简介: |
Abstract：
In this presentation, we will demonstrate that heterogeneous computing with processors of different architectures used concurrently for different algorithms may give both good scalability and efficiency for large-scale simulation of complex flow and transport processes. Meso-scale models are introduced to consider the sub-grid-scale structures in continuum methods and coarse-grained discrete methods. Highly parallel codes for continuum-based finite volume method, discrete element method and lattice Boltzmann method are implemented on the central processing units (CPUs), many-integrated cores (MICs) and graphics processing units (GPUs), respectively. Quasi-realtime simulations of lab-scale systems with particle-scale resolution are thus obtained and virtual-reality (VR) style simulations are under development. Some preliminary results can be found in our recent publications [1-6] and the most recent developments will be introduced in this presentation.
1. Bo L, Zhou G, Ge W, et al. 2014. Particuology 11:160-169
3. Ge W, Lu L, Liu S, et al. 2015. Chem. Eng. Tech. 38:575-584.
4. Lu L, Xu J, Ge W, et al. 2016. Chem. Eng. Sci. 155:314-337.
5. Zhang C, Shen G, Li C, et al. 2016. Mol. Simul. 42 (14): 1171-1182.
6. Ge W, Wang L, Xu J, et al. 2017. Rev. Chem. Eng. DOI 10.1515/revce-2015-0079 (Online).
Bio:
葛蔚：男，1970年3月生，浙江杭州人。1992和1998年先后在哈尔滨工业大学热能工程专业获工学学士和博士学位。现任多相复杂系统国家重点实验室主任；中国科学院特聘核心骨干研究员，科技创新交叉合作团队负责人；中国科学院过程工程研究所介尺度科学研究部主任，学术委员会、学位委员会委员；中国科学院大学岗位教授、博导；天津大学外聘博导、天津化学化工协同创新中心成员；中国化工学会模拟与仿真专业委员会主任委员，中国颗粒学会理事；《过程工程学报》副主编，Chemical Engineering Science、《化学反应工程与工艺》、《计算机与应用化学》、《数值计算与计算机应用》编委，Particuology顾问。曾获国家自然科学基金委杰出青年基金，周光召基金会首届“杰出青年基础科学奖”、中国颗粒学会“宝洁青年颗粒学奖”、中国科学院首届超级计算应用奖及中国科学院“先进工作者”称号。主要研究对象为气固、颗粒与散料、多孔介质及纳微多相系统，内容包括流动、传递、反应及其耦合过程的介尺度机理与多尺度建模、分子模拟及离散-连续耦合模拟、高性能计算及其工业应用，致力于实现虚拟过程工程(Virtual Process Engineering)。 |
