From Multi-Scale Methods to Virtual Process Engineering --- Efficient and Accurate Simulation of Multi-Phase Complex Systems

Dynamic simulation of the equipments and reactors in process engineering typically requires very long time and enormous computational cost. The major challenge lies in the multi-scale nature of the structures and behaviors of these systems. The structural consistency between the system simulated, the algorithm employed and the computing hardware used may provide an effective way to speedup the computation and to improve the accuracy, which can be well implemented in a multi-scale discrete computing mode with stability constraints. This presentation will give an overview of the work at our Institute in this direction, focusing on its development and extension in a variety of complex multi-phase systems, and establishment of the corresponding supercomputing software and hardware. The possibility of achieving real-time simulation, that is, virtual process engineering in this computing mode is prospected finally.