Dynamic Analysis and Shape Optimization of Automotive Gage Panel
International Journal of Mechanical Engineering and Automation 2014
Aleksandrs Januševskis, Anatolijs Meļņikovs, Jānis Januševskis

Design of mechanical products nowadays extensively utilizes simulation models which allow fast design procedure. FEM models are often used for optimization in metamodel based methods. In this work automotive gage panel (GP) is designed using FEM and metamodel based optimization. The complex geometrical model of the GP initial design is created by CAD software and simplified FEM model is used for required dynamic and static FEM calculations. The lowest 10 natural frequencies, distribution of stresses, strains and displacements under static loading of the GP, as well as the stationary and transient behaviors under deterministic and random excitations of the GP are investigated. In the next step these results are taken into account for optimization problem formulation. The optimum cross section shapes of the GP brackets are found using metamodel based shape optimization combined with effective parameterization. The employed parameterization is based on knot points of non-uniform rational B-splines allowing reduction of the number of design parameters. Calculation of the responses of the full FEM models is based on Latin Hypercube design of experiments. Quadratic locally weighted polynomial metamodels are used for fast optimization procedure. The bracket optimum design improves overall characteristics of the GP as it has higher fundamental frequency, improved strength and stiffness properties.


Atslēgas vārdi
gage panel, dynamic analysis, metamodel, shape optimization, finite element, design of experiments
Hipersaite
http://www.ethanpublishing.com/index.php?m=content&c=index&a=show&catid=159&id=209

Januševskis, A., Meļņikovs, A., Januševskis, J. Dynamic Analysis and Shape Optimization of Automotive Gage Panel. International Journal of Mechanical Engineering and Automation, 2014, Vol.1, No.4, 193.-202.lpp. ISSN 2333-9179. e-ISSN 2333-9187.

Publikācijas valoda
English (en)
RTU Zinātniskā bibliotēka.
E-pasts: uzzinas@rtu.lv; Tālr: +371 28399196