Optimization of microwave-halogen lamp baking of bread

Thesis Type: Postgraduate

Institution Of The Thesis: Orta Doğu Teknik Üniversitesi, Faculty of Engineering, Department of Food Engineering, Turkey

Approval Date: 2004




The main objective of this study was to optimize the processing conditions of breads baked in halogen lamp-microwave combination oven by using response surface methodology. It was also aimed to construct neural network models for the prediction of quality parameters of bread as a function of processing conditions. Different baking time and power combinations were used in order to find the optimum baking conditions of bread in halogen lamp-microwave combination oven. The independent variables were the baking time (4, 4.5, 5, 5.5, and 6 min), power of upper and lower halogen lamps (40, 50, 60, 70, and 80%), and power of the microwave (20, 30, 40, 50, and 60%). As control, breads baked in conventional oven at 200ðC for 13 min were used. The measured quality parameters were the weight loss, color change, specific volume, porosity, and texture profile of the breads. Baking time, upper halogen lamp power, and microwave power were found to be significant on affecting most of the quality parameters. On the other hand, lower halogen lamp power was found to be an insignificant factor for all of the responses. For the optimization process, Response Surface Methodology (RSM) was used. The optimum baking conditions were determined as 5 min of baking time at 70% upper halogen lamp power, 50% lower halogen lamp power, and 20% microwave power. Breads baked at the optimum condition had comparable quality with conventionally baked ones. When halogen lamp-microwave combination oven was used, conventional baking time of breads was reduced by 60%. Artificial neural network models were developed for each of the quality parameters in order to observe the effects of the baking time and different oven conditions on the quality of the breads. High regression coefficients were calculated between the experimental data and predicted values showing that this method is capable in predicting quality