Statistical Analysis and Monte-Carlo Simulations of Warm Prestressing of Pre-Cracked Small Single Edge Notch Bend Specimens

Abstract

Warm prestressing is widely acknowledged as being able to enhance material toughness, especially in steels that exhibit lower shelf cleavage fracture. The enhancement in toughness has a significant impact on the integrity of pressure vessels, particularly during severe loading conditions, such as pressurised thermal shock. In this paper, we undertake detailed statistical analyses of experimental data provided via a comprehensive programme of fracture tests at UJV (Ústav jaderného výzkumu Řež a.s.). A warm prestressing model, developed by Chell, is used to predict the change in toughness probabilistically, using Monte-Carlo methods to predict the distribution in toughness following different warm prestressing cycles. The results obtained from this model are also compared to predictions made by the Wallin approach. Experimental data was generated, at UJV for WWER 440 RPV steel, using small single-edge-notched bend SEN(B) specimens (or pre-cracked Charpy) across a range of different fracture temperatures, warm pre-stress temperatures, and levels of preload, in both as-received and irradiated conditions. In this paper, experimental data obtained only from tests on unirradiated specimens were statistically treated. A three parameter Weibull distribution was used to map the scatter observed in the virgin toughness. The statistical significance of increase in apparent fracture toughness due to warm prestressing was evaluated using the Mann-Whitney test. It was further shown by Monte-Carlo simulations that the Chell and Wallin models provide slightly conservative predictions of the resulting fracture toughness. Both, the experimentally measured and predicted values of the resulting fracture toughness, depend on the specific tests conditions, especially on the level of preload.