Response to macro-synthetic fibre critique Nov 2014

Dear TunnelTalk,

Thank you for providing me the opportunity to respond to the recent criticism published in TunnelTalk by Mr Pierre Rossi^[1] as Critique of synthetic-fiber FRS paper. The criticism concerned a paper I published at the WTC 2014 conference at Iguassu Falls in Brazil in May 2014^[2]. The original article is titled The Use of Macro-synthetic FRS for Safe Underground Hard Rock Support but unfortunately was not included with the criticism thereby making assessment by others somewhat difficult. I would like to respond to the criticisms levelled against that paper through the following series of points.

The article by Mr Rossi comes as no surprise given that he has published previous papers promoting steel fibres^[3] and is a consultant to Bekaert. His criticisms essentially mirror the stock standard arguments that steel fibre manufacturers have promoted for many years. However, there appears to be a discontinuity between the theory Mr Rossi likes to base his criticisms upon and reality, given that several million m³of macro-synthetic fibre reinforced shotcrete (MSFRS) is now successfully performing its role as ground support in underground hard rock excavations around the world, with a further 1.5 million m³ being added annually. In addition, recent civil tunnels in Norway have seen macro-synthetic fibres used in more than 80% of the shotcrete used for ground support, with the remainder comprising steel fibres or rebar. The great success that macro-synthetic FRS has enjoyed internationally suggests the basis of the steel fibre industry’s arguments against macro-synthetic FRS needs revision.

As noted in my paper at the WTC 2014, the main drivers for the adoption of macro-synthetic fibres as reinforcement in shotcrete for hard rock support have been its demonstrated performance in the field, together with economic and corrosion-based advantages over steel fibre reinforcement. Its effectiveness in the field has defied the many naysayers who fret about alleged creep problems. The simple fact is that engineers employed in the hundreds of mines and civil tunnels internationally that now use MSFRS do not care what material the reinforcement is made of provided it performs the roles required of it. But unlike engineers in the field, advocates for steel fibres prefer to rely on theory as a basis for claims against macro-synthetics rather than actual performance in the field. Numerous papers have also been published on the problems steel FRS has exhibited in relation to corrosion at cracks or embrittlement in concrete of 40MPa 28 day strength and upwards[4-9], yet these studies are conveniently ignored.

In relation to technical criticisms, one can firstly ignore the comments about ‘retaining walls’ and 'ductility in the interior of cracks' since neither of these topics was mentioned in the WTC 2014 paper. However, Mr Rossi made the following explicit claim in relation to macro-synthetic FRS:

"With this type of fiber reinforced concrete, the only way to ensure the stress hardening behaviour of the material, and therefore to permit redistribution of forces, is to have a hyperstatic mechanical system. This is the case when the underground support interacts with rocky ground with hard rocks."

This passage implies that ‘stress-hardening behaviour’ is a requirement for shotcrete to perform successfully as ground support in hard rock applications. I challenge anyone to point out an example of the successful use of a ‘stress hardening’ FRS anywhere in the world! One of the most successful and widely used approaches to the design of ground support with FRS is the Q-system^{[10, 11]}, which has been shown to work very effectively with strain-softening FRS exhibiting an energy absorption of only 1,000 Joules at 25mm deflection in the EN14488.5 square panel test^[12] (or 400 Joules at 40mm deflection in the ASTM C1550 round panel test^[13]). There is no requirement for stress hardening behaviour in this or any other widely used design method for hard rock FRS linings. This is one of the reasons why dosages of macro-synthetic fibres are seldom required to exceed 7kg/m³, and steel fibre dosages seldom exceed 40kg/m3, to achieve satisfactory support of hard rock. In shotcrete there is a diminishing return in post-crack performance as fibre dosage increases since fibre rebound increases rapidly as dosages exceed the levels indicated above. Dosages of 50-60kg/m³ of steel fibres are also exceedingly difficult to spray, which anyone familiar with fibre reinforced shotcrete would know.

On embrittlement (or post-crack performance loss with age), Mr Rossi says:

"I assert that the steel fibers usually used in shotcrete, having a length of around 30mm, can be used in a matrix for which compressive strength can reach 90MPa with only a small percentage of them breaking."

This unsubstantiated claim flies in the face of numerous experimental studies conducted on both shotcrete and cast FRC with steel fibres, incorporating thousands of test specimens, including recent research published at the WTC 2014 in Brazil^[9], an international symposium on shotcrete held in Norway in June 2014^[8], and a tunnelling conference held in Sydney in September 2014^[7]. One has to ask the question: how many times does the same investigation have to be performed, resulting in the same damaging outcome, before the reality is accepted that steel FRC loses performance with age? The continued denial by the steel fibre industry of the insidious problem of performance loss with age reminds me of the tobacco industry’s response to lung cancer in the 1970s.

In relation to corrosion, I agree with Mr Rossi’s claim that the small diameter of steel fibres means that disruption of the concrete matrix is minimal when corrosion occurs. However, the small diameter also means that only a shallow degree of surface corrosion is required before substantial loss of fibre cross-section occurs, leading to rapid loss of structural capacity at cracks, as demonstrated by Kaufmann in his recent research performed in Switzerland^[9], and Nordstrom in earlier work performed in Sweden^[14]. It merely needs to be pointed out that ground water leaks through shotcrete tunnel linings in road tunnels or metro tunnels are widely observed and evident to most members of the travelling public. Cracks can therefore allow water to pass through a lining directly to exposed fibres at a crack. The claim that cracks in shotcrete linings heal themselves and are therefore nothing to worry about is commercially-driven wishful thinking. Pressure gradients give rise to transport of oxygen and aggressive agents directly to steel fibres at cracks rather than via slow diffusion-based transport mechanisms. This is why corrosion occurs much faster in real tunnels than in lab tests.

In relation to creep, users of steel fibres remain constrained by crack widths to 0.30mm or less because of concerns about corrosion of steel and mistakenly assume the same constraints apply to MSFRS. These concerns are misplaced because there is no steel in a MSFRS lining, and thus there is no need for crack width limits related to corrosion. Use of macro-synthetic FRS is a 'liberation' from the constraints implied by the use of old corrosion-sensitive technology like steel reinforcement. The underground mining industry discovered this many years ago and has consigned steel fibres and mesh to the dustbin of ‘technologies we no longer need’^[15]! The more conservative civil tunnelling industry is taking longer to come to the same conclusion but is clearly moving in the same direction as evidenced by the increasing number of civil tunnels using MSFRS in preference to steel reinforcement^{[16, 17]}.

In summary, the criticisms levelled by Mr Rossi represent the viewpoint of steel fibre advocates that are primarily based on theoretical arguments against macro-synthetic fibres that do not match the reality seen in hard rock ground support internationally. The reason for this reliance on theoretical arguments is that these same advocates cannot point to any actual tunnels lined with MSFRS that have suffered distress in the form of excessive convergence or any other form of unsatisfactory in-service performance of the type they have frequently claimed will happen. To the contrary, macro synthetic FRS has performed very well, justifying its near universal adoption in underground mining in countries such as Australia, and rapidly advancing adoption elsewhere^[15]. Steel fibres were an advance over alternatives when they first emerged commercially in the 1970s, but macro-synthetics represent a further advance in fibre technology by providing complete protection against corrosion and other serviceability-related concerns.