This guest post on Innovation Intelligence is written by Agata Abramowicz Sokoll, CEO of Impact Design Europe, developer of Crash Cad Calculate and Cross Section Creator. CCC and CSC are available through the Altair Partner Alliance.
Constantly growing crash safety requirements are presenting the need for development, testing and implementation of various strategies in vehicle body design. A novel approach can be introduced to vehicle structure design in the form of new materials and related geometrical specification.
Test procedures involve laboratory experiments as well as numerical simulations, and the process of structure development is time consuming and costly, therefore more effective procedures are necessary. To address this problem in the automotive industry, automatic crashworthiness optimization of separate members is now broadly recognized as a leading trend and necessity.
The possibility of studying the influence of various parameters on crushing response is a tempting perspective. Complex pre-processing tools are used, equipped with various optimization algorithms, and combined with finite element codes. Although they are providing adequate results, they are extremely demanding in terms of computation costs and calculation time. At early stages of product development, when numerous design variants are to be considered, that time cost is a problem. A simple and fast methodology would be the key to success.
The Macro Element Method (MEM) fits in optimization processes perfectly and is a great complement to the Finite Element Methods (FEM). The macro element approach is especially useful in the field of large deformations of thin walled members made from a variety of ductile isotropic materials and provides the highest efficiency at the early stages of vehicle development.
MEM is based on simplified modeling and the concept of super folding element at the cross sectional level, and super beam element in the case of a 3D structure, enable creation of simplified models for a variety of design concepts. Most significant for the optimization purpose is the possibility to easily modify a simplified structure (coordinates of points, plate thickness, assigned material). Calculation time, which does not exceed several seconds in the case of various cross sections, and several minutes for complex beam structures, guarantees numerous analyses of an optimization loop will be performed in a reasonable amount of time.
Any optimization “arrangement”, including input data modifier tool + solver + output collection + results visualization, can be combined with a MEM solver without great effort thanks to its transparent file format. The calculation results are comparable with the ones received in laboratory tests and FE calculations, and most importantly, they come with low computation cost.
Use Case Example: B-Pillar
A practical example of MEM is strength testing of a B-Pillar, where the bending response of the assembly is tested.
One of the greatest challenges faced by engineers is reducing the mass of the part while still satisfying all crash requirements. MEM can be implemented in such a design process on both 2D and 3D levels to help achieve this.
The first involves cross section bending analysis for various materials and geometry variants.
The simplified MEM model of a b-pillar cross section can be modified and tested for best crush performance, where the simplicity of the model enables quick and easy modification of the geometry. Each plate of a cross section can be defined individually which enables testing of various thicknesses and materials.
The time of calculation at the 2D level usually takes less than a fraction of second on standard PC. As a result, several blocks of results are available including bending response of the analyzed cross section. There is also the possibility of using HyperStudy and macro element batch mode processor for an automatic optimization routine.
On the 3D level of a structure, MEM allows definition of a simplified b-pillar beam model build of nodes and super beam elements. This provides the capability to test various geometrical and material specifications in many configurations, and the modeling procedure can be conducted semi-automatically using the import from FE tool. Initial boundary conditions are defined in the form of kinematic constraints assigned to the nodes. The presented B-Pillar structure consists of 11 cross sections, 14 Nodes and 13 Super Beam Elements.
In the case of the presented load case (350 ms of simulation time, 87,500 time steps) calculations take only 6 seconds.
Results are obtained in the form of visualization and charts available of each element in the solution.