Predicting local surface heat transfer coefficients by different turbulent k-ϵ models to simulate heat and moisture transfer during air-blast chilling

A numerical investigation using a computational fluid dynamics (CFD) code is carried out to predict the turbulent flow field, and heat and moisture transfer in a three-dimensional air-blast chiller with cooked meats of cylindrical and elliptical shapes. Three turbulent models [standard, low Reynolds number (LRN) and RNG k-ϵ model] have been used in these simulations. Based on local heat transfer coefficients on the surface of the meat calculated by CFD code, the unsteady heat and mass transfer were simulated which took into account of the effects of conduction within the meat, forced and natural convection, radiation and moisture evaporation on the surface of the cooked meat joint. The model allowed the simultaneous CFD prediction of both temperature distribution and weight loss of the meat throughout the chilling process. Good agreement with experimental results was obtained. The effect of using different models on the accuracy of the simulation of local heat transfer coefficient is presented.