CAREER: A Systematic Data-Analytics Approach to the Design of Interface-Rich Materials

界面丰富的材料在所有工程系统中都很普遍，包括用于能量收集的涡轮盘，以及医学成像设备中的粘合层状复合材料。控制富界面材料性能的机制体现在中尺度上，即纳米和宏观系统（毫米到米）尺度之间的长度尺度。因此，为了有效地设计这些材料，以达到工程系统所需的性能指标（即x射线产生的热阻），需要对制造过程及其与这种“中间”长度尺度上的机制的关系进行定量理解。该学院早期职业发展（Career）奖支持整合介结构绩效理论和基于统计的“大数据”分析所需的基础研究。该设计范例将专门应用于锻造镍基高温合金，并有可能对新材料的快速插入和制造工艺产生广泛影响，从而降低界面丰富材料的工程系统的成本和上市时间。研究的目标是测试这样一个假设，即统计数据分析可以为制造满足特定性能要求的界面丰富的材料提供一个通用的设计范例。这些数据衍生模型的预测能力将被评估为Inconel-706锻造镍基高温合金，具有高温强度和低循环疲劳寿命的性能指标。这种合金对于制造高效能量收集应用具有重要意义。本研究的数据包括过去10年由美国铝业锻造研究小组获得的传统数据中的制造、细观结构和性能指标。研究团队将通过培养一个开放、健壮的数据基础设施来验证这一假设，该基础设施超越了简单的电子“文件柜”，允许分析工具无缝访问数据，并允许分析结果也与所有相关的元数据一起存储。探索性数据分析将指导团队确定需要哪些额外的数据集来提高数据派生模型的统计有效性。该研究计划通过材料基因组计划和综合计算材料工程，通过产生一个强大的开源数据框架，支持材料科学界更广泛的努力，该数据框架可推广到其他富含界面的材料的制造路线。

Interface-rich materials are pervasive in all engineered systems, including turbine discs for energy harvesting, and bonded layered composites in medical imaging equipment. The mechanisms that control performance of interface-rich materials manifest at the mesoscale, a length scale between the nanometer and macroscopic system (millimeter to meters) scales. Therefore, efficient design of these materials to achieve the performance metrics necessary for an engineered system (i.e., thermal resistance for x-ray generation) requires a quantitative understanding of the manufacturing processes and their relationship to the mechanisms at this "in-between" length scale. This Faculty Early Career Development (CAREER) award supports fundamental research needed integrate the theory of mesostructure performance, and statistical-based analysis of "big data." The design paradigm will be applied specifically to a forged nickel-based superalloy, and has the potential for broad impact on the rapid insertion of new materials and manufacturing processes to reduce cost and time-to-market for engineered systems which interface-rich materials.The research goal is to test the hypothesis that statistical data analytics can lead to a versatile design paradigm for the manufacturing of interface-rich materials for particular performance requirements. The predictive capabilities of these data-derived models will be assessed for Inconel-706 a forged Ni-based superalloy, with the performance metrics of high temperature strength and low cycle fatigue life. This alloy is significant for the manufacturing of efficient energy harvesting applications. The data for this research includes manufacturing, mesostructure and performance measures from legacy data obtained over the past 10 years by Alcoa Forging Research Group. The research team will test this hypothesis by cultivating an open, robust, data infrastructure that goes beyond a simple electronic "filing cabinet" and allows for seamless access to the data by analysis tools and allows for the analysis results to also be stored with all the associated metadata. Exploratory data analysis will guide the team in determining what additional datasets are needed to increase the statistical validity of the data-derived models. This research program supports broader efforts of the Material Science community through both the Materials Genome Initiative and Integrated Computational Materials Engineering by producing a robust, open-source data framework, which is generalizable to manufacturing routes of other interface-rich materials.