comparable to that of a non-catalyst vehicle. The three-way
catalyst equipped vehicle would also require precise control of engine
A/F ratio so that the three-way catalyst could be maintained within its'
operating window. A prototype system for electronic control of engine
A/F ratio developed by the Robert Bosch Corporation was chosen for A/F
control on this vehicle. The Bosch fuel injection system (K-JETRONIC) was
modified with an oxygen sensor and closed loop feedback control to
modulate the amount of fuel injected to the engine; and thus it could
maintain a set A/F ratio value, even during transient driving modes.
This system made it possible to use a three-way catalyst on a vehicle to
simultaneously control CO, HC and NOX emissions. The requirements of
Tasks 1 and 2 were fulfilled by these two designs. One made use of the
low oxygen partial pressure effect, and also included a V-8 engine,
while the other incorporated a three-way catalyst into a prototype
vehicle system.
Once conceptual design of the systems and their expected
emissions were established, hardware development and optimization on the
vehicles began.
2.1 LOW EXCESS AIR VEHICLE
A 1975 Chevrolet Malibu equipped with a 350 cubic inch displace-
ment V-8 engine was chosen as the vehicle to be modified. The low
excess air system that was installed consisted of: a "flat" air-fuel
ratio (A/F) carburetor, super early fuel evaporation system (SEFE),
close-coupled (warm-up) and main oxidation catalysts, and backpressure
modulated (proportional) exhaust gas recirculation (EGR). Carburetion
was jetted to provide the proper "slightly lean" A/F necessary to obtain
-------
- 5 -
low excess air conditions over the catalysts. In addition, the carburetor
was modified to give constant A/F over the full range of engine air
demand. This "flatness" of A/F insured that the proper catalyst environ-
ment would be maintained regardless of driving mode. To meet the low
gaseous emission levels required by the contract, the system design
required extremely rapid opening of the carburetor choke. The removal
of A/F enrichment due to rapid choke pull off coupled with an already
fuel lean mixture was expected to create serious cold driveability
problems. To offset this, a SEFE system was included in the vehicle
design. Its' function was to divert all of the hot engine exhaust gas
through passages in the intake manifold. The heat supplied to the
manifold runners and the carburetor base was expected to prevent fuel
condensation, and enchance cold driveability. Construction of the SEFE
system involved extensive modifications to the intake manifold. To
ease fabrication (welding and machining), an aftermarket aluminum
manifold was substituted for the production cast iron unit.
Once the low excess air system hardware was installed, optimi-
zation of the vehicle began. Poor driveability was experienced upon
cold start. The vehicle repeatedly surged, backfired, and stalled
during the first (cold) bag of the FTP cycle. The poor cold driveability
suggested that the SEFE system was not operating properly. Many problems
were identified, for instance, SEFE (Diverter) heat riser valves that
did not seal, a defective EGR valve and maldistribution of EGR flow
within the intake manifold. All these problems were resolved success-
fully, yet the vehicle still did not exhibit good cold driveability.
Even with the choke set rich to prevent stalls, severe hesitation and
sluggish accelerations persisted during warm-up. Exhaust emissions
control was also poor. FTP emissions were well above the contract 0 km
targets, in fact, they exceeded the 80,000 km limits. The lowest FTP
emissions observed were with the SEFE and starter oxidation catalyst out
of the system and normal choke opening during cold start operation.
However, even under these conditions, emission levels exceeded 0 km
targets.
It had become apparent that the vehicles' SEFE system impared
cold start driveability, rather than enhancing it. It was suspected
that the aluminum manifold conducted too much heat away from the carburetor
base, and fuel condensation was causing the repeated stalls. The hesitation
and sluggish acceleration performance were also attributed to the SEFE
system; since with it disabled, the vehicle performed normally. During
SEFE operation the entire engine's exhaust gas was diverted through
passages of about one fifth the crossectional area of the vehicles
exhaust system. This flow restriction could cause high exhaust back-
pressures during accelerations, and result in sluggish performance.
At this point work on the low excess air vehicle was stopped.
A new design, or considerable modifications to the original vehicle were
required to continue testing. Discussions were held with the EPA to
assess alternative low excess air design strategies. Thus Task 1, the
-------
- 6 -
conceptual system design, and Task 2, vehicle modification to incorporate
the design, were considered to be completed on this vehicle. Task 3,
the 80,000 km emissions and durability testing, was never begun due to
the high emissions levels and poor driveability experienced during
preliminary FTP testing.
2.2 THREE-WAY CATALYST VEHICLE
A 1975 Volvo equipped with a fuel injected 2.0 litre L-4
engine was chosen for modification. Since both the production Volvo,
and the advanced feedback controlled fuel injection systems were built
by Bosch, little hardware design and fabrication was expected in the
fuel metering system modifications needed to incorporate a three-way
catalyst. The production fuel injection system could be easily converted
to the advanced closed-loop form by the addition of some components,
which Bosch agreed to supply.
Components supplied by Bosch were installed on the 1975 Volvo
to convert its' fuel injection to the closed loop feedback control
system. They included: an oxygen " *" sensor, electronic control unit
(ECU), oxygen sensor threshold voltage trimbox, recalibrated warm-up
regulator, fuel distributor with frequency modulated (solenoid) pressure
control valve, and a revised air flow rate sensing unit. To complete
the system an Engelhard TWC-9 monolithic three-way catalyst was installed
in place of the production oxidation catalyst.
In preliminary testing, during which ECU control settings and
oxygen sensor threshold voltage levels were optimized, cold engine
starting problems were discovered. The engine would start, run about
two seconds, and then stall. Restarting required extended engine cranking,
Once started, the engine performed poorly (stalls, hesitation, and back-
fire) until completely warmed up. Hot starting presented no difficulty,
and hot engine performance was very good. Because of this behavior, a
defective warm-up regulator was suspected. After many recalibrations by
Bosch, which resulted in no significant improvement, the entire fuel
injection system was returned to Bosch in Chicago. It was found that an
incorrect component installation diagram had been furnished with the
system. Because of this, the warm-up regulator had been installed back-
wards, with its' inlet and outlet connections reversed. With the system
installed correctly, the engine started and performed smoothly even when
cold. Emissions were below 0 km contract target levels.
In addition to the contract tasks, the three-way catalyst
Volvo was one of about 70 different vehicles selected by the EPA to be
used in another study at ER&E to set up a sulfate test. From a timing
point of view it was desirable to complete these studies before initiating
the 80,000 kilometer accumulation. It was also felt that the 800 km
conditioning required for the sulfate test study would provide a shakedown
for the system. After the 800 km accumulation the car stalled repeatedly
-------
- 7 -
during start-up. A number of problem areas were indentified: the air
bellows between the airflow control sensor and the intake manifold had
loosened, the flange at the exhaust manifold had come loose, an intermittant
short had developed in the oxygen sensor lead, two fuel injectors leaked,
an electrical connection had come loose at the cold start injector,
spark plugs that were marginal, and an oxygen sensor that was defective.
With these problems repaired the vehicle started normally, and emissions
were below 0 km target levels.
Before initiating the 80,000 km accumulation the TWC-9 catalyst,
at the EPA's request, was replaced by an Engelhard TWC-9B monolithic
three-way catalyst. Emission tests were carried out as required after
2,000 km, 6,000 km, and 15,000 km were accumulated on the vehicle. NOX
levels during cold start FTP's remained constant, while CO and HC levels
tended to increase with accumulated kilometers. CO emissions at the
15,000 km point exceeded the 80,000 km target levels. S04S emissions
were well below contract target levels for 96 km/hr cruise conditions.
A tune-up reduced the CO level to 2.0 gm/km, and km accumulation was
resumed. At 27,000 km emissions were checked to see if any deterioration
had occurred which could be corrected before the scheduled 30,000 km
test sequence. CO again exceeded the maximum target level, while HC and
NOX increased to the 80,000 km limit. Even after replacement of the $2
sensor, emission levels were very close to the 80,000 km limits. These
results indicated that the three-way catalyst had lost much of its'
activity.
At this point, for the three -way catalyst vehicle, Tasks 1
and 2 were considered to be completed. Task 3, however, was never
completed to the full 80,000 kilometers due to the evident catalyst
deterioration seen at 30,000 kilometers.
CONTRACT MODIFICATION
This juncture in the Volvo three-way catalyst vehicle work
corresponds to the point in time when contract work was discontinued on
the low excess air vehicle. A contract modification was made (September
1976) stating that all contract work on the original vehicles/emission
control systems stop immediately, and two new vehicles/emission control
systems be chosen for the low excess air and three-way catalyst approaches.
It was decided to restart the three-way catalyst part of the
contract using a factory built prototype three-way catalyst equipped
Volvo. Through negotiations between the EPA and Volvo of America, a
prototype 1977 "Lambda-Sond" Volvo was made available to ER&E for testing.
In addition to determining the vehicles' 80,000 km durability for regulated
emissions, the levels of unregulated emissions (S04=, NH3, and HCN)
would be monitored during the accumulation.
-------
- 8 -
A new conceptual design for the low excess air vehicle was
agreed upon. It called for:
(1) An advanced fuel metering system (carburetor or fuel injection)
to maintain engine A/F slightly lean of stoichiometric.
(2) Either a prototype mine-mix three-way catalyst or oxidation
catalyst.
(3) Back-pressure modulated EGR system.
(4) The limited use of excess air (air pump) during cold start
operation.
(5) Optional (depending on the fuel metering system) exhaust gas
oxygen sensing and closed loop feedback control.
2.3 THREE-WAY CATALYST VEHICLE - DESIGN #2
A prototype 1977 Volvo with a redesigned 2.0 litre L-4 cylinder
engine (overhead camshaft versus the first three-way catalyst vehicle's
push-rod operated 2.0 litre engine) and "Lambda-Sond" three-way catalyst
system was selected as the vehicle to be tested. The Lambda-Sond
system, which is supplied to Volvo by Bosch, was an improved version of
the fuel injection system that was installed on the first three-way
catalyst demonstration vehicle. It uses electronic feedback control
with an oxygen sensor to modulate the vehicle's fuel injection system.
A/F ratio is maintained at the desired value by varying the amount of
fuel delivered to the engine.
The only vehicle modification made was the substitution of a
mine-mix three-way catalyst containing about 5% rhodium in place of the
original rhodium enriched certification catalyst (~17% Rh).
The vehicle passed 0 km emissions targets for HC and NOX with
both the certification and mine-mix three-way catalysts. Although CO
emissions (with both catalysts) were slightly above the 0 km target,
they were well below the 80,000 km maximum level, and durability testing
was initiated. The entire 80,000 km of durability testing were success-
fully completed with only minor malfunctions in the Lambda-Sond system
and a few mechanical vehicle failures. For instance, the water pump and
engine cooling fan clutch had to be replaced. A small fluid leak developed
in the automatic transmission which required a gasket replacement.
Finally, near the end of the accumulation interval the right rear wheel
bearing seized and caused the right axle shaft to break. After repairing
the axle and bearings, kilometer accumulation continued with no further
mechanical trouble to the 80,000 km finish. Problems with the Lambda-
Sond system included a cracked fuel injector retainer which caused an
air leak, one leaking fuel injector, and broken wires to the frequency
control solenoid valve on the fuel distributor.
-------
- 9 -
Throughout the 80,000 km durability testing interval both HC
and NOX emissions (as measured on the FTP) remained below the maximum
target levels. CO, however, exceeded the maximum target level at 15,000
km and continued to increase with accumulated kilometers. At the final
80,000 km test point, CO emissions were 60% greater than the maximum
allowable limit. Raw exhaust emission tests revealed that although CO
had increased greatly, conversion efficiency for the catalyst was greater
than 60%. Efficiencies for HC and NOX conversion were about 80% and 60%
respectively,
Sulfate (S04=) emissions remained well below the maximum
target levels (at the 96 km/hr cruise test mode) over the entire test
interval.
Thus the prototype 1977 Volvo three-way catalyst vehicle
completed the contract tasks. The entire kilometer accumulation was
completed with the full set of required emissions measured at the
specified intervals. The vehicle, equipped with a mine-mix three-way
catalyst met the 80,000 km contract targets of 2.1 CO/0.25 HC/1.2 NOX
(g/km) and 10 mg S04=/km, except for CO. Emissions of NH3 and HCN were
also monitored. Both remained at low levels over the durability interval
2.4 LOW EXCESS AIR VEHICLE - DESIGN #2
A new low excess air vehicle was designed for testing under
the contract modification. It consisted of a 1977 California emissions
certified Ford Pinto (2.3 litre overhead camshaft L-4 engine, back-
pressure modulated EGR, air pump and oxidation catalyst), that was
modified for low excess air operation with a Holley closed loop feedback
controlled carburetor with oxygen sensor, and an Engelhard monolithic
oxidation catalyst. The advantages of this particular design were, (1)
installation of the Holly feedback carburetor system on the 1977 Pinto
would require a minimum of mechanical work, since 1978 California three-
way catalyst Pinto1s were to be produced with the same system, (2)
Holley agreed to modify the electronic control unit for the feedback
carburetor so it could be set for low excess air conditions, rather than
three-way catalyst operation, (3) Holley agreed to furnish technical
assistance as required, and (4) Engelhard would supply a monolithic
catalyst of suitable size and "composition" to provide good conversion
activity under low excess air conditions.
Installation of the feedback carburetor system proceeded
smoothly after an incorrect installation schematic diagram was replaced
by Holley. Optimization was begun to determine the correct electronic
control unit (ECU) biasing for low excess air operation. With the
production catalyst operating under low excess air conditions, FTP CO
emissions were an order of magnitude higher than for the vehicle in
stock configuration. Adjustment of automatic choke "pull-off" rate and
recalibration of the ECU's cold start enrichment circuit lowered first
bag (cold start) CO levels. Weighted FTP CO emissions, though reduced,
still exceeded the both 0 and 80,000 km contract targets. Further
adjustments would not lower CO emissions.
-------
- 10 -
Catalyst temperature traces indicated that the production Ford
monolith required over three minutes to achieve "light-off" on cold
start. Substitution of an Engelhard PTX-514 monolith (a known active
oxidation catalyst) reduced light-off time to about one minute. Further-
more, first bag CO levels were decreased threefold, and total weighted
FTP CO and HC levels were below the 0 km targets. NOX emissions,
however, increased and exceeded the 80,000 km maximum limits. This NOX
increase was attributed to the PTX-514 catalyst's different composition.
That is, during rich transients the PTX-514 reduced less NOX than did
the production Ford catalyst. Adding credence to this argument was the
fact that raw exhaust (no catalyst) FTP data showed even higher NOX
levels than those seen with the PTX-514.
Increasing EGR rate reduced NOX to within contract targets,
but CO emissions increased slightly. Although CO was slightly above the
0 km target level, it was well below 80,000 km maximum limits. Because
of this, it was decided to start kilometer accumulation. The PTX-514
(68 in3) catalyst was replaced by an Engelhard PTX-516 monolith (102 1n3).
It was expected the larger catalyst volume would help offset the effects
of catalyst deterioration, making achievement of the 80,000 km goal more
certain. With the PTX-516, the vehicle met all contract targets at
zero kilometers, and was sent out for the first 2,000 km durability
interval.
The feedback carburetor system was plagued with operating and
durability problems from the time it was installed. Vacuum control
regulator valves which interface the ECU to the feedback carburetor were
very susceptible to plugging. Many times during calibration and optimi-
zation runs, the regulator valve would become plugged, causing the feed-
back system to fail full rich. This problem was attributed to the small
(0.030 inch) air bleed orifices in the valve which could easily be plugged.
In addition to this problem, the particular feedback carburetor
used in the program suffered durability problems. At the end of the
first 2,000 km durability interval CO emissions had increased, and the
system would not control A/F properly. Holley agreed, at ER&E's request,
to send an engineer and repair the failure. He recommended that the
carburetor be returned to the factory. It was found that the carburetor
had shifted from its original calibration and had to be readjusted.
In repairing the carburetor a transition region instability
appeared. In the region of transition from the idle to main jet circuits,
raw fuel dripped from the main jet nozzle and puddled on the throttle
plate. As a result CO emissions oscillated sharply, and the system
could not maintain A/F control. FTP CO emissions were 1n excess of the
80,000 km limits. The carburetor was returned to Holley for idle/main
jet reproportioning. However, they remarked that it could take several
iterations to solve the problem. Testing of the repaired carburetor
revealed that although the Instability speed range had been narrowed, CO
emissions were still above contract limits. Another iteration would
have to be made to solve the problem.
-------
- 11 -
It was jointly agreed by ER&E and the EPA that contract work
on the second low excess air vehicle be terminated based upon achieve-
ment of 0 km emission goals.
Thus for the second low excess air vehicle, Tasks 1 and 2 were
completed. Task 3, however, was dropped from the contract requirements
because of recurring feedback carburetor system failures. The vehicle
met all 0 kilometer contract targets, including that for S04=. This
shows that the low excess air concept does inhibit oxidation catalyst
sulfate production, and is technically feasible for application to motor
vehicles.
TASK ACCOMPLISHMENT
A summary of the vehicles designed and fabricated, as well as
the progress toward completion of required contract tasks is found in
Table 2-1.
TABLE 2-1
SUMMARY OF VEHICLES TESTED AND
CONTRACT TASKS ACCOMPLISHED
Vehicle
Task 1
Task 2
Task 3
Low Excess
Air Vehicle
Design #1
(1975 Chevrolet
Malibu)
Three-Way
Catalyst Vehicle
Design #1
(1975 Volvo)
Three-Way
Catalyst Vehicle
Design #1
(1977 Prototype
"Lambda-Sond"
Volvo)
Low Excess Air
Vehicle Design #2
(1977 Ford Pinto)
(Conceptual
Design)
Completed
Completed
Completed
Completed
(Modification &
Fabrication)
Completed
Completed
Completed
Completed
(80,000 km.
Durability
Testing)
Deleted with
EPA Approval
Carried to
30,000 km.
Stopped with
EPA Approval
Completed
Terminated
with EPA Approval
-------
- 12 -
SECTION 3
CONCLUSIONS
A total of four vehicles, two low excess air-oxidation catalyst
systems and two three-way catalyst systems, were built to demonstrate
the feasibility of low sulfate production automotive emission control
technology. General conclusions drawn from the testing of these vehicles
are summarized below.
Low Excess Air Vehicles
Application of the "low excess air" concept to automotive
emission control systems that utilize oxidation catalysts is technically
feasible, and controls sulfate emissions to levels comparable to those
of non-catalyst vehicles.
To successfully implement the low excess air concept on a
vehicle, two key system design criterion have to be met. First, the
engine's fuel metering system must be capable of precisely controlling
A/F ratio at the required "slightly lean" value regardless of vehicle
driving mode. This is necessary so that a low oxygen partial pressure
atmosphere is consistently maintained over the oxidation catalyst to
minimize sulfate formation. Second, the oxidation catalyst bed must
light-off quickly from cold start conditions. This is especially
important since catalyst CO and HC conversion efficiencies are reduced
when operating under low excess air conditions. The compounding of
reduced catalyst conversions with slow light-off would result in unac-
ceptably high tailpipe emission levels.
The design criteria outlined above are a result of experience
gained from fabricating and testing of two low excess air demonstration
vehicles. It should be kept in mind that although individual system
components will vary depending on the particular vehicle application,
these two basic design premises must be followed to insure technical
success.
Three-Way Catalyst Vehicles
Three-way catalyst's are a viable technology for low sulfate
production automotive emission control systems. Testing of the two con-
tract demonstration vehicles has shown that three-way catalyst systems
simultaneously control all three regulated emissions (CO, HC, NOX) to
stringent levels, while emitting very low levels of sulfate, ammonia,
and hydrogen cyanide.
Vehicle fuel metering is a primary concern in the design of a
three-way catalyst system. Unless the three-way catalyst is maintained
within it's operating window by precise control of the engine at a
stoichiometric A/F ratio, losses in conversion efficiency will be
experienced.
-------
- 13 -
In testing of the demonstration vehicles both three-way
catalyst's (one preparation rhodium enriched, the other mine-mix)
suffered deactivation with age. In general, for both catalysts CO
emissions increased significantly with accumulated kilometers, while HC
and NOX emissions also increased, but to a lesser extent.
-------
- 14 -
SECTION 4
RECOMMENDATIONS
Two approaches for low sulfate production automotive emission
control systems were evaluated under this contract. Recommendations for
future work in these areas are given below.
LowExcess^ _Air-Oxidation Catalyst Systems
Although the low excess air concept was successfully demonstrated,
recurring hardware failures in the demonstration vehicle's fuel metering
system prevented durability testing. It is expected that the durability
problems encountered with the prototype fuel metering system (Hoiley
feedback carburetor system) will be resolved as system production and
vehicle application increases.
Future work, then, should be directed toward studying the
long-term effects of kilometer accumulation on low excess air system
components. In particular, the effect of operating under low excess air
conditions on oxidation catalyst deterioration over 80,000 km remains
undetermined. Remaining questions to be answered include: the effect
of catalyst composition on conversion activity under low excess air
conditions, and, the effect of increasing catalyst bed oxygen levels on
vehicle tailpipe CO, HC and S04= emission levels.
Three-Uay Catalyst Systems
The substantial deactiyation seen during durability testing of
both rhodium enriched and mine-mix rhodium ratio three-way catalysts
pointed out a need for catalyst development work. Since work on this
contract was completed several manufacturers have commercialized three-
way catalyst vehicles with good 80,000 km. durability. System hardware
and catalysts necessary for three-way catalyst operation are durable,
and are currently available in production vehicles. The application of
feedback carburetor technology to three-way catalyst systems has no
technical barriers, other than the durability problems, mentioned in
connection with the low excess air catalyst systems.
-------
- 15 -
SECTION 5
LOW EXCESS AIR VEHICLE - DESIGN NO. 1
5.1 DESIGN AND FABRICATION
A 1975 Chevrolet Malibu equipped with a 350 cubic inch displace-
ment V-8 engine and automatic transmission was chosen as the vehicle to
be modified. The low excess air system design consisted of:
(1) Research Carburetor
(2) Super Early Fuel Evaportion System (SEFE)
(3) Proportional Exhaust Gas Recirculation (EGR)
(4) "Starter" and "Main" Oxidation Catalysts
A discussion of each of the systems components follows.
Research Carburetor
In order to maintain the low excess air conditions that are
necessary for low sulfate production from an oxidation catalyst, the A/F
ratio inducted to the engine must be maintained slightly lean of stoichio-
metric, at approximately A/F = 15/1. Exxon Research and Engineering
obtained from General Motors a "research" carburetor that could be
jetted to provide a wide range of A/F ratios, was repeatable, and could
maintain the set A/F ratio over the full range of throttle openings.
This "flatness" of A/F ratio was a necessity for the operation of the
low excess air vehicle. Rich deviation from the desired 15/1 A/F ratio
would place the catalyst in a reducing regime where conversion levels of
CO and HC were low, resulting in increased tailpipe emissions. Lean
deviation would not affect conversion levels of the oxidation catalyst,
but would increase sulfate production.
The research carburetor was installed on the vehicle in place
of the production carburetor. A number of main jets were tested to find
the size that gave the desired A/F ratio for low excess air conditions.
As can be seen from Table 5-1 A/F ratio was fairly sensitive to changes
in main jet diameter. For a fixed main jet diameter, however, A/F ratio
was relatively constant across the speed range tested. This A/F flatness
is especially apparent when the speed-A/F variations of the production
carburetor are compared to those of the research carburetor.
-------
Research
Carburetor
- 16 -
Table 5-1
EFFECT OF MAIN JET SIZE ON A/F RATIO
Jet Diameter A/F Ratio At
(Inches)Idle 48 kph 80 kph
0.0415 16.2 16.3 16.5
0.0425 15.4 15.5 15.6
0.0432 15.0 15.6 15.5
0.0440 15.9 14.8 14.7
Sbureto? °-040 17'3 16'5 15-5
A main jet diameter of 0.0425" was chosen for the test vehicle since it
provided the flattest A/F-speed response and was relatively close to the
desired A/F ratio.
Super Early Fuel Evaporation System
Since the design of the low excess air vehicle did not include
an air pump, catalyst light-off from a cold start would be slow. To
compensate for the resultant loss of CO control in Bag 1 (cold bag) of
the FTP, it was desired to open the carburetor choke as rapidly as pos-
sible. This reduction of A/F enrichment due to rapid dechoking coupled
with a nominally lean A/F mixture was expected to create serious cold
driveabillty problems. To offset this a super early fuel evaporation
system (SEFE) was designed for the vehicle.
Figure 5-1 shows a pictorial diagram of the SEFE system. In
the cold start mode early fuel evaporation (EFE) valves 1 and 2 are
closed, and EFE valve 3 1s open. Hot exhaust gases from the combustion
chambers are blocked from flowing out through the exhaust manifolds.
Instead, all of the exhaust gas is diverted internally through the
intake manifold's crossover into a plenum chamber. This chamber is
vented by an auxiliary exhaust pipe which rejoins the vehicle's exhaust
system downstream of the exhaust manifolds. The rapid heating of this
plenum, which forms most of the intake manifold floor under the carburetor
base, promotes the vaporization of condensed fuel puddles and thus
should enhance cold drlveability with lean A/F mixtures. In hot mode
operation EFE valves 1 and 2 are opened, while EFE valve 3 is closed.
Engine exhaust then follows the normal path out of the exhaust manifolds
and into the vehicle's exhaust system. The three EFE valves used to
divert exhaust gas flow were standard production units with vacuum motor
actuators. They were modified for remote switching from the passenger
compartment by placing electrically operated solenoid valves in their
vacuum supply lines.
-------
- 17 -
EXHAUST HEATED
PLENUM (MANIFOLD
FLOOR)
INTAKE MANIFOLD
EXHAUST CROSSOV
STARTER MONOLITHIC
CATALYST
\ EFE
VALVE 3
EFE
VALVE
TO MAIN
MONOLITHIC CATALYST
AIR AND FUEL MIXTURE (INTAKE CHARGE)
HOT EXHAUST GAS
FIGURE 5-1
SEFE SYSTEM PICTORIAL DIAGRAM
-------
- 18 -
The SEFE system differed from production early fuel evapor-
ation systems in that all of the engine exhaust was diverted through the
intake manifold. In production systems only one exhaust manifold is
partially blocked off during cold start operation. Exhaust gas is
diverted up through the intake manifold and out through the opposite
cylinder head and exhaust manifold. Production EFE systems, therefore,
do not have an auxiliary exhaust passage from the intake manifold.
Because there is less total exhaust gas flow, plenum heating rates are
slower, and final intake manifold floor temperature is lower than for
the SEFE system.
Construction of the SEFE system involved extensive modifica-
tions to the intake manifold. To ease fabrication (e.g. welding and
machining), a cast aluminum manifold - Edelbrock "Streetmaster" Model
3025 was substituted for the production cast iron manifold. Figure 5-2
is a top view of the aluminum manifold before SEFE modifications were
made. In Figure 5-3 the entrance port to the intake manifold exhaust
crossover passage is shown within the circle. This passage leads to the
exhaust plenum chamber, which forms the heated intake manifold floor
under the carburetor base. The plenum chamber, of course, is not
visible in the photograph since it is an internal cavity. The chamber
boundaries, however, are approximately those shown by the dotted lines.
The auxiliary exhaust passage required to vent the plenum chamber is
shown in Figure 5-4 and 5-5. The original design specification for a
5.08 cm (2 in.) pipe size could not be met due to space constraints.
Instead, a 3.81 cm (1.5 in.) exhaust line was installed. Also shown in
Figure 5-5 are the locations of the exhaust crossover passages and the
plenum chamber. A bottom view of the aluminum intake manifold is shown
in Figure 5-6. In this photograph the transition piece that was fabricated
to connect the plenum chamber to the auxiliary exhaust pipe is visible.
The routing of hot exhaust gas through the SEFE system is
shown in Figure 5-7. During cold start operation the exhaust manifolds
are blocked-off by closed EFE valves. Exhaust is diverted up through
the cylinder heads, and into the intake manifolds exhaust crossover
passage. This hot gas flows through (and thus heats) the plenum chamber
and is vented from the manifold by the auxiliary exhaust passage. The
passages and plenum shown in the preceeding figures are internal cavities
located beneath the intake manifold floor and runners.
Proportional Exhaust Gas Recycle (EGR) System
The low excess air vehicle design specified exhaust gas re-
cycle (EGR) for control of NOX emissions to within contract target
levels. An EPA supplied EGR valve was installed on the vehicle. Valve
calibration was "proportional", that is, the amount it opened was a
direct function of the airflow through the engine. The valve was
adjustable, in that different springs could be placed in the actuator's
-------
EDELBROCK ALUMINUM MANIFOLD
BEFORE MODIFICATION
-------
INTAKE MANIFOLD EXHAUST
CROSSOVER AND PLENUM CHAMBER
I
o
-------
AUXILIARY
EXHAUST
PASSAGE
SEFE AUXILIARY EXHAUST PIPE
ro
i
-------
EXHAUST CROSSOVER
PASSAGE
PLENUM
CHAMBER
EXHAUST CROSSOVER
PASSAGE
ro
FIGURE 5-5
EDELBROCK ALUMINUM MANIFOLD-
AFTER SEFE MODIFICATION
-------
TRANSITION
PIECE
PLENUM CHAMBER
ro
OJ
FIGURE 5^j
SEFE HEATED PLENUM-
CONNECTION TO AUXILIARY EXHAUST PASSAGE
-------
AUXILIARY
EXHAUST PASSAGE
EXHAUST CROSSOVER
PASSAGE
PLENUM CHAMBER
EXHAUST CROSSOVER
INJ
FIGURE 5-7
EXHAUST GAS FLOW THROUGH SEFE MODIFIED INTAKE MANIFOLD
-------
- 25 -
vacuum diaphragm assembly, so the EGR schedule could be optimized for
NOX control and driveability. Mixing of the recycle stream with the A/F
mixture was under the carburetor base at the EGR entrance port. The
location of this port is in the intake manifold floor and is shown in
Figure 5-8.
"Starter" and "Main" Oxidation Catalyst System
One can place a small starter oxidation catalyst close-coupled
at the engine exhaust manifold when there is slow light-off of the main
catalyst bed. The small bed volume, high exhaust gas space velocity,
and high input gas temperature, promote rapid light-off of the starter
catalyst. Thus, initial emissions on cold start are controlled by the
starter catalyst (while the main catalyst bed lights-off).
Slow catalyst light-off was expected for the low excess air
vehicle, since its' design did not include an air pump (air injection on
cold start). Therefore, the starter/main catalyst configuration was
adopted to increase control of cold start emissions. The starter oxida-
tion catalyst bed was located in the auxiliary exhaust passage that
vents the intake manifold's plenum chamber. (See Figure 5-1). On cold
start, with the SEFE system in operation, the starter catalyst is placed
in series with the main catalyst bed. During hot operation, the starter
catalyst and SEFE system are not activated, and, therefore, only the
main catalyst bed receives exhaust from the engine. Catalysts chosen
for use in the system were Matthey-Bishop type 20G platinum monoliths.
The catalyst's specifications and bed dimensions are listed in Table 5-2,
Table 5-2
CATALYST SPECIFICATIONS
Noble Metal
Loading
Substrate
Monolith
Dimensions
(Diameter X
Length)
Catalyst
Volume
Starter Catalyst
Pt, 14.1 gm/litre
(40 gm/ft3)
Corning Monolith -
47 Holes/cm?
(300 holes/in2)
D - 9.30 cm X L = 7.60 cm
(3.66 in) X (3.0 in)
516.0 cm3
(31.5 in3)
Main Catalyst
Same
Same
D = 11.84 cm X L = 15.24 cm
(4.66 In) X (6.0 in)
1678.0 cm3
(102.3 in3)
-------
- 26 -
f
CARBURETOR
MOUNTING
FLANGE 2
MANIFOLD
FLOOR
EGR ENTRANCE PORT LOCATION
-------
- 27 -
5.2 TEST RESULTS
Due to various production difficulties Matthey-Bishop was
unable to supply the starter and main catalysts for a period of six
months from the order date. During this time initial optimization of
the low excess air vehicle was made with sections of blank exhaust pipe
in the locations that were to be occupied by the starter and main catalysts.
These preliminary tests revealed several problem areas. It was found
that the SEFE section of the intake manifold warmed-up very slowly from
cold start conditions. The slow heating rate was found to be due in
part to poor closure of the EFE valves at the exhaust manifolds. Inspection
of the production EFE valves revealed that they had 1.6 mm (1/16 inch)
clearance around the edge of the butterfly plate when closed. Stainless
steel butterfly plates with reduced edge clearance were fabricated and
installed to decrease the amount exhaust gas that could blow-by in the
closed position.
Tests with the modified valves showed that they reduced the
time required to reach a given under-carburetor temperature to about
half that obtained with production EFE valves. The rate of temperature
rise (measured with a thermocouple installed in intake manifold floor
under the carburetor) observed was 1.2°C/sec. (2.2°F/sec.) for the cold
start FTP. Initial temperature of the manifold was 21.1°C (70°F) and
final stabilized hot temperature was 132.3°C (270°F) after SEFE opera-
tion for approximately 240 seconds. This was only about 1/5 of the rate
(5.6°C/sec. = 10°F/sec.) which others (6) had indicated as required for
good driveability with rapid choke pull-off. It was suspected that this
poor warm-up performance was caused by rapid heat dissipation from the
SEFE section due to the aluminum manifold's high thermal conductivity.
Driveability of the low excess air vehicle was poor. During
the cold start FTP there were repeated stalls, surging, sluggish acceleration
performance, and backfires. Some loss of cold driveability was expected,
due to the poor SEFE warm-up characteristics of the aluminum intake
manifold. However, even with the engine fully warmed up if the SEFE
system was activated, severe engine hesitation, sluggish performance,
and backfiring would occur. The backfiring was found to be a result of
EGR maldistribution. This was established by installing taps at the
left front, left rear, and right rear of the intake manifold. Samples
were withdrawn with and without EGR while the engine was run at various
steady state conditions. The percent of recycled exhaust (EGR) at the
right rear of the intake manifold was generally about twice that of the
left front, with the left rear being intermediate. The maldistribution
was eliminated by redesigning the EGR inlet so it would introduce the
recycled exhaust directly under the carburetor throttle plates.
-------
- 28 -
As previously shown (Figure 5-8) the Edelbrock manifold intro-
duced EGR through a "floor port" which was about 9 cm (3.5 in.) below
the carburetor base. A 9 cm (3.5 in.) long by 1.27 cm (0.5 in.) ID
aluminum tube with two top ports was threaded into the original EGR
port. A schematic diagram of the modified port, and an installed view
are shown in Figure 5-9. The redesigned inlet directed EGR flow across
the incoming A/F mixture to insure good mixing and distribution would
take place. The improvement made in EGR distribution is shown in Table
5-3, which compares percent recycle measured before and after installing
the modified inlet.
TABLE 5-3
EGR DISTRIBUTION (AT 48 KPH: 30 MPH)
Sample Location Percent EGR Based on % C02
Before New After New
Inlet Installed Inlet Installed
Left Front 6.7 8.9
Left Rear 9.9 9.6
Left Reat 12.8 9.9
Avq. 9.8 9.5
Improving the EGR distribution eliminated the backfiring
problem with the SEFE system activated; however, stalling and poor
acceleration performance continued to occur. A/F ratio enrichment by
resetting the choke helped to reduce the tendency to stall on cold start
and during warm-up on the FTP. The additional enrichment was necessary
to compensate for poor fuel vaporization resulting from slow SEFE heating,
Vehicle performance during both cold and hot start driving
modes while the SEFE system operated was poor. During deep throttle
acclerations in the FTP (such as the 162 sec.-332 sec. "Big Hill" in
Bags 1 and 3) severe hesitation, low engine power output, and sluggish
acceleration were experienced. The low excess air vehicle, however,
performed normally with the SEFE system switched off. During hot engine
operation (no SEFE) exhaust was vented into two 6.35 cm. (2.5 in.) pipes
of the production exhaust system. In SEFE mode, however, all exhaust
was vented through a 3.81 cm. (1.5 in.) auxiliary exhaust passage.
Operation of the SEFE system created approximately a five fold reduction
in exhaust flow area. [Total available SEFE flow area: (3.81 cm]2 /4 =
11.42 cm2, as compared to non-SEFE area: (6.35 cm)2 /4 = 31.68 cm? x 2
exhaust pipes for V-8 engine = 63.35 cm2 total area]. It was felt that
a high level of exhaust backpressure developed during deep throttle
transients and reduced engine power output, causing hesitation and
sluggish accelerations.
-------
- 29 -
CARBURETOR |
MOUNTING FLANGE
THERMOCOUPLE
(FLOOR TEMPERATURE)
FIGURE 5-9
MODIFIED EGR INLET
0
c
8 ."
(3 ]
.64 c
J.251
1
1
t
) cm
/2»)
1
1
~~
^
^
s^*^
\
L
s
_
/
X
^32 cm
J(0.125")
' 1 .27 cm
t(0.50")
\
One rm
(3/8")
1.27 cm_
(0.50")
-------
- 30 -
Thus, preliminary testing revealed a number of problems with
the low excess air vehicle. In particular, the SEFE system did not
perform as expected. The function of the system, to promote intake
manifold warm-up and fuel vaporization, was never fully realized. The
aluminum intake manifold, chosen to ease SEFE system fabrication, had
high thermal conductivity and dissipated enough heat to prevent rapid
warm-up. Furthermore, due to space constraints during fabrication, the
SEFE auxiliary exhaust passage had to be reduced in diameter (crossectional
area). The increased exhaust backpressure resulting from this caused
poor vehicle performance. The net effect was that the system instead of
enhancing cold driveability, degraded it. It was decided to determine
the emissions level that could be attained with the low excess air
vehicle, before further modifications were made to the SEFE system or
intake manifold.
Matthey-Bishop was able initially to supply only the "starter"
catalyst bed. Testing proceeded with an Engelhard PTX-IIB platinum-
palladium monolith, until the Matthey-Bishop 20G main catalyst bed
became available. During testing various adjustments such as leaning
the choke and shortening of the accelerator pump stroke were made in an
effort to reduce exhaust emission levels. Table 5-4 shows the test
results obtained.
TABLE 5-4
LOU EXCESS AIR VEHICLE EXHAUST EMISSION LEVELS
FTP Emissions, g/km SEFE
"CO""TTCT TO System
80,000 km. emission targets 2.1 0.25 1.2
0 km. emission targets 1.0 0.12 1.2
PTX-IIB main catalyst alone 8.3 0.32 2.3 On
PTX-IIB main & M-B starter catalyst 10.5 0.43 2.0 On
M-B main catalyst alone 4.4 0.36 2.1 On
M-B main & starter catalysts 5.2 0.32 2.2 On
M-B main & starter catalysts 4.8 0.14 2.2 On
(air injection into exhaust manifolds)
M-B main catalyst alone 3.5 0.25 2.2 Off
(normal crossover heating)
When the SEFE system and starter catalyst were used, the main
exhaust lines were closed. All exhaust was vented through the auxiliary
passage for the first 240 seconds in Bag 1, and for the first 135 seconds
in Bag 3 of the FTP. In this configuration, the starter and main catalyst
beds were placed in series. When the main catalyst alone was used, the
auxiliary passage was closed (SEFE System Off) and exhaust flowed through
the main exhaust lines to the catalyst for the entire FTP.
-------
- 31 -
As shown in the table, lower emission levels were obtained
without the SEFE system/starter catalyst, both with the PTX-IIB and
Matthey-Blshop main catalysts. The latter tended to give the lowest
emission levels overall. The lowest emission levels were obtained with
the vehicle operating in the normal manner. That is, using only the
main catalyst, without SEFE operation, allowing the exhaust to cross
over through the intake manifold from the driver to passenger side of
the engine for the first 135 seconds of both Bags 1 and 3. However, it
should be noted that in all cases, exhaust emission levels were above
both 0 and 80,000 km target levels. The poor emissions performance of
the low excess air vehicle was attributable to a number of factors such
as; design shortcomings in the SEFE system, lack of air injection on
cold start, and relatively small catalyst volumes.
Poor control of CO, for example, had a number of causes.
Enrichment of A/F ratio to reduce cold start stall-outs, which reduced
the oxygen content in the exhaust, tended to increase catalyst bed
light-off time. Operation of the SEFE system/starter catalyst itself
created problems of high exhaust backpressure. The higher level of CO
consistently seen with the starter catalyst was suspected to be a result
of having to run at near wide open throttle conditions on FTP accelerations.
The addition of an airpump to supply secondary air injection at the
exhaust manifold reduced catalyst light-off time and improved the control
of cold start CO and HC emissions (airpump operating during the first
150 seconds of Bag 1). However, as shown in Table 5-4, overall FTP
emissions remained above contract targets. At this time, an instability
in the research carburetor was found. During testing carburetion often
became unstable, and frequently went rich. The problem was traced to a
temperature instability in the carburetor, which was aggrevated by the
SEFE system increasing the carburetor body temperature. In Table 5-5
A/F ratios for the low excess air vehicle at two temperatures are displayed.
Note that although A/F ratio for a particular temperature remained
"flat" with speed variations, as temperature increased the A/F became
rich.
Table 5-5
RESEARCH CARBURETOR TEMPERATURE INSTABILITY
Speed A/F Ratio
Ikphl(mphT 24°C (75QF)29°C (84°F)'
IDLE IDLE 15.4 14.8
16 10 14.8 14.7
32 20 15.3 14.6
48 30 15.1 14.8
64 40 14.9 14.6
80 50 14.9 14.4
96 60 15.0 14.5
-------
- 32 -
Thus, the low excess air vehicle could not meet any of the
contract emission level targets. A number of problem areas remained
unresolved, and would require extensive redesign and modifications to
the vehicle. Rather than pursue this path, it was decided to stop work
on the vehicle. A second low excess air vehicle was built, and is
described in Section 8 of this report.
-------
- 33 -
SECTION 6
THREE-WAY CATALYST VEHICLE - DESIGN NO. 1
6.1 DESIGN AND FABRICATION
A 1975 Volvo equipped with a fuel injected 2.0 litre displace-
ment L-4 cylinder engine was chosen for modification. A prototype
K-JETRONIC fuel injection system, developed by the Robert Bosch Corpor-
ation, using closed loop electronic feedback control with an oxygen
sensor was installed on the vehicle. Such a system was expected to
provide the precise control of engine A/F ratio necessary for efficient
operation of a three-way catalyst within its' conversion window. The
conversion to the advanced closed loop feedback control system consisted
of replacing production Volvo fuel injection system parts (also Bosch
built) with prototype parts supplied by Bosch. Components installed in
the modification included: an oxygen (X) sensor, electronic control
unit (ECU), oxygen sensor threshold voltage trimbox, recalibrated warm-
up regulator, fuel distributor with frequency (solenoid) pressure control
valve, and revised air flow rate sensing unit. To complete the system,
an Engelhard TWC-9 monolithic three-way catalyst was installed (replacing
the production oxidation catalyst).
A simplified pictorial diagram of the Bosch closed loop feedback
control fuel injection system is shown in Figure 6-1. A functional
description of each of the components follows:
1. ELECTRIC FUEL PUMP - a roller cell fuel pump driven by an electric
motor, pumps fuel from the fuel tank into the fuel injection
system.
2. FUEL ACCUMULATOR - holds the fuel pressure constant for an extended
length of time after the engine has been turned off. This prevents
the formation of gasoline vapor bubbles (fuel percolation: vapor
lock), and as a result improves hot engine starting.
3. FUEL FILTER - protects the fuel distributor and the fuel injector
nozzles against clogging and damage from dirt.
4. FREQUENCY SOLENOID PRESSURE CONTROL VALVE - signals from an
electronic control unit, based upon exhaust oxygen content
(determined by an oxygen sensor), open and close this magnetic
solenoid valve. Air-fuel mixture is varied by changing the primary
circuit control pressure in the Mixture Control Unit. Decreasing
primary control pressure increases fuel flow (enriching the A/F
mixture), and corresponds to an increase in solenoid valve duty
cycle or "% on" time. Decreasing the solenoid valve duty cycle,
increases primary control pressure and leans the A/F mixture, due
to decreased fuel flow.
-------
- 34 -
5. WARM-UP REGULATOR - controls the pressure acting against the top
of the control plunger. During cold start operation the control
plunger pressure is reduced enriching the A/F mixtures to aid
cold driveability. An electrically heated bimetallic strip
switches the regulator off after the warm-up period, and prevents
regulator operation under hot start conditions.
6. MIXTURE CONTROL UNIT - consists of the air-flow sensor and the
fuel distributor. The air drawn into the engine, the volume flow-
rate of which depends on the position of the throttle plate,
lifts the air-flow sensor plate, and at the same time the control
plunger in the fuel distributor is lifted by a shorter lever arm
against the hydraulic primary control pressure. The amount of fuel
required for the volume of air flowing through the air-flow sensor
is metered in this way and is fed through the metering slits to the
individual injection valves.
7. PRIMARY CIRCUIT PRESSURE REGULATOR - holds the primary fuel circuit
pressure at a constant value. Adjustment of the primary circuit
regulator affects the amount of fuel flow for a given air-
flow rate.
8. DIFFERENTIAL PRESSURE VALVE - designed to assure that the volumetric
flow of fuel depends only on the crossectional areas of the
metering slits.
9. METERING SLITS - the fuel flows through the metering slits, one
for each cylinder of the engine, depending only on the crossectional
area of the slits opened by the control plunger as it is moved up
and down by the air-flow sensor plate.
10. START VALVE - sprays additional fuel into the intake manifold
during cold start operation to compensate for reduced fuel
vaporization.
11. AUXILIARY AIR DEVICE - feeds more air to the engine during warm
up (to increase engine idle speed), then closes the by-pass channel
around the throttle plate by means of an electrically heated
bimetallic strip.
12. THERMO-TIME-SWITCH - controls the length of time the cold start
valve operates upon cold start, as well as preventing valve opening
above a certain temperature limit.
13. INJECTION VALVE - sprays the precisely metered fuel into the intake
manifold, and is continuously open after the engine is started.
-------
Injection Valve
Metering ( q
Slit v
Idle Speed
Adjusting Screw
Throttle Plate
Air Funnel
Air-Flow Sensor Plate
Start Valve MB
Control
Plunger
Fuel Distributor
Differential-
Pressure Valve
Common Intake
Manifold
Primary Circuit
Pressure Regulator
©
Auxiliary-air Device
il
Thermo-time Switch
12
Air-Flow Sensor
Frequency Solenoid \
Pressure Control ^-^~^~.
Valve s~^ III]
Warm-up Regulator
Mixture Control Unit
6
To Electronic
Control Unit
(oxygen)
Fuel Accumulator
2
Electric Fuel Pump
©
FIGURE 6-1
BOSCH K-JETRONIC FEEDBACK FUEL INJECTION SYSTEM PICTORIAL DIAGRAM
tn
i
-------
- 36 -
It is not within the scope of this report to describe in great detail
how the Bosch feedback fuel injection system operates. Rather, the
reader is referred to Bosch technical papers (7,8) for the design and
operation of the continuous fuel injection system, as well as detailed
descriptions of the feedback control system's electronics (9). A simplified
"action circle" of the feedback control fuel injection system's logic is
found in Figure 6-2. It shows how the system compensates for variations
in the air-fuel mixture, so that a desired A/F value is maintained.
6.2 TEST RESULTS
During preliminary testing, in which ECU control settings and
oxygen sensor threshold voltage levels were optimized, a problem with
cold engine operation was found. Upon cold start, the engine would run
about two seconds, and then stall out. Restarting of the engine required
extended starter cranking. After restart, vehicle performance was poor
with stalling, hesitation and backfire until the engine was fully wanned
up. Hot starting and hot engine performance, however, were normal.
This type of behaviour indicated that the warm-up regulator (supplied
with the feedback fuel injection system) was not operating properly.
Bosch felt that although the warm-up regulator had been set to provide a
lean cold start A/F ratio (to minimize CO emissions), this should not
cause the poor driveability being experienced. They agreed that the
warm-up regulator could be malfunctioning and suggested it be returned
for a calibration check.
Bosch recalibrated the regulator and returned it with some
additional components (e.g. acceleration enrichment system with manifold
absolute pressure sensing) to help improve cold start engine performance.
Testing after installation of these modified parts showed that the
start-up and cold driveability difficulties previously experienced were
corrected. The feedback system, however, could not maintain stable A/F
ratio control and FTP emissions were above maximum contract target
levels. The entire feedback fuel injection system was returned to Bosch
in Chicago for diagnostic testing. It was discovered that an incorrect
component installation diagram had been furnished with the system.
Because of this the Warm-up regulator had been installed backwards, with
its' inlet and outlet connections reversed. With the warm-up regulator
installed correctly the feedback system maintained A/F ratio control,
and emissions were below 0 km contract targets, as shown in Table 6-1.
TABLE 6-1
THREE-WAY CATALYST VOLVO EXHAUST EMISSIONS
PRELIMINARY TEST
FTP Emissions, g/km
CO HC
80,000 km. emissions target 2.1 0.25 1.2
0 km. emissions target 1.0 0.12 0.6
Measured emissions 0.87 0.12 0.52
-------
- 37 -
Injected Fuel
Quantity
Increases
Rich
Mixture
No Oxygen In
Exhaust Gases
\
Regulated Fuel
Pressure
Decreases
High Sensor
Voltage
\
Open/Close Ratio
Of Frequency Valve
Is Decreasing
Open/Close Ratio
Of Frequency Valve
Is Decreasing
\
Low Sensor
Voltage
\
Regulated Fuel
Pressure
Increases
Oxygen In
Exhaust Gases
Lean
Mixture
Injected Fuel
Quantity
Decreases
FIGURE 6-2
"ACTION CIRCLE" OF FEEDBACK
CONTROLLED FUEL INJECTION SYSTEM'S LOGIC
-------
- 38 -
In addition to the contract tasks the three-way catalyst Volvo
was one of approximately seventy vehicles selected by the EPA to be used
in a study to standardize sulfate testing. It was desirable, from a
timing point of view, to complete these studies before starting the
80,000 km durability accumulation. It was felt that the 800 km of
conditioning required for the sulfate test study would provide a good
initial shakedown for the vehicle. After accumulating 800 km on the
modified AMA driving schedule, as required for the test procedure and
sulfate standard study, the car stalled repeatedly during start up and
the feedback fuel injection system often went out of control during FTP
and SET testing.
A number of problem areas were found upon inspection of the
vehicle. (1) The air bellows between the air flow sensor unit and the
intake manifold had come loose, which could affect the A/F ratio. Since
air leaks after the airflow sensor do not increase fuel flowrate, a net
leaning-out of the air fuel mixture inducted to the engine would take
place. If the air leak rate was substantial the feedback control system
could become "pinned" at its full-rich limit trying to bring the A/F
ratio back to a stoichiometric value, causing a loss of A/F control.
(2) The flange connection of the exhaust header pipe to the engine
exhaust manifold had come loose and opened slightly, probably due to
vibration. Any air leakage into the exhaust pipe would be seen by the
oxygen sensor (located about 2 cm upstream of the flange) as a lean A/F
mixture, and could result in the same condition as previously described.
And (3) an intermittant short had developed in the output lead from the
oxygen sensor.
With these areas repaired the feedback system was able to
maintain A/F control, however, the vehicle still stalled on start up. A
complete engine tune-up was carried out in an effort to determine the
reason for the repeated stall-outs. Three possible contributing factors
were found: (1) two fuel injectors leaked, one sprayed fuel continuously,
while the second leaked fuel at a rate three times faster than Bosch
specifications. Two new injectors were installed. (2) The spark plugs
(which had accumulated a total of 3830 km) were marginal on a firing
voltage breakdown test and were replaced. (It is interesting to note
that the spark plug from the cylinder with the leaking injector was
fouled, confirming the failure of the injector). And (3) an electrical
lead between the thermo-time switch and the cold start injector was
loose, and was giving intermittant contact. Thus on cold start, the
injector necessary to supply additional enrichment may not have operated.
Following these repairs the vehicle started smoothly, and the feedback
control system was able to maintain A/F control. Emissions levels for
the FTP, however, were above those seen before the 800 km shakedown
accumulation. Testing revealed that the feedback control system's
oxygen sensor had begun to lose activity. Installation of a new oxygen
sensor reduced FTP emissions to previous levels (seen before the shakedown
run).
-------
- 39 -
Before initiating the 80,000 km durability accumulation the
TWC-9 catalyst was replaced, at the EPA's request, by an Engelhard
TWC-9B Monolithic three-way catalyst. FTP emissions remained well below
the 0 km contract targets for kilometer accumulation initiation, and are
found in Table 6-2.
TABLE 6-2
THREE-HAY CATALYST VOLVO EXHAUST EMISSIONS
0 KILOMETER TEST
FTP Emi ss i ons , 9/km
CCT ~HC~
80,000 km emissions targets 2.1 0.25 1.2
0 km emissions targets 1.0 0.12 0.6
Measured emissions: TWC-9B 1.0 0.09 0.41
A summary of the three-way catalyst Volvo emission data for the
entire kilometer accumulation is found in Table 6-3. Kilometer accumu-
lation was carried out using the modified AMA cycle and a commercial
unleaded fuel containing 312 ppm of sulfur, 0.01 g Pb/gal , and 0.70 mg
P/gal. Emission tests at the specified accumulation intervals were run
using Indolene fuel containing 299 ppm of sulfur.
Emissions tests were carried out at 2000 km, 6000 km, and
15,000 km accumulation points (as well as a preliminary test at 150 km).
The NOX levels during cold start FTP's remained fairly constant, while
CO and HC tended to increase with accumulated kilometers. CO emissions
at the 15,000 km point exceeded the 80,000 km maximum contract targets,
and FTP fuel economy was about 10 to 20 percent lower than for FTP tests
at lower kilometer distances. A check showed that the breaker point
dwell angle had increased from the desired 61° to 71° due to rubbing
block wear. This 10° change in dwell resulted in a substantially retarded
ignition timing. An oil and filter change, as well as a complete tune-up
was performed (spark plugs, ignition points, condenser) on the vehicle.
An FTP was run after the tune-up and showed that emissions and fuel economy
had returned to previous levels. Kilometer accumulation was resumed, and
cold start FTP's were run at approximately 27,000 km to determine if any
degradation in emission control had occurred which could be corrected
before the testing required at the 30,000 km accumulation interval. Based
on an average of two tests CO emissions had increased to about 2.8 g/km,
while HC and NOX had increased to the maximum target levels of 0.25 and
1.2 g/km respectively. Replacement of the oxygen sensor reduced CO and
HC emissions to 1.8 g/km and 0.20 g/km respectively, however, NOX remained
unchanged at the maximum level of 1.2 g/km. These results indicated that
the three-way catalyst had begun to lose its' NOX conversion activity.
-------
- 40 -
Emissions testing at the 30,000 km Interval confirmed the three-way
catalyst's loss of activity, and kilometer accumulation for the three-
way catalyst vehicle was stopped.
In general, sulfate emissions during the kilometer accumulation
ranged between less than 0.1 to 2.3 percent of the sulfur in the fuel,
depending on the test cycle being run. Conversion levels of fuel sulfur
to sulfate were highest on the FTP driving cycle and were in the range of
0.6 to 2.3 percent, while the levels at 96 km/hr cruise were the lowest,
ranging from less than 0.1 to a maximum of 0.8 percent. Sulfate emissions
for 96 km/hr cruise conditions ranged from 0.2 mg/km to 0.8 mg/km, and
were well below the contract target of 6.2 mg/km. In fact, sulfate emissions
for any driving cycle regardless of accumulated kilometers remained well
below the maximum contract target.
Although kilometer accumulation had to be terminated at 30,000 km
because of catalyst deactivation, the three-way Volvo demonstrated that low
levels of exhaust emissions could be attained without a sulfate emissions
penalty. It was decided to restart the three-way catalyst phase of the
contract work with a prototype 1977 Volvo "Lambda-Sond" vehicle to demon-
strate the full 80,000 km durability. A discussion of this second three-way
catalyst vehicle follows immediately in Section 7 of this report.
-------
TABLE 6-3
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST-EQUIPPED VOLVO
km
150
2 000
(Test Fuel
Date
2/19/76
n
11
11
11
2/24/76
it
11
11
2/25/76
3/2/76
11
»
n
"
"
11
11
11
i'
3/5/76
Test
FTP
SET 1
SET 2
FET
SET 3(a)
FTP (b)
SET 4
96-1
96-2
96-3
96.4
FTP (c)
FTP
SET 1
SET 2
FET
SET 3
SET 4
96-1
96-2
96-3
96.4
FTP (d)
Emi
CO
1.21
0.03
0.05
0.02
0.06
1.36
0.05
0.03
0.01
0.02
0.01
0.89
1.62
0.09
0.07
0.04
0.10
0.09
0.02
0.01
0.01
0.02
8.63
= Indolene with 299 ppm
Emissions,
ssions, g/km
HC
0.27
0.01
0.02
0.02
0.02
0.15
0.01
0.03
0.02
0.02
0.02
0.08
0.16
0.02
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.66
NOX
0.60
0.94
1.03
1.16
0.82
0.48
0.68
1.43
1.60
1.48
1.43
0.80
0.69
0.78
0.87
0.98
0.83
0.83
1.80
1.90
1.90
1.83
2.71
mgykm
S04=
1.7
0.8
1.0
0.9
0.8
1.6
0.7
0.4"^
0.6
0.8
0.6>
1.4
0.9
0.4
0.3
0.3
0.2
0.2
0.4^
0.3
0.4
0.5^
1.4
S02
21
61
61
72
52
31
63
\
67
J
32
32
53
46
51
57
51
I
\ 56
/ bb
1
(e)
S)
% Fuel S As
S04=
1.7
1.0
1.4
1.3
1.0
1.7
1.0
0.6^
0.8
1.1
0.8 _
1.6
0.9
0.5
0.4
0.4
0.3
0.3
0.6"^
0.5
0.5
0.7^
SO?
31
122
122
154
105
49
133
\
)142
J
52
47.9
109
95.8
116
121
109
|
)119
;ny
;
% Fuel S
Recovered
32.7
123
123
155
106
51
134
~\
) 143
(
j
54
48.8
109
96.2
116
121
109
"\
/ 120
/ 1 I \s
)
Fuel
Economy,
mpg
15.5
20.8
20.8
22.3
21.0
16.3
22.1
22.0
16.8
15.8
21.5
21.5
23.6
22.1
22.1
22.1
17.0
-pi
I
"(IT) Test series interrupted following this test due to dynamometer breakdown.
(b) Test series was resumed with a cold start FTP.
(c) Cold start FTP made with 400 mv. setting on oxygen sensor control compared to 450 mv. used for all
prior tests in this series. 400 mv. setting used in all subsequent tests.
(d) Run without catalyst
(e) Sample lost
-------
TABLE 6-3 (Continued)
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST-EQUIPPED VOLVO
(Test Fuel = Indolene with 299 ppm S)
km
6 000
Emissions,
Emissions, g/km
Date
3/17/76
"
11
11
"
11
11
"
"
Test
FTP
SET 1
SET 2
FET
SET 3
SET 4
96-1
96-2
96-3
96-4
CO
1.68
0.23
0.18
0.08
0.23
0.21
0.03
0.001
0.02
0.02
HC
0.13
0.01
0.01
0.13
0.01
0.02
0.002
0.0
0.0
0.01
NOx
0.79
0.88
0.96
1.14
1.09
1.04
1.81
1.68
1.79
1.62
mg/km
504"
1.5
1.0
0.4
0.9
0.8
0.7
0.5^
0.2
0.4
0.2
S02
32
54
42
57
51
50
^
57
)
% Fuel
S04=
1.6
1.4
0.6
1.5
1.1
1.0
0.7 ^
0.3 i
0.5 j
0.4 J
S As
SOp
55.1
117
115
137
103
103
) 126
i
% Fuel S
Recovered
56.7
118
116
139
104
104
^
> 126
J
Fuel
Economy,
mpg
17.7
22.4
23.1
24.9
20.9
21.3
23.0
6 325 3/18/76 Oil, Oil Filter Changed, PI
15 000 3/31/76 FTP (f)
SET 1
SET 2
FET
SET 3
SET 4
96-1
96-2
96-3
96-4
2.86
1.10
0.93
0.62
1.00
1.05
0.73
0.79
0.75
0.95
0.19
0.02
0.02
0.02
0.03
0.02
0.004
0.001
0.002
0.004
ugs Replaced
0.78
0.78
0.76
0.53
0.73
0.71
0.82
0.77
0.65
0.65
2.2
1.0
0.6
0.7
0.5
0.5
0.7^
0.3 1
0.3 ,
0.2 J
66
64
58
55
48
41
\ 64
/ O't
1
1.9
1.2
0.7
0.9
0.6
0.6
1 .0 ">
0.4 I
0.4 /
0.2 J
85.2
119
105
108
86.5
72.6
11 9((
\ L.Q
87.1
120
106
109
87.1
73.1
126
13.5
19.4
19.0
20.5
18.8
18.6
20.3
-P.
ro
-------
TABLE 6-3 (Continued)
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST-EQUIPPED VOLVO
(Test Fuel = Indolene with 299 ppm S)
Emissions,
Emissions,
km
15 400
15 410
15 420
26 780
26 840
30 000
Date
4/2/76
4/4/76
4/6/76
5/7/76
5/10/76
5/12/76
5/13/76
6/17/76
6/17/76
6/17/76
6/17/76
6/17/76
6/17/76
6/17/76
6/17/76
6/17/76
6/17/76
6/17/76
Test
FTP (f)
Oil, Oil
FTP (g)
FTP
FTP
FTP (h)
FTP
FTP
SET-1
SET-2
FET
SET-3
SET-4
96 km
96 km
96 km
96 km
FTP
CO
2.66
HC
0.16
Filter Changed,
2.05
2.94
2.58
1.95
1.71
1.28
0.38
0.34
0.15
0.35
0.33
0.17
0.30
0.327
0.541
1.90
0.22
0.28
0.22
0.21
0.19
0.20
0.04
0
0.03
0.05
0.05
0.012
0.012
0
0.22
0.19
_a/km
NO*
1.09
Plugs and
0.80
1.25
1.06
1.30
1.17
1.19
1.35
1.40
1.31
1.31
1.28
1.80
1.74
1.96
1.77
1.04
mg/km
so4-
1.4
Points Repl
0.6
2.1
2.3
0.9
1.0
0.9
0.4
0.4
0.6
0.5
0.4
0.5^
0.3 I
0.3 f
0.2 }
1.0
S02
43
aced,
(e)
65
34
64
52
56
65
54
28
55
55
62
48
%Fuel
S04=
1.2
Timing
0.6
2.0
2.3
0.8
1.0
0.9
0.5
0.5
0.9
0.6
0.6
O.T>
<0.1 1
<0.1 j
<0.1 J
1.0
S As
S02
55.3
and Dwell
93.6
51.1
95.1
77.3
85.6
135.6
114.6
64.0
114.9
117.1
>
^133.3
w*
72.6
% Fuel S
Recovered
56.4
Set
95.6
53.4
95.8
78.3
86.5
136.1
115.1
64.9
115.5
117.7
^
\ 133.4
/
73.6
Fuel
Economy,
mpg
13.4
15.4
15.1
15.4
Sample lost.
(f) A second FTP was carried out on 4/2 at the 15 000 km point because the cycle follower stuck several
times during the initial FTP on 3/31 which made the run about 2 minutes longer than normal.
(g) This FTP was carried out after tune-up.
(h) A new oxygen sensor was installed prior to this test.
15.
15.
15.
21.
21.
24.
21.
22.0
22.4
15.9
CO
-------
- 44 -
SECTION 7
THREE-WAY CATALYST VEHICLE - DESIGN NO. 2
7.1 DESIGN AND FABRICATION
A prototype 1977 Volvo with 2.0 litre overhead camshaft L-4
cylinder engine and "Lambda Sond" three-way catalyst system was selected
to demonstrate 80,000 km of durability under the contract modification.
Through EPA negotiation, the vehicle was supplied to ER&E for testing by
Volvo of America. The Lambda Sond system, which is supplied to Volvo by
Bosch, is an upadated and improved version of the feedback fuel injection
that was installed on the first Volvo three-way catalyst demonstration
vehicle. It uses electronic feedback control with an oxygen sensor to
modulate the vehicle's fuel injection system. The desired A/F ratio is
maintained by varying the amount of fuel injected into the engine. The
Volvo Lambda Sond System is virtually unchanged from the Bosch feedback
fuel injection system previously described in both system components,
and in system operational logic. See Figure 7-1. Rather than repeat
the description already given, the reader is redirected to the previous
section (6.1), as well as to Volvo's technical literature (10,11) for
complete details of the Lambda Sond System.
The only modification made to the vehicle was the substitution
of a "mine-mix" three-way catalyst containing about 5% rhodium in place
of the original certification catalyst, about 17% rhodium, supplied with
the vehicle.
7.2 TEST RESULTS
The vehicle passed 0 km FTP emissions targets for HC and NOX
with both the certification and mine-mix catalysts. Although FTP CO
emissions (with both catalysts) were slightly above the 0 km target
level, they were well below the 80,000 km maximum, and durability
testing was begun. See Table 7-1.
Run
FTP#1
FTP#2
FTP#3
Targets
TABLE 7-1
PRELIMINARY (0 km.) FTP EXHAUST EMISSIONS
Catalyst Type Emissions (g/km)
Certification
Mine-Mix
Mine-Mix
0 km
80,000 km
CO
1.39
1.22
1.55
1.00
2.10
HC
,11
.13
12
.12
0.25
NOX
0.10
0.13
0.20
0.60
1.20
-------
- 45 -
line
pressure return
to
tcnk
air and fuel control unit
>xygen sensor
FIGURE 7-1
VOLVO LAMBDA-SOND FEEDBACK CONTROL
FUEL INJECTION SYSTEM
-------
- 46 -
Kilometer accumulation was carried out using the modified AMA
cycle and a commercial unleaded fuel containing 312 ppm of sulfur, 0.01
g Pb/gal, and 0.70 mg P/gal. Emission testing was carried out using
Indolene containing 299 ppm of sulfur. The test sequence consisted of
the Federal Test Procedure (FTP), two Sulfate Emissions Tests (SET), a
Highway Fuel Economy Test (HWFET), two SET's, and two hours of 96 km/hr
steady state cruise. In addition to the 80,000 km durability testing
for regulated emissions, sulfate (S04=), and sulfur dioxide (SO?), the
contract modification added hydrogen cyanide (HCN) and ammonia (NHs) to
the test list of unregulated pollutants.
Routine maintenance was performed according to manufacturer's
recommendations and was scheduled with respect to the absolute mileage
of the engine, which was 9,520 km (5,916 miles) when the catalyst was
changed and testing began. Table 7-2 contains a compilation of all
scheduled and unscheduled maintenance performed to the vehicle.
In addition to routine maintenance a number of repairs were
made. At 6,000 km a new water pump was installed because the one on the
vehicle was making noise. It was decided to adopt a policy of preventative
part replacement and maintenance, rather than wait for a total failure.
That is, in the case of the water pump, to replace the noisy unit before
outright failure occurred. For the same reason the cooling fan clutch
was replaced at 15,000 km when it began to make noise and show signs of
excessive vibration.
At 37,814 km, during a routine check, it was noticed that the
vehicle had developed an unsteady idle condition. Engine idle speed
varied by about 300 rpm about the set-point, with stable periods of one
to two minutes. Inspection of the fuel injection system revealed two
cracked fuel injector retainers. The idle speed cycling that was observed
was a result of rich-to-lean A/F excursions of the Lambda-Sond feedback
control system trying to compensate for air leakage at the fuel injectors.
Volvo of America replaced the fuel injectors and their retainers, and
recalibrated idle speed and carbon monoxide settings to manufacturers
specifications. The oxygen sensor was also replaced after installation
of the new injectors as a precaution against the possibility of sensor
burn-out during calibration adjustments. The catalyst was also removed
from the vehicle during system troubleshooting and recalibration to
protect it from damage.
At 60,000 km vehicle kilometer accumulation was stopped for
emission testing. At this time it was discovered, after noting poor
driveability and abnormally high FTP hydrocarbon levels, that one spark
plug was completely bridged by a piece of engine deposit. Exhaust
temperature traces before and after the catalyst indicated that the
spark plug bridging occurred at approximately 56,200 km. The problem
was not discovered until 3,800 km later because of a four day (holiday)
weekend. During the misfire condition, the catalyst overtemperature
safety did not trip. It was set for engine shutdown at 1550°F which
Engelhard Industries (the catalyst manufacturer) specified as a safe
upper limit catalyst operating temperature. Before the misfire condition
-------
- 47 -
occurred, the temperature gradient across the catalyst (outlet-inlet gas
temperatures) was approximately 400°F. During misfire this gradient
increased to 650°F. Inlet temperatures to the catalyst during misfire,
which were generally lower than those before misfire, varied over a
range from 550°F to a maximum 850°F. The exit temperatures during
misfire varied from 1150°F to a maximum of 1450°F. These exotherms did
not change from the onset of plug misfire to the discovery of the problem
3,800 km later. The fact that catalyst exotherms were unchanged throughout
misfire, coupled with a maximum catalyst temperature one hundred degrees
below a safe upper limit specification, was an assurance that catalyst
damage did not take place. However, as a precaution, the oxygen sensor
was changed because of its' possible sensitivity to a misfire condition.
Near the end of the 60,000 km test sequence two more malfunctions
occurred. The vehicle began operating in an open loop mode and would
not control A/F ratio. It was found that the feedback control system
wires that lead to the frequency control solenoid valve had been cut and
shorted by rubbing against the fuel return line. The wires were replaced
and rerouted away from any contact points to prevent future reoccurence.
With the vehicle operating properly the 60,000 km test sequence was
completed. At this time a fluid leak from the automatic transmission
was noted. The transmission kickdown cable and gasket were replaced,
and the old transmission fluid was flushed and replaced. The vehicle
was then sent out to complete the durability accumulation.
Finally, at 77,806 km, the right rear axle bearing seized and
caused the right rear axle shaft to break. No apparent damage occurred
to the rear axle carrier housing, so the shaft and bearings were renewed.
After repairing this failure, kilometer accumulation continued to the
full 80,000 km with no further trouble.
A summary of the emission test data for this vehicle during
the 80,000 km durability accumulation is found in Table 7-3. Emissions
tests were performed at accumulation points of 2,182 km, 6,000 km,
15,000 km, 30,000 km, 60,000 km, and 80,000 km (as well as an initial
test at 0 km). In general, CO and NOX emission levels on the FTP driving
cycle increased with accumulated kilometers. HC emissions, though not
as severely affected, were also increased by the end of durability
testing. The vehicle exceeded the maximum FTP CO contract target (2.1
g/km) at 15,000 km, with CO increasing to 3.36 g/km at the final 80,000
km test point. HC and NOX emission levels, however, remained below
contract targets throughout the entire durability accumulation. A test
run made without the three-way catalyst mounted (engine-out emissions)
at 80,000 km showed that the catalyst was still more than 60% active for
CO oxidation, with HC and NOX conversion levels of 80% and 60%, respectively.
-------
- 48 -
It is not felt that the deterioration seen in the performance
of the three-way catalyst was related to, or influenced by, the engine
misfire condition that occurred at 56,200 km. As shown in Figure 7-2,
the emissions of all three regulated pollutants increased smoothly with
accumulated kilometers. The fact that there was no significant emissions
discontinuity at the 56,200 km point, in addition to the constant catalyst
exotherms during misfire (mentioned previously) are excellent assurances
that the emissions degradation seen was due solely to catalyst deterioration
with increasing kilometers.
-------
- 49 -
Table 7-2
MAINTENANCE PERFORMED TO THREE-WAY CATALYST
VOLVO DESIGN #2
km Durability 0 km = 5,916 miles Services, etc.
374 Oil and filter change
2 000 Adjusted toothed belt
Installed "625" 02 sensor
in lieu of "4209"
6 000 0 Installed new water pump
0 Replaced engine coolant
15 000 Installed new fan clutch
15 320 Oil and filter change
Installed new spark plugs
0 Adjusted toothed belt
20 134 0 Replaced oil fill cap gasket
30 353 0 Installed new 0? sensor
37 814 0 New injectors and retainers,
new 02 sensor
42 294 0 Oil and filter change
56 134 0 Oil and filter change
60 000 0 New plugs, new 02 sensor,
Fixed fuel regulator and replaced
connections to fuel return lines
and fuel distributor
60 400 0 Repair kick-down cable, drain
and recharge transmission.
62 534 0 New timing belt, plugs, evaporative
control filter
65 972 0 Oil and filter change
77 806 0 Repaired or replaced right rear
axle, brake lining, etc. as
necessary
-------
TABLE 7-3
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST EQUIPPED VOLVO-DESIGN #2
(Test Fuel = Indolene with 299 ppm S)
Emissions,
Km Date Test
0 8/10/76 FTP
SET-1
SET-2
FET
SET-3
SET-4
96 km-1
96 km-2
96 km-3
96 km-4
2nd FTP
CO
1.220
0.099
0.111
0.080
0.071
0.072
0.059
0.129
0.136
0.153
1.550
HC
0.130
0.031
0.023
0.025
0.029
0.022
0
0
0.014
0.009
0.120
g/km
NOx
0.130
0.041
0.038
0.065
0.075
0.043
0.016
0.010
0.007
0.010
0.200
Emissions, mg/km
HCN NH3 SOiT
<1.4 42 2.3
<1.2 28 0.84
<1.7 40 0.69
<1.2 18 0.76
<1.2 25 0.68
<1.3 37 0.58
<0.15 6
0.27
0.62
2
0.96
0.42
<1.5 54 0.72
S02
6.3
31.9
68.2
45.6
47.3
108.1
48.9
30
% Fuel S As
so^i
2.32
0.12
0.09
1.19
0.93
0.82
0.95
0.71
S02
9.36
66.96
139.60
106.16
98.54
232.06
123.43
45.15
% Fuel S
Recovered
11.68
67.08
139.69
107.35
99.47
232.88
124.38
45.86
# Fuel Used
(Carbon
Balance)
4.36
3.84
3.94
2.60
3.88 ,
u
3.76 c
i
4.69*
4.68*
4.68*
4.76*
4.34
* 20 minute fuel consumption
-------
TABLE 7-3 (Continued)
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST EQUIPPED VOLVO-DESIGN #2
(Test Fuel = Indolene with 299 ppm S)
Emissions,
Km Date Test
2182 8/26/76 FTP
SET-1
SET-2
FET
SET-3
SET-4
96 km-1
96 km-2
96 km-3
96 km-4
8/27/76 2nd FTP
8/28/76 3rd FTP**
CO
2.189
0.475
0.449
0.480
0.420
0.387
0.201
0.195
0.137
0.127
1.351
1.861
HC
0.184
0.059
0.046
0.032
0.043
0.047
0.010
0
0
0
0.117
0.132
g/km
NOx
0.287
0.218
0.176
0.137
0.258
0.211
0.165
0.150
0.155
0.153
0.320
0.486
Emissions, mg/km °i
HCN
2.9
<1 .2
<1 .2
<1 .6
<1 .9
<1.8
<0.14
1.8
"
NH3
13
27
51
57
55
52
0.34
34
S0i,=
0.87
0.57
0.61
0.55
0.70
1.2
0.40
0.44
0.27
0.27
0.88
, Fuel S
As
S02 S04~ S02
38.2 0.
62.6 0.
48.0 0.
39.6 0.
44.3 0.
46.6 1.
50.5 0.
273.0
88 60
78 132
88 104
86 93
97 93
60 98
53 121
42
.12
.2
.21
.63
.88
.89
.84
.46
% Fuel S
Recovered
60.99
133.07
105.09
94.49
94.85
100.50
122.37
43.35
# Fuel Used
(Carbon
Balance)
4.29
3.88
3.78
2.57
3.87
in
3.86 ,
5.01*
5.11*
5.08*
5.05*
4.34
4.69
* 20 minute fuel consumption
** 02 sensor changed before run
-------
TABLE 7-3 (Continued)
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST EQUIPPED VOLVO-DESIGN #2
(Test Fuel = Indolene with 299 ppm S)
Emissions,
Km Date Test
6000 9/8/76 FTP
SET-1
SET-2
FET
SET-3
SET-4
96 km-1
96 km-2
96 km-3
96 km-4
CO
1.718
0.196
0.121
0.122
0.218
0.169
0.213
0.240
0.322
0.275
HC
0.115
0.026
0.035
0.015
0.027
0.024
0.016
0.004
0.026
0.022
g/km Emissions, rug/km
NOx HCN NH3 SO^ S02
0.314 <1.6 7
0.373 <1.2 15
0.243 <1.2 11
0.182 <1.6 20
0.300 <1.2 13
0.282 <1.2 16
0.189
0.196
<0.14 81
0.208
0.175
.9 1.51 30.9
.1 0.47 78.1
.4 0.30 50.2
.5 0.50 51.2
.6 0.49 50.1
.6 0.076 53.3
0.28 _
0.41
.7 55.7
0.40
0.46
% Fuel S As
SO.T
1.62
0.67
0.44
0.81
0.69
0.12
0.60
S02
81.08
169.91
112.39
127.79
109.28
118.25
133.24
% Fuel S
Recovered
82.70
170.58
112.84
128.60
109.96
118.36
133.84
$ Fuel Used
(Carbon
Balance)
4.05
3.77
3.66
2.46
3.76
3.86
5.19*
5.03*
5.11*
5.10*
in
tVJ
20 minute fuel consumption
-------
TABLE 7-3 (Continued)
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST EQUIPPED VOLVO-DESIGN #2
(Test Fuel = Indolene with 299 ppm S)
Emissions,
Km Date Test
15 000 9/22/76 FTP
SET-1
SET-2
FET
SET-3
SET-4
96 km-1
96 km-2
96 km-3
96 km-4
CO
2.508
0.362
0.413
0.303
0.414
0.414
0.108
0.094
0.094
0.094
HC
0.162
0.035
0.033
0.024
0.019
0.041
0.0
0.02
0.012
0.013
g/km Emissions, mg/km
NOx HCN
NH3 SO.T S02
0.375 <1.5 <7 1.6 88.4
0.290 <1.2 <6 0.66 73.4
0.272 <1.2 18 0.55 45.2
0.237 <1.5 14 0.59 44.2
0.326 <1.2
8 1.1 42.0
0.333 <1.2 23 0.56 44.9
0.308
0.301
0.2 1
0.295
0.280
0.30
0.20
6 52.6
0.19
0.21
% Fuel S As
S0i»
1.42
0.94
0.79
0.93
1.47
0.78
0.23
S02
120.20
160.47
99.21
105.81
88.27
95.34
123.61
# Fuel Used
% Fuel S (Carbon
Recovered Balance)
121.62
161.41
100.00
106.74
89.74
96.12
123.84
4.97
3.75
3.74
2.54
3.91
3.86 S
5.19*
5.28*
5.18*
5.16*
* 20 minute fuel consumption
-------
TABLE 7-3 (Continued)
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST EQUIPPED VOLVO-DESIGN #2
(Test Fuel = Indolene with 299 ppm S)
Emissions,
Km Date Test
30 000 10/12/76 FTP
SET-1
SET-2
FET
SET-3
SET-4
96 km-1
96 km-2
96 km-3
96 km-4
10/13/76 FTP
(2nd)**
3.
0.
0.
0.
0.
0.
0.
0.
0.
0.
2.
CO
000
802
640
245
882
690
132
095
096
097
327
HC
0.135
0.037
0.040
0.028
0.073
0.046
0.019
0.004
0.021
0.015
0.132
g/km
Emissions, mg/km
NOx HCN
0.
0.
0.
0.
0.
0.
0.
1.
1.
1.
0.
NH, SO.T
549 <1.4 12 2.0
562 <1.1
9 0.87
571 <1.1 22 0.50
660 2.7 15 0.46
668 <1.1 16 1.1
627 <1.2 23 0.51
974
050
<0.1
199
121
641 --
0.41
0.33
3
0.48
0.40
S02
26.8
49.2
34.3
41.9
34.8
24.2
55.7
_ _
% Fuel S As
SOiT
2.21
1.23
0.72
0.68
1.48
0.72
3.56
~ ~
S02
43.26
104.63
75.45
93.96
72.13
51.70
182.65
~ ~
% Fuel S
Recovered
45.47
105.86
76.17
94.64
73.61
52.42
186.21
~
# Fuel Used
(Carbon
Balance)
4.04
3.79
3.67
2.70
3-89 i,
->
3.77
5.45*
5.43*
5.52*
5.41*
4.17
* 20 minute fuel consumption
** after new Op sensor
-------
TABLE 7-3 (Continued)
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST EQUIPPED VOLVO-DESIGN #2
(Test Fuel = Indolene with 299 ppm S)
Emissions,
Km Date Test
60 000 12/9/76 FTP
SET-1
SET-2
FET
SET-3
SET-4
96 km-1
96 km-2
96 km-3
96 km-4
CO
2.727
0.747
0.751
0.679
0.841
0.706
0.663
0.771
0.788
0.724
HC
0.176
0.057
0.055
0.050
0.053
0.052
0.041
0.038
0.031
0.029
g/km
NOx
0.765
0.786
0.755
0.698
0.791
0.785
0.922
1.001
1.000
1.025
# Fuel Used
Emissions, mg/km % Fuel S As % Fuel S (Carbon
HCN
7.8
3.2
4.2
3.1
4.0
<2.2
2.3
NH3
6.8
10.5
22.1
31.6
13.4
2.7
54.1
'
SO., S02 SOu S02 Recovered Balance)
3.4 0.0** 3.88
1.1 56.61 1.62 121
0.93 -- ** 1.38
0.69 34.63 1.12 84
0.93 0.0** 1.37
3.81
79 123.41 3.75
3.61
50 85.62 2.48
3.66
0.76 24.07 1.13 53.64 54.77 3.62
0.60
0.27
52.09 0.13 180.
0.19
0.33
5.06*
5.10*
21 180.74
5.25*
5.17*
* 20 minute fuel consumption
** bubbler not good or no inflection in titration
01
en
-------
TABLE 7-3 (Continued)
SUMMARY OF KILOMETER ACCUMULATION EMISSION DATA FOR THREE-WAY CATALYST EQUIPPED VOLVO-DESIGN #2
(Test Fuel = Indolene with 299 ppm S)
Emissions,
Km Date Test
80 000 2/4/77 FTP
SET-1
SET-2
FET
SET-3
SET-4
96 km-1
96 km-2
96 km-3
96 km-4
CO
3.356
0.980
0.636
0.645
1.088
0.958
0.642
0.742
0.790
0.750
HC
0.200
0.066
0.049
0.054
0.070
0.056
0.047
0.046
0.046
0.043
g/km
NOx
0.954
1.037
1.106
1.080
1.163
1.085
1.533
1.573
1.573
1.613
Emissions, mg/km
HCN
4.21
17.50
2.02
1.48
4.21
4.30
0.31
NH3
30.45
14.22
5.56
3.56
40.59
46.92
43.11
so,-
8.072
0.54
0.58
0.43
0.61
0.29
0.20
0.45
0.063
0.087
S02
29.0
33.9
50.8
71.8
30.9
32.8
44.0
% Fuel S As
SO.T
9.84
0.83
0.88
0.71
0.87
0.43
0.14
S02
52.98
78.44
116.42
176.49
65.86
73.82
145.80
% Fuel S
Recovered
62.82
79'. 2 7
117.30
177.20
66.73
74.25
145.94
# Fuel Used
(Carbon
Balance)
3.57
3.49
3.52
2.46
3.79
3.58
5.37*
5.37*
5.37*
5.37*
in
i
2/4/77
FTP**
8.899 0.888 2.301
3.83
* 20 minute fuel consumption
** without a catalyst
-------
- 57 -
A detailed examination of tailpipe CO emissions is also useful to illustrate
the three-way catalyst's deactivation history. FTP CO emissions began
to increase steadily starting at the 6,000 km accumulation point (well
in advance of the misfire condition). The percent increase in CO emissions
(a function of catalyst deactivation) within each successive accumulation
interval, as well as a normalized percent increase per 1,000 km (which
normalizes the varying kilometer length of each accumulation), is presented
in Table 7-4.
TABLE 7-4
FTP CO EMISSIONS: THREE-UAY CATALYST VEHICLE - DESIGN #2
Accumulation
Interval
(km)
6000
15000
30000
30000*
60000
80000
FTP CO
Emissions
(g/km)
718
508
000
2.327*
2.727
3.356
AGO Emissions
During Interval
+45.98
+19.62
+17.19
+23.07
Normalized AGO
Emissions
(A%/1000 km)
+5.11
+1.31
+0.573
+1.15
*FTP after new Op sensor installed
The largest CO increase seen was during the accumulation interval from
6,000 km to 15,000 km. Successive intervals show continued catalyst
deactivation for CO oxidation, but at a reduced rate. It should be
noted that the deactivation (normalized percent) for the 60,000 to
80,000 km interval after the misfire is lower than those of the 6,000
km-15,000 km and 15,000 km-30,000 km intervals. That CO emissions did
not increase abnormally after the misfire occurrence further reinforces
the fact that the emissions increases seen were due to catalyst deactivation
with accumulated kilometers.
Sulfate emissions during the kilometer accumulation ranged
between less than 0.1 to 9.8 percent of the sulfur in the fuel, depending
on the test cycle being performed. Conversion of fuel sulfur to sulfate
was highest during the FTP driving cycle and was in the range of 0.88 to
9.84 percent, while conversion levels at 96 km/hr cruise were generally
less, ranging from 0.13 to 3.56 percent. Sulfate emissions for 96 km/hr
steady state cruise conditions were in the range of 0.06 to 0.96 mg/km,
and were well below the contract target of 6.2 mg/km.
-------
1.6 -
NOX '*
(g/km)
0.8
0.4
0.0^
1
0.30
HC °'20
(g/km)
0.10
0.0
3.0
CO *'°
(g/km) ^
1.0
0.0
80KKM CONTRACT TARGET
^^A- ^^A
i i i i i i i i
3 10 20 30 40 50 60 70 80
80KKM CONTRACT TARGET A
,A-A- * A *
D 10 20 30 40 50 60 70 80
A
./£± 80KKM CONTRACT TARGET
. ACCUMULATED KILOMETERS (XI 000)
l i l i i i 1 1
0 10 20 30 40 50 60 70 80
FIGURE 7-2
REGULATED EMISSIONS VS. ACCUMULATED KILOMETERS
CJl
CXI
THREE-WAY CATALYST VEHICLE DESIGN NO. 2
-------
- 59 -
Emissions of ammonia (NHs) and hydrogen cyanide (HCN) remained
low throughout the 80,000 km of durability testing. Ammonia emission
rates ranged from less than 0.34 mg/km to a maximum of 81.7 mg/km.
Hydrogen cyanide emission rates were in the range of less than 0.1 mg/km
to a maximum of 17.5 mg/km. There was no apparent correlation for
either NH3 or HCN emission rates with the type of driving cycle performed,
or with accumulated kilometers. There was, however, an increase in HCN
emissions at the 60,000 km and 80,000 km test intervals, as compared to
previous test points. No other emission increased as dramatically at
these test points, and there is no obvious explanation for the discon-
tinuous increase in HCN emission rate observed.
In conclusion, the 1977 three-way catalyst Lambda-Sond Volvo
completed the full 80,000 km of durability testing. The vehicle met all
of the 80,000 km contract targets, except for CO. It is interesting to
note that although the vehicle exceeded the 80,000 km CO contract targets,
it met the 1977-1979 California certification targets of 5.6 gpk CO,
0.26 gpk HC, and 0.95 gpk NOX, and with a mine-mix catalyst. Emissions
of NHs and HCN were also monitored during durability testing, and in
general remained low. Testing of this vehicle confirmed that three-way
catalysts were a viable method of automotive emission control. Three-
way catalyst systems can control regulated pollutants (CO, HC and NOx)
to stringent levels and emit low levels of unregulated pollutants such
as sulfate.
-------
- 60 -
SECTION 8
LOW EXCESS AIR VEHICLE - DESIGN #2
8.1 DESIGN AND FABRICATION
A new low excess air vehicle was designed and tested under the
contract modification. A 1977 California emissions certified Ford Pinto
(2.3 litre overhead camshaft L-4 cylinder engine, backpressure modulated
EGR, alrpump and oxidation catalyst) was modified for low excess air
operation with a Holley closed-loop feedback controlled carburetor with
oxygen sensor, and an Engelhard monolithic oxidation catalyst. The
advantages of this particular design were (1) the Installation of the
Holley vacuum modulated feedback carburetor system on the 1977 Pinto
would require a minimum of mechanical modification, since 1978 California
three-way catalyst Pinto's were to be built by Ford with the Holley
system as production equipment, (2) Holley agreed to modify the electronic
control unit for the feedback carburetor so 1t could be set for low
excess air conditions, rather than for three-way catalyst operation, (3)
Holley agreed to furnish technical assistance as required, and (4)
Engelhard agreed to supply a monolithic catalyst of suitable size and
"composition" to provide good conversion activity under low excess air
conditions.
A schematic of the Holley closed-loop feedback controlled
carburetor system is found in Figure 8-1. As was described for the
previous feedback systems, a desired A/F ratio is maintained by varying
the amount of fuel delivered to the engine. In the Holley system,
feedback controlled airbleeds and main fuel jets are modulated by an
electronic control unit in response to changing signals from an oxygen
sensor (located in the exhaust stream). A major difference of the
Holley system compared to other feedback control fuel metering systems
is that it retains the carburetor as the basic fuel metering device.
Feedback carburetor metering elements are modulated by a control vacuum
signal. An electro-pneumatic Interface, the vacuum control regulator
valve, converts the electronic signals from the ECU to control vacuum
levels that set the carburetor's airbleeds and metering jets to give the
desired A/F ratio. A high ECU "on-time" (high duty cycle) results 1n an
Increased control vacuum level, closing off the feedback main jets and
leans the A/F ratio. A description of the feedback control system's
logic and operating sequence is found in Figure 8-2. A schematic diagram
of the feedback carburetor and vacuum control regulator valve is shown
in Figure 8-3.
In addition to the exhaust gas oxygen sensor, the ECU receives
input signals from two other sensors. A throttle operation sensor
signals the ECU to provide a rich A/F ratio during wide open throttle
operation for power enrichment. An engine coolant temperature sensor
switches the ECU to operate 1n an "open-loop" mode during cold start
operation. In this mode ECU signals are preset to a level that enriches
the A/F ratio and gives enchanced cold drlveabiHty. Spurious signals
-------
VACUUM REGULATOR-
CONTROL VALVE
ELECTRONIC
CONTROL UNIT
VACUUM
STORAGE-
CANISTER
FEEDBACK
CARBURETOR
r*\
THROTTLE
^OPERATION
SENSOR
CATALYST
\
ENGINE WATER
TEMPERATURE
SENSOR
AIR PUMP
en
FIGURE 8-1
HOLLEY FEEDBACK CARBURETOR SYSTEM
PICTORIAL DIAGRAM
-------
ENGINE
OPERATING-*
CONDITION
RICH
of
Stoichiometry
OXYGEN
-SENSOR
OUTPUT
High
Output
Voltage
(>1.0
Volt)
-*- ECU -*-
Directs
Vac. Sol. -Reg.
To Greater
"On Time"
(50- -^100%)
VACUUM
SOLENOID
REGULATOR
VALVE
Increased
Duty Cycle
Results
In Higher
Output
Vacuum
VACUUM SIGNAL POSITION
TO
-*- CARBURETOR
F.B. CIRCUIT
>2.5in. Hg
Signal Pulls
Metering
Rod Up
(Smaller Orifice)
OF FUEL
->- METERING-*- FLOW ->-
ROD
Higher Decreased
Position
For Decrease
In F.B.
Fuel Flow
RESULTANT
CORRECTIONS
Toward
LEAN
LEAN
of
Stoichiometry
Low
Output
Voltage
(<0.35
Volt)
Directs
Vac.Sol. Reg.
To Less
"On Time"
(50%»-0)
LU
OXYGEN
SENSOR
(>2.5 in. Hg)
Decreased
Duty Cycle
Results
In Lower
Output
Vacuum
«2.5 in. Hg)
<2.5 in. Hg
Signal Moves
Metering
Rod Down
(Larger Orifice)
O
.N
N
RPM
RICH
LEAN
DUTY CYCLE
VACUUM CONTROL
REGULATOR
Lower
Position
For Increased
F.B. Fuel
Flow
, FUEL
H /
/////// x ///////
'/ »
Increased
Toward
RICH
en
ro
\\x\x\\vi\\\\\\\\
METERING ROD
FIGURE 8-2
HOLLEY FEEDBACK CARBURETOR SYSTEM OPERATIONAL LOGIC
-------
FIXED IDLE
AIR BLEED
FEEDBACK CONTROLLED
IDLE AIR BLEED
CONTROL VACUUM
CONNECTION
COIL
ARMATURE
MAIN METERING JET
FEEDBACK CONTROLLED
MAIN SYSTEM FUEL
ARMATURE RETURN
SPRING
VENT
CO
I
FIGURE 8-3
HOLLEY FEEDBACK CARBURETOR AND VACUUM
CONTROL REGULATOR VALVE
SOLENOID
ZJOUTPUT TO CARBURETOR
iVACUUM INPUT
:J(MANIFOLD VACUUM)
VACUUM
REGULATOR
-------
- 64 -
from t!ie oxygen sensor, which has not yet heated up to operating temperature,
are disregarded. When the engine coolant temperature reaches approximately
38°C (100°F) the sensor opens and the ECU assumes closed-loop control.
Signals from the oxygen sensor are then used to maintain the engine A/F
ratio at the desired value. A more detailed treatment of the Holley
system's operation can be found in reference (12).
Holley modified the electronic control unit, at ER&E's request,
so that the control reference voltage (above which a lean correction
signal is generated/below which a rich correction signal is generated)
could be set at any value desired. Typical oxygen sensor voltage response
is 1n the overall range of 0 to 1000 millivolts, with approximately 500
mv. representing a stoichiometric A/F ratio. The oxygen sensor's charac-
teristic response curve is found in Figure 8-4. Bv setting an ECU
reference voltage of 500 mv (point C on Figure 8-4), a stoichiometric
A/F ratio would be maintained, since any sensor output >500 mv drives
the system lean and <500 mv drives the system rich. Thus, by a proper
choice of the ECU reference voltage the A/F ratio can be biased slightly
lean or rich of stoichiometric. For example a reference level of ^750
mv (A) would result in an A/F slightly rich of stoichiometric (any time
the sensor voltage is <750 mv the ECU would send out a rich correction
signal, and thus would spend a greater percentage of it's time rich),
while a reference level of M50 mv (B) would result 1n a slightly lean
A/F ratio. A practical limit to the range of adjustability is when the
reference voltage reaches either "shoulder" of the oxygen sensor curve.
Sensitivity (change in mv output for A/F change) becomes low on the
shoulder as compared to the almost infinite slope near the Inflection
point, making A/F control very unstable and imprecise.
Uhen the feedback carburetor system was first installed on the
vehicle it would not control A/F ratio. Troubleshooting revealed that
the vacuum control regulator valve (VCRV) was not providing a control
signal to the carburetor's feedback input port. Inspection of the
valve by Holley revealed that 1t had become plugged with rubber particles.
It was suspected that as rubber vacuum tubing was forced on the valve's
ports Us inner surface was abraded. The resultant rubber dust plugged
the regulator's airbleeds when vacuum was applied, rendering the valve
inoperative.
Upon installation of a new VCRV, baselining of the feedback
carburetor system began. A correlation between the electronic control
unit's reference voltage setting and A/F ratio was determined and the
ECU was biased "slightly lean" for low excess air operation. Previous
work done at ER&E had shown that the "break-point" of low sulfate
production with high levels of CO and HC conversion occurred over an
oxidation catalyst at about 0.5% oxygen in the exhaust gas. As seen in
Table 8-1, a setting of "200" on the reference voltage potentiometer
(arbitrary 0-1000 scale on a ten-turn helipot) resulted in the desired
after-catalyst oxygen level, and corresponded to an A/F ratio of approx-
imately 15.2.
-------
- 65 -
UJ
t/)
z:
o
Q.
LO
UJ
LU
o:
o
CO
z
LU
CO
X
o
lOOOr
14.514.7 14.9
A/F RATIO
RICH
STOICHIOMETRIC
LEAN
FIGURE 8-4
OXYGEN SENSOR VOLTAGE RESPONSE CURVE
-------
- 66 -
Table 8-1
REFERENCE VOLTAGE POTENTIOMETER SETTING VS. A/F RATIO
40 mph - Steady State Conditions, Air Pump Off
Potentiometer Setting A/F Ratio 0?% After-Catalyst
490 13.84 0
470 14.10 0
400 14.65 0
300 14.95 0.15
250 15.05 0.20
200 15.20 0.51
100 15.35 0.65
0 15.75 0.9
A check was also made to insure that as vehicle speed varied, carburetor
stoichiometry remained close to the desired set-point. As seen in Table
8-2, the A/F ratio variation with speed was slight, indicating the
feedback system was tracking properly.
Table 8-2
A/F RATIO VS. VEHICLE SPEED
Vehicle Speed, mph A/F Ratio
0-Idle 14.90
20 15.15
30 15.20
40 15.20
50 15.35
8.2 TEST RESULTS
FTP emissions with the feedback system installed were an order
of magnitude higher for CO, when compared to the stock "as received"
vehicle. HC and NOX emissions were not as severely affected. It should
be noted that installation of the feedback carburetor system increased
FTP fuel economy by almost 0.5 km/liter (1 mpg) as shown in Table 8-3.
Both tests were conducted with the production Ford catalyst mounted on
-------
- 67 -
the vehicle. The low excess air vehicle used air injection only on cold
start to aid catalyst light-off, while the production vehicle had continuous
air injection to promote oxidation over the catalyst. It was suspected
that the lack of continuous air injection during the FTP cycle contributed
to the increase in CO and HC emissions seen for the feedback carbureted
vehicle.
TABLE 8-3
FTP EMISSIONS (g/km)
(Average of Three Tests)
Vehicle With Pro- Vehicle With
duction Carburetor Feedback Carburetor
Air Pump Operating 1 Min Air Injection, ECU @ 200
CO 0,60 2.53
HC 0.12 0.18
NOX 0.92 1.25
Fuel Economy 9.9 (23.6 mpg) 10.3 (24.4 mpg)
(km/liter)
Raw emissions concentrations indicated that almost all of the
CO produced during the test ( 85%) was being emitted during the FTP's
first bag. Re-setting the automatic choke for lean operation (faster
opening), however, did not significantly reduce first bag CO emissions.
To check the electronic control units' open-loop cold start enrichment
circuit a switch was installed so that the feedback system could be
placed in a cold start mode at any time desired. With a fully-warmed up
engine engaging the cold start enrichment circuit resulted in a lean A/F
ratio of about 16.0/1, rather than giving A/F enrichment. Driveability
of the vehicle was poor due to engine misfire from the lean mixture. CO
and HC emissions were increased while operating in this open-loop mode.
The electronic control unit was reset slightly rich (A/F = 14.5) to
augment normal choke action during cold start operation. Recallbration
of the electronic control unit's open-loop circuit coupled with the
leaned choke reduced FTP Bag 1 CO emissions. However the total weighted
CO emissions remained above 0 km targets as shown in Table 8-4. Further
adjustment would not lower CO emissions.
-------
- 68 -
TABLE 8-4
FTP Emissions (g/km)
RECALIBRATED COLD START CIRCUIT
ECU @ 200, 1 MIN. AIR INJECTION
(Average of Three Tests)
CO
HC
NO.
Production Ford Catalyst
1.50
0.12
1.26
EPA 0 KM. Contact Targets
1.00
0.12
1.20
Catalyst temperature traces indicated that the production Ford
catalyst took over three minutes to light-off during cold start operation.
The Ford monolith was removed and replaced by a fresh Engelhard PTX-514
monolith (a known active oxidation catalyst) so that light-off times for
the two could be compared. Light-off time for the PTX-514 was approximately
one minute, which confirmed the suspicion that the Ford catalyst had
deactivated. Most importantly, Bag 1 CO emissions were decreased by a
factor of three, while total weighted FTP CO and HC emissions were below
contract targets. See Table 8-5. Repeated FTP tests insured that the
low emissions seen were not just a result of the "edge" of a fresh
catalyst. It was felt that the Ford catalyst deactivated during the
calibration of the feedback carburetor system's cold start circuits
(choke duration, electronic control unit A/F enrichment circuit, air
pump injection time). The A/F ratio during these tests became very rich
(<13/1) and air was being injected to the catalyst. It was suspected
that the high heat levels generated from oxidizing the CO deactivated
the catalyst sufficiently to produce slow light off.
TABLE 8-5
FTP EMISSIONS (g/km)
ENGELHARD PTX-514 MONOLITH
ECU @ 200, 2 MIN. AIR INJECTION
(Four Test Average)
CO
HC
NOX
Fuel Economy
(km/liter)
0.72
0.11
1.89
10.3 (24.4 mpg)
-------
- 69 -
Although the vehicle met CO and HC 0 km contract targets, NCL
emissions exceeded the 80,000 km maximum limits. The NOX increase was
attributed to the PTX-514 catalyst's different composition. That is,
during rich transients the PTX-514 reduced less NOX than did the pro-
duction Ford catalyst, raising tailpipe NOX levels. Adding support to
this argument was the fact that raw exhaust (no catalyst) FTP data
showed higher levels of NOX than with the PTX-514 mounted. To bring NO
emissions back to within acceptable levels, the EGR recycle rate was
increased. Connecting the EGR system to the carburetor's "spark port"
(instead of the "EGR port" normally used) gave higher amplitude vacuum
signals at the EGR valve. This pulled the valve further open, and
increased EGR rate. This solution avoided complicated machining of the
valve body, which could have resulted in a leaky EGR valve. It should
be noted that even though vacuum amplitude was increased, EGR vacuum
modulation depended only on exhaust backpressure. Since the backpressure
transducer was not modified, the schedule of EGR valve operation was
unchanged, although recycle rate was increased. Replicate FTP testing
showed that NO* was reduced by this modification, HC and fuel economy
were unaffected, and CO was increased. Sulfate emissions for these runs
were below the 6.21 mg/km contract target, as shown in Table 8-6.
Driveability (as perceived during emissions test runs) suffered noticeably
from the increased EGR rate, particularly during accelerations where
engine roughness and surging occurred.
TABLE 8-6
FTP EMISSIONS (g/km)
ENGELHARD PTX-514 MONOLITH
ECU @ 200. 2 MIN AIR INJECTION. SPARK PORT EGR
FTP A FTP B
CO 1.33 1.31
HC 0.12 0.09
NOX 1.15 1.12
Fuel Economy 10.3 (24.4 mpg) 10.4 (24.8 mpg)
(km/liter)
$04 = (mg/km) 4.1 3.2
Although FTP CO emissions were above 0 km contract targets,
they were well below the 80,000 km maximum limits, and the EPA gave
approval for initiating kilometer accumulation. The PTX-514 catalyst
(68 in3) used during vehicle optimization and baseline testing was
replaced by the PTX-516 (102 in3) monolithic catalyst purchased for the
contract work. It was felt that the larger catalyst volume would better
offset the effects of aging and deterioration, making attainment of the
80,000 km goals more certain. Out of the two experimental preparations
purchased, Engelhard recommended Serial #112178-029 for our application.
With this catalyst the vehicle met all 0 km contract targets and was
sent out for the first 2,000 km of durability testing. See Table 8-7.
-------
- 70 -
TABLE 8-7
FTP EMISSIONS (g/km)
ENGELHARD PTX-516 MONOLITH
ECU @ 200,~2 MIN. AIR INJECTION.~SPARK PORT EGR
(Single Test)
CO 0.70
HC 0.08
NOX 0.91
Fuel Economy 10.21 (24.3 mpg)
(km/liter)
S04 = (mg/km) 2.6
The Holley feedback carburetor system was plagued with operating
and durability problems from the time 1t was Installed. The vacuum
control regulator valve (VCRV) which Interfaces the ECU to the feedback
carburetor was extremely susceptible to plugging with partlculate matter.
Many times during calibration and optimization runs, the valve would
plug and cause a loss of A/F control. The result of this was high CO
production due to the "full-rich" failure mode of the feedback carbu-
retor. Holley attributed the valve's partlculate sensitivity to two
very small air bleeds (0.6350 mm = .025 1n., 0.7620 mm = .030 1n) in the
vacuum regulator assembly.
In addition to this problem, the particular feedback carbu-
retor used in the program also suffered from a lack of durability. At
the end of the first 2,000 km accumulation interval, CO emissions had
Increased drastically (Table 8-8). In fact, they exceeded the 80,000 km
maximum limit. The system, once again, would not control A/F ratio at
the desired set-point. The VCRV was operating, however its control
signals were abnormal and asymmetric, with extended time spent at high
vacuum (full lean signal) and short periods of low vacuum (full rich
signal). Time averaged A/F ratios were rich (<14/1), which correlates
with the high CO, low NOX> and low sulfate emissions observed. The
TABLE 8-8
FTP EMISSIONS (g/km)
AFTER 2,000 KM DURABILITYINTERVAL
PTX-516, ECU
-------
- 71 -
engine was checked for ignition system problems, but none were found.
In addition, replacement of the oxygen sensor had no effect on the loss
of A/F control or abnormal control vacuum pattern.
Holley agreed, at ER&E's request, to send an engineer to
diagnose and repair the failure. It was found that as underhood tempera-
ture increased, A/F ratio control became erratic. After studying the
airbox curve (A/F ratio vs air flow) of the carburetor sent with the
system, the engineer felt the loss of A/F control was due to a carburetor
malfunction. The feedback carburetor is prone to idle circuit percolation
which causes a lean A/F. The feedback system responds by going rich.
However, the airbox curve shows A/F control is nonlinear (Figure 8-5) in
the Idle/off-idle air flow regions (up to 30 scfm). A change from
stoichiometric (2.5" control vacuum) to full rich (0" control vacuum)
produced a much larger A/F correction than going from stoichiometric to
full lean (5.0" control vacuum). Thus, the loss of A/F control could be
explained in this way. The carburetor is driven rich by the ECU to
compensate for the percolation induced lean A/F. However, the A/F goes
too rich. The ECU drives the carburetor full lean for an extended time
in an attempt to remain at the set point A/F. Since the exhaust 1s now
lean, the ECU drives the carburetor rich, restarting the entire cycle.
This correlates with the high, then low control vacuum cycling observed.
Although the carburetor was nonlinear when originally received, there
were no cycling problems. Because of this it was suspected that the
carburetor had shifted stoichiometry and become even more nonlinear than
shown on the original airbox curve. It was recommended the carburetor
be returned to Holley for inspection and recallbration.
Figure 8-6 is the airbox curve of the defective feedback
carburetor. Note that the stoichiometric curve (2.5" control i/acuum) is
displaced lean from its original calibration shown on the previous
airbox curve. As air flow increases, the 2.5" control vacuum curve
begins to approach a stoichiometric A/F ratio. The full lean (5" con-
trol vacuum) curve was shifted so far lean it was off the A/F scale and
does not appear on the airbox figure. Thus, the feedback carburetor had
shifted drastically in calibration during the 2,000 km durability accum-
ulation.
Figure 8-7 is the airbox curve for the recalibrated feedback
carburetor. The three control vacuum curves are linearly spaced in idle
and off-idle air flow regions and have been aligned to provide the
correct A/F ratios. Steady states were run to check the ECU reference
voltage setting vs. A/F ratio for the recalibrated carburetor. In the
speed range of 10 mph to about 40 mph CO emissions were very high, and
there was no A/F control. CO emissions oscillated wildly, with peaks up
to 1% above mean levels. CO emissions only became stable above 45 mph.
Because of these oscillations, A/F ratio could not be determined accu-
rately. An FTP was run at previous reference voltage settings to see if
the CO oscillations had a great Impact on weighted emissions. As seen in
Table 8-9, the CO emission level of the vehicle was higher than when the
feedback carburetor first failed.
-------
AKLiUHtl Of!
? 8S25f)
Mo
PSI I
PSI Customer
"HG Engine __
"HG
Man. Vac. Idk-
Mi,n V.ic. W O.T.
! ! ! !Tn r
tA/F Curve
Stoichiometric i _._
-- 2.5" *-<--
A/F Curve
!_. Control Vacuum
T Control Vacuum
1
I. Full Lean A/F Curve
I ORIGINAL CARBURETOR AIRBOX CURVE
PHOJtCT ENGR,
10 20
AIRFLOW - scfM
!73 LB FT'DEMSllV)
APPV'O 8Y
140 IrO 130 20C ?20 ?-)0 ?30 3^0 360 400
200 250 300 400 500 C03 700 800
4 CO GOO 300 100C 1700
-------
HOLLEY CARBURETOR DIV., WARREN, MICHIGAN
I-LUW ItSI KtUUKU
_ OPERATOR ..i
Idle Bleed A.H.
Idle Bleed Vent
Idle Chin. Rest.
Idle Feed
Thro. Bore
Venturli
Pump Cep. C.C
Fuel Level
Fuel Pre«
Men. Vec. Idle /"7
Man. Vec W.O.T. _
Air Cleener
Hi-Speed Bleed
Idle Le»n
C.T.A.F.
Test Fuel
0" Control
Full Rich A/F Curve
Stoichiometric A/F Curve
Control Vacuum
DEFECTIVE CARBURETOR AIRBOX CURVE
5 10 20
AIRFLOW - SCFM
nit l o 'irr r>CN«ITV\
180 200 220 240
200 25C 300
-------
3
U
A|F
20|0.0
18
1611.0-
u||.B
12IS.5-
,0
8 ma
u
p
e
Ul
£
6IB.3-
.11)
.09
on
.08
.07
.06
OS
I DATP Y ~ & ' 7 7 AIRBOXNO. '// OPERATOR 'V A tf3fl h.' ft ^? £ ^ N" *
r
;r
2
C
5
C
0
"0
U
1
.5'
ont
.0'
ont
1
\
V
i
e-..,,..
.
acuum
i
4
9
4
rol Vacu
n
3l Vacui
I
E
1
j .
» fc
. , a *
1 s
jm " ^
Thro. Bora
L - Vmturli
"" " X
* 5 . ,
1 * fc
r
* * j
s
ss .-""'
1
1
die Bleed A.M.
die Bleed Vent.
Idto Chan. R*tt.
|
die Feed
" L Ml. Tram.
1 ' , (
* "! '
Loc
>t
....... ! i IdU Rich
Eliii.l.....
t ' ' '
Full Rich A/F
Stoichiometri
Full Lean A/F P
1 III II II j JIIH
1111 III II III)
RECALIBRAT
'"y
CL
M
:
at
~=v.
\
rv
*
V
;i(
F
S
"e
M
F
)
IG
V
«
)
«
i»»
L
>
(^
Curve
:u
Uf
-
rv
?F
^
e
f
.
^
<
^/
}-
V
y
^
/
\
*
mm
.
'
!s
n
'
^
\
1
r<
t «
BS
\
\
\_
**
1
CARBURETOR JjRBOX
^«
(
V
\
Pump Car
Fuel Leve
Fuel Pres
Man. Vac
Man. Vac
Air Clean
Hi-Spaed
Idle Lean
- i
;
s
^
\
i
w^
fea.
\
^
:
\
\
/
CURVE
MB
s
1
-
Idle
W.O.T.
Blaad
C.C
At
, PS
-^' PSI
J.?2.
1
"HG
"HG
w
_i
1
/
/
\
1
5 10 20 30 40 5C SO 70 80 200 120 140 :?J 180
AIRFLOW - SCFM
(.073 LB, FT 'DENSITY) , av PpOJ[r<-r FMCP
j
/
F
^
\
\
y
/
i
/
/
r^
V
V
i
/
r
/
/
^
V
\
^
*
1
ft
1
/»
f
"*
ft-
i~
r
X
£
6
If
fl
S
b
P
/
I/
20C 22C 240
*vw
f\
1.
/
,
.
I
n
Model 4l^ L
Custom
Engine
Serial f
C.T.A.F
TestFu
n
- -
^
f
ftv
'
-,
>
.r-
: J
- .
-j
dl
t
h
l
d
V
f
if
,L
i
$
I-
<;
V
}*
~y
/
^
-
*
P
f
-
r.
»«
£
v
LA
U
d
^>
V
T
}(
,±-
/y
--£_
_3
1,
3
t;
i
4 ^5 T3* // ^/
f, 1
V
>
^
6
-
i
^
f
ft
Q
/;
^
i
^
^
i
j
;
c
f.
/
\
t
'l
-V
k)
)
F
"^
V
<
t
1
I1-
»
.
*d
<:
i
-4
/
a
>
*
n
el
.,
i
>
J
j
IM
.
/
i
*«
^
/
5
3
*
n
/
^
..
If
1
S^
J »
1
t
1
1
1 ';
j
4)
i
X,
1
2?0 320 360 JOC J4C
400 &00 SCO 7TO -80fl
>2oo-L
u-»4t-i.pe
Vlv.a-72
-------
- 75 -
TABLE 8-9
FTP EMISSIONS (g/km)
RECALIBRATED CARBURETOR
PTX-516, ECU (? 200, 2 MIN. AIR INJECTION,
LEAN CHOKE. SPARK PORT EGR
(Single Test)
CO 3.04
HC 0.14
NOX 0.63
Because of the speed range Involved, a transition problem
between the idle and main jet circuits was suspected. This was con-
firmed by looking down the carburetor venturi while the vehicle was held
at steady speeds. At approximately 12 mph, fuel started to drip out of
the main nozzle and puddle on the throttle plate. The frequency of the
CO oscillations correlated with the fuel drop rate out of the main
nozzle. As vehicle speed (engine rpm) was increased, the main nozzle
continued to drip fuel until at 38 mph, atomization began to take place.
By 45 mph, the main nozzle was spraying fuel, and coincidentally the CO
oscillations smoothed out considerably. At 60 mph, only the main jet
was in operation. CO emissions were steady with very little oscilla-
tion, and A/F control was good, as shown in Figure 8-8.
Holley agreed with the diagnoses of a transition region
instability, remarking that this "sloppy main-jet startup" problem had
been seen on other feedback carburetors. The carburetor was returned to
Holley for reproportioning of the idle/main jet circuits. They remarked,
however, that it could take several iterations to resolve the instability.
Testing of the reproportioned carburetor revealed that although the
instability speed range had been narrowed, CO emissions were still above
80,000 km contract limits. Another iteration would have to be made to
solve the problem.
It was agreed by both ER&E and the EPA that contract work on
the second low excess air vehicle be terminated based upon its achievement
of 0 km emission goals.
A number of factors influenced this decision. Optimization
of the feedback carburetor system for low excess air operation was
hampered by recurring hardware failures. A significant portion of the
contract performance time period was spent waiting for new system com-
ponents, or for repairs and recalibration of failed pieces. These
operating difficulties were not expected to improve during kilometer
accumulation. It was expected that repair turn-around time would increase,
as Holley had to meet committments to manufacture feedback carburetors
for Ford and General Motors. The fact that the 80,000 km durability
accumulation would be extremely time-consuming and costly, coupled with
the general de-emphasis of vehicular $04* emissions, led to the decision
to drop durability testing from the contract requirements.
-------
- 76 -
35 mph
&
FIGURE 8-8
CO EMISSION STABILITY AT VARIOUS
VEHICLE SPEEDS
-------
- 77 -
Although kilometer accumulation was never completed due to
recurring system hardware failures, 1t 1s Important to realize that the
low excess air vehicle met all 0 km contract targets for exhaust emissions,
Even though system durability data was not obtained, the application and
demonstration of the low excess air concept for vehicle sulfate emission
control technology was a large percentage of the contract requirements;
and, in Itself, a substantial technical achievement.
-------
- 78 -
SECTION 9
REFERENCES
1. W. R. Pierson, R. H. Hammerie, J. T. Kummer, Sulfuric Acid Aerosol
Emissions from Catalyst Equipped Vehicles. Society of Automobile
Engineers Technical Paper No. 740287, 1974.
2./3. M. Beltzer, R. J. Campion, J. Marian, and A. M. Hochhauser, The
Conversion of S02 Over Automotive Oxidation Catalysts, Society of
Automobile Engineers Technical Paper No. 750095, 1975.
4. Holt, E. L., and Keirns, M. H., Hydrogen Cyanide Emissions from
a Three-Way Catalyst Prototype Vehicle, Environmental Protection
Agency Report No. 460/3-77-023, December 1977.
5. Griffith, M. 6., et. al., Assessment of Automotive Sulfate Emission
Control Techno!ogy. Environmental Protection Agency Report No.
460/3-77-008, June 1977.
6. Bond, W. D., Quick-Heat Intake Manifolds for Reducing Cold Engine
Emissions, Society of Automotive Engineers Technical Paper No.
720935, 1972.
7. R. Schwartz, G. Stumpp and H. Knapp. The Bosch Continuous Full
Injection System - A Mechanically Operating System for Continuous
Gasoline Injection. Bosch Technische Berichte, Robert Bosch GMBH,
Vol 4 #5, 1973 p 200-214.
8. U. Adler, E. Kaufmann, J. Warner, M. Scott, Fuel Injection - Bosch
Continuous Injection System (CIS), Technical Instruction Manual,
Robert Bosch GMBH, February 28, 1974 edition.
9. I. Gorille, N. Rittmannsberger, P. Werner, Bosch Electronic Fuel
Injection with Closed Loop Control, Society of Automotive Engineers
Technical Paper No. 750368, 1975.
10. 1977 New Car Features, USA and Canada - Technical Service Manual;
240 and 260 Model Series. TP ni583-US 9.76, Volvo of America,
Rockleigh, N.J.
11. Engh and Wallman, Development of the Volvo Lambda-Sand System,
Society of Automotive Engineers Technical Paper No. 770295, 1977.
12. R. E. Seiter, R. J. Clark, Ford Three-Way Catalyst and Feedback
Fuel Control System, Society of Automotive Engineers Technical
Paper No. 780203, 1978.
------- |