Prediction Of Fracture Toughness In The Ductile-to-brittle Transition Region Of Pressure Vessel Steels

Ferritic steels used for construction of pressure vessels undergo a ductile-to-brittle transition (DBT) with decreasing temperature, which is a potential threat to the integrity and safety of pressure vessels, especially reactor pressure vessles. The fracture toughness of ferritic steels shows a strong temperature dependence on temperature and is highly scattered in the DBT region, which make it very diffcult to describe and measure the fracture toughness data. The scientific assessment and prediction of the ductile-to-brittle transion of materils pose great challenge to both the academic circle and the engineering circle. The main contents and conclusions of the thesis are as follows:(1) The applicability of ASTM E1921to the domestic16MnR steel and the domestic A508-Ⅲ was investigated. The reference temperature To for16MnR was determined to be-63℃and the To value for A508-Ⅲ was-61℃in accordance with ASTM E1921. The applicability of ASTM E1921to the two types of steel was validated.(2) The determination of the reference temperature To using small specimens of A508-Ⅲ and16MnR was carried out. The experiment results showed that small specimen may lead to non-consevative To estimate even when the test was performed according to E1921requirements. The reference temperatures calculated from the small specimens resulted in To estimates that were30℃to40℃below the estimates obtained from the larger specimens. To avoid the incorrect To estimate, four auxiliary methods for censoring To were suggested, including a more stringent Mlimit value, the evaluation of Weibull slope, the evaluation of the Kjc data validity in terms of Jp/Jc or CMODP/CMOD and the evaluation of To validity using the empirical correlation between To and TCVN from Charpy impact testing.(3) Based on random sampling from experimental data and Monte Carlo simulation of data from the scatter band described by the Master Curve, it was proved that the calibration method proposed by Minami would result in large uncertainty in the calibrated Beremin’s parametes, m and σu, when a finite number of data are available. It was found that m and σu vary regularly, and they constitute a unique curve for a specimen.(4) A new calibration method was proposed to find Beremin’s parmeters determined by the intersection of m～σn curves for high and low constraint specimens. Compared with the existing calibration method based on toughness scaling model (TSM), the new calibration method is characterized by low computational cost, the same accuracy and visual display of the calibration process.(5) Monte Carlo simulation was employed to produce a large number of fracture toughness data randomly drawn from the scatter band described by the Master Curve to determine Beremin’s parameters in the DBT region. In terms of16MnR steel, the results revealed that m decreases with temperature in the lower transition region and remains independent of temperature over the lower-to-mid transition. The changes in the scatter of macroscopic fracture toughness data and cleavage fracture mechanism are responsible for the temperature dependence of m. The Weibull stress scale parameter, σu, increases with temperature reflecting a rise in the microscale toughness.(6) The Beremin’s parameters for16MnR and A508-Ⅲ were calibrated using the calibration method based on the intersection of m-σu curves and the calibration method proposed by Ruggieri et al.(RGD method), respectively. After the fracture toughness data from small specimens were scaled to the results of1T-SE(B) by using Beremin’s toughness scaling model, the To values were estimated. The underestimated T0b measured from small specimens, which exhibited loss of constraint, were corrected.