Optimization And Prediction Of Low Phosphorus Electroless Nickel Plating On Cr12MoV Die Steel

There is a large demand for die steel,which in mould,automobile,mechanical equipment manufacturing industry.Factories generally plate thick chromium on the surface of products to enhance the hardness and wear resistance of the substrate.Studies have shown that lowphosphorus electroless nickel plating has a significant effect on improving the surface hardness of die steel.Generally,the selection of electroless nickel coating with appropriate phosphorus content depends on the working condition.However,the method of determining the best process parameters through extensive testing is inefficient and cannot quickly predict the phosphorus content and performance of the coating.In this paper,Cr12MoV die steel is taken as the research object,the high-pressure steam degreasing method and the lowphosphorus chemical nickel plating alloy coating are studied.On the basis of obtaining the process parameters and performance data of Ni-P coating with different phosphorus content,the neural network optimized by particle swarm optimization algorithm was used to establish the process optimization model of low phosphorus electroless nickel plating.1.The Cr12MoV die steel was rinsed in the high-pressure steam cleaner for 10 min,and then put in a degreasing agent at a ratio of 1:20 and ultrasonically cleaned for 5 minutes.The process method has effectively removed the oil film adhering to the steel surface.The detected COD value of the treated oil wastewater is around 65mg/L and the pH is 10.1.For the treatment of oily wastewater produced by this process,the solution proposed in this paper can reduce the pollutant discharge and reasonably recycle the water resources in the process of electroless nickel plating.2.The influencing factors of deposition rate and phosphorus content of the coating alloy were studied by surface fitting.The results showed that the deposition rate was mainly affected by temperature and pH,and the highest plating rate reached 10.54 ?m/h.The phosphorus content is mainly affected by the concentration of the reducing agent of the reactant and the pH value,and the lowest phosphorus content of the coating is 3.01 wt.%.In this paper,through SEM surface morphology observation,it was found that the cracks and voids of the lowphosphorus coating are more than that of the medium and high-phosphorus coating.According to the XRD diffraction pattern,when the temperature reaches 450 ?,most of the amorphous components of the coating have turned into crystals.At this time,the hardness of the high phosphorus coating has exceeded hardness of the low phosphorus coating,reaching the maximum value of 1022 HV.Therefore,the best heat treatment condition selected in this experiment is to keep the steel parts under nitrogen protection at 450 ? for 1h.3.The multiple linear regression model of phosphorus content and plating rate was established by MATLAB.According to the model fitting results,the average relative errors of phosphorus content and plating rate prediction are 19.2% and 9%,respectively.A neural network prediction model was adopted.The research shows that the prediction results of a single neural network for two performance indicators were unsatisfactory.Therefore,we report a brand-new variance analysis and that was used to determine the main influencing factors of phosphorus content and plating rate.Temperature,pH and reductant were used as new neural network inputs.The final results of the relative errors of the phosphorus content and the plating rate are 16.58% and 9.64%,respectively.By adding optimization algorithmcross validation,genetic algorithm,particle swarm optimization,we further completed the optimization of neural network model.According to this comparison,the neural network optimized by the particle swarm optimization algorithm has significantly improved the fitting ability of the experimental results of electroless nickel plating.The average relative errors of the predicted phosphorus content and plating rate are 7.13% and 5.45%,respectively.The prediction errors of its best state are 4.3% and 3.53% respectively,and the prediction results are basically consistent with the experimental values.By the neural network of particle swarm optimization,the new experimental group was predicted,and the best process parameters of low phosphorus electroless nickel plating were obtained: temperature 78 ?,pH=6.2,concentration of lactate 16 g/L,concentration of sodium hypophosphite 16 g/L.In addition,the best parameters of the optimization model were also given in this paper,which can be used to reference in other prediction models or production practices.