Evap 75-5
Technical Support for Regulatory Action
In-House Test Program
Report No. 2-
Vehicle Preconditioning:
LA-4 vs. HFET
November 1975
Thomas Rarick
Notice
Technical support reports for regulatory action do not
necessarily represent the final EPA decision on regulatory
issues. They are intended to present a technical analysis
of an issue and recommendations resulting from the assumptions
and constraints of that analysis. Agency policy considerations
or data received subsequent to the date of release of this
report may alter the recommendations reached. Readers are
cautioned to seek the latest analysis from EPA before using
the information contained herein.
Standards Development and Support Branch
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air and Waste Management
U.S. Environmental Protection Agency
-------
Contents
Page
1. Introduction 1
2. Summary and Conclusions 3
3. Technical Discussion 4
3.1 Facilities and Equipment 5
3.1.1 Test Vehicles 5
3.1.2 Test Fuel 7
3.2 Test Procedures 7
3.2.1 Vehicle Preconditioning 7
3.2.2 Evaporative Emissions 9
3.2.3 Exhaust Emissions ... 9
4. Test Results 10
4.1 Evaporative Diurnal Emissions 10
4.2 Canister Weights . . 11
4.3 Exhaust Emission Results 14
5. Data Analysis 18
6. References 19
Appendix A - Individual Test Results
Appendix B - Analyses of Variance
-------
1. Introduction
The preconditioning vehicle operation described in the Federal
Register (1) requires that the test vehicle be driven over a prescribed
mileage accumulation route for a period of one hour (this is usually
called the AMA road route). The fuel tank is then drained and a speci-
fied type of test fuel is added. The vehicle is then driven over a
simulated trip on a chassis dynamometer. The Urban Dynamometer Driving
Schedule (usually called an LA-4) is used for this simulation. In the
In-House Report No. 1 (2) the current preconditioning cycle consisting
of both the AMA and LA-4 was evaluated to determine if the AMA segment
of the cycle could be eliminated. It was shown in that report that
similar exhaust and diurnal emission levels existed whether or not the
AMA was conducted. Since the LA-4 is considered a typical urban driving
cycle, it was concluded that the LA-4 prep cycle alone would be suf-
ficient to precondition the vehicle. Further support of this conclusion
can be drawn from another in-house study (3) conducted to determine the
equivalency of an AMA + LA-4, three LA-4's + one LA-4, or one LA-4
preconditioning drive with respect to exhaust emissions. That study,
based on ten (10) replicate tests on three vehicles concluded that
equivalent exhaust results are achieved with any of the three precon-
ditioning drives.
The purpose of vehicle preconditioning is to provide a relatively
consistent starting base for all vehicles involved in emission testing.
It is intended that this desired starting base be achieved by a pre-
conditioning drive which tends to simulate a real-life condition that a
vehicle would normally experience in its day-to-day operation. The basic
vehicle preconditioning sequence involves soaking the vehicle (cold
soak) prior to the diurnal test which allows the vehicle to reach a
stabilized ambient temperature. The preconditioning drive is conducted
prior to the cold soak, which conditions the evaporative control system
(i.e. canister) and flushes out the old fuel. In addition, the precon-
ditioning (prep) drive may standardize the vehicle if it has not been
operated within a reasonable length of time.
The Highway Fuel Economy Test (HFET) cycle is a typical driving
cycle for highway driving. The method of conducting the HFET test
requires driving two identical cycles; the first intended to warm up the
vehicle and the second cycle is used to measure the vehicle's highway
fuel economy. If the LA-4 prep cycle and the two HFET prep cycles can
be shown to condition the vehicle equivalently, then the HFET cycle
could be used to replace the LA-4 prep.
Replacing the LA-4 prep with the HFET prep would likely have an
economic advantage. Currently, if a HFET cycle is run, it is run at the
end of the Hot Soak test and is an additional test. This additional
test adds time to the overall procedure and requires further manpower.
-------
-2-
An alternative to replacing the LA-4 with the HFET for vehicle
preconditioning is the use of an LA-4 as a warm up cycle followed by one
HFET data gathering cycle. This alternative would have the advantage of
simulating both urban and rural driving.
The purpose of this study is, therefore, to examine whether or not
the HFET driving cycle is equivalent to the LA-4 preconditioning cycle.
This test program was designed to reveal if any significant differences
existed in diurnal emissions or exhaust emissions when the vehicle
underwent different preconditioning drives. It was assumed that, if
significant differences in diurnal losses were exhibited, the effect of
the different cycles on the evaporative control system could be respon-
sible. For all vehicles tested the evaporative control system utilized
a charcoal canister as a storage media for the generated fuel vapors.
In order to determine what effect the prep cycles had on the generation
of .fuel vapors, canister weights were taken before and after the prep
drive and before and after the diurnal test.
A small portion of the test program was designed to gain insight
into the alternative of using one LA-4 plus one HFET as a precondition-
ing cycle. By recording canister weights before and after this combined
preconditioning cycle, the cycle could be evaluated to determine if the
purge characteristics were similar to either a single LA-4 cycle or 2
HFET cycles. Only the Matador was used during this portion of testing
as its evaporative control system was the most sensitive to the differ-
ent preconditioning cycles.
-------
-3-
2. Summary and Conclusions
The purpose of this study was to determine whether or not the two
HFET cycles used for measuring fuel economy would be suitable for a
preconditioning driving cycle in place of one LA-4. The main concern
with using the HFET was the possibility that the purge characteristics
of the evaporative emission control system would allow for purging
during preconditioning and not during the exhaust emission cycle which
is essentially the same as an LA-4. If this were to occur then suf-
ficient canister capacity would enable a vehicle to pass the evaporative
emissions test, and the exhaust test would be easier to pass due to the
fact that hydrocarbons from the canister would not be purging into the
engine.
The results of the analysis indicate that this does in actuality
occur, and that on the average lower diurnal evaporative emission levels
occur following HFET preconditioning for all 5 vehicles. The alterna-
tive of using an LA-4 plus one HFET for preconditioning appears to have
the same effect on canister purging for one of the test vehicles and
thus its use as a preconditioning cycle would probably also result in
abnormally low evaporative emission levels. Thus, the use of 2 HFET
cycles or one LA-4 plus one HFET does not seem suitable as a replacement
for the LA-4 cycle for preconditioning. These cycles could possibly be
given further consideration for use in preconditioning if they were
followed by a hot soak period followed by an additional LA-4 cycle.
This possibility should in the future be given more consideration.
A
-------
-4-
3. Technical Discussion
Generally, a test procedure attempts to typify or simulate an
actual condition. A concerted effort was expended to establish the
Urban Dynamometer Driving Schedule (LA-4) as an average driving cycle
(4). There was also a similar effort put forth to design the Highway
cycle as a typical driving cycle (5). Each cycle is typical of a
particular type of driving; the LA-4 is typical of urban driving, and
the HFET is typical of rural driving. Of particular interest for either
cycle is how the evaporative control system (i.e., canister) is con-
ditioned. This conditioning is accomplished by purging (or loading,
possibly) the canister during the prep drive to its normal working
capacity. This allows the canister to accomodate the evaporative losses
generated by the subsequent diurnal heat-build. If the LA-4 and HFET are
to be considered equivalent preconditioning drives then equivalent
diurnal losses and exhaust emissions should be exhibited for the two
test sequences.
The exhaust emission test used during the Federal Test Procedure
simulates a cold and hot start. The cold start test consists of running
one LA-4 on the dynamometer. The hot start test consists of the first
505 seconds of the LA-4 cycle and is conducted after the cold start test
with a 10 minute soak between tests. Since both the LA-4 and Federal
Test Procedure exhaust cycles are basically the same, equal amounts of
purging should occur during the preconditioning drive and exhaust test.
The HFET on the other hand is an entirely different driving cycle from
the exhaust cycle used, with the HFET cycle having a 29 mph higher
average speed. It is conceivable that a system could be designed that
would purge during the preconditioning drive and not during the exhaust
test. This condition should be avoided.
The economic advantage of the HFET being used as a preconditioning
driving cycle stems from the fact that the HFET, when conducted, is run
after the hot soak and thus adds time to the overall sequence. It also
requires the vehicle to be scheduled for dynamometer operation three
separate times (preconditioning, exhaust test, and HFET). The HFET is
currently an optional test which the manufacturers may choose to have
run. It is conceivable that in the future it may be a standard part of
the Federal Test Procedure. In either case, it would be cost effective
to be able to gather data during the preconditioning drive which must be
conducted regardless. This would mean that only 2 dynamometer operations
would need to be scheduled (preconditioning and exhaust test). One
issue that would still need to be resolved is whether or not the vehicle
would need any additional preconditioning prior to the HFET precondition-
ing drive. The Federal Register (6) currently requires extensive pre-
conditioning prior to the HFET if the vehicle has not been operated
within a 24 hour period. If it is determined that this preconditioning
is necessary, then the additional time to do this may make the use of
the HFET a less cost-effective alternative.
-------
-5-
It is reasonable to assume that either of the preconditioning
cycles is capable of flushing out the old fuel in the system and also
capable of standardizing vehicles which have not been operated within a
reasonable length of time. Hot soak tests were not conducted.
The use of two HFET cycles as a preconditioning cycle would only
simulate rural driving during preconditioning. The use of an LA-4 cycle
to warm up the vehicle followed by an HFET data gathering cycle would
have the same economic advantages as two HFET cycles but it would also
simulate both urban and rural driving during preconditioning. This
alternative combined cycle would, however, be somewhat longer than a
single LA-4 cycle or 2 HFET cycles by approximately 10 minutes. This
additional time during the preconditioning could allow the canister to
abnormally purge.
The Federal exhaust test simulates normal urban driving over a 7.5
mile trip. The amount of purging which takes place during this trip
should be roughly the same as the amount of purging during the pre-
conditioning cycle if the preconditioning and exhaust emission cycles
are to be considered compatible for use in the same test procedure. If
an abnormally large amount of purging is expected to take place during a
preconditioning cycle, the vehicle could possibly be soaked for one hour
following the HFET to load the canister again and an additional single
LA-4 preconditioning drive conducted after the soak. This additional
cycle should purge the canister equivalent to the purge expected during
the exhaust emission cycle and thus the canister would be properly
preconditioned. The use of this additional preconditioning would,
however, greatly reduce the cost savings advantage of using the HFET for
preconditioning.
3.1 Facilities and Equipment
The LDV Evaporative Enclosure as shown in Figures 3-1 and 3-2 was
used for all evaporative emission tests. The SHED is nominally 8 feet
high x 10 feet wide x 20 feet long and has a measured volume of 1540
ft3. Calculation of the enclosure volume with a propane injection and
recovery test compared within + 2 percent. Propane retention tests of 2
and 4 hours were performed periodically and indicated a leakage rate of
less than 0.1 g/hr.
3.1.1 Test Vehicles
Five 1975 MY vehicles were used in this evaluation. The criteria
for selecting the vehicles was that they had accumulated 4000 miles and
had been in use for over 90 days. Additional criteria were engine fuel
tank configuration and exhaust control system. Specifics of each vehi-
cle are shown in Table 3-1.
-------
-6-
Figure 3-1 Evaporative Enclosure (front view).
Figure 3-2 Evaporative Enclosure (rear view).
-------
-7-
Hake
Model
VIN
Disp/Cyl
Displacement
Transmission
Air Cor.d.
Ign. Timing
Idle RPM
Tires
Carb. Model
Venturis
Fuel Bovl Size
Fuel Tank Vol.
Inertia Wt.
Dyno H.P.
Exhaust Sys.
Evap. Sys.
Chevrolet
Vega
IV77B5U113062
140-14
4-speed
no
10°.BTDC
700
A78-13
Holley
2
38.5 cc
16.0 gal
2750
9.9
XCR
Catalytic
Reactor
Canister
Chevrolet
Cataaro
IQ87H5N511341
350-V8
Automatic
yes
6°BTDC
600
FR-7814
Rochester
2
72 cc
21.0 gal
4000
12.0
EGR
Catalytic
Reactor
Car.ister
Chrysler
New Yorker
LS23T5C110951
440-V8
Automatic
yes
8°BTDC
750
JR78-15
Carter
4
160 cc
26.5 gal
5000
13.4
EGR
Catalytic
Reactor
(dual)
Canister
AKC
Matador
A56167P15041
360-V8
Automatic
yes
5°BTDC
700
HR78-14
Motorcraft
4
24.5 gal
4500
12.7
EGR-AIR
Catalytic
Reactor
(dual)
Canister
Volkswagen
Beetle
1352038245
97 - 14 .
4 - speed'
No
5° ATDC
875
6.00 - 15L
-
-
-
11.0 gal
2250
8.3
EGR
Fuel injection
Canister
_ Table 3-1 Test Vehicle Descriptions
3.1.2 Test Fuel
Indolene Type HO lead-free test fuel was used throughout the pro-
gram including vehicle preconditioning.
3.2 Test Procedures
Each of the five (5) vehicles underwent identical test sequences
according to the flow diagram shown in figure 3-3. The comparative
tests involved changes in the preconditioning drive only. In addition 4
preconditioning drives consisting of one LA-4 followed by one HFET cycle
were conducted back to back with a one hour hot soak between each drive
to evaluate the effect of the combined cycle. Only the Matador was used
for this evaluation and only canister weight data were recorded during
this evaluation with the Matador.
3.2.1 Vehicle Preconditioning
All vehicles except the 4 tests run on the Matador using the
combined preconditioning cycle underwent a preconditioning drive fol-
lowed by an 11 to 20 hour soak period at a temperature of 76 to 86°F.
-------
OQ
UJ
I
LO
n>
J3
1. Drive LA-4 or HFET
Record '
canister wt before & after
driving cycle
tank, bowl, underhood temps
after driving cycle
1. Drive vehicle to soak area.
2. Soak for 11-20 hrs..
Record
* tank, bowl, underhood &
soak area temps for first
2 hrs
lv Drain fuel.
2. Add fresh fuel to 40Z
capacity.
3.. Move vehicle to_ Shed.
Record ' '-'
' tank & fuel temps
time to fuel vehicle
time to move vehicle to
Shed
canister weight
ca.
r>
(D
I
1. Push vehicle into Shed.
2. Heat fuel from 60'f + 2*F to
. 84*F + 2*F in 60 mlnutM
3. Fur$«~8h«d.
Record
' HC concentration
* time vehicle is in Shed
time of tank heating
tank, tank skin, underhood,
bowl, soak area & Shed temps
canister wt @ end of cycle
1. Push vehicle onto dyno.
2. Conduct '75 FTP
Record
time between end diurnal &
start FTP
tank, bowl temp 8 end of
cycle
-------
-9-
Since test fuel was being continuously used for the entire procedure,
including the prep cycle, the fuel was not drained and replenished prior
to the dynamometer prep cycle. Two different preconditioning driving
cycles were used.
The first method was the 1975 Federal Procedure (1) without the AMA
road route. The second was two Highway Fuel Economy Test (HFET) cycles
as described in the Federal Register (6).
Canister weights were measured before and after each precondition-
ing cycle.
3.2.2 Evaporative Emissions
Evaporative emission measurements were determined using the SHED
technique in accordance with the SAE J171a Recommended Practice (7) with
the following modifications:
1. Vehicle background emissions were considered negligible based
upon a previous study (8), and were therefore not determined.
All test vehicles were over three (3) months old and had
accumulated 4000 miles.
2. The 60 minute heat-build for the diurnal was held to + 2
minutes.
3.2.3 Exhaust Emissions
Exhaust emissions were determined using the 1975 Federal Test
Procedure. Since it is intended to integrate the evaporative procedure
along with the exhaust test, it was decided to use the present (cold-
hot) exhaust test procedure driving cycle rather than the cycle proposed
in the SAE procedure.
-------
-10-
4. Test Results
A total of 58 tests was conducted. Nineteen (19) tests were
conducted with an LA-4 prep, thirty-five (35) tests were conducted using
2 HFET prep cycles, and four (4) tests using one IA-4 plus one HFET prep
cycles were conducted. For a few tests, only the diurnal data were used
due to void CVS tests and in a few other cases only the CVS test data
were used due to void diurnal tests. The individual test results for
the diurnal heat-build tests and exhaust emission tests are shown in
Appendices A-l and A-2 respectively. In addition to the diurnal test
results, Appendix A-l contains the canister weight data for all tests.
4.1 Evaporative Diurnal Emissions
The individual test results for evaporative emissions are shown in
Appendix A-l and the means of the diurnal losses for each of the vehicles
along with a composite mean are shown in the bar graph, Figure 4-1.
The bar graph shows that for the Camaro, Vega and Volkswagen
Beetle very little difference in diurnal loss levels was observed
between the two prep cycles. However, for the Matador and New Yorker a
sizable difference existed, with the test conducted with an HFET prep
exhibiting lower diurnal losses. This would indicate that the HFET
cycle may condition the canister more than does the LA-4 driving cycle.
This would seem logical as the average speed of the HFET is approximate-
ly 29 mph higher than the LA-4 and therefore a greater potential for
purging the canister exists with the HFET. Whether or not this will
have an effect on the amount of purging is dependent on the character-
istics of the individual purge systems, however.
CO
B
cd
M
60
\
\
\
\
s
\
rs ra
-^
\
\
\
\
\
Camaro Matador New Vega Volks- 5 Vehicle
Yorker wagon Mean
Figure 4-1 Diurnal HC Losses for LA4 and HFET Preps
-------
-11-
4.2 Canister Weights
Figure 4-2 shows the mean canister weights for each vehicle and
prep cycle through the test sequence. It should be remembered that
different control systems and purge systems are represented by the
different vehicles. This fact makes it somewhat difficult to evaluate
differences between the two prep drives as the different control systems
may react differently to either of the prep cycles. Three important
facts do, however, emerge from studying these figures. First, the amount
of purging appears to depend on the initial level of canister weight.
In comparing the Camaro and Vega diagrams it can be seen that for the
Camaro the HFET cycle purges the canister more than the LA-4 cycle, but
for the Vega just the opposite is true. However, for all vehicles a
greater amount of purging takes place when the canister is at an initial-
ly higher weight. Secondly, it is apparent that in spite of different
amounts of purging some systems are still capable of storing roughly the
same amounts of diurnal losses. The Camaro, Vega, and Volkswagen show
different trends in the amount of purging for the two prep drives, but
the diurnal losses appear very nearly the same. For the Matador and New
Yorker, the diurnal loss levels are larger than for the other three
vehicles and they both appear to do better with the HFET preconditioning
drive. This fact may be due to the particular system designs. These
facts indicate that although the two driving cycles are not totally
equivalent some vehicle systems will react equivalently to them, whereas
other vehicle systems will not with respect to their diurnal losses.
The third important fact that can be drawn from Figure 4-2 is that
the Matador purge system reacts differently to the LA-4 and 2 HFET
preconditiong cycles. The Matador canister is loaded during the LA-4
driving cycle and is purged during the HFET cycle. The figure shows,
however, that the initial weight of the canister is higher on the
average at the beginning of the tests conducted using 2 HFET cycles than
at the beginning of tests in which one LA-4 preconditioning cycle was
used. Closer examination of individual tests indicate that the canister
always loads during an LA-4 preconditiong cycle regardless of its
initial weight. The amount of purging appears to be more sensitive to
the initial weight of the canister during"the 2 HFET cycles, however.
The greatest amount of purging generally takes place when the initial
weight is the highest. Thus, the LA-4 cycle does not appear to be
capable of purging the Matador's canister whereas the 2 HFET cycles have
the ability to purge the canister and the amount of purging is somewhat
dependent on the amount of hydrocarbons already stored in the canister.
Canister weights during the combined preconditioning driving cycle
of one LA-4 cycle plus one HFET cycle are presented in Table 4-1. These
data are plotted as changes in canister weight during the precondition-
ing drive in Figure 4-3 along with the canister weight changes during
one LA-4 cycle, two HFET cycles and an AMA drive plus one LA-4 cycle.
The AMA + LA-4 preconditioning cycle data were taken from the "In-house
Test Program Report No. 1" (2). Only data for the Matador are shown as
-------
-12-
w 1120
i
M 1110
*J
M 1100.
S
n 1090
o
jj
3 1080
u
a
rl
III
1120'
1110.
1100
1090
1080.
goak ^ Diurnall
Test Sequence
Figure 4-2a Cainaro
~] Prep ( Soak ["DTurnal |
Test Sequence
Figure 4-2b Matador
u
.e
1120
1110
3 1100.
1090
1080
| Prep | Soak | Diurnal |
Test Sequence
Figure 4-2c New Yorker
1180
1170 .
4? 1160
S3 H50
c
CO
u
"1 Prep I Soak I Diurnal I
Test Sequence
Figure 4-2d Vega
1000.
990-
H 980
-------
-13-
Test No.
1
2
3
4
Canister Weights, e
Before Prep.
1114
1123
1122
1122
After Prep.
1113
1113
1114
1114
After Hot Soai
1123
1122
1122
1120
00
Table 4-1 Canister Weights for LA-44HFET Preconditioning
Tests on Matador.
e
o
4 r
ig Precondit:
o NJ
Wi
H
T3
60 -2
1
O
I '4
H
-------
-14-
the Matador was the only vehicle which had a purge system which reacted
differently to different cycles. However, the fact that a system can be
designed which reacts as the Matador system does is an important con-
sideration. The figure shows that the Matador system is purged to
similar levels for either the LA-4 + HFET or 2 HFET preconditioning
drives which implies that only a single HFET is needed to purge the
canister. The canister weight data for the AMA + LA-4 prep showed that
purging did take place during the AMA road route drive but due to the
loading of the canister during the LA-4 cycle which followed, the
overall effect on the canister during the preconditioning was a slight
increase in canister weight. This fact reemphasizes the fact that the
Matador canister cannot purge during an LA-4 cycle.
4.3 Exhaust Emission Results
It is felt that in order to conclude that the two preconditioning
driving cycles are equivalent, similar exhaust results should also be
exhibited regardless of prep cycle used. It was assumed that if any
differences in exhaust emission levels existed they would be most
noticeable in Bag 1 of the exhaust test. Therefore, both Bag 1 and
composite exhaust results were analyzed. Mean Bag 1 exhaust levels are
shown in Figure 4-4 and mean composite exhaust levels are shown in
Figure 4-5.
The results from the Bag 1 exhaust analysis shown in Figure 4-4
indicate that, for the Camaro, Vega and Volkswagen, no large differences
in exhaust emissions exist. This is also the case for those vehicles
for composite exhaust results as indicated by Figure 4-5.
The Matador and New Yorker do show somewhat larger differences for
some exhaust components between the two prep cycles. The New Yorker has
higher HC and CO emissions for both Bag 1 and composite tests when the
HFET is used as a preconditioning drive when compared to emissions
levels seen when an LA-4 prep is used. Looking closer at individual
data shows that the differences in CO levels are not statistically
significant. However, differences in HC levels are statistically
significant. A look at the individual test data reveals that HC levels
increased with successive tests run with an LA-4 prep and that the HC
level measured during the last test was as high as HC levels measured
during the HFET preconditioning tests which were run some time later.
The differences in exhaust HC and CO levels could have been due to a
deterioration with time or other unaccounted for influences. The effect
of the different preconditioning drives cannot be discounted as a pos-
sible cause, but an adequate explanation of why this should have the
effect indicated cannot be found. The HC and CO emissions for Bag 1 and
composite samples for the Matador show just the opposite trend with the
higher emissions resulting when an LA-4 preconditioning drive is used.
This result is strongly influenced by test No. 142 in which high HC and
CO emissions occurred. The day after this test was conducted a recall
notice was received for the Matador describing the need to have the
-------
1.0.
o
I -8-
a
| .6.
oo
00 ~
c
3 -2-
|
~
\
\
\
\
\
fl rR
^ [^
s
\
s
\
\
s
\
s
\
\
s
s
\
\
s
V
^
\
s
\
> Camaro Matador New Vega Volks- 5 vchic;
Yorker wagen mean
w
&
V)
Figure 4-5a HC emissions
a 15
I
2
VI
3 10.
p"
5
m
s
\
^\i
i *
! |\
j [\
! (^
i i i
i N
j N!
\j i i_ i 1 "I
^ Mj ! y
N i ,>: ! -0
Canaro
S
\
S
\
N
\
\
\
1
i r\s *,
ra ^
1
1
i
j:Kt;-.(ic.r Now Vega Volks- 3 vehicle
"pvker wagen mean
Lr:!.g"jre 4-5 b CO emissions
I
ta
g
rt
CO
0)
5-
CM
O
U
1000
800-
600'
400'
200
Camaro
Volks-
wagen
Figure 4-5c CO. emissions
Matador New Vega
Yorker
5 vehicle
mean
2.5
o>
i-l
B ^"
2 1.5
M
.z
.5
-
R
S
s
s
s
s
\
S
S
s
s
s
\
s
\
s
F
s
\
\
\
\
\
\
\
X
N
N
\
s
s
\
\
\
\
s
s
H
\
\
\
\
\
\
Camaro Matador New Vega Volks- 5 vehicle
Yorker wagen mean
II LA-4 Prep
ESS HFET Prep
Figure 4-5d N0x emissions
-------
OQ
Oi
10
\
s
s
s
s
s
s
s
s
s
s
s
s
s
s
s
s
y
s
s
S
s
s
\
s
s
s
\
s
\
s
s
\
Camaro Matador
New
Yorker
Figure 4-4a
Veg
Volks-
wagen
HC emissions
200-
150
100.
50.
R
f~i\
i \!
rfl
! f\
i L
1
^
\
\
K
fe
K
ki f\
Ki
N
i\
N
f m
s
s
\
\
\
\
\
\
\
s
-
s
\
\
\
s
\
\
^
i
5 vehicle
mean
Canaro
New
Worker
Vega
Volks-
wagen
"inure 4-4b CO emissions
5 vehicle
mean
fl>
M
0)
CD
O
cn
M
ft
CD
O
O
Figure 4-^c CO. emissions
3000 .
2500
2000
1500 .
1000 .
500
\
S
S
S
s
s
s
s
s
s
V
v
V
s
s
V
s
s
s
s
V
T
s
V
\
s
s
\
s
s
\
s
s
N
S
S
s
V
T
s
s
\
S
s
s
,.
^
^
s,
Camaro Matador New Vega Volks- 5 vehicle
Yorker wagen mean
10
8
i
2 6
oo'
s
\
s
\
s
V
s
\
\
\
\
\
s
\
s
\
s
N
s
s
\
S
S
S
\
S
S
-
s
\
\
\
s
\
s
\
s,
\
1 1
^
\
\
\
\
\
X,
Canaro
I 1 LA-4 Prep
HFET Prep
Matador New Vega
Yorker
Volks-
wagen
Figure 4-4d NO emissions
5 vehicle
mean
-------
-17-
carburetor fuel inlet sealing plug and secondary throttle lockout level
replaced. One effect of this condition is to cause the throttle plates
to stick open if a cold engine is accelerated to an open throttle
position during the choking cycle. This particular test was run some-
time after the first two tests on the Matador were conducted using an
LA-4 prep. For this reason the exhaust emissions results of the Matador
tests are somewhat questionable. The only other large difference in
exhaust emissions was the difference in NOx levels between the two preps
for the Matador. Again, the above mentioned problem could have been
responsible for this result.
The figures also show that an average of all 5 vehicles tested shows
very little difference for the different preconditioning drives.
-------
-18-
5. Data Analysis
In order to determine if a statistically significant difference
exists for mean diurnal or exhaust emission levels for tests conducted
with either an LA-4 or HFET preconditioning drive, an analysis of var-
iance test was performed. The analyses are shown in Appendix B to this
report.
The analyses were based on the following hypotheses:
HOa: There is no statistically significant difference in emission
levels between the five vehicles tested.
HOb: There is no statistically significant difference in emission
levels between the two test sequences; one with an LA-4 and
the other with an HFET preconditioning drive.
HCc: There is no statistically significant difference in emission
levels due to the interaction of the test vehicle and the test
sequence used.
The analysis of variance test requires that the number of replicate
tests be the same for all tests. Three test results were chosen for
each vehicle and test sequence. When it was necessary to eliminate test
data, the middle values were used because only three tests were avail-
able for some of the vehicles. For example, if there were five test
results, the high and low values were omitted and the remaining three
values used.
The results of the analysis of variance tests are shown in Table 5-
1. For both evaporative and exhaust emission tests the levels exhibited
by the different vehicles were found to be different. A high level of
confidence (90%) can be placed on this conclusion. The type of precon-
ditiong drive used was found to significantly affect the diurnal emis-
sion levels only. A high level of confidence (95%) can be placed on the
conclusion that the type of preconditioning drive used does have an
effect on the diurnal losses encountered.
Test Type
Diurnal
HC
M CO
«>C02
£ NOx
0)
X HC
g CO
§-C02
,3 NOx
Hypothesis
HOa
Rejected
Rejected
Rejected
Rejected
Rejected
Rejected
Rejected
Rejected
Rejected
HOb
Rejected
Accepted
Accepted
Accepted
Accepted
Accepted
Accepted
Accepted
Accepted
HOC
Accepted
Accepted
Accepted
Accepted
Rejected
Accepted
Accepted
Accepted
Accepted
Table 5-1 Summary of Analysis of Variance Evaluation
-------
-19-
6. References
1. Federal Register. Vol. 39, No. 133, Section 85.076-12, July 10,
1974.
2. EPA In-House Test Program Report No. 1 - Vehicle Preconditioning:
AMA + LA-4 vs. LA-4, June 1975.
3. Vehicle Preconditioning Study, EFA-MSAPC In-House study conducted
by MSAPC Ann Arbor, Michigan.
4. R. E. Kruse and T. A. Huls, "Development of the Federal Urban
Driving Schedule." Paper 730553 presented at SAE Automotive
Engineering Meeting, Detroit, May 1973.
5. T. C. Austin, K. H. Hellman, and C. Don Paulsell, "Passenger Car
Fuel Economy During Non-Urban Driving". Paper 740592 presented at
SAE West Coast Meeting, Anaheim, California, August 1974.
6. Federal Register, Volume 39, No. 200, "Voluntary Fuel Economy
Labeling," October 15, 1974.
7. "Measurement of Fuel Evaporative Emissions from Gasoline Powered
Passenger Cars and Light Trucks Using the Enclosure Technique", SAE
Recommended Practice, SAE J171a, 1973 SAE handbook.
8. S. W. Martens and K. W. Thurston, "Measurement of Total Vehicle
Emissions". Paper 680125 presented at the SAE Annual Meeting,
Detroit, Michigan, January 1968.
-------
Appendix A-la Evaporative Emission Results
and Canister Weights for Tests Conducted
with an LA-4 Prep.
Test
No.
0014
0015
0016
0017
0024
0025
0026
0142
0030
0032
0034
0036
0134
0102
0114
0117
0094
0097
0101
Vehicle
Camaro
Camaro
Camaro
Camaro
Matador
Matador
Matador
Matador
New Yorker
New Yorker
New Yorker
New Yorker
New Yorker
Vega
Vega
Vega
Volkswagen
Volkswagen
Volkswagen
Diurnal
Loss,
grams
0.47
0-80
1.48
4.19
4.37
5.07
2.47
6.49
6.81
3.94
3.10
1.47
.692
.450
.480
1.075
.931
.835
Canister Weights, grans
Initial
LA-4
1080
1088
1030
1088
1073
1086
1085
1104
1090
1108
1108
1110
1097
1167
1168
1161
985
983
978
Final
LA-4
1076
1081
1082
1085
1081
1088
1090
1110
1075
1078
1084
1085
1088
1157
1150
1149
982
979
974
Initial
Diurnal
1079
1081
1084
1092
1078
1078
1082
1105
1094
1104
1108
1092
1108
1170
1163
1162
984
979
.974
Final
Diurnal
1096
1084
1103
1109
1102
1103
1105
1124
1116
1124
1128
1114
1128
1182
1176
1175
996
989
986
Final
Hot Soak
1088
1082
1090
1086
1089
1096
1106
1112
1109
1168
983
977
974
-------
Appendix A-lb Evaporative Emission Results
and Canister Weights for Tests conducted
with an HFET Prep.
Test
No.
0031
0035
0044
0047
0049
0060
0063
0033
0037 -
0046
0048
0050
0061
0065
0059
0062
0067
0119
0120
0123
0126
0112
0113
0115
,0116
0118
'0121
0122
0127
Vehicle
Camaro
Camaro
Camaro
Camaro
Camaro
Camaro
Camaro
Matador
Matador
Matador
Matador
Matador
Matador
Matador
New Yorker
New Yorker
New Yorker
Vega
Vega
Vega
Vega
Volkswagen
Volkswagen
Volkswagen
Volkswagen
Volkswagen .
Volkswagen
Volkswagen
Volkswagen
Diurnal
Loss,
grains
1.73
1.08
0.54
0.53
0.61
,1.01
0.51
5.34
2.91
4.05
2.50
5.30
3.99
1,77
2.51
3,68
2.68
0,52
0.43
0,52
0,45
1.16
0.96
0.75
0.91
0.78
0.77
0.82
0.72
Canister Weights, grains
Initial
HFET
1114
1092
1099
1097
1100
1101
1088
1091
1098
1109
1110
1100
1094
1098
1099
1080
1165
1154
1151
1148
964
974
970
968
969
970
971
976
Final
HFET
1090
1113
1088
1086
1084
1091
1091
1090
1088
1090
1089
1099
1092
1079
1081
1076
1152
1148
1145
1142
961
969
966
964
967
968
969
971
Initial
Diurnal
1098
1094
1092
1093
1087
1102
1095
1088
1085
1089
1087
1094
1096
1093
1097
1102
1106
1167
1159
1160
1158
961
969
966
966
968
968
969
972
Flnnl
Diurnal
1115
1110
1110
1105
1117
1112
1113
1110
1116
1112
1119
1119
1114
1120
1124
1125
1179
1171
1174
1171
973
980
978 '
978
980
980
981
983
Final
FTP
1098.
1100
1110
1103
1103
1104
1109
1110
1112
1118
1110
1097
1096
1093
1154
1151
1151
1149
963
970
968
968
970
971
971
973
-------
Test
No.
0014
0015
0016
0017
0024
0026
0142
0030
0036
0134
0102
0114
0117
0094
0097
0101
Vehicle
Camaro
Camaro
Camaro
Camaro
Matador
Matador
Matador
New Yorker
New Yorker
New Yorker
Vega
Vega
Vega
Volkswagen
Volkswagen
Volkswagen
Bag 1 Exhaust results (grams}
HC
4.88
4.54
4.83
4.96
3.35
3.50
5.46
4.04
4.06
4.79
9.27
7.27
6.64
4.61
3.83
4.06
CO
84.8
.72.6
93.6
84.6
48.9
49.0
82.0
173.3
157.6
203.6
117
105
115
35.4
36.1
34.0
co2
2334
2317
2325
2301
2380
2331
2478
3083
3102
3100
1210
1218
1188
1223
1299
1241
NOx
7.27
7.71
7.22
7.89
9.19
8.69
9.29
8.13
9.61
7.63
5.61
6.10
5.89
8.02
6.26
6.71
Composite exhaust results (g/mi)
HC
.528
.459
.524
.507
.284
.283
.703
.210
.372
.462
.765
.675
.679
.892
.832
.941
CO
9.32
.7.44
11.10
9.91
3.06
3.01
7.29
10.4
14.5
21.6
11.5
10.6
13.3
4.27
4.39
4.31
co2
647
643
645
637
649
637
671
818
879
858
358
353
348
361
368
358
NOx
1.24
1.38
1.16
1.21
2.27
1.99
2.28
1.72
2.46
1.88
1.46
1.60
1.46
1.79
1.37
1.42
i?
s
B.
8
Ni
MHO)
5> (t>
"\ en w
HOC
mow
T3 3 rt
o-
rt H-
(D CO
Q. CD
PI CD
3 rt
-------
Test
No.
0031
0035
0044
0047
0060
0063
0037
0045
0046
0048
0050 .
0061
0063
0066
0052
0054
0059
0062
0064
0119
0123
0124
0126
0112
0113
0115
0116
0118
Vehicle
Camaro
Camaro
Camaro
Camaro
Camaro
Camaro
Matador
Matador
Matador
Matador
Matador
Matador
Matador
Matador
New Yorker
New Yorker
New Yorker
New Yorker
New Yorker
Vega
Vega
Vega
Vega
Volkswagen
Volkswagen
Volkswagen
Volkswagen
Volkswagen
Bag 1 exhaust (grams)
HC
4.17
5.16
5.99
4.79
4.62
4.24
3.61
3.83
3.36
3.66
3.23 -
3.24 '
3.59
3.76
6.60
5.46
6.03 .
5.29
4.21
6.60
7.02
9.01
9.01
5.17
4.14
4.44
4.02
4.35
CO
76.2
86.7
115.2
91.1
91.6
73.6
-55.0
57.0
55.7
60.6
54.0
40.8
51.3
45.0
222
205
232
177
161
111.9
116.2
100.4
128.2
36.9
35.0
37.0
39.2
39.2
C02
2335
2340
2372
2397
2435
2303
2466
2520
2390
2463
2548
2450
2351
2404
3128
3041
3092
3073
3003
1208
1212
1227
1173
1265
1262
1251
1254
1298
NOx
7.84
7.22
7.89
8.15
6.96
7.46
9.52
10.33
10.90
11.17
12.23
10.01
10.61
9.64
10.07
8.86
7.43
7.53
7.86
5.84
6.19
5.19
5.75
6.20
5.95
6.98
6.61
6.28
Composite exhaust (grams/mile)
HC
.370
.514
.568
.524
.431
.374
.297
.310
.302
.294
.274
.268
.303
.296
.525
.419
.489
.388
.359
.630
.606
.748
.754
1.07
.891
.741
.930
.860
CO
6.80
9.23
14.3
13.4
10.5
8.07
, 3.57
3.51
3.48
3.55
3.99
2.41
3.07
2.60
20.4
15.8
20.2
14.8
14.3
. 11.3
11.7
11.9
12.9
4.73
4.49
4.50
5.00
4.78
C02
649
640
640
657
674
fi-57
657
644
651
667
669
661
637
677
882
853
867
868
848
358
362
359
354
360
365
353
361
369
NOx
1.31
1.23
1.21
1.45
1.12
1 .17
2.44
2.34
2.70
2.77
2.85
2.52
2.60
2.54
2.61
2.37
1.89
1.88
1.89
1.48
1.54
1.47
1.52
1.34
1.34 ,
1.51
1.45.
1.40
as
01 «
Ml
o >
H I
S3
H rt X
CO P*
n n £
(D O CO
*o 3 rt
rt H-
(D CO
a- n
H- 0
rt
"I?
P> 0)
P rt
-------
Appendix B Analysis of Variance
Analysis, of .Variance For Diurnal .Loss Results
Vehicle
Test
LA-4 Prep
HFET Prep
Tc
Canaro
0.47
0.80
1.48
0.61
1.01
0.54
'4.91 '
Matador
4.19
4.37
5.07
2.91
4.05
3.99
24.'58
New Yorker
6.49
3.94
3.10
2.51
3.68
2.68
22.40
Vega
0.69
0.45
0.48
0.52
0.52
0.45
3.11
Volkswagen
1.08
0.93
0.84
0.78
0.82
0.91
5.36
Tr
34.38
25.98
T= 60.36
No. of columns, c = 5
No. of rows, r = 2
.1 2/N -121 .'44
.JE.«* ° 208.66
Tjc2 - 1168.45
I Tr2 «* 1856.94
£ Tcr2 =598.91
SS,
ZTc 2/n-r - T2/N = 73.30
SSr = ETr 2/n'c - T2/N
SS
'cr
*cr
2/n - T2/N - SSC - SSr > 2.54
SS.
SS
'res
EX2 - T 2/N - 87.22
= . SSt - SSC - SSr - SS
9.03
No. of replicates, n = 3
Total No. of tests, N = 30
Source of
Variation
Vehicles
Test Type
Vehicle-Test
Interaction
Rostdual
Total
SS
73.30
2.35
2.54
9.02
87.22
DF
5-1 = 4
2-1=1
(4) (1) = 4
29 - 9 = 20
30 - 1 - 29
M S,
(SS/DF)
18.33
2.35
0.64
0.45
MSR,
(MS/MS res)
40.73
5.22
1.42
<
>
>
>
<
F
(a- 0.05.)
2.87
4.35
2.87
Hoa: Rejected a 95% C.L.
Hob: Rejected a 95% C.L.
Hoc: Accepted
-------
-2-
Analysis of Variance for Bag 1 HC Results
Vehicle
Test
LA-4 Prep
HPET Prep
TC
Ctaaaro
4.88
4.83
4:96
5.16
4.79
4.24
28.86
Matador
3.35
3.50
5146
3.61
3.66
3.59 i
23.17
Hex* Yorker
4.04
4.06
4; 79
5.46
6.03
5.29
29.67
Vef>a
9.27
7.27
6.64
7.02
9.01
9.01
48.22
Volkswagen
4.61
3.83
4.06
4.14
4.44
4.35
25.43
Tr
75.55
79.80
T-15S.V5
No. of columns, c 5
No. of rows, r 2
T 2/N = 804.45
£ *? "884.49
Tic2 - 5221.9
£ Tr2 = 12076
E Tcr2 = 2621.5
SSC = ETC 2/n-T - 1'2/N - £5_»8J_.
SS.
SS
cr
ETi: 2/n'c. - T?/N = Q.62
-rT 2/n - T2/K - SSC - s
No. of replicates, n ° 3
T^tal No. of tests, N = 30
2.89
SS
'res
. SSt - SSC - SSr - S3cr - 10.66
Source of
Variation
Vehicles
'.''esr. T\T)e
Vehir.le-Tost
.Interaction
^*-sidual
Total
SS
65.87
0.62
2.89
10.66
80.04
OF
5-1 - 4
2-1 " 1
(4) (1) ." 4
29-9-20
30 - 1 = 29
M S,
CSS/DF)
16.47
0.62
0.72
0.53
MSR,
CMS/MS res)
30.90
1.17
1.36
<
>
>
<
<
F
(a- 0.10 )
2.25
2.97
2.25
lion: Rejected a 90% C.L.
Hob: Accepted
Hoc: Accepted
-------
-3-
Analysis of Variance for Bag 1 CO Results
Vehicle
Test
LA-4 Prep
HFET Prep
TC
Caciaro
84.8
93.6
84.6
86.7
91.1
91.6
532
Matador
48.9
49.0
82.0
55.0
55.7
54.0
344
New Yorker
173
158
204
222
205
177
1T?Q
Vega
117
105
115
112
116
128
693
Volkswagen
35.4
36.1
34.0
36.9
37.0
39.2
219
Tr
. 1420
1507
T- 2927
No. of columns, c - 5
No. of rows, r » 2
.1 2/n _ 2.8558 x 105
£
- 3.7533 x 105
T'Tc2 - 2.2269 x 106
E Tr2 -
E Tcr2
SSC
SS.
SScr -
sst
SS
res
. 4.2874 x 106
- 1.1163 x 106
ETc 2/n'r - T2/N - 8.557 x 10*
ETr 2/h-c - T2/N - 2.467 x 102
ETcr 2/n - T2/N - SSC - SSr - 7.0330 x 102
EX2 - T 2/N - 8.975 x 104
« . SSt - SSC - SSr - SScr " 3.230 x 103
Mo. of replicates, n - 3
Total. Ho. of tests, N - 30
Source of
Variation
Vehicles
Test Type
Vehicle-Test
Interaction
Residual
Total
SS
85570
246.7
703.3
3230
89750
DF
5-1 - 4
2-1 - 1
(4) (1) - 4
29-9-20
30-1-29
M S,
(SS/DF)
21390
246.7
175.8
161.5
HSR,
(MS/MS res)
132.4
1.53
1.09
<
>
>
<
<
F
(a= 0.10 )
2.25
2.97
2.25
Hoa: Rejected a 90% C.L.
Hob: Accepted
Hoc: Accepted
-------
-4-
^^Analysis of Variance for Bag l.Cp2 Results _
Vehicle
Test
LA-4 Prep
HFET Prep
Tc
Camaro
2334
2317
2325
2340
2372
2397
14085
Matador
2380
2331
2478
2466
2463
2450
14568
New Yorker
3083
3102
3100
3041
3092
3073
18491
Vega
1210
1218
1188
1208
1212
1227
7263
Volkswagen
1223
1229
1241
1265
1262
1254
7474
Tr
30759
31122
T" 61881
No. of columns, c = 5
No. of rows, r - 2
.T 2/N -1.276 x 108
.A*7 "1.
436 x 10
8
O0? "8.611 x 108
E Tr2 =1.915 x 109
I Tcr2
SS
= 4.306 x 10'
,8
No. of replicates, n ** 3
Total No. of tests, N - 30
2/n-r - T2/N = 1.502 x in7
SSr ° ZTr 2/h-c - T2/N - 4.392 x 1Q3
SScr =
£Tcr 2/n - T2/N - SSC - SSr - 8.941 x 103
SSt = EX2 - T 2/N = 1.600 x 10^
SS
res
. SSt - SSC - SSr - SScr - 6.667
Source of
Varintion
Vehicles
Test Type
Vehicle-Test
Interaction
t^UM.ll
Total
SS
1.592 x 10?
4.392 x 103
8.941 x 103
6.667 x 10*
1.600 x 10?
DF
5-1 = 4
2-1 = 1
(4) (1) - 4
29 - 9 = 20
30-1-29
M S,
(SS/DF)
3.980 x 106
4.392 x 103
2.235 x 103
3.333 x 103
MSR,
(MS/MS res)
1194
1.318
.671
<
>
»
<
<
F
(a- 0.10 )
2.25
2.97
2.25
Hoa: Rejected a 90%'C.L.
Hob: Accepted
Hoc: Accepted
-------
-5-
-Aaalysis of Variance for Bag.1 NO Results
X
Vehicle
Test
LA-4 Prep
HFET Prep
TC
Caaaro
7.27
7.7.1
7.89
7.84
7.89
7.46
46.06
Matador
9.19
8.69
9.29
10.3
10.9
10.6
58.97
New Yorker
8.13
9.61
7.63
8.86
7.53
7.86
49.62
Vena
5.61
6.10
5.89
5.84
6.19
5.75
35.38
Volksw.ipon
8.02
6.26
6.71
6.20
6.61
6.28
40,08
Tr
114.00
116.11
C= 230.11
No. of columns, c = 5
No. of rows, r = 2
.T 2/N ° 1765.02
E x^T = 1830.06
Tjc2 =10919
Z Tr2 = 26477
S Tcr2 "5472.86
STc 2/n-T - T2/N = 54.81
SS
SS
SS,
'cr
fit 2/h'c - T2/N = 0.148
£Tcr 2/n - T2/N - SSC - SSr - 4.309
SSt = £X2 - T 2/N = 65.090
SS
res
= . sst - ssc - ssr - ss(
"cr
5.823
V-.i. of replJchtfes, n = 3
Totf.l So. c~. tests, N « 30
Source of
Variation
Vehicles
Test Type.
Vehicle-Test
Interaction
jp-^^idual
Total
SS
54.81
0.148
4.309
5.823
65.090
DP
5-1 " 4
2-1 = 1
f/0 Q) = 4 '
>9 - 9 - 20
30 - 1 - 29
M S,
(SS/DF)
13.70
0.148
1.077
0.291
MS«,
(MS/MS res)
47.06
0.509
3.702
<
>
>
<
>
P
(** 0.1O)
2.25
2^97
2.25
Hoa: Rejected a 90% C.L.
'Hob: Accepted
Hoc: Accepted
-------
-6-
Analysis of Variance for Composite HC Results
Vehicle
Test
LA-4 Prep
HFET Prep
Tc
Camaro
.528
.524
.507
.514
.524
.431
3.028
Matador
.284
.283
.703
.297
.302
.296
2.165
New Yorker
.210
.372
.462
.419
.489
.388
2.340
Vega
.765
.675
.679
.630
.748
.745
4.242
Volkswagen
.892
.832
.941
.891
.930
.860
5.346
Tr
8.657
8.464
TaL7.121
No. of columns, c " 5
No. of rows, r " 2
T 2/N "9.771
z"xT-.li.2Q3
IQc2 "65.906
Z Tr2 "146.583
Z Tcr2 "33.059
SSC - ZTc 2/n'T - T2/N " -1.213
SSr " ZTr 2/ixrc - T2/N - .00124
SScr - ZTcr 2/n - T2/N - SSC - SSr - .03443
SSr - ZX2 - T 2/N = 1.4320
No. of replicates, n " 3
Total No. of tests, N 30
SS
'res
. SSt - SSC - SSr - SScr - .1833
Source of
Variation
Vehicles
Test Tvpe
Vehicle-Test
Interaction
pesidual
Total
SS
1.213
1.24 x 10-3
J.443 x 10~2
.1833
1.4320
DF
5-1 - 4
2-1 = 1
(4) (1) " 4
29 - 9 - 20
30-1-29
M S,
(SS/DF)
0.303
1.24 x lor3
J.608 x 1(T3
9.17 x 10~3
HSR,
(MS/MS res)
33.06
.135
.939
<
>
>
<
<
F
(a-0.1 )
2.25
2.97
2.25
Hoat Rejected a 90% C.L.
Hob: Accepted
HOC. Accepted
-------
-7-
J.Aua,lxsia_Qf~2axiance. .for Composite..CO .Results....
Vehicle
Tfist
LA-4 Prep
HKRT Prep
Tc
Camaro
9.32
ll.l'
9.91
9.23
13.4
10.5
63.46
Matador
3.06
3.01
7.29
3.51
3.48
3.55
23.90
New Yorker
10.4
14.5
21.6
15.8
20.2
14.8
97,30
Vega
11.5
10.6
13.3
11.7
11.9 .
12.9
'71.90
Volkswagen
4.27
4.39
4.31
4.73
4.50
4.78
26.98
Tr
138.56
144.98
T" 283. 54
No. of columns, c = 5
No. of rows, r = 2
T 2/M «. 2679.8
E x2" ° 3441.5
Tic2 - 19963
E Tr2 = 40218
£ Tcr2 = 9999.9
SSC = Etc 2/n-r - T2/N - 647.4
SSr = ETr 2/n-c - T2/N - 1.374
SS
cr
sse
r,s
'res
ETcr 2/n - T2/N - SSC - SSr - 4.726
EX2 - T 2/N = 761.8
= . SSt - SSC - SSr - SScr
108.3
No. of replicates, n 3
Total No. of tests, N <* 30
Source of
Variat ion
Vehicles
Test Type
Vehicle- Test
Interaction
Residufi.1
TM.&1
SS
647.4
1.374
4.726
108.3
761.8
DP
5-1 - 4
2-1 '" 1
(1) (1) - A
29-9-20
30 - 1 > 29
M S,
(SS/DF)
161.8
1.374
1.182
5.415
MSR,
(MS/MS res)
29.9
.254
.218
<
>
>
<
<
F
(a*0.1 )
2.25
2.97
2.25
Has. Rejected a 902 C.L.
Hob- Accepted
Hoc: Accepted
-------
-8-
Aoalygi_s_Qf ^Variance, for Composite. CCL Results
, Vehicle
Test
LA-4 Prep
HFET Prep
TC
Camaro
647
643
645
649
640
657
3881
Matador
649
637
671
657
667
661
3942
New Yorker
818
879
858
853
867
868
5143
Vepa
358
353
348
358
362
359
2138
Volkswagen
361
368
358
360
365
361
2173
Tr
8593
8684
T=17277
No. of columns, c ** 5
No. of rows, r = 2
T 2/N ° 9.9498 x 106
. E. x2 - 1.1Q6 x 107
jEJc2 = 6.6345 x 107
E Tr2 = 1.4925 x 108
I Tcr2 = 3.3174 x 107
SSC = ETC 2/n'T - T2/H = 1.1077 x 106
SSr = ETr 2/h-c. - T2/N " 276.03
SScr = ETcr 2/n - T2/N - SSC - SSr = 90.00
SSt = EX2 - T 2/N = 1.1102 x 10^
SSres - . SSt - SSC - SSr - SScr = 2.1340 x
No. of r^nliccitee, n - 3
Total No. of tests, N - 30
Source of
Variation
Vehicles
Test Type
Vehicle-Test
Interaction
Residual
Total
SS
1.1077 x 106
276.03
90.00
2134
1.1102 x 106
DF
5-1 = 4
-2rl .- 1
(4) (1) -= 4
29 - 9 - 20
30 - 1 - 29
M S,
(SS/DF)
2.769 x 105
276.03
22.50
106.7
HSR,
(KP/MS rea)
2595
2.59
.2 x 109
<
}
»
<
<
< .
( a-O. IP )
2.25 j
,., . , |
2.97 !
2.25
Hoa: Rejected a 90% C.L.
Hob: Accepted
Hoc: Accepted
-------
-9- '
Analysis..of.Variance for Composite NO Results
Vehicle
Test
LA- A Prep
HFET Prep
TC
Camaro
1.24
1.38
1.21
1.31
1.23
1.37
7.74
Matador
2.27
1.99
2.28
2.70
2.60
2.54
14.38
New Yorker
1.72
2.46
1.88
2.37
1.89
1.89
12.21
Vepa
1.46
1.60
1.46
1.48
1.54
1.52
9.06
Volkswagen
1.79
1.37
1.42
. 1.34
1.45
1.40
8.77
Tr
25.53
26.63
T-52.16
No. of columns, c = 5
Mo. of rows, r " 2
T 2/N - 90.69
E x? =96.78
~E Tc2 = 574.77
L Tr2 - 1360.94
£ Tcr2 = 288.31
S:>c. = ETC 2/ii'c - T2/N = 3.11
SSr = ZTr 2/n-c - T2/H « O.OA03
SScr = "£Tcr 2/a - T^/N - SSC - SSr - 0.2658
SSt - ZX2 - T 2/N - 6»09
ssix-s = - SS£ - SSC - S£r - SScr - 0.6739
No. of replicates, n 3
Total Mo. of tests, N - 30
Source of
Variation
Vehicles
Test Type
Yehi-lo-Tesc
J-.itcractior.
_g.csidual
Total
SS
5.11
0.0403
0.2658
0.6739
6.09
DF
5-1 <* 4
?-i - 1
(4) (1) = '
29 - 9 = 20
30 - 1 - 29
H S,
(SS.'DF)
1.28
0.0403
0.0665
0.0337
>;SR,
CMS/MS re-a)
38.0
1.20
1.97
<
>
>
<
<
F
(« = 0.10 >
2.25
2.97
2.25
lloft: Rejected <» 90Z C.L.
Hob: Accepted
H"c: Accepted .
------- |