Ultrascalable Modelling of Advanced Materials with Complex Architectures

目前，发电和运输系统的技术进步受到材料方面的限制。理想的材料不适用于这些新设计所需的日益极端的环境(高温、高热流密度、压力、辐射损伤、疲劳等)。目前对这些问题的许多建议解决方案都是复合材料。然而，将具有不同组成材料性能的两个或多个物理上不同的相组合成单一材料充满了制造和建模的困难。基于组成材料的性能和微观结构形态对复合材料的体积性能进行可靠的预测是相当有兴趣的，因为这显然使新材料能够设计出具有特定要求(如韧性、硬度)的新材料。再加上快速成型和对复合材料制造过程的更好控制，这可能会带来新一代高性能材料。高分辨率的三维成像技术，如x射线微层析成像(XMT)，在材料科学上相当于医学CAT扫描，现在可以探测到亚微米级，与数值解算器结合在一起，原则上可以为模拟物理过程提供交钥匙解决方案。然而，采用基于图像的分析有两个主要的技术障碍：(1)稳健而准确地将3-D数据转换成适合于求解器的计算网格；(2)以适当的分辨率和大小研究区域以连接微观到宏观长度尺度所需的计算问题的大小，从而使模拟的体积代表材料的大部分。这两个问题将在项目中通过结合和进一步开发申请者为解决大规模问题(新型迭代求解器)和从3D图像进行网格划分而开发的最先进技术来解决。为了为将要开发和实施的解决技术提供佐证，将考虑两个问题：陶瓷基复合材料和开孔泡沫，一些申请者在这两个问题上有经验，它们代表了非常广泛的工业和生物重要结构。本课程将探讨结构和热学性质。这些材料是范例，因为它们代表了计算方法的两个不同挑战：复合材料具有多相复杂的结构；泡沫经历了非常大的应变变形，随后是单元接触和应变局部化。这个雄心勃勃的项目解决了三个密切相关的问题，这些问题需要团队中包含的技能的结合，他们的解决方案将在计算和材料工程中得到深远的应用，即：预测具有复杂结构的材料的行为；并行模拟具有大应变和接触的问题；以及重新划分可变形介质的高效算法。

Technological advances in power generation and transport systems are currently materials limited. Ideal materials are not available for use in the increasingly extreme environments (high temperatures, high thermal fluxes, pressures, irradiation damage, fatigue etc) these novel designs require. Many of the current suggested solutions to these problems are composites. However, the combination of two or more physically distinct phases with different constituent material properties to form a single material is fraught with fabrication and modelling difficulties.There is considerable interest in the reliable prediction of the bulk properties of composite materials based on the properties of the constituent materials and the microstructural morphology as this clearly enables novel materials to be designed with specified requirements (e.g. toughness, stiffness). Coupled with rapid prototyping and greater control of composite fabrication processes, this could deliver a new generation of high performance materials. High resolution imaging in 3-D such as x-ray microtomography (XMT), the materials science equivalent of medical CAT scans, can now probe at the sub-micron scale and coupled with numerical solvers could in principle provide turnkey solutions for modelling physical processes. However there are two main technical hurdles to the adoption of image based analysis: (1) robustly and accurately converting the 3-D data into computational meshes suitable for solvers; and (2) the size of computational problem required to study domains at suitable resolutions and size to bridge the micro to macro length scales such that the volumes modelled are representative of the bulk of the material. Both of these problems will be addressed within the project by combining and further developing state of the art techniques developed by the applicants for solving large scale problems (novel iterative solvers) and for meshing from 3-D images.In order to provide corroboration for the solution techniques to be developed and implemented, two problems with which some of the applicants have experience and which typify a very broad range of industrially and biologically important structures will be considered: ceramic matrix composites and open-celled foams. Both structural and thermal properties will be explored. These materials are exemplars as they represent two different challenges to the computational approach: the composite has a multiphase complex architecture; the foam undergoes very large strain deformation followed by element contact and strain localisation. These challenges are common to a wide range of materials.This ambitious project addresses three intimately linked problems that require the combination of skills contained within the team whose solutions will have far reaching application in computing, and materials engineering, namely: predicting behaviour of materials with complex architectures; parallel simulation of problems with large strains and contacts; and efficient algorithms for remeshing deformable media.