Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
97 CHAPTER 6. VEHICLE TESTING ON SLOPE As part of this research, several full-scale vehicle traversability tests were performed on a 1V:3H slope. These tests were performed to evaluate the performance of the simulation vehicle models on slopes. The results were used to determine if changes should be made to the vehicle models. The testing was performed using the MASH small car and pickup truck vehicles. Three tests were performed for each of these vehicles (a total of six tests). Results of the simulations performed earlier were used to select the test conditions. In this chapter, details of the test matrix are presented along with additional analyses performed to evaluate the range of different test outcomes that could be observed due to differences in terrain properties. This is followed by the results of the traversability tests and their comparison to the simulation analyses. 6.1 TEST MATRIX DESIGN It should also be noted that for any terrain slope, including a flat terrain, a range of vehicle stability outcomes can occur in the real world, including rollovers. It is well known from crash data that rollovers occur even on flat and level terrain. This can also be seen from the simulation results presented in Chapter 5. In those simulations, a broad range of vehicle stability outcomes were observed for all foreslopes analyzed (see Figure 5.22). Thus it is important to note that a vehicle rollover in a test encroachment on a particular slope does not necessarily establish that the slope is non-traversable. Similarly, a test vehicle returning to the roadway after encroaching on a foreslope under a particular set of encroachment conditions does not necessarily establish that the slope is recoverable. While selecting the encroachment conditions for testing, it was desired to achieve various vehicle stability outcomes in the tests; more specifically, vehicle recovery or return to the roadway; spinout or side-slipping in which the vehicle cannot return to the roadway, but is able to traverse to the bottom of the slope without rolling over; and vehicle rollover. While the variation in test outcomes was desired, it was noted that this can be challenging to achieve due to differences in the terrain properties. As further discussed in Section 6.2, each encroachment condition can result in a range of different outcomes depending on the friction coefficient of the terrain. Thus, the full-scale encroachment testing primarily served to determine if the vehicle models sufficiently capture the vehicle dynamics over a sloped terrain within a range of outcomes, or if some changes should be made to these vehicle models. Using the results of the simulations performed earlier (Chapter 5), the researchers selected six (6) encroachment conditions for testing: three (3) for the pickup truck and three (3) for the small car. All of these encroachments were on 1V:3H foreslope. The results of the simulations were used to ensure that these encroachments could be successfully performed on a practically sized roadside ditch that was to be constructed for testing purposes. The full-scale test matrix and the test site cross-section is presented in Figure 6.1. The encroachment speeds varied from 35 mi/h to 55 mi/h and the encroachment angles varied between 5 and 10 degrees, as shown in the figure. Using the results of the simulations, it was determined that the overall length of the roadside ditch should be 400 ft to allow for sufficient evaluation of the vehicle dynamics for all of the selected encroachment conditions. Similarly,
98 the results of the simulations were used to guide selection of the width of the foreslope and the ditch bottom. Test Vehicle Speed (mi/h) Angle (degrees) Driver Input* 1 Pickup 55 5 3 2 Pickup 35 5 3 3 Pickup 35 5 2 4 Small Car 45 10 3 5 Small Car 55 5 3 6 Small Car 45 5 2 * Driver inputs are defined in Figure 5.21 Figure 6.1. Full-scale slope traversability test matrix and test site cross-section. 6.2 EXPECTED TEST RESULT VARIATIONS As discussed in Section 5.7, vehicle dynamics simulations of the encroachment events use a high maximum lateral friction coefficient between the tire and the terrain as a surrogate for incorporating soil furrowing forces. This high surrogate lateral friction coefficient applies higher surrogate soil forces on the tires when the vehicle is side slipping. The surrogate friction coefficient value was determined through a sensitivity study in which the lateral friction coefficient was varied and the probability of rollover in the simulations was compared to rollover probability estimates derived from crash data. Based on this sensitivity study, a maximum lateral friction coefficient of 2.0 was used in the friction ellipse model to represent the tire-terrain forces acting on the vehicle when soil furrowing is present. However, in the full-scale encroachment tests, the true friction coefficient between the tires and terrain is unknown and is likely to be less than the 2.0 value used as a surrogate for soil furrowing forces. The lateral coefficient of 2.0 is more representative of a terrain with soft or saturated soil than a dry grassy surface. In accident reconstruction analyses, a dry grassy roadside slope is typically represented with a friction coefficient of 0.5. The constructed ditch for vehicle testing did not have saturated wet soil. For this reason, vehicle kinematics in the full- scale encroachment tests were expected to be different from the simulations performed with a lateral friction coefficient of 2.0. To evaluate the effects of the variation of the terrain in the test ditch and the surrogate soil in the simulations, the researchers simulated the selected test configurations with a range of maximum lateral friction coefficients. The minimum value of the coefficient was selected to be
99 0.5, which is typical for grassy surface. The maximum value was selected to be 2.0, which is what was previously selected as a surrogate for soil furrowing conditions. In between these values, simulations were performed with friction coefficients of 1.0 and 1.5. The actual response of the vehicle for a specific encroachment condition was expected to be covered by the range of these maximum lateral friction coefficient values. Results of these additional simulations are presented next. Shown in Figure 6.2 are the sequential images of the results of simulations with the pickup truck model. The simulations results are overlaid. The only difference between the simulations is the value of maximum lateral friction coefficient (MuY), which varies as described previously (i.e. 2.0, 1.5, 1.0, and 0.5). The encroachment conditions are those associated with Test 1 in Figure 6.1. The vehicles start off at the same location on the roadway (Figure 6.2a). After encroaching on the foreslope, the vehicles start turning back to the roadway. As a result, the vehicles start to sideslip (Figure 6.2b). As the simulations progress, the differences in vehicle kinematics become more apparent. In Figure 6.2c, the vehicle with MuY of 2.0 starts developing high roll. The last sequential image in Figure 6.2d shows the final outcomes of the four simulations, with vehicle rollover occurring in one case (for MuY of 2.0) and the vehicle spinning out in the other three cases. In Figure 6.3, the final results of the same four simulations are shown along with the travel path of the vehicleâs CG and the roll angle of the vehicle for the four simulations. It can be seen that differences in the MuY result in differences in the stopping distances for the vehicles. In Figure 6.4, final results of the simulations for the pickup truck encroachment in Test 2 (defined in Figure 6.1) are shown. The path and vehicle roll angle are also shown for the different maximum lateral friction coefficients. In this case, while none of the vehicles roll over, the vehicles have various levels of vehicle rotation. Simulations with higher MuY resulted in the vehicle stopping sooner and yawing less compared to the simulations with lower MuY values. Similar to the pickup truck simulations, results of the simulations with the MASH small passenger car are shown in Figures 6.5 and 6.6, which show simulated results for encroachment conditions of Test 3 and 4, respectively. The vehicles in the simulations exhibit a range of kinematic behavior depending on the friction of the simulated terrain.
100 (a) Vehicle simulations start. Results of four vehicle simulations are overlaid. (b) Vehicles are starting to turn back to the roadway after encroaching the slope. The vehicles are starting to sideslip. (c) Simulations start differing as the vehicles sideslip further. The vehicle starts to rollover in the simulation with 2.0 lateral friction coefficient. (d) Results of all four simulations. Rollover occurred in simulation with maximum lateral coefficient (MuY) of 2.0. For each vehicle, the MuY value used in simulation is shown. Figure 6.2. Variation in vehicle kinematics due to different terrain friction. Four simulations are shown overlaid with same vehicle model but varying maximum lateral friction coefficient.
101 Figure 6.3. Vehicle kinematics of pickup trucks encroaching at 55 mi/h and 5-degree angle with different terrain coefficients.
102 Figure 6.4. Vehicle kinematics of pickup trucks encroaching at 35 mi/h and 5-degree angle with different terrain coefficients.
103 Figure 6.5. Vehicle kinematics of small cars encroaching at 45 mi/h and 10-degree angle with different terrain coefficients.
104 Figure 6.6. Vehicle kinematics of small cars encroaching at 55 mi/h and 5-degree angle with different terrain coefficients.
105 In summary, the results of the simulations show a range of expected outcomes in the full scale encroachment tests for the different terrain frictions modeled. This is not much different from the real world where the kinematics of a vehicle will greatly depend on the condition of the terrain the vehicle is traversing. A wet soft soil condition, in which the vehicle tires are likely to dig in as the vehicle sideslips, is more likely to cause a soil tripped rollover. Hence we observed in the simulations that when a higher lateral friction coefficient was used, the vehicle rollover is more likely. Similarly, in the real world, a dry terrain with some vegetation is less likely to trip a vehicle due to soil furrowing. In this case the vehicle is more likely to continue side slipping and may spin out without rolling over. Hence, with lower maximum lateral friction coefficients, the simulation results show the vehicles travelling farther and spinning out compared to the higher friction values. Keeping in mind that the actual terrain of the test site had a road base top surface, the opportunity for a soil-tripped rollover is low. The results of the simulation were, therefore, used to evaluate the vehicle models and determine if any further modifications to the vehicle models is considered necessary. This process involved comparing results of the tests to a range of simulations for each encroachment case. 6.3 VEHICLE TRAVERSAL TESTS To perform the vehicle traversal tests, a ditch was constructed with a 1V:3H foreslope as previously described in Figure 6.1. The cross section of the ditch and its photograph are shown in Figure 6.7. The foreslope was 32 ft wide. The bottom of the ditch was flat and had a width of 12 ft. The backslope had a 1V:1H slope and was not intended to be engaged in the vehicle traversal tests. The foreslope and the ditch bottom were lined with geogrid to retain the roadbase. A 6-inch layer of crushed limestone road base was then compacted on the geogrid to construct a compacted surface. The backslope of the ditch was left non-compacted with native soil.
106 (a) Cross section of the test ditch (b) Photo of the testing ditch Figure 6.7. Design and photo of the ditch with 1V:3H foreslope for vehicle traversal testing. 6.3.1 Testing Procedure The test vehicles were equipped with servo motors that controlled the steering input to the vehicle. A hydraulic cylinder, which could be electronically activated, was used to apply the braking input to the test vehicles. The vehicles were reverse-towed onto the slope for each of the encroachment tests. The tow and guidance system was disengaged prior to the vehicle entering the slope. TTI developed an on-board micro controller which was programmed to control the steering servo motor and the hydraulic cylinder for applying the brakes. The micro controller applied the steering and braking inputs using the same perception reaction times and steering rates used in the simulations. Figure 6.8 shows the inside of the test vehicles with the driver input control hardware installed.
107 Figure 6.8. Steering and braking control installed on small car (left) and pickup (right). As the vehicle approached the time-zero marker, a signal was sent to the micro-computer to execute the pre-programmed steering and braking inputs. Prior to performing the full-scale tests, the researchers performed several test runs on a flat concrete surface to ensure proper functionality and robustness of the driver input control hardware. As in the computer simulations, a perception-reaction (PR) time of 1 second was used prior to applying any steering or braking input. The one-second perception-reaction time window began once the vehicleâs leading bumper corner was a shoulder width away from the top edge of the ditch foreslope. 6.3.2 Test Vehicles The first three tests were performed with a MASH 2270P pickup truck and the last three were performed with a MASH 1100C small car. The MASH 2270P pickup was a 2008 Chevy Silverado, which was reused for the second and third tests. The test inertial weight of the vehicle was 5,068 lb. Other vehicle properties and dimensions of the test vehicle are presented in Figure 6.9. The MASH 1100C small car was a 2009 Kia Rio, which was also reused for the second and third small car tests. The test inertial weight of this vehicle was 2,577 lb. Other vehicle dimensions and properties are presented in Figure 6.9.
108 Date: 2015-07-26 Test No.: 479310-1-1 thru 3 VIN No.: 3GCEC13J58G211693 Year: 2008 Make: Chevrolet Model: C1500 Tire Size: P275/55R20 Tire Inflation Pressure: 35 psi Tread Type: Highway Odometer: 176414 Note any damage to the vehicle prior to test: None Geometry: inches A 80.0 F 39.4 K 20.2 P 3.0 U 29.0 B 72.5 G 28.5 L 29.8 Q 31.0 V 26.0 C 230.1 H 57.0 M 68.5 R 21.5 W 56.9 D 47.2 I 12.5 N 67.5 S 16.0 X 82.5 E 46.5 J 27.2 O 45.5 T 45.5 Wheel Center Height Front 14.75 Wheel Well Clearance (Front) 5.25 Bottom Frame Height - Front 17.125 Wheel Center Height Rear 14.75 Wheel Well Clearance (Rear) 12.00 Bottom Frame Height - Rear 25.125 RANGE LIMIT: A=78 Â±2 inches; C=237 Â±13 inches; E=148 Â±12 inches; F=39 Â±3 inches; G = > 28 inches; H = 63 Â±4 inches; O=43 Â±4 inches; M+N/2=67 Â±1.5 inches (Allowable Range for TIM and GSM = 5000 lb Â±110 lb) Mass Distribution: lb LF: 1551 RF: 1505 LR: 1013 RR: 999 Figure 6.9. Pickup truck test vehicle properties. ï· Denotes accelerometer location. NOTES: None Engine Type: V-8 Engine CID: 5.3 liter Transmission Type: x Auto or Manual FWD x RWD 4WD Optional Equipment: None Dummy Data: Type: None Mass: NA Seat Position: NA GVWR Ratings: Mass: lb Curb Test Inertial Gross Static Front 3650 Mfront 3033 3056 Back 6950 Mrear 2150 2012 Total 6800 MTotal 5183 5068
109 Date: 2015-07-20 Test No.: 479340-1-4 thru 6 VIN No.: KNADE223096524486 Year: 2009 Make: Kia Model: Rio Tire Inflation Pressure: 32 psi Odometer: 110248 Tire Size: Describe any damage to the vehicle prior to test: None Geometry: inches A 66.38 F 33.00 K 12.50 P 4.12 U 14.00 B 58.00 G ------ L 25.00 Q 27.19 V 21.00 C 165.75 H 61.62 M 57.75 R 15.38 W 44.00 D 34.00 I 8.50 N 57.12 S 7.75 X 108.50 E 98.75 J 21.50 O 31.50 T 66.12 Wheel Center Ht Front 11.00 Wheel Center Ht Rear 11.00 Allowable TIM = 2420 lb Â±55 lb | Allowable GSM = 2585 lb Â± 55 lb Mass Distribution: lb LF: 858 RF: 761 LR: 456 RR: 502 Figure 6.10. Small car test vehicle properties. ï· Denotes accelerometer location. NOTES: None Engine Type: 4 cylinder Engine CID: 1.6 liter Transmission Type: x Auto or Manual x FWD RWD 4WD Optional Equipment: None Dummy Data: Type: None Mass: NA Seat Position: NA GVWR Ratings: Mass: lb Curb Test Inertial Gross Static Front 1918 Mfront 1584 1619 Back 1874 Mrear 864 958 Total 3638 MTotal 2448 2577
110 6.3.3 Test Results The target and the actual encroachment speeds and angles are listed in Figure 6.11. Also listed are the shoulder offset and the steering and braking combinations used for each test, which are based on the previously presented test matrix (Figure 6.1). Sequential images of the vehicle traversals are shown in Figure 6.12 through 6.17. Test Number Vehicle Speed (mi/h) Angle (degrees) Shoulder Offset (ft.) Steering Applied Brake Applied Target Actual Target Actual 479340-1-1 Pickup 35 34.83 5 5.31 2 Yes No 479340-1-2 Pickup 45 45.08 5 4.99 2 Yes Yes 479340-1-3 Pickup 55 55.47 5 5.53 6 Yes Yes 479340-1-4 Small Car 45 45.13 5 5.65 6 Yes No 479340-1-5 Small Car 55 55.8 5 - 6 Yes Yes 479340-1-6 Small Car 45 45.48 10 10.63 8 Yes Yes Figure 6.11. Test target and actual encroachment conditions. In test 479340-1-1, the pickup truck encroached onto the 2-ft shoulder at a speed and angle of 34.83 mi/h and 5.31 degrees, respectively. The right-front tire of the pickup truck entered the dich at 0.460 seconds. Steering was applied at 1 second. No brakes were applied in this test. As the steering developed, the vehicle started to recover toward the shoulder. However, it could not fully recover and subsequently began to spin out. The vehicle yawed 129.9 degrees. The front tires of the vehicle stopped outside of the ditch, while the rear tires were still on the foreslope. Since the brakes were not applied in this test, the vehicle then started rolling backwards into the ditch and reached the ditch bottom. The traversal of the vehicle in this test is shown in Figure 6.12. In test 479340-1-2, the pickup truck encroached onto the 2-ft shoulder at a speed and angle of 45.08 mi/h and 4.99 degrees, respectively. The right-front tire of the pickup truck entered the dich at 0.38 seconds, and the left-rear tire entered the ditch at 1.38 seconds. Steering and braking input were applied at 1 second. As the steering developed, the vehicle started yawing toward the shoulder and then slid sideways on the foreslope. Maximum vehicle yaw was 73.5 degrees when the vehicle came to a stop on the 1V:3H foreslope. The traversal of the vehicle in this test is shown in Figure 6.13. In test 479340-1-3, the pickup truck encroached onto the 6-ft shoulder offset at a speed and angle of 55.47 mi/h and 5.53 degrees, respectively. The right-front tire of the pickup truck entered the dich at 0.906 seconds, whereas the left-rear tire entered the ditch at 1.662 seconds. Steering and braking input was applied at 1 second. As the steering developed, the vehicle started yawing toward the shoulder as it slid mostly sideways on the foreslope. The maximum vehicle yaw was 101.6 degrees when the vehicle came to a stop on the 1V:3H foreslope. The traversal of the vehicle in this test is shown in Figure 6.14. In test 479340-1-4, the small passenger car encroached onto the 6-ft shoulder offset at a speed and angle of 45.13 mi/h and 5.65 degrees, respectively. The right-front tire of the vehicle entered the dich at 0.9 seconds. Steering was applied at 1 second. No brakes were applied in this test. As the steering developed, the vehicle started yawing toward the shoulder. After the vehicle was on the foreslope, it started to recover due to the steering, but eventually started spinning out.
111 The vehicle then side slipped with the front two tires on the shoulder offset area outside the ditch and the rear two tires on the foreslope. The vehicle eventually yawed 144.7 degrees and had zero heading velocity. Since no brakes were applied, the vehicle then started moving backwards on the foreslope and reached the bottom of the ditch. The traversal of the vehicle in this test is shown in Figure 6.15. In test 479340-1-5, the small passenger car encroached onto the 6-ft shoulder at a speed of 55.8 mi/h. The high-speed overhead camera did not trigger and the precise encroachment angle of the vehicle could not be determined from the film analysis. The right-front tire of the pickup truck entered the dich at 0.714 seconds, and the left-rear tire entered the ditch at 1.470 seconds. Steering and braking input were applied at 1 second. As the steering developed, the vehicle started yawing toward the shoulder and slid mostly sideways on the foreslope. The maximum vehicle yaw was 117.1 degrees when the vehicle came to a stop on the 1V:3H foreslope. The traversal of the vehicle in this test is shown in Figure 6.16. In test 479340-1-6, the small passenger car encroached onto the 8-ft shoulder offset at a speed and angle of 45.48 mi/h and 10.63 degrees, respectively. The right-front tire of the vehicle entered the dich at 0.648 seconds, and the left-rear tire entered the ditch at 1.144 seconds. Steering and braking input were applied at 1 second. As the steering developed, the vehicle started side slipping on the foreslope and reached the bottom of the ditch in this manner. The rear right tire of the vehicle reached the ditch bottom first, followed by the front right tire. The vehicle came to a stop with the left front and rear tires on the foreslope. The maximum vehicle yaw was 61.0 degrees toward the shoulder at 3.404 seconds, which is approximately the time the rear right tire reached the bottom of the ditch. After this, the vehicle started yawing in the opposite direction before coming to rest with a final yaw of 14 degrees toward the shoulder. The traversal of the vehicle in this test is shown in Figure 6.17.
112 0.000 s 1.400 s 0.350 s 1.750 s 0.700 s 1.823 s 1.050 s 2.100 s Figure 6.12. Sequential photographs for Test No. 479340-1-1 (rear view).
113 2.450 s 3.065 s 2.458 s Figure 6.12. Sequential photographs for Test No. 479340-1-1 (rear view) (continued).
114 0.000 s 1.400 s 0.350 s 1.750 s 0.700 s 2.100 s 1.050 s 2.450 s Figure 6.13. Sequential photographs for Test No. 479340-1-2 (rear view).
115 2.800 s 4.200 s 3.150 s 4.550 s 3.500 s 4.900 s 3.850 s 5.250 s Figure 6.13. Sequential photographs for Test No. 479340-1-2 (rear view) (continued).
116 0.000 s 1.400 s 0.350 s 1.750 s 0.700 s 2.100 s 1.050 s 2.450 s Figure 6.14. Sequential photographs for Test No. 479340-1-3 (rear view).
117 2.800 s 4.200 s 3.150 s 4.550 s 3.500 s 4.900 s 3.850 s 5.250 s Figure 6.14. Sequential photographs for Test No. 479340-1-3 (rear view) (continued).
118 0.000 s 1.400 s 0.350 s 1.750 s 0.700 s 2.100 s 1.050 s 2.450 s Figure 6.15. Sequential photographs for Test No. 479340-1-4 (rear view).
119 2.800 s 4.200 s 3.150 s 4.550 s 3.500 s 4.900 s 3.850 s Figure 6.15. Sequential photographs for Test No. 479340-1-4 (rear view) (continued).
120 0.000 s 1.400 s 0.350 s 1.750 s 0.700 s 2.100 s 1.050 s 2.450 s Figure 6.16. Sequential photographs for Test No. 479340-1-5 (rear view).
121 2.800 s 3.850 s 3.150 s 4.200 s 3.500 s Figure 6.16. Sequential photographs for Test No. 479340-1-5 (rear view) (continued).
122 0.000 s 1.400 s 0.350 s 1.750 s 0.700 s 2.100 s 1.050 s 2.450 s Figure 6.17. Sequential photographs for Test No. 479340-1-6 (rear view).
123 2.800 s 3.850 s 3.150 s 4.200 s 3.500 s Figure 6.17. Sequential photographs for Test No. 479340-1-6 (rear view) (continued).
124 6.3.4 Terrain Deterioration and Its Effects The testing phase of this project was delayed due to an extended rainy season. After the construction of the test ditch, the full-scale tests could not be run for several months due to the extended rainfall. After each rain, finer soil on the shoulder and foreslope of the ditch washed down and accumulated at bottom of the ditch. The research team pumped out the accumulated water and excavated the silt after the rain, however, prior to being able to run the tests, more rainfall occurred. This cycle repeated three times before the tests could be run. As discussed previously in Section 6.2, it was expected that the surface of the test ditch would have a lower lateral friction coefficient compared to a saturated soil condition that often leads to soil tripped rollover due to soil furrowing. For this reason, additional simulations were performed prior to running the full-scale tests to evaluate the range of test outcomes that could be expected. However, due to the loss of finer soils on the ditch surface, there was a high ratio of loose aggregate present, which further influenced the vehicle traversal tests. Due to the high ratio of loose aggregate on the foreslope, the ABS brakes of the vehicles were not very effective. ABS brakes are designed to release if the wheel locks and the tire slips in the longitudinal direction. Once the wheel unlocks, the brakes are reapplied, but the cycle repeats if the wheels continue to slip and lock on reapplying the brakes. Videos of the ditch testing showed that since the tires could easily slip on the loose gravel when the brakes were applied, the ABS brakes were released and reapplied in a repeated manner. This reduced the effectiveness of the braking input. 6.4 TEST AND SIMULATION COMPARISON After completing the vehicle traversal tests, the researchers performed the simulations of the six tests using the actual test encroachment speeds and angles listed in Figure 6.11. The simulations were performed using the previously developed vehicle dynamics models of the MASH small car and pickup up truck vehicles. The objective was to compare the results of the tests and the simulations. Presented next are the comparisons of the simulations and the tests, showing that a reasonable correlation was achieved and that the vehicle dynamics models are capable of capturing vehicle kinematics on slopes. As described in Section 6.2, prior to performing the full-scale tests, simulations were performed using the target encroachment speed and angle for each test. A range of maximum lateral friction coefficient (from 0.5 to 2.0) was used in these simulations. This was done because the test ditch was expected to have a lower lateral friction coefficient than the soft soil that typically results in vehicle rollover due to soil buildup at the tire-terrain interface as the vehicle slides laterally (i.e. soil furrowing). In the full-scale tests, there was no meaningful soil buildup at the tire-terrain interface as the vehicles slid laterally. Furthermore, the presence of loose gravel affected the terrain friction and braking capacities as described previously. To account for these factors, the researchers performed the simulations with a maximum lateral friction coefficient of 0.55. This coefficient is slightly higher than the 0.5 friction coefficient typically used for a grassy dry surface. However, it is still representative of a surface that has low friction without enough soil buildup â which was the case in the ditch traversal tests.
125 A total of six new simulations were performed, one for each ditch traversal test, using the actual encroachment speeds and angles measured in the tests. To compare the results of the testing and simulation, the position and orientation of the vehicles during the ditch traversal were compared. These comparisons are shown in Figure 6.18 through Figure 6.23. Additionally, the path of the vehicle in the horizontal plane was also compared for the test and simulation vehicles. This path was recorded from the test vehicles using a high-speed GPS device. It was also calculated for each of the simulations. The simulation and test vehicle paths during the slope traversal are compared in Figure 6.24 for the three tests with the pickup truck, and in Figure 6.25 for the three tests the small passenger car. Due to the changes in the terrain features described earlier, which affected the terrain friction and the effectiveness of the ABS braking, some differences between the simulation and test results were expected. Even with these differences, the results of the simulations compared reasonably well with the test results. Some of the differences observed are discussed next. In Tests 479340-1-1 and 479340-1-4, which were with the pickup truck and the small car, the stopping distances of both simulation and test were very similar as can been seen in Figure 6.24a and Figure 6.25a, respectively. The simulations were stopped once the vehicleâs CG returned to the edge of the roadway since this was set as one of the simulation stopping criteria. In the tests, however, there was more vehicle yaw and the vehicles partially returned to the road, as can be seen in Figure 6.18 and Figure 6.21 for the pickup truck and the small car, respectively. The comparison of the simulation and test results also showed that the stopping distances and cornering forces were slightly different. The simulation model uses friction on a smooth surface to determine the lateral and longitudinal tire forces, whereas in the test there was a high ratio of loose aggregate. The performance of the ABS braking was also, therefore, different in the tests and simulations. Both of these factors contributed to some of the differences. However, even with some of the differences described above, the overall results of the simulations matched reasonably well with the testing. The pickup truck model spun out and slid laterally on the foreslope before reaching the bottom, similar to the behavior observed in Test 479340-1-2 and Test 479340-1-3 (see Figure 6.19 and Figure 6.20, respectively). Similarly, the small car model spun out while sliding laterally on the foreslope, and eventually reached the bottom of the ditch. Similar behavior was observed in Test 479340-5 and Test 479340-1-6 (see Figure 6.22 and Figure 6.23, respectively). In conclusion, the traversability testing and the subsequent simulation comparisons in this project demonstrate the ability of the vehicle dynamics models to simulate vehicle encroachments on slopes with a reasonable degree of accuracy. The vehicle dynamics models used in this project allowed the research team to evaluate 43,200 unique cases of vehicle encroachments on various slopes. The prime advantage of using vehicle dynamics models was the ability to perform a large number of simulations in a manageable time period and limited resources. This would not have been possible using other simulation techniques, such as finite element analysis, that require significantly larger computational time and resources.
126 1.050 s 1.750 s 2.100 s 3.065 s Figure 6.18. Sequential photographs for Test No. 479340-1-1 (rear view).
127 1.050 s 2.450 s 3.500 s 5.250 s Figure 6.19. Sequential photographs for Test No. 479340-1-2 (rear view).
128 1.400 s 2.450 s 3.500 s 4.900 s Figure 6.20. Sequential photographs for Test No. 479340-1-3 (rear view).
129 0.700 s 1.750 s 2.800 s 3.500 s Figure 6.21. Sequential photographs for Test No. 479340-1-4 (rear view).
130 0.700 s 1.750 s 2.800 s 4.200 s Figure 6.22. Sequential photographs for Test No. 479340-1-5 (rear view).
131 0.700 s 1.400 s 2.450 s 4.200 s Figure 6.23. Sequential photographs for Test No. 479340-1-6 (rear view).
132 (a) (b) (c) Figure 6.24. Simulation versus test comparison of vehicle path for tests with pickup. Test Simulation
133 (a) (b) (c) Figure 6.25. Simulation versus test comparison of vehicle path for tests with small car. Test Simulation