Research On Simulation Of Dynamic Thermodynamic Coupling And Optimization Of Process Parameters During The SLM

Selective Laser melting(SLM)has been widely used in small-batch,difficult-toprocess and high value-added product as a process of additive manufacturing.However,it has a decisive influence on the forming quality to select the process parameters.Once the process parameters do not match,the unbalanced temperature field distribution will produce large thermal stress,which will cause the crack,warping,porosity,deformation and other quality problems of the formed parts.In view of the unknown mechanism of cyclic thermodynamic coupling and the difficult matching and optimizing of process parameters are becoming a scientific problem.Therefore,a dynamical SLM simulation method based on moving heat source algorithm and circulating thermal transfer model is proposed to simulate and analyze the coupling mechanism of cyclic temperature field and cumulative thermal stress in the SLM processing.Furthermore,a "compound algorithm" is developed to optimize and matching intelligently the main process parameters of SLM,which is a method to obtain the evaluation of thermal results under different process parameters and a more reasonable parameter matching scheme.This study can provide theoretical reference for prediction of SLM forming thermal stress,matching and optimization of process parameters,and control forming quality.The full text is divided into six chapters,and the main research contents and innovation of this dissertation are summarized as follows:(1)A dynamic thermodynamic coupling simulation method for SLM forming driven by process parameters is proposed.Firstly,the new FEM method according to the actual process parameters is created based on the dynamic heat source,the cyclic heat transfer model,the elastic-plastic strain matrix,and the basic equation of thermal elasticity theory with the SLM process characteristics.Second,the dynamic thermodynamic coupling simulation of SLM forming process is developed based on the dynamic cyclic thermodynamic coupling algorithm on the simufact software.And through a really model with the actual SLM process parameters,the multi-layer and multi-channel forming simulation is carried out.Finally,the SLM dynamic thermodynamic coupling output results were verified by means of the molten pool monitoring and the XRD residual stress test data in the same process parameters with experiment,so as to prove the effectiveness of the algorithm and the reliability of the simulation.Compared with the existing methods,this simulation method can realize the highly efficient dynamic simulation of the multichannel SLM forming process,and which can achieve the data tracking of the cyclic thermal action process with the actual process characteristics.(2)A SLM process parameter matching method based on dynamic thermal coupling simulation is proposed.Based on the developed SLM dynamic thermodynamic coupling algorithm,the input method of multi-process parameters is extended to obtain the matching analysis of thermodynamic coupling results of several main process parameters.Firstly,the input parameter module of SLM dynamic thermodynamic coupling simulation is extended,and the multi-process parameter driven model simulation is carried out.Secondly,the simulation results were verified by the same combination experiment of process parameters,and the matching requirements of major process parameters such as laser power,scanning speed and scanning trajectory which were demonstrated by the simulation and test results of molten pool size and residual stress,so as to reduce forming defects or forming failures.At the same time,it is further proved that the multi-process parameters simulate algorithm is developable through the simulation of multi-process parameters.In addition,due to the interaction between process parameters is dynamic and non-linear,it is proved that intelligent iterative solution is needed to obtain the process parameter matching of qualified SLM forming parts.(3)SLM process parameter optimization based on compound algorithm is also proposed.Due to the nonlinearity and discreteness of the relationship between the input and output of the SLM multi-process parameters,the fuzziness of the objective function,a compound algorithm(named CA)is proposed.First,CA based on particle swarm optimization(PSO)combined with neural network algorithm-Error Back Propagation(BP)is designed,the serial parameters of SLM dynamic thermodynamic coupling simulation software and experimental measurements were used as learning and training samples in CLA.Second,the improved methods of multilayer subdivision structure of BP and variable inertia weight and variable learning factor of PSO were established which can avoid the compound algorithm from falling into the local optimal solution.Then the framework of CA was completed through Matlab software platform programming.Finally,the UI and debugging of SLM process parameter optimization system is carried out.Through the optimization of several groups of process parameters and the analysis of experimental test results,it is proved that the CA is convergence in evaluating the thermodynamics results of SLM main process parameters,and feasible in the optimizing the SLM main process parameters with boundary conditions.(4)The thermodynamic coupling mechanism and the rule of SLM process parameters are revealed.The SLM thermodynamic distribution trend and interaction mechanism were obtained according to analyze the main process parameters of different series in the simulation results due to calculate the temperature and stress cycle based on tracking of numerical points was set up in the dynamic simulation process.Furthermore,the corresponding relations of SLM’s four main process parameters under the change of cyclic thermal temperature field,the evolution of thermal stress(XRD residual stress test and EBSD grain statistics)and so on were studied by combining experimental measurements with operating data of compound algorithm.Therefore,the influence law of the SLM main process parameters on the thermodynamic effect was preliminarily obtained.Which can provide a theoretical reference for the design and analysis of multi-process parameters in the SLM engineering applications.