Dough sheeting model verification by flow visualization

This thesis concerns the rheological characteristics of wheat flour dough that control sheeting. Data and models are lacking in the literature that would allow scale up and fundamental explanation of commonly observed phenomena occurring during industrial sheeting of doughs in making croissants, pastries, pizza crusts and other sheeted products. Much has been learned in polymer science and rheology that can be applied to wheat flour doughs with the proper cautions and boundary conditions. By careful examination of the sheeting process and fundamental rheological measurements of the subject dough, models were evaluated that will be of value to the food processing industry. It was hypothesized that tracers in dough can be used to visualize deformation during a sheeting process and can be useful in validating a dough visco elastic model. There were three major objectives of this work: (1) Develop techniques for incorporating tracers for flow visualization, (2) Measure dough relaxation times using a fundamental method, (3) Comparison of tracer measurements to a visco elastic model of a rolling system. Techniques for injecting tracers into wheat flour dough were developed and can be shown to follow the deformation of the dough during the sheeting operation. An ARES rheometer from Rheometrics was used to conduct step shear experiments to find the dough relaxation times. The resulting generalized Maxwell model was then extended to a non-linear form by separation of variables techniques. Using physical models to estimate the slip boundary conditions, the sheeting process model was integrated and the measured deformations (tracers) were compared to the calculated values. The predicted values were within three standard deviations of the measured values except at the extreme corners of the experimental work that was done. Future work was proposed on extensional measurements and controlling the initial sheeting preparation to reduce modeling and experimental errors to reduce prediction error.