This paper presents a physics-consistent hybrid surrogate framework for simulating the mechanical behavior of reinforced concrete beams that utilize waste fired clay (WFC) as a partial substitute for cement. The main contribution is the integration of empirically observed deformation behavior with physics-informed learning to produce an interpretable, mechanically valid surrogate model. Full-field surface deformation fields were measured using Digital Image Correlation (DIC) under monotonic loading and processed through a convolutional neural network (CNN) to extract deformation- and crack-sensitive features. These features were integrated with experimentally measured stress–strain data within a Physics-Informed Neural Network (PINN) in which equilibrium and conditional constitutive monotonicity constraints were enforced through the loss function. A Non-Dominated Sorting Genetic Algorithm II (NSGA-II) was utilized as a downstream parametric exploration tool to examine trade-offs among maximum load capacity, material cost, and embodied CO₂ inside a constrained mixture-design space. Model interpretability was assessed by SHapley Additive exPlanations (SHAP), indicating that deformation-driven kinematic factors predominantly influence stress prediction, whereas WFC content and reinforcement parameters have a secondary, mixture-level impact. The resulting framework achieves enhanced predictive accuracy (R² = 0.969) relative to its individual components and operates as an offline, physics-calibrated surrogate rather than a real-time digital twin, providing a reliable and interpretable basis for structural assessment and sustainability-oriented design evaluation of WFC-modified reinforced concrete beams.

Throughout their service life, concrete buildings are subjected to a number of significant degradation processes, one of which is exposure to high temperatures. This degradation degrades the mechanical and physical properties of concrete, resulting in a reduction in its strength. Consequently, it is essential to enhance the qualities of concrete at elevated temperatures. Therefore, this study examines the synergistic effects of WBA content and temperature on the mechanical properties of concrete, emphasizing sustainability and high-temperature durability. WBA substituted fine aggregate at 0–50% by mass, and specimens were subjected to ambient and elevated temperatures up to 800 °C prior to testing for compressive strength (CS), flexural strength (FS), and splitting tensile strength (STS). Two-way ANOVA established that both WBA and temperature had statistically significant effects (p < 0.05) on all strength measures, with WBA accounting for the bulk of the variation. At 24 °C, augmenting WBA from 0% to 50% enhanced CS, FS, and STS by 37.26%, 40.63%, and 32.86%, respectively. Elevated temperatures diminished all strengths, with STS exhibiting the most significant relative decline, especially beyond 400 °C. response surface methodology (RSM) models exhibited exceptional prediction accuracy (R² > 0.97) and indicated that WBA mitigates strength loss due to elevated temperatures.

The construction sector is a huge contributor to carbon emissions in the world, mostly because of the heavy applications of Ordinary Portland Cement (OPC). Geopolymer concrete with low carbon content has become a viable alternative to concrete; nevertheless, accurate prediction and evaluation of its CO2 emissions have not been fully achieved, especially through combined experimental and data analysis. The literature on this study focuses mainly on mechanical performance, and there is limited understanding of environmental emission modeling. To measure and forecast CO2 emissions from low-carbon geopolymer concrete, this paper develops an experimental and machine-learning system. An array of geopolymer blends comprising fly ash and ground granulated blast furnace slag (GGBS) was developed at different binder proportions, alkaline activator ratios, and curing conditions. The amount of CO2 was determined using a cradle-to-gate evaluation method (kg CO2/m3). Data were subsequently trained on and tested on machine learning models, such as Artificial Neural Networks (ANN), Random Forest (RF), and Extreme Gradient Boosting (XGBoost) using experimental data. The R², RMSE, and MAE measures were used to assess model performance. Findings show that the XGBoost model has the best prediction accuracy (R² > 0.95), indicating good generalization ability. The highest CO2 emission reduction was 45–65 percent for geopolymer mixtures compared to traditional OPC concrete. The construct can offer a trustworthy decision support system in geopolymer concrete design that is environmentally friendly and further sustainability in construction methods with predictive modeling of emissions.