Fatigue crack growth in welded joints in seawater

A pipe-to-plate specimen has been developed to study the influence of seawater on the fatigue behaviour of welded tubular joints. DC potential drop techniques have been used to detect fatigue crack initiation, and to monitor the subsequent growth of fatigue cracks. Results for three specimens, tested in air, are compared with similar data for tubular and T-plate joints. These comparisons indicate that the pipe/plate is a reasonable model of a tubular joint.

Testing was performed on a further six specimens in artificial seawater; two each with free corrosion, optimum cathodic protection, and cathodic over-protection. Fatigue life reduction factors compared with corresponding tests in air were 1.8 and 2.8 for free corrosion, 1.7 and 1.1 with cathodic protection, and 4.2 and 3.3 with cathodic over-protection. These fatigue life reduction factors were comparable to results on T-plate specimens, and were strongly dependent on crack shape development. The beneficial effects, if any, of cathodic protection are dependent on favourable initiation and crack shape development; no benefits should be given in design.

Linear elastic fracture mechanics techniques appear suitable for the calculation of fatigue crack propagation life. Three approximate solution techniques for crack tip stress intensity factors show reasonable agreement with experimentally derived values. It is recommended that forcing functions be used to model crack aspect ratio development in welded joints. Such forcing functions are influenced by the initial stress distribution and the environment.