Fundamental mechanisms and modeling of microstructure evolution during beam and powder bed-based additive manufacturing

粉末床金属的基于光束的增材制造（AM）不仅提供了构建高性能材料的复杂定制组件的机会，而且还可以通过熟练的加工来调整局部材料特性。凝固条件的变化可以改变微观组织的长度尺度。此外，最新的研究结果表明，在制造过程中，组件的纹理也是可调节的。因此，对于轻量化组件的优化，开辟了全新的视角，因为不仅拓扑结构，而且材料的纹理都可以根据组件上的局部负载进行调整。为了理解和控制织构演变，需要从根本上了解水动力非平衡凝固过程（晶粒生长、选择和成核）。实验研究表明，特别是在AM极端条件下的成核机制没有得到充分的解析，也不能用经典模型再现。本文的目的是识别、从根本上了解和物理模拟微观组织的演变，特别是在特殊凝固条件下的成核。这个模型应该在现有的软件中实现，这是在我们的主持下开发的。以IN718的增材制造样品为基础，进行了建模和验证实验。在项目结束时，该模型应预测梁式和粉床式增材过程中的凝固组织、晶粒组织和织构演变。该项目利用我们的软件来模拟基于梁和粉末床的AM过程中的固结过程。该软件包含晶格玻尔兹曼方法来描述在熔化和凝固过程中的水力和热力学。该方法结合元胞自动机模拟凝固过程中的晶粒结构演变，忽略了当前晶粒的成核过程。我们的新理论分析除了包含温度梯度和凝固前沿速度外，还首次包含了关于前一层的织构（取向、细胞/枝晶间距、偏析）和当前加工层界面处熔体的局部成分的附加信息。需要研究的是，凝固前沿的取向变化与快速熔化材料中的偏析（熔体记忆）是如何通过局部过冷诱导晶粒成核的。这些发现在数学上被用于晶粒成核模型。

Beam-based additive manufacturing (AM) of metals in a powder bed not only offers the opportunity to build complex, custom-made components of high-performance materials, but also to adjust the local material properties by proficient processing. The variation of solidification conditions enables the modification of microstructure length scales. Additionally, latest research results indicate, that also the texture of the components is adjustable during manufacturing. Therefore, entirely new perspectives are opened regarding optimization of light weight components, because not only the topology, but also the texture of the material is adjustable to the local loads on the component. In order to comprehend and control the texture evolution, the hydrodynamic non-equilibrium solidification process (grain growth, selection and nucleation) needs to be fundamentally understood. Experimental investigations show that especially the mechanisms of nucleation under the extreme conditions of AM are insufficiently resolved and are not reproduced by classical models.The aim of this proposal is to identify, to fundamentally understand and to physically model the microstructure evolution, especially the nucleation under the special solidification. This model should be implemented in existing software, which is developed at our chair. Modeling and verification are experimentally substantiated basing on additively manufactured samples of IN718. At the end of the project the model should predict the solidification structure, grain structure and texture evolution during beam and powder bed-based AM.The project draws on our software for simulating the consolidation process during beam and powder bed-based AM. The software contains a lattice Boltzmann method to describe the hydro- and thermodynamics during melting and solidification. This method s coupled to a cellular automaton modeling the grain structure evolution during solidification neglecting currently grain nucleation. Our new theoretical ansatz contains besides the temperature gradient and the solidification front velocity for the first time additional information about the texture of the previous layers (orientation, spacing of cells/dendrites, segregation) and the local composition of the melt at the interface to the currently processed layer. It should be investigated, how orientation changes at the solidification front in combination with the present segregations in the rapidly melted material (memory of melt) induce grain nucleation by local undercooling. These findings are mathematically utilized for a grain nucleation model.