In the case of DSAW, finite element calculations revealed that the stress level in the vicinity of the weld toe can be significantly higher than that in the pipe body. Figure 3 shows the stress profile for a typical seam weld geometry. Only half of the pipe wall thickness is indicated in the figure, as the weld was assumed to be symmetrical across the mid-wall line. Under an applied stress of 340 MPa (77% of SMYS for the pipe in question), there is a zone a few mm width in

Figure 3 Results of finite element calculation of stress levels in the vicinity of a weld.

which the actual local stress is close to the SMYS of the base steel. At the very toe, the stress is above the actual yield point of the material.

In one series of measurements of residual stress in pipes retrieved from service, tensile residual stresses in the range of 20% SMYS were often found to exist in the pipe wall up to a depth of about 1 mm [11], and the level of residual stress varied as a function of distance from the pipe surface. Thus if the nominal operating stress is at 72% of SMYS, the total net stress could be at 92% SMYS in the metal at this depth under the surface; such a stress level is conceivably high enough for crack initiation for many SCC systems.

EFFECTS OF STRESS FLUCTUATION

As in the case of pipeline SCC in carbonate-bicarbonate environment, the severity of transgranular SCC is not only affected by stress level per se, but also the degree of stress fluctuation. In a CANMET study on crack initiation [6], detectable cracks could be produced when stress was applied in a cyclic wave with a maximum of 90% SMYS and R=0.6. The cracking severity was much increased when the R-value was reduced to 0.4, under the same environmental conditions, maximum stress, load frequency and wave form. While these R-values are not typical of many gas transmission pipelines, the results do show the effects of R-values.

Laboratory results on the growth of deep SCC cracks demonstrate dramatic effects of pressure fluctuation. Figure 4 shows typical growth behaviour of cracks in a full-scale test. In such tests, sections of full-size pipes containing sharp fatigue pre-cracks were buried in soil,

Figure 4 Typical growth behaviour of cracks during full-scale tests showing the effect of pressure fluctuation. ("P"- pressure in psi, "S" - static hold period (min.) and "Dyn" - Dynamic load period (min.))

It should be noted that corrosion grooves, or "linear corrosion" as it is known, forms on pipe surface when the tape coating wrinkles to form long and narrow pockets of disbondment and the subsequent corrosion takes on the appearance of the coating wrinkles.  

PAGE 1 2 3 4 5 6 7 next