A new method is proposed combining multiple synchronized digital image correlation setups (multi-DIC) and finite element model updating to identify the hardening behaviour and anisotropy of 23.5 mm thick X70 line pipe steel. Curved tensile samples have been cut from a coil. While performing a tensile test on those samples, the force was obtained from the load cell and the back and front surface strain fields were measured by means of two synchronized stereo digital image correlation setups. The tests on the curved samples are reproduced with FE simulations, applying the same boundary conditions as the experimental setup to obtain the numerical force and strain fields. While simultaneously minimising the discrepancy between the experimentally and numerically obtained force and strain fields, the strain hardening behaviour is identified beyond the point of maximum uniform elongation.
A profound understanding of the anisotropy is also mandatory because the hot rolling operation develops substantial anisotropy which has an important influence on the line pipe performance. Due to the 23.5 mm thick steel that is used in this work, it is possible to measure the front and side surfaces with two synchronized stereo digital image correlation setups. Because full field information is available in all 3 material directions (lateral, longitudinal and through thickness direction), a 3D anisotropic yield criterion can be identified. A prerequisite for stable and accurate identification of the yield locus parameters is that the governing parameters are sufficiently sensitive to the experimentally measured response. For this purpose, a double perforated specimen has been designed which includes a side perforation. The latter guarantees the necessary through-thickness information to inversely identify the 3D anisotropic yield function through multi-DIC and finite element model updating.
The presented procedure could potentially be used by line pipe manufactures to verify whether the mechanical properties meet the specified requirements. The proposed approach has some advantages compared to conventional methods to determine mechanical properties of large diameter pipe. The curved specimen geometry is modelled in the FE simulation, hence the detrimental effects of flatting the tensile specimen can be avoided. Further, the new approach enables to consider the complete wall thickness as opposed to conventional testing with round bar samples of which a part of the wall thickness is removed during manufacturing.
Identification of Local Elastic Parameters in Heterogeneous Materials Using a Parallelized Femu Method
