Cooling via AI

Look at the internals of an ordinary radiator or cooling jacket and you’ll probably see rows of tubes or straight channels. But that’s not what you’ll see inside a jacket designed by Diabati in Belgium. The firm uses what it calls generative thermal design run by an artificial intelligence program to optimize the flows through heat exchangers. The result usually looks like the ant-hill-like structure visible on the screen behind Diabati CEO Lieven Vervecken, or the example at the bottom right. In one case, Diabati used its program to optimize the water jacket around the hub motor on a U. of Belfast student race car, bottom left. The result was a motor operating 20% cooler than before. Vervecken holds a plastic version of the water jacket his AI program designed.
NEXT PAGE: Motor controllers double as dynamometers