Design of pipeline composite repairs : from lab scale tests to FEA and full scale testing
Résumé
Composite repair systems are used for many years to restore locally the pipe strength where it has been affected by damage such as wall thickness reduction due to corrosion, dent, lamination or cracks.
Composite repair systems are commonly qualified, designed and installed according to ASME PCC2 code or ISO 24817 standard requirements. In both of these codes, the Maximum Allowable Working Pressure (MAWP) of the damaged section must be determined to design the composite repair. To do so, codes such as ASME B31G for example for corrosion, are used. The composite repair systems is designed to “bridge the gap” between the MAWP of the damaged pipe and the original design pressure.
The main weakness of available approaches is their applicability to combined loading conditions and various types of defects. The objective of this work is to set-up a “universal” methodology to design the composite repair by finite element calculations with directly taking into consideration the loading conditions and the influence of the defect on pipe strength (whatever its geometry and type).
First a program of mechanical tests is defined to allow determining all the composite properties necessary to run the finite elements calculations. It consists in compression and tensile tests in various directions to account for the composite anisotropy and of Arcan tests to determine steel to composite interface behaviors in tension and shear. In parallel, a full scale burst test is performed on a repaired pipe section where a local wall thinning is previously machined. For this test, the composite repair was designed according to ISO 24817.
Then, a finite element model integrating damaged pipe and composite repair system is built. It allowed simulating the test, comparing the results with experiments and validating damage models implemented to capture the various possible types of failures. In addition, sensitivity analysis considering composite properties variations evidenced by experiments are run.
The composite behavior considered in this study is not time dependent. No degradation of the composite material strength due to ageing is taking into account. The roadmap for the next steps of this work is to clearly identify the ageing mechanisms, to perform tests in relevant conditions and to introduce ageing effects in the design process (and in particular in the composite constitutive laws).