PSI - Issue 26

Vladimír Chmelko et al. / Procedia Structural Integrity 26 (2020) 417–421 Chmelko et al. / Structural Integrity Procedia 00 (2020) 000 – 000

420

4

destruction. In case, if we use the actual stress as the value of the strength limit or recalculated stress according to the relationship (Iľkaev et a l. 2005, Chen et al. 1998) ( )    = + 1 con true (1) this means that for a strain corresponding to a stress at 10% strength, the actual value of the ultimate strength will be (2) The selected condition can represent the destruction of the pipeline only for the case of ideal coat geometry without defects. Geometric variances of the pipeline from roundness and defects in welds and coats cause uneven stress distribution in these locations. The stress peaks have the notch effect characterizing the stress concentration in a small volume of material. They need to cooperate with the stresses in their surroundings and around the defect for the destruction of the pipeline. For the development of plastic collapse, it is necessary that the surroundings of the point where the stress has reached the value of the true strength limit allows the development of plastic deformation. In the case of small defect dimensions (see Fig.1: welded roots of welds, offset of weld faces, hole corrosion, ...) it is necessary to achieve in the computational simulation a state where the stress in the whole defect surface reaches the real strength limit and the defect surroundings is also in plastic state (Fig.4). Such a situation will allow the development of plastic collapse into the destruction of the pipeline and can be considered as corresponding to the destruction pressure. The condition of destruction must also take into account the multiaxial state of stress in the pipe casing. In Chmelko et al. (2019) the best results were obtained using equivalent stresses of Mises-Hubert-Hencky. Experimental verification of proposed numerical simulation procedures was performed in laboratory conditions on samples of small pipes DN151x4. Individual pipes had a defect in the longitudinal direction of the coat of 41 x 25 mm with three different wall thickness decreases: 24%, 35% and 75%. The results and comparison with the computational simulations are in Table 1. m,con m,true . R R 1 1 =

Table 1. Experimentally obtained burst pressures and results of FEM simulations Specimen No. Depth of defect (%) Burst pressure (MPa) p experimental p FEM 1 24 32.1 32 2 35 30 29.8 3 75 18.5 18

Fig. 4. The state stress in pipeline from the streel X52 with defect at the moment of destruction.