Monolith多尺度模型前沿进展:从有限元方法到机器学习  

Recent progress in multiscale model of monolith:from finite element method to machine learning

作　　者：单斌[1] 陈荣昕 杨家强 Chaitanya Sampara SHAN Bin;CHEN Rongxin;YANG Jiaqiang;Chaitanya Sampara(School of Materials Science and Engineering,Huazhong University of Science and Technology,Wuhan 430074,Hubei,China;Viridis Chemicals Private Limited,54A Mittal Towers,Nariman Point,Mumbai 400021,India)

机构地区：[1]华中科技大学材料科学与工程学院,湖北武汉430074 [2]Viridis Chemicals Private Limited,54A Mittal Towers,Nariman Point,Mumbai 400021,India

出　　处：《金属功能材料》2022年第3期1-15,共15页Metallic Functional Materials

摘　　要：Monolith催化器被广泛用于移动源和固定源的废气净化,是气相催化体系中的关键优化对象。计算模拟的方法有助于进一步理解monolith中宏微观耦合的多尺度物理与化学过程,实现催化体系的快速评估,促进废气后处理催化体系的更新迭代与技术升级。总结了monolith多尺度模型的总体框架,详细讨论了建模过程中涉及到的催化器内部物质与能量的扩散、反应动力学、流动均匀性等基本原理与过程,综述了不同尺度模型的有限元求解方法特点与近年来结合机器学习模型的前沿进展。最后对monolith多尺度建模领域面临的挑战与发展趋势进行了展望。Monolith catalytic converter is widely used in the exhaust gas purification of mobile and stationary sources,which is a key optimization target for gas-phase catalysis systems.The simulation approach could help understand the multiscale physical and chemical processes coupling macroscale and microscale in the monolith,as well as enable the rapid evaluation of catalytic systems.It is also of vital importance for the iteration and technology development of the exhaust gas aftertreatment catalytic system.The overall framework of monolith multiscale model is summarized and fundamental principles and chemical processes involved in the modeling were discussed in detail,such as the diffusion of matter and energy inside the reactor,reaction kinetics and flow uniformity.Further,the recent progress about the finite element solution methods for different scale models and machine learning-assisted modeling was reviewed.Finally,we gave perspectives on the challenges and development trends in the field of monolith multiscale modeling were given.

关 键 词：环境催化 monolith催化器 多尺度模型 模拟 

分 类 号：X701[环境科学与工程—环境工程] TP181[自动化与计算机技术—控制理论与控制工程]