Effect of moisture content on alkali-activated stabilisation of expansive soils

This study presents a comparative evaluation of two slag-based alkali-activated stabilisation approaches for expansive soils, distinguished by the initial moisture contents of the treated soil: (i) optimum moisture content, OMC, (in a moist state) and (ii) liquid limit, LL, (in a wet state). The soil at OMC can be dynamically compacted as part of the stabilisation process, whereas the soil at LL is not suitable for dynamic compaction and hence allowed to consolidate either statically or through vibration, similar to the casting methods used in cement-based systems. A direct comparison was undertaken to assess the influence of these moisture conditions on swelling behaviour, mechanical performance and microstructural development, using a combination of one-dimensional swelling tests, unconfined compressive strength (UCS) tests, Brazilian tensile strength (BTS) tests, scanning electron microscopy with energy-dispersive spectroscopy (SEM–EDS), powder X-ray diffraction (PXRD), and Fourier transform infrared spectroscopy (FTIR) analyses. Results revealed that both stabilisation approaches in moist and wet states outperformed conventional lime stabilisation in terms of swelling control and strength enhancement. Treated soil samples in a moist state exhibited superior strength gains, attributed to the development of highly polymerised, interlocked C-(N-)A-S-H gel structures. In contrast, treated soil samples in a wet state, while forming less polymerised and weaker gel networks due to excessive moisture and competing carbonation reactions, displayed more uniformly distributed gel phases, resulting in more effective swelling control at lower binder slurry dosages. These findings offer valuable insights for optimising the tailored application of alkali-activated systems in expansive soil stabilisation, highlighting the trade-offs between strength and swelling performance as a function of moisture regime and binder dosage.