LEACHING EXPERIMENTS

In order to validate the results of the numerical simulations, six different cement paste mixtures were prepared with the three cements described in Table (1) and at two w/c ratios (0.4 and 0.6). Only the results obtained for the 0.6 w/c ratio mixture made of the CSA Type 10 cement (cement A) will be reported. Data obtained for the other mixtures will be discussed in a forthcoming publication.

All mixtures were prepared using deionized water. The mixtures were batched in a high-speed mixer placed under vacuum (at 10 mbar) to prevent, as much as possible, the formation of air voids during mixing. Mixtures were cast in plastic cylinders (diameter = 7.0 cm, height = 20 cm). The molds were sealed and rotated for the first 24 hours in order to prevent any bleeding of the mixtures. At the end of this period, the cylinders were demoulded and sealed with an adhesive aluminum foil for an 18-month period at room temperature. This period was selected in order to get mature and well-hydrated cement pastes.

After the 18-month curing period, samples of each mixture were subjected to migration tests, porosity measurements pore solution extractions and thermal analyses (to assess the degree of hydration of each system). The experimental procedures for the migration tests and the pore solution extractions have been described elsewhere^{28, 29}. Porosity measurements were carried out according to the requirements of ASTM C 642²⁷. The water diffusion properties of these mixtures had been previously determined by Nuclear Magnetic Resonance Imaging (NMRI) as part of a previous project³⁰.

The remaining parts of the cement paste cylinders were sawn in thin disks. The thickness of the disks varied from 12 mm to 15 mm. The disks were then vacuum saturated in a sodium hydroxyde solution (prepared at 300 mmol/L) for a 24 h period prior to the degradation experiments. The latter were performed during three months under saturated (series 1) and unsaturated conditions (series 2) using deionized water.

The series 1 samples were first coated with an epoxy resin (on all their faces except one) and then immersed in the test solutions (see Figure (1)). For the samples of series 2 (unsaturated conditions), a relative humidity gradient was created between the two faces of the disks (see Figure (1)). One face was directly placed in contact with water and the other was placed in contact with a CO₂ free environment at a relative humidity close to 65 %. In order to avoid carbonation, nitrogen was added on a daily basis in the compartment.

  

Figure 1: Degradation test set-up.

At the end of the degradation experiments, samples were broken in small pieces and then immersed in isopropyl (propan-2-ol) alcohol for a minimum period of 14 d. After this period, samples were dried under vacuum for 7 d. Once the drying process was completed, the samples were impregnated with an epoxy resin, polished, and coated with carbon.

Microstructural alterations of the cement paste samples were investigated by means of electron microprobe analyses. The polished sections were observed using a microprobe (Cameca SX-100 ¹) operating at 15 kV and 20 nA. For each sample, measurements were performed at a maximum interval of 13 µm on four distinct imaginary lines extending from the external surface in contact with the aggressive solution toward the internal part of the samples. At each point of measurement, the total content of calcium, sulfur, sodium, potassium, silicon, and aluminum was determined.

¹

Certain commercial equipment is identified by name in this paper to adequately specify the experimental procedure. In no case does such identification imply endorsement by the National Institute of Standards and Technology, nor does it imply that the products are necessarily the best available for the purpose.