Multi-objective optimization of beef roasting

S.M. GOÑI; V.O. SALVADORI

Atenas

Congreso; 11 International Congress on Engineering and Food; 2011

Roasting of beef in convective ovens leads to a significant weight loss of the final product. Two mechanisms of water loss can be identified: evaporation and dripping, depending on operative conditions. To achieve a certain degree of doneness, it is commonly accepted that a high oven temperature produces a reduction in the cooking time, however such condition can produce a high weight loss. This situation presents an opposite behavior between both objectives, i.e. the cooking time can not be reduced without increasing the weight loss and vice versa. In this sense, it could be possible to select the oven temperature depending on which objective is more important for the process. The objective of this work was to analyze the behavior of cooking times and weight losses as a function of oven temperature, using a beef roasting model developed and validated in earlier studies. The predicted values were compared with commonly accepted recommendations. The mathematical model considers conductive heat transfer inside the sample and the evaporative weight loss is taken into account at the surface. In order to predict the dripping losses, the model incorporates a simple mass balance which establishes that water content variation is directly proportional to water demand. This value is the difference between the instant water content and the water holding capacity. The geometries used for model simulation were six 3D irregular domains constructed from beef semitendinosus muscle samples; their weights were between 0.49 and 1.08 kg. The model was solved using the finite element method, with constant oven temperature, between 150 and 230ºC. In all cases, the initial temperature and water content of beef were set to 4ºC and 75% (wet basis), respectively. The cooking time and weight loss were computed when the coldest point reached 72ºC. Minimum and maximum predicted cooking times were 49 and 106 minutes, for the smallest sample at the highest oven temperature and the largest sample at the lowest oven temperature, respectively; for such conditions the total weight losses were 39.3 and 21.7%. For all tested samples an opposite behavior between cooking times and weight losses was predicted by the model, i.e. an increase of oven temperature results in a decrease of cooking times (20-26 minutes) and an increase of weight loss (14-22%). The analysis of these results indicates that it is not possible to select an operative condition that has favorable effect on both predicted quantities, i.e. all of them are optimal in the Pareto sense. The trends of both variables predicted by the beef roasting model (cooking time and weight loss) are comparable to published results. Because of the mentioned behavior, a compromise situation is encountered. In consequence, the operator must select the best operative condition in an arbitrary way or from additional information. The last one can be related to energy supply for roasting, and also can be related to quality features as texture and color, among others. Also, such information could be coupled to the roasting model.