6/17/2014 Jonathan Damery, ECE ILLINOIS
Written by Jonathan Damery, ECE ILLINOIS
Innovative research happens at the crossroads of talent and timing. For Professor Jianming Jin, who recently received the ACES Technical Achievement Award from the Applied Computational Electromagnetics Society, one of these crossroads was encountered in the late 1980s, as a doctoral student working on a new electromagnetics modeling method at the University of Michigan. It was an innovative project, begun at a turning point in electromagnetics — the ideal timing.
For the past several decades, researchers had been developing and refining models for electromagnetic signals — primarily the moment method and the FDTD method
“Stealth airplanes are always coated with multilayer, absorbent materials. That’s not metallic. That’s not homogenous,” Jin said, explaining one failing of the methods.
To account for those faults, researchers, in the early 1980s, began devoting attention to an approach known as the finite element method. It was first developed for the aerospace industry in the 1950s where it was used for structural analyses, but it seemed to hold promise for complex electromagnetics. Translating the method between structural and electromagnetic applications, however, proved to be difficult. An airplane occupies a defined, physical space, whereas radar signals bouncing off of the airplane’s fuselage are vast and nebulous. There are no boundaries.
As a doctoral student, entering at this key juncture, Jin developed an approach known as the finite-element boundary-integral method — essentially a fusion of the finite element method and the moment method. This hybrid approach deals with the problem of undefined space by delimitating a fixed sample and using the sample to model the whole.
“At that time, I thought I had solved a nice problem, but I didn’t know ... that the method becomes so powerful,” Jin said. “Right now, the finite-element boundary-integral method is probably the most popular and well-established method to apply finite element method to scattering and antenna and any open region problems.”
Jin joined the ECE ILLINOIS faculty in 1993 and continued to refine the finite-element boundary-integral method for more than a decade. That same year, his first book was published: The Finite Element Method in Electromagnetics (Wiley). It was republished in its third edition this year and is widely regarded as the foremost text on the subject.
Jin’s major contributions to the finite element method, which are cited collectively as the impetus for the ACES Technical Achievement Award, are not confined to the development of the hybrid approach. In addition to demonstrating novel applications, including for MRI technology, Jin and his team of graduate students and postdocs have demonstrated two further breakthroughs, each encompassing more than 10 years of stepwise discoveries.
As the finite element method was first gaining traction for electromagnetics in the 1980s, most of the research focused on the frequency domain, meaning that the models simulated signals at a single frequency, without factoring in changes over time. In the late 1990s, however, Jin and his students turned their attention to modeling signal propagation in the time domain. This is particularly effective for radar scattering and nonlinear devices like diodes and transistors, which have variable electromagnetic properties.
“My students overcame a lot of difficulties that were specifically associated with time-domain analysis and eventually developed the method to a state that it can be used for almost all the problems that a frequency-domain analysis can deal with,” Jin said.
Although their time-domain research is ongoing, with recent papers on nonlinear devices submitted to top journals, their earlier work was incorporated into Jin’s 2008 book Finite Element Analysis of Antennas and Arrays (Wiley). That work paved the way for the military’s predominant use of the finite element method for antenna and radar analyses.
Jin’s team also made key contributions to the development of domain decomposition. This is an approach that breaks a massive and complex problem into manageable subunits for analysis, enabling rapid parallel computation when the subunits are divided and analyzed on different processing units.
“Computation time for engineering applications is very important,” Jin said. “If you want to do a design, and you have to wait for 10 hours ... it would be very difficult. ... But if you have something that you put in, and within a couple of minutes, you know the results, then you can design much more effectively.”
The Applied Computational Electromagnetics Society recognizes Jin for these “numerous contributions” to the finite element method, and Jin, in turn, points to his lab members, who have provided over two decades of insights and indefatigable research. “This award is really recognition of the contributions by my former students and former postdocs,” Jin said. “I feel happy in that sense.”