Premium
Intelligent modeling of rheological and thermophysical properties of nanoencapsulated PCM slurry
Author(s) -
Hashemi Jirandeh Mohhammad Reza,
Mohammadiun Mohammad,
Mohammadiun Hamid,
Dubaie Mohammad Hosein,
Sadi Meisam
Publication year - 2020
Publication title -
heat transfer
Language(s) - English
Resource type - Journals
eISSN - 2688-4542
pISSN - 2688-4534
DOI - 10.1002/htj.21709
Subject(s) - materials science , slurry , rheology , thermal conductivity , mass fraction , melting point , miniemulsion , phase change material , composite material , chemical engineering , polymerization , thermal , polymer , thermodynamics , engineering , physics
Nanoencapsulated phase change material slurries (NPCMS) combine properties of carried fluid and phase change material (PCM). Usage of NPCMS instead of water as a working fluid has a lot of advantages in many industrial fields. The costly and time‐consuming determination of thermophysical properties of NPCMS through the experimental analysis led the current investigations to use soft computing methods like correlating, artificial neural network (ANN), and ant colony optimization (ACO R ). In this study, the application of ANN, empirical correlations, and ACO R for modeling the thermophysical properties of NPCM slurry, which has been synthesized through a facile and eco‐friendly procedure, has been investigated. PCM nanocapsules have been synthesized using a miniemulsion polymerization method. Nancapsules consist of AP‐25 as core and a Styrene shell, which is modified with graphene oxide nanosheets as an extra protective screen. The morphology and thermal properties of nanocapsules were characterized and analyzed, respectively. Results revealed that minimum average particle‐size values result in a melting latent heat of 146.8 J/g. In case of NPCM slurry, the results showed that the thermal conductivity of MPCS decreased with particle concentration for the temperatures below the melting point. The NPCMS can be considered a Newtonian fluid within the test region (shear rate > 200/seconds and mass fraction < 0.25). The ANN‐ACO R model consists of two neurons in the input layer, six neurons in the hidden layer, and two neurons in the output layer. The input layer consists of two nodes (PCM concentration and temperature) that correspond to parameters found essential and sufficient for thermophysical properties prediction. Upon comparison, the results show that the presented model, which is a combination of the ACO R algorithm and an artificial neural network, is compatible with experimental work.