Evaluation of different laboratory methods for testing resistance of concrete to chloride ingress

Paper i proceeding, 2007

In the past years a European cooperative project called ChlorTest was carried out with the intention to arrive at laboratory test methods applicable to EU states based on non-biased approaches and evaluations, and to recommend the practical use of the laboratory performance tests based on the verification of the in-field performance of concrete under different chloride exposure environments. This paper presents the results from the evaluation of different laboratory methods for testing resistance of concrete to chloride ingress. In the evaluation two inter-laboratory tests were carried out: one for pre-evaluation and another for final evaluation. In the pre-evaluation six different test methods including prEN-13396, a natural immersion test NT BUILD 443, a rapid migration test NT BUILD 492, two steady-state migration tests and a resistivity test were evaluated by eight laboratories. Four types of concrete manufactured with Portland cement (PC) and PC blended with silica fume, fly ash and blast furnace slag, were used in the inter-laboratory test. Based on the pre-evaluation results and taking other factors into account, four methods, that is, NT BUILD 443, NT BUILD 492, a steady-state migration test based on the conductivity measurement, and an indirect method (resistivity test), were selected for the final evaluation, in which a total of fifteen laboratories participated in order to produce adequate precision data. Six types of concrete manufactured with the combination of four types of binder (PC, and PC blended with silica fume, slag and fly ash) were used in the final evaluation. Based on the evaluation results, three methods were recommended for further standardisation, i.e. a natural immersion test (NT BUILD 443), a rapid migration test (NT BUILD 492), and an indirect method (resistivity test), all of which have the precision in an acceptable range, i.e. repeatability coeffifient of variation (COV) ≤20% (11% to 20%) and reproducibility COV ≤30% (24% to 28%). Therefore, they are considered to be suitable for data exchanges and industrial applications.