This is most likely caused by snow falling back into
Table 6. Analysis of undercarriage drag, tests
the rut, which is common for deep snow covers
0125h and 0125i.
(Fig. 12) and may have happened in this test.
Unfortunately, this effect was not identified early
Test 0125h
Test 0125i
enough during the testing to be fully investigated.
These data are shown graphically in Figure 13. The
Total average vertical force
25,371 N
25,280 N
data of test 0126 seem to agree well with the theory
Longitudinal forcesa:
that succeeding tires should have decreasing resis-
tance, the average values for the third and fourth
Average left front
454.6
492.7
"axles" being 0.0094 and 0.00685 respectively.
Average right front
634.0
771.0
The implication of this analysis is that, for ve-
Average left rear
368.3
368.3
Average right rear
347.8
456.4
hicles with more than two axles such as the HEMTT
and LAV, a rationale needs to be determined for
Hard surface resistanceb
507.4
505.6
modeling these additional trailing tires. It does not
appear that individual wheel resistance on large
134.9c
Net force (available) w/o gravity
66.6
vehicles will be directly measurable in the near
Average force due to gravityd
869
837
future.
Net force available with gravitye
734
770.4
SHALLOW SNOW RESISTANCE MODEL
aFrom
Table 3.
bEstimated using an average coefficient of 0.02.
cA negative value infers that the vehicle would be immobilized
As mention earlier, Richmond et al. (1990) ana-
(no tractive reserve).
lyzed resistance data from a large number of tests
dThe CIV tilt sensor measured an average angle (φ) of 1.96 and
and vehicles, obtaining eq 4 from data for both
1.89, indicating that the CIV was moving downhill. The
wheeled and tracked vehicles. Richmond (1990)
force due to gravity was calculated by multiplying the total
average vertical force by sin φ.
combined the data into groups of like vehicles and
eSince the vehicle was assumed to be traveling at a steady
developed a set of equations of the same form as eq
speed, this value must be attributed to undercarriage drag.
4, but no significant improvement in the correla-
tion coefficients resulted. In these earlier analyses,
only the data from the CIV could be considered as
attributable to the four wheels and the average
leading wheel data. The data obtained from the
undercarriage resistance results in an undercar-
other vehicles were "whole" vehicle resistances
riage resistance that is 22% of the total vehicle
resistance or 1.14 times Rs. This is interesting in
(from snow) values. To try and reduce the data to
that Richmond et al. (in prep) use a multiplier of
fit one equation, the whole vehicle resistance was
1.25 to estimate undercarriage drag for this condi-
simply divided by the number of wheels for each
tion.
vehicle.
Appendix B contains a discussion of the defor-
mation of snow by a wheel and concludes along
Multiple passes
Data sets 6 and 8 were obtained by towing the
with eq 6 that wheel width may not be as an
CIV such that the wheels traveled through the ruts
important parameter as arc length and initial den-
created by the tow vehicle. The forces measured
sity. Additionally, the ratio of wheel radius to
may be assumed to be estimates of third and
sinkage should be considered to account for in-
fourth trailing tires. It was further assumed that
creased deformation (more bulldozing) in deeper
snow covers, possibly in the form (4a/rπ) noted
tests 0126g0126i could represent the first and
second wheels; together, these data represent a
earlier. With these ideas in mind, and with the new
complete set of resistance data for axles 1 through
trailing tire information, an attempt was made to
4 (Table 7). Tests 0202f0202h were analyzed simi-
improve the currently used algorithm (eq 4 and 1)
larly.
for predicting motion resistance for wheeled ve-
Comparing the average resistance coefficients
hicles in snow.
between the two data sets shows that the resis-
After several attempts were made to extend the
tance data measured during the 0202 tests (snow
CIV trailing tire data to the HMMWV data, with-
depth 15.8 cm with a density of 250 kg/m3) are
much higher than the data from the 0126 tests
base obtained during the Wheels vs. Tracks shal-
(snow depth 12.6 cm with a density of 110 kg/m3).
low snow tests was reexamined. Appendix C con-
13