Real-time Quality Research Based On Spectral Diagnostic In Laser Additive Manufacturing

In the process of the laser additive manufacturing laser,the radiation intensity,temperature and the electron density of the plasma have a great influence on transmission efficiency of laser energy,molding quality and structure property.At the same time,defects produced in the process also correspond to the radiation intensity of the plasma.So there will be great importance in analysis of plasma.The findings of this research would be helpful to understand the formative mechanism and influencing factors of laser induced plasma during laser direct metal deposition,and lay foundations for automatic quality control of laser additive manufacturing process.Aiming at the quality control in the laser additive manufacturing process,some defect diagnosis and component identification experiment were conducted in our research.On one hand,the defects produced in the manufacturing process were recognized by spectral analysis and vision sensing system.On the other hand,the functionally graded material parts were produced by direct metal deposition technology.Meanwhile,the composition changes during the process were monitored by spectroscopic diagnostic technology.At first,the forming defect experiments of 316 L steel were conducted.At the same time,the spectral information were collected,selected and analyzed during the process.It was found that spectral intensity and plasma temperature would fluctuate obviously when the parameters changed or the defects occurred.Besides,a coaxial visual sensing system was established.And the defects results from the changes of parameters were diagnosed through monitoring the changes of molten pool area.Further,in order to improve the diagnostic method,Weighted Average Algorithm was used to combine the advantages of both approaches.A more accurate and stable diagnosis result of forming defects was obtained during in single-tract-mufti-layers experiments.Single-tract-single-layer and single-tract-mufti-layers parts were produced in use of 316 L steel and AlSi10 Mg alloy with different gradients.The possibility that detecting component changes by means of monitoring spectral intensity and intensity ratio changes was proved.Besides,the mathematical relationships among spectral information,powder components added and elements of forming parts were obtained,which confirmed the feasibility of ingredient identification by spectral diagnosis.It was also found that the plasma temperature would increased first and decrease then when the additive composition were gradually changed from 316 L steel to AlSi10 Mg alloy and the peak value occurred as the concentration of 316 L steel accounted for 80%~90%.In addition,it was found that there was a obvious corresponding relationship between the intensity ratio changes and mass fraction changes of 316 L steel.Single-tract-mufti-layers parts were manufactured in use of 316 L steel and Inconel718 alloy with different gradients by direct metal deposition technology.The laser additive manufacturing technology was researched and optimized.On the other hand,through the collection,selection of different spectral lines,FeI,NiI,NbI,CrI,TiI,and the calculation of plasma temperature,electron density and intensity ratio,the mathematical relationship between spectral information and the concentration of Inconel718 were obtained.Further,with the aid of Weighted Average Algorithm and BP Artificial Neural Network,the quantitative relationship between the fusion information of various spectral intensity ratios and composition changes was established.More accurate diagnosis and prediction of ingredient changes were realized...