This presentation was made at the NAFEMS European Conference on Simulation-Based Optimisation held on the 15th of October in London.
Optimisation has become a key ingredient in many engineering disciplines and has experienced rapid growth in recent years due to innovations in optimisation algorithms and techniques, coupled with developments in computer hardware and software capabilities. The growing popularity of optimisation in engineering applications is driven by ever-increasing competition pressure, where optimised products and processes can offer improved performance and cost-effectiveness which would not be possible using traditional design approaches. However, there are still many hurdles to be overcome before optimisation is used routinely for engineering applications.
The NAFEMS European Conference on Simulation-Based Optimisation brings together practitioners and academics from all relevant disciplines to share their knowledge and experience, and discuss problems and challenges, in order to facilitate further improvements in optimisation techniques.
The perspective on local hardening effect coming with laser-weld is changing. Was it an unasked-for feature in the past, it is developing as a useful mechanism to adjust part proper-ties locally. This approach can help to meet part requirements even late in the development process, without the costly change of global properties. Research shows that it is possible to improve formability and crash performance. Comparable options addressing local or graded properties are CFK patches, local strain hardening, rolling with variable sheet thickness or tailored blanks. The application of local hardening by blind laser-weld seams is restricted by few technological constraints and selection of applicable material only. The consequent open design space introduce the question of optimal pattern of laser-weld seams. Due to the specic characteristics of the optimisation problem typical approaches like topology or parameter optimisation can not be applied.
The papers presents and compares three strategies to derive optimal laser-weld patterns for a plate with hole in the first step. Aim of the strategies is to find weld pattern with at less hardened elements to meet part requirements like a maximum plastic strain, node displacement or a more homogenous internal energy density. Based on the fact that, the presented approach targets an optimisation bases on local effects, even the results of the academic load-case should
be applicable on industrial parts.
Findings of previous research show that laser-weld lines in the direction of the first principal stress are very efficient. This heuristics is implemented within the strategy "follow principal stress direction". Every combination of parameters and constraints like number of weld lines, minimum distance between weld lines etc. gets simulated. Eventually the optimal pattern is chosen based on pattern efficiency.
Furthermore the application of predefined patterns seems promising. The first step is to derive application zones with cluster algorithms. In the following step, parametrised base patterns like "horizontal lines" or "star" are combined. Different DOE schemes are used to reduce the number of combinations. To get a proposal for an optimal pattern two options are available.
The best combination can be taken directly from the DOE samples. Alternatively a proposal for an optimal pattern is derived from a cluster of best designs.
Eventually the evolutionary algorithm like strategy "iterative propagation" is presented. Be-ginning from automatically selected start elements, the laser-weld grow in different directions within each generation. With an adaptive "survival of the fittest" approach patterns are selected for further growth.
|Date||15th October 2019|