Researcher Wins NSF Award for Electric Machine Modelling
Electric machines are in the cars that carry us, the planes poised to transform air travel and the turbines generating renewable energy. At the heart of these machines lies a design challenge that engineers have wrestled with for decades: how to balance performance, efficiency and cost while accounting for the complex electromagnetic, structural and thermal behaviours of motors and generators.
Traditionally, the answer has been finite element analysis (FEA), a computational method capable of simulating nearly every nuance of machine performance. But while FEA is reliable, it comes with a drawback: It can often take a long time. For researchers and companies alike, this means fewer design variations can be tested, fewer optimizations can be explored and innovation often gets bottlenecked at the simulation stage.
Now, researchers at The University of Texas at Dallas are offering a solution. A team in the Department of Electrical and Computer Engineering has developed a parameterized nonlinear magnetic equivalent circuit (MEC) model that achieves the speed engineers crave without sacrificing the accuracy they depend on. Their findings, presented at the IEEE International Electric Machines & Drives Conference (IEMDC) 2025, show that the approach can run simulations 21 times faster than FEA, while keeping prediction errors remarkably low.
Transforming Electric Machine Design
This work is part of a larger research vision led by Dr. Matthew Gardner, assistant professor of electrical and computer engineering in the Erik Jonsson School of Engineering and Computer Science, who was recently awarded the prestigious National Science Foundation CAREER Award. The CAREER Award, one of the NSF’s highest honors for early-career faculty, recognizes both groundbreaking research and commitment to education.
“[Gardner] conducts groundbreaking research in electric powertrains, focusing on the development of smarter and more efficient systems that enhance renewable energy and electric mobility,” said Dinesh Bhatia MS’87, PhD’90, department head and professor of electrical and computer engineering. “His research significantly contributes to various aspects of Power and Energy Systems, positioning the ECE department among the leading groups in this field.”
“We’re opening the door to designs that were previously out of reach due to computational limits.”
Dr. Matthew Gardner, assistant professor of electrical and computer engineering
Specifically, Gardner’s $500,945 award supports his research on High-Resolution Lumped Parameter Network Multiphysics Models for Electric Machine Topology Optimization. His focus is on enabling electromagnetic, structural and thermal optimization of electric motors simultaneously. These models, developed in collaboration with colleagues and students, provide a bridge between the possibilities of additive manufacturing and the demands of next-generation applications such as electric aircraft, renewable energy and robotics.
“This recognition validates the importance of making electric machine design not just more efficient, but transformative,” Gardner said. “We’re opening the door to designs that were previously out of reach due to computational limits.”
Breaking the Bottleneck
The team’s published paper was co-authored by Gardner, electrical engineering doctoral students Danial Kazemikia and Salek Khan, and Manuel De Jesus Contreras BS’25. It focuses on a workhorse of modern engineering: the surface permanent magnet synchronous machine (PMSM). PMSMs are widely used in electric vehicles and other high-performance systems where efficiency, torque density and reliability are paramount.
The group’s high-resolution, parameterized MEC approach discretizes the PMSM into node cells, each representing reluctances (magnetic resistance) and magnetomotive force sources. By dynamically adjusting the number of angular and radial layers based on machine geometry, the model captures detail where it’s needed, without wasting computation where it’s not.
“MECs offer precise magnetic analysis with drastically reduced computational cost, enabling complex optimization and innovation in electric machine design that were once computationally impossible,” Kazemikia said.
In design terms, engineers can now explore dozens or even hundreds more variations within the same timeframe it would take FEA to analyze only a handful.To validate their model, the team conducted a parametric sweep across multiple PMSM designs, then refined the chosen design with practical modifications. These included fillets on magnets and tooth tips and a step skew rotor to reduce torque ripple.
The refined design was checked against FEA and then fabricated into a prototype. Testing showed that the prototype’s measured back electromotive force (back EMF) closely matched the predictions, confirming the MEC model’s real-world reliability.
Bridging Theory and Practice
For industries pushing the limits of electric machines, design bottlenecks can mean the difference between progress and stagnation. Aerospace is a prime example. The promise of electric aircraft and air taxis hinges on developing machines that are lighter, more efficient and capable of delivering high power in compact packages.
CAREER Awards
Dr. Matthew Gardner was among five UT Dallas faculty members who received 2025 Faculty Early Career Development Program (CAREER) awards from the National Science Foundation. These five-year grants will support research to design innovative electric machines, improve robotic-system security, create temperature sensations for virtual reality users, advance digital-storage technology and synthesize new drugs.
Similarly, the automotive sector relies on motors that can deliver higher torque per unit weight while maintaining reliability. Renewable energy systems, such as wind turbines, also stand to gain from faster design cycles that make scaling and deployment more cost-effective. The CAREER Award emphasizes not just research, but education. As part of his project, Gardner is developing teaching modules that bring computational modeling into engineering coursework. Students will be able to run simulations, debug code and check results. This will bridge the gap between classroom theory and real-world engineering.
Undergraduate and graduate students will contribute by coding models and building prototypes. High school students will also participate through UT Dallas’ STEM Bridge program, gaining early exposure to advanced engineering challenges.
Gardner’s innovative approach has the potential to fundamentally change the way electric machines are designed and developed, said Dr. Babak Fahimi, professor of electrical engineering, Distinguished Chair in Engineering and founding director of the Renewable Energy and Vehicular Technology (REVT) Lab.
“Dr. Gardner’s approach in modeling and design is very suitable for teaching electric machines and power electronic converters to undergraduate and graduate students and has the potential for providing new design software that can be used by many industries,” Fahimi said. “I am very impressed by his accomplishments to date and consider him a leading researcher who is destined to create seminal contributions for many years to come.”
Powering the Future
The researchers anticipate that their work will evolve into a comprehensive platform for topology optimization, where electromagnetic, structural and thermal functions can be optimized simultaneously. With additive manufacturing making complex geometries possible, tools like these will be crucial for pushing the boundaries of machine design.
The potential applications are wide-ranging. In aerospace, lighter and more efficient motors could enable practical electric vertical take-off and landing (eVTOL) aircraft. In renewable energy, optimized generators could boost efficiency and reduce costs. Even robotics and medical devices stand to gain from motors that are smaller, cooler and more precise.
For Gardner and his collaborators, the recognition from NSF and the promising results presented at IEMDC 2025 are not the endpoints. Ongoing work aims to expand the models, integrate more physical domains, and refine optimization strategies.
“This research is about more than just faster simulations,” Gardner said. “It’s about enabling the next generation of electric machines, the ones that will power the future.”