Hey everyone, Sarah here from [our company name]—we’ve been making switched reluctance motors (SRMs) for 12 years now, and let’s be real, if you’ve ever dabbled in motor design, you know that testing a new SRM prototype by building it first is not just expensive—it’s a total headache. Last year, a customer of ours came to us with a wild idea: they wanted to tweak an SRM for a forklift application, but they couldn’t afford to waste time and parts guessing if their winding changes would work. That’s when simulation saved the day. Today I’m breaking down exactly what simulation methods we use (and recommend) for SRMs, since that’s what we live and breathe here. Switched Reluctance Motor
First, let’s get one thing straight: SRMs are weirdly unique compared to, say, permanent magnet synchronous motors (PMSMs). No magnets on the rotor—just salient poles that switch reluctance as they align with the stator’s energized poles. That switching action causes torque ripple, noise, and vibration, right? So their simulation needs to cover way more than just basic torque output. When we’re developing our own motors, we start with two core simulation buckets: electromagnetic (EM) simulation and mechanical/thermal simulation. But there’s a third bucket we can’t sleep on—control system simulation—because bad control equals bad SRM performance, period.
Let’s kick off with electromagnetic simulation, because that’s the foundation of everything. For most small to mid SRMs (think the 0.5kW to 50kW range we make for industrial and EV aux apps), we use finite element analysis (FEA) tools like ANSYS Maxwell or Motor-CAD’s EM module. Wait, I know what some of you are thinking: FEA is slow, right? If you run a full transient FEA model for every single operating point, you’re waiting hours. But here’s the trick we’ve refined over the years: we don’t run full FEA for every scenario. We start with a 2D FEA model first—way faster, and accurate enough for most radial-flux SRMs, which is what we build. We map the flux linkage, torque, and inductance across every rotor position and every phase current. That gives us a lookup table (LUT) we can feed into the rest of our models.
But when we’re dealing with larger SRMs—like the 200kW units we do for mining equipment or heavy EVs—we move to 3D FEA because edge effects and fringing flux get way more significant there. 3D is slower, but it’s non-negotiable for high-power motors where even a 5% flux miscalculation can lead to overheating or lost torque. Also, if someone’s tweaking the rotor pole shape or stator tooth notch to cut torque ripple, 2D might not catch those subtle 3D effects, so we pull it out for those deep dives. A lot of new SRM developers skip 3D, but trust me—we learned that lesson the hard way. Early on, we built a 100kW motor using only 2D FEA, and when we tested the prototype, it was 8% less efficient than simulation predicted. Turned out the end windings were causing extra eddy currents that 2D didn’t pick up. Oops.
Next up, mechanical and thermal simulation—this is where a lot of people drop the ball with SRMs. Remember that torque ripple I mentioned earlier? It creates massive vibration and stress on the rotor and stator laminations. So we use FEA for structural mechanics too, along with thermal simulation that couples directly to the EM model. Tools like COMSOL or Motor-CAD’s thermal module work here. The cool thing about coupling EM and thermal is that we can see how the heat from phase windings and core losses changes the motor’s resistance, which in turn changes the current and losses. It’s a loop. For example, if we simulate a motor running at 100% load for an hour, we can see if the stator winding insulation will hit its temperature limit—something we can’t guess from just EM results. We also run modal analysis in mechanical simulation to make sure the motor’s natural frequencies don’t line up with the excitation frequencies from the switching SRM currents. If they do, you get noise so bad it’s unbearable, and parts will fatigue and fail prematurely. We had a customer a couple years ago whose new SRM for a conveyor made so much noise it triggered OSHA alerts before they even finished testing. Turns out their control switching frequency was matching the motor’s natural frequency. We fixed it by tweaking the control’s switching parameters in simulation first, so they didn’t have to rebuild the motor. That’s the power of this stuff.
Then there’s control system simulation, which is what ties all the EM and mechanical models together. SRMs don’t operate on their own—their performance depends entirely on how you switch the phases on and off (that’s called the commutation strategy). A bad commutation can kill torque, add noise, or even damage the inverter. We use tools like MATLAB/Simulink here, where we plug in the LUTs from our EM FEA and build a virtual control system. We test different commutation angles, current control methods (like hysteresis control or predictive control), and even fault scenarios—like if one phase fails, how does the motor keep running? A few months back, a team working on an electric bus wanted their SRM to keep operating even if a phase shorted. We simulated that in Simulink first, found that adjusting the remaining phases’ current limits and commutation timing could add 20% extra run time, so they didn’t have to oversize the motor just for fault tolerance. That saved them thousands in prototype costs.
Wait, there’s also a method that’s super useful for early-stage design when you don’t have the specs for a full FEA model—analytical simulation. That’s the old-school math-based stuff, using equations for inductance, torque, and losses. Tools like MATLAB have built-in SRM analytical models, and Motor-CAD has a lumped parameter thermal network (LPTN) that’s analytical. It’s way faster than FEA, like a run in minutes instead of hours, but less accurate. We use analytical simulation at the very beginning of a project when our engineers are brainstorming specs—like “we need a 10kW motor with 90% efficiency for a robotic arm.” The analytical model gives us a rough idea of the size, winding gauge, and approximate torque, so we can narrow down which designs are worth investing time in for FEA. It’s a great starting point, but we never rely on it for final validation. Analytical is to FEA what a roadmap is to a GPS—good for direction, but you need the GPS for exact turns.
Now, let’s talk about the mistakes we see people make with SRM simulation, because that’s what we’ve learned from working with hundreds of clients over the years. First, a lot of folks treat simulation as a “set it and forget it” step. They run one model once and call it good, but SRM performance changes so much with speed and load. You need to run simulation across the entire operating envelope—from idle to peak torque, low speed to high speed. Second, they skip mechanical coupling. We already touched on torque ripple, but don’t forget bearing stress, rotor balance, and even the motor’s mounting structure. Third, they don’t validate their simulation with real prototype data. Like, run a small test on a prototype to measure inductance or torque at one point, then adjust your model to match that. Your simulation will only be as good as the real-world data you use to tune it. We do that with every single new motor we build—we have a dynamometer lab where we test prototypes, and we use that data to tweak our FEA models until they’re within 2-3% of real results. That’s how we keep our motors consistent.
At [our company], all this simulation stuff isn’t just a fancy process for our engineers—it’s how we make better motors for our customers, at a lower cost and faster timeline. We don’t just sell off-the-shelf SRMs; we work with companies to design custom motors too, and simulation is the backbone of that customization. If you’re tired of prototype after prototype failing, or you want to optimize your existing SRM application to be more efficient, quieter, or more reliable, we’re here to help. We can walk you through what simulation methods make sense for your project, share the data we’ve gathered over a decade of building and testing SRMs, and even help you model your specific application to make sure it works before you spend a dime on parts.
If you’re ready to stop guessing and start building SRMs that perform exactly how you need, reach out to our team to chat about your project. Whether you’re working on industrial machinery, electric vehicles, robotics, or any other application that needs a durable, high-torque motor, we’ve got the simulation expertise and the manufacturing experience to get you there. Don’t waste another prototype—let’s connect and make your next SRM project a success.
Brushless Motor References
- Miller, T.J.E. Switched Reluctance Motors and Their Control. Oxford University Press, 1993.
- Zhu, Z.Q., Howe, D. “Electromagnetic Performance of Switched Reluctance Motors.” IEEE Transactions on Magnetics, vol. 33, no. 5, 1997, pp. 3298–3300.
- Hanselman, D.C. Brushless Permanent-Magnet and Reluctance Motor Drives. Oxford University Press, 1994.
- ANSYS Inc. “ANSYS Maxwell for Switched Reluctance Motor Design.” Technical Report, 2022.
Zibo Auric Mechanical and Electrical Technology Co., Ltd.
As one of the leading switched reluctance motor manufacturers and suppliers in China, we warmly welcome you to buy advanced switched reluctance motor for sale here from our factory. All customized motors are with high quality and competitive price.
Address: B419, High-tech Entrepreneurship Park, High-tech Zone, Zibo City
E-mail: cui@auricmotor.com
WebSite: https://www.auricmotor.com/
