PSI - Issue 70

Sahil Sehrawat et al. / Procedia Structural Integrity 70 (2025) 394–400

395

CO₂ emissions, making it a key target for decarbonisation efforts (Andrew, 2018; Habert et al., 2020). Regrettably, the manufacturing of OPC results in significant greenhouse gas emissions such as CO₂, which play a role in climate change (Bayasi & Zhou, 1993; Li et al., 2021). OPC manufacturing is projected to produce 1.354 billion tonnes of greenhouse gas emissions annually (Cleetus et al., 2018). Research efforts have focused on reducing OPC usage by investigating alternative binders. Geopolymers, introduced by Davidovits in the 1980s, have emerged as sustainable alternatives due to their lower carbon footprint and high performance in structural applications (Davidovits, 2008; Provis & van Deventer, 2014). Alkali-activated geopolymers possess the capability to surpass conventional cement based concrete (CBCC) (Ahmad et al., 2022a; Zhang et al., 2020; Zakka et al., 2021; Babu, 2018). Alkali-activated compounds are derived from a blend of precursors and activators. The chemical reaction results in geopolymers characterised by either low or high calcium levels, with a Ca/(Si + Al) ratio surpassing one (Marvila et al., 2021). Geopolymers seek to serve as a substitute for OPC in concrete (Farooq et al., 2021a; Muttashar et al., 2018; Farhan et al., 2020; Hosan et al., 2016). The objective is to develop a building material that removes the need for OPC, emphasises eco-friendliness, and fosters sustainability. As industries and populations expand, landfills accumulate a variety of waste materials, such as fly ash, rice husk, discarded glass, and powdered granulated blast furnace slag. These waste materials, rich in aluminosilicates, are ideal precursors for geopolymer concrete and promote circular economy principles (Singh et al., 2015; Chindaprasirt & Rattanasak, 2017). The disposal of these contaminants in landfills negatively impacts the environment (Herath et al., 2021; Ahmad et al., 2021a; Alyousef et al., 2021; Khan & Ali, 2019; Anjos et al., 2020; Mehta & Ashish, 2020). Waste materials include high-aluminosilicate raw substances suitable for geopolymer concrete (GPC), and their repurposing helps to mitigate pollution (Tchakouté et al., 2016; Ahmad et al., 2021b; Reddy et al., 2018). Figure 1 illustrates the GPC components alongside the associated curing procedures. It depicts the prevalence of these wastes and emphasises that the need for affordable housing will increase with population growth, indicating that their use will be advantageous for both the environment and the economy (Van Deventer et al., 2012; Jindal, 2019). Global research suggests that GPC may emerge as the foremost green-building material (Wong et al., 2020; John et al., 2021; Pilehvar et al., 2018; Shi et al., 2011). GPC has the potential to enhance cement-based construction and technology. ML models have been utilised by a number of researchers (Nikhil et al. 2023, Malik et al. 2025) in order to solve difficulties that are associated with civil engineering.

Fig. 1. Schematic of geopolymer manufacture, Yang et al., 2022 Supervised machine learning for predicting material properties is extensively utilised in contemporary artificial intelligence. Ahmad et al. (2022a) employed decision trees, AdaBoost, and bagging regression to forecast the properties of fly ash GPC and its compressive strength. The BR model exhibited the greatest precision. Ahmad et al. (2020) described genetic programming and artificial neural networks to predict the compressive strength of recycled aggregate concrete, with GEP surpassing ANN ML in accuracy. Similarly, Song et al. (2021b) used waste materials and artificial neural networks (ANN) to estimate the compressive strength (CS) of concrete. GEP outperformed ANN ML in predicting the desired outcome. Song et al. (2021a) utilised waste material and artificial neural networks (ANN)