Strength prediction model considering material, ultrasonic and stress of cemented waste rock backfill for recycling gangue

Recycling gangue as the aggregate of cemented waste rock backfill (CWRB) to filling mining is the most economical, environmental and friendly method, which solves the environmental problems caused by gangue discharge and reduces the mining damages. Evaluating the strength property of CWRB is of great significance for maximizing the utilization of gangue recycling and improving the economic benefits of filling mining. In this paper, the triaxial compression and ultrasonic detection tests were carried out on 208 CWRB specimens to determine the relations between its strength parameters and ultrasonic pulse velocity (UPV). The coupling relations between the confining pressure, dosage of cementing material and particle size distribution (PSD) of aggregates and the strength parameters of CWRB under three-dimensional compression were established. Three models were proposed to predict the compressive strength, cohesion and internal friction angle of CWRB, in which Model I was built through the UPV, Model II was established based on the input of influencing parameters, Model III was constructed by combining material properties with on-site UPV and stress tests. The differences among these three models and their accuracies were discussed to obtain an optimal prediction model. The results show that the strength parameters of CWRB are positively correlated with its UPV, a modified exponential function can be used to describe those relationships. The strength parameters of CWRB increase with the confining pressure and dosage of cementing material, while increase firstly and then decrease with the PSD of aggregate particles, these coupling relationships between the multiple influencing factors and the strength parameters are optimized by the constructed genetic algorithm (GA). Compared with Model I using single UPV and Model II with the only input of influencing parameters, Model III combining ultrasonic technology that is conducive to field verification and intelligent technology GA for parameterizing work have higher evaluation accuracy, which eliminates the prediction distortion of the strength parameters of CWRB in three-dimensional stress state. This accurate and timely assessment is a pioneering work for ensuring the backfill stability, reducing the waste of the filling material and the filling costs. (C) 2020 Elsevier Ltd. All rights reserved.