MC 75-03
Technical Support Report for Regulatory Action
The Effect of Dynamometer Inertia Increment and
Power Setting on Emissions and Fuel Economy
for Motorcycles
August, 1975
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 conclusions and/or
recommendations resulting from the assumptions and constraints of that
analysis. Agency policy constraints or data received subsequent to
the date of release of this report may alter the conclusions 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
-------
ABSTRACT
The effects of dynamometer inertia increment and horsepower
are determined from dynamometer tests of two motorcycles. An inertia
increment of 10 kg is recommended based on the data. Data on the
effects of actual driving distance on emissions and fuel economy, as
measured by a dynamometer roll counter, are also presented.
Prepared by
Approved - Project Manager
Motorcycles
Approved - Branch,Chief, SDSB
Approved - Division Director
-------
INTRODUCTION
In developing the motorcycle regulations, the increment of inertia
weight to be set on the dynamometer was picked as 10 kg.* For example,
any motorcycle with a loaded vehicle mass of 256 to 265 kg would be
tested at a dynamometer inertia setting of 260 kg. Corresponding to
the inertia setting, a road load horsepower also would be set on the
dynamometer; the horsepower increment is approximately .04 horsepower
per 10 kg inertia increment.
Several comments were received concerning the inertia increment.
Concern was voiced within EPA that the inertia increment might need to
be smaller, 5 kg for example, in order to minimize any bias to the fuel
economy measurement. In a comment to the ANPRM, Honda presented an
opposing point of view. They felt a 20 kg increment was sufficient,
and stated their data showed little effect on emissions. No data were
presented, however.
To evaluate these comments, and to provide data to support the
choice of an inertia increment, the effect of inertia on emissions and
fuel economy needed to be quantitatively defined. And since road load
horsepower setting, as currently defined in the draft regulations,
is based on inertia increment, its effect on emissions and fuel economy
required evaluation. A test program involving measuring emissions and
fuel economy at various inertia and road load horsepower settings was
conducted. Two motorcycles, a Honda CB360 and a Yamaha 125 Enduro,
were tested on a Shibaura dynamometer using the Urban Dynamometer Driving
Schedule (UDDS). The Honda was chosen as representative of a middle
weight 4-stroke motorcycle, the 2-stroke Yamaha representing the lighter
weight motorcycles. The inertia weights of the motorcycles, with an
80 kg rider, are
Honda CB360 245 kg
Yamaha 125 Enduro 195 kg
An attempt was made to test the lightest motorcycle available at EPA,
the Honda CT70 at 135 kg, but it could not come close to following
the driving trace.
A secondary objective was to evaluate the magnitude of variations
in the actual distance covered while driving the UDDS, and determine
the effects on emissions and fuel economy. It was felt that actual
driven distance may vary, and that if emissions and economy were cal-
culated using the standard distance, an error would result. Therefore,
a dynamometer roll revolution counter was installed and driving distance
data were generated.
Summary and Conclusions
. For a 5 kg change in inertia and a .02 change in dynamometer
horsepower (the maximum combined test deviation from the actual
*The smallest inertia increment available on the Shibaura dynamometer
is also 10 kg.
-------
-2-
inertia and road load horsepower for a 10 kg inertia increment),
the effect on fuel economy is less than 1/2 percent of nominal
value.
. The maximum effect on emissions for a 5 kg change in inertia and
a .02 change in road load horsepower is less than 2 percent for
NOX. The effects on CO and HC are less than 1 percent.
. The results of this test series are expected to be applicable to
other uncontrolled motorcycles and to controlled motorcycles with
the exception of the percentage change in NOX mass emissions,
which could increase with certain control techniques.
. Based on these data, a 10 kg inertia increment is recommended.
. Variations of up to 3 percent of the standard test distance were
encountered. It is expected that similar variations in emissions
and fuel economy could occur if calculated using the standard
test distance. Future evaluation is recommended.
Discussion
A. Effect of Inertia Increment
j
Bag 1 and 2 (hot start) tests were performed using the Honda CB360
and the Yamaha 125 Enduro. Inertia setting was varied with 3 repeat
tests performed at each inertia setting, as shown in Table 1.
Table 1
Test Series-Dynamometer Inertia
// of
Inertia Settings
11
9
// of Tests at
Each Setting
3
3
Range of Inertias
Tested
200 - 300 kg
150 - 230 kg
Dyno
Horsepower
5.8
4.9
@ 50 mph
Motorcycle
Honda CB360
Yamaha 125
Enduro
The test to test repeatability at a given inertia setting was good,
as shown in the Appendix.
The measured dynamometer roll distance varied up to 3 percent of
the standard distance. The data, calculated using the standard and
measured distances, were compared; the data using measured distance
showed the higher degree of correlation. Therefore measured distance
was used in all the calculations for this test series.
To determine the relationship between emissions (economy) and dyna-
mometer inertia, a linear regression analysis was performed. The means
-------
-3-
of the three tests at each inertia setting were calculated, and the
regression equation was determined using the means. The regression
equation has the form,
Emissions
a. * inertia + aQ
where a is the slope, i.e. the rate of change of emissions per unit of
inertia.
With inertia set on the dynamometer in increments of 10 kg, the
largest possible deviation of actual inertia from the test value is
5 kg, i.e., one half the increment. Table II presents the effects on
emissions and fuel economy of a 5 kg change in inertia. In interpreting
the results, the level of correlation, as presented in the Appendix,
should be taken into consideration.
Table II
Change Due to a 5 kg Change in Inertia
Motorcycle
Honda CB360
Yamaha 125 Enduro
Parameter
CO
HC
NO
„ x
Economy
CO
HC
NO
« x
Economy
% Change from
Nominal Value
.30
.65
1.1
-.25
(1)
.58
(2)
-.33
Magnitude of Change
.11 gins/km
.01 gms/km
.002 gms/km
-.06 km/£
(1)
.05 gms/km
(2)
-.08 km/fc
(1) Insufficient correlation at a 95% confidence level to allow
prediction.
(2) The NOX emissions from current 2 strokes are so low
that any effect could not be measured.
B. Effect of Dynamometer Horsepower
ppm)
Bag 1 hot start tests using the Honda CB360 were run to determine
the effect on emissions and fuel economy of dynamometer horsepower.
Tests on a lighter, low power motorcycle were not completed because
the motorcycle could not operate ^ver a wide enough range of dynamometer
horsepower settings. Table III describes the tests performed using
the Honda.
-------
-4-
Table III
Test Series-Dynamoiaeter Horsepower
it of tests
// of HP at each Range of dyno HP Inertia
Motorcycle jSejtirigs^ HP Setting Settings Tested kg
Honda CB360 7 5 4.7-7.7 250
A linear regression analysis was also performed on these data to
determine the relationship between emissions (fuel economy) and dynamometer
horsepower. The regression statistics are shown in the Appendix.
The road load horsepower increment corresponding to the proposed
10 kg inertia increment is .04 horsepower. The maximum variation of
the actual road load horsepower from the dynamometer setting is one
half the increment, or .02 horsepower. Table IV presents the effects
on emissions and economy for a .02 horsepower change. There was insuffi-
cient correlation of HC emissions to predict a relationship.
Table IV
Change Due to a .02 Change in Dynamometer HP Setting
% Change from
Motorcycle Parameter Nominal Value Magnitude of Change
Honda CB360 CO .04 .015 gm/km
N0x .44 .001 gm/km
Economy -.06 -.014 km/2,
C. Combined Effect of Road Load Horsepower and Inertia
If a motorcycle is tested at a higher than actual inertia, the
corresponding dynamometer horsepower setting may also be high. Therefore,
the maximum effect on the motorcycle's emissions and fuel economy is
the superposition of the independent effects. For the proposed 10 kg
inertia increment, the combined effects on emissions and economy are
small,, less than 2 percent for all emission parameters and less than 1/2
percent for economy.
The magnitudes of the effects do not support the necessity of a 5 kg
inertia increment. A larger increment, such as 20 kg, is also not
recommended due to the uncertainty of the applicability of the NOX results
to controlled motorcycles. Therefore, based on these data, the 10 kg
inertia increment, as proposed in the ANPRM, is the best choice.
-------
-5-
D. Applicability to Other Motorcycles
It is anticipated that the percentage change in emissions and fuel
economy due to inertia increment will be of the same magnitude for all
uncontrolled motorcycles. Comparison of the Yamaha and Honda data
show the same order of magnitude effects. The Yamaha is representative
of the lightest class of motorcycles, and the effect of inertia increment
should be the worst case for this motorcycle,, In any case the ratio of
inertia increment to total motorcycle inertia only varies between 1 1/2
to 3 1/2 percent for the heaviest (HD 1200) to lightest (Honda CT 70)
motorcycle. Thus, the effect of inertia increment is a small percent
of the total inertia for all motorcycles, and the effects determined
in this test series should be of the same order of magnitude for all
motorcyles.
For controlled motorcycles the percentage change in emissions and
economy is difficult to predict due to the uncertainty of the control
techniques. A likely control method which may increase the effects
determined from these data is leaning the air fuel ratio. At leaner
mixtures the rate of formation of NOX increases and could result in
percentage changes in NOX mass emissions due to inertia increment
several times larger than determined from these tests.
E. Physical Interpretation of the Effect of Horsepower and Inertia
The increase in HC and CO mass emissions with increased load
determined in this test program is primarily a result of the increased
mass flow of fuel air mixture through the engine. Other uncontrolled
gasoline engines have shown a similar effect due to load (^). An
increase in load will cause an advance in throttle position and
possible change in air fuel ratio; however, the magnitude of the
change in load being considered is small and any air fuel ratio
effects should be minimal.
An increase in load increases both the concentration and the mass
emissions of NO. The increased load increases the combustion temper-
ature, on which the formation of NO is highly dependent. The data
support these statements.
Inertia acts as a load to be overcome during acceleration and
deceleration, therefore, the description of the horsepower effects
applies to inertia. An increase in throttle setting during acceleration
and decrease during deceleration occurs, and with large advances in
throttle, the carburetor may deliver a richer air fuel ratio. An examin-
ation of the HC and CO concentration data showed no change in concen-
trations at either end of the inertia range tests, indicating no air
fuel ratio caused bias occurred. This supports the relationships obtained
by linear curve fitting. The HC and CO data both show a linear increase
in mass emissions with increased inertia. The N0x data also show the
expected increase with higher inertia.
The trend is not clear. Data from some gasoline engines have
shown no change in HC and CO mass emissions with increases in load.
-------
-6-
Fuel economy for both increased inertia and horsepower shows a
decrease, as would be expected from the increased mass flow of fuel.
F. Distance Variations
When driving the Federal Urban Dynamometer Driving Cycle, variations
in the actual driving distance can be expected due to the driver not
being able to exactly follow the driving trace. To determine the
magnitude of the variations, dynamometer roll distance was measured on
most of the tests. Table V presents the range of distance variations
encountered.
Table V
Variations in Dyno Roll Distance
Motorcycle Standard Distance, km Range Mean a.
Yamaha 125 Enduro 4.665 Bag 1 4.525-4.761 4.651 .066
Yamaha 125 Enduro 6.210 Bag 2 5.998-6.334 6.163 .089
Honda CB360 5.779 Bag 1 5.724-5.858 5.774 .029
Two sigma variations from the standard distance are as large as 3 per-
cent for the Yamaha 125 Enduro.
It is expected that emissions and economy, if calculated using the
standard distance, will vary in proportion to the deviation of actual
distance from the standard distance. In other words if the UDDS distance
was 3 percent long, calculated mass emissions and economy could vary on
the order of 3 percent.
The measured effect of distance on emissions was not determined
due to an insufficient amount of data. Most of the deviations from the
standard distance were small, with only a few values approaching the
3 percent deviation mentioned above. A regression line was calculated
but due to the heavy weighting near the mean, it indicated no relation-
ship between distance and emissions existed.
It is recommended that roll distance be measured on future tests
and that distance variations and measurement techniques be investigated
in the variability study.
-------
APPENDIX
Presented in this Appendix are the statistics of the regression
analyses and a description of the application of the regression equations.
At a given inertia and horsepower setting, the test to test repeat-
ability was good, as shown for the Honda CB360 in Table A-l. The test
repeatability of the Yamaha was similar.
Table A-l
Typical Test Repeatability at Constant Inertia and Horsepower Setting
Honda CB360*
Inertia
kg
260
210
210
290
1
1
2
2
HC
Mean
1.77
1.65
1.57
1.83
£
.10
.07
.01
.04
CO
Mean
36.30
36.78
33.40
35.85
a_
.90
1.02
.56
.57
NOx
Mean £
.18 .01
.15 .01
.05 .00
.05 .00
Fuel Economy
Mean a_
.2
.4
.6
.2
^Emissions in gms/km
fuel economy in km/A
The regression statistics for the inertia test series are presented
in Table A-2. The bag showing the highest degree of correlation is pre-
sented in the table. In many cases the other bag showed insufficient
correlation at the 95 percent level. The largest amount of variation
explained by any regression equation is 85% (r2 for the Yamaha HC is .85)
with all other relations showing a lesser degree of correlation. If
determining an exact relationship between emissions and inertia was
required, these data would have only limited value due to the low con-
fidence, in the correlation.
Table A-2
Inertia Regression Statistics
Regression Parameter.
HONDA CB360
r reqd. for
95% confidence
Significant @ 95% ?
Slope, a-i
HC
.71
.60
yes
.0024
2
CO
.64
.60
yes
.0224
2
N°-x
.86
.60
yes
.00036
1
Economy
-.71
.60
yes
-.01265
2
-------
-A-2
Table A-2 (continued)
YAMAHA 125 ENDURO
r .92 .56 0 -.92
r reqd. for
95% confidence .67 .67 .67 .67
Significant @ 95% ? yes no no yes
Slope, a^ .0100 .0117 0 -.0163
Bag 2221
But the objective is to evaluate the inertia increment size, and to do
this an exact relationship is not required. The data can provide an
estimate of the size of the change in emissions (economy) as a function
of inertia. The level of correlation, however, should be kept in mind
when using the specific results.
The results of the regression analysis are used as follows. The
first derivative of the regression equations gives the change in emissions
(economy) per unit change in inertia. For example, the regression equa-
tion for the Honda NOX is
N0v = .00036 * inertia + .0758
* km
and the slope (first derivative) is
.00036 gms/km _
1 kg al
The percent change of nominal NOX emissions for the Honda is
.00036 ems/km * 100 _„„ ,
1 kg * 1.6 gms/km " 'iih *B
where 1.6 gms/km is the average NOX mass emissions for this motorcycle.
This calculation procedure is repeated for the other pollutants and fuel
economy, and the results are presented in Table 11 of the discussion
section.
The regression statistics for the horsepower test series are pre-
sented in Table A-3.
-------
A-3
Table A-3
Horsepower Regression Statistics
Motorcycle
Honda CB360
Regression
Parameter
r
r reqd. for
95% conf.
Significant
at 95% ?
Slope, a..
HC
.376
.75
No
.0308
CO
.850
.75
Yes
.7424
.942
.75
Yes
.0351
Economy
.959
.75
Yes
-.6931
The regression statistics show there is sufficient correlation at a 95
percent confidence level for all parameters except HC.
The regression equations are used in the same manner as described
for the inertia tests. The results are presented in Table IV of the
discussion section.
-------
Distribution List
Motorcycle Technical Support Reports for Regulatory Action
D. Alexander
E. Brune
T. Cackette
J.P. DeKany
C.L. Gray
D. Hardin
K. Hellman
W. Houtman
T. Huls
R. Jenkins
E. Rosenberg
R. Stahman
E.O. Stork
G. Thompson
M. Williams
------- |