LASCISO · Large Scale Industrial Structural Optimisation for Advanced Applications

Large Scale Industrial Structural Optimisation for Advanced Applications

1 FINAL PUBLISHABLE SUMMARY REPORT 1.1 SUMMARY DESCRIPTION OF THE PROJECT OBJECTIVES The goal of LaScISO (Large Scale Industrial Structural Optimisation for Advanced Applications) is to enhance computational structural optimisation methods to be capable of optimisation with current industrial state-of-the-art simulation techniques, i.e. to “optimize what can be simulated”. With the initialization of the project no appropriate optimization capabilities existed to optimize problems in the highly challenging regime of non-linear and/or coupled multiphysic applications of computational shape and topology optimization. The scientific and technology objectives consequently deal with aspects regarding efficiency and effectiveness of methods to be developed or enhanced to be able to reach the goals. They address most latest scientific results in the fields of topology optimization as well as sensitivity based shape optimization for mechanically nonlinear, coupled, multiphysic problems. From the point of practical application issues of process- and CAD-integration for large optimisation problems are addressed with special focus on manufacture constraints and processes. The partners are leading in their special fields and provide transfer of state-of-the art methods. 1.2 WORK PERFORMED DURING THE PROJECT, LaScISO had been working on some of the most challenging subjects of large scale, industrial compatible structural optimization. Consequently, work was performed on parallel lines dealing with non-linear, multi-physic design objectives, topology optimization strategies, large scale sensitivity techniques (“exact” semi-analytical and adjoint) for shape optimization (parametric and node-based) and the related computational aspects including CAD-reconstruction and the implementation in an industrial environment. 1.3 FINAL RESULTS, THEIR POTENTIAL IMPACT AND USE The optimisation techniques developed by LaScISO support the computer based analysis and design of everyday industrial products. Optimization techniques are most important ingredients of an overall computer based implementation of the design chain. CAD-based geometric design, finite element based mechanical simulation and computational optimisation are integrated under an unique user interface. The goal is to reduce the time-to-market but, at the same time, to improve product quality and sustainability. Also, optimisation methods help to propose new and more efficient designs which are beyond actual technical knowledge or traditions. Mechanical and physical properties may be tested by virtual experiments with a minimum of physical components by use of simulation and optimisation. The specific methods as developed by LaSciSO provide an automatic and assisted way to implement design changes and improvements within the existing CAE Environment of the user and within the existing used FEM solvers like ABAQUS, ANSYS or NASTRAN which are widely available. This is an essential action to decrease the time spent on product development and reduce the amount of faulty interpretations of results. Every important industry is already on the way to reorganize their development chains by the use of simulation techniques. As an example, the implementation by Dassault is to be mentioned which also may serve as the proof for the importance, the future relevance and the economic success of the general trend as well as of the methods developed by the consortium. LaScISO has transferred knowledge in a very powerful way from academia to industry and vice versa. The gained knowledge from the advanced simulations is a major competitive advantage towards non-European competitors. This applies to a number of very diverse product types, such as cars, trains, aero planes, mobile phones, wind turbines, power-tools, heat exchangers or electric motors where European industry is leading worldwide and in steady competition. 1.4 MAIN RESULTS ACHIEVED MAIN RESULTS FOR TOPOLOGY OPTIMIZATION: • Topology optimization of multiphysic problems: consistent continuous formulation • Efficient solvers for transient and multiscale systems and topology optimization. MAIN RESULTS FOR SHAPE OPTIMIZATION: • Adjoint and “exact” semi-analytical method for linear and non-linear shape optimization developed. • Surface control and regularization methods for large scale shape optimization MAIN RESULTS FOR CAD RECONSTRUCTION, DATA PROCESSING AND INDUSTRIAL APPLICATIONS • Improved CAD-transfer methods • Rapid surfacing to transform very detailed point clouds into highly efficient models SOFTWARE IMPLEMENTATIONS • Sizing optimization for multiphysics, i.e. thermal and static loading as well as coupled vibro-acoustic • Topology optimization for conduction and convection • Sensitivity based shape optimization for mechanically non-linear problems in SIMULIA / TOSCA (2016X) • Reconstruction methods including sub-division surfaces as well as rapid CAD-reconstruction. www.LaScISO.eu 1.5 PRINCIPAL INVESTIGATORS: Prof. Dr.-Ing. Kai-Uwe Bletzinger, coordinator and leader WP 2 Technische Universität München, Germany, kub@tum.de Prof. Ole Sigmund, leader WP 1 Topopt Group, Danish Technical University (DTU), Lyngby, Denmark, sigmund@mek.dtu.dk Prof. Dr. Stoyan Maleshkov, Prof. Dr. Georgi Todorov, leaders WP3 Technical University of Sofia (TUS), Bulgaria, maleshkov@tu-sofia.bg Dr. Jürgen Sauter, Peter Clausen, leaders WP 4 Dassault Systemes Germany (DS); formerly Fe-Design (FED), Karlsruhe, Germany Juergen.SAUTER@3ds.com, Peter.CLAUSEN@3ds.com

Data: CORDIS, © European Union