Embankment Widening and Slopes for Energy Absorbing End Terminals

009408422003Final ReportPDFTech ReportTRP-03-126-02SPR-3(017)Highway safetyEnergy absorbing materialsEmbankment foundationsSlopesEmbankmentsGuardrail terminalsRollover crashesSpeedHighwaysEmbankment WideningHVOSMSafety and Human FactorsEnergy Absorbing End TerminalsI85: Safety Devices used in Transport InfrastructureUnited StatesNebraskaMidwest States Regional Pooled Fund ProgramPaulsen, Tony J.Sicking, Dean L.Holloway, James C.Midwest Roadside Safety FacilityMidwest States Regional Pooled Fund ProgramUniversity of Nebraska--LincolnNebraska. Department of RoadsUnited States. Federal Highway AdministrationUS Transportation CollectionFull-scale crash testing has shown that W-beam guardrails do not perform well when installed on even modest roadside slopes. In order to provide the maximum clear recovery area, guardrails are installed as far as possible from the traveled way. This practice causes guardrail terminals to be placed adjacent to roadside slopes and embankments. Vehicles that strike guardrail terminals are often allowed to pass through the systems at a relatively high speed. There is a concern that even moderate embankments placed near the end of a guardrail could cause these vehicles to rollover. Safety performance evaluation criteria contained in NCHRP Report 350 recommend that guardrail terminals be tested on flat ground. Hence, there is a great deal of uncertainty regarding acceptable terrain configurations near a guardrail terminal. One important finding from this study is that energy-absorbing terminals greatly reduce the velocity of small vehicles. Energy-absorbing terminals were found to reduce speeds by 75% during head-on impacts and approximately 50% when the terminal is struck at an angle of 15 degrees. These large reductions in vehicle velocity greatly reduce the risk of vehicle rollover as it moves onto roadside slopes and ditches. The HVOSM modeling effort predicted that energy absorbing terminals, when installed adjacent to roadside slopes of 4:1, would not be likely to cause rollovers. Further, the simulation effort did not predict rollover for nominal levels of friction, even when 3:1 slopes were placed behind the guardrail. The simulation study only predicted rollover when the tire ground friction was increased to artificially high levels to model wheel rutting in soft soil. Even under that circumstance, rollover was predicted for only one of the four crash tests included in the study.75