Abstract
In the laser cleaving of brittle materials, using a controlled fracture technique, thermal stresses are used to induce the crack and the material is separated along the cutting path by extending the crack. One of the problems in laser cutting of glass with this technique is the cut path deviation at the leading and the trailing edges of the glass sheet. Previous work has shown this deviation to be partly due to the high magnitudes of thermal stresses generated near the edges of the sheet. This paper compares the performance of an artificial neural network (ANN) and finite element (FE) model to predict thermal stresses at the leading and trailing edges of the glass sheet for varying thickness and laser cutting speed. The mathematical model for thermal stresses was expressed as a function of the temperature, glass thickness and time using ANN. From the experimental data sets for glass thickness of 3mm, 5mm and 10mm the ANN model was trained. The testing accuracy was then verified with extra experimental data sets for 8mm glass thickness. Statistically assessed as adequate, the ANN model was then used to investigate the effect of the laser cutting speed for the varying glass thickness to obtain stress fields at the leading and trailing edge of the glass. FE modeling of the cutting process has also been used to simulate the transient effects of the moving beam and predict thermal fields and stress distribution. These predictions are validated against the experimental data and it was revealed that predicted results from ANN are better than the FE model.