NEURAL NETWORK MODELING OF THERMODINAMIC PROPERTIES OF HIDROFLUYORCARBON BASED REFRIGERANTS

• Grant number: 24/10447-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: September 01, 2024
• End date: August 31, 2025
• Field of knowledge: Engineering - Chemical Engineering
• Principal Investigator: Tiago Dias Martins
• Grantee: Vitor Ravazzi Hidalgo
• Host Institution: Instituto de Ciências Ambientais, Químicas e Farmacêuticas (ICAQF). Universidade Federal de São Paulo (UNIFESP). Campus Diadema. Diadema , SP, Brazil

Abstract

Thermodynamic parameters are usually hard to obtain. Therefore, numerical methodsbecome valuable tools for this task. This study aims to obtain one artificial neural network(ANN) to calculate four thermodynamic properties for the modeling of the refrigerationprocesses: thermal conductivity, viscosity, heat capacity and density. Experimental data fromfive different hydrofluorcarbon refrigerants in the liquid phase, available in the literature, willbe used for training the ANNs. There will be used as the input data: critical temperature,critical pressure, operational temperature and pressure, as well as the acentric factor and molarmass of each individual refrigerant. ANNs with one and two hidden layers will be trained.Different optimization methods, combinations of activation functions and neurons numberwill be considered. Since ANNs are efficient and fast to obtain results, this paper aims atreducing the time to calculate the parameters of the refrigerant fluids, granting more agility tothe productive process, as well as reducing gaps in production and increasing productivity.With that, this study also aims to contribute within the national and international scope.