Reliability analysis has been advocated to account for the uncertainty in geometric design and to evaluate the risk associated with various design options. Most of the previous studies using reliability-analysis in highway design evaluated only one-mode of noncompliance. This study assesses the performance of horizontal curves using a system of multi-modal noncompliance (insufficient sight distance, vehicle skidding, and vehicle rollover). Five case studies of a highway in British Columbia are considered. Two approaches were used: (1) second-order reliability-bounds with FORM analysis (First-Order Reliability Method), (2) Monte-Carlo Simulation (MCS). A calibrated design chart that accommodates heavy-trucks on horizontal-curves with sharp-radii is provided. The results show that the differences in the system probability of noncompliance between one-mode and system of multi-modes of noncompliance are more pronounced for heavy-trucks. Results also show that the probability of noncompliance associated with vehicle rollover is significantly affected by the stability-ratio compared to height-ratio and roll-rate.

Additional Metadata
Keywords multi-modal reliability analysis, risk-based design, road safety, rollover, Surrogate safety measures
Persistent URL dx.doi.org/10.1080/23249935.2020.1720859
Journal Transportmetrica A: Transport Science
Citation
Alsaleh, R. (Rushdi), Sayed, T. (Tarek), Ismail, K, & AlRukaibi, F. (Fahad). (2020). System reliability as a surrogate measure of safety for horizontal curves: methodology and case studies. Transportmetrica A: Transport Science, 16(3), 957–986. doi:10.1080/23249935.2020.1720859