Ground Control in Hard Rock Using Fibre Reinforced Shotcrete

Dr. E. Stefan Bernard, TSE P/L, Australia.

Introduction

Fibre Reinforced Shotcrete (FRS) linings have been used for ground support in tunnels and mines for many years. Several guides exist on the use of shotcrete in general construction [1-3] and these provide advice on issues ranging from mix design to spraying technique. However, load determination and resistance design for ground support in underground applications have seldom been addressed in a comprehensive manner. FRS linings for civil tunnels are commonly designed using methods such as that provided by the DBV [4], but ground control using FRS linings in mines is more commonly based on simple methods developed from basic engineering principles [5, 6]. This paper provides a review of methods of design for FRS linings based on the limited number of available papers and guides together with experience gained in the Australian mining industry.

The reasons why FRS linings are used for ground support include the demonstrated ability of shotcrete linings and rockbolts to effectively stabilize a wide range of ground conditions with little more that a variation in strength, toughness, or geometry necessary to accommodation different levels of stability. This makes FRS linings possibly the most adaptable method of ground control available. However, there are limits to its effectiveness and economy that mean it is not suitable for all circumstances. Another motivation for the use of shotcrete is that it conforms to the profile of the ground and does not require formwork. This increases its effectiveness for immediate support of ground and greatly reduces construction costs. Indeed, once the initial hurdle of establishing a capacity for delivery and spraying is overcome, shotcrete has been found to be economical when compared to alternatives in most ground conditions, especially when varying conditions or geometry are encountered. In addition, the durability of shotcrete is good compared to alternatives, although problems of corrosion remain a concern for steel fibres and conventional reinforcement, especially at cracks.

Stabilization of Ground

To understand the strengths and limitations of FRS in hard rock tunnels, it is important to understand the means by which shotcrete functions to stabilize the ground. When an opening is first excavated, stresses in the surrounding ground are redistributed leading to a relaxation in the immediate vicinity of the opening. This generally leads to an expansion and unravelling of the surface of the excavation which can, in many cases, cause the opening to converge and eventually collapse. Depending on the degree of latent instability, FRS applied immediately to the freshly excavated ground can secure the rock surface, limit unravelling and, to a lesser extent, control relaxation of surrounding rock (Figure 1). When combined with rock bolts of appropriate length and stiffness, the zone of relaxation and potential instability around the opening can usually be secured against further movement for protracted periods of time (Figure 2).


FRS linings do not 'hold the ground up' in hard rock applications but instead secure locally unstable regions, primarily between the rock bolts, so the ground supports itself. The lining does this through a combination of flexural, shear, and tensile action depending on ground conditions. To work effectively, it must be well bonded to the surface, possess sufficient strength and toughness, and must not degrade over the design life of the opening. Alternative ground control strategies, such as the use of mesh in the absence of shotcrete, simply catch falling rock fragments and do not control the zone of relaxation. However, in highly unstable ground, mesh placed over a tough FRS lining has often proven to be an effective means of providing short to medium-term stabilization.

Design Considerations

Very few ground conditions are 'impossible' to excavate through, but the cost of doing so can vary enormously. It is therefore necessary to plan the route of an excavation carefully and be prepared to modify this if poor conditions are found. In civil applications, weathered ground and sediments often represent problematic areas and considerable attention is paid to the transition zone between such regions. In mines, the problem of ground high stress and severe re-distribution of stress due to deep excavation or nearby stoping, flat-backing, etc. generally pose the most significant challenges. In civil tunnels, rehabilitation is very expensive and inconvenient, thus every effort is made to permanently stabilize the ground during construction. In mines, rehabilitation is inconvenient but nevertheless possible, and can usually be implemented if and when required. Long-term access ways are therefore regularly monitored for signs that rehabilitation is necessary. The widespread use of regular monitoring and feedback on ground control effectiveness is one aspect of shotcrete use in mines that has contributed substantially to the rapid evolution of ground control strategies involving FRS in Australia. In combination with a flexible regulatory environment, and the profitability of the metalliferous mining industry, this has helped make Australia one of the world's leading innovators in shotcreting practices and technology.

One of the primary motivators for the use of FRS for ground control in hard rock is its adaptability and effectiveness across a wide range of conditions. Without changing equipment or supplier, a geotechnical engineer can select bolts of different types, lengths, and spacing, and then specify FRS of a suitable thickness that can readily be varied to achieve effective stabilization. High utilization rates for staff, spraying rigs, and associated plant results in improved economy. This combination of adaptability and economy is difficult to match using any alternative means of ground control.

Apart from utilization rate, other factors that influence the economy of ground control using FRS obviously include the cost of materials and their performance efficiency. Fibres and set accelerators constitute a large proportion of the total material cost, and their influence on performance is readily assessable, so particular attention should be focused on the efficacy of these ingredients. Rebound and the quality of spraying technique are other important issues, and the water demand of aggregates used represents a common problem. If these issues can be brought under control, then good overall economy should be readily achieved.

Ultimate Strength

The peak load resistance of a FRS lining can be compared to the ultimate strength requirements commonly imposed on other civil engineering structures. The load resistance of a lining depends on many factors, including the strength of materials used, their toughness and thickness, the characteristics of bolts used in combination with the FRS, the bond strength between lining and substrate, and interactions between the ground and lining. The complexity of these interactions is so involved that fully engineered design methods for the ultimate strength FRS linings have yet to be developed. Instead, the designer is dependent on either an observational approach to ground support, in which the thickness and toughness of the FRS are varied together with bolt lengths and spacing, or several simple deterministic design equations intended to model localized failure modes.

The observational approach to design is commonly based on the Q-system developed and expanded upon by Barton and co-workers over many years [7-9]. A minimum grade and thickness of FRS is selected according to the so-called Q-chart that represents many years of accumulated experience regarding effective ground control. Local spot bolting and thickening of the lining is commonly used in response to localized poor ground that must be identified through constant monitoring as excavation proceeds. There are no deterministic models of load resistance used in this approach, and the actual margin of safety against collapse is unknown. This method is simple and conservative, but suffers the problem that innovations in material performance and ground control strategies cannot be fully exploited because it is based on experience gained using 'old' techniques and materials. A designer that relies too strongly on this approach is therefore bound to remain 'stuck in the past'. Moreover, the experience underlying this approach is somewhat Euro-centric and does not extend to deep high stress environments. This approach should therefore be regarded as a conservative starting point for ground control.

If a designer wishes to go beyond the conservatism of the Q-chart then there are several options available including alternative observational approaches, numerical analyses, or simple deterministic models of ground and lining behaviour.. A tried-and-tested method of specifying FRS performance for temporary linings is outlined in the Guide to Shotcrete in Australia [3]. This approach is similar in concept to the Barton approach but incorporates a lower level of conservatism and uses FRS with higher levels of toughness. Numerical methods will not be covered here, but it must be noted that most numerical methods have not been calibrated against field observations and therefore should be used cautiously. Simple models of localized lining failure have been reviewed by Barrett & McCreath [4] and Pells & Bertuzzi [5] and are described below. It must also be noted that all the simple models are based on the premise that rock bolts constitute the principle means of global ground control and that a FRS lining only supports local loads between the rock bolts. In common with many conventional ultimate strength analyses for suspended structures, these models are based on yield line theory and require substantial ductility to exist in a lining in order that stress actions be redistributed when the peak load resistance is sustained. Brittle FRS should not be used with these methods as the peak load resistance of brittle materials is lower than predicted by yield line theory [10].

The first and most common ultimate strength failure mode usually considered involves flexural failure under localized loads that may comprise either an unstable block or region of fractured ground (Figure 3). The lining is generally assumed to be de-bonded and therefore acts in simple bending rather than as a composite beam incorporating elements of rock acting in compression. This conservative case normally governs the thickness and strength requirements for a lining.

A second mode of lining failure that must be considered is out-of-plane shearing of the lining (Figure 4). This has been observed in the field to occur most commonly during the first few hours after spraying and is thus an important issue when safe re-entry times are considered. The punching resistance of a lining is found using expressions commonly used for suspended reinforced concrete floors except that a more realistic model of the shear strength of shotcrete [11] based on field tests is used. It should be noted that the shear strength of FRS is not affected by the type or dosage rate of fibre used for economically viable grades of shotcrete. Moreover, essentially all FRS displays strongly strain-softening behaviour in a punching mode hence it is important to avoid this type of lining failure.

Serviceability

Serviceability requirements for a FRS lining in hard rock are normally much more stringent for civil tunnel applications than in mines, and extend to issues including water-tightness, acceptability of cracking, surface roughness, lighting, and appearance. Of these issues, water tightness is normally the most difficult to achieve and can govern the design of the lining. Linings of normal thickness (100 mm) cannot support a substantial hydrostatic head, and thus the provision of adequate and long-lasting drainage behind the lining to control inflow is essential. Cracks in a lining are the most common route for water to ingress, thus it is unrealistic to believe that inflow can be controlled through use of low permeability concrete. Measures to limit the incidence of cracking may be necessary, such as the use of low shrinkage concrete and full stabilization of ground movement before the final lining is applied. Attention should also focus on the elastic modulus and ductility of the FRS since high strength mixes are also more rigid, less ductile, and creep less than low strength mixes, and thus are more prone to cracking in the event of continued ground movement.

The durability of FRS is normally considered in the context of either the concrete matrix or the fibre reinforcement. The durability of a shotcrete matrix has only occasionally been studied specifically because an extensive amount of information is available based on studies of cast concrete suggesting that the high cement concrete and low permeability typical of shotcrete will lead to good durability. The most frequent incidence of shotcrete matrix degradation reported is in environments combining freeze-thaw action and deicer salts. General experience suggests that the matrix durability of most shotcrete mixes is excellent.

The situation concerning the durability of fibres is more ambiguous. A substantial number of studies have examined corrosion and degradation of post-crack performance in FRS both in the cracked and uncracked states. These have indicated that uncracked steel FRS suffers little damage due to corrosion of fibres other than surface staining [12-14], but may suffer serious embrittlement over time as a result of increasing anchorage strength around fibres [15, 16]. This problem is most apparent in high strength grades of shotcrete and can lead to a substantial loss of toughness at large deflections. Low deformation performance is not affected by this phenomenon. Embrittlement is of concern in situations where ductility is required to be retained in a structure for extended periods of time. For example, if seismic activity or a substantial nearby underground construction are anticipated over the life of a FRS lining then a mix must be used that has a demonstrated capacity to retain ductility at large deflections over many years. Unfortunately, most steel FRS does not conform to these requirements and the user should then consider macro-synthetic FRS linings as these have been demonstrated to retain toughness with age (see Figure 5). Post-crack performance for steel FRS may fall substantially after 28-56 days age, but macro-synthetic FRS linings have been shown to steadily increase in performance with age and are therefore highly suitable for seismic environments.


Corrosion of steel fibres at cracks greater than 0.20-0.30 mm width is also a serious cause of performance loss in steel FRS exposed to exterior environments. Crevice crack corrosion, even in inland environments free of salt, can lead to complete loss of steel fibre continuity after only a few years [17]. Experience of steel FRS corrosion in Scandinavian tunnels [18] has led to steel fibres being banned in sub-sea tunnels in Norway and restrictions have been placed on their use in Sweden.

References

• European Specification for Sprayed Concrete, 1996. European Federation of National Associations of Specialist Contractors and Material Suppliers for the Construction Industry (EFNARC).
• American Concrete Institute 506R-05 Guide to Shotcrete, ACI, Farmington Hills, 2005
• Guide to Shotcrete in Australia, 2^nd Edition, Australian Shotcrete Society/CIA, 2010.
• DBV 2001. Design Principles of Steel Fibre Reinforced Concrete for Tunnelling Works, Deutscher Beton-Verein.
• Barrett, S. & McCreath, D.R. "Shotcrete Support Design in Blocky Ground - Towards a Deterministic Approach," Tunnels and Deep Space, 10(1), pp79-88, 1995.
• Pells, P. & Bertuzzi, R. "Rock Engineering for Tunnels and Underground Structures Short Course," UNSW, 30 Aug.- 1 Sept. 2004.
• Barton, N. Lien, R. & Lunde, J., "Engineering classification of rock masses for design of tunnel support," in Rock Mechanics, 6(4), pp189-236, 1974.
• Barton, N. & Grimstad, E. 1994. "The Q-System following twenty years of application in NMT support selection." Felsbau 12 (6), 423-436.
• Barton, N. & Grimstad, E. 2004. "The Q-system following thirty years of development and application in tunneling projects." In Rock Engineering - Theory and Practice, Proceedings of the ISRM Regional Symposium EUROCK 2004. Salzburg, Austria 2004. Pp. 15-18.
• Bernard, E.S. 2006. "Influence of toughness on the apparent cracking load of Fibre Reinforced Concrete slabs," ASCE Journal of Structural Engineering, Vol. 132, No. 12, p1-8.
• Bernard, E.S. 2008, "Early-age load resistance of fibre reinforced shotcrete linings," Tunnelling and Underground Space Technology, 23, pp451-460.
• Lankard, D.R. & Walker, H.J. 1978. "Laboratory and field investigations of the durability of Wirand concrete exposed to various service environments," Battelle Development Corp., Columbus Laboratories, Ohio, 26p.
• Kosa, K. & Naaman, A.E., 1990. "Corrosion of Steel Fiber Reinforced Concrete," ACI Materials, 87 (1), Jan-Feb, pp27-37.
• Schupack, M. 1985. "Durability of SFRC exposed to severe environments," Proceedings, U.S.-Sweden Joint Seminar on Steel Fiber Concrete, Swedish Cement and Concrete Research Institute, Stockholm, pp479-496.
• Bernard, E.S., & Hanke, S.A. "Age-dependent behaviour of fibre-reinforced shotcrete," Proceedings of the Fourth International Symposium on Sprayed Concrete, Davos, Switzerland, 22-26 September, 2002, pp11-25.
• Bernard, E.S., 2008. "Embrittlement of Fibre Reinforced Shotcrete," Shotcrete, Vol. 10, No. 3, pp16-21.
• Bernard, E.S. 2004. "Durability of cracked fibre reinforced shotcrete," Shotcrete: More Engineering Developments, Bernard (ed.), pp 59-66, Taylor & Francis, London.
• Nordström, E., 2001. "Durability of steel fibre reinforced shotcrete with regard to corrosion," Shotcrete: Engineering Developments, Bernard (ed.), pp213-217, Swets & Zeitlinger, Lisse.