252
Y. He et al.
as material sensitive analysis, shape design and so on [2–5]. Sensitivity analysis of
automobile lightweight design was also a hot topic recently. Zuo et al. [6] made
stress analysis and optimization of rectangular tubes. And Chen et al. [7] proposed
a method of lightweight design under static and dynamic stiffness demands.
Material structure optimization is a vital approach for automobile lightweight
design. Topology optimization method [8] is widely used in cellular material. And
surrogate model has been an effective way to replace the material structure model in
FE software [9], which has been proved to reduce optimization time while guarantee
high accuracy.
Among such metamodeling methods, the Kriging method is one of the most
accurate and efficient methods even for highly nonlinear performance functions. The
Kriging method generates a surrogate model using a mean structure and a Gaussian
stochastic process, and the desired prediction value becomes different depending
on the type of the mean structure and correlation function, and parameters of the
stochastic process. Therefore, in this paper, Kriging surrogate model and NSGA-II
are employed to solve crashworthiness and lightweight optimization problem.
20.2 Finite Element Model
The FE model of mini-bus’s body-in-white (BIW) established in this study is
constructed for simulation of frontal impact test against deformable barrier with
40% overlapping in Hypermesh software. Much computational time is needed to
simulate a car collision since the entire vehicle model consisted of many parts. And a
reduced FE model is developed to decrease the computational time [10]. The B-pillar
front part of BIW is detailed constructed in FE model while other parts are replaced
by mass points, and the reduced FE model is shown in Fig. 20.1. Referring to the
regular in the C-NCAP [11], the front passenger seat can be replaced by 77 kg weight
Fig. 20.1 Reduced FE model of mini-bus’s BIW
Y. He et al.
as material sensitive analysis, shape design and so on [2–5]. Sensitivity analysis of
automobile lightweight design was also a hot topic recently. Zuo et al. [6] made
stress analysis and optimization of rectangular tubes. And Chen et al. [7] proposed
a method of lightweight design under static and dynamic stiffness demands.
Material structure optimization is a vital approach for automobile lightweight
design. Topology optimization method [8] is widely used in cellular material. And
surrogate model has been an effective way to replace the material structure model in
FE software [9], which has been proved to reduce optimization time while guarantee
high accuracy.
Among such metamodeling methods, the Kriging method is one of the most
accurate and efficient methods even for highly nonlinear performance functions. The
Kriging method generates a surrogate model using a mean structure and a Gaussian
stochastic process, and the desired prediction value becomes different depending
on the type of the mean structure and correlation function, and parameters of the
stochastic process. Therefore, in this paper, Kriging surrogate model and NSGA-II
are employed to solve crashworthiness and lightweight optimization problem.
20.2 Finite Element Model
The FE model of mini-bus’s body-in-white (BIW) established in this study is
constructed for simulation of frontal impact test against deformable barrier with
40% overlapping in Hypermesh software. Much computational time is needed to
simulate a car collision since the entire vehicle model consisted of many parts. And a
reduced FE model is developed to decrease the computational time [10]. The B-pillar
front part of BIW is detailed constructed in FE model while other parts are replaced
by mass points, and the reduced FE model is shown in Fig. 20.1. Referring to the
regular in the C-NCAP [11], the front passenger seat can be replaced by 77 kg weight
Fig. 20.1 Reduced FE model of mini-bus’s BIW