Alfonso William Mauro

Department of Industrial Engineering, 

Federico II University of Naples, Naples, Italy

Alfonso William Mauro is associate professor at the Department of Industrial Engineering of Federico II University of Naples. Educational and research activities deal mainly with the increase of the efficiency in energy conversion related to the refrigeration and air-conditioning fields. In particular, he is leading the refrigeration lab, promoting researches at basic and applied level. At basic level the main topics are related to the optimal design of high performance evaporators and heat sinks; at a system level the focus is related to the introduction of new refrigerants and technologies for heat pumps and refrigeration systems. In the last five years the main keywords of the research are: low GWP refrigerants, carbon dioxide, propane, flow boiling, two-phase ejectors,  heat driven hybrid refrigeration systems. 

He is author of 83 papers indexed by Scopus. Vice-President of the B2 Commission of the International Institute of Refrigeration. Member of several associations in the field of refrigeration and air-conditioning at national and international level. 

Title: PHYSIC BASED MODELS VS ARTIFICIAL INTELLIGENCE TOOLS: PERFORMANCE AND LIMITS IN PREDICTIONS OF PRESSURE GRADIENTS IN TWO-PHASE FLOWS

Abstract: Prediction of pressure gradients of two-phase flows is of primary importance for sizing and simulations of heat exchangers and piping used in a number of applications: refrigeration and air-conditioning systems, nuclear plants, chemical and process industries.

Several approaches are available: numerical resolution of fundamental equations (continuity, momentum, energy) via computation, analytical correlations for the frictional term of the momentum equation, artificial intelligence (AI) tools. Each of them has its pros and cons depending on the scope of the calculation and on the number of calculations required.

In this presentation the focus is on the comparison of the analytical correlations with AI tools. In particular, a large dataset of data is used to assess the prediction capability of several quoted method from literature while a specific tool based on AI is introduced. The results of the case study are used to drive some general conclusions about accuracy of calculations, robustness and time consumption of each class of methods.