In this paper an effective successive multi-objective discrete optimization algorithm is developed to find the optimum design of Portable Concrete Barrier (PCB) systems. Multi-objective discrete optimization algorithm is attained by coupling Taguchis Orthogonal Arrays (OAs) with Grey Relational Analysis (GRA) method while nonlinear explicit dynamic Finite Element Analyses (FEA) are employed to generate response points and evaluate the performance of the PCB systems in accordance with National Cooperative Highway Research Program (NCHRP) Report No. 350 guidelines. GRA is used to convert multiple performance characteristics into a single performance criterion. Taguchis method is used to generate Design of Experiment (DOE) tables and prediction of the optimum design is accomplished by using Analysis of Means (ANOM) concept. Longitudinal and lateral ride down accelerations, the vehicle roll angle and dynamic barrier displacement are used as the design objectives to minimize while barrier safety shape, length, width, opening gap and hook distance in the connection are considered as discrete design variables.
Abstract