Next: Determination of the
Up: Experimental Procedure
Previous: Water Permeability
Forty-two hours before the start of the test, a cylinder was removed from the curing water and allowed to dry for 30 minutes. The cylinder was brushed and cleaned with a compressed air jet, then placed in a 110 x 225 mm cylindrical steel mold which was specially machined to form specimens that would fit tightly in the acrylic receptacles of the diffusion cell. A mixture of epoxy and fine sand was cast in the annulus around the sample. The epoxy was allowed to cure overnight; then the sample was cut into three 50 mm thick slices. The samples were then vacuum saturated following the procedure specified by ASTM C 1202 .
The rate of chloride diffusion was determined using diffusion cells similar to those described by Detwiler et al. , except that KCl and KOH rather than NaCl and NaOH solutions were used in order to avoid interference with the chlorine peaks in the neutron activation analysis used for determination of the Cl- ion concentration. This test method is based on the ASTM C 1202- 91 test , which uses a 60 V DC power supply to drive the Cl- ions through the sample and infers a resistance to diffusion indirectly by measuring the electrical charge passed in six hours. The test used in this study, though more laborious to perform than the ASTM test, avoids certain of its problems. In this test, the Cl- ion concentration at the downstream face is measured directly, thus eliminating the need to infer a degree of ion transport from an indirect measurement. Also, the driving force is only 12 V, which is not enough to increase the temperature of the specimen. Since temperature affects the rate of ion diffusion, this is a significant advantage.
The exact thickness of each specimen was measured. The sharp edges were scraped and filed down as needed so that the cylinder could form a good seal against the O-ring of the acrylic receptacle of the diffusion cell. The edges of the cylinder were coated with a high vacuum grease to improve the seal. Stainless steel wire mesh electrodes were placed so as to cover the exposed concrete surfaces. A receptacle was force fitted over the electrode onto each end of the cylinder and the whole assemblage bolted together. The cathodic compartment was filled with 3% KCl solution and the anodic compartment with 0.3 M KOH solution. The volume of solution in each side was 260 ml. The stainless steel electrodes were then connected to a 12 ± 0.1 V DC power supply. Samples of the KOH solution were taken for chloride content analysis daily or twice daily depending on the expected rate of ion transport. The KCl concentration was also monitored on one sample from each set. It was observed that the Cl- concentration decreased very rapidly during the first 30 - 40 hours in the upstream side, while later the decrease was slower. To approximate constant concentration of the upstream KCl solution, it was decided to change this solution every 24 hours for one sample in each set. The test had to be stopped when rust occurred in the downstream (KOH) part of the cell. Neutron activation analysis (NAA) was used to determine the chloride content of the liquid samples . With this method, concentrations as low as five ppm could be determined.
Next: Determination of the Up: Experimental Procedure Previous: Water Permeability