Smart Simulation to Cool Power Electronics

28 February, 2017

6SigmaET Artikelbild Banner

Power electronics developers are working harder than ever before. How can thermal simulation help?

If you’re an engineer working on the development of power electronics, the chances are you’ll have seen your workload double over the past few years. Here’s the good – or bad – news. It’s likely to continue to increase rapidly, if the latest research from BCC Research is to be believed. It shows the market growing from around $11.5 billion in 2014 to nearly $16 billion by 2019.  

You’ll be well aware of the challenges. Not only do your customers want maximum efficiency – they also want maximum reliability and minimum total cost of ownership. That wouldn’t be so hard if it weren’t for the substantially increased power densities you’re being asked to deal with as a result of ever higher levels of integration. Increased densities, of course, mean increased heat – and increased heat means greater chances of unreliable operation and premature component failure. Additionally, despite the high efficiency of power electronics, the enormous amounts of electrical power transmitted in many applications mean very large heat loads must be managed.

Dissipating heat is, perhaps, one of the most significant challenges facing the electronics industry today. The acronym popular with electronics engineers is ‘SWaP’ – size, weight and power (or ‘SWaP-C’ for size, weight, power and cost) – and there are few new products being developed that are not ‘SWaP-constrained’. Electronics subsystems must be smaller and lighter and consume less power – while delivering more performance. SWaP-constrained systems are, by definition, difficult to cool with restricted areas for air to move around and fans to drive air around simply add weight, power and are a mechanical source of potential unreliability.


Innovative cooling strategies

These days, R&D focuses on deploying innovative cooling strategies. A case in point is the work being done by GE in conjunction with DARPA (the US Defense Advanced Research Projects Agency) with technologies such as Thermal Management Technology Bridge, Nano Thermal Interface, Thermal Ground Plane and Dual Cool Jets.

Cooling challenges are only exacerbated by the harsh environments in which increasing numbers of electronic subsystems – and especially power electronics – are deployed. Classic examples of such challenging environments are subject not only to significant changes in temperature but also to shock and vibration – include automotive and other forms of transportation; industrial equipment; military vehicles; and power generation and networks.

When BCC Research released its findings, the company explained the fundamental objectives driving innovation in power electronics are thermal management, form factor efficiency and cost-effective synthesis. This is not too different from the larger electronics industry. However, implications of mismanaging any of these objectives are unquestionably graver for power electronics as compared to the larger electronics domain.


Getting thermal analysis correct

So getting thermal management wrong has potentially significant consequences and can lead to unpredictable or unreliable performance – with associated through-life costs – and also create surfaces that are hazardous to touch; that consume more energy; that cost more; and that are larger and/or heavier than they need to be.  

If you’ve familiar with power electronics already, you will already be familiar with how it’s possible to simulate power electronics circuits using simulation programs such as PSIM and MATLAB/Simulink. Circuits are simulated before they are produced to test how they respond under certain conditions. After all, creating a simulation is both cheaper and faster than developing a prototype to use for testing.

The same principles apply equally to thermal management. In the same way that the operating characteristics of a circuit can be simulated or modelled, so too can its thermal characteristics – using thermal simulation software like 6SigmaET.

Thermal simulation gives a better understanding of the temperature distribution in an electronic system which can be used to fine-tune designs and enable the creation of products that are more reliable. The technology can also significantly cut product development time by reducing design iterations and ensuring that a final product operates within its required thermal parameters, enhancing operating life and reliability.

Abb2. Kühldose Ergebnis.jpg


Significant Cost Savings

Simulation can also enable significant cost savings and return-on-investment for OEMs and developers. It can reduce project delays and negate the need to redesign product enclosures and PCBs. Simulation can prove the differentiator between a successful and timely release to market, and a missed opportunity.

To be made ready for thermal simulation, component or system models are represented as a mesh of grid cells created from intelligent modelling objects, CAD models and PCB layout information. However, there are many ways to generate a suitable grid and its quality and resolution is crucial to achieving an accurate solution.

6SigmaET Release 11 incorporates a new and revolutionary CFD (computational fluid dynamics) solver that removes the burden of having to manually optimize the grid by navigating the trade-off between computing performance and grid detail to provide optimum accuracy and performance. The powerful, parallel solver within Release 11 handles complex geometries and provides accurate and sophisticated simulations – potentially in a matter of minutes.

The latest version of 6SigmaET software also incorporates a user interface which makes it easier to create a thermal model using intelligent objects such as power supplies or heat sinks, and PCB and CAD data. A new object panel in the user interface displays all available intelligent modelling objects and allows them to be dragged and dropped into the model. Powerful search functions make it easier to select the appropriate object to model a given component or subsystem. 

If the BCC forecast is correct, you’re going to be working even harder. Using sophisticated simulation tools like 6SigmaET for thermal modelling represents one way of achieving more and better – but in less time, with less effort, and at lower cost.