Study Of The Experiment And Process Simulation On YAG Solid-state Laser Drilling

Laser drilling technology has many advantages, such as widely processing range, no tools loss, highly efficient, remarkably cost-effective, micro hole machining. At present, it has been widely applied in the manufacturing field. But the machining mechanism of laser drilling is a complicated and changeable thermal etching process, and the reasonable process conditions is the security to obtain qualified products. Therefore, based on the research results at home and abroad, this paper analyzed and researched the processing technology of laser drilling from theory, experiment and simulation.Firstly, this paper simply summarized the development and research status of laser drilling at home and abroad, explained the basic principles and characteristics of laser drilling, and analysed the mainly process index and process parameters of laser drilling on YAG solid-state laser.Secondly, the 304 stainless, H62 brass,1060 aluminium alloy with 1 mm thickness and the YAG laser processing equipment were chosed to make laser drilling experiment by copy method, and the tool microscope was used to measure the upper and lower hole diameter. Through the single factor experimental method, discussed the effect of process parameters and process conditions on drilling quality during laser drilling, including laser voltage, pulse frequency, pulse width, focus condition, assistant gas and so on. Using 1 mm stainless steel as research object, designed three factors four level orthogonal experimental L16(43). According to the intuitive and variance analysis of test results, found the influence degree of laser voltage, pulse frequency and pulse width to upper and lower hole diameter and hole taper, and proposed the method of relevant parameters selection at machining different small hole.Finally, aiming at the strong nonlinear processing ability of neural network and strong optimization ability of GA, presented intelligent modeling method combining GA and neural network. With the help of MATLAB neural network and genetic algorithm toolbox, built laser drilling process model based on BP neural network and GA-BP. Then the two models were respectively used to train and predict the hole diameter under different machining parameters. The results showed that GA-BP model have higher fitting precision and better forecasting precision, and the model reflects nonlinear mapping relationship between the laser's parameters and the hole's diameter. It also provide a new way to select process parameters in the process of laser drilling.