ABSTRACT Cement, 2.8x109t of which was produced worldwide in 2007, results in global concrete usage approach-ing 1.5m3 per person per year (USGS, 2008), mak-ing it our most widely used man made material. The constituent raw materials of cement are easily found and extracted, and are unlikely to be depleted in the foreseeable future. However, cement has a high em-bodied energy of approximately 4.6MJ/kg (Ham-mond, 2008) and its manufacture is estimated to ac-count for 3% of global CO2 emissions (WRI, 2005).

This suggests that concrete should be cast in op-timised structures that minimise material use and take advantage of the mouldability of concrete. This fluidity is rarely exploited and concrete structures tend to be cast in rigid, material intensive formwork systems. The optimisation of concrete structures such that they use less material, are structurally effi-cient and easy to construct would bring real innova-tion.

Research at the University of Bath has been look-ing at this issue for some time now. Using fabric formwork, it is possible to cast architecturally inter-esting, structurally optimised shapes based on simple design rules that are analogous to the growth of bone. When a bone is overstressed, it grows; when it is under stressed, it atrophies. In a similar way, fab-ric formwork can be used in the construction of con-crete beams to place material only where it is re-quired, and research has shown that this is both a predictable and practical approach that can achieve material reductions of up to 50% (Garbett, 2008) when compared to an equivalent orthogonally cast beam construction.

This paper explores the basis of fabric formed structures, beginning with work undertaken in the fields of offshore and geotechnical engineering in the early 20th Century. Architectural interest, which began with the Spanish architect Miguel Fisac and has since been growing around the world, is then discussed. Most notable is the work undertaken by

Professor Mark West and his team at the Centre for Architectural Structures and Technology (CAST), at the University of Manitoba, where the design and construction of beams, trusses, columns, façade ele-ments and shells has been developed.

Comparisons to conventional concrete construc-tion are made, and the advantages of fabric form-work illustrated. Focusing on fabric formed beams, design methods used by previous researchers are outlined before optimisation processes are discussed. Four methods for the construction of fabric formed beams are then presented.

The limited available structural test data is out-lined and areas of current research, including inno-vative anchorage methods for advanced composite longitudinal reinforcement, are discussed. The future of fabric formwork is then considered, focusing on the use of active reinforcement, advanced compo-sites and participatory external formwork with addi-tional thought given to improved design, optimisa-tion and construction processes.

Figure 1. Fabric formed beams (Garbett, 2008)

REFERENCES

Garbett, J., 2008. Bone growth analogy for optimising flexibly fomed concrete beams. Thesis (MEng). University of Bath: Bath.

Hammond, G.P., Jones, C.I., 2008. Embodied energy and car-bon in construction materials. Proc. Instn Civil Engrs: En-ergy, in press.

USGS, 2008, Cement statistics, in Kelly, T.D., and Matos, G.R., comps., Historical statistics for mineral and material commodities in the United States: U.S. Geological Survey Data Series 140, available online at http://pubs.usgs.gov /ds/2005/140/ (Accessed 13/01/10)

WRI, 2005. Carbon Dioxide Emissions by Source 2005. Earth-trends Data Tables: Climate and Atmosphere, available on-line at http://tinyurl.com/ybdohdn (Accessed 13/01/10)

Innovative concrete structures using fabric formwork

J.J. Orr, T. J. Ibell & A.P. Darby University of Bath, Bath, England.

1 INTRODUCTION

The use of fabric formwork for concrete structures can be traced back to the early 1900s, and the meth-ods involved mainly stem from work in offshore and geotechnical engineering. In 1922 it was proposed to use concrete filled fabric bags in the construction of underwater concrete structures but it was not until the late 1960s that any real headway was made in this field, precipitated by the new availability of low cost, high strength, durable synthetic fibres that al-lowed the forming of complex shapes (Lamberton, 1989). Initial interest in the architectural possibilities of fabric formwork can be attributed to the Spanish architect Miguel Fisac, who in 1969 completed the Centro de Rehabilitación para la Mutualidad del Papel (MUPAG) in Madrid. It was here that the first patented method for pre-fabricated fabric formed wall panels was developed.

Subsequent developments have occurred simulta-neously yet independently of each other. Whilst both Kenzo Unno and Rick Fearn have developed suc-cessful systems and techniques for fabric formed structures, the most influential work has come from Mark West, founder of the Centre for Architectural Structures and Technology (CAST) at the University of Manitoba in Canada, which is the first research centre dedicated to the development and promulga-tion of fabric formwork for concrete structures. It is this architecturally led work that has formed the basis for previous research at the University of Bath.

This paper begins by considering the principles behind fabric formwork, before focusing on the cur-rent state of the art in design, optimisation and con-struction of fabric formed beams.

1.1 Traditional practice Concrete has been primarily cast in orthogonal tim-ber or steel moulds since the mid-1800s, resulting in the well established formwork practices that exist today. Rigid formwork systems tend to be simple to construct, but consume more material than an equi-valent variable section member, increasing both cost and structural dead weight. Variable section mem-bers can feasibly be produced on an industrial scale, but their geometry remains governed by primarily prismatic forms.

1.2 Fabric formwork Forming concrete in a flexible membrane (typically a high strength polyester fabric) provides a simple method for the construction of efficient, optimised and aesthetically pleasing concrete structures that of-fer several advantages for engineers.

The material required for a fabric formed struc-ture is lightweight, cheap and ubiquitous – 12m span beams have been formed using under 10kg of fabric (West, 2007). Pouring concrete into a permeable fabric results in a filtering effect in which air and water are allowed to bleed from the structure, the ef-fect of which is twofold. First, small increases in concrete compressive strength occur that are attribu-table to the reduction in water:cement ratio. More significant are the increases in surface density that occur as a result of there being very few air bubbles trapped between the formwork and concrete. In-creases in surface density prevent in-service mois-ture and air ingress into the section, thereby slowing corrosion processes and potentially allowing for a reduction in cover to steel in fabric formed struc-

Innovative concrete structures using fabric formwork

J.J. Orr, T. J. Ibell & A.P. Darby University of Bath, Bath, England.

ABSTRACT: This paper examines the state of the art in fabric formed beam design, providing a summary of previous work, experimental data and optimisation processes. The future of fabric formwork is considered and current work at the University of Bath is presented.

tures when compared to their conventional counter-parts. The distance to which this ‘case hardening’ ef-fect extends into the member is unclear at present.

Fabric formwork can be stitched into almost any configuration and the boundary conditions, includ-ing support locations and degree of pretensioning, can be altered to achieve the desired form. The con-struction of façade panels, columns, trusses, shells and beams has already been achieved, as illustrated in Figure 1. This paper will investigate the current state of the art in fabric formed beams.

Figure 1. Fabric formwork (after West, 2007; Garbett, 2008)

2 DESIGN

Design methods for fabric formed beams are cur-rently in a state of flux. Work at CAST has previ-ously taken an empirical approach, and many beams were not reinforced or structurally tested. The final shape of such a beam is determined by the material properties of the fabric and boundary conditions im-posed during construction (§3).

The hydrostatic shape obtained from a given set of these boundary conditions can be accurately pre-dicted using elastic theory, although dynamic relax-ation has also been used to model the interaction be-tween fabric and concrete (Veenendaal, 2008). In addition, Bailiss (2006) and Garbett (2008) used an empirical method to determine the area, perimeter and shape of fabric structures that was moderately successful.

The design of fabric structures has, up to now, been approached primarily from an architectural perspective. Structural verification is now required, and this has been the focus of previous work at the University of Bath. Garbett (2008) implemented a sectional analysis method, as outlined in Figure 2, to design singly reinforced beams that were unre-inforced in shear to BS 8110-1 (1997).

Whilst the flexural strength of a reinforced con-crete member can accurately be predicted using the plane section hypothesis, which forms the basis of many concrete design codes including BS 8110, shear design is widely considered to be an unre-solved area of concrete technology (Bentz et al., 2006). The complex cross sections of fabric formed beams make conventional shear links difficult to de-

tail, yet omitting shear reinforcement relies on an accurate prediction of the unreinforced section’s shear capacity. In BS 8110-1, the unreinforced ca-pacity is predicted using empirical relationships based on many hundreds of beam tests, none of which were carried out on fabric formed sections, thereby making the accuracy of this method ques-tionable.

Both BS EN 1992-1-1 (2004) and ACI 318 (2005) use similar empirically based approaches, while the use of modified compression field theory (Bentz et al., 2006) in the Canadian CAN/CSA S6 (2006) make this code better suited to the design of variable section members.

Longitudinal reinforcement in fabric formed beams has previously been limited to single or bun-dled bars, pre-bent to the desired profile. The accu-rate placement of, and provision of anchorage to such bars has proven to be difficult (Garbett, 2008). The welded end plate (Figure 3) is the most common anchorage connection, yet this leaves longitudinal steel susceptible to corrosion. The use of advanced composite reinforcement and post-tensioning offers a potential solution to this, as discussed in §4.

Figure 2. Sectional design method.

Figure 3. Anchorage in bending moment shaped beams.

2.1 Optimisation Optimisation can be considered as the process by which variables are used to determine the best op-tion for a given set of parameters. Physical model-ling techniques used in the past have now been all but replaced with numerical simulation methods such as evolutionary structural optimisation, solid isotropic material penalisation and sensitivity analy-sis that are described in detail elsewhere (Rozvany, 2008).

!"#$%&'(%#)*+(

,"#$%&'(%#-.+/)

,+"#012#)3%+4#+(/#5.5%(6#%(7%-.*%)

8"#$'7'/%#'(6.#)%96'.()

:"#;(+-<)%

;**-'%/#!#=#"+6#)%96'.(##

>+-9?-+6%#)3%+4#+(/#5.5%(6#4%@?'4%5%(6)#6.#A2#B!!CD!

E%@?'4%/#)%96'.(#/%*63#/%&'(%/

F</4.)6+6'9#)3+*%G'7%(#%5*'4'9+--<

H6%4+6%#&.4##$)%96'.()

I"#J-.6#.*6'5')%/#)3+*%

%#&'()*+,&%*#$-'&.''#$/'0&%*#/$('12%('3

Structural and material optimisation is a key component of fabric formwork. Up to now bending moment shaped beams (Figure 1) have been opti-mised using the previously described sectional an-alysis method and material reductions of up to 50% have been obtained when compared to an equivalent rectangular beam. This is remarkable given the sim-plicity of both the design and construction of these beams and offers a real opportunity for material use reductions in entire building systems.

More complex approaches utilising evolutionary optimisation are currently being considered that of-fer two opportunities to further reduce material usage. First, through more accurate modelling of the hydrostatic shape and concrete-fabric interaction during pouring and second through improved analy-sis of the reinforced concrete section under loading to ensure that material is provided only where it is required. However, computational methods must al-ways consider construction processes to ensure the optimised beam design can feasibly be built using fabric formwork.

3 CONSTRUCTION

Construction methods for fabric formed beams are continually improving. This section details four methods for the construction of variable section beams, three of which were developed by research-ers at CAST (Veenendaal, 2008).

The spline method (Figure 4) uses a metal bar to vertically pretension a single rectangular sheet of fabric held on a timber forming table. Pretensioning the fabric reduces the volume of concrete in the ten-sion zone, thus providing an optimised design. Beams constructed using this method have previ-ously had a parabolic elevation, although the final layout is determined by varying the locations and magnitude of the applied pretension.

The keel mould (Figure 5) uses two sheets of fab-ric, held vertically and secured between sheets of plywood (the keel) that are cut to the desired beam elevation. The fabric is then draped over a forming table and pretensioned to both obtain the desired shape and to prevent wrinkling during construction.

The pinch mould (Figure 6) is used to create beams and trusses by sandwiching two sheets of fab-ric between a rigid timber mould. At designated lo-cations protrusions from the timber mould ‘pinch’ the fabric to create openings in the web of the beam. The method allows the rapid construction of opti-mised beam elements, but constructing the form-work is more labour and material intensive than other methods. In addition, the structural behaviour of these beams is governed by Vierendeel action, which somewhat complicates their analysis.

The fourth method, developed by Bailiss (2006) and Garbett (2008) utilises solely the wet concrete

weight to form the beam. By predicting the shape of the fluid filled fabric, fixing points along its perim-eter are determined. The fabric is then hung between two supports before being filled with concrete to ob-tain the desired forms, some of which are illustrated in Figure 1.

Figure 4. Spline mould.

Figure 5. Keel mould.

Figure 6. Pinch mould.

3.1 Structural tests A total of six 2m span singly reinforced beams, de-signed as described in §2, have been tested in five point bending at the University of Bath (Ibell et al., 2007). Of these, five were found to fail in shear close to the supports, although two tests failed to reach their design load due to incorrect positioning of the longitudinal reinforcement.

Fabrication of the beams was generally success-ful, and empirical methods were employed to predict the hydrostatic section shape. In general, elastic and plastic methods for the prediction of failure loads were accurate as was the prediction of load-deflection responses by double integration of curva-tures along the length of each beam.

The testing highlighted two areas that require fur-ther work. The predominance of shear failures in

!"#$%&'()*+,-!*+#%.

/0,%&-

12(3"4#%&'

!*+#%.(0%&&-5*)(-5'-6

/0,%&-(04,,-5(5"7&()"(8"#$(69*0-

!"#$%&'

:2(;*<"4)

!"#$%&'&()"*%"+',-#.(./,)&

!"##$%&'()*+),-.%+($/%&)*+0*)/$

0"%1-#2(',+3&

4&#.-"#-#2(%")&.

0,+%-$(.,#56-$/&5(+&'6&&#(7&&3.(,#5(8-9&5(-#()".-'-"#

:&&3($*'('"(5&.-%&5+&,1()%"8-3&

!$(%'-/ !$(%'-/:&&3

!"#$%&'($")*('+,*$-+$./+$01'&+2&$*3*)+4$*(

!"#$%&

!+$./+$0'-+$'3$455

!6"3'#+"$(

!"##$%&'()*+),-.%+($/%&)*+0*)/$

!$(%'-/

!$(%'-/

bending moment shaped beams was unexpected and highlights deficiencies in the current design proced-ure. In addition, anchorage methods for longitudinal reinforcement (Figure 3) leave steel bars exposed to corrosion and cannot be used if advanced composite reinforcement is to be used, an area of future devel-opment that is discussed below.

4 THE FUTURE

Fabric formed beams have, to date, been fairly simi-lar in form, reinforcement and scale. Alternative de-signs incorporating some recent innovations in this field are presented in Figure 7.

Figure 7. Future design directions.

The addition of easily manufactured and installed shear reinforcement is key for the future of fabric formed beams. The use of Carbon Fibre Reinforced Polymer (CFRP) sheets, acting as reinforcement and permanent formwork, is currently being investigated at the University of Bath. The high tensile strength and durability of advanced composites also makes them a logical alternative to steel as longitudinal re-inforcement, but introduces new design problems.

As welded end plates cannot be used for the an-chorage of longitudinal FRP reinforcement, current research is investigating the use of an innovative wedging anchorage method for Carbon Fibre Re-inforced Polymer (CFRP) bars. The concept has been proven in cube pull out tests (Darby et al., 2007) where order of magnitude increases in load and displacement capacity were seen and verifica-tion by beam tests is now required.

However, advanced composites have high work-ing strains and are therefore inefficient when used in passively reinforced structures. Burgoyne (2001) ar-gues that advanced composites are most effectively used in prestressed structures, where greater moment capacity can be obtained and the full tensile strength of the tendon utilised. Post tensioned fabric formed beams are an exciting prospect, and offer potential improvements in moment and shear capacity. By sewing ducts into the formwork, tendon positioning within the section can be ensured, and the use of un-bonded advanced composite rope bypasses potential corrosion concerns.

4.1 Alternative areas The field of fabric formwork is by no means limited to beams. Shells, working efficiently in membrane action, offer great advantages, but their design and construction is more complex than bending elements and they are rarely used in commercial building sys-tems. However, the construction of shell structures using fabric formwork is well established at CAST, and work is being undertaken by the authors to in-vestigate the potential for large scale uses of fabric formed shell structures that will bring further mate-rial and cost savings to concrete construction.

5 CONCLUSIONS

Fabric formed concrete beams offer significant ad-vantages for designers, including material reduc-tions, ease of construction and aesthetic appeal. The forms are predictable and the development of robust methods for their design and optimisation is well under way. New materials, including advanced composites, prestressed reinforcement and fibre re-inforced concrete offer additional advantages for fabric formed beams that will be investigated in fu-ture work.

REFERENCES

ACI 318:2005. Building code requirements for structural con-crete. ACI.

Bailiss, J., 2006. Fabric-formed concrete beams: Design and analysis. Thesis (MEng). University of Bath: Bath

Bentz, E.C., Vecchio, F.J., Collins, M.P., 2006. Simplified Modified Compression Field Theory for Calculating Shear Strength of Reinforced Concrete Elements. ACI Structural Journal 103 (4): 614-624

BS 8110-1:1997. Structural use of concrete - Part 1: Code of practice for design and construction. BSI.

BS EN 1992-1-1:2004. Eurocode 2: Design of concrete struc-tures - Part 1-1: General rules and rules for buildings. BSI.

Burgoyne C.J., 2001. Rational Use of Advanced Composites in Concrete, Proc. ICE Structures & Buildings, 146, 253-262

CAN/CSA S6:2006. Canadian Highway Bridge Code. CSA. Darby, A.D., Ibell, T.J., Tallis, S., & Winkle, C., 2007. End

Anchorage for Internal FRP reinforcement. In: Triantafil-lou, T.C., ed. FRPRCS-8, July 16-18, University of Patras.

Garbett, J., 2008. Bone growth analogy for optimising flexibly formed concrete beams. Thesis (MEng). University of Bath: Bath.

Ibell, T.J., Darby, A.P., Bailiss, J.A., 2007. Fabric formed con-crete beams. In: Darby, A.P., ed. ACIC 2007, April 2-4 2007, Bath: University of Bath.

Lamberton, B.A., 1989. Fabric forms for concrete. Concrete International. 11(12): 58-67

Rozvany, G.I.N., 2008. A critical review of established meth-ods of structural topology optimization. Struct Multidisc Optim. 37: 217-237

Veenendaal, D., 2008. Evolutionary Optimization of Fabric Formed Structural Elements. Thesis (MS). Delft University of Technology: Delft.

West, M., 2007. Construction. Winnipeg: Centre for Architec-tural Structures and Technology, University of Manitoba.

!