Computational design of layered barrier system for vehicle impact attenuation

There has been significant work going on in the field of crash-worthiness of the vehicle and safety of the passengers occupying the vehicle. This includes the research in developing barrier systems for additional attenuation of impact. In this thesis, such a barrier system has been investigated. This barrier called SAFER was developed by the researchers at the Midwest Roadside Safety Facility at the University of Nebraska-Lincoln. SAFER stands for Steel and Foam Energy Reduction. This system has a steel guardrail with layered foam cushioning between the side-wall and guardrail.;The first stage of the project was to look at the system from a broader perspective to see how the acceleration of the impacting body is affected by different configurations of this system. Finite Element models were made on LS-DYNA for this investigation. Effect of different geometrical parameters and foam arrangements were investigated.;The next stage was to look into the design and optimization of this system using simplified models coded in Fortran. For that purpose, each foam layer was initially approximated by a mass-spring-damper system. Each of these are connected in series with other layers, hence we have the representation of the complete system. For this simplified system, springs of constant stiffness were used. However, under compression, the stress-strain response of polystyrene foams has three regimes: linear-elastic, plastic yielding and hardening. So, in order to use foam as the material in the model, a mathematical representation (constitutive model) for stress-strain behavior has been developed. This model incorporates the deformation response in all the three regimes as mentioned before and is based in part on a formulation for foams presented in the ABAQUS Theory Manual. In the constitutive model, some unknown parameters needed to be estimated. So, the next stage of the work was parameter estimation using experimental results. The algorithm of the constitutive model was started with an initial guess of the parameters. It was then fitted with the experimental data using a Marquardt algorithm to find the optimized value of parameters.;After the development of the constitutive model, the linear springs in the mass-spring-damper system were replaced by non-linear springs having the response from the constitutive model. Also, a mathematical representation of the guardrail was used for the final analysis. The final stage of the work focused on optimization of this system in order to minimize the accelerations that the impacting body undergoes after hitting the barrier. The three design aspects that have been looked into are the material density, frontal area of each foam layer and the thickness of each layer. A Genetic Algorithm (GA) was used for this purpose.