Mechanical Behavior of Textile Reinforced Concrete (TRC ...

Journal of Advanced Concrete Technology Vol. 8, No. 1, 35-47, February 2010 / Copyright © 2010 Japan Concrete Institute 35

Scientific paper

Mechanical Behavior of Textile Reinforced Concrete (TRC) / Concrete Composite Elements Catherine G. Papanicolaou1 and Ioannis C. Papantoniou2

Received 29 August 2009, accepted 23 December 2009

Abstract In this paper the response of composite structural elements cast against thin-walled stay-in-place (SiP) formwork ele-ments made of Textile Reinforced Concrete (TRC) is experimentally investigated and analytically approached. TRC comprises an innovative composite material consisting of fabric meshes made of long fibre yarns (e.g carbon, glass, aramid or basalt) arranged in at least two (typically orthogonal) directions and embedded in a cementitious fine-grained matrix. Two types of reinforced concrete specimens were considered: the first one included 22 beam-type specimens incorporating flat TRC formworks, whereas the second included 11 prismatic column-type specimens cast into perma-nent precast TRC shafts. Moment and deflection values at first-crack, steel yielding (where applicable) and ultimate for the beam-type specimens were analytically derived based on a proposed simplified approximation of strain distribution across a fibre roving. Based on the results of this study SiP TRC formwork elements comprise an attractive system for hybrid construction practices.

1. Introduction

The minimization of two inter-related factors, namely the duration of construction and the total (life-cycle) cost of a structure has always been the focal point of construction management. The use of Stay-in-Place (SiP) – or permanent, or integrated – formwork ele-ments addresses this goal allowing for hybrid concrete construction practices to be developed that combine all the benefits of precasting (e.g. speed, controlled quality, accuracy, flawless finish), with all the benefits of in-situ construction (e.g. economy, continuity and robustness).

SiP formwork is a structural element that is used to contain the placed concrete, mould it to the required dimensions and remain in place for the life of the struc-ture (Wrigley, 2001). Permanent formwork elements are distinguished into participating and non-participating ones; the former contribute to the strength of the struc-ture through composite action with the cast-in-place parts of it, while the latter make no strength contribu-tions. Several different SiP formwork systems have been developed incorporating a variety of materials, such as steel, timber and fiber reinforced polymers (FRP). High corrosion susceptibility, poor durability and low fire resistance are, respectively, some marked drawbacks of the aforementioned systems. The alleviation of these drawbacks may be realized by using cementitious com-posite materials consisting of inorganic matrices (ce-

ment-based mortars or micro-concrete) reinforced with non-corrosive fiber yarns arranged in a grid structure. The generic term for these materials is Textile Rein-forced Concrete (TRC) (RILEM, 2006). In this case, composite behavior is achieved mainly through me-chanical interlock between the matrix and the grid open-ings; furthermore, the fiber properties may be fully ex-ploited since the quantity and the orientation of the yarns can be selected according to the design require-ments (contrary to the fiber reinforced concrete, in which fibers are randomly distributed and oriented). The flexibility of the textile meshes allows for the design of thin SiP formwork elements with complex geometry. This study presents the results of an experimental inves-tigation conducted on concrete elements cast against thin-walled TRC SiP formwork and gives an analytical insight of the response of TRC-concrete composite beams under four-point bending.

2. Experimental program

2.1 Beam-type specimens Beam-type specimens measuring 1500 mm in length, 150 mm in width and 100 mm in total height, were sub-jected to four-point bending. The following parameters were examined in order to investigate their influence on the flexural behavior of the composite elements: (i) Fi-ber reinforcement ratio (corresponding to 1 and 2 lay-ers); (ii) Fiber rovings’ coating (polymeric and none); (iii) Treatment of the TRC/cast-in-situ concrete interface (smooth and rough); (iv) Spacing of the textile rein-forcement layers (1 mm and 6 mm); and (v) Filaments’ material (glass and carbon).

Specimens’ fabrication was divided in two stages. Ini-tially, a thin flat TRC element (measuring 1500 mm x 150 mm x 12 mm, as in length x width x height) was

1Lecturer, Structural Materials Laboratory, Civil Engineering Department, University of Patras, Greece. E-mail:[email protected]. 2Civil Engineer, M.Sc., Ph.D. Candidate, Structural Materials Laboratory, Civil Engineering Department, University of Patras, Greece.

36 C. G. Papanicolaou and I. C. Papantoniou / Journal of Advanced Concrete Technology Vol. 8, No. 1, 35-47, 2010

produced by application of two 6 mm thick mortar lay-ers, within which the textile reinforcement (either in 1 or 2 layers) was placed. In the case of double-layer tex-tile reinforcement, a thin mortar layer (1 mm in thick-ness) was used to separate the textiles and to ensure that an adequate quantity of matrix material would protrude through the grid openings in order to provide better bond conditions. In most specimens, the exposed sur-face of the TRC element was roughened while in fresh state by the formation of longitudinal grooves using a hand tool. In order to investigate the influence of the SiP formwork’s surface treatment two TRC elements re-ceived smoothening. Following a 24-hour curing period the specimen’s main part (88 mm in thickness) was cast on top of the TRC and the completed beam was stored in a curing chamber (20°C, 98%RH) for 28 days (until testing). For 15 specimens the main part consisted of plain concrete, while for the remaining specimens rein-forced concrete was used. Steel reinforcement com-prised two hook-ended longitudinal bars (with a diame-ter of 8 mm), positioned at an effective depth of 75 mm (resulting in a reinforcement ratio equal to approx. 9‰). No transverse reinforcement was used. The specimen geometry is shown in Fig. 1.

For the TRC matrix material a commercial inorganic binder was used consisting of a blend of cementitious powders and a low fraction of polymers. The water to cementitious materials ratio was equal to 0.23; the com-pressive and flexural strengths (measured at 28 days) were equal to 24 MPa and 6.3 MPa, respectively. The mean 28-day compressive strength of the concrete was equal to 21.2 MPa. The average yield stress of the steel reinforcement was equal to 571.6 MPa.

Two basic types of commercial nonwoven textiles with equal quantity of fiber rovings in two orthogonal directions (0°/90°) were used. The first type comprised high-strength carbon fiber rovings and the second one glass fiber rovings. In either textile the width of each roving was equal to 3 mm and the clear spacing between rovings was 7 mm. The weight of fibers in the textiles was 348 g/m2 and 480 g/m2 for carbon fibers and glass

fibers, respectively. The nominal thickness of each layer (based on the equivalent smeared distribution of fibers) was 0.095 mm and 0.092 mm for the carbon fiber textile and the glass fiber textile, respectively. The resulting volumetric ratio of fibers in the TRC element (fiber vol-ume of a single textile layer per TRC unit volume) was approximately equal for both carbon (7.9‰) and glass fibers (7.6‰). The guaranteed tensile strength of the fibers in each direction was taken from data sheets of the producer equal to 3350 MPa (carbon) and 1600 MPa (glass). The elastic modulus of the fibers was 225 GPa (carbon) and 70 GPa (glass). The fiber rovings in most of the textiles were uncoated, but in a small number of textiles they were polymer-coated. The textile structure is shown in Fig. 2 while Fig. 3 illustrates various stages of specimens’ construction.

A total of 22 specimens were constructed: 12 speci-mens comprised plain concrete main parts, 7 specimens comprised reinforced concrete main parts and 3 were used as control specimens. The specimens designated as “Con” and “Conm” consisted of plain concrete and were cast against a retrievable and a SiP TRC formwork, re-spectively. Specimen “ConS” was also cast against a retrievable formwork and consisted of reinforced con-crete. Specimens are given the general notation XF, where X denotes the number of textile layers of the TRC element (1 or 2) and F denotes the fiber type of the

1500 mm In Situ Concrete

Steel Reinforcement

TRC Formwork Element

150 mm

75 m

m

94 m

m

Textile Reinforcement

100

mm

12 mm

(a) (b)

Fig. 1 Beam-type specimen geometry (with reinforced concrete main part): (a) 3D and (b) section view.

7 mm

10 m

m

10 mm

7 m

m

Fig. 2 Textile structure.

C. G. Papanicolaou and I. C. Papantoniou / Journal of Advanced Concrete Technology Vol. 8, No. 1, 35-47, 2010 37

textiles (C for carbon and G for glass). Indices “c” and “s”, where applicable, denote polymer-coated textiles and smooth TRC/concrete interface, respectively. The following specimens with plain concrete main parts were constructed: 1C, 2C, 1Cc, 2Cc, 1Ccs, 1G, 2G, 1Gc, 2Gc and 1Gcs. The notation of specimens with rein-forced concrete main parts is complemented with the letter “S”. These specimens are the following: 1CS, 2CS, 1CcS, 1GS, 2GS and 1GcS.

Three additional specimens with plain concrete main parts were constructed in order to investigate the effect of the textile inter-layer spacing along with the com-bined use of carbon and glass fiber textiles and the com-bined use of coated and uncoated glass fiber textiles. More specifically, specimens CGl and CGp were both cast on top of TRC elements each containing one layer of carbon and one layer of glass fiber textiles, the only differences being the spacing of the two layers (6 mm for CGl and 1 mm for CGp) and the resulting TRC thickness (18 mm for CGl and 12 mm for CGp). Speci-men GcG was cast on top of a TRC element reinforced with two layers of glass fiber textile, one coated (closest to the bottom of the specimen) and one uncoated. Table 1 summarizes design parameters for all beam-type specimens.

All specimens were tested under monotonic four-

point bending at a total span of 1400 mm using a stiff steel frame. The displacement-controlled loading proto-col was applied using a vertically positioned actuator. The load application points were 50 mm distant from the mid-span ensuring flexure-dominated response of the tested specimens. The displacement application rate was equal to 0.016 mm/sec. Displacements were meas-ured at mid-span using an external rectilinear displace-ment transducer (of 50 mm stroke capacity) mounted on one side of the specimen (Fig. 4).

2.2 Column-type specimens

Prismatic column-type specimens with square cross section (224 x 224 mm) and 500 mm height were tested under concentric monotonic compressive loading. The construction procedure of these specimens resembled the one of the beam-type specimens. First, a thin-walled TRC SiP shaft was constructed measuring 500 mm in height and 12 mm in thickness. Following a curing pe-riod of 28 days the main part of the specimen was cast in the TRC formwork consisting of reinforced concrete. The reinforcement was of the same grade as the one used for the beam-type specimens and comprised a 460 mm high steel cage formed by 4 hook-ended longitudi-nal bars (12 mm diameter) and square ties (8 mm di-ameter) with a spacing of 100 mm (Fig. 5).

Displacement transducer

Hydraulic Actuator

650 mm

1400 mm

Fig. 4 Test apparatus for beam-type specimens.

Textile

150 mm

Fig. 3 Specimen construction stages.


The parameters under investigation were: (i) Fila-ments’ material (glass and carbon); (ii) Fiber rovings’ coating (polymeric and none); (iii) Matrix material; and (iv) Fiber rovings’ orientation.

The TRC column-type formwork elements were cast in a custom-made modular metallic mould measuring 500 mm in height. Two sets of shells (each comprising four steel plates) formed a 12 mm gap in which the ce-ment-based mortar was hand-pumped through a plug at the bottom of the mould. The plates’ edges (at the mould’s corners) were chamfered at a curvature radius of 15 mm. Prior to fixing the plates together a single-layer textile was put in place (in the middle of the space between the inner and the outer parts of the mould) al-lowing for a 200 mm overlap.

Four different textiles were used in the TRC shafts. The first two were identical to the ones used for the production of the flat TRC SiP formwork elements, while the other two comprised polymer-coated fiber rovings (carbon for one of the textiles and glass for the other) oriented at +45°/-45°. Apart from roving orienta-tion the textiles differed in no other aspect.

Two types of inorganic mortars were used: the first one (mortar I) was identical to the mortar used for the flat SiP formwork elements (measured 28-day compres-sive and flexural strength were equal to 23.1 MPa and 6.9 MPa, respectively); the second mortar (mortar B) was of higher compressive strength (51.7 MPa) and of lower flexural strength (2.5 MPa) than mortar I. Both mortars included a moderate quantity of super-plasticizer so that high fluidity and pumpability could be achieved.

A total of 11 specimens were constructed. One speci-men (designated as “Con”) was cast in a retrievable formwork and was used as the control specimen. The remaining specimens were cast in the TRC moulds. The mean 28-days compressive strength of the concrete used in this specimen series was equal to 21 MPa. Mortar I was used for the production of half of the TRC shafts, whereas mortar B was used for the remaining ones. The specimens receive the general notation MFO, where M denotes the mortar type (I or B), F denotes the type of fibers in the textile (C for carbon and G for glass) and O denotes the roving orientation (90 for 0°/90° and 45 for +45°/-45°). Index “c” is used to distinguish the poly-mer-coated textiles. According to the above the follow-ing TRC-concrete composite specimens were produced: BC90, BCc45, IC90, ICc45, BG90, BGc90, BGc45, IG90, IGc90, IGc45, all described in detail in Table 1. All specimens were stored in a curing chamber (20°C, 98% RH) for a time span of 28 days (i.e. until testing). Axial concentric monotonic compression was exerted on all specimens in a displacement control mode at a rate of 0.035 mm/s using a 4000 kN capacity loading machine. Axial deformations were measured by an ex-ternal LVDT (with a gauge length of 250 mm) posi-tioned at mid-height.

Table 1 Design parameters for all test specimens. Specimen

designation Reinforcement

TRC formwork element Beam-typespecimens

Num-ber of layers

Type of fiber rovings (1) Concrete

part

Con N.A. N.A. Conm - N.A. 1C 1 uncoated carbon 2C 2 uncoated carbon 1Cc 1 coated carbon 2Cc 2 coated carbon 1Ccs (2) 1 coated carbon 1G 1 uncoated glass 2G 2 uncoated glass 1Gc 1 coated glass 2Gc 2 coated glass 1Gcs (2) 1 coated glass GcG 1 + 1 coated glass + uncoated glass

CGp 1 + 1 uncoated carbon + uncoated glass

CGl (3) 1 + 1 uncoated carbon + uncoated glass

-

ConS N.A. N.A. 1CS 1 uncoated carbon 2CS 2 uncoated carbon 1CcS 1 coated carbon 1GS 1 uncoated glass 2GS 2 uncoated glass 1GcS 1 coated glass

2∅8

Column -type

specimens

Con N.A. N.A. BC90 (4) 1 uncoated carbon BCc45 (4) 1 coated carbon (+45°/-45°)IC90 1 uncoated carbon ICc45 1 coated carbon (+45°/-45°)BG90 (4) 1 uncoated glass BGc90 (4) 1 coated glass BGc45 (4) 1 coated glass (+45°/-45°) IG90 1 uncoated glass IGc90 1 coated glass IGc45 1 coated glass (+45°/-45°)

4∅12 (longitu-

dinal bars) +

∅8 (square

ties) @100mm

N.A.: Not applicable; (1) Orientation of fiber rovings: (0°/90°) for all specimens, unless otherwise specified; (2) Smooth TRC/concrete interface (rough TRC/concrete interface for all other beam-type specimens); (3) 6 mm spacing between the two textile layers (1 mm spacing between textile layers for all other beam-type specimens with double-layered TRM elements); (4) properties of mortar used in TRC element: 28-day compressive strength = 51.7 MPa (= 23.1 MPa for all other specimens) & flexural strength = 2.5 MPa (= 6.9 MPa for all other speci-mens).


3. Results and discussion

3.1 Response of beam-type specimens Key results and load versus mid-span deflection curves of all beam-type specimens are presented in Table 2 and in Fig. 6, respectively. Higher load and deflection at failure was recorded for all composite specimens in comparison to the control ones. Specimens Con and Conm responded in an almost identical manner (failing at similar load and deflection) due to a single crack that formed at mid-span.

Most plots for composite beams with plain (unrein-forced) concrete main parts are characterized by an as-cending linear branch (in the uncracked state) followed by a “saw-like” second one of reduced mean slope. The initial linear branch terminates at first-crack where – depending on the textile reinforcement ratio – slight or significant load reduction occurs; this reduction amounts to 50% of the load achieved at first-crack crP (used for the calculation of crM in Table 2) for speci-mens with lightly reinforced TRC formworks (i.e. 1C, 1Cc, 1Ccs, 1Gc & 1Gcs), to 25% of crP for specimens with moderately reinforced TRCs (i.e. CGp & CGl) and is negligible for specimens with heavily reinforced TRCs (i.e. 2C, 2Cc, 2G, 2Gc).

After first-crack tensile stresses in its vicinity are car-ried solemnly by the textile. Failure at this point (unre-coverable or partially recoverable loss of load-bearing capacity – e.g. specimens 1G, 1Gc & 1Gcs) is associated to inadequate textile reinforcement. Theoretically, the minimum textile reinforcement area, minf,A , should be that required to resist the tensile force carried by the cross-section just prior to first cracking; minf,A is ap-proximated by Eq.(1).

mtct.eff,fu fAf = A minf, (1)

where .eff,fuf is the effective tensile strength of the tex-tile reinforcement equal to a fraction of the tensile strength of the individual fiber, fuf , so that

fu.eff,fu kff = ; ctA is the part of the beam’s cross-section in tension (roughly equal to gA.40 , gA being the gross cross-section of the beam); and mtf is the direct tensile strength of the mortar. The reduction factor k accounts for the progressive fiber damage at the cracks’ edges where either fiber slippage and wear or fiber kinking and crimping occurs. Brameschuber et al. (2008) sug-gested a k value equal to 0.75; in this work, k is taken equal to 0.65 and 0.75 for carbon and glass fibers, re-spectively. By taking mtf equal to approximately 50% of the measured flexural tensile strength of the mortar (according to the supplier’s data) the calculated minf,A is found equal to 8.7 mm2 and 15.8 mm2 for carbon and glass textiles, respectively. Based on the above discus-sion, indeed, single-layer glass TRCs (with textile rein-forcement area equal to 13.8 mm2) comprise marginally reinforced formworks that fail to uptake higher load than that corresponding to first cracking. “Pseudo-ductile” behavior of specimens 1Gc and 1Gcs is attrib-uted to polymeric coating of the textiles, as explained below.

The post first-crack response of the beams is depend-ent on both the textile reinforcement ratio and the rov-ings-to-mortar bond characteristics. More specifically, the use of polymer-coated textiles and higher textile reinforcement ratios led to an increase of the load-carrying and deformation capacity; this becomes evident by comparison of cru /MM and cru /ΔΔ ratios in Table 2 of ‘coated’ versus ‘uncoated’ counterpart specimens (effect of textile coating) and single- versus double-layered counterpart specimens (effect of reinforcement ratio). Comparisons with specimen 2Cc in regard to deformation capacity should not be taken into consid-

Fig. 5 Column-type specimens: geometry and fabrication process (all dimensions in mm).


eration as this specimen failed due to shear. Fiber coat-ing also led to denser cracking patterns (Fig. 7) with smaller crack widths. This fact is indicative of homoge-neous stress distribution in the TRC element and of high stress transfer capacity between the mortar and the tex-tile (shear failure of specimen 2Cc confirms this obser-vation). Coating of the rovings seems to improve tex-tile/mortar bond conditions and fiber alignment by sta-bilizing the fiber bundles at the textile joints and by pre-venting their relative slippage in these areas. Further-more, the coating procedure results into a high degree of impregnation of the fiber bundles, providing strain compatibility between the inner and the outer filaments.

Smooth post first-crack load-deflection curves are as-sociated to either: (i) well distributed cracking (speci-mens 2Cc and 2Gc); and (ii) relative slippage at the in-terface between the mortar matrix and the uncoated car-bon filaments found at the perimeter of the rovings (specimens 1C and 2C). It is noted that in a cementi-tious carbon fiber composite the bond between the fi-bers and the cementitious matrix is a limiting parameter. ‘Saw’-like load-deflection curves in the post first-crack region depict the formation of individual cracks for the rest of the specimens.

Failure of plain concrete specimens cast on top of un-coated carbon TRC formworks was attributed to pro-gressive pull-out of fiber rovings from the inorganic

matrix until central cracks propagated to the full height of the cross-section. Improved bond conditions for coated carbon textiles led to combined shear/flexure failure for specimens 1Cc and 1Ccs, and to shear failure for specimen 2Cc. Failure of specimens with uncoated glass textiles was the result of sleeve filaments’ damage (i.e. of the fracture of individual fibers found on the perimeter of the rovings) caused by stress concentra-tions at the cracks’ edges until an inadequate reinforce-ment area remained to withstand the imposed load. Tex-tile rupture was the typical failure mode for specimens incorporating coated glass TRC formworks.

According to the experimental results the effect of the TRC/concrete interface roughness is rather small but deserves more experimental attention since it is strongly dependent on: the roughness scheme (i.e. the magnitude of surface undulations); the mechanical properties and the maximum aggregate size of cast-in-place concrete; and the elapsed time between TRC and concrete casting (i.e. the effect of chemical bond). Regardless of the sur-face treatment that the SiP formworks received in this investigation all TRC/concrete interfaces remained in-tact during testing (i.e. no debonding was observed).

By comparison of the response of specimens CGp and CGl it is concluded that the increase of the matrix thickness between the textile layers (and the consequent increase of the TRC thickness) has small influence in

Table 2 Key experimental and analytical values for all beam-type specimens.

Experimental Values First crack Section Analysis Modified Section Analysis(M.S.A.)

Failure Modes

Beams crM kNm

yM

kNm uM

kNm crΔ

mm

yΔ

mm

uΔ mm

u

cr

MM

u

y

MM

u

cr

Δ

Δ u

y

Δ

Δ

,

u

u CON

MM

,

u

u CON

Δ

Δ

,1crM

kNm ,1

cr

cr

MM

, 2crM

kNm , 2

cr

cr

MM

.calcyM

kNm

.calcuM

kNm .

y

calcy

M

M

.u

calcu

MM

.calc

yM

kNm

.calcuM

kNm .

y

calcy

M

M

.u

calcu

MM M.S.A. Test

Con 0.77 0.77 0.72 0.72 1.00 1.00 1.00 1.00 1.02 0.75 1.02 0.75 0.60 1.28 0.60 1.28 F.F. F.F.Conm 0.83 0.83 0.94 0.94 1.00 1.00 1.08 1.31 1.02 0.81 1.02 0.81 0.90 0.92 0.90 0.92 F.F. F.F.

1C 0.83 1.45 0.98 19.41 1.75 19.81 1.88 26.96 1.05 0.79 0.94 0.88 4.14 0.35 1.47 0.99 T.R. F.F.2C 0.86 3.32 0.78 28.93 3.86 37.09 4.31 40.18 1.08 0.80 0.97 0.88 5.41 0.61 3.31 1.00 T.R. F.F.1Cc 1.19 3.08 1.05 36.87 2.59 35.11 4.00 51.21 1.05 1.13 0.94 1.27 4.14 0.74 3.18 0.97 T.R. S/F2Cc 1.18 4.19 1.33 29.17 3.55 21.93 5.44 40.51 1.08 1.09 0.97 1.21 5.41 0.77 5.41 0.77 C.C. F.F.1Ccs 1.06 2.28 1.00 37.35 2.15 37.35 2.96 51.88 1.05 1.01 0.94 1.13 4.14 0.55 3.18 0.72 T.R. S/F1G 0.82 0.82 1.00 1.00 1.00 1.00 1.06 1.39 1.03 0.80 0.91 0.90 2.04 0.40 0.83 0.99 T.R. F.F.2G 0.92 1.17 0.83 5.55 1.27 6.69 1.52 7.71 1.04 0.89 0.92 1.00 3.35 0.35 1.61 0.73 T.R. F.F.1Gc 0.93 0.93 0.83 9.42 1.00 11.35 1.21 13.08 1.03 0.90 0.91 1.02 2.04 0.46 1.55 0.60 T.R. T.R.2Gc 0.94 2.90 0.83 31.62 3.09 38.10 3.77 43.92 1.04 0.91 0.92 1.02 3.35 0.87 2.98 0.97 T.R. T.R.1Gcs 1.04 1.30 1.22 12.45 1.25 10.20 1.69 17.29 1.03 1.01 0.91 1.14 2.04 0.64 1.55 0.84 T.R. T.R.GcG 0.97 1.14 1.11 10.28 1.18 9.26 1.48 14.28 1.04 0.94 0.92 1.05 3.35 0.34 1.77 0.64 T.R. T.R.CGp 1.13 1.90 0.94 19.78 1.68 21.04 2.47 27.47 1.06 1.07 0.95 1.19 4.57 0.42 3.35 0.57 T.R. S/FCGl 0.96

N.A.

1.56 0.89

N.A.

25.88 1.63

N.A.

29.08

N.A.

2.03 35.94 1.06 0.91 0.95 1.02

N.A.

4.49

N.A.

0.35

N.A.

3.25

N.A.

0.48 T.R. S/FConS 0.66 2.98 3.15 0.78 12.55 26.25 4.77 1.06 33.65 2.09 1.00 1.00 1.09 0.61 1.09 0.61 3.74 3.76 0.80 0.84 3.74 3.76 0.80 0.84 C.C. C.C.1CS 0.87 3.65 4.64 0.95 12.60 21.48 5.33 1.27 22.61 1.70 1.47 0.82 1.12 0.78 1.02 0.85 4.74 5.46 0.77 0.85 4.26 4.83 0.86 0.96 C.C.* C.C.2CS 0.92 4.32 5.69 0.83 13.10 24.17 6.18 1.32 29.12 1.85 1.81 0.92 1.15 0.80 1.06 0.87 5.74 6.33 0.75 0.90 4.57 5.50 0.95 1.03 C.C.* C.C.1CcS 0.99 4.23 5.14 0.94 13.52 20.26 5.19 1.22 21.55 1.50 1.63 0.77 1.12 0.88 1.02 0.97 4.74 5.46 0.89 0.94 4.56 5.46 0.93 0.94 C.C. C.C.1GS 0.89 3.66 4.14 0.94 12.57 18.07 4.65 1.13 19.22 1.44 1.31 0.69 1.10 0.81 1.00 0.89 4.03 4.46 0.91 0.93 3.90 4.16 0.94 1.00 C.C. C.C.2GS 0.89 3.74 4.49 0.72 13.31 20.75 5.04 1.20 28.82 1.56 1.43 0.79 1.10 0.81 1.01 0.88 4.33 4.95 0.86 0.91 4.04 4.48 0.93 1.00 C.C. C.C.1GcS 0.84 3.47 4.21 0.94 13.36 25.48 5.01 1.21 27.11 1.91 1.34 0.97 1.10 0.76 1.00 0.84 4.03 4.46 0.86 0.94 4.04 4.46 0.86 0.94 C.C. C.C.

F.F.: Flexural failure; T.R.: Textile reinforcement rupture; S/F: Combined Shear/Flexure failure; C.C.: Concrete crushing after steel reinforcement yielding; C.C.*: Concrete crushing before steel reinforcement yielding.

(a) (b)

(c) (d)


regard to ultimate load-bearing capacity and deform-ability. The observed differences are clearly attributed to the change of the textile’s effective depth.

Crashing of concrete in the compression zone for specimens with steel reinforced main parts was the

cause for load-bearing capacity reduction. The steel reinforcement (acting as crack arrestor) led to dense crack patterns with small crack widths. The effect of polymer coating of the textiles in this case was also beneficial but less pronounced.

0 4 8 12 16 20 24 28 32 36 4002468

101214161820

1GcS

ConS1GS

1CcS

2CS

2GS

Load

(kN

)

Mid-span deflection (mm)

1CS

0 4 8 12 16 20 24 28 32 36 400

2

4

6

8

10

12

14

0 1 2 3 40

1

2

3

4

Conm

Con

Load

(kN

)


1G

2Gc

GcG1Gcs

2G

Load

(kN

)


1Gc

0 4 8 12 16 20 24 28 32 36 400

2

4

6

8

10

12

14

CGp

Load

(kN

)


CGl

0 4 8 12 16 20 24 28 32 36 400

2

4

6

8

10

12

14

1Ccs

1Cc2C

2Cc

Load

(kN

)Mid-span deflection (mm)

1C

Fig. 6 Load versus mid-span deflection curves.

Fig. 7 Crack patterns for all beam-type specimens at failure.


3.2 Theoretical assessment of post cracking behavior of beam-type specimens (1) Flexural capacities Calculations were performed in order to predict first-crack, steel yielding (where applicable) and ultimate moments of the beam-type specimens based on strain compatibility and equilibrium of forces at the cross-section. The analysis (implementing a fiber model) took into account idealized material constitutive laws for concrete in compression and carbon or glass fiber textile (and steel where applicable) in tension (Fig. 8), classical assumptions of plane cross sections and negligible con-tribution of the concrete and of the fine-grained cemen-titious matrix below the neutral axis. Initially, uniform stress distribution across each roving’s cross section and perfect TRC-to-concrete bond was assumed. Possible failure modes included textile fracture (prior to or after steel yielding, if applicable) and concrete compressive crushing.

The acting flexural moment at first-crack, crM , was calculated by Eq. (2) based on the section modulus, S , of: (i) the total uncracked transformed cross-section, gS [concrete section plus cE/rE times the reinforcement area, rE and cE being the moduli of elasticity for the reinforcement (fibers and/or steel) and concrete, respec-tively]; and (ii) the uncracked transformed cross-section neglecting the textile cover (here equal to 6 mm), red_gS , (a similar suggestion was made by Matthys and Taerwe, 2000). Both approaches led to good approximation of

crM (Table 2).

SfM mtcr = (2)

Calculated values of ultimate flexural moment, uM , but also moment at steel yielding, yM (where applica-ble), are tabulated in Table 2 along with the respective ratios of experimental values over calculated ones. For specimens comprising plain concrete main parts, the section analysis overestimates the actual flexural capac-ity, this overestimation being higher for specimens with uncoated textiles. Lack of full exploitation of the textile

properties in these specimens is attributed to the inabil-ity of the matrix material to penetrate and impregnate the fiber yarns. This fact leads to the inhomogeneous stress distribution between the outer and the inner fila-ments and thus the former carry larger tensile stresses in comparison to the latter (this has also been reported by Voss and Hegger, 2006). As a result, the well-bonded glass outer filaments fail in tension and rapid crack width opening occurs until the dry core filaments are fully stressed. For the case of carbon rovings, differen-tial stress distribution results in telescopic sliding of the filaments. The calculated values of yM and uM for reinforced concrete composite beams were in fairly good agreement with the experimentally derived ones.

In an attempt to achieve better predictions of the flexural capacities of the beam-type specimens the fol-lowing alterations in the cross-section analysis were made: (i) nominal values of ultimate strains for carbon,

cfuε , and glass fibers, g

fuε , were substituted by effective ones following the rationale given in the previous sec-tion (thus c

fuc

.eff,fu ε.ε 650= and gfu

g.eff,fu ε.ε 750= ); (ii) the

effect of rovings’ coating was taken into account by introduction of two coefficients: bk quantifies the part of the textile reinforcement area that is in good contact with the cementitious matrix and, therefore, exhibits high bond conditions with the latter, and sk accounts for the ‘strain lag’ between inner and outer filaments in uncoated rovings. More specifically, for textiles with coated rovings coefficients bk and sk are taken equal to unity denoting full resin impregnation of the fiber bun-dles, perfect bond of the textile to the cementitious ma-trix and uniform strain profile in each fiber roving (or, equivalently, over the thickness of the textile reinforce-ment). For textiles with uncoated rovings coefficients

bk and sk are taken equal to 0.25 and 0.4, respectively, denoting that perfect filament-to-matrix bond conditions (attributed to mortar penetration in the rovings) are achieved within only a fraction (1/4) of the textile rein-forcement area, fA (located at the periphery of the rov-ings in practice) and that during loading the ‘core’ fiber

Fig. 8 Idealized material constitutive laws for: (a) concrete in compression and (b) carbon or glass fiber textile (and steel where applicable) in tension.

MPa.122===

≤=

≤≤=⎥⎥⎥

⎦

⎤

⎢⎢⎢

⎣

⎡

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛

cmf 0.002c2ε & 2n

cuεcε for cmfcσ

c2εcε 0 for n

c2εcε-1-1 cmfcσ

εcu

fcm

εc2

0 15 30 45 60 75 90 105 120 135 1500

500

1000

1500

2000

2500

3000

3500

4000

Glass

Carbon

Stre

ss

(MPa

)

ε ( ‰ )

Steel

(a) (b)


reinforcement is under lower strain conditions compared to the reinforcement comprising the outer well-bonded ‘sleeve’ (Fig. 9).

In Fig. 9 the limiting value of the ‘sleeve’ filaments’ strain, max,fε , is equal to .eff,fuε . The proposed coeffi-cients bk and sk are to be seen as “boxed” values, which may be altered in the future, should more ex-perimental data be systemically gathered and organized. The above considerations originate from the ideas in-cluded in the ‘adhesive cross linkage model’ [Schorn (2004)] and the ‘plug model’ [Lepenies et al. (2007)].

It should be underlined that fiber-matrix interaction is a poly-parametric problem relying on a large number of factors. The use of cementitious matrices (which alone comprise composite materials) introduces a higher de-gree of complexity in understanding and adequately modeling this interaction; matrix compositions, filament surface treatments, types of rovings’ coatings used [see also Mäder et al. (2004), Keil and Raupach (2007)] re-sult in an infinite number of possible combinations ren-dering the derivation of a global micromechanical model even more difficult. Significant experimental and analytical work focusing on bond mechanisms in TRC has been conducted by Krüger et al. (2001), Krüger et al. (2002a), Krüger et al. (2002b), Banholzer (2004), Chudoba et al. (2007) and Jesse et al. (2008), to men-tion a few. A combination of the micro- and mesoscopic models leading to the so-called Micro-Meso-Macro-Prediction Model (MMM-PM) has been proposed by Lepenies et al. (2008). In this work, and in sake of gen-eralization and simplification that should characterize any design tool, textile-to-matrix interaction is consid-ered to be primarily dependent on the rovings’ coating (or its absence thereof).

Table 2 provides the calculated values of ultimate flexural moment, uM , and moment at steel yielding,

yM (where applicable), based on the above-described modified section analysis. It is evident that better agreement is achieved for both plain concrete and rein-forced concrete composite beams.

(2) Deflections This section provides predictions of the mid-span de-flection of the tested beams at first crack, steel yielding (were applicable) and ultimate. The mid-span deflection of a beam loaded in four-point bending (including the shear component), as derived from linear elastic theory, is given by Eq. (4).

shearflexurespanmid ΔΔΔ +=− (3)

( )e

aa

ec

aspanmid GI

hMLLIE

M10

4324

222 +−=−Δ (4)

where: aM is the applied moment at the mid-span cross-section, L is the length between supports (here 1400 mm), aL is the shear span (distance from the load application point to the support, here 650 mm), cE is the modulus of elasticity of concrete (here 19.4 GPa),

eI is the effective moment of inertia ( eI provides a transition between the gross and cracked moments of inertia and is assumed to be constant along the length of the member), h is the height of the cross-section, and G is the shear modulus of concrete. The shear compo-nent of the mid-span deflection may be neglected from Eq. (4), as it comprises a very small portion of spanmid −Δ (namely, less than 1% of spanmid −Δ ).

Mid-span deflections at first crack, crΔ , were calcu-lated from Eq. (4) taking into account the second mo-ment of area of: (i) the total transformed cross-section,

gI ( 1,crΔ in Table 3); and (ii) the transformed cross-section neglecting the textile cover, red_gI ( 2,crΔ in Ta-ble 3). Use of red_gI in the calculations resulted in slightly better agreement with the experimentally de-rived values of crΔ , which were generally underesti-mated by both approaches, as shown in Table 3.

For the calculation of mid-span deflections at steel yielding yΔ , and at ultimate uΔ (for specimens with reinforced concrete main part), eI is taken from Eq. (5) and Eq. (6), respectively.

max,fmin,f ε.ε 40= ft

ft.1250

ft0.125

fε

min,fε2ft

max,fε max,fε

(a) (b) (c)

Fig. 9 Idealization of the strain profile across the cross-section of a fiber roving: (a) Uniform strain distribution in a coated fiber roving; (b) Actual strain distribution in an un-coated fiber roving; and (c) Simplified approximation of strain distribution in an un-coated fiber roving.


))xd/(ε(EM

φEM

Iysyc

y

yc

yy,e −

== (5)

)x/ε(EM

φEMI

ucuc

u

uc

uu,e == (6)

where: yφ & uφ are the beam curvatures at yielding and ultimate, respectively; yε is the yield strain of the steel reinforcement (2.86‰); cuε is the ultimate com-pressive concrete strain (3.50‰); sd is the effective depth of the steel reinforcement (75 mm); and yx & ux are the neutral axis depths corresponding to steel yield-ing and ultimate beam capacity, respectively. The latter were taken from the modified section analysis described previously. Use of Eq. (6) is justified by the failure mode of all reinforced concrete beams (due to concrete crushing). Both Eq. (5) and Eq. (6) imply a fully cracked state for the beams from the onset of steel yield-ing and onwards; visual observation of the crack pattern development in all reinforced concrete specimens be-tween yielding and specimen failure validates this as-sumption. The ratios of experimental to computed val-ues of yΔ and uΔ are given in Table 3. The systematic underestimation of yΔ ( .calc

yΔ in Table 3) – by a rela-tively small percentage ≈15% – indicates a slightly smaller degree of textile reinforcement activation at steel yielding than the one assumed by the modified section analysis. Computed deflections at ultimate ( .calc

uΔ in Table 3) show reasonable agreement to the experimentally derived ones.

A tri-linear approximation of the overall moment – mid-span deflection behavior of the specimens with reinforced concrete main parts is given hereafter based on the expression proposed by Benmokrane et al. (1996) for the calculation of eI :

3

⎟⎟⎠

⎞⎜⎜⎝

⎛⎟⎟⎠

⎞⎜⎜⎝

⎛−+=

a

crcr

red_gcre M

MaΙβ

ΙaII (7)

where: crI is the moment of inertia of the cracked trans-formed section calculated by the modified section analysis presented earlier, red_gI is the moment of iner-tia of the uncracked transformed cross-section neglect-ing the textile cover, crM is the cracking moment, a & β are empirical constants; for the tested specimens

650.a ≈ (0.62 for carbon textiles and 0.68 for glass) & β =2. Figure 10 illustrates the comparison between the experimental and the tri-linear approximation of the overall moment – mid-span deflection behavior of the above-mentioned specimens. In terms of yΔ / .calc

yΔ and uΔ / .calc

uΔ values the use of Eq. (7) results in an accuracy comparable to the one presented in Table 3. 3.3 Response of column-type specimens All specimens exhibited similar response under concen-tric monotonic compressive loading. The normalized axial stress versus axial strain curves are shown in Fig. 11. Normalized stress is defined as the ratio of the com-

pressive stress fcc over the specimen’s concrete com-pressive strength fc (specimens were cast with concrete of slightly different compressive strengths, with a mean value equal to 21 MPa). Mortar B TRC shafts developed a pattern of fine cracks prior to testing due to shrinkage. These specimens failed under lower maximum load (by 10-18%) compared to the control specimen. Better re-sponse was observed for specimens cast in mortar I TRC shafts. Yet, these specimens (but one) also failed under lower maximum load (by 6% on average) com-pared to the control specimen. This was due to early local buckling of the TRC element, as explained hereaf-ter.

From the early stages of loading vertical cracking and flaking of the matrix material was evident on the TRC elements, resulting in minor contribution of the thin-

Table 3 Predicted values of mid-span deflections at first crack, yielding (where applicable) and ultimate for all beam-type specimens.

Beams 1,crΔ

mm 1,cr

cr

ΔΔ 2,crΔ

mm 2,cr

cr

ΔΔ

.calc

yΔ mm

.calcy

y

Δ

Δ

.calcuΔ

mm .calcu

u

ΔΔ

Con 0.75 0.96 0.75 0.96 0.65 1.11Conm 0.75 1.26 0.75 1.26 0.65 1.451C 0.75 1.31 0.79 1.24 19.23 1.012C 0.76 1.03 0.81 0.97 28.35 1.021Cc 0.75 1.40 0.79 1.32 36.57 1.012Cc* 0.76 1.75 0.81 1.65 27.74 1.051Ccs 0.75 1.34 0.79 1.26 26.40 1.411G 0.75 1.34 0.79 1.26 0.69 1.452G 0.75 1.11 0.79 1.05 12.10 0.461Gc 0.75 1.11 0.79 1.05 7.08 1.332Gc 0.75 1.11 0.79 1.05 34.00 0.931Gcs 0.75 1.63 0.79 1.54 14.64 0.85GcG 0.75 1.48 0.79 1.40 10.09 1.02CGp 0.75 1.26 0.81 1.17 20.80 0.95CGl 0.75 1.19 0.81 1.11 16.88 1.53

Average: 1.29 1.22 1.11St. deviation: 0.22 0.20

N.A. N.A.

0.28ConS 0.76 1.03 0.76 1.03 10.78 1.16 28.39 0.921CS 0.77 1.23 0.83 1.15 11.22 1.12 22.43 0.962CS 0.77 1.08 0.83 1.00 11.49 1.14 19.65 1.231CcS 0.77 1.22 0.83 1.14 11.47 1.18 19.83 1.021GS 0.77 1.22 0.82 1.15 10.92 1.15 25.89 0.702GS 0.77 0.93 0.82 0.88 11.04 1.21 24.04 0.861GcS 0.77 1.22 0.82 1.15 11.04 1.21 24.23 1.05

Average: 1.13 1.07 1.17 0.96St. deviation: 0.12 0.10

0.03 0.17

* Failed in shear


walled TRC elements to the load-bearing capacity of the specimens. Until the peak load was reached, growing of the vertical cracks width was observed (especially at the specimens’ corners where stress concentration was pre-sent) and local buckling of the formwork occurred at the vicinity of the load application areas (top and bottom parts of the columns); this was accompanied by mortar spalling. The confining action of the TRC formwork

was activated after maximum load was reached, result-ing into the increase of post-peak deformability. Buck-ling of the vertical steel bars was observed at a later stage of the test (at 50% load carrying capacity reduc-tion) causing textile rupture and total failure of the specimens.

High compressive strength / low shrinkage – or even moderately expansive – matrices are recommended in

0 4 8 12 16 20 24 28 32 36 400

1

2

3

4

5

6

7

8 Experimental curve Trilinear approximation based on Eq. (7)

Mom

ent (

kNm

)


1CS

0 4 8 12 16 20 24 28 32 36 400

1

2

3

4

5

6

7

8

1GS

Mom

ent (

kNm

)


0 4 8 12 16 20 24 28 32 36 400

1

2

3

4

5

6

7

8

1CcS

Mom

ent (

kNm

)

Mid-span deflection (mm)0 4 8 12 16 20 24 28 32 36 40

0

1

2

3

4

5

6

7

8

1GcS

Mom

ent (

kNm

)


0 4 8 12 16 20 24 28 32 36 400

1

2

3

4

5

6

7

8

2CS

Mom

ent (

kNm

)

Mid-span deflection (mm)0 4 8 12 16 20 24 28 32 36 40

0

1

2

3

4

5

6

7

8

2GS

Mom

ent (

kNm

)

Mid-span deflection (mm) Fig. 10 Comparison between the experimental and the tri-linear approximation of the overall moment – mid-span deflec-tion behavior of beam-type specimens with reinforced concrete main parts.


order to avoid the premature failure of TRC element. Additionally, the installation or the formation of shear connectors on the inner side of the TRC elements is expected to improve composite action between the SiP formwork and the core concrete. It is pointed out that the main purpose of use of these innovative composite elements is not augmenting the load carrying capacity of axially loaded structural members. The main structural advantage of these systems, if used as participating formwork, lies in the increase of deformability. In this study (using low textile reinforcement ratios), the ex-perimental results indicate an increase in deformability by 25% to 75% (corresponding to the axial strain at a post-peak axial stress equal to 80% of the specimen’s compressive strength) compared to the monolithic specimen.

4. Conclusions

The experimental investigation presented in this study provides useful information on the behavior of compos-ite TRC/concrete structural elements which can be used for the conduction of large scale tests. For beam-type specimens, the use of polymer-coated textiles and higher textile reinforcement ratios results in higher load-carrying and deformation capacities. The effect of the TRC/concrete interface roughness (based on the limited experimental data presented herein) was found to be rather small. Textiles in participating SiP TRC form-work elements may serve as the composite beams’ pri-mary reinforcement substituting discrete steel rein-forcement bars and resulting in mass savings. The ide-alization of the strain profile across the cross-section of a fiber roving proposed in this work renders the fiber model-based section analysis a reliable tool for the pre-diction of the ultimate flexural capacities of composite TRC/concrete beams. The use of precast SiP TRC shafts as formwork elements for columns comprises an inter-esting prospect provided that in-situ construction prac-tices will inhibit local buckling of the shells.

References Banholzer, B. (2004). “Bond behaviour of a multi-

filament yarn embedded in a cementitious matrix [online].” Thesis (PhD). Institut für Bauforschung, RWTH Aachen, Aachen, Germany. Available from: <http://deposit.ddb.de/cgi-bin/dokserv?idn=97401495 8&dok_var=d1&dok_ext=pdf&filename=974014958.pdf>[Accessed 25 Aug 2009].

Benmokrane, B., Chaallal, O. and Masmoudi, R. (1996). “Flexural response of concrete beams reinforced with FRP reinforcing bars.” ACI Structural Journal, 93(1), 46-55.

Brameschuber, W., Koster, M., Hegger, J. and Voss, S. (2008). “Integrated formwork elements made of textile-reinforced concrete.” ACI SP-250 (CD) “Textile-Reinforced Concrete.” A. Dubey Ed.

Chudoba, R., Konrad, M., Vořechovskŷ, M. and Roye, A. (2007). “Numerical and experimental study of imperfections in the yarn and its bond to cementitious matrix.” ACI SP-244 (CD) “Thin Fiber and Textile Reinforced Cementitious Systems.” C.-M. Aldea Ed.

Jesse, F., Will, N., Curbach, M. and Hegger, J. (2008). “Load-bearing behaviour of Textile Reinforced Concrete.” ACI SP-250 (CD) “Textile-Reinforced Concrete.” A. Dubey Ed.

Keil, A. and Raupach, M. (2007). “Improvement of the Load-Bearing Capacity of Textile Reinforced Concrete by the Use of Polymers.” In: Yeon, K.-S., Ed. 12th International Congress on Polymers in Concrete – ICPIC'07, Chuncheon, Korea 26-28 September 2007. Korea Science and Engineering Foundation, 873–881.

Krüger, M., Reinhardt, H. W. and Fichtlscherer, M. (2001). “Bond behavior of Textile reinforcement in reinforced and prestressed concrete elements.” Otto-Graf Journal, 12, 33-50.

Krüger, M., Ozbolt, J. and Reinhardt, H. W. (2002a). “A discrete bond model for 3D analysis of textile reinforced and prestressed concrete.” Otto-Graf Journal, 13, 111-128.

0 2 4 6 8 10 12 14 16 18 200.0

0.2

0.4

0.6

0.8

1.0

1.2

ε ( ‰ )

f cc

/ fc

BCc45BC90

ICc45

Con

IC90

0 2 4 6 8 10 12 14 16 18 200.0

0.2

0.4

0.6

0.8

1.0

1.2

f cc

/ fc

ε ( ‰ )

IGc90

BGc45

BG90 BGc90

IGc45

Con

IG90

Fig. 11 Normalized axial stress versus axial strain curves for composite RC column-type specimens under concentric compression cast in: (a) carbon-TRC and (b) glass-TRC shafts.

(b) (a)


Krüger, M., Xu, S., Reinhardt, H. W. and Ozbolt, J. (2002b). “Experimental and numerical studies on bond properties between high performance fine grain concrete and carbon textile using pull out tests.” In: Beitrage aus der Befestigungstechnik und dem Stahlbetonbau (Festschrift zum 60. Geburtstag von Prof. Dr.-Ing R. Eligehausen), Stuttgart, 2002, 151-164.

Lepenies, I., Meyer, H., Schorn, H. and Zastrau, B. (2007). “Modeling of load transfer behavior of AR-glass-rovings in textile reinforced cconcrete.” In Thin Fiber and Textile Reinforced Cementitious Systems. ACI SP-244 (CD) “Thin Fiber and Textile Reinforced Cementitious Systems.” C.-M. Aldea Ed.

Lepenies, I., Richter, M. and Zastrau, B. (2008). “A multi scale analysis of textile reinforced concrete structures.” In: 79th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM), Bremen, Germany 31 March-4 April 2008, Weinheim: Wiley Publications, 10553–10554.

Mäder, E., Plonka, R., Schiekel, M. and Hempel, R. (2004). “Coatings on alkali-resistant glass fibres for the improvement of concrete.” Journal of Industrial

Textiles, 33(3), 191-206. Matthys, S. and Taerwe, L. (2000). “Concrete slabs

reinforced with FRP grids. I: One-way bending.” Journal of Composites for Construction, 4(3), 145-153.

RILEM Technical Committee 201, (2006). “Textile Reinforced Concrete.” W. Brameshuber, Ed. Bagneux: RILEM Publications S.A.R.L. RILEM report 36.

Schorn, H. (2004). “An adhesive cross-linkage model for textile reinforcement in concrete.” In: 2nd International Conference on Structural Engineering, Mechanics and Computation SEMC, Cape Town, South Africa 5-7 July 2004, 1563–1567.

Voss, S. and Hegger, J. (2006). “Dimensioning of textile reinforced concrete structures.” In: J. Hegger, W. Brameshuber and N. Will, Eds. 1st International RILEM Symposium Textile Reinforced Concrete, Aachen, Germany, 9-14 September 2006. Bagneux: RILEM Publications S.A.R.L., 151-160.

Wrigley, R. G. (2001). “Permanent formwork in construction.” London: Construction Industry Research and Information Association (CIRIA).

Mechanical Behavior of Textile Reinforced Concrete (TRC ...

Documents