Estimation of effect of low-frequency vibrations on the fouling mitigation and performance of double-pipe heat exchanger using fuzzy inference system
Savitri Vemireddi, Sewan Das PatlePurpose
The present study develops an artificial intelligence (AI)-based prediction tool to estimate the performance of a double-pipe heat exchanger under low-frequency vibrations. For this purpose, an experimental setup has been developed to generate several data points that can be used to train the AI model.
Design/methodology/approach
This study starts with development of an experimental setup consisting of a double-pipe heat exchanger and a low-frequency vibratory module capable of producing vibrations up to 1,250 kHz. Further, 0.3% CuO nanoparticles are mixed with water as a cold fluid flowing through the inner pipe, at varying low-frequency vibrations. The generated data points are then fed into fuzzy inference system to obtain the relationship among the input and output process parameters.
Findings
It is found that the heat transfer enhancement is raised to 40% under the presence of low-frequency vibrations as compared to the experiments carried out without the presence of vibrations. However, the relation between the presence of low frequencies with respect to the heat exchanger performance is non-uniform. Further, the developed fuzzy-based prediction tool allows for optimisation and found five optimal data points where the heat exchanger operates at the maximum enhancement factor of 1.6252.
Originality/value
The current investigation is extended to the implementation of a fuzzy inference system to obtain the relation between the input parameters, i.e. cold fluid flow rate, hot fluid flow rate and frequency and the output parameter as the performance of a double-pipe heat exchanger. A wide range of experimental datasets at 16-fluid flow combinations are collected to train and validate the prediction model.