58-9

Limitation of turbulence models in simulation of heat transfer and flow in air convection food processing

A. SARKAR, B. A. Anderson, and R. P. Singh. Dept. of Biological & Agricultural Engineering, Univ. of California, Davis, 1 Shields Ave., 2030 Bainer Hall, Davis, CA 95616-5294

Time averaged Navier-Stokes equations for turbulent flow and heat transfer have an inherent closure problem. This requires introduction of simplifying turbulence models for obtaining numerical solutions to air convection food processing phenomena. The k-e turbulence model is often used for simulating such processes. The resulting solutions however do not always agree with experimental results. Other than numerical inaccuracies an important reason for such errors is the limitation of the turbulence model itself due to features such as transition, wall roughness characteristics and change in boundary conditions introduced due to the food product in the flow field.

The objective was to evaluate the applicability of numerical solution schemes based on the k-e turbulence models in air convection food processing operations that involve boundary layer flow, stagnation and transition.

Numerical solution for free jet flows, impinging jets with stagnation, and boundary layer flows typical to food processing applications were obtained. The solutions were compared to published data and experiments for boundary layer and stagnation flow over food products. The simulated results differed from predict the experimental results in certain cases which were investigated further using visualization techniques.

The results showed 15 to 30 % variations between stagnation heat transfer coefficient measurements and k-e model predictions. The model predictions failed to predict characteristic variation of heat transfer coefficients due to transition and vortex shedding. The results also showed that for smooth or intermediate boundary layer flows errors of up to 40 % may result in heat transfer coefficient predictions. Wall roughness estimates influence the value of the closure constant in the turbulence model and inaccurate assumptions can cause significant errors.

The results indicate that the assumption of turbulence models for solution of air convection processes requires careful consideration of limitations of the turbulence models and estimate of roughness and transition parameters.

Session 58, Food Engineering: Transport processes and kinetics
2:30 PM - 5:30 PM, Monday PM

2003 IFT Annual Meeting - Chicago,