Calibrating Constitutive Models to Laboratory Data: an Innovative Framework for the Numerical Reproduction of Triaxial System Tests

ABSTRACT: The geomechanical simulation is a routine tool employed in the Oil & Gas Industry to devise safe and efficient field development plans. The quality of the outcomes depends on the simulator inputs. Here, the quality and ease of the constitutive model calibration is assessed. The rock mechanics tests in this industry are typically performed on Triaxial Systems. The tests range from simple, single-stage tests, towards tests based on the concatenation of several loading stages. The literature on constitutive model calibration is, otherwise, generally based on limited sets of homogeneously-controlled tests. The current work demonstrates an innovative framework for the numerical reproduction of drained, quasi-static, isothermal tests executed on Triaxial Systems. It is based on two steps - a test segmentation; and the numerical replication of the boundary conditions. This framework intends to fulfill a literature gap of a versatile, prone-to-automate solution of numerical test reproduction. As shown in a brief literature review, the reproduction of a test is a common stage to all optimization techniques. When coupled to these optimization routines, the proposed framework is likely to transform the constitutive model calibration routine.