Abstract:
Shotcrete has gained significant popularity as an effective support material for underground excavations owing to its fast application, early gain of strength and suitability for complicated excavations. There are, however, very few studies that have developed mathematical models for assessing the effectiveness of shotcrete support system including sustainable materials. In this study, an attempt has been made to develop a sustainable dolerite-Xorex fibre reinforced shotcrete system based on experimental, numerical and multivariate approach.
The rationale behind this research comes from the depletion of natural sand sources and the increase in mining wastes, mainly dolerite rocks. Fibre reinforced shotcretes improve the properties of shotcrete system; but ordinary steel fibres are not durable owing to their corrosive nature. The Xorex fibre system, which is a non-corrosive one, needs more investigation.
A total of 48 specimens were prepared using dolerite aggregate, Xorex fibres (0.02 wt%) and water reducing superplasticizer. The samples were then tested after 4 days, 7 days, 28 days and 56 days of curing using uniaxial compressive strength, triaxial compressive strength, tensile strength and shrinkage test. Numerical evaluation was done using Rocscience software such as Unwedge, RocSupport, RS2, and Ground Reaction Curve analysis. The Pearson correlation and multiple regression analyses were used to determine the main factors affecting the performance of shotcrete.
It was evident that there was progressive mechanical improvement due to increased curing time. The compressive strength increased from 10.85 MPa to 22.07 MPa, elastic modulus from 12.44 GPa to 20.34 GPa and tensile strength from 1.67 MPa to 4.21 MPa after 4 and 56 days. In the triaxial test, it was evident that there was improved confined behaviour where cohesion increased from 3.5 MPa to 5.3 MPa and friction angle from 46.6° to 51.6°. From the numerical modelling, it was clear that with increase in depth of excavation, there would be deformation and with increase in curing time,
vii
there would less displacement. The regression models have excellent prediction capability (R= 0.99937-0.99967; R²>0.998). Limitations of the study include lack of a control mixture, lower fibre content and using laboratory and numerical modelling. Future studies need to look at optimising the dosage of fibres, their durability and performance through field tests conducted underground.