The gathering of opinions and evidence was commenced with the issue of the independently verified questionnaire sent out to both specifiers and suppliers covering those fibre properties most often quoted as advantages or disadvantages of the two main types. This asked for opinions of which properties were relevant, advantageous or disadvantageous, for each in the applications of sprayed concrete, pre-cast lining egments, and cast in situ linings. Respondents were also invited to expand on the views with more reasoning and to provide any supporting evidence in the form of relevant testing or national and international standards met.
Important properties
Supplier respondents to the questionnaire usually indicated that major properties were mportant whatever the type of fibre, albeit perhaps from different perspectives resulting in major differences of opinion. Promotors of steel fibres questioned the design life of synthetic fibres, maintaining that this is not proven, and the validity of the values of Young’s Modulus of Elasticity (E) for the synthetic fibres. The latter is applicable to stretching of the fibre or to the bending of a (test) beam and is defined as the ratio of stress (tensile or compressive) over strain (tensile or compressive). Steel fibre proponents draw attention to the different Young’s Modulus values for concrete (around 30GPa), steel (200GPa) and synthetic fibres (about 9GPa). However Andrew Ridout of ElastoPlastic Concrete (EPC) says, “The Young’s Modulus is not an issue with fibre-reinforced concrete as this is a composite material. As such it is the performance of all the components working together to produce a performance that is of interest.”
Environmental
On the hazards that fibres may face once in place, both camps maintain that the other is not effective (alone) in a tunnel fire. In the case of macrosynthetic fibres it is claimed by steel fibre producers that when they melt in the high temperatures, all reinforcement effect is lost. On the other hand, the beneficial effect of (nonreinforcing) microsynthetic fibres is widely accepted in that their melting creates microscopic passages for the release of vapour build-up in the heated concrete, thus deterring spalling damage. Therefore they are recommended for use whatever the type of reinforcement. For their part, synthetic fibre producers claim that steel fibres can actually aggravate the spalling problem through heat transmission and expansion, whereas macrosynthetic forms have similar effects to microsynthetics.
It is in the long-term (time-dependent) properties of the fibres in situ that there are major disagreements, leading to them being the subject of several programmes of research. One such property is corrosion. It is well known that synthetic (polymer) fibres are resistant to most acid or alkaline environments. Synthetic fibre producers cite rusting or other corrosion of normal steel fibres as a problem, but steel fibre producers deny this except for aesthetic reasons, and state that galvanised fibres are available if required.
Benot de Rivaz of Bekaert says, “steel fibres need only a concrete cover of 1- 2mm compared to 30-40mm for normal rebar and mesh. Corrosion of the (steel) fibres may cause discoloration but does not affect the mechanical properties of reinforced structures. BRITE EURAM shows that fibres in crack openings smaller than 0.25mm do not corrode.”
To counter this, EPC cites work on subsea tunnels, particularly in Norway, showing corrosion of steel fibres. According to Ridout the Norwegain Public Roads Authority has carried out a lot of work on the subject and has concluded that steel fibre should not be used in subsea tunnels.
Another example of long-term performance is that steel fibre producers maintain that macrosynthetic fibres will ‘creep’ over cracks in concrete, effectively losing most of their reinforcement/loading effect. This applies to both short-term and long-term ground movements that may distort the lining.
Design life
Macrosythetic fibres have a disadvantage in that their track record is not so extensive as that of steel fibre reinforcement due to their later introduction and the lack of meeting international standards so far. Proponents are sure that this problem will be corrected soon and, say that in any case, is made irrelevant by projectspecific performance testing. This difference applies mainly to sprayed concrete. In the case of pre-cast segmental lining, the difference is not so pronounced with steel fibre producers still working on the replacement of more and more conventional rebar reinforcement. Where fibres are accepted, only partial replacement of the bar reinforcement cage is the norm, although there are some cases of total replacement by steel fibres. Apart from cost-saving, the main advantage is prevention of damage during fresh segment handling as most in situ forces are in compression.
Bekaert’s de Rivaz points out that design methods are only validated for steel fibre so far. For the macrosythetic fibres the claim to longer design life is based mainly on corrosion resistance. Ridout states that the fibre only starts to work when the concrete cracks, before which it does very little in terms of structural capacity. He says, “any crack will possibly allow moisture to come into contact with fibre reinforcement. The degree of degradation of the fibre will be dependent on the crack width, the concrete’s environment and fibre type. Using synthetic fibre will substantially improve the structure’s design life particularly in aggressive environments.”
Tensile strength
The tensile strength of the fibre, especially across a crack, is sometimes cited as another factor in favour of steel fibre. However, Ridout states, “the tensile strength of individual fibres is not relevant as we are dealing with a composite material that needs to attain a performance in either a plate or beam test. The performance in the test is determined by how much ductility is produced from the composite and this is determined by the bond/pull-out of the fibre in the concrete. From a quality perspective, macrosynthetic fibres should have a minimum tensile strength of 500MPa, which ensures that high quality materials are used to manufacture the fibre, and for steel fibres a minimum of 1,100MPa.”
De Rivaz says that fibre tensile strength should be matched to the concrete mix formula and its compressive strength, with a ‘non-brittle’ criterion.
Anchors and bonds
The efficiency of fibre performance in situ depends on adequate anchorage in or bonding with the cured concrete mix. For example smooth fibres of any type would present problems such as inadequate structural use. The Bekaert Dramix and HIC type of fibres offer hooks at each end for anchorage while most of the fibre is friction free. EPC BarChip macrosynthetic fibres are embossed along the full length. Advantages are claimed for both but in any case pull-out tests can determine the efficiency of any fibres in this regard, as well as by the usual beam or plate test of the concrete.
EPC’s Ridout points out, “the principle is not for the fibre to be bonded so much that it snaps, but to pull out from the concrete to provide ductility. This is a combination of getting the fibre length, thickness and anchorage optimised to provide the best ductility.”
The finer the fibre per unit weight, the more fibres will cross a crack, accordingto Ridout. However fineness also means that the fibres can snap more easily, and can cause mixing problems. Therefore a balance has to be struck.
Fibre length
Related to bonding is the fibre length. In theory the fibre should be as long as possible, but the length is usually determined by dosing and mixing performance. Bekaert says that the fibre length should be three times longer than the biggest aggregate particle to ensure anchorage. EPC claims that steel fibre length is limited by the size of the concrete spray nozzle whereas macrosynthetic fibres are not.
Creep
One of the most controversial alleged features of macrosytehtic fibres is the phenomenon of creep, or the lengthening of fibres over a crack beyond its loadbearing limit.
In structurally cracked linings, such as in asymetrically moving ground, this may not be an issue at all.
One of our questionnaire referees (subsequently known as Referee C), who wishes to remain anonymous due to promixity to a major project currently in design, pointed out, “in civil engineering we don’t normally expect tunnel linings to work that hard, especially if the loading is uniform. Asymmetrical forces could, however, give higher bending moments that can produce serious crack problems.”
Embrittlement?
Another aspect of long-term performance is embrittlement (of steel fibres). De Rivaz of Bekaert says that this is a totally incorrect concept and not a design problem, and produces an expert critical analysis to back up this claim. However, this has not deterred various research bodies from including the subject in studies on long-term, fibrereinforced concrete behaviour. According to de Rivaz the toughness of the fibre and concrete mix, plus its post-crack residual strength are much more important to consider.
Cost
A possible error in drawing up our original questionnaire is that it did not distinguish between the temporary (primary) support use of fibre-reinforced sprayed concrete, and its use as permanent (final) lining. This obviously determines the desirability of properties affecting the longer term efficiency of the fibres. Also, depending on who decides the composition of temporary support mixes, cost of fibres may become more important. Contractors left to determine temporary support (not eventually forming part of the designed final support) will naturally go for cheaper materials if they are effective over the time period required and acceptable on health and safety grounds. In moving ground the choice of fibre may be critical.
The costs of fibres are greatly affected by market forces and alternative supplies. The effect on the design of permanent lining is minimal compared to the potential savings compared to traditional reinforcement and the importance of a robust design. In sprayed primary lining, the contractor generally has to meet a test performance specified by the designer. It is then the contractor’s choice how to meet this, and that is generally a question of balance between cost and quality, necessitating a higher dosage of lower cost/lower quality fibres to meet performance requirements. There are supplies of cheaper fibres available from Asia, normally China, both of steel and polymer materials, but these should be assessed on the basis of quality as well as costs as all fibres are far from the same, even when ignoring likely lack of technical support.
Bekaert’s de Rivaz comments, “the cost per cubic metre should be for a clear, given performance, and technical requirements should be evaluated project by project.”
Ridout agrees that cost is dependent on the quality of the fibres being compared. “Generally a comparison of the top end of the market for steel fibres and synthetic fibres shows a possible saving of about five per cent by using macrosynthetic fibre.” However, he comments that steel prices have been very volatile of late whereas polymer pricing offers more stability even with fluctuations in oil stock prices. “In 2008 there was a lot of pressure on steel prices,” he says, “as China took up most of the iron supply and this is threatening to happen again. This will make macrosynthetic fibres more costs effective.” He claims that the A3 Hindhead Tunnel had to change to macrosysthetic fibre as no steel was available at the time.
Density & handling
The low density of synthetic fibres is often quoted as a factor in easier materials handling and safety, but with the correct equipment, any fibre handling should not be a problem. Bekaert emphasises the use of correct dosing equipment.
Ridout points out that CE marking of fibres shows the amount of kilos of fibre required to meet the CE specification; a ratio of 1kg of macrosynthetic to 5kg equivalent of steel fibre is stated. Although macrosynthetic fibres are lighter than water, and so will float, Ridout says that they do not float on the concrete mix as they are held by the cement and aggregates in the mix, that are able to hold the fibre homogeneously.
Packaging can be important in correct handling and mixing as, for example, Bekaert Dramix fibres are glued together in line for easy dispersal in the mixing as the glue dissolves.
Puncture hazard
Another alleged hazard of steel fibre reinforcement raised by synthetic fibre producers is that of puncturing, whether of workers or materials. From the point of view of tunnel construction this has to be addressed when using waterproofing membranes, but steel fibre producers point out that this is not a problem when a fleece/geotextile layer is employed between the membrane and shotcreted rock surface, as is usually the case.
A puncturing/scratching hazard of steel fibre has been accepted by many mining operations now using synthetic fibre reinforced sprayed concrete, although the circumstances of mining are usually a bit different to ‘civils’ tunnelling.
The possibility of a puncture hazard from fibres is non-existent from macrosynthetic fibres but, according to Bekaert, restricted to only fibre of diameter less than 0.2mm as used in Ultra-High-Performance Fibre- Reinforced Concrete (UHPFRC) otherwise known as bomb-blast concrete. EPC claims that experience in Australian mining shows that steel fibres can cause short-circuits in trailing cables and tyre punctures. Alleged damage to waterproofing membranes and operators seem to be less of an issue.
Further studies
Indicative of the less than clear current information on reinforced concrete mix design is the high level of current and likely future studies being undertaken.
The continued activity of ITA Working Group 12 (Shotcrete Use) announced, under new animateur Atsumu Itshida of Denei Kagaku Kogyo Co of Japan, a test programme for evaluation of fibrereinforced sprayed concrete, and discussion and information collection on mix design and durability to compare the durability of sprayed concrete and cast concrete. There has been agreement on the concrete mix and testing programme for short-duration loading, but long duration load tests were still under discussion.
Prof Dr Tarciscio Celestino at the Sao Carlos Engineering School of the University of Sao Paulo is conducting research into shotcrete for temporary and final linings focussing on time-dependent behaviour of both the rock mass and shotcrete, toughness of fibre-reinforced shotcrete, and the use of acoustic emission to identify the interaction between fibres an shotcrete.
The International Tunneling Consortium at the University of Texas, under the direction of Dr Fulvio Tonon, is conducting research into fibres for final linings and shotcrete, including timedependent behaviour such as creep tests on structural synthetic fibres—both single fibres and bending creep—in beams. Embrittlement over time of steel fibres and quantifying of loss of reinforcing strength caused by concrete strengthening are also covered. Other work covers the interaction between normal stress and shear resistance, further work on validation of algorithms developed for the normal force bending moment interaction diagram for steel, synthetic and fibre blends, and a review of the ASTM 1550 text (‘pizza test’) and the use of ASTM C 1609 on portions of rectangular panels to provide design input.
In Norway the COIN study conducted by NTNU covers many types of performance improvement in concrete including the use of fibre reinforcement. One aspect is the use of glass-fibre or carbon-fibre reinforcement to avoid steel corrosion. Plus more efficient, lessexpensive solutions to correct mixing of fibres into the concrete.
Referee C commented that there is better research and development on synthetic fibres year on year, thus improving its performance knowledge position. On both counts specifiers have to keep up to date with the latest findings to ensure that correct choices are made. The importance of performance-related testing was emphasised.
Applicable performance
If possible testing should relate to the conditions of the tunnel application. Therefore it is necessary to determine not only the performance of the fibre within a selected concrete mix, as in the various standard panel and beam testing methods, but also the performance of the concrete under in situ conditions, as near as can be simulated.
Performance-related specifications are therefore preferred except for very commonplace tunnel conditions, if such things exist. There are accepted tests of panel or beam samples of the concrete mixes that can assess correct ‘manufacture.’ However, for permanent (final) lining these designers also need to take into consideration the long-term conditions such as possible changes in ground loading, corrosion and alleged embrittlement.
For excavations subject to strong ground movements the extra considerations may be for the relatively short term as well as long term.
Most agree that it is a great challenge for designers and other interested parties, including suppliers, to demonstrate how the various mixes of fibre-reinforced (especially sprayed) concrete will work in situ, except perhaps in full-scale trials. Even then long-term performance involving cracking and ground movement (if any) may not be clear.
Conclusion
It is not the place of the author to decide for you which type of reinforcement fibre is the correct one for you, for your application. There is a great variety of tunnel support requirements, so therefore generalised testing seems to have limited relevance if tunneling conditions are not clear.
As above, while total simulation of materials performance under tunnel conditions is extremely difficult, the correct fibre and concrete mix choice has to be based on a complete understanding of the concrete’s in situ performance.
Acknowledgements
We wish to thank the referees that reviewed our original questionnaire, suggested amendments and additional questions, and provided additional information. These included Prof Tarcisio Celestino manager of civil engineering at Themag Engenharia, Sao Paulo, Brazil, and immediate past animateur of the ITA Working Group 12: Shotcrete Use; and Prof Fulvio Tonon of the University of Texas, Austin, US who is leading a research programme on reinforced sprayed concrete under the auspices of the International Tunneling Consortium, based there.
A mine tunnel supported throughout by arches and fibrereinforced sprayed concrete (Photo: BASF Meyco) Forming pre-cast segments using EPC BarChip macrosynthetic fibre-reinforced concrete for a tunnel in Malaga, Spain
