Large Scale Computing in Computational Electromagnetics——Past, Present, and Future

Title: Large Scale Computing in Computational Electromagnetics——Past, Present, and Future


Speaker : W. C. CHEW (University of Illinois, Urbana-Champaign, IL, USA)


Date: 10:30-11:30am, December 16, 2013


Place: Conference Room A304, Department of Mathematical Sciences


Abstract: Computational electromagnetics (CEM) research is important for producing simulation software that have been used for virtual prototyping and the design of major electrical and electronic components. Solving electromagnetics problem is a challenging task, especially when the structure is electrically large and involves multi-scale structures. This kind of structures is often encountered in circuits in electronic packaging, small antenna designs, RFID sensor designs, and antennas on complex platforms.  


In this presentation, we will give a brief introduction to the physics of electromagnetic fields. Then we will give an overview of past and recent progress in large scale computing in electromagnetics by our research group, and by other groups around the world. We will give a brief overview the fast multipole algorithm. We will discuss various methods to overcome multi-scale problems such as some domain decomposition method using equivalence principle algorithm (EPA), and some difficulties associated with it. Various methods to stabilize and improve the accuracy of numerical methods will be discussed. We will propose some high-frequency techniques when ray-physics becomes important. If time permits, we can also present the idea of the perfectly matched layer from a coordinate stretching viewpoint.


ContactZhongyi Huang