- March 16, 2017
- Posted by: Ian Gray
- Category: News
The models and methodologies for determining rock properties and predicting rock behaviour are numerous, but are they all equally effective?
Designing tunnels, mines or rock slopes requires a model with which to predict their behaviour so that failure may be avoided.
Failure may be defined in all cases as unacceptable deformation, but what defines acceptable deformation? It depends on the application.
A tunnel for a mass transit rail system, for example, is likely to have a far lower tolerance of deformation than the back of a stope, where failure means collapse to the extent that it leads to dilution of the ore being mined.
Models of deformation are dependent on the properties of the rock.
Most of these models are based on the elastic analysis of a material that has a constant, single modulus up until the point of failure. This applies to numerical as well as to analytical models.
Where numerical models incorporate non-linear behaviour there is little experimental basis for determining what properties should be used.
Unfortunately, an engineer is restricted to designing on limited laboratory data, assisted by back analysis of structures that already exist.
The use of back analysis presumes that the engineer has an adequate model of stress-strain behaviour for which to derive suitable parameters. Such models are lacking.
Complicating the behaviour of rock is a general lack of understanding of the role that fluid plays in its behaviour.
While the concept of effective stress is generally applied in soils, it is not in rock—which is troubling, as it’s no less important.
Within rock there are significant differences between the rock matrix and joints, and the way in which they respond to fluid pressure.
What’s Wrong With Uniaxial Testing?
Frequently the only tests undertaken on rock to determine its properties are uniaxial tests for strength.
Some of these tests may be enhanced by measurement of strain, but they’re rarely presented to show how the material properties vary with stress levels.
Where triaxial testing is employed, it is for the purpose of determining the ultimate strength of the rock, rather than its constitutive properties prior to failure.
To overcome these shortcomings, Sigra has built equipment that permits it to load HQ (63.5 mm diameter), and HQ-3 (60.9 mm diameter) core with axial and confining pressure, and with internal fluid pressure while measuring the rock deformation.
A significant number of tests have been undertaken on siltstones and sandstones.
Young’s Modulus, Poisson’s Ratio
The findings are that Young’s modulus is highly dependent on the mean stress, while Poisson’s ratio is dependent on the shear and mean stress. Anisotropy can also be highly important, with some samples showing ratios of 1:3 for stiffness transverse to the bedding compared to that parallel to it.
The action of fluid pressure within the rock also needs to be determined. In some rock its effect approaches that of soil, with a Biot’s coefficient of near-unity, while in other it is virtually zero.
For a better understanding of the effects of fluid within rock, we strongly recommend the paper ‘Effective Stress in Rock’ by Dr Ian Gray, which was presented at the Eighth International Conference on Deep and High Stress Mining in Perth on 28 March.
Joints behave differently to the host rock. Their mineralogy and degree of openness mean that they have their own unique properties.
Accuracy in modeling and testing is absolutely crucial to any project. Sigra is equipped with the academic discipline and resources to accurately predict rock behaviour based on a variety of project types—namely civil tunnelling projects and mine design.