Multiscale Modeling of Transport Phenomena and Dendritic Growth in Laser Cladding Processes
- PDF / 1,560,805 Bytes
- 13 Pages / 593.972 x 792 pts Page_size
- 46 Downloads / 239 Views
INTRODUCTION
LASER cladding is used widely in the rapid fabrication of parts and surface modification. A comprehensive understanding of this process is essential for the prediction and control of the final microstructure and properties of the deposited layer. Because of various constraints and the complexity involved, experimental observation cannot provide detailed information on the dynamics of this process, and numerical simulation may serve as a better means of investigation. The process inherently includes multiscale phenomena, and therefore, a multiscale model is required. The heat/mass transport and free-surface evolution can be modeled on a macroscale, whereas the dendritic growth and solute redistribution should be modeled on a microscale. Extensive efforts have been made on the macroscale modeling of the heat/mass transport in the cladding process. Some models have been developed based on a heat conduction equation with the consideration of latent heat of phase transformation,[1–5] whereas other models have included a convection term in the formulation to have a more complete consideration of the underlying physics.[6–12] The most challenging issue in the modeling of cladding processes is to track the freesurface evolution caused by material deposition. Some models are based on a simplified formulation with a predefined geometry of clad track,[13,14] which can give only qualitative results. Recently, more advanced techniques have been introduced to track the dynamic variation of the surface. The element activation method has been used in finite-element models,[1,2,5] whereas the volume-of-fluid method[7] and the level-set method[6,8–12] WENDA TAN, SHAOYI WEN, and NEIL BAILEY, Graduate Students, and YUNG C. SHIN, Professor, are with the Center for Laser-Based Manufacturing, Purdue University, West Lafayette, IN 47907. Contact e-mail: [email protected] Manuscript submitted March 16, 2011. Article published online July 1, 2011. 1306—VOLUME 42B, DECEMBER 2011
have been used in finite-volume and/or finite-difference models in which the thermofluid flow has been considered. An improved level-set method was proposed by Wen and Shin[11,12] to capture the evolution of the free surface of clad tracks. Additional source terms have been incorporated into the governing equation set to consider more rigorously the effects of continual additions of mass and energy from the deposited powder. The model has been validated by actual cladding experiments with different laser-density profiles and powder distributions.[11,12] In contrast, different models have been developed to investigate the microscale dendritic growth and solute redistribution in the molten pool. The Monte Carlo (MC) model has been used to by Yang et al.[15] and Koseki et al.[16] to predict the grain size and distribution in the molten pool as well as in the heat affect zone, and a good agreement with experimental results was obtained. However, the MC model can describe only the microstructure on the grain level and no feature on the dendritic level can be captured. To inve
Data Loading...
