您当前位置: 首页  >  学术讲座

讲准字23号:Future Challenges in Modelling and Simulating Electrochemical Energy Conversion Technologies

发布时间:2018-03-13|浏览次数:

题目:Future Challenges in Modelling and Simulating Electrochemical Energy Conversion Technologies

主讲:Akeel Shah

时间:2018年3月21日 10:00

地点:能源研究院1517报告厅

主办:能源研究院


主讲简介:Akeel Shah,英国华威大学教授。研究专长:储能技术,液流电池。Dr Akeel Shah is a Reader in School of Engineering, University of Warwick, UK. His current research interests lie in the development of energy storage systems, with an emphasis on organic flow batteries and computational and multiscale modelling of complex systems and materials including predictive modelling/analytics approaches. He got all his degrees from University of Manchester, UK. Before his current position, he was a Senior Lecturer in University of Southampton and an Associate Professor in University of Warwick. He was Awarded a joint Mathematics Fellowship of Information Technology and Complex Systems/Pacific Institute of Mathematical Sciences (MITACS/PIMS). He has published more than 60 peer-reviewed journal papers with a citation of 2613 (H-index=23).

 

主讲内容:In this talk we discuss present and future challenges involved in modelling electrochemical energy conversion storage technologies, with a focus on fuel cells, flow batteries and metal-air batteries. Traditional modelling approaches are based on continuum models, which are valid at the macroscopic scale. Such approaches, however, are severely limited in key aspects, especially in terms of: capturing fundamental phenomena such as the deposition of a metal on an electrode surface; designing and selecting new materials and active species; and predicting the performance under extreme conditions (leading to degradation and possibly failure). There is also an issue of computational cost; complex, multi-diemnsional models are burdensome in terms of memory and simulation time, which renders applications such as sensitivity analysis and optimization extremely challenging, and indeed usually unfeasible. We provide details of a number of methods developed in our group to overcome such issues using data-driven (machine learning) methods combined with nonlinear dimensionality reduction (manifold learning). We show how these methods can be used for expensive tasks such as uncertainty quantification relating to a hydrogen fuel cell. We further outline approaches to screening new materials (e.g., organic species in flow batteries) using ab-initio methods, where the challenge is to predict not only electrochemical activity but also properties such as solubility. Finally, a kinetic Monte Carlo (kMC) approach to simulating metal deposition is outlined and we present preliminary results, alongside an emulation approach to predict the deposition profile in a fraction of the time taken for the full kMC.


欢迎广大师生参加!