Multiobjective Optimization of 316L Laser Cladding Powder Using Gray Relational Analysis
- PDF / 1,871,117 Bytes
- 14 Pages / 593.972 x 792 pts Page_size
- 82 Downloads / 189 Views
©ASM International 1059-9495/$19.00
Multiobjective Optimization of 316L Laser Cladding Powder Using Gray Relational Analysis Mingsan Xu, Chunhui Zhou, Xu Huang, Zheng Zhang, and Tao Wang Submitted: 8 June 2020 / Revised: 5 October 2020 / Accepted: 17 October 2020 During laser cladding of repair parts, multiple objectives must be synthesized; however, the multiobjective optimization of laser cladding is difficult. In this paper, the gray correlation of microhardness, residual stress and wear rate is established to transform these three parameters into a single-objective optimization algorithm to perform a gray correlation. By also using response surface methodology to obtain a mathematical model of the gray correlation degree, the model was analyzed and verified by analysis of variance, and the model accuracy was 90%. The optimal process parameters were determined using the gray correlation model, and the maximum gray correlation was 0.9204832, the microhardness reached 203.65 HV, the residual stress was 45.72 MPa, and the wear rate was 1923.56 μm3/N m. The phases and microstructure of the laser-cladded coating were characterized by scanning electron microscopic analysis of a sample prepared under the optimal processing parameters. This study provides guidance for the multiobjective optimization of process parameters for 316L powder cladding.
Keywords
gray relational grade, laser cladding, process parameter optimization
1. Introduction Laser cladding is a surface modification technology that improves the surface mechanical properties of parts with a cladding layer that exhibits better mechanical properties than those of the bulk material (Ref 1-3). This type of cladding is widely used in the metallurgy, aerospace, automobile, ship and chemical industries. Laser cladding can effectively repair damaged parts and improve the microhardness (MH) (Ref 4), wear resistance, corrosion resistance and other properties of the parts used in the field of remanufacturing (Ref 5, 6). Because of the fast-cooling and fast-heating solidification processes involved in laser cladding, high residual stress (RS) exists in the clad parts, which affects their shape accuracy and service life (Ref 7, 8). At present, the research methods used to optimize the laser cladding parameters are mainly divided into neural networks (Ref 9), finite element analyses (Ref 10), response surface experiments (Ref 11) and orthogonal experiments (Ref 12). Li et al. used the nonlinear processing capacity of a neural network to establish a model for predicting laser cladding process parameters for the cladding-layer morphology; when this model was combined with the extremum learning optimization model, the error rate was between 10 and 20% (Ref 13). Li et al. used the finite element method to establish a threedimensional thermomechanical coupling model to simulate the Mingsan Xu, Chunhui Zhou, Xu Huang, Zheng Zhang, and Tao Wang, School of Mechanical and Automotive Engineering, FuJian University of Technology, Fuzhou 350118, China; and Productivity Promotion Center
Data Loading...
