Research On The Finite Element Simulation And Process Optimization In Fused Deposition Modeling

Fused deposition modeling (FDM) is one of the technologys on rapid prototyping and manufacturing (RPM) that develops in recent years. With the development of modern industry, fused deposition modeling is widely used in developing new product, rapid mold manufacting, design and development of medical devices, making prototype of human organs and so on. Fused deposition modeling becomes a important direction of development. However, the development is limited by the lower accuracy and intersity parts on FDM. Therefore, this paper studies the numerical simulation and process optimization in fused deposition modeling to improve the accuracy on FDM parts. It not only has important theoretical significance, but also has important practical value on engineering.The influencing factors of the accuracy on FDM parts are analysed. They are the machine error on modeling and manufadting, the error on CAD model, the error on slicing, the error on spinneret width, the error on materials shrink and so on. This paper puts forward some methods which can improve the precision and surface quality of parts.The finite element simulation technology of temperature field on FDM is studied. Based on the process characters of FDM, suitable hypothesis are introduced. The model of finite element simulation of temperature field on FDM is founded. Considering the latent heat of phase change and the theremal physical property parameters varied with the temperature's changing, the temperature field on FDM is analysed and simulated. The distribution of temperature field is obtained and the concrete analysis ars discussed. The conclusions have some referenced value on theoretical analysis of thermal process on FDM and following research.The finite element simulation technology of stress field on FDM is studied. Based on the simulation of temperature field on FDM, the model of finite element simulation of stress field on FDM is founded through thermodynamic couple method indirectly. The simulation program is compiled by language APDL (ANSYS Parametric Design Language). The simulation analysis of stress field on FDM is completed and some profitable conclusion is given.The parts'precision predidtion technology based on wavelet neural network (WNN) of stress field on FDM is studied. Based on analyzing the principle of artificial neural network and its concrete algorithm, the pretiction model of FDM parts is founded by BP neural network. Concidering the defect of BP neural network, the paper presents a new method that combines the wavelet with ANN in order to overcome some defect on ANN which convergence is difficult and its training speed is slow. Then, the prediction model of FDM parts based on wavelet neural network (WNN) is founded. The subsystem of FDM parts'prediction based on wavelet neural network (WNN) is developed by M language and compiled on MATLAB. It is practically applied on FDM.The process parameter optimization technology based on genetic algorithm (GA) on FDM is studied. Based on analyzing the genetic algorithm (GA) and its basic operation and the subsystem of wavelet neural network (WNN) for FDM parts' prediction, the subsystem of genetic algorithm (GA) is developed. It is practically applied on FDM to optimize the concrete process parameters.The software system of FDM is developed and perfected. Based on studying the algorithm of slicing and intelligent support on FDM, the software system of FDM is perfected by object oriented language VC++2005.net based on the earlier research work of our work team. The software can slice and add support automatic on any three-dimensional parts with STL format. Some parts are selected to test the software and the products were manufactured which were qualified. The results demonstrate that the software can run reliabilitily.