A new practical model to calculate the reduced compressive strength of pre-damaged lightweight concrete subjected to freezing-thawing cycles

Many existing lightweight concrete (LWC) buildings have reached the end of their service life across Eastern Europe and Central Asia. This is manifested as progressive damage in the LWC, which in turn is reflected in a reduced compressive strength (RCS) and in several building collapses. To assess the structural condition of such vulnerable buildings, an accurate estimation of the RCS is necessary. However, limited research exists on the progressive damage and RCS of LWC. This article aims to investigate experimentally and analytically the influence of progressive damage on the compressive strength and thermal properties of LWC. The study also proposes a new thermal-based model to predict the RCS of damaged LWC subjected to freezing and thawing (FnT) cycles. To achieve this, 108 concrete cubes (size 100 mm) were subjected to different pre-damage levels (DL = 0%, 40%, 60%, 75%, 85%, 95%) and to subsequent FnT cycles (30, 60, 90, 120, 150, 180). After this, the thermal conductivity (TC) of the LWC cubes was determined using an innovative ad-hoc test rig. The cubes were finally tested in compression until failure to determine their RCS. The experimental results indicate that the relationship between the TC and RCS follows an approximate linear trend regardless of the number of applied FnT cycles. The TC of the tested cubes consistently reduced as the level of pre-damage and number of FnT cycles increased. At 30 FnT cycles, the TC of severely damaged cubes with DL = 95% (0.518 W/mK) was 58% lower than that of cubes with DL = 0% (0.891 W/mK). Based on the experimental results, a new and practical thermal-based RCS prediction model was derived for damaged LWC, adopting a modified version of Maxwell's equation for homogeneous materials. The experimental results are used to calibrate the new model, which calculates the RCS of LWC based on measured TC values. The proposed model accurately predicts the RCS of LWC cubes, with a Test/Prediction ratio of 1.0 and a Std.Dev. = 0.15. The accuracy of the proposed model is evaluated against LWC cored cylinders from a building located in Karaganda (Kazakhstan), and the model is proven to predict well the RCS of the cored cylinders with a mean T/P ratio of 1.13 (Std.Dev. = 0.26). The findings of this study contribute towards the development of more accurate models to assess progressive damage in LWC and the structural condition of existing buildings exposed to harsh continental climates.