Delayed hydride cracking-model investigation

Pressure tubes of cold-worked Zr-2.5%wt Nb alloy are the primary containment for the coolant of CANDU-6 NPP (Nuclear Power Plant). Under certain conditions, zirconium and its alloys are susceptible to a localised embrittlement process called Delayed Hydride Cracking (DHC). DHC is caused by the preferential and repeated accumulation and fracture of hydride at stress raisers such as cracks, causing the cracks to slowly increase in length over a period of time. Because understanding and controlling DHC initiation and growth is of great importance in ensuring the safe and economic operation of CANDU NPP, the DHC mechanism continues to be a subject of extensive experimental and theoretical investigations. The purpose of the present paper is to implement the Dutton - Nutal - Puls (DNP) DHC model into a computer code and to check it with the experimental data, obtained in INR-Pitesti. After this step certain analytical expressions were replaced with available empirical correlations in order to improve the agreement between theoretical predictions and experimental DHC data. The theoretical DNP model gave an expression for the DHC velocity as a function of stress intensity factor, material yield stress, temperature, hydrogen content, and bulk hydride distribution. Plots of calculated crack velocity versus inverse temperature are presented as well as a plot of experimental data, DNP calculated data and modify model data. A positive result of the modified model is that the arrest crack growth temperature is evaluated at about 600 K, a value closer to the experimental range value (425 - 540 K) than the DNP value which is 770 K. Concerning the difference between experimental and calculated plot that still remains, a better evaluation can be obtained if we consider the changes in DHC velocities involved by hydride spacing, L. The velocities supposing that L depends on the hydrogen content present in the alloy matrix were calculated and the results suggest that the changes of this far field distance to smaller values will produce a better fit for both calculated values of velocities and arrest temperature. This changes can be explained by the fact that the hydrogen content from samples is greater than terminal solubility. So, for a value of L about 100 μm the arrest temperature is about 588 K. But, this will be a subject of future researches