Sensitivity Analysis Of Temperature Field And Visualization Of Prediction Model For Selected Laser Melted Titanium Alloy

As an emerging processing method that is currently soaring,additive manufacturing technology is receiving more and more attention in both academic and industrial circles due to its outstanding features of high degree of freedom and high precision.At present,the biggest difficulty that hinders the industrial application of this technology is the difficulty of temperature control,which leads to uneven stress distribution,resulting in the degradation of quality and mechanical properties of molded parts.Therefore,the study of temperature field of selective laser melting has important theoretical significance and practical value for improving the quality of molded parts.In this paper,Ti-6Al-4V is used as the research object,and the temperature field of selective laser melting is mainly studied.A three-dimensional temperature prediction model for a multilayer single channel in a semi-infinite area is established,using the moving point heat source method,considering the effects of convection,radiation,and latent heat,as well as the effects of material properties with temperature.The predicted values are compared with the experimental values,and the results show that the predicted temperature distribution and melt pool size are in good agreement with the experimental results.The prediction model provides a fast and effective method for predicting the temperature of metal additive manufacturing.Based on the validated temperature prediction model for the selective laser melting,a sensitivity analysis was conducted to investigate the factors influencing the temperature field.First,the factors affecting the temperature were weighted and ranked using principal component analysis;then,the melt pool size during processing was selected as the dependent variable indicator to compare the sensitivity of four main processing parameters,namely,laser power,scanning speed,powder layer thickness and ambient temperature,to the melt pool size using Sobol global sensitivity analysis and Morris local sensitivity analysis.The results show that the scanning speed has the greatest effect on the melt pool width,followed by the laser power and powder layer thickness.The ambient temperature has the least effect on the melt pool width.For the melt pool length,laser power has the greatest effect,followed by scanning speed and powder layer thickness,and ambient temperature has the least effect.For the depth of the melt pool,the scanning speed has the greatest influence,followed by the laser power and ambient temperature,and the powder layer thickness has the least influence.In order to apply the above theoretical research results to practical production,this paper develops a visualization system for laser melting temperature prediction in selected areas.Based on the open source Python program,the prediction model is connected in a slot function and the data is cascaded to a database.The predicted temperature field history and distribution can be visualized by entering the selection laser processing parameters and material property parameters in the main operation page.The system is simple and easy to operate,not limited by workers' theoretical level and experience,and can be further extended in the future to provide theoretical basis and practical guidance for the industrial promotion of the theoretical research results of selected-area laser melting.