Font Size: a A A

Simulation Of Microstructure Evolution During Welding Solidification Of Nickel-based Alloy

Posted on:2011-11-11Degree:DoctorType:Dissertation
Country:ChinaCandidate:R MaFull Text:PDF
GTID:1101360332456506Subject:Materials Processing Engineering
Abstract/Summary:PDF Full Text Request
Solidification behavior during welding process controls the morphologies of microstructure, the extent of segregation, the distribution of inclusions, affects the formation of porosity and hot cracks, and ultimately influences the properties of the weld metal. To date, it's still hard to observe the solidification process by experiment method, and the dynamic and instantaneous features of weld solidification process can't be described accurately by experiment method either. So that it is of great importance to analyze solidification process of molten pool through numerical simulation method. In this dissertation, a coupled model of cellular automaton, finite difference and finite volume method is developed to reproduce dendrite growth behavior in molten pool. The characteristics of molten pool such as epitaxial growth and fluid flow are taken into consideration. The effects of fluid flow on dendrite morphology and primary dendrite spacing are studied, and solute redistribution in molten pool is also deeply researched.First, the dendrite grain growth model is established based on grain growth kinetics theory. In addition, the method of"diagonal simulating angle"is modified and the computational efficiency is improved to simulate the grain growth with random crystallographic orientations, which could describe the epitaxial growth in molten pool better. Based on this model, the simulation of grain growth of full-sized molten pool is carried out, and the microstructure evolution during welding solidification process is reproduced. The simulated results indicate that the competitive growth among grains is severe during solidification process, and the grains with preferred orientations could grow constantly, while other grains are eliminated. Ultimately, a complicated morphology is formed in molten pool. The simulated results agree well with experiments, which confirm the reasonability and feasibility of the simulation model in this dissertation.Flow field model is established to further study the effects of fluid flow in molten pool on grain growth, and the model is solved by finite volume method. Staggered-grid finite difference approach and SIMPLER algorithm are also employed to solve the problem of the discrete of pressure gradient term and the lacking of independent equation of pressure term, respectively. After the establishment of the flow field model, it is coupled with grain growth model, and the solute diffusion equation is replaced with solute transfer equation to solve the solute distribution in liquid.The simulation of the equiaxed grain growth and columnar growth with fluid flow in molten pool is carried out based on the model mentioned above, and the results are also compared with the grain growth without fluid flow. The behavior of fluid flowing at the dendrite tips and its effects on dendritic morphologies and solute distribution is reproduced. The simulated results present that fluid flow alters solute distribution and constitutional undercooling tendency around the dendrite tips, accordingly, the morphologies of grains exhibit assymetrical. The dendritic morphologies are different with different inlet flow direction, but they all have the same characteristic that the dendrite arms on the upstream side grow fast, while the growth of dendrite arms on the downstream side is much delayed. With inlet velocity increasing, the asymmetry of dendrite growth becomes even severe. The effects of fluid flow on primary dendrite spacing are also examined. The results show that fluid flow changes the solute distribution at the dendrite tips of columnar grains that grow relatively slower and makes them eliminated in the competitive growth more quickly, accordingly, accelerates the primary dendrite spacing adjustment process.Microsegregation in weld seam could induce the formation of second phase, pore and hot crack. Therefore, quantitative forecasting of the microsegregation could serve as basis for the prediction of the mechanical properties of weld metal. The characteristics of solute redistribution and microsegregation of columnar grain growth with and without flow are studied, and the effects of welding speed and nucleation density are also examined. The results indicate that for grain growth without flow, solute concentration increases from the bottom of the dendrite to the dendrite tips along the grain growth direction in the interior of a columnar grain, but increasing extent slows down. Eventually the solute concentration tends to a certain value. Along the direction normal to grain growth, the solid composition at the center of the main stem is low, while near the edge and at the secondary arms it is high. For grain growth with flow, the solute distribution along grain growth direction has the same tendency as that without flow, while the extent of solute segregation is greater along the direction normal to grain growth direction. With other conditions constant, with welding speed increasing, the extent of solute segregation becomes more severe, and with nucleation density increasing, the extent of solute segregation reduces.A system for simulating grain growth in molten pool is designed based on this study, and it consists of 4 parts including fluid field computation, temperature field computation, grain growth simulation and data post-processing. The system provides a simple graphical user interface, by which the simulation of grain growth in molten pool could be completed easily and the results could also be post processed. The system is designed to achieve a goal of providing reference to actual welding process.
Keywords/Search Tags:molten pool, dendrite grain growth, fluid flow, solute segregation, cellular automaton
PDF Full Text Request
Related items