CFD Analysis and Optimization of Effect of Shroud with Multi-outlets on Airflow Uniformity in a Frost-Free Refrigerator

Authors

1 School of Mechanical Engineering, Southeast University, Nanjing 211189, China

2 School of Mechanical Engineering, Dalian University of Technology, Dalian 116023, China

Abstract

The shroud is a key component of the frost-free refrigerator and its geometric parameters have great influence on the aerodynamic performance of the whole system. Previous researches mainly focused on the effect of other components, such as the fan, shelves, or plate-evaporator. In this paper, the influence of the shroud with multi-outlets on the flow distributions of a frost-free refrigerator is studied thoroughly with the help of Computational Fluid Dynamics (CFD) tools. A 1/2 3-D CFD model is developed, where the verification of turbulence models and mesh independence tests are performed by comparing the mass flow rate obtained by different model configurations. The standard k-epsilon is deemed as the most suitable turbulence model choice and a mesh with Fine level is considered as mesh independence. To obtain the boundaries of the developed CFD model, an airflow velocity test rig is built and constructed. To convert the measured data to CFD model, Structural Response Vector (SRV) method is implemented for velocity profile fitting, and the fitted surface is assigned by User Defined Functions (UDF) macros in simulations. A series of simulations are carried out with the developed model, and the results indicates that no streamline in the middle two cavities of the original freezer compartment and the airflow velocities at the three outlets of the investigated shroud show a certain difference. To optimize the flow distribution, the agent model based on the BP neural network is established, in which four critical parameters of the shroud are adopted as design variables. The results show that the velocity streamlines in the middle two cavities are significantly increased after optimization and the value of the mean square error model constructed in optimization has a reduction of 61.09% compared with the original design.

Keywords