ELECTRONICS
Fast thermal estimation for
liquid-cooled electrical components
CLIENT
Developer of electrical systems and components, with thermal analysis needs in the design and optimisation of cooling systems
CHALLENGE
Estimate heat exchange inside liquid-cooled electrical components quickly enough to be useful in preliminary design
RESULT
A tool the client runs in-house: reliable temperature estimates, fewer full CFD simulations, faster time to market
01
Designing a liquid-cooled electrical system means knowing the operating temperatures accurately, because performance and reliability both depend on them.
Full CFD answers that question very accurately, and is the wrong instrument for the moment the question is first asked: it is complex to set up, expensive in time and computing resources, and poorly suited to a preliminary design phase where configurations are still changing.
What the client needed was a faster route to a reliable estimate — something usable in the early stages of development and in iterative analysis, where the value is in comparing options rather than in the third decimal place.


02
We delivered software built on proprietary algorithms, which estimates the heat exchange in an electrical component from a handful of input parameters.
The code lets the user define the geometry of the system, set the operating conditions — the coolant flow rate, for instance — and compute the temperature distribution through the circuit, whether its branches run in series or in parallel.
It was designed to be used, not admired: input and output are handled through structured files, so it drops into the client’s existing workflow rather than demanding one of its own.

03
With the software in hand the client can run thermal analysis quickly and systematically: reliable estimates of operating temperature, a rapid comparison of different design configurations, and thermal problems identified while the design is still preliminary.
It also cut the number of full CFD simulations needed — not by replacing them, but by deciding which ones are worth running. The result is a more efficient design process and a shorter time to market.
