International Review on Modelling and Simulations | 2019

Reynolds Averaged Navier–Stokes Simulations of the Airflow in a Centrifugal Fan Using OpenFOAM

 
 
 

Abstract


The use of numerical models through computational tools allows studying the inner flow of centrifugal fans; this is a normal practice in engineering and scientific communities. When an experimental work is expensive, numerical methods may be used to obtain necessary data. Computational Fluid Dynamics is the name given to computer simulations that enable the visualization and the inspection of flow phenomena in zones that are not easily accessible or measurable, such as blades of the impeller in a turbomachine. Important progresses have been made in Computational Fluid Dynamics over the years: accurate mathematical equations of numerical models, software with more efficient algorithms, and flexible solver codes that reduce the computational effort through user modifications. OpenFOAM is an open-source software developed in C++ library that offers different advantages over commercial computational tools, for instance, over 80 solver applications that can simulate specific cases in fluid mechanics and thermal sciences and engineering. In order to solve the Navier–Stokes equations for incompressible, and turbulent flows in a centrifugal fan with backward inclined blades, the steady-state MRFSimpleFOAM solver has been used. This study has aimed to simulate the flow behavior in the centrifugal fan and to compare the results of three turbulence models (standard k-(, RNG k-(, and realizable k-() with the air flow – pressure curve acquired from experimental measurements using LabVIEW software. A good agreement has been reached between experimental and simulated data. An error rate of 1% between simulated and experimental data has been reached; it is shown that the standard k -epsilon is a semi-empirical model with reasonable accuracy for a wide range of practical applications of flow phenomena.

Volume 12
Pages 230-239
DOI 10.15866/iremos.v12i4.17802
Language English
Journal International Review on Modelling and Simulations

Full Text