| Slurry blasting technology can not only improve the roughness of the metal surface but also strengthen the metal surface,and has received widespread attention in recent years.However,most of the current research is limited to the exchange regulation of matrix surface roughness after affect of the parameters of the slurry blasting procedure,and the control of the surface roughness of the metal matrix after the impact of the slurry blasting impact depends more on manual experience,and there is no accurate control model,which seriously affects the promotion and application of the slurry blasting technology.Therefore,304 stainless steel was used as the research object,through theoretical analysis,simulation and experimental research,the influence of pulping process parameters on surface roughness was studied,and combined with the BP neural network model,the BP neural network prediction model of the pulping process was established.The main research work is as follows:(1)By analyzing the impact motion process of the slurry,the ejection velocity and velocity attenuation rate model of the slurry were established;Based on the conservation of energy theory,the impact force of the slurry on the surface of the matrix is analyzed;With the help of the Vickers hardness calculation formula,The relationship between have an effect on pressure and plastic deformation of matrix is studied.Then,the relationship between the plastic deformation depth of the matrix and the abrasive particle size,ejection speed and ejection angle was analyzed This paper provides theoretical support for the influence of slurry blasting process on the surface roughness of the matrix.(2)The ANSYS-AUTODYN was used to simulate the effects of different abrasive particle size,ejection velocity and ejection angle on the surface impact of the substrate.It was found that when the ejection speed is 35m/s,the ejection angle is 90°,and the abrasive particle size is 1.2mm,the maximum impact temperature at the moment of contact is only 95°C,which will not affect the deformation of the matrix,indicating that the slurry blasting can take away the impact heat and prevent the thermal deformation of the matrix.It is determined that the have an impact on range of the substrate surface will increase with the expand of abrasive particle size and ejection velocity after impact,and the vogue of first growing and then reducing with the make bigger of the ejection angle;The impact depth is proportional to the abrasive particle size,ejection speed and ejection angle.It is proved that the surface stress of the metal matrix is positively correlated with the abrasive particle size,ejection velocity and ejection angle.(3)Through the grouting have an impact on test,the affect of abrasive particle size,ejection speed and ejection angle on the surface roughness and surface residual stress of the metal matrix after slurry blasting was obtained.At the same time,the interaction law of the two process parameters on the surface roughness of the matrix was analyzed,and the different influences of the process parameters on the surface roughness of the substrate were found,which provided a basis for the subsequent roughness prediction model.The finite element simulation and test results were compared,and the maximum error of surface stress was 3.6%,indicating that the finite element simulated slurry impact results were feasible.(4)Using the test results of abrasive particle size,ejection speed and ejection angle as process parameters,the BP neural network training method was selected,and the BP neural network prediction model of the slurry blasting process parameters was established,and the maximum error between the roughness test result and the target roughness was 2.4%.Through the above research,the control model of the surface roughness of the metal matrix after the impact of the slurry blasting can predict the roughness of the surface of the metal substrate under different impact process parameters,which provides certain reference significance for the development of the slurry blasting process technology. |