PSI - Issue 13

Adam Smith et al. / Procedia Structural Integrity 13 (2018) 566–570 Smith / Structural Integrity Procedia 00 (2018) 000 – 000

567

2

Health and Safety Executive (HSE, 1999) provided a recommendation for the value of the ERF on the fatigue life of welded joints based on investigations concerning offshore steel structures subjected to wave loading (0.15 Hz to 0.5 Hz). It is recommended in HSE (1999) that a ERF of 3 should be applied to assess the fatigue lives of welded joints in sea water subjected to freely corroding conditions, and this recommendation is used as the basis for guidance in the British Standard BS 7608:2014+A1:2015. However, the experimental results reported in HSE (1999) contain considerable scatter (ERFs range from about 1 to 5), and there are limited data in this report (and the open literature) concerning the effect of different variables on the ERF (e.g. applied stress range, cycle frequency, and weld quality). Furthermore, the ERFs in HSE (1999) are only applicable to structures subjected to wave loading (cycle frequency > 0.15 Hz), and it is indicated in BS 7608 (2015) that the detrimental effect of sea water could be greater for structures loaded at lower cycle frequencies (e.g. submarine pressure hulls). Minimal information is provided in relevant standards or the literature concerning the ERF for welded joints subjected to low cycle frequencies (< 0.15 Hz), presumably because fatigue testing of welded joints is expensive and time consuming (especially at low cycle frequencies). Thus, it is difficult to accurately assess the fatigue life of welded steel structures in marine environments subjected to these low cycle frequencies. One relevant study, Kilpatrick (1997), measured the fatigue life of freely corroding full-penetration tee-butt welded joints of high strength Q2(N) steel (690 MPa) in aqueous 3.5 wt.% NaCl using a constant stress range (518 MPa) and various constant cycle frequencies (from 1 Hz to ~0.001 Hz) (Fig. 1). The results of this study indicate that the ERF varies from about 1 (i.e. little reduction) at high cycle frequencies (1 Hz) to 5.8 at very low cycle frequencies (~0.001 Hz). However, it was emphasised by the researchers in the study that the test data were limited and that more work on the environmental effects on fatigue was required.

Fig. 1. (a) Plot of stress-range versus number of cycles-to-failure for tests conducted in air on tee-butt welded joints (Kilpatrick (1997)). (b) Plot of number of cycles-to-failure (left axis) and the corresponding environmental reduction factor (right axis) versus cycle frequency for tests conducted in aqueous 3.5 wt.% NaCl (an analogue for sea water) on tee-butt welded joints showing that the deleterious effect of a salt-water environment on fatigue life increases with decreasing cycle frequency (Kilpatrick (1997)). In the present study, the experimental fatigue results, reported by Kilpatrick (1997), have been modelled using a fracture-mechanics based approach in conjunction with the finite-element method. The main focus was to model the fatigue crack-growth behaviour in aqueous 3.5 wt.% NaCl under freely corroding conditions at various cycle frequencies so that a better understanding could be obtained concerning the detrimental effect of the environment, particularly at low cycle frequencies (< 0.15 Hz). This type of fracture-mechanics based analysis is now possible because a more comprehensive corrosion-fatigue crack-growth-rate dataset, describing the behaviour over a broad range of cycle frequencies, compared with previous studies (e.g. Vosikovsky (1975), Shimodaira et al. (1990)) has been recently published (Knop (2015)).