Search Results

A passenger airbag is an important part of a vehicle restraint system which provides supplemental protection to an occupant in a crash event. New Federal Motor Vehicle Safety Standards No. 208 requires considering multiple crash scenarios at different speeds with various sizes of occupants both belted and unbelted. The increased complexity of the new requirements makes the selection of an optimal airbag shape a new challenge. The aim of this research is to present an automated optimization framework to facilitate the airbag shape design process by integrating advanced tools and technologies, including system integration, numerical optimization, robust assessment, and occupant simulation. A real-world frontal impact application is used to demonstrate the methodology.

In vehicle safety engineering, it is important to determine the severity of occupant injury during a crash. Computer simulations are widely used to study how occupants move in a crash, what they collide during the crash and thus how they are injured. The vehicle motion is typically defined for the occupant simulation by specifying a crash pulse. Many computer models used to analyze occupant kinematics do not calculate both vehicle motion and occupant motion at the same time. This paper presents a framework of response surface methodology for the crash pulse prediction and vehicle structure design optimization. The process is composed of running simulation at DOE sampling data points, generating surrogate models (response surface models), performing sensitivity analysis and structure design optimization for time history data (e.g., crash pulse).

This paper is on the application of reliability-based design optimization of automotive structures. Several foundmental approaches aimed at simulations, optimization and robustness analysis are discussed. They include: Sampling techniques and nonlinear response surface methodologies; Optimization and robustness assessment. Basic approaches of reliability-based design optimization such as Hasofer-Lind method, single loop single vector method and mean value first order reliability method; Robust design formulation. The major focus is on the implementation of reliability-based design optimization methodologies to vehicle crash safety design. Some recent applications are presented to demostrate how these approaches can be used for vehicle structure design.