Characterization and Mapping of Mechanical Properties in Complex Materials and Geometries

该研究计划建立在20多年的生物组织机械性能表征和映射经验的基础上，通过开发一套连贯和集成的材料模型以及用于机械性能表征的数学和计算方法，以促进这些技术在广泛的应用中的使用。该研究计划的主要目的是创建一个强大的，可重复的和可量化的技术工具箱，能够跨多个领域运行，从医学和生物医学应用，如神经成像，癌症筛查和细胞生物学到工业设计和制造应用，如航空航天，运输和体育行业的先进复合材料结构的表征。为了实现这一目标，本研究计划将集中在三个主要研究领域：建模，方法和模式（应用）。在建模领域，有两个主要目标。第一，将用于杆、梁、板和壳建模的结构有限元集成到现有的表征方法中，以便于应用于更简单的2D几何形状和制造结构。该领域的第二个目标是开发多尺度建模程序，用于模拟生物组织和复合材料行为的复杂，微尺度流体-固体和纤维-基质相互作用。在方法领域，第一个主要的研究目标是扩展现有的表征方法，包括先进的预处理和后处理数据分析方法，以加快计算时间和提高准确性。这一领域的第二个目标是为所产生的财产地图制定量化的不确定性措施。这些不确定性测量将与现有的体内特性图和模拟数据相结合，为基于深度学习的直接表征方法开发无偏的完整训练方法。在应用模式领域，该研究计划的主要目标是标准化用于不同表征技术的软件工具，以便各种模型和方法易于在不同应用中进行测试和应用。这包括使用不同的数据源（从数码相机到磁共振成像），在不同的物理机制下操作（从细胞尺度的高频成像到结构面板的准静态成像），并提供不同的结果（例如，频率相关材料特性的详细地图与均匀结构特征的单一估计）。

This research program builds on more than 20 years' experience in the characterization and mapping of mechanical properties in biological tissues by developing a coherent and integrated set of material models as well as mathematical and computational methods for mechanical property characterization to facilitate the use of these techniques across a wide range of applications. The principal aim of this research program is to create a toolbox of robust, repeatable, and quantifiable techniques capable of operating across multiple domains, from medical and biomedical applications such as neuroimaging, cancer screening, and cell biology to industrial design and fabrication applications such as the characterization of advanced composite structures for the aerospace, transport and sporting industries.******To achieve this aim, this research program will focus on three main research areas: Modeling, Methods and Modes (of Application). In the area of Modeling, there are two major objectives. The first, to integrate structural finite elements used for the modeling of bars, beams, plates, and shells into the existing characterization methods to facilitate application to simpler, 2D geometries and manufactured structures. The second objective in this area is to develop multi-scale modeling procedures for simulating the complex, micro-scale fluid-solid and fiber-matrix interactions that underlie the behavior of biological tissues and composite materials.******In the area of Methods, the first major research objective is to expand existing characterization methods to include advanced pre- and post-processing data analysis methods to accelerate computation times and increase accuracy. The second objective in this area is to develop quantified uncertainty measures of the resulting property maps. These uncertainty measures will be coupled with existing in-vivo property maps and simulated data to develop unbiased, complete training methods for deep learning based direct characterization methods.******In the area of Modes of Application, the major objective of this research program is to standardize the software tools used for different characterization techniques, such that the various models and methods are easily tested and applied across different applications. This includes the use of different data sources (from digital cameras to Magnetic Resonance Imaging), operating in different physical regimes (from high frequency imaging at the cellular scale to quasi-static imaging of structural panels) and providing different outcomes (e.g. detailed maps of frequency dependent material properties vs. single estimates of homogeneous structural characteristics).**