Machine learning algorithms for two-phase heat transfer and pressure drop estimation for Joule Thomson Cryocoolers Venkatesh Dasari1, Venkatarathnam G.1,* 1Refrigeration and Air-Conditioning Laboratory, Department of Mechanical Engineering, Indian Institute of Technology, Madras, Chennai, 600036, India *E-mail ID: gvenkat@iitm.ac.in
Online published on 4 December, 2023. Mixed Refrigerant Joule Thomson (MRJT) Cryocoolers operating with Nitrogen Hydrocarbon (N2-HC) mixtures are quite popular for low-temperature applications in the range of 80-230 K. MRJT cryocoolers require very high effectiveness heat exchangers for the system to operate at the design temperature. Estimation of two-phase heat transfer coefficients and two-phase pressure drop is critical for the design of heat exchangers used in MRJT cryocoolers. The application of machine learning algorithms is quite promising in the estimation of two-phase heat transfer coefficients and pressure drop even though they are applicable for a limited data range. In this paper, Artificial Neural Networks (ANNs) are used to regress the two-phase heat transfer and pressure drop for the experimental data available in the literature. A large data set containing more than 10,000 data points are taken from literature and a part of the data is used for training the ANN and another part is used for validation. The results are compared with experimental values as well as existing conventional methods. The mean magnitude of relative error (MMRE) of proposed models with experimental data is presented. Top Keywords ANN, Cryocoolers, Two-phase heat transfer, Pressure drop, Mixed refrigerants. Top |