It is now well recognized that pre-loading a ferritic steel structure at a temperature above its ductile-brittle transition range may result in an effective increase in toughness on loading at a lower temperature where fracture occurs by cleavage. Of particular significance is the fact that cleavage failure has never been reported in a cracked structure at a low temperature, following pre-loading at a higher temperature, for cases where the applied crack driving force was constant or decreasing with time. This is the basis of the so-called “conservative” warm pre-stressing (WPS) principle. Although this principle is of considerable practical importance, much work is still needed to further understand all aspects of WPS phenomena in order that analysts be better able to assess their effects quantitatively in safety cases.
In the present work, WPS experiments have been carried out using 20-mm-thick, plain-sided compact specimens sectioned from 150-mm-thick A533 Grade B Class 1 pressure vessel steel. The effects of load-unload-cool-fracture (LUCF) and load-cool-fracture (LCF) pre-loading cycles (KWPS ≈ 100 to 350 MPa √ m; TWPS = 25 or 290°C) on cleavage fracture resistance have been investigated over a range of temperatures spanning the ductile-brittle transition range.
A particular objective of the work was to investigate the combined effects of WPS and strain ageing on transition toughness behavior; therefore, in addition to WPS experiments, tests were carried out on “virgin” specimens to obtain base-line data and on specimens sectioned from material which had been uniformly pre-strained by 2% at ambient temperature and then aged for 1.25 h at 290°C to obtain control data.
The main results were as follows. The validity of the conservative WPS principle was confirmed: In every test performed, an increase in loading was required to cause cleavage fracture at low temperature following a severe pre-load at a higher temperature. Moreover, in the case of pre-loading carried out at 25°C, conferred WPS margins were observed even for those tests where pre-loading had resulted in ductile tearing (0 ⩽ Δa ⩽ 1.0 mm) under conditions of ligament yielding; and the Case 1 and Case 2 solutions of Chell's Q1 -integral theory of WPS yielded reasonably accurate predictions of final fracture. A deleterious effect of pre-loading at 290°C on effective transition toughness below a level of ≈ 200 MPa√m was demonstrated by a combination of control and LUCF and LCF WPS data, and found to be consistent with the expected embrittling effect of strain ageing. However, all toughness data generated by the study were bounded by the ASME KIc versus temperature curve indexed to the value of RTNDT (=-30°C) appropriate to the material in the non-strain-aged condition.