PSI - Issue 43

Lukáš Trávníček et al. / Procedia Structural Integrity 43 (2023) 148 – 153 Author name / Structural Integrity Procedia 00 (2022) 000 – 000

151

4

grades). However, going higher with the load range causes the pipe specimens to break with a ductile failure, which is illustrated by the fracture surface in the Fig. 4, where a pipe specimen made of material 1 (PE 80) loaded by Δ σ = 12 MPa failed after N f =13694 cycles by ductile failure – no SCG occurred. This happens, even though the stress intensity factor range (Δ K I = 0.59 MPa·m 1/2 , obtained by FEM) at this load level is lower than for CRB specimens (for Δ σ = 10 MPa it is Δ K I = 0.65 MPa·m 1/2 ), which still perform well and produce representative results at this level even for less SCG-resistant materials. This is caused by the relatively thin wall of the pipe – the thin wall makes constraint in the vicinity of the crack larger, which prevents SCG initiat ion at a lower Δ K I , but when the propagation starts, the amount of SCG is low because it transfers into the ductile failure mode very early or almost instantly as the ligament becomes thinner, making the results not representative. To be able to sort the materials according to SCG resistance, the loading range of Δ σ = 10 MPa was chosen for all the pipe specimens. The results of fatigue testing are presented in the next section.

Fig. 2. Photo of fracture surface of a CRB specimen made of PE100RC – material 8 taken by optical microscope. The areas of SCG and final ductile failure are depicted.

Fig. 3. Photo of fracture surface of an extruded pipe specimen made of PE100RC – material 8 taken by optical microscope. The areas of SCG and final ductile failure are shown in the detail.

3.2. Fatigue performance Fig. 5 summarizes the material ranking based on the number of cycles to failure N f of extruded pipe specimens and CRB specimens, respectively. For the pipe specimens, the mean value of N f for every material is depicted by black dots. Results from both types of specimens divide the materials into the polyethylene pipe grades groups correctly. However, the results of pipe specimens suffer by high scatter within every tested pipe grade, which is caused by the lack of constraint and lack of space for the crack to propagate. This makes the crack propagation more susceptible to any defects and irregularities in the geometry as opposed to the CRB specimens, where the results are notably less