Microwave Convective Drying and Predictive Modelling of Compacted Unbound Granular Materials

Drying of wet unbound granular pavement layers to a lower moisture content or degree of saturation (Sr) after compaction is one of the important processes in road construction. The efficiency of this so-called dry-back process is crucial for ensuring pavement stability before primer sealing. Traditional drying methods rely on solar radiation, which is slow and weather-dependent, leading to delays in construction timelines. Consequently, microwave drying has been explored as a rapid alternative; however, its effectiveness is limited by non-uniform moisture removal and shallow penetration depth (~ 50—60 mm). Hence, this study investigated the potential of microwave convective drying, integrating microwave energy with hot airflow to enhance the drying efficiency of a compacted unbound granular material (UGM). Laboratory-scale experiments were conducted to assess the influence of air temperature, airflow speed, airflow angle, and drying duration on moisture removal. Results showed that microwave convective drying improved surface drying efficiency compared to microwave drying alone, as evidenced by a larger dried area and wider surface temperature distribution. Nevertheless, both approaches exhibited similar limitations in deeper moisture removal, resulting in non-uniform drying. Two machine learning models, Support Vector Machine (SVM) and Gaussian Process Regression (GPR), were trained to predict the final Sr based on experimental data. The models predicted the drying trends reasonably well, with SVM yielding a mean absolute error (MAE) of 0.924% and GPR producing 0.006% in terms of Sr ratio. These findings highlight the benefits and limitations of microwave convective drying, suggesting the need for further optimisation for road construction applications.