Catsius Clay Project. Calculation and Testing of Behaviour of Unsaturated Clay as Barrier in Radioactive Waste Repositories. Stage 2: Validation Exercises at Laboratory scale

Stage 2 of CATSIUS CLAY Project: Validation Exercises at Laboratory Scale includes two Benchmarks, Benchmark 2.1: Oedometer Suction Controlled Tests on Samples of compacted Boom Clay and Benchmark 2.2: Small Scale Weltting-Heating Test on Compacted Bentonite. BM 2.1 had two parts: BM 2.1A (volumetric deformation upon wetting-drying cycles) and BM 2.1 B (swelling pressure test). In BM 2.1A, participants were asked to model the results of a series of five tests on samples of compacted Boom clay. In BM 2.1B, a swelling pressure test in which suction, vertical and horizontal stresses were monitored, was proposed as a blind exercise. Participants were asked to use, without further changes, the models calibrated in BM 2.1A. This exercise provides an evaluation of the capabilities of current mechanical constitutive models for unsaturated clay behaviour. It was found that, even if a calibration exercise on the basis of known experimental data is satisfactory, blind predictions of tests involving different paths may prove difficult. The test set up for BM 2.2 consisted of a stainless stell cell filled with highly expansive compacted bentonite (S2 clay from Almeria, Spain). The clay was subjected to a simultaneous central heating and a progressive water inflow through the botton plate. Temperature at various locations within the sample and the boundary radial stress were monitored throughout the test. Water content distribution was also measured at the end of the experiment. Predictions for this benchmark required the solution of field equations for flow, temperature distribution and mechanical analysis. Model parameters were derived from the extensive set of available experiments on this clay. Comparison between model predictions and measurements revealed the significance of water transport in vapour phase, the difficulties to predict boundary stresses and the general good agreement between measured and calculated temperatures. The report provides a detailed accojnt of the models used by participants, the procedures to determine model parameters and discussions on the physical reasons which may explain agreements or discrepancies between model predictions and the actual behaviour of the two expansive clays involved in the two benchmarks proposed for solution. (Author)