Burst ductility of zirconium clads: the defining role of residual stress

Abstract

Closed end burst tests, using room temperature water as pressurizing medium, were performed on a number of industrially produced zirconium (Zr) clads. A total of 31 samples were selected based on observed differences in burst ductility. The latter was represented as total circumferential elongation or TCE. The selected samples, with a range of TCE values (5 to 35 pct), did not show any correlation with mechanical properties along axial direction, microstructural parameters, crystallographic textures, and outer tube-surface normal (σ 11) and shear (τ 13) components of the residual stress matrix. TCEs, however, had a clear correlation with hydrostatic residual stress (P h), as estimated from tri-axial stress analysis on the outer tube surface. Estimated P h also scaled with measured normal stress (σ 33) at the tube cross section. An elastic–plastic finite element model with ductile damage failure criterion was developed to understand the burst mechanism of zirconium clads. Experimentally measured P h gradients were imposed on a solid element continuum finite element (FE) simulation to mimic the residual stresses present prior to pressurization. Trends in experimental TCEs were also brought out with computationally efficient shell element-based FE simulations imposing the outer tube-surface P h values. Suitable components of the residual stress matrix thus determined the burst performance of the Zr clads.