Efficient Simulation and Optimization for Tandem Press Lines

Licentiatavhandling, 2012

Today, simulation is needed as one of the tools for optimum utilization of a press line. Utilization includes: high throughput, minimum ware of the line and high quality. High quality includes quality of the produced parts, quality of the stamping dies and quality of the tooling manufactured. In order to perform efficient geometrical simulation and virtual commissioning in stamping, two main fields are investigated namely simulation time and optimization time. Thus, reducing computation time is the main theme of this thesis. An efficient press line simulation model is built and verified, resulting in reduced simulation building time due to the modularization of the model. To reduce simulation time, collision detection time is reduced by a method based on 3D to 2D geometrical collision detection. The method is based on pre-calculation of all collision points in the environment of interest, and then using a simplified collision detection model in a simulation based optimization. This is less resource consuming than collision checking the original 3D objects for all optimization evaluations. The suggested approach reduces the collision detection from a 3D to a 2D problem, where collision between simplified but moving curves is used in the repeated simulation for optimization. This collision detection approach, together with a simplified implementation of the control code, results in ~200 times reduction of the computation time, compared to the original simulation based on standard 3D collision detection. The variable parameters in a press line exceed 100, resulting in time/computationally demanding computations. There is also a need of fast optimizations for die design, line tryout and ramp-up. Since the number of evaluations grows exponentially with the number of dimensions in an optimization problem, optimization time is reduced by a decomposition strategy aiming at dimension reduction. Two simulation/optimization strategies were chosen and tested to decrease calculations. The presented results mean that simulation and virtual commissioning can be performed not only for press stations but also for complete press lines, where the complexity increases linearly with the number of stations in the line. KEYWORDS: Press Line Simulation, Simulation based optimization, Virtual manufacturing, Virtual commissioning, Parameter tuning, Optimization strategies.