Robust Design Of Sheet Metal Forming Process And Topology Optimization Die Structure High Strength Steel

High strength steel (HSS) is one of the most widely used material for automotive industry at present. Compared with mild steels, it has much higher yield stress, which can maintain structure stiffness of vehicle body in white with thinner material, and dramatically reduce the vehicle weight. However, application of HSS brings challenges to sheet metal part manufacturing. Firstly, as a new material, experience of design process and die structure is not sufficient. Secondly, higher yield stress leads to worse formability, higher die load, bringing difficulties to process and die structure design. Thirdly, springback is more severe. Fourthly, variations of material properties of HSS exist due to unstable sheet rolling process, making the stamping quality and die load quite unpredictable. In order to solve the above problems, the research work funded by National 863 program (Grant # 2007AA04Z130) was conducted as follows:1. Considering the impact of noise factors on stamping quality, a robust design method for sheet metal forming process based on support vector machine and adaptive importance sampling was proposed, the method can not only find a process to achieve best stamping quality, but also guarantee the satisfied quality reliability and minimum quality variations. Robust designs of forming process for galvanized steel and high strength steel were realized. Compared with results from deterministic optimal results, though value of quality objective decreases, robustness of forming process is guaranteed and probability of quality problems was greatly reduced. Moreover, thickness distribution of forming part by robust designed process was improved, and springback magnitude after trimming was also reduced.In order to evaluate the reliability of forming process during robust design, an adaptive sampling method was proposed to quantify impact of variations of noise factors on quality index such as split, wrinkle and springback. Compared with conventional Monte Carlo sampling method, the new method reduces sampling scale and improves probability calculation efficiency. Compared with other approximation methods for probability calculation, the new method overcomes prediction inaccuracy due to high nonlinearity of constraint functions.In order to reduce optimization iteration time and keep accuracy of robust design, a response surface method based on support vector machine was proposed. With respect to characteristics of optimization objective and constraints, difference types of support vector machine were used to approximate relationship between forming process and quality index. The method overcomes over-fitting and poor nonlinear capability of other approximation models. 2. Due to insufficient design experience related with stamping die structure of high-strength steel forming, topology optimization algorithm for stamping die structure based on SIMP method was developed. Improvement information reflected from results of die structure analysis was fully utilized, and optimization algorithm was used to search the most suitable material distribution for the structure, providing reasonable reference for die design.Binder structure design of a hyperbolic cup drawing die was studied to validate the algorithm. Based on the specific feature of the binder, an optimization model which targets to minimize structure volume with displacement constraints on the binder surface was established. Modeling techniques concerning force boundary conditions and optimization parameter selection were discussed. The final optimal structure achieved 27% weight reduction compared with the original one.3. A step-bottomed cup drawing die of high-strength sheet was further designed to validate effectiveness of topology optimization of the stamping die structure. Impact of die assembly error on forming quality was analyzed at first. It is pointed out that guiding structure could eliminate the impact of those errors, ensuring the robustness of the forming process. Characteritics of die with wearplate were further studied. It is indicated that both total amount and position of wear plates had great impact on the horizontal forming load. Due to the wear plate, side force exerted on punch from sheet metal during stamping is exerted indirectly on binder, making its load distribution much more complex.The proposed topology optimization method was used to optimize the binder part of the testing die. Both compressive load and side load were taken into account. Results of comparing experiment with the old design shows: 1) Optimized die structure saves 28% material. However, local stiffness of the die structure is less affected, and more reasonable material distribution strengthens the die structure; 2) Displacement constraints on binder surface guarantees that even though die structure topology is changed, the stamping part quality was not largely influenced. The maximum difference of part thickness on a section between the original one and optimal design is only 0.06mm, 3.6% of the initial thickness of the sheet metal.Finally, the proposed topology optimization algorithm was used to optimize the binder of a drawing die for high-strength floor member part. Compared with the old design, the new structure with same material volume shows better performance while maximum structure displacement is only 45.4% of that on the old one. Die weight is reduced by 50% with the same performance requirement. The research results indicate the proposd methods are valuable to both forming process robust design and light-weight die structure design of high strength steel.