Modeling the effects of sulphate and curing temperature on the strength of cemented paste backfill using artificial neural networks

Abstract

The effects of sulphate and curing temperature play an important role on the strength development and durability of Cemented Paste Backfill (CPBs). Depending on the application of the CPB, different strength values, measured as unconfined compressive strength (UCS), are targeted. There is a lack of proven theory to predict the UCS for a specific CPB mix due to the complexity of the interactions between the variables that affect the CPB strength. This thesis presents an approach to use the artificial neural network (ANN) methodology in order to develop two models that can predict the effects of sulphate and curing temperature on the UCS of CPBs. The ANN models here developed illustrate an outstanding accuracy in the UCS prediction for the simulation of sulphate and its coupled effect with curing temperature. The ANN models provide a better understanding of the effects of sulphate and/or temperature on the strength of CPBs.