Transitioning from Reactive to Proactive Thermal Management

Helix Silicon Carbide Inverter. (Image: Helix)

In an electrical circuit, flowing electrons bump into the atoms in the conducting material and cause those atoms to vibrate. Thermal energy is the total kinetic and potential energy of particles within a region of space, so this transfer of energy from electrons to particle as kinetic energy manifests as heat.

The greater the resistance, the more frequent the bumps and this means greater heat generation. Not only does greater resistance generate more heat, but for metals, this increase in heat generates more resistance.

It’s the job of thermal management to find ways to route this heat away and prevent this feedback loop from significantly reducing electrical efficiency and performance of a system. It’s also key to longevity for permanent magnets, since the particles making up magnetic material will lose their magnetic alignment if they pick up too much kinetic energy — that is, they will demagnetise.

This all makes thermal management critical for an EV’s motor. But along with the basic problem of resistance and heating, EV motors also need to contend with other sources of heat such as the eddy currents within the motor’s stator iron, viscous losses in bearings, and high frequency losses resulting from fast-switching inverters.

Real-Time Intelligence

Along with the feedback loop of temperature and resistance (copper’s resistance increases by 40 percent with every 100 degrees of temperature) there’s also the fact that the electrical insulation around stator windings degrades quickly if it gets too hot. One rule of thumb is that for every 10 degrees of temperature rise, insulation resistance halves, and at some temperature the insulation breaks down entirely.

This makes motor thermal management in EVs a very complex task, and one that needs to be countered early. A single hot spot that’s left undetected can cause an EV’s motor to deteriorate incredibly rapidly, with permanent results.

This means EV thermal management is an intervention that requires real-time intelligence on motor conditions, even where active liquid cooling is deployed, as is commonplace. Along with improving the actual motor performance, and more significantly battery performance, this sort of real-time intelligence can enable innovations in all-vehicle thermal management, heating and cooling — emblemized in technologies like the Tesla Octovalve, which integrates multiple coolant and heating systems to centralize and rationalize an entire vehicle’s heat management.

Coolant channel fluid pressure contour plot. (Image: Helix)

But where exactly does this real-time intelligence come from? Gathering precise heat readings of a whole EV motor is not an easy task, especially at the resolution needed to identify hot spots as quickly and accurately as is required to make them useful. Numerous live temperature readings inside the motor, and of the motor’s coolant are needed, covering the rotor and stator windings.

Direct and Indirect Methods

There are two ways to acquire this temperature information. The first is the direct method, by placing sensors directly in-situ to provide a real-time temperature measurement, i.e., using thermistor or thermocouple sensor in direct contact, or rotor measurements via infra-red sensors. This is the most intuitive way to go about the task, but it comes with some major complications: an increase in complexity, points of failure, and the cost of quality for the sensors themselves.

Helix Scalable Core Technology Motors. (Image: Helix)

Installing and wiring numerous sensors within a motor presents a wiring, packaging, and maintenance challenge that requires significant extra engineering to address and may require trade-offs in efficiency and performance to accommodate. And the failure of just a single sensor can result in false readings that throw thermal management strategies into disarray, resulting in losses of performance, efficiency, and customer dissatisfaction.

This brings up the indirect method: the use of coupled electromagnetic and thermal modelling. By developing a sophisticated model of how a motor behaves in a variety of electrical and thermal conditions, we can pair this with sensors that are already deployed in and around a motor. For example, we can use information from current and position sensors that are otherwise required and must also meet tough functional safety standards, coupled with a single coolant temperature measurement. From these, we can then infer from real-time models what the heat distribution looks like across the system at any point in time.

Electrothermal Modeling

The challenge with relying on electrothermal modeling instead of sensors is that it’s an indirect method. This means that you now depend on the applicability and accuracy of your model, which in turn requires enough testing to develop and refine it, along with enough computational power to run the model and make the necessary temperature inferences.

Stator temperature contour plot. (Image: Helix)

As a result, it’s an approach that relies particularly heavily on the sophistication of the engineering teams behind a motor and a vehicle. Plus, as with any means of measurement, it requires an understanding of its accuracy, and limits setting appropriately.

But the benefits of this approach, if done right, are considerable. There are no trade-offs required to accommodate sensors and their wiring within and around a motor, which means that thermal management data doesn’t need to compromise efficiency or performance.

This approach also reduces points of failure in a motor and an EV. Critically, if a model is sophisticated enough, it can also enjoy significant predictive ability and be well-placed to instruct the thermal management system to fully preempt the formation of hotpots and keep a motor in a continuous steady state of optimal performance.

This shift from reactive to preemptive thermal management is particularly important for improving performance, efficiency, and service life. By maintaining a uniform, stable temperature and minimizing even small momentary disruptions to such a steady state, thermal management systems backed by coupled electrothermal modeling will be key to pushing the envelope for EV motor quality. In fact, rather than being an indirect substitute for a temperature sensor, it may be better to think of temperature sensors as a suboptimal substitute for good electrothermal modeling.

This article was written by Andrew Cross, Chief Innovation Officer, Helix (Milton Keynes, U.K.). For more information, please visit here  .