Development of a material model suited for topology and shape optimization of crash structures made of thermoplastic fiber reinforced composites

有可能设计出纤维增强塑料（FRP）的碰撞结构，其在可调节和恒定的力水平上吸收动能。只要设计适合FRP材料，它们就能获得比传统金属解决方案更好的结果。这对于由热塑性塑料制成的FRP材料尤其重要，因为与流行的热固性材料相比，它们在回收和机械连接方面具有优势。然而，FRP部件的设计复杂性增加是具有挑战性的。为此，基于非线性有限元计算的数学优化任务可以提供支持，并有助于改进设计过程。这些优化任务的执行需要关于所使用的材料模型的特殊要求。在本研究项目的范围内，将开发一个灵活的和优化兼容的热塑性单向FRP材料的材料模型。最后，该模型可用于应用于碰撞结构的自动拓扑和形状优化任务。因此，应用混合优化过程，其使用基于规则或启发式方法来修改所选择的结构描述，并通过数学优化算法来控制其它结构描述。通过数学图形来实现几何表示，是一种灵活的方法。因此，每个设计都有一个可评估的有限元碰撞模型。在每一个中，材料模型和启发式方法都得到了更精确的阐述。因此，材料模型的质量和开发过程在每个步骤中使用给定边界条件的结构优化进行基准测试。第一步是设计一个材料模型，该模型代表单向热塑性FRP的刚度、强度和损伤。采用不同的试样级测试方法对材料常数和功能进行了实验评价。通过已经定义的拓扑学和通用材料模型同时进行拓扑优化。这种方法产生关于单个参数的灵敏度的信息。第二步包括材料模型的升级，以计算渐进损坏（损坏后/破碎）。为此，制造并测试组件和试样级的样本。此外，不同的生态学与给定的目标进行了研究：产生强大的FRP结构，集成的FRP制造约束和管理的材料模型。与步骤1和步骤2中的结果一起，在步骤3中执行示例性部件优化（多功能支撑部件）。借助实验验证，可以验证所开发的结构优化材料模型的质量和适用性。毕竟它也可以用于热塑性FRP材料。

It is possible to design crash structures out of fiber reinforced plastics (FRP), which absorb kinetic energy on an adjustable and constant level of force. Provided that the design is suitable for FRP materials they achieve better results than conventional metallic solutions. This counts especially for FRP materials made out of thermoplastics because they have advantages regarding recycling and mechanical joining compared to the popular thermosets. Nevertheless the increased design complexity of FRP parts is challenging. For this purpose mathematical optimization tasks based on non-linear finite element calculations can be supportive and help to improve the design process. The execution of those optimization tasks need special requirements regarding the material model used. Within the scope of this research project a flexible and optimization compatible material model for thermoplastic unidirectional FRP materials will be developed. In the end this model can be used for automated topology and shape optimization tasks applied on crash structures. Therefore a hybrid optimization procedure is applied, which uses a rule based or heuristic approach to modify selected structure descriptions and controls the other structure descriptions by mathematical optimization algorithms. The geometry representation is realized through mathematical graphs, which is a flexible approach. Consequently an evaluable finite element crash model exits to each design.The development is carried out in three steps. In each of them the material model and heuristics are stated more precisely. Therefore the quality and development process of the material model is benchmarked within each step using structural optimization regarding given boundary conditions. The first step consists of the design of a material model that represents stiffness, strength and damage of a unidirectional thermoplastic FRP. The experimental evaluation of the material constants and functions is carried out by using different coupon-level test methods. By means of already defined heuristics and the generic material model topology optimizations are carried out simultaneously. This approach results in information about the sensitivity of the single parameters. The second step includes an upgrade of the material model in order to calculate progressive damage (post-damage/crushing). Specimens on component and coupon level are manufactured and tested for this purpose. Additionally different heuristics with given objectives are investigated: Generation of robust FRP structures, integration of FRP manufacturing constraints and management of the material model. Together with the results in step 1 und step 2 exemplary part optimizations (multifunctional support part) are performed in step 3. With help of experimental validation the quality and applicability of the developed material model for structural optimization can be verified. After all it can also be used for thermoplastic FRP materials.