EPA-AA-EOD-84/1
CATALYST ACTIVITY TEST SEQUENCE (CATS)
January 1984
by
Donald D. Danyko
William W. VJatson
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Mobile Sources
Engineering Operations Division
Testing Services Group
2565 Plymouth Road
Ann Arbor, Michigan 48105
-------
CATALYST ACTIVITY TEST SEQUENCE (CATS)
Abstract
The primary objective of this program was to establish vehicle
operating modes during which a temperature rise across the
catalytic converter could be measured and used for determining
the catalyst activity. The results of twenty-five (25) vehicle
tests are reported. Seventeen (17) vehicle tests were
conducted using known good converters, six (6) using dead
converters and two (2) using partially-active (lead-poisoned)
converters. The test sequence consisted of a series of five
engine operating modes. In general, with careful thermocouple
attachment to the exhaust pipe surface at the inlet and output
of the converter, active converters show a large enough
temperature rise that they can be distinguished from dead
converters.
Introduction
The primary objective of this program was to establish
vehicle operating modes during which a temperature rise across
the catalytic converter could be measured and used for
determining the catalyst activity. In theory, unburned
hydrocarbons, carbon monoxide and nitric oxides (three-way
converters) from the engine will react in an active converter
and produce heat. The procedure was intended to be quick,
simple and cheap, and the equipment used, easy to install and
durable. The potential application of the procedure would be
as a secondary or referee test in cases where tampering or
misfueling had been established by other means. This report
details a procedure which has been developed emphasizing these
objectives. Temperature data for seventeen (17) vehicles with
good or active catalysts and eight (8) vehicles with inactive
or partially-inactive catalysts is presented. The discussion
also includes details on the general applicability of the
hardware and instrumentation employed during the course of this
effort.
Background
We were directed by Field Operations and Support Division,
the sponsors of this project, to develop a non-idle procedure
for assessing catalyst activity by measuring exhaust system
skin temperatures. Procedure development was divided into two
phases. Both phases included extensive collection of
temperature data from vehicles being run on a chassis
dynamometer. During the initial data collection period, the
temperatures on the exhaust system skin near the inlet and
outlet of the converter were measured during exhaust emissions
-------
-3-
tests. The vehicles being tested represented a sample of
in-use cars procured from the public. Attempts were made to
gather temperature data from as many different vehicle types as
possible (Table 1). No attempt was made to create a
statistical or weighted sample of any kind. The exhaust
emissions tests being performed included the FTP, KWFET and
various other short cycle tests encompassing several different
vehicle operating modes.
Following analysis of the temperature data from ten (10)
vehicles, a procedure was designed to yield a temperature rise
across the converter in as short a time as possible without
operating the vehicle in an unusual manner. Temperature data
was then gathered on vehicles using this test cycle.
Throughout this second phase of testing, variations of the
procedure were examined in order to resolve ambiguities in the
temperature data.
Vehicle exhaust system configurations presented many
challenges in attachment of the temperature measurement
instrumentation. Hose clamps were used to fasten K-type
thermocouples to the exhaust pipe as close as possible to the
converter inlet and outlet. Special attention was paid to the
location on the pipe (top, bottom or side) for each
thermocouple. Visual estimates of the heat transfer properties
of the pipe at the inlet and outlet to the converter were made
in order to best yield comparable temperatures. A strip chart
recorder was used to record the temperatures throughout the
test cycle.
Discussion
The CATS procedure was designed to meet the following
criteria:
1. incorporate vehicle operating modes which create high
converter loading;
2. minimize time to identify an active converter;
3. minimize operating stress on the vehicle; and
4. make no alterations to vehicle hardware.
High converter loading results from rich fuel/air
mixtures. Analysis of typical engine dynamics indicates rich
mixtures are likely during start-up idle and high power modes.
Consequently, the test segments of the cycle were designed so
the vehicle would be operating at a rich fuel/air ratio.
-------
-4-
Figure 1 (attached) shows the test cycle that was employed
in . the second phase of the data collection to achieve a
temperature rise across the converter. All driving is done at
FTP dyno settings for each vehicle. The cycle is initiated
with a 30 mph steady state to warm up the engine, exhaust
system and catalytic converter. As the temperature at 'the
inlet to the converter reaches 400CF, the vehicle is ready to
begin the actual testing segment of the procedure. The first
two modes are "crowds", that is, accelerations under conditions
of constant manifold vacuum. Crowds are a means of
automatically adjusting engine loading for each engine/vehicle
combination. The 7" crowd will give a temperature rise across
the converter for most vehicles with active catalysts. The 4"
crowds, in general, will supply a rich enough fuel/air mixture
to assure catalyst activity in vehicles where the catalyst may
be partially deactivated. The next mode, a 10-minute idle, is
representative of. another rich mixture mode of operation. Some
vehicles which do not show catalyst activity during crowd
conditions will demonstrate activity during this mode.
Finally, a misfire condition is initiated at 2500 rpm by
removal of a spark plug wire. This final test, by supplying an
unburned charge to the converter, should yield good evidence of
catalytic activity.
There are vast differences in the thousands of converter
emission calibrations on the road today. A high percentage of
good converters will likely show negative results in any one of
the modes. Fewer will show negative results in two or more
modes, and so on. Although the entire procedure requires
twenty-five (25) minutes on the dyno, most vehicles with good
converters will pass within the first three modes. Only the
dead converter vehicles will have to run through the entire
cycle. Although this procedure is not likely to identify 100%
of the active converters correctly, the passing criteria for
each mode can be adjusted to minimize errors of commission,
omission or both.
Disadvantages of the procedure are:
1. Temperature measurements 'are highly influenced by
thermocouple placement.
2. Many exhaust system configurations do not have
'equivalent1 inlet and outlet thermocouple
installation points.
3. Many exhaust system configurations make thermocouple
installation difficult.
4. Each unique exhaust/converter system has its own heat
capacity, heat transfer properties and catalyst
loading design.
-------
-5-
The first three disadvantages concern the thermocouple
measurement technique and will be discussed later. The fourth
addresses the high degree of variation observed between the
individual temperature traces of the data fleet. This item
identifies some factors responsible for the differences in
temperature traces from vehicle to vehicle. These factors and
the relatively different exhaust flow rates and temperatures of
each mode have been used in the refinement of mode selections,
their order and time length.
Data Analysis
Figure 2 (attached) is a typical trace of exhaust system
skin temperature versus time. The outlet temperature of the
active converters is shown higher than the inlet for each mode
and the outlet of the inactive converters is shown lower. This
representation reflects the results of the test fleet data
summarized in Table 1. For example, on the average, the good
converters on the seventeen (17) vehicles tested show about
60°F higher outlet temperature at the end of the 7" crowd
mode. The range for the good converters is from -150°F to
+320°F. Each mode has a similar high degree of variation.
Using the pass/fail criteria of Figure 1, all the good and
both lead poisoned converters .(about half-active) passed and
the dead converters all failed. However, several of the
vehicles were re-run due to non-equivalent thermocouple
placement. Incorrect thermocouple installation caused the
temperature rise measured across the converter to shift to the
extent that the results were contrary to the actual converter
condition (available FTP results). Results from the
correctly-placed thermocouple trace were included in the table
and the other results omitted.
Examining Table 1, we find, the majority of the seventeen
(17) good converter test vehicles demonstrating higher outlet
temperatures during the crowd modes. This is the clearest
demonstration of converter activity considering that both the
inlet and outlet temperatures are rising (Figure 3). Figures 3
and 4 are temperature traces taken from the same vehicle. The
difference is that the converter beads were removed (no
activity) for the test trace of Figure 4. There is a 70°F
crossover at the end of the 4" crowds for the good converter
and no temperature crossover for the dead one.
Analysis of the idle portion of these traces is more
complicated than the simple crowd analysis. Note that there is
a temperature crossover for the dead converter (Figure 4)
during the idle mode. During the crowd modes (high power
output) the converter interior and shell are being heated by
the high temperature exhaust. The engine exhaust temperature
during the following idle is much cooler, so both the inlet and
-------
-6-
outlet skin temperatures fall during this mode. The outlet
skin temperature is partially maintained during this time by
convective heat transfer from the converter interior to the
cooler idle exhaust gas. Note also that in both figures
neither temperature has reached an equilibrium (steady)
condition after 10 minutes. This behavior is a result of (1)
the heat capacity (retained heat from crowd modes), (2) the
heat transfer (heat loss of the converter to its surroundings),
and (3) the amount of heat being generated during the idle (HC,
CO and NOx loading from the engine).
Early in the development of this procedure, equilibrium
idle temperature comparisons were sought but cycle-time
requirements limited this possibility. Reviewing the idle data
in Table 1 indicates there are distinguishable temperature
differences in the good and dead converter groups in spite of
the complicated factors at work during this mode. Therefore a
100°F temperature difference at the five-minute idle point was
chosen as a pass/fail criteria. It is important to recognize
that the temperature condition of the converter at the
beginning of the idle period shown in Figures 3 and 4 has an
integral part in determining the five-minute point temperature
difference. That is to say, each mode is a preconditioning for
the following mode. The outlet pipe temperature is lower than
the inlet for the dead converter (Figure 4) at the instant the
idle mode begins, and higher for the good (Figure 3). At the
end of the idle mode (beginning misfire) the outlet
temperatures are very different in the two figures. The
misfire portion of Figure 3 (good converter) may look very
different if the vehicle were operated in a manner to bring the
inlet and outlet temperatures together before the misfire
begins.
It is possible to divide this procedure into discrete
segments for special purposes, however the existing analysis
criteria would no longer apply, and the diagnostic ability of
any section will be less than the whole. Further development
and analysis could also be directed into any given mode to
determine its single maximum usefulness.
Hardware Applicability
Application of the temperature measurement hardware to the
different exhaust system configurations created many
complicated installations.
The equipment used for this measurement process was:
K-Type Thermocouples, 1/16" Diameter (Calibrated)
Vacuum Gauge (Calibrated)
- Worm-gear type hose Clamps
Temperature Recorder, able to read from 0-2000°F.
(Calibrated)
-------
-7-
The primary problem areas were shields (attached to converter
and exhaust pipes), converter location and exhaust system
routing.
Shields protecting the exhaust pipes and converter created
major complications. The imported vehicles appeared to have
the most occurences of shields, whereas most domestic vehicles
avoid the use of shields. For example, Subaru vehicles which
posed the worst case, have shields welded on the entire exhaust
system except for the converter. The converter has a bolted
shield, where the removal of six bolts will expose the surface
areas necessary for proper temperature measurement. In most
cases however, the exhaust shields are bolted on, such as,
Mazda GLC and Toyota Tercel. Once the nuts and the shields are
removed, the inlet and outlet areas necessary for temperature
measurement are adequately exposed. Problems that arose during
installation on domestic vehicles were also with catalyst
shields, and shields on mechanical components. Ford Escorts,
for example, have a two piece shield welded around the
catalyst. However, adequate space is allowed to properly
attach the thermocouples if care is used during attachment.
Chrysler Omni and Horizon models generally have a two catalyst
system, with the first catalyst located six inches after the
exhaust manifold along the fire wall. To properly expose the
catalyst inlet pipe, removal of a shield above the right
transaxle shaft is required.
Exhaust system routing and converte'r location created the
second problem area. In most cases this problem was
non-existent, but where it did exist, major obstacles were
created. Some specific cases are Chrysler K cars and
Omni-Horizon models, late model Hondas, all Subarus, Ford
Escort-Lynx, Ford Mustang-Capri, and Ford Fairmont models.
The Chrysler K cars and Omni-Horizon models are equipped
with a two-catalyst system. A splash shield blocks the access
and one of the converters is located up and behind the engine
against the fire wall, six inches below the exhaust manifold.
Late model Hondas, although none were tested, have the
converter mounted to the outlet of the exhaust manifold. This
will create a special problem, in that no inlet pipe mounting
area exists.
All Subaru models have two seperate inlet pipes into the
converter, with only one outlet pipe. And as mentioned
earlier, welded shields encompass the inlet and outlet pipes,
with a two piece shield bolted around the entire converter.
Finally, the aforementioned Ford models have catalysts with
sharp angled inlets and outlets. The Fairmont models also have
heavy steel flanges at the converter inlet and outlet areas.
-------
-8-
During development of the procedure, differences as high as
70°F were measured simply by locating the thermocouples on
different sides of the exhaust pipe. Therefore certain
precautions must be observed when attaching the thermocouples.
During tightening of the hose clamp, the tip of. the
thermocouple must be covered by the band. The axes of the
thermocouple and the exhaust pipe must be parallel or the clamp
will not apply adequate pressure. The worm gear on the clamp
should be opposite the thermocouple. Otherwise, the metal in
the worm gear creates ,a heat sink, giving an inaccurate
temperature reading. The thermocouple wires must be kept away
from moving parts and the hot exhaust system. All thermocouple
mounting surfaces should be sanded to remove excessive
corrosion, allowing for good contact and proper heat transfer.
The hose clamps should be as tight as possible, holding the
thermocouple firmly against the pipe. There should be no
exhaust leaks near the thermocouples, or there will be an
impact on measured temperatures. Thermocouples should be
mounted as close as possible to the converter inlet and outlet,
the farther away from the converter inlet and outlet the more
heat that will be lost through radiation. If a bend exists in
the inlet and outlet pipes, thermocouples should be mounted on
the outside of the bend, allowing measurement at the hottest
point on the pipe. Lastly, the thermocouples should be at
least one to two inches away from overlapping pipe joints,
inlet and outlet connection flanges, mounting brackets or
hangers, and shields. All of these are heat sinks and can
radiate heat away from the thermocouples. Note that in some
cases one to two inches will not apply because of interfering
brackets, etc. In these special cases, a brief analysis of the
mounting problem and engineering judgement will allow for the
proper thermocouple location.
Conclusion
Measurement of exhaust system skin temperatures, if done
carefully, may indicate whether a catalytic converter is good
or bad. Temperature measurement across the converter indicates
thermal activity "but it does not establish whether it is a
result of hydrocarbon, carbon' monoxide and nitric oxide
conversion or of any combination of those three conversions.
Assessment of temperature data taken from the exhaust system
skin at the outlet and inlet of the converter is complicated by
the variety of exhaust system configurations and the unique
properties of each converter. The CATS procedure, by operating
the vehicle in rich fuel/air mode, provides a test that was
successful in identifying active and inactive catalysts. The
procedure does 'require a large amount of equipment including a
chassis dynamometer. Extensive engineering judgment was
employed in interpretation of the data generated during this
program, but with proper training the use of the procedure is
clearly feasible.
-------
-9-
Recommendatlons
The procedure should be performed as designed, since the
differing exhaust system configurations and their corresponding
heat transfer and heat capacity characteristics will alter the
temperature data if a different sequence is used. The simplest
and most effective means of thermocouple attachment is by hose
clamp. Strip charts should be used to record the temperature
data to allow analysis. A relatively small sample fleet of
screened tampering/misfueled vehicles should be CATS tested and
then FTP tested to determine real-world CATS accuracy prior to
full scale implementation.
-------
TABLE 1. Converter Activity Test Sequence Fleet Data
VEH
# MANUFACTURER/^DEL
ENGINE
CLASS/TYPE
GOOD
10
14
4
16
5
6
17
1
2
9
CONVERTERS
GM Camaro
GM Cutlass
GM Regal
GM Grand Prix
GM Skylark
GM Citation
GM Citation
Ford Escort
Ford Escort
Chrysler Horizon
^
A106
A110
A110
A69
A69
-
A3
A3
A86
(Two Converters
15
3
7
8
11
12
13
DEAD
10
14
16
17
1
9
LEAD
16
17
Chrysler Champ
Nissan Datsun 210
Toyota Tergel
Honda Accord
Toyota Mazda
Isuzu Subaru'
Isuzu Subaru
CONVERTERS
A109
A73
A125
A83
Alll
A105
A105
v-e
V-8
V-6
V-6
1-4
1-4
1-4
1-4
1-4
1-4
FTP (GM/MI)
HC CO
0.2 3
0.2 3
0.8 8
0.4 7
0.6 14
0.3 3
0.2 4
0.2 3
0.6 9
0.5 8
NOx
0.9
0.8
2.9
0.6
0.5
1.1
0.4
0.3
0.7
1.0
in Series)
1-4
1-4
1-4
1-4
1-4
H-4
H-4
SAME VEHICLES AS ABOVE
POISONED CONVERTERS
SAME VEHICLES AS ABOVE
0.6 6
0.4 8
0.3 5
0.5 7
0.8 10
0.3 6
0.2 3
2.5 24
2.2 22
2.3 18
- -
2.5 31
2.9 42
2.9 42
1.0 7
- -
1.0
1.2
0.5
0.4
0.6
0.7
0.8
3.4
2.1
2.5
-
0.9
1.0
1.0
1.3
-
TEST MODE/TEMPERATURE RISE ACROSS CONVERTER °F
CROWDS
7" 4"
10 -10
50 70*
80* 150
-30 70*
50* 130
-20 20
-150 -110
85* 260
320* 275
80* -220
-65 -110
80* 110
100* 130
120* 180
60* 110
170* 270
-20 -15
30 90*
-115 -145
-100 -45
-80 -65
-190 -135
-50
-340 -300
-100 -45
-30 -30
-80 -80
IDLE
START 5 MIN
-10
70
150
70,
130
-20
-110
130
320
-220
-110
110
190
60
115
440
-15
90
-145
-45
-65
-135
-150
-300
-45
-30
-80
185*
170
170
200
85
80
50
210
85
160
-160
240
270
115
140
430
60
170
-30
15
60
-30
-50
5
55
140*
130*
MISFIRE
END
215
155
160
230
0
0
0
80
20
45
-70
220
280
90
-
480
-20
140
-30
-20
35
-65
-40
-90
45
165
50
START END
150
145
-
200
-
115
0
40
-
0
-60
200
-
-.
-
400
-15
70
-60
-20
20
-55
-
-60
0
95
55
65
175
-
130
-
330*
50*
190
-
-20
-60
330
-
- -
-
540
85*
150
-150
-20
-95
-70
-
-230
-20
-10
70
Refers to 'pass' point according to Figure 1 criteria.
-------
Figure 1
11
Converter Activity Test Sequence (CATS) Flowchart
'FAIL'
END OF
TEST
START 30 MPH
3 MINUTE WARM-UP
Yes
No
RUN 7 CROWDS
20/40 MPH - 4 MIN MAX
Yes
No
RUN 4" CROWDS
20/50 MPH - 4 MIN MAX
Yes
No
RUN IDLE
10 MINUTES
Yes
No
RUN MISFIRE
2500 RPM-2 MIN MAX
.* AT is the temperature rise across the converter (outlet-inlet),
-------
MII, lulu
-------
H
o
o
pa
VO
OO
I-*
O
O
c
•XI
•n
M
O
pa
tt
u>
o
H
n
o
I
Fl
?a
o
o
o
o
o
|
PI
8
o
H
r1
PI
H
z
H
S3
PI
200
20
20
20
J
Ff=//,r
;' 400
KCIO I/I
400 \
\i
80
1200
T*-
400
: 'YtOO
1200
-tTT-T
,2 .}
Kite in . !
1200
20 .••••' 400 600
— .g£G-|N
MVO ri
800
1200
1600 •
1400
1600
1400
• 1SOO
1400
1600
i i
'• 1800
1
1800 ' N>
T
M
f
O
w
1400 ! ' 1600 • i : 1800
i . i
pa
o
1800
ei
-------
a
73
H >
73 O
> H
O C
o n
p. O
o H
00 PI
O l-l
a z
50 o
M G
s
CO
o
PI H
n PI
i g
< H
PI C
8 g
------- |