This thesis investigates the aging of dolomitic concretes and evaluates thermogravimetric analysis as a fast, relatively simple and cost-effective approach for qualitative and quantitative assessment of cement–carbonate systems. Since concrete is a multiphase material, exposure to aggressive environments alters both the hydrated cement paste and carbonate aggregates, which is reflected in the evolution of TG/DTG curves with time. The aim of this work was to relate TG/DTG features to phase changes confirmed by X-ray powder diffraction, and to test a procedure for estimating phase composition from thermal data.
Two concretes prepared with Gunte dolomite aggregate (Slovenia) and two different binders (C1 and C3) were studied. After 28 days of curing, specimens were aged in 1 M NaOH at 60 °C for up to 12 months, with intermediate sampling after 28 days and after 3, 6 and 12 months. Phase composition was determined by XRD using Rietveld refinement, while thermal behaviour was monitored by simultaneous TG/DTG up to 1200 °C. For selected samples, mass spectrometry was coupled to the thermal analysis; the m/z = 18 and m/z = 44 signals were used as indicators of H$_2$O and CO$_2$ release, which proved particularly useful in temperature ranges where a single broad DTG peak can conceal several overlapping processes. DTG peak deconvolution was applied to quantify individual contributions, followed by stoichiometric conversion of mass losses into phase contents. Importantly, the calculation was restricted to phases confirmed by XRD, as introducing additional phases would be speculative.
The results show systematic aging-related changes in both XRD phase assemblage and DTG peak contributions. In the carbonate range, calcite contributions generally increase while dolomite decreases, consistent with progressive transformations in carbonate-rich systems and changes in pore solution chemistry. Secondary Mg- and Al-bearing phases (hydrotalcite-like phases, brucite and hydrogarnets) become more pronounced with aging, whereas portlandite decreases. Sulfate–aluminate phases exhibit a trend of ettringite destabilisation in carbonate-rich conditions, with minor formation or persistence of monocarbonate AFm. Comparison between TG/DTG-based estimates and XRD results indicates that the approach captures major trends, while larger deviations occur in the low-temperature H$_2$O region and in the carbonate region, where multiple carbonate contributions overlap and cannot always be uniquely separated from DTG shape alone.
A key limitation was identified in the treatment of less-ordered carbonate contributions: a broad carbonate peak attributed to less-ordered calcite may also include less-ordered dolomite. Determining the ratio between less-ordered dolomite and less-ordered calcite (e.g., via additional XRD constraints or calibration) would likely improve the agreement between calculated and measured phase contents and yield more robust stoichiometric back-calculations. Overall, the work confirms TG/DTG as a useful rapid method for assessing the broad structure of aged concretes, while combination with XRD (and MS when needed) reduces ambiguity in peak assignment and interpretation.
|
