Comportamento mecânico de um solo siltoso de Curitiba-Paraná melhorado com cimento e pó de vidro

Abstract

The soils of the Guabirotuba Formation, predominant in Curitiba, exhibit high expansibility and low mechanical resistance, resulting in unfavorable geotechnical conditions for the implementation of engineering works. In this context, the present study investigated the influence of adding cement and glass powder on the strength of a local silty soil. For this purpose, undisturbed samples were collected from a street exhibiting pavement defects, revealing the collapsible behavior of the soil. The samples underwent initial characterization tests to determine the mineralogical composition of the fine and coarse fractions, texture, structure, physical indices, consistency limits, and mechanical resistance tests: tensile strength by diametral compression and simple compression. Mixtures with varying proportions of glass powder (0%, 1%, 2%, 3%, and 4%) in combination with cement proportions (0%, 2%, 4%, 6%, and 8%) were tested. The results of the mixture compaction tests showed similar behaviors under different applied energies, with a decrease in apparent dry bulk density and an increase in optimum moisture content compared to the natural soil, along with increased mechanical resistance. The research comprised two phases, both involving tests of simple compression and tensile strength by diametral compression. The first, more simplified phase encompassed three mixtures with a curing period of 28 days and optimal moisture content. Tests were conducted in quadruplicate to obtain the initial resistance parameters. In the second phase, nine mixtures with fixed moisture content were tested, with curing times of 28 and 90 days, and samples in quintuplicate. A comprehensive statistical analysis was conducted to assess the impact of additives on sample resistance. Subsequently, regression models were developed to relate mechanical strength to the variables of cement and glass powder. The results indicated an extremely strong correlation for tests with a 28-day curing period and a moderate correlation for samples with a 90-day curing period. The coefficients associated with cement and glass powder in the regression equations revealed that both additives positively contribute to tensile and compressive strength, although this influence tends to diminish over time, being more pronounced in the initial 28-day period. Even after 90 days, the regression model still indicated an increase in the strength of all samples.