EVALUATION OF EMPIRICAL NUSSELT CORRELATIONS FOR NUMERICAL PREDICTION OF FROST THICKNESS GROWTH OVER A COLD CYLINDER

This paper evaluates numerically some of the parameters involved in modeling the process of frost formation over a cold cylinder surface subject to the flow of humid air. The model employs one-dimensional transient formulation based upon the local volume averaging technique. Was utilized for numerical predictions the empirical Nusselt correlation from the literature, obtained in experimental studies on frost formation phenomena of frost grow over a cold cylinder surface. To predict frosting process a numerical solution was utilized, and a new correlation for Nusselt number based on the experimental correlation of Kim was estimated. For the new Nusselt correlation an optimization method that adjusts the numerical solution of modeling the frost formation process with experimental results of the frost layer thickness was used. The calculation procedure allows the estimation of the parameter K of equation. The modeling process was validated by comparison with available experimental data.