This study proposes a hybrid generative–ensemble framework to predict key mechanical properties of recycled aggregate concrete from mix proportions. An established database of 112 mixes was used to model compressive strength, split tensile strength, flexural strength, and elastic modulus. To mitigate data scarcity, a conditional variational autoencoder was trained on the training data only and used to generate additional physically plausible input samples, after which seven supervised learning algorithms were trained and compared using cross-validation. Gradient boosting and support vector regression achieved the most accurate and stable predictions across all targets, outperforming baseline linear models and commonly used empirical correlations. Feature-attribution analysis was used to identify the dominant drivers of each property, showing that binder-related variables primarily govern strength, while aggregate-related variables dominate stiffness. The results support practical, data-driven screening of recycled aggregate concrete mixes and provide interpretable guidance for sustainable mix design.

Incorporating recycled polyethylene terephthalate (PET) into asphalt mixtures offers a sustainable approach to enhance pavement performance while reducing plastic waste. However, the mesoscale mechanisms governing the influence of PET on stiffness, deformation resistance, and fracture behavior remain unclear. In this study, a three-dimensional Discrete Element Method (DEM) framework was developed to investigate the constitutive response of PET-modified asphalt concrete through the explicit representation of aggregates, asphalt mortar, PET inclusions, and air voids. Two bonding schemes, the Contact Bond Model (CBM) and Parallel Bond Model (PBM), were implemented and compared in terms of stiffness, tensile strength, damage evolution, and crack propagation. The experimental dynamic modulus (|E*|), indirect tensile strength (ITS), resilient modulus (Mr), rutting, and moisture susceptibility tests were conducted for mixtures containing 0–10% PET by volume. The DEM microparameters were calibrated using |E*| and ITS data, whereas Mr, rut depth, and tensile strength ratio (TSR) were used for independent validation. The results show that PET incorporation increases the mixture stiffness, with the dynamic modulus rising from 3500 to 5159 MPa and improves the resilient response under repeated loading. ITS increased from 0.44 MPa for the control mixture to a peak value of 1.15 MPa at 6% PET before decreasing to 0.89 MPa at 10% PET due to interfacial weakening. The rut depth decreased consistently with increasing PET content, indicating enhanced resistance to permanent deformation, whereas the TSR values confirmed acceptable moisture durability. Mesoscale analyses revealed that PET modified the force-chain distribution and promoted interface-controlled damage at the PET–mortar contacts. Compared with CBM, PBM more accurately reproduces progressive stiffness degradation and distributed cracking. An optimum PET content of approximately 6% was identified, providing the best balance between stiffness enhancement, tensile resistance and durability. These findings provide mechanistic insights into PET-modified asphalt mixtures and support the development of performance-based sustainable pavement materials.

The increasing generation of plastic waste and the growing demand for sustainable pavement materials have encouraged the incorporation of recycled polymers into asphalt mixtures. This study evaluates the engineering performance, microstructural characteristics, numerical response, and preliminary environmental implications of recycled polyethylene terephthalate (RPET)-modified asphalt concrete. RPET obtained from post-consumer plastic bottles was incorporated into asphalt mixtures through the dry process at dosages of 0–9% by weight of binder. Marshall stability, indirect tensile strength (ITS), repeated load dynamic creep (RLDC), scanning electron microscopy (SEM), and finite element modelling (FEM) were employed to assess the influence of RPET content on mixture behavior. Experimental results showed that increasing RPET content improved stiffness-related properties and rutting resistance. Marshall stability increased from 5.5 kN for the control mixture to 14.3 kN at 9% RPET, while ITS increased from 0.72 MPa to 1.02 MPa. RLDC results indicated a reduction in accumulated permanent strain from 3.20% to 1.85%, demonstrating enhanced resistance to deformation under repeated loading. SEM observations revealed comparatively uniform RPET dispersion at moderate dosages (3–5%), whereas higher contents showed localized particle agglomeration. FEM simulations demonstrated reduced surface deflection and improved stress distribution with increasing RPET-related stiffness. Preliminary life cycle assessment indicated modest embodied carbon reduction and potential cost savings. The findings suggest that RPET incorporation can enhance the mechanical and deformation-resistant characteristics of asphalt mixtures while contributing to plastic waste valorization and sustainability objectives. However, the results should be interpreted as comparative laboratory and numerical indicators rather than direct predictors of long-term field performance.