Numerical optimization strategies for springback compensation in sheet metal forming
| Autor: | Luís Menezes, Marta Oliveira, Ana F. Maia, E. Ferreira, António Andrade-Campos |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Engineering
Mathematical optimization 021103 operations research business.industry 0211 other engineering and technologies Automotive industry Process (computing) Forming processes 02 engineering and technology Structural engineering 021001 nanoscience & nanotechnology Finite element method Compensation (engineering) Metamodeling visual_art Benchmark (computing) visual_art.visual_art_medium 0210 nano-technology business Sheet metal |
| DOI: | 10.1016/b978-0-85709-481-0.00003-3 |
| Popis: | The industrial sheet metal forming process is known for being influenced by undesired effects such as springback, fracture, or wrinkling, leading to a lack of quality of the final part. Springback is a critical problem, particularly in the automotive industry where mass production is frequently used and geometric inaccuracies result in heavy economic losses. Therefore, when designing a forming process, it is mandatory to take into consideration the springback behavior of the part in order to compensate or reduce it, through the adjustment of the process parameters such as external forces or tools geometry. The aim of this work is to discuss efficient strategies to compensate springback. To this end, two different approaches are compared: the Finite Element Method Updating and the Metamodeling Optimization. These approaches can use distinct solution evaluation strategies, namely Experimental and Finite Element Analyses, using both implicit and explicit time integration schemes. These solution evaluations are used to quantify the efficacy of the solution for the springback compensation problem, which is then used as input in the optimization algorithms to obtain new designs for the forming process. Considerations are also given regarding the use of different forming tools’ parameterizations in the springback compensation problem: NURBS versus classical geometrical parameterization. All different approaches are applied to the springback compensation problem of the U-shaped rail, a benchmark that shows large elastic recovery after the tool removing. The results are also compared with the ones obtained by other authors, highlighting that, even for this simple example, metamodeling optimization requires a large number of evaluations. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|