Neutron diffraction residual stress measurements on girth-welded 304 stainless steel pipes with weld metal deposited up to half and full pipe wall thickness

• Published in: The International journal of pressure vessels and piping 2013-01, Vol.101, p.1-11
• Main Authors: Haigh, R.D., Hutchings, M.T., James, J.A., Ganguly, S., Mizuno, R., Ogawa, K., Okido, S., Paradowska, A.M., Fitzpatrick, M.E.
• Format: Article
• Language: English
• Subjects: Deposition; Finite element method; Finite element modelling; Girth welding; Mathematical analysis; Mathematical models; Neutron diffraction; Pipe; Residual stress; Wall thickness; Welded joints
• ISSN: 0308-0161; 1879-3541
• DOI: 10.1016/j.ijpvp.2012.08.003

The residual stress distribution has been measured in two girth-welded austenitic stainless steel pipe weldments using time-of-flight neutron diffraction. One had weld filler metal deposited up to half the pipe wall thickness, and one had weld metal deposited up to full pipe wall thickness. The aim of the work is to evaluate the evolution in residual stress profile on filling the weld, on which there is little experimental data, and where the selection of the correct hardening model used in finite element modelling can benefit greatly from an understanding of the intermediate residual stresses partway through the welding operation. The measured residual stresses are compared with those calculated by finite element modelling and measured using X-ray diffraction. The results show a change in the measured hoop stress at the weld toe from tension to compression between the half- and fully-filled weld. The finite element results show an overprediction of the residual stress, which may be a consequence of the simple isotropic hardening model applied. The results have implications for the likely occurrence of stress corrosion cracking in this important type of pipe-to-pipe weldment. ► 304 steel girth welded with weld metal to half and full pipe wall thickness. ► Residual stresses measured by neutron and X-ray diffraction, and modelled by FE. ► Weld toe residual σhoop changes from tensile to compressive from half to fully-filled. ► FE model for the fully-filled weld gives higher stress levels than those measured. ► Discrepancy is attributed to the isotropic hardening model used.