Title: Multiobjective reliability-based and robust design optimisation for crashworthiness of a vehicle side impact

Authors: Sandeep Shetty; Larsgunnar Nilsson

Addresses: Volvo Car Corporation, Gothenburg SE-405 31, Sweden ' Division of Solid Mechanics, Linköping University, Linköping SE-581 83, Sweden

Abstract: Optimisation of vehicle design is necessary to meet increased safety requirements, new emission regulations, and to deal with competition in the global market, etc. However, optimised design using classical optimisation techniques with deterministic models might not meet the desired performance level or might fail in extreme events in real life owing to uncertainties in the design parameters and loading conditions. Consequently, it is essential to account for uncertainties in a systematic manner to generate a robust and reliable design. In this paper, an approach to perform multiobjective, reliability-based, and robust design optimisation is presented using a vehicle side impact crashworthiness application. Metamodels have been used in the optimisation process to decrease computational effort. Variations in material properties, thicknesses, loading conditions, and B-pillar heat-affected zone material strength have been considered for the stochastic optimisation. A comparative study of deterministic, reliability-based, and robust optimisation approaches is performed.

Keywords: multiobjective optimisation; robust optimisation; reliability based optimisation; crashworthiness; Monte Carlo analysis; metamodelling; boron steel; fracture risk factor; robust design; vehicle side impact; vehicle design; vehicle safety; vehicle crashes.

DOI: 10.1504/IJVD.2015.070410

International Journal of Vehicle Design, 2015 Vol.67 No.4, pp.347 - 367

Received: 11 Feb 2014
Accepted: 12 Jul 2014
Published online: 05 Jul 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article