Optimization using sequential approach for triangular tube structure in crashworthiness



Journal Title

Journal ISSN

Volume Title


Texas Tech University


Crashworthiness technology strives to protect occupants by maintaining the structural integrity and by converting the kinetic energy into other forms; and at the same time, the lower crashing force the occupants are being undertaken, the more safety the occupants would obtain. This is a dissertation research that intends to design a component (front horn) of a vehicle’s bumper through optimizing the shape, dimension and parameters of a triangular tube structure in deterministic and probabilistic method in order to accomplish such purpose. The design has combined two new characteristics into the front horn: triangular tube structure and Reliability-Based Design Optimization (RBDO), using explicit finite element code Ls-Dyna as a simulation tool. The research is based on the fact that a triangular thin tube possesses the lowest crushing mean force, and the front horn is an important tubular structure for reducing injuries of front collision. RSM is used to identify the optimal shape of the triangular tube, applying the lowest Load Difference LD as objective function. The result shows that the equilateral triangular tube possesses the lowest LD value. The dimension optimization is carried out to determine the optimal dimensions of triangular tube under the same impacting condition, taking the load uniformity LU as the objective function and mean crushing force and absorbed energy as constraints based on Least Square method and nonlinear programming. The optimal dimension is 80 (mm) and 3 (mm) for lateral length and thickness respectively. Using the result and the formulation of dimension optimization, the tapered part of front horn is designed and optimized on the bases of RBDO, which assumes design variables with normal distribution. With addition of Monte Carlo method as another simulation tool, performance functions of reliability are generated in the same design domain. The final result is genetic shape and dimensions of a front horn possessing LU = 1.36, Fm = 52.47 (KN) with the reliability 0.94 of less than 55.04 (KN), and E = 9247 (J) with the reliability 1.0 larger than 8000 (J). Meanwhile a novel method of RBDO with a mapping technique for optimization is developed instead of using a safety index or performance objective, through determination of reliability domain.