Modeling Of Microstructure In TIG Welding Heat Affected Zone And Study On Mechanical Constitutive Relation For TA15Alloy

In current study, metallurgical and structural behaviors are concerned to study the mechanical performance of welded joint. Metallurgy refers to grain size and phase constituent, while structure includes structural stress and transformation stress. For titanium welding, grains in HAZ can grow rapidly, and widmannstatten may be formed at cooling stage due to microstructure heredity. When welding stress, hydrogen and microstructure are interacted, the formation of cold cracks will damage the whole welded part. Therefore, solid phase transformation plays the role of intermediate link. It is essential to couple the simulation of microstructure evolution with macroscopic stress and deformation.In this paper, cellular automaton model of grain coarsening and solid phase transformation are established. Simulations are performed to uncover the ambiguous physical mechanisms of solid state microstructure evolution in welding metallurgy. Firstly, cellular automaton of grain growth is modified considering thermodynamic energy difference due to large temperature gradient in welded joint, as well as activation energy of atomic jumping. Coarse mesh and similarity principle methods are applied for mapping of temperature to grain field, overcoming the difficulty of uncoordination of different space scale. The influence factors and characteristics of grain coarsening are studied. Results demonstrate small heat conductivity coefficient of titanium makes wide heat affected zone. Low atomic activation energy of β phase results in fast migration of grain boundary that grain growth is fierce. Grain growth with low, moderate and high heat input are compared. Higher heat input results higher temperature and smaller temperature gradient, accordingly, the thermal pinning effect is weaker and grain coarsening is more severe.Cellular automaton model of β to α phase solid transformation based on thermodynamic, evolution kinetics and crystallographics is presented. The driving force which determines the dynamics of new phase growth is calculated by evaluation and selection from current existing methods, according to its application field and solid solution type. Virtual front tracking (VFT) modified method is newly put forward to simulate solid phase transformation with specific crystallographic preferred orientation in TA15alloy, to eliminate the dependence of traditional CA method on space meshing and reflect the habit relation of β and α phase.Using the established model, the dynamic microstructure evolution during β to α phase transformation is simulated. The effects of thermodynamic factors including temperature and cooling rate, kinetic factors including solute diffusion, interface mobility and Gibbs Thomson coefficient, as well as crystallographic factors including preferred growth direction, anisotropic modules and coefficient on the phase growth velocity, morphology and solute concentration are analyzed. Simulation results exhibit desired agreement with the prediction result of analytical equation like JMA and classic phase transformation theory like M-S instability. Moreover, microstructure evolution heredity under thermal cycling is vividly reflected, showing the effect of grain size of parent phase, initial solute concentration, initial phase constituent and solute distribution coefficient in solid/liquid phase and in two solid solutions. Results demonstrate that grain size of parent phase and primary phase constituent take effect on solid phase transformation through changing the space for nucleation and growth of new phase. Micro-segregation due to solidification leads to new phase growth “hastening after benefit”.Furthermore, diffusion/mixed/interface controlled transformations are distinguished by a numeric definition. Solute diffusion and interface migration characteristics are compared to distinguish each phase transformation type. By simulation, it is indicated that during cooling transformation, diffusion controlled transformation at high temperature changes to interface controlled transformation at low temperature. High cooling rate corresponds to high transformation type transition temperature.To overcome the difficulty of expression and solution of traditional JMA equation for continuous cooling transformation, an explicit expression is transformed. Pattern search method was implemented to calculate the unknown parameters in the JMA equation according to the CA simulated phase transformation fraction at different cooling rate. A dedicated programming code for finite element simulation of solid phase transformation was developed. The elastic-plastic constitutive model for TA15titanium alloy welding which describes the mechanical effect of volume change during diffusional solid phase transformation and transformation induced plasticity (Trp) during diffusionless solid phase transformation has been established. The model in combination of mixed strain hardening and mechanical effect of solid phase transformation is coupled in MARC software by secondary development. The constitutive model considering phase transformation effect is applied for finite element numerical simulation of the welding process. It is indicated that strain hardening effect can lower the welding residual deformation at cost of enlarging the residual stress. Volume expansion in tensile stress zone and volume shrinkage in compression stress zone can lower the residual equivalent stress. Trp strain which is positively related to stress can moderate the stress distribution, make the high tensile stress and high compressive stress converge to zero and finally lower the residual equivalent stress. Maximal temperature for solid phase transformation with phase transformation strain and Trp strain are finally obtained to control the welding residual stress.