Damage identification using inverse analysis for 3D coupled thermo-hydro-mechanical problems

In this paper, location and degree of damages in massive masonry structures are identified by a multi-field based inverse analysis which relies on a series of measurements such as transient displacements, temperatures and water pressures. As it is typical for the multi-field problems, the existence of damage leads to local changes in parameters of the different physical fields. The degree of the damage is defined by one primary variable, from which other quantities are derived. For fluid-flow problems in deformable porous media under non-isothermal boundary conditions such a quantity is the porosity of the material. The inverse analysis bases on a global search method, in which a dual-level parallel-computation is applied to improve computational cost. The effects of uncertainties in measurements and the size of the damage on the accuracy of the solutions are also discussed in the paper.