FEpX is a modeling framework for computing the elastoplastic deformations of polycrystalline solids. Using the framework, one can simulate the mechanical behavior of aggregates of crystals, referred to as virtual polycrystals, over large strain defor
mation paths. This article presents the theory, the finite element formulation, and important features of the numerical implementation that collectively define the modeling framework. The article also provides several examples of simulating the elastoplastic behavior of polycrystalline solids to illustrate possible applications of the framework. There is an associated finite element code, also referred to as FEpX, that is based on the framework presented here and was used to perform the simulations presented in the examples. The article serves as a citable reference for the modeling framework for users of that code. Specific information about the formats of the input and output data, the code architecture, and the code archive are contained in other documents.
A methodology is presented for estimating average values for the temperature and the frictional traction over a tool-workpiece interface using measured values of force and torque applied to the tool. The approach was developed specifically for fricti
on stir welding and friction stir processing applications, but is sufficiently general to be of use in a variety of other processes that involve sliding contact and heating at a tool-workpiece interface. The methodology works with a finite element framework that is intended to predict the evolution of the microstructural state of the workpiece material as it undergoes a complex thermomechanical history imposed by the process tooling. We employ a three-dimensional, Eulerian, finite element formulation; it includes coupling among the solutions for velocity, temperature and material state evolution. A critical element of the methodology is a procedure to estimate the tool interface traction and temperature from typical, measured values of force and torque. The procedure leads naturally to an intuitive basis for estimating error that is used in conjunction with multiple meshes to assure convergence. The methodology is demonstrate for a suite of three experiments that had been previously published as part of a study on the effect of weld speed on friction stir welding. The probe interface temperatures and torques are estimated for all three weld speeds and the multi-mesh error estimation methodology is employed to quantify the rate of convergence. Finally, comparison of computed and measured power usage is used as a further validation. Using the converged results, trends in the material flow, temperature, stress, deformation rate and material state with changing weld conditions are examined.
