Classification SystemsPublished on February 15 2015
Classification systems are probably among the most frequently tools used in the design of underground excavations in rock. The primary object of all rock mass classification systems is to quantify different engineering properties of, or related to, the rock mass based on past experience. Their main core is the assessment of the rock mass quality that, preferably, can be used as an indicator for rock engineering. From a set of parameters of the rock, rockmass, groundwater and stresses, the quality of a rockmass with respect to strength, deformability and stability can be estimated.
Another important task is that they serve as a kind of checklist.
Three different types of output can be distinguished from the rock mass classification systems:
- Characterisation of the rock mass expressed as overall rock mass quality, incorporating the combined effects of different geological parameters and their relative importance for the overall condition of a rock mass. This enables the comparison of rock mass conditions throughout the site and delineation of regions of the rock mass from 'very good' to 'very poor', thus providing a map of rock mass quality boundaries.
- Empirical design with guidelines for tunnel support compatible with rock mass quality and the method of excavation. Traditionally, this is often seen as the major benefit from the use of rock mass classification systems.
- Estimates of rock mass properties. Rock mass characterisation expressed as an overall rock mass quality has been found useful for estimating the in situ modulus of rock mass deformability and the rock mass strength to be used in different types of design calculations.
- The accuracy of these existing empirical design methods is not established.
- Contractual problems are often created when unforeseen geological conditions have been encountered, and where the system has not been applicable. Typical conditions that are not covered sufficiently are swelling, squeezing, ravelling, flowing, or popping ground.
In the early stages of a project, the existing quantitative rock mass classification systems (empirical design methods) can be applied as a useful tool to establish a preliminary design. At least two systems should be applied (Bieniawski, 1984, 1989). Classification systems are, however, unreliable for rock support determinations during construction, as local geometric and geological features may override the rock mass quality defined by the classification system. Restrictions on their use here are also pointed out by Bieniawski (1997).
Several classification systems have been developed over time. Some have been developed specifically for a project. Classification systems are probably among the most frequently used tools in support design of underground excavations in rock.
To view a paper on the ability and use of classification systems, see the paper 'Classification as a tool in rock engineering'.
2. The RMR classification system
The RMR (or Geomechanics) system was launched by Bieniawski in 1973. It was a further development of the RSR system by Wickham, Tiedemann and Skinner (1972). Later, the system has been revised/updated by Bieniawski in 1974, 1975, 1976 and 1989. A short description of the RMR is shown here.
Numerous papers have been presented and published on the RMR system and the system is currently being used by many practitioners. The input parameters to RMR and the RMR support table are shown here.
In applying this classification system, the rock masses are divided into a number of structural regions. The boundaries of the structural regions usually coincide with major structural features (Bieniawski 1984, 1989). Bieniawski strongly emphasises that a great deal of judgement is needed in the application of rock mass classification to support design.
The RMR value has also been used to estimate rock mass properties. Bieniawski (1984, 1989) and Serafim and Pereira (1983) have given a relationship between the RMR and the rock mass deformation modulus (Em). The RMR value is also used as one way to estimate the m and s factors in the Hoek Brown failure criterion (Wood, 1991; Hoek, 1994; Hoek and Brown, 1997) as well as the GSI value to evaluate the rock mass strength properties.
The RMR-values can be estimated using a computer spreadsheet together with Q-values and RMi-values.
3. The Q classification system
The Q system was launched in 1974 by Nick Barton, Reidar Lien and Johny Lunde of the Norwegian Geotechnical Institute (NGI). The Q-system is as an empirical design method for rock support. Together with the ratio between the span or height of the opening and an excavation support ratio (ESR), the Q value defines the rock support. This is described here.
Over the years, some developments have been introduced by:
- a new design of the support diagram (Grimstad and Barton, 1988), see list of published paper here.
- adjustments of the input parameter SRF (Grimstad and Barton, 1993), inclusion of the rock strength (Barton, 1995), inclusion of shotcrete ribs as a support element (Barton and Grimstad, 2004).
A paper presenting Q and NATM systems gives some information on the two systems.
Grimstad and Barton (1993) have also presented an equation to use the Q value to estimate the rock mass deformation modulus (for values of Q > 1). The Q value is also used as one method to estimate the m and s factors in the Hoek Brown failure criterion (Hoek, 1983; Hoek and Brown, 1988).
In 1999 the originator of the Q system has increased the limitation of the original Q to also incorporate excavation by TBM (tunnel boring machines) introducing QTBM (Barton, 1999) and to also use Q-values for estimates of the effect of grouting (Barton et al, 2001), estimate Q-values from sound velocities in the ground (found from refraction seismic measurements).
Some comments on the limitations of the use of classification systems are presented in the paper 'Use and misuse of classification systems with special reference to the Q system'. This paper was first presented at the annual Norwegian conference on Rock and Soil excavation in 2002. A summary in English of this paper can be read here.
The Q-values can be estimated using a computer spreadsheet (/files/Q-RMR-RMi_v2-1.xls) together with RMR-values and RMi-values. See also Section 6 below.
4. The RMi classification system
The rock mass index, RMi, is a volumetric parameter indicating the approximate uniaxial compressive strength of a rock mass. The system was first presented by Palmström (1995) in his PhD. and has been further developed and presented in several different papers, see the tag 'Why RMi?'
The application of RMi is two-fold:
- RMi is an approximate value for the compressive strength of the rockmass (MPa)
- RMI is used as input to other calculations or estimates of rock properties, like rock support, TBM progress assessment, deformation modulus of rockmass
RMi makes use of the uniaxial compressive strength of intact rock (UCS) and the reducing effect of the joints penetrating the rock (JP) has. This is expressed as RMi = UCS x JP
A short introduction to the RMi system is shown here and a more detailed description here where also the input parameters to RMi are shown. RMi requires more calculation than the RMR and the Q system, but spreadsheets can be used from which RMi-values can be found together with RMR-values and Q-values from the same input observations, see also Section 6 below.
Based on RMi combined with the geometrical features of the underground opening and rock stresses, the rock support can be estimated using support charts or from theexcel spreadsheet ( /files/rmi_support_3-1.xls). RMi can also be applied as input to other rock engineering methods, such as numerical modelling, the Hoek-Brown failure criterion for rock masses. The paper 'Deformation modulus of rock masses' shows how RMi is used.
The system applies best to massive and jointed rock masses where the joints in the various sets have similar properties. It may also be used as a first check for support in faults and weakness zones, but its limitations here are pointed out by Palmström (1995).
For special ground conditions like swelling, squeezing, ravelling, the rock support should be evaluated separately for each and every case. Other features to be separately assessed are connected to project specific requirements such as the life-time of the project.
There is more to be found on RMi in the tag 'Ph.D. Thesis on RMi'.
5. Other classification/characterization systems
Several other classification systems have been developed over time. Some have been developed specifically for a project.
The first one to present a practical classification system for use in rock engineering was Dr. Karl Terzaghi when he in 1946 wrote the famous article Introduction to tunnel geology (Terzaghi, 1946) as a chapter in the book 'Rock Tunneling with steel supports' by Robert V. Proctor and Thomas L. White after visits to several tunnelling projects in America and in Europe. Many of these projects experienced tunnel excavation and stability problems.
In the 1940s, the rock support was often performed by steel arches, on which the support evaluation in the Terzaghi system was developed. Today, the Terzaghi classification system has lost much of its interest. However, the engineering geological part of the book is very interesting and pinpointing important geological features with respect to tunnel construction.
The NATM (New Austrian Tunnelling Method) is a sequence of tunnel planning, design and follow-up, and hence not a classification system.
6. Combination of RMR, Q and RMi classification systems
In two papers Comparing the RMR, Q and RMi classification systems, a Comparison on the RMR, Q and RMi systems is presented in Part-1 (files/correlation_Q-RMR-RMi-1.pdf) and Part-2 (files/correlation_Q-RMR-RMi-2.pdf). Shorter versions are presented in the papers 'Combining the RMR, Q and RMi classification systems' and in a Technical note.
An overview of the input parameters used in the RMR, Q, and RMi systems are presented here ( /files/parameter_class-systems.pdf)
A computer spreadsheet has been worked out where the RMR, Q and RMi values are calculated independently by combining the input parameters as described in the papers mentioned above.
See also references on classification systems.