As electric motor performance requirements increase, engineers face a familiar challenge: the losses predicted during design do not always match what happens in operation. 

Pulse width modulation (PWM) is one reason why. 

PWM induced losses can contribute to unexpected hotspots and reduced efficiency, yet they are often addressed through empirical correction factors or considered relatively late in the motor design process. 

At CWIEME Berlin 2026, Dr. Thomas Vyncke, Managing Director and Founder of DRV Solutions, presented an alternative approach based on first principles modelling. The aim is to give engineers a faster way to understand these losses before designs reach later development stages. From Build Factors to Physics -… 

When simulation and reality diverge 

The significance of PWM losses becomes particularly clear when simulation results are compared with physical measurements. 

During the session, Vyncke presented an automotive case study in which there was a 30% discrepancy between simulated and measured losses when PWM effects were not adequately represented. 

For engineers developing high performance electric drives, that gap can have consequences beyond efficiency figures. Unaccounted losses can influence thermal behaviour, cooling requirements and ultimately the decisions made around motor design. 

The challenge is therefore not simply calculating PWM induced losses accurately. Engineers also need a method fast enough to use throughout design optimisation. 

Moving from correction factors to physics 

The methodology presented at CWIEME Berlin combines transient finite element analysis (FEA) with pre calculated datasets and post processing. 

Generic current harmonic calculations are separated from motor specific transient FEA, while frequency domain transformations support faster loss estimation. This allows engineers to assess PWM losses without repeatedly performing computationally intensive simulations for every operating condition. 

The potential time saving is significant. 

According to the session, the approach can support evaluations across full efficiency maps and drive cycles within hours, rather than the far longer timescales associated with applying conventional transient simulation across every operating point. 

Experimental validation also showed close agreement with measurements, although some inaccuracies remained at high modulation indices due to assumptions around harmonic superposition and higher order effects. 

That distinction matters. The objective is not to remove engineering judgement, but to provide a practical modelling approach that makes detailed PWM analysis more usable during development. 

Faster analysis creates more room for optimisation 

The methodology has been integrated into DRV Solutions’ Taurus toolkit, which can interface with established engineering software including Ansys and JMAG. 

That integration points to a wider opportunity for electric motor development. 

When loss calculations become faster, engineers can evaluate more design options. They can investigate the effect of switching strategies, explore cooling requirements and assess performance across operating conditions without treating PWM losses as a final correction. 

It shifts the question from “How much should we add to account for PWM?” towards “What does the physics tell us about this design?” 

Bringing modelling closer to real operating conditions 

As electrification progresses, motor developers are being asked to deliver higher efficiency, performance and reliability while managing increasingly demanding development cycles. 

Better modelling will be an important part of meeting those requirements. 

CWIEME Berlin brings together the motor design, materials, manufacturing and engineering communities working across these challenges. Discussions such as this demonstrate why collaboration between simulation specialists and motor engineers matters: improvements in modelling can directly influence decisions made long before a motor reaches production. 

For PWM induced losses, considering the physics earlier could help engineers identify problems sooner, compare more design options and develop motors that perform closer to expectations when simulation becomes reality.