In the full scale impact test the depth of the ditch was 1.0 meter. The front slope was built to grade of 1:3 and the back slope to the grade of 1:2. The height of the back slope was 2.0 meters to avoid unnecessary large damage to the vehicle. There was also a need to get information if it is possible to avoid the vehicles hitting, for example, rock cuts by high back slopes. The soil was hard clay. The slopes were very drivable. The wheels left only minor marks to the soil. The main interests were (1) if the vehicle crosses the ditch; (2) the severity of the impact compared to the road barrier; (3) if there is significant rolling, pitching or yawing; and (4) what if the barrier is placed further from the road border. The impact speed was 80 or 100 km/h and the angle was 5, 10 or 20 degrees. Some of the vehicles were equipped with the steering. First connecting a weight to the side of the steering wheel made the steering effect. The vehicle hardly reached the ditch and on the following tests the weight replaced with radio control devices. With the impact angle 5 degrees the vehicles run smoothly. Three tests were made with impact speed about 80 km/h and the vehicles climbed up to the edge of the back slope but stayed inside the ditch. Only one test was made with impact speed 100 km/h. At this time the vehicle again rose up the back slope but then turned back to the road across the ditch. Without the high back slope all the vehicles except one would cross the ditch. The impact angle was lower in that exception and the vehicle travelled along the ditch bottom. All the measured accelerations of the vehicles were very low ¡ there was no clear impact during the test. With the impact angle 10 degrees the wheels of the vehicle did not yet loose the contact with the ground. The vehicles, except those with steering effect, run over the ditch in a relative straight line. Two attempts were made with the remote steering. First the vehicle tried to drive back to the road. The steering manoeuvre was made too late and was over reacted. The vehicle rolled over. Second time the steering was more moderate and the vehicle continued smoothly along the ditch bottom to the end of the ditch. One test conducted with concrete barrier installed in the back slope to the height of 0.7 m from the bottom of the ditch. The barrier redirected the vehicle but the vehicle rolled over. At a 20 degree approach angle the vehicles severely hit the back slope with the right front corner. The vehicles flew over the front slope and tires contacted the ground at the bottom of the ditch about the same time as they hit the back slope. This resulted in two out of three vehicles rolling over. The one that remained upright slid sideways along the back slope after the impact. Two tests were made at the impact speed 80 km/h and one at 100 km/h. The vehicles at higher impact speed exceeded the ditch and those at lower speed remained in the ditch. Of course the number of tests is small, but it is not easy to turn a vehicle over while it is on the track ¡ no sliding before the impact. It is managed only by high approaching angle or by giving the vehicle "driver response". One program is used for modelling impacts as far (PC-crash). The program gives reasonable results as far as the vehicle is on the track. The models response to the declination of the ditch is not yet as quick as in real life. And if it is difficult to overturn the vehicle in full scale crash tests it is more difficult to overturn the vehicle at simulation ¡ it has not succeeded yet with reasonable parameters. (A) For the covering abstract of the conference see ITRD no 207828. The reprints are also available at the web - http://www.vti.se/pdf/reports/K18APart1.pdf; http://www.vti.se/pdf/reports/K18APart2.pdf and http://www.vti.se/pdf/reports/K18APart3.pdf.
Abstract