Drag Factor Tests on Grass Using ABS and Non-ABS Braking
Trooper Robert Frost, NYSP and Prof. John Kwasnoski

A series of tests were done on March 12, 2002 in the Town of Farmington, NY using a 2001 Ford Crown Victoria. Several types of grassy surfaces were tested, and the tests were done with the ABS activated, and again without the ABS. The results are as follows:

Mowed Grass - Approximately 2-3 inches High
Braking / Avg. f (Using VC2000)	Braking / Avg. f (Using VC2000)
ABS / .376 ABS / .367 ABS / .381 ABS / .371 ABS / .387 ABS / .379	Non-ABS / .376 Non-ABS / .370 Non-ABS / .384 Non-ABS / .332 Non-ABS / .328 Non-ABS / .340
Average of all tests = .377	Average of all tests = .355

The tests done on the longer, unmown grass show consistent results within each test scenario, and indicate that the non-ABS drag factor is higher than with the ABS. This may be because of greater rutting that occurs with the non-ABS braking action.

There seems to be little variation in the measured average drag factor for the two different types of grass in the ABS tests.

The tests on the shorter, mown grass with the non-ABS braking shows inconsistent results and cast doubt on the average value derived from all six tests. The differences within this category suggest the need for additional testing to resolve this scatter in the measurements. It may be that the first three tests and the second three tests reflected actual differences in the soil/grass surface.

These tests were undertaken as a starting point for information about the effective drag factor of a vehicle braking on grass as there seemed to be very little information in the published literature on drag factor values for this particular surface.

Tests Show Drag Factor Not Velocity-Dependent
Trooper Robert Frost, NY State Police, Sgt. Steve Sklenar, Wayne County Sheriff Dept., Prof. John Kwasnoski, Prof. Emeritus, Western New England College

Published drag factor charts may imply that the drag factor of a road surface is dependent upon the speed of the vehicle, and that an adjustment should therefore be made to lower speed tests or measurements when the result is applied to a high-speed vehicle motion. The authors have used the multiple-interval measurement capacity of a VC2000 accelerometer to study the following questions:

1. Is the average VC2000 drag factor value valid when skid tests are done at low vehicle speeds ?
   
2. Is the drag factor value significantly lower at high speeds ?
   
3. Is any adjustment needed to apply drag factor values to high-speed vehicle motion analysis ?

Tests were performed with a VC2000 instrumented vehicle braking in both the ABS and non-ABS modes. The tests were all done on the same road surface. The results of the field tests, done at various initial pre-braking speeds are shown below.

The tests lead to several conclusions. In every test the inclusion of measurements during the buildup of braking capacity affects the average drag factor value since values well below peak are being added to determine the average. When the number of measurements is fewer (at lower test speeds) the average drag factor clearly shows the mathematical result of averaging in several low drag factor measurements. The tests indicate that when drag factors are determined in relatively low speed tests the average value is an underestimate of the drag factor because of this mathematical averaging of the initial readings during the individual test.

Within any test the individual drag factor measurements made at 0.1 second intervals at the higher speeds are not lower than those at lower speeds. Data shows some scatter, which is probably a result of the measuring system itself, but there is no clear increase of the road's friction capacity as vehicle speed decreases during a test run.

There is no evidence in these tests that the average drag factor should be reduced when a test result is being applied to a higher speed vehicle motion. On the contrary, the low-speed test results are lower than those done at higher speeds, and therefore would underestimate vehicle speed calculations.

Anyone performing additional tests can communicate with
Prof. Kwasnoski by e-mail at: kwasnoski@aol.com

###

Return to June/July 2003 Newsletter