Quantitative analysis of local deformation and direct correlation with microstructural and mechanical parameters during equal channel angular pressing in tools with variable geometries

等通道角挤压（ECAP）是一种加工技术，可以使金属材料发生严重的塑性变形，从而导致大量的晶粒细化。由此产生的超细颗粒材料具有优异的机械性能（高强度与良好的延展性相结合），使其在轻量化设计等应用中具有潜在的吸引力。虽然近年来ECAP对许多合金的微观组织和性能的影响已经进行了非常详细的研究，但在不同ECAP工具的塑性变形区域的局部变形方面仍然存在基本的知识差距：通常，等效应变仅基于简化模型进行估计，而不考虑例如实验记录的材料硬化或通常应用的背压的影响。然而，为了定量地描述微观组织变形机制和晶粒细化之间的相互关系，需要更详细地了解局部（塑性）变形，迄今为止只进行了定性分析，并从根本上研究和预测物理冶金过程作为材料和加工参数的函数。首席研究员最近扩展了一种分析模型，该模型可以准确地描述90°-ECAP工具沿流线的局部变形，因此它也可以用于分析任意工具的几何形状。结合同样在初步工作中开发的高分辨率视觉塑性技术，现在存在一套工具集，可以研究该建议的关键假设-假设实际等效应变（由流线指数表示，其在工件中的梯度），这些参数随着ECAP通道数量的增加而演变)代表了一种物理相关的测量，可以与许多不同的现象相关（甚至用于定量预测），如微观结构和纹理演变、材料特性或损伤积累。在这个项目中，将使用三种模型合金和四种不同的ECAP工具来实验表征不同加工条件下的塑性变形区域以及塑性变形如何沿着不同的流线进行。将开发基于人工智能的图像识别算法，并用于在三个互补的层面上精确拟合局部变形。这也将首次允许评估不同流线模型的质量，从而为目前在ECAP社区建立的不同建模方法提供基准测试。最后，实验和理论工作还将提供明确识别相关加工和材料参数的手段，并将其与所得的宏观性能直接联系起来。

Equal Channel Angular Pressing (ECAP) is a processing technique that allows to subject metallic materials to severe plastic deformation which leads to massive grain refinement. The resulting ultrafine-grained materials are characterized by exceptional mechanical properties (high strength in combination with good ductility), making them potentially attractive for applications in, for instance, light-weight design. While the effect of ECAP on microstructure and properties of many alloys has been studied in great detail in recent years, there still exists a fundamental knowledge gap with respect to local deformation in the plastic deformation region of different ECAP tools: typically, equivalent strains are only estimated based on simplified models that do not consider, for example, the experimentally documented effects of material hardening or of a commonly applied back-pressure. A more detailed understanding of local (plastic) deformation, however, is needed in order to quantitatively describe the interrelationships between microstructural deformation mechanisms and grain refinement, which have only been analyzed qualitatively so far, and to fundamentally study and predict physical metallurgy processes as a function of material and processing parameters. The principal investigator has recently extended an analytical model that accurately describes the local deformation in 90°-ECAP tools along flow lines such that it can be also used to analyze arbitrary tool geometries. In combination with a high-resolution visioplastic technique, also developed in preliminary work, there now exists a tool-set that allows to study the key hypothesis of this proposal – the assumption that the actual equivalent strain (as expressed by the flow line exponent, its gradient in the work-piece, and the evolution of these parameters with increasing number of ECAP passes) represents a physically relevant measure that can be corelated with (or even used to quantitatively predict) many different phenomena, like microstructural and texture evolution, material properties, or damage accumulation. In this project, three model alloys and four different ECAP tools will be used to experimentally characterize the plastic deformation region under different processing conditions and how plastic deformation proceeds along different flow lines. Image recognition algorithms, based on artificial intelligence, will be developed and used to accurately fit local deformation on three complementary levels. This will for the first time also allow to evaluate the quality of different flow line models and thus to provide a benchmark test for different modeling approaches that are currently established in the ECAP community. Finally, the experimental and theoretical work will also provide the means to clearly identify relevant processing and material parameters, and to directly relate them with the resulting macroscopic properties.