Return to search

Models for designing pipe-grade polyethylenes to resist rapid crack propagation

Plastic pipeline systems have now become dominant for fuel-gas and water distribution networks. Although they have an impressive service record failures do occur, with Rapid Crack Propagation being characterised as the least probable but most potentially catastrophic one. This study investigates the effect of structural morphology and bulk residual strains on the RCP performance of polyethylene pipes, and proposes a new methodology for predicting a safe service envelope. During crack propagation in PE pipes, the fracture surface has two distinct regions; plane strain and plane stress. In addition to the Instrumented Charpy, Reversed Charpy, High Speed Double Torsion, Dynamic Mechanical Analysis and uniaxial tensile testing, S4 tests of extruded pipe specimens were employed in order to evaluate the structural and fracture parameters of pipe grade resins in these two fracture modes on pipe. A new experimental technique, which modified the pipe bore crystallinity without altering the residual strain field (as evaluated from slit ring tests) showed that the bore surface layer properties had much less influence on RCP than previously thought. Parallel with the experimental work, modeling of the fracture mechanisms was also undertaken. Using previous models in the field, such as the adiabatic decohesion model, the plane strain fracture toughness was evaluated while the plane stress fracture toughness was evaluated either from the Reversed Charpy or from the stability of adiabatic drawing in a tensile test. A mixed mode, temperature sensitive toughness was finally evaluated, leading to an overall fracture properties assessment for polyethylene pipes which could be compared directly to the crack driving force during RCP in pipe. By employing a new mathematical approach, which incorporated both the effects of residual strains and pipe stiffness behind the pressure decay length, a previous basic analytical RCP model was further developed and compared to more elaborate finite element and finite volume solutions. The new results were also compared to S4 experiments using high-speed photography and showed that the new methodology could be employed by the end user even when testing facilities are not directly available

Identiferoai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:513525
Date January 2010
CreatorsArgyrakis, Christos
ContributorsLeevers, Pat
PublisherImperial College London
Source SetsEthos UK
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Sourcehttp://hdl.handle.net/10044/1/5564

Page generated in 0.0104 seconds