Response surface methodology-based characterization and optimization of fibre reinforced cemented tailings backfill with Slag

ABSTRACT

With the increasing depths of underground mines due to the scarcity of near-surface ores, the introduction of fibre-reinforced cemented paste backfill (F-CPB) has emerged as a novel solution to address the demanding geomechanical conditions in deep mining operations. However, the widespread adoption of F-CPB in mining and industrial backfill operations necessitates a comprehensive understanding of its key engineering properties, including strength, yield stress, modulus of elasticity, and cost. Moreover, it is crucial to investigate the influence of the constituent materials (water, fibres, binders, tailings) and their interactions on these properties. This research paper presents the application of response surface methodology (RSM) to model the effects of binder content (Portland cement/Slag), water content, fibre content, tailings and their interactions on the mechanical and rheological properties, as well as the cost of F-CPB. Central Composite Design (CCD) experiments were conducted, and a high degree of agreement was observed between the experimental and predicted responses. The RSM approach proves suitable for accurately estimating the responses and assessing the interactions between the model parameters and the properties of F-CPB. Furthermore, a combination of RSM and the desirability approach enables the development of an optimisation tool for F-CPB, facilitating the formulation of optimal backfill mixtures. The results obtained from this study highlight the effectiveness of the combined RSM and desirability approach in F-CPB mix proportioning, offering an advanced engineering approach to F-CPB mix design. The proposed design method has the potential to reduce the laboratory testing protocol required for determining the optimal mix composition.

KEYWORDS:

• Cemented paste backfill
• tailings
• fibre
• deep mine
• response surface method
• optimisation

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Natural Sciences and Engineering Research Council of Canada; University of Ottawa.