Modelling Microstructure Evolution Phase Separation

我们的研究计划旨在开发将描述相行为的热力学与化学反应和支化聚合物的分子运动的动力学相结合的计算模型。我们的工作是受到最近的实验观察的推动，这些实验观察目前缺乏物理解释。这样的解释是设计和优化新配方的关键。最令人惊讶的是，正如提出者在最近完成的索尔维资助的博士项目中进行的中子散射实验所揭示的那样，在液态的混合物中存在预先存在的纳米级结构。这种结构似乎与里昂Solvay/CNRS联合研究中心的合作者的另一个令人惊讶的观察结果相关联，即即使在形成网络之后，混合物仍在以与相前分离结构相对应的长度尺度继续显著进化。现有的模型无法解释这两种观测结果，因为现有模型未能捕捉到分支如何影响热力学和动力学的物理后果。我们最近开发了一个蒙特卡罗模型，它能够模拟支化/线性聚合物的共混，并证明它能够预测平衡现象。我们在这个项目中的目标是：1.将模型从平衡过程扩展到非平衡过程，以便对相分离动力学进行建模并预测共混物的流动响应。后者对于比较在威尔顿进行的固化样品的理论和实验是必不可少的。使用蒙特卡罗模型对较粗颗粒的方法进行参数化，以模拟结构和存在界面时的结构演变，这在将共混物浸渍到碳纤维中时尤为重要。通过使用图形处理单元的并行化显著提高代码效率，以便为更大的系统模拟更长时间的过程，使结果在实验上更具相关性。4.支持在索尔维与里昂Long教授的研究小组合作进行的中子散射实验，这将是验证和改进我们的模型所必需的。

Our research program aims to develop computational models that couple the thermodynamics describing the phase behaviour with the kinetics of both chemical reactions and the molecular motion of branched polymers. Our work is motivated by recent experimental observations that currently lack a physical interpretation. Such interpretations are key to the design and optimisation of new formulations. Most surprisingly, as revealed by neutron scattering experiments conducted by the proposer in a recently completed Solvay funded PhD project, there are pre-existing nanoscale structures in the blend in the liquid state. Such structures appear to be correlated with another surprising observation, by collaborators at the joint Solvay/CNRS research centre in Lyon, that the blends continue to evolve significantly, even after the formation of a network, at length-scales corresponding with that of the pre-phase separated structures. Neither of these observations can be explained by existing models, which fail to capture the physical consequences of how branching impacts on the thermodynamics and kinetics. We have recently developed a Monte Carlo model that is able to model blends of branched/linear polymers and demonstrated that it is able to predict equilibrium phenomena. Our aims in this project are to build upon this work:1. To extend the model from equilibrium to non-equilibrium processes in order to model the kinetics of phase separation and predict the flow response of the blends. The latter is essential for comparing theory with experiments conducted on curing samples at Wilton.2. To use the Monte Carlo model to parameterise coarser grained methods for modelling structure and structure evolution in the presence of interfaces, which is of particular importance when the blends are impregnated into carbon fibres.3. To significantly increase code efficiency through parallelisation using Graphical Processing Units, in order to model longer time processes for larger systems, making results more experimentally relevant. 4. To support neutron scattering experiments, being conducted at Solvay in collaboration with the research group of Professor Long in Lyon, which will be essential to verify and refine our modelling.