The changes in the air velocities caused by the resistance for the air flow due to fruit pulp model packaged (100 grams) and conditioned in multi layers boxes during freezing process were evaluated and air flow rate were estimated using a method of treatment of the experimental data. The air velocities were measured before the air pass through the pillage of multi layer boxes in the air stream. For the measurements processing was used a non linear regression routine. Air flow rate measured by the present method resulted of the numerical integration of air velocities adjusted profile. Results presented a relative difference 10 % higher than the standard average procedure, which consists in averaging the air velocity measurements performed at each point of the section. In the same fan operation conditions, the results shown for the 7 layers arrays of product in the boxes which the air velocity was 62% lower than the 3 layers arrays and 50.9 % lower than the 5 layers arrays of product. These results were proportional to the bulk area for the air flow.

The effective surface heat transfer coefficients during air blast freezing of plastic polyethylene packages containing fruit pulp models conditioned inside multi layer boxes were evaluated under conditions encountered in commercial practice. The results were presented as dimensionless correlations based on hydraulic diameter of the rectangular ducts. The effects on the surface heat transfer coefficient of the air temperature, air velocity, transducer position inside the boxes and box position in the pillage on the surface were analyzed. The results show: the effect of air-cooling temperature on the surface heat transfer is negligible in the range encountered in blast freezing practice. The effective surface heat transfer coefficients predicted by the non-dimensional correlations based on hydraulic diameter was successfully used to represent the bed irregularities, mainly in arrays of 5 and 3 layers. The effective surface heat transfer coefficients varied according to the position between the top and the bottom of the boxes and was influenced by the number of layers in the arrays.