
Russian scientists have found a way to reduce sensor testing time
Scientists from Russia and other countries have proposed a new method for modeling heat transfer in nanofluids, which can significantly reduce the testing time of devices developed for industry, the press service of South Ural State University (SUSU) reported.
As university representatives explained, traditional CFD modeling methods require enormous computing power, with which modeling heat transfer, for example, in a developed industrial heat exchanger can take at least two or four weeks. Innovative models proposed by a team of scientists make it possible to obtain similar data literally in hours and at minimal cost.
“The innovation of the method lies in the fact that it combined CFD methods with machine learning algorithms (GPR, KNN and MLP). It’s like moving from manual calculations to smart forecasting,” TASS quotes the opinion of university specialists.
This development, according to the authors of the idea, can find application in the energy sector, petrochemicals, and climate control systems.