Abstract

In this paper, the behaviour of the tantalum subjected to a complex range of thermomechanical solicitations is studied and modelled. The objective of this work is the prediction of microstructure evolution of a tantalum part that is cold flow-formed, then stress-relieved and softened by heat treatments. Flow forming is a cold chipless process that elongates and thins the wall of a tubular part by applying a free rotating roller on the wall of the rotating part. (See Figure 1) This process leads to large strain up to 5 and one material point may be deformed at a fluctuating strain rate. To model this process, it is thus recommended to use a constitutive law that takes into account the history of the material. Microstructure evolution is controlled through the whole process by monitoring the state variable, that is, the dislocation density, within a FEM code. The 3D FEM software FORGE2007® is used to model the entire process. The code is enhanced with physical constitutive laws relative to the tantalum and derived from works of Klepaczko and Buy et al. [1,2,3]. The formalism of Klepaczko and Buy et al. is useful since it is based on laws regulating physical mechanisms of dislocations multiplication, annihilation and kinetics of glide depending on strain rate and thermal activation. Static recovery is modelled subsequently with an incremental approach where the softening is related to the dislocation density evolution. A set of thermomechanical experiments and treatments are done to identify the constitutive law and associated microstructure evolution. Compression, dynamic and static torsion tests covering a wide range of strain rates are used to fit the mechanical constitutive law and the dislocation density evolution law both inspired from Klepaczko and Buy et al. works. The same samples are then annealed with variable temperatures and times. Microstructure of annealed samples is characterized by means of micro-hardness. They give information to evaluate the dislocation density evolution. The data is used to model static recovery.0.

Details

Actions