Dominique GUEUGNAUT, Manuel TESSIER, Romuald BOUAFFRE
Papers # 2018 Las-Vegas
Given the virtual absence of available up-to-date results regarding the acceptance criteria for defects at electro-welding interfaces, a study has been launched by GRTgaz RICE with the two-fold aim of confirming the accuracy of the PAUT technique for the detection and sizing of calibrated pit-shaped defects implanted in an electrofusion(EF) tapping saddle on the one hand, and of evaluating the harmfulness of these volume defects on the mechanical strength of a pipe-saddle EF assembly through hydrostatic pressure testing, on the other hand. This study reveals the good reliability level of PAUT in detecting and sizing the volume defects - even the smallest ones - implanted in the saddles on the one hand, and the fair correlation existing between PAUT and X-Ray Tomography as a complementary technique on the other hand. The hydrostatic pressure tests do not reveal any detrimental influence of the defects – in terms of both location and size - on the mechanical resistance of the welded assemblies in these laboratory conditions. At this stage, pending complementary investigations, no correlation can be established between the observed failure times and the position and size of the volume defects implanted in the tapping saddles.
Phased Array Ultrasonic Technique (PAUT) is currently recognised as the reference technique for non-destructive evaluation of electrowelded assemblies. Therefore, over the past five years, CRIGEN has managed a research pogram comprising various test campaigns both on assemblies recovered from the field and on lab specimens. The primary aim of these campaigns was to evaluate the accuracy of the technique, on the one hand in terms of the detectability of defects mainly located at the pipe-fitting interface and, on the other hand, in terms of the sizing of detected defects. Moreover, an attempt was made to correlate the proportion and distribution of surface defects with the results of peel decohesion tests; but the low sensitivity of these mechanical tests made it difficult to establish a reliable link between the energy required for decohesion and the distribution and proportion of the surface defects. Now the implementation of acceptance criteria for defects – either surfacic or volumic - in welded assemblies requires the establishment of strong correlations between PAUT data and mechanical tests capable of representing long-term performances. But to date, an extremely limited number of relevant studies have been published in this field and the subject is still a matter of debate, as seen during the NDT workshop of the Plastic Pipes XVIII. Consequently GRTgaz Research and Innovation Center for Energy has launched a new campaign in 2016 focused on the volume defects, with the two-fold aim of confirming the accuracy of PAUT for the detection and sizing of volume-type defects implanted in a pipe- saddle electrowelded assembly on the one hand, and evaluating the harmfulness of volume defects on the mechanical resistance of assemblies through hydrostatic pressure testing on the other hand. For this purpose, four pit-shaped defects were hot-machined in the heating area of a 63 mm saddle at three different locations (outer, middle and inner crown). For a given configuration, the defects had four different heights in the saddle body. PAUT examination during the welding cycle allowed evaluating the changes in height of the defects. X-ray tomography analysis was used to complete the PAUT evaluation and to evaluate the final shape of the implanted defects. Hydrostatic pressure testing was then applied to the welded assemblies up to failure, with regard to the NF EN 1555 standard requirements. Finally peel tests were carried out in order to confirm the crack paths. This study reveals the good reliability level of PAUT in detecting and sizing the volume defects in the saddles on the one hand, and the fair correlation existing between PAUT and XRay Tomography as a complementary technique on the other hand. The hydrostatic pressure data do not reveal any detrimental influence of the defects – in terms of both location and size - on the mechanical resistance of the welded assemblies in these laboratory conditions. Duplicate tests should be made in order to confirm this point.
