Investigation of postweld heat treatment of quenched and tempered pressure vessel steels

Author

Zoran Sterjovski, University of WollongongFollow

Year

2003

Degree Name

Doctor of Philosophy

Department

Materials Engineering - Faculty of Engineering

Recommended Citation

Sterjovski, Zoran, Investigation of postweld heat treatment of quenched and tempered pressure vessel steels, PhD thesis, Materials Engineering, University of Wollongong, 2003. http://ro.uow.edu.au/theses/464

Abstract

Longitudinal and circumferential welds in transportable pressure vessels are produced by submerged-arc welding using a single vee preparation and multiple weld runs. Quenched and tempered (QT) steels, which are commonly used for transportable pressure vessels, require mandatory postweld heat treatment (PWHT) regardless of the plate thickness. During their life transportable pressure vessels may have up to four PWHT cycles, and concerns have emerged about possible effects on material properties such as hardness, tensile strength, impact toughness and fracture toughness. This thesis reports on the weld procedure, microstructural evaluation and various mechanical properties (bend, yield strength, tensile strength, elongation, fatigue, impact toughness, CTOD fracture toughness and hardness) for 11 mm, 12 mm and 20 mm QT steel weldments. The 11 mm and 20 mm base plates investigated were BIS80PV, which is a pressure vessel grade steel. The 12 mm plate was BIS80, which is structural grade steel but deemed a possible candidate for pressure vessels due to superior impact and fracture toughness properties over currently used QT pressure vessel steels. The parent metal, heat affected zone and weld metal regions of each weldment were examined, and then exposed to temperatures and times in the PWHT range. Although there was no apparent change in microstructure at an optical level and little change in hardness for up to four postweld heat treatments, there was a marked decrease in hardness of the parent metal for more extensive heat treatments (increasing Holloman parameter). There was also evidence of minor secondary hardening in the 11 mm and 20 mm BIS80PV parent plate following short heat treatment times. The weld metal (WM) and HAZ hardness typically decreased with one PWHT cycle and subsequently stabilised with further PWHT cycles. For all test plates, results are also presented for Charpy V-notch impact tests in the parent metal, HAZ and weld metal region, and CTOD fracture toughness tests in the PM region. The effect of exposure to multiple PWHT cycles on these properties is discussed. A decrease in impact energy and fracture toughness with an increase in the number of heat treatments was evident in the parent metal. In contrast, the weld metal showed a decrease in impact energy after two PWHT cycles, and then an increase towards the original impact energy after a further two cycles. In PM samples, which have been extensively tempered in the manufacturing process, the mechanism by which toughness properties are affected by cumulative PWHT holding time is through the coarsening and coalescence of second phase carbide particles. These particles decrease impact energy by the formation of larger voids in the plastic zone ahead of the crack tip or by the initiation of cleavage fracture. These two phenomena decrease the energy required for fracture. In WM with an as-solidified structure, ductile failure in the form of void coalescence is initiated by non-metallic inclusions. This leads to an initially high impact toughness, which then decreases after PWHT because of the nucleation and coarsening of metastable Fe3C precipitates that promote quasi-cleavage type fracture. Upon exposure to further PWHT cycles the impact energy begins to increase again due to the dissolution of the metastable carbides and the formation of finer, more stable carbides based on elements such as Cr, Mo, Ni and Nb. This change in the form of the carbide promotes localised fracture by micro-plasticity, with void formation and coalescence. Additionally, cross-weld root bend (180°) and tensile tests were carried out before and after PWHT. None of the bend samples showed any evidence of cracking or tearing, hence confirming the ductility of the weldment. The tensile properties of the BIS80PV cross-weld samples complied with the Australian Standards, but the tensile properties of the BIS80 cross-weld samples only complied in the as-welded condition. Tensile properties of the PM showed no significant trends with the number of PWHT cycles and the fatigue crack growth rate increased slightly in relation to the number of PWHT cycles or PWHT holding time. One of the primary roles of PWHT is to reduce residual stresses caused by the welding process. Residual stress measurements using the hole drilling method were made to ascertain the need for PWHT. Residual stresses, measured in the weld centre-line by the hole drilling technique, were compressive along both the longitudinal and transverse directions and were no greater than 250 MPa (between 0.2-0.5 times the yield strength at room temperature). Although, the presence of residual stresses (tensile and compressive) of greater magnitude is not excluded, the residual stress measurements demonstrated that PWHT significantly reduced the magnitude of the residual stresses. The ability to predict the impact energy of a pressure vessel steel subjected to various cumulative PWHT holding times is beneficial to pressure vessel inspectors, who are misguidedly instructed to be wary of the mechanical properties of vessels after 6 hours of cumulative PWHT holding time. The original impact energy of the steel, which is determined by its chemical composition, microstructure and thermo-mechanical treatment determines the cumulative PWHT time before the impact properties of the QT steel decrease below the specified 40 J limit at –20°C. An artificial neural network (ANN) model has been developed using an artificial neural network to predict the effect of PWHT time and other variables such as composition and test temperature on impact energy. This ANN model provides a valuable design tool for predicting impact toughness as a function of composition, heat treatment and testing conditions. Finally, the performance of BIS80 was evaluated and its superior impact oughness and fracture toughness properties over BIS80PV were confirmed. However, the tensile properties of postweld heat-treated 12 mm BIS80 cross-weld samples did not comply with AS3597, the Australian Standard for QT steels; and the PM stress ratio exceeded the specified limit of 0.93. These ‘deficiencies’ could be remedied by (i) discarding PWHT or (ii) using a high strength WM; and (iii) abandoning or relaxing the stress ratio requirement, which is of questionable significance, under conditions of high strain rate impact loading (road collision).