In the self-sensing materials area, the cement-based matrix has been sufficiently studied during recent over 20 years. Considering the low-carbon target in 30-50 years all over the world, alkali-activated materials can be applied as a substitute for cement-based materials. The strong alkaline liquids on which the two-part alkali-activated materials rely are not safe and stable during transportation and storage. One-part alkali-activated materials with solid activators were proposed in recent several years to solve this problem. Therefore, this study can be divided into two parts, namely one-part alkali-activated materials and self-sensing materials.

To select a suitable matrix, the physical, fresh and mechanical properties of 27 AAM mixes with varying mix proportions were examined. Key design parameters include activator/binder ratio (A/B) from 0.05-0.15, binder content from 460-1050 kg/m³, and fly ash/GGBS ratio from 50/50-10/90. Flow table and setting time evaluated the workability, and compressive and flexural tests assessed the mechanical properties. Phase analysis in micro level, carbon emissions, and cost assessments were also performed. One-part AAMs achieved compressive strengths of 66 MPa at 7 days and 76 MPa at 28 days. Mixes with low binder content (460 kg/m³) showed 4-28% lower compressive strength at 84 days compared to 28 days. Higher binder content (650-1050 kg/m³) and A/B (0.1-0.15) had increased compressive strength by 6-21% at 84 days compared to that at 28 days. The mix with 0.15 A/B, 650 kg/m³ binder, and 30/70 fly ash/GGBS ratio performed best in a multi-criteria assessment that considers mechanical properties, embodied carbon, and cost. To understand the effects of environment and sand moisture content, the sand moisture (0-12%) and curing condition in the best two mixes were also explored. To compare the best two mixes with the cement-based material, two cement-based mixes were designed, tested and evaluated.

After obtaining the most suitable matrix mixing design in this study, the graphite (5-10% by weight) and metal fibres (0.5-5% by volume) were added as functional fillers to improve conductivity for self-sensing capability. The metal fibres with 0.2mm diameter and 6-13 mm length include recycled tire steel fibres, brass-coated steel fibres and stainless steel fibres. As the self-sensing sensors can be part of structural members, the fresh and harden performance of them was also examined, as well as phase analysis. According to the trial tests under monotonic loading to mixes with recycled tire steel fibres, 2-3% by volume were selected as the more suitable fibre contents, considering the balance between mechanical and piezoresistance performance. According to the results, the mixes with all kinds of metal fillers with 2-3% by volume and graphite fillers with 5-10% by weight were selected and tested under 100-cycle loading with 0.05 Hz. The loading ranged from 0.3-0.5 of the maximum loading of each mix. The plain matrix and cement-based matrix were also tested for comparison. The results show that the mixes with graphite and shorter metal fibres have less cracks after loading. The mixes with metal fibres can show better self-sensing performance than the graphite mixes. The mixes with brass-coated steel fibres can show less self-sensing sensitivity to stress compared with mixes with other steel fibres. The alkali-activated matrix could be better than the cement-based one due to more ion paths for conductivity.