Chinese Journal of Lasers, Volume. 48, Issue 6, 0602112(2021)

Parameter Optimization of High Deposition Rate Laser Cladding Based on the Response Surface Method and Genetic Neural Network Model

Yifan Pang, Geyan Fu*, Mingyu Wang, Yanqi Gong, Siqi Yu, Jiachao Xu, and Fan Liu
Author Affiliations
  • Laser Manufacturing Technology Institute, School of Mechanical and Electrical Engineering, Soochow University, Suzhou, Jiangsu 215021, China
  • show less
    Figures & Tables(16)
    Schematic diagram of laser cladding with powder feeding
    Diagram of experimental equipment
    Schematic diagram of weld passage section
    Microstructure of laser cladding Fe314 with different deposition rates. (a) Sample 1; (b) sample 9
    BBD experimental parameters and results
    Random parameter experimental parameters and results
    Interactive influence of process parameters on deposition rate. (a) Powder feeding velocity and power; (b) defocus and power; (c) scanning velocity and powder feeding velocity
    Topological structure of BP neural network
    Diagrams of error iteration. (a) Evolution of genetic fitness; (b) iteration of neural network error
    Predicted comparison results of RSM and GA-BP models. (a) Model of RSM; (b) model of GA-BP
    Comparison of RSM and GA-BP generalization ability
    Comparison of optimization results between RSM and GA-BP
    • Table 1. Composition of Fe314 powder unit: %

      View table

      Table 1. Composition of Fe314 powder unit: %

      ElementFeCCrBSiNi
      Mass fractionBal.0.115.01.01.01.0
    • Table 2. Orthogonal experimental parameters and results

      View table

      Table 2. Orthogonal experimental parameters and results

      No.P /Wvf /(g·min-1)vs /(mm·s-1)D /mmRd /(g·min-1)
      1320043.58-1027.6
      2320058.010-1149.5
      3320072.512-1257.6
      4350072.510-1066.0
      5350058.08-1244.4
      6350043.512-1136.0
      7400043.510-1242.0
      8400058.012-1055.8
      9400072.58-1168.4
    • Table 3. Orthogonal experimental range analysis results

      View table

      Table 3. Orthogonal experimental range analysis results

      ParameterP /Wvf /(g·min-1)vs /(mm·s-1)D /mm
      K144.935.246.849.8
      K248.849.952.551.3
      K355.464.049.848.0
      η10.528.85.73.3
    • Table 4. Analysis of variance of deposition rate predicted by RSM model

      View table

      Table 4. Analysis of variance of deposition rate predicted by RSM model

      SourceSum of squaresMean squareF valuep valueSignificance
      Model3069.27383.6645.49<0.0001Yes
      A-P414.19414.1949.11<0.0001
      B-vf2293.572293.57271.97<0.0001
      C-vs41.0741.074.870.0392
      D-D43.3243.325.140.0347
      AB36.0036.004.270.0520
      AD27.5627.563.270.0857
      BC123.21123.2114.610.0011
      SourceSum of squaresMean squareF valuep valueSignificance
      B290.3590.3510.710.0038
      Residual168.678.43
      Lack of fit143.478.971.420.3985No
      Pure error25.206.30
    Tools

    Get Citation

    Copy Citation Text

    Yifan Pang, Geyan Fu, Mingyu Wang, Yanqi Gong, Siqi Yu, Jiachao Xu, Fan Liu. Parameter Optimization of High Deposition Rate Laser Cladding Based on the Response Surface Method and Genetic Neural Network Model[J]. Chinese Journal of Lasers, 2021, 48(6): 0602112

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Laser Material Processing

    Received: Jun. 22, 2020

    Accepted: Sep. 24, 2020

    Published Online: Mar. 15, 2021

    The Author Email: Fu Geyan (fugeyan@suda.edu.cn)

    DOI:10.3788/CJL202148.0602112

    Topics