Sulfate attack on portland cement-based materials: Mechanisms of damage and long-term performance

Through the novel applications of x-ray microtomography (microCT) and energy dispersive x-ray diffraction (EDXRD) as techniques of microstructural characterization, and by combining this information with measurements of expansion and compressive strength, mechanisms of damage by sulfate attack on portland cement-based materials were investigated. Specifically, the influence of cation type (i.e., Na₂SO₄ vs. MgSO₄), sulfate ion concentration (i.e., 33,800 ppm, 10,000 ppm, and 1000 ppm), cement composition (i.e., general use Type I vs. sulfate-resisting Type V), w/c (i.e., 0.485 vs. 0.435), and the presence of aggregates on sulfate attack were monitored over time by microCT and EDXRD and through complementary measurements of compressive strength and expansion. Some key findings include: (a) under sodium sulfate attack damage manifested in the forms of expansion, cracking and spalling, while under magnesium sulfate attack damage was primarily in the form of loss of material and compressive strength; (b) the onset and progress of damage was more rapid at higher sulfate concentrations for both sodium and magnesium sulfate exposures; (c) the onset and progression of cracking under sodium sulfate attack was more rapid in lower w/c samples, than that at the higher w/c; possible explanations for this unanticipated result are proposed; and (d) the inclusion of aggregates led to a more rapid onset and progress of damage as compared to corresponding cement paste samples. Based on the analysis, it was observed that ASTM C1012, which is a standard test method to evaluate sulfate performance of portland cement-based mortar bars, does not adequately capture different manifestations of sulfate damage. It is recommended to augment this method by including measurements of compressive strength, and to modify the sulfate environment in the test method to accommodate two possible scenarios of sulfate attack on the field. A model for sulfate transport through saturated, portland cement paste, and ensuing damage is proposed. The model predicts that, when exposed to an external sulfate solution with constant sulfate ion concentration and pH, different sulfate attack products will be predominantly deposited in different zones, and that onset of damage will occur in or near the outer zone.