"Вестник МГСУ"

The analysis of a large set of samples by means of several different methods - petrography, optical microscopy, IR- and Raman spectroscopy, porosimetry, DSC - is very common in practice of material science. After carrying out all the experiments, the groups of researchers obtain a wealth of raw data. The required final result, though, in most cases is to answer several - or even one - question concerning the state of the construction. Obviously, the transition from empirical information to the final decision can be done by means of non formal operations, for example expert appraisal. However, even for most intelligent experts it is quite difficult to perform such an evaluation. In order to condense the raw experimental data we propose simple and formal procedure. The offered method consists of several steps. The first step is to arrange data in such a way, that the rectangular matrix (of size M by N, where M and N are the number of samples and methods, respectively) is formed. This matrix can be called matrix of defectiveness. Then, for all pairs of columns of the mentioned matrix, we compute the Pearson's product-moment (correlation) coefficient; the result is the symmetric N by N matrix of accordance of methods. By means of summation over the rows of the later matrix we obtain information concerning the mutual correspondence of the methods - vector of significance (third step). And finally, at the fourth step, we compute the M scalar products of vector of significance and row of the matrix of defectiveness. The M obtained values are subject to further application by the descriptive statistics, and on the basis of this statistics the final decision can be made. The offered method was successfully applied in the practical task of identification of alcali-silica reaction.

Portland cement-based concrete is widely used in civil engineering. Therefore, it is very important to determine the preconditions of corrosion of the cement concrete. The service life of concrete structures can be substantially reduced by the alkali-silica reaction. It is well known that this reaction causes formation of the sodium silicate hydrogel. Thus, by identifying this gel, a researcher can make an assumption about the reasons for the corrosion. Obviously, macroscopic quantities of sodium salts can be discerned using analytical chemistry methods. Unfortunately, determinant values of such salts in the concrete structure are usually very small. Thus, there is a need for special research methods.Raman spectroscopy is an advanced method based on the analysis of instantaneous two-photon non-elastic light scattering. This method is applicable even in case of small quantities of chemicals under research. The first successful study of silicates using Raman spectroscopy methods was performed in the 20ies of the 20th century. In this work the authors have proven that sodium hydrogels can be easily identified in the concrete using the Raman spectroscopy. In the course of the analysis of the interphase boundary between the cement stone and the aggregates, the authors observed, at least, one spectral peak which did not belong to cement or to the disperse phases of the concrete. At the same time, this peak can be classified as a peak of the sodium silicate. Thus, sodium silicate gel is generated during the service life of the structure under research, and this research has revealed the presence of the alkali-silica reaction.