Shear behaviour of SFR-UHPC I-shaped beams

0 downloads 0 Views 3MB Size Report
ing, Technical University of Cluj Napoca, Romania. E-mail address: ... beams, two times 5 I-shaped beams casted with SFR-UHPC (steel fibre reinforced ...
Construction and Building Materials 124 (2016) 258–268

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Shear behaviour of SFR-UHPC I-shaped beams Raul Zagon a,b,⇑, Stijn Matthys b, Zoltan Kiss a a b

Department of Structures, Faculty of Civil Engineering, Technical University of Cluj Napoca, Romania Magnel Laboratory for Concrete Research, Dep. of Structural Engineering, Ghent University, Belgium

h i g h l i g h t s  A mix of short fibres and long fibres has been chosen to be added in the concrete.  All the tested beams showed a shear failure through the web.  The conducted tests confirm the feasibility of SFR-UHPC for I-shaped concrete beams.

a r t i c l e

i n f o

Article history: Received 18 November 2015 Received in revised form 3 July 2016 Accepted 15 July 2016

Keywords: Shear capacity Web openings UHPC Steel fibre SFR

a b s t r a c t The paper presents the details of an experimental study on the shear behaviour of SFR-UHPC (steel fibrereinforced ultra-high performance concrete) I-shaped beams with or without web openings. In addition, partial replacement of shear links is investigated by comparing steel fibres with or without a single additional diagonal rebar. The object of the research was to study the shear capacity of 10 I-shaped beams made from SFR-UHPC that were tested in shear until failure with two a/dratios. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction The use of SFR-UHPC (steel fibre reinforced ultra-high performance concrete) has gained interest during the last years in relation to more advanced structural applications [1], such as the shear resistance of reinforced concrete beams. In order to avoid the time consuming process of assembling stirrups (shear links) for reinforced concrete beams, research has focussed on finding alternative methods. In several cases [1–8], an effective solution was found to be the use of fibre reinforced concrete. With an appropriate fibre dosage and the possible use of ultra-high performance concrete, it is proven that traditional reinforcement can be replaced or reduced. Moreover, in case of a total replacement, the thickness of the web can be reduced, as the concrete cover for the stirrups is not needed anymore. Especially the combination of steel fibre and ultra-high performance concrete is of interest, as SFR-UHPC is a concrete mix with superior properties compared to conventional concrete [9]: a compressive strength higher than ⇑ Corresponding author at: Department of Structures, Faculty of Civil Engineering, Technical University of Cluj Napoca, Romania. E-mail address: [email protected] (R. Zagon). http://dx.doi.org/10.1016/j.conbuildmat.2016.07.075 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.

150 MPa, higher tensile strength, a non-brittle behaviour and a low water binder ratio. Combining these aspects new structural solutions can be achieved. Due to the fact that modern buildings need more space for utilities such as HVAC (heating, ventilation and air-conditioning), this requires more pipes and ducts to be integrated in the structure. By placing these utility pipes through the beams the total construction height can be reduced. The web openings modify however the mechanical behaviour of the beams by concentrated stresses and earlier cracking of the concrete in those areas, resulting in reduced shear capacity. Special reinforcement detailing normally needs to be applied for a better crack control and to prevent the failure of the beam through web openings [10]. Given the above aspects a test programme has been conducted to study the replacement of stirrups by steel fibres and to compare beams with or without web openings. In addition, partial replacement of shear links is investigated by comparing SFR-UHPC beams with or without a single additional diagonal rebar. A mix between short straight fibres and long hooked fibres has been chosen to ensure the best contribution to the shear resistance of the beams, given the fact that small fibres work better on micro-cracks while long fibres are starting to work after the cracks appeared [11]. The

259

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268

pull-out behaviour of these types of fibres has been studied in e.g. [12]. Based on the experimental and analytical work reported in this paper, the feasibility is demonstrated of the novel type of I-shaped beams in SFR-UHPC. 2. Experimental program 2.1. Test specimens To investigate the shear capacity of fibre reinforced concrete beams, two times 5 I-shaped beams casted with SFR-UHPC (steel fibre reinforced ultra-high performance concrete) were tested in shear until failure. The test specimens (Fig. 1) have a total length of 4000 mm, a cross section of 140 mm width, 400 mm height and a web thickness of 60 mm. The bottom flange has a height of 80 mm while the upper flange has a height of 60 mm. The five different types of beam, their designation and test parameters are given in Table 1. In order to assure a shear failure, a high amount of longitudinal rebars was placed, 3 rebars of diameters 22 mm, corresponding to a geometrical reinforcement ratio of 3.16%. For the beams with a single diagonal shear link, rebars (/12) inclined at 45° were adopted, while single leg stirrups with a diameter of 8 mm, placed each 100 mm, have been used for the beam type with vertical shear links. In order to anchor the longitudinal reinforcement at the ends, welded transverse bars of 10 mm diameter were present. All the casted elements have the same SFR-UHPC composition, and where manufactured in identical pairs. This allowed doing each test configuration two times, to obtain an indication of the test variability. The five different types of beams are designated as follows (Fig. 1):  Type F: SFR-UHPC with only longitudinal reinforcement;  Type FD: type F with single diagonal shear link;

Table 1 Test matrix. Test designation

a/d

Web openings (mm)

Diagonal shear link (mm)

Stirrups (mm)

F_25.1 and F_25.2 F_23.1 and F_23.2 FO_25.1 and FO_25.2 FO_23.1 and FO_23.2 FD_25.1 and FD_25.2 FD_23.1 and FD_23.2 FOD_25.1 and FOD_25.2 FOD_23.1 and FOD_23.2 FS_25.1 and FS_25.2

2.5 2.3 2.5 2.3 2.5 2.3 2.5 2.3 2.5

– – 160 160 – – 160 160 –

– – – – /12 /12 /12 /12 –

– – – – – – – – /8 @ 100 mm

 Type FO: type F with web opening;  Type FOD: type FO with single diagonal shear link;  Type FS: type F with traditional stirrups. 2.2. Test set-up The beams were tested in shear according to the test set-up outlined in Fig. 2. Hereby the position of the supports has been chosen to test the specimens at two opposite beam ends and for two different shear span to effective depth (a/d) ratios. A hinge and roller support were used to avoid horizontal constrain of the beam. The load was applied on a metal plate with a spherical hinge, and between the jack and the beam a loading cell with an accuracy of 0.3 kN was placed. In this way the load is applied on a load path of 260 mm diameter. An overview of the test matrix is given in Table 1. Beams F_25, FD_25, FO_25, FOD_25 were tested with a span of 2.73 m and a ratio a/d = 2.5 and beams F_23, FD_23, FO_23 and FOD_23 were tested in the opposite shear span, with a span length of 2.53 m and a ratio a/d = 2.3. The load was gradually increased until failure, using load steps of 10 or 20 kN. During the tests at each load level, the following

Fig. 1. Schematic of beam configuration (dimensions in mm).

260

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268

Fig. 2. Test scheme and instrumentation of the beams (dimensions in m).

manual and electronic measurements were conducted. The vertical deflection was measured with displacement transducers, while concrete strains were measured with digital gauges and a Hugenberger deformeter. Crack evolution was recorded at every loading step in terms of crack appearance and crack widths. For a better view of the cracks, one half of the beam was painted in white. Strain readings of the diagonal shear links has been conducted by means of strain gauges installed at mid-height. To monitor the strains of the longitudinal reinforcement, a digital gauge was fixed on it. To do this, before casting of the beams, 2 screws were welded perpendicular to the reinforcement directed to the bottom part at a distance of 200 mm, and they were covered with two small pieces of polystyrene, which were removed later and the gauge could be easily fixed. 2.3. Materials The UHPC mix is based on research conducted at RWTH-Aachen University [13]. The mix has been further optimised, in order to allow the use of local constituent materials and in order to upscale the concrete mix for use in an industrial mixing plant. The final mix design is shown in Table 2. The concrete casting was done in a precast concrete plant at CON-a Sibiu. All the beams were casted in 2 batches, 5 beams per batch. Fibre reinforcement was added at the beginning of the

Table 2 Concrete composition (1 m3). CEM I 52.5R Andesite (4–8 mm) Sand (0,125-4 mm) Silica Fume (SIKA) Quartz powder Steel fibres Water Superplastifiant

mixing process together with the andesite aggregates. In this way a good dispersion of fibres has been assured. To determine the concrete properties, compressive strength (fcm,cube), Young’s modulus (Ec), splitting tensile strength (fctm,sp) and bending tensile strength (fctm,fl), cubes with side length of 100 mm and prisms of 100 mm  100 mm  300 mm and 100 mm  100 mm  450 mm were produced. The quality control specimens were tested at 7 days, 28 days, age of beam testing and after 128 days. The obtained properties of the hardened concrete are reported in Table 3. The values represent the mean value of 3 control specimens, as well as the overall average and standard deviation of the 2 batches together. The concrete strength obtained for the 2 batches was very similar. At the age of testing the standard deviation on the concrete strength remained limited to 2.5% of the mean compressive strength. The fibres used are a mix of 1/3 from the total quantity of short straight fibres with a length of 9 mm (type WHS-9/0.175/S), and 2/3 long hooked-end fibres with a length of 25 mm (type WMS25/0.4/H/430). The total amount of fibres for 1 m3 of concrete was 78.5 kg. This corresponds with a volume fraction of 1%, which can be regarded low to moderate for fibre reinforced concrete (FRC). The properties of the steel fibres as given by the manufactures are reported in Table 4.

Table 3 Concrete properties.

Microfibres Macrofibres

650 kg 598 kg 354 kg 184 kg 456 kg 26 kg 52.5 kg 178 l 31 l

Batch

1 2 Average St. dev.

fctm,sp

Ec

7 days [MPa]

fcm,cube 28 days [MPa]

Testing day [MPa]

128 days [MPa]

Testing day [MPa]

Testing day [MPa]

85.7 80.6 83.2 3.6

141.6 135.9 138.8 4.0

135.1 140.0 137.6 3.5

159.5 146.3 152.9 9.3

12.2 12.2 12.2 0.0

43,470 43,005 43237.5 328.8

261

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268 Table 4 Steel fibre properties.

Table 5 FRC properties from bending tests at 128 days.

Properties

Macrofibres

Microfibres

Batch

Shape Length (Lf) [mm] Diameter (/f) [mm] Aspect ratio [Lf//f] Density [kg/m3] Tensile strength [MPa] E-modulus [MPa]

Hooked 25 0.4 62.5 7850 >1500 200,000

Straight 9 0.175 51.4 7850 >2200 200,000

fctm,fl [MPa]

fRm,1 [MPa]

fRm,2 [MPa]

fRm,3 [MPa]

fRm,4 [MPa]

1 2 Average St.dev. between batches

11.73 12.32 12.02 0.41

11.38 11.96 11.67 0.41

6.28 6.57 6.43 0.20

4.15 4.55 4.35 0.28

3.06 3.20 3.13 0.09

Table 6 Steel properties.

Fig. 3. Test set-up for FRC bending tests (dimensions shown for batch 1, in mm).

For the post cracking behaviour of the SFR-UHPC, crack opening controlled bending tests have been performed at an age of 128 days according to RILEM [15], on prisms of 100 mm  100 mm  450 mm (first batch), and 100 mm  100 mm  300 mm (second batch). The span for the tests was equal to 400 mm and 280 mm, respectively. Each element was turned 90° from the casting surface and a notch with a width of 2 mm and a depth of 27 mm was provided at the mid-span of the elements (Fig. 3). The specimens were loaded under three-point bending. During the tests, deflection, load and CMOD (crack mouth opening displacement) were recorded (Fig. 4). For the measurement of the crack mouth opening a sensor was fixed at each side of the prism. The experimental results in terms of flexural tensile stress-CMOD diagrams are given in Figs. 5 and 6. The post-cracking behaviour of the SFR-UHPC can be further represented by considering the bending tensile strength of the maximum load (fctm,fl), as well as the residual bending tensile strength fRm,1 till fRm,4. Hereby, fRm is defined at a CMOD of 0.5 mm, 1.5 mm, 2.5 mm, and 3.5 mm, respectively. The values are given in Table 5 and represent the mean value of 3 CMOD specimens per batch, as well as the overall average and standard

Rebar Type

Diameter [mm]

fy [MPa]

ft [MPa]

eu [%]

S355 S355 S355 S345

8 10 12 22

502 447 371 365

712 634 561 558

18 18 16 29

deviation of all specimens. A consistent FRC material behaviour is obtained between the average values of the 2 batches, with low values of standard deviation for the different crack mouth opening displacements (Table 5). The standard deviation between all the individual tests is given in Table 5 as well, and can be noted from Figs. 5 and 6. As can be observed from Figs. 5 and 6 the SFR-UHPC mix demonstrates a strain hardening between the cracking stress of about 7 MPa and the maximum tensile strength of about 12 MPa. The latter value occurs at a CMOD of about 0.4 mm. At higher CMOD values a strain softening is observed whereby on average fRm,3 (corresponding with a CMOD of 2.5 mm and often used in design formulations) equals 4,35 MPa. To characterize the tensile properties of the steel reinforcement, tensile coupon testing has been performed. The results in terms of tensile yield strength (fy), tensile strength (ft) and strain (eu) at ft of the used mild steel rebars are reported in Table 6. The stressstrains diagrams for the /12 mm diagonal shear links and for the /8 mm stirrups are given in Fig. 7. 3. Test results and discussion The main test results for the beams are given in Table 7 in terms of load at which a first shear crack starts to appear (Vcr), obtained shear capacity (Vexp), ratio (Vexp/Vexp.F) of obtained shear capacity with respect to the considered reference situation of beam type F_25 (no additional shear links, no web opening, a/d = 2.5), and failure aspect. The shear load values (V) are obtained from the applied point load (Q) on the beam, considering the shear span (a/d = 2.3 and 2.5, respectively) versus span length and the self-weight of the beam (see also Fig. 2). The formula used for the calculation of shear load values is:



Fig. 4. Displacement sensor for CMOD.

Q  ðl  aÞ l

ð1Þ

where l is the distance between the supports. The observed values Vcr are based on visual inspection at the different load intervals for manual measurements (load steps of 10 or 20 kN), while at that point flexural cracking was already presented in [14]. Flexural cracking first occurred under the point load, at about 20 kN, which is less than 10% of the maximum load. Shear cracking occurred at 50–80 kN (Table 7) by means of diagonal tension web shear cracks. This corresponded with 30–93% of the maximum load. As is made clear in Fig. 16, showing the crack patterns, both web shear cracks and flexure-shear cracks occurred. The beams failed in shear (except beams type FS, which failed in

262

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268

Fig. 5. Tensile stress-CMOD diagram for the first batch.

Fig. 6. Tensile stress - CMOD diagram for the second batch.

bending) by means of a web shear crack developing into a critical shear crack. In the beams with diagonal rebars, rupture of the latter has been observed.

The load-deflection response of the beams is given in Figs. 11–15. As can be noted from these figures, some of the beams failing in shear also passed the point of yielding of the longitudinal

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268

263

Fig. 7. Tensile stress-strain diagram for the shear links.

Table 7 Main test results.

F_25 FD_25 FO_25 FOD_25 FS_25** F_23 FD_23 FO_23 FOD_23

Vcr.1 (kN)

Vcr.2 (kN)

Vcr (kN)

Vexp.1 (kN)

Vexp.2 (kN)

Vexp (kN)

Vexp/ Vexp.F

Failure aspect*

80 60 50 80 – 80 80 70 80

60 60 50 60 – 80 80 70 80

70 60 50 70 – 80 80 70 80

149 152 70 133 >168 114 206 75 154

93 168 70 125 >199 123 172 85 154

121 160 70 129 >184 119 189 80 154

1.00 1.32 0.58 1.06 >1.52 0.98 1.56 0.66 1.27

S S S S F S S S S

* S: shear failure by means of web shear crack developing into a critical shear crack; F: flexural failure (test stopped at extensive flexural cracking and steel yielding, yet before concrete crushing). ** For those beams failing in bending the reported shear capacity should be regarded as lower bound value.

steel. A further discussion of the obtained test results is given per beam type in the appendix. 3.1. Comparison between test series In Fig. 8 a comparison is made between the average capacities of the tested beams failing in shear. On the vertical axis the ultimate shear force can be found for the different beam types and a/d ratios. The figures also provides an indicator of the variability between the 2 identical specimens tested each time. By adding one diagonal shear link, the shear capacity of beams FD_25 and FD_23 increased with 32% and 59% compared to reference beams F_25 and F_23. For the beams with web opening (FO_25 and FO_23) the shear capacity is 42% and 33% less compared to the beams without

Fig. 8. Comparison of capacity for tested beams failing in shear.

web opening. Beams FOD that differ from beams FO by adding one diagonal shear link had an increase with 84% and 93% the ultimate shear load. The shear capacity of the beams with web opening and single shear link (FOD) exceeds that of the reference beams (F) by 7– 29%. Furthermore, it can be noted that the shear capacity of the beams tested with a/d = 2.3 is systematically higher (14–32%) than the beams with a/d = 2.5. This indicates that the shorter shear span resulted in a higher degree of direct load transfer to the support through arching action [16]. As an exception, this trend is not cleary observed for beam type F (only FRC to take the shear load along the span length). This is explained by the high fibre variability of beams type F_25, which appeared to be of more governing influence than the arching effect. As one of the main observations from the tests, the replacement of the traditional stirrups by dispersed steel fibres is feasible for

264

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268

moderate shear load levels (the shear capacity of beam type F being about 65% of that of beam type FS with high amount of stirrups, such that flexural failure was obtained), while the use of one additional diagonal shear link in combination with dispersed steel fibres allowed a shear capacity close to the beams with stirrups failing in flexure (the shear capacity of beam type FD being about 95% of beam type FS). As such, the combined use of dispersed steel fibres and single diagonal shear links appeared also efficient in case of web openings (the shear capacity of beam type FOD being about 118% of beam type F). 4. Shear capacity The calculation of the shear capacity of a beam without stirrups and with disperse fibres involves a number of parameters like volume of fibres used, shear span to depth ratio and the size of the beam section. In this section a comparison is made between the shear capacity of beams type F and two analytical models: Narayanan and Darwish [17] and Model Code 2010 [18]. Narayanan and Darwish [17] observed that a volume of 1‰ fibres used in a beam without stirrups can increase the ultimate shear strength with up to 170%. In their calculation method, a fibre factor (F) is introduced:

 F¼

L  q  df D f

 ð2Þ

where L is the length of the fibre, D is the fibre diameter, qf is the fibre volume fraction and df is a factor that counts for the bond characteristic of the fibres. Based on a large number of tests a df value of 0.5 was found for straight fibres, 0.75 for crimped fibres and 1 for indented fibres. The shear capacity of a FRC beam can be calculated by superposition of the different phenomena contributing to the actual shear capacity:

Vu ¼ Va þ Vb þ Vc þ Vd

ð3Þ

where Va is the vertical component of the aggregate interlocking force, Vb is the vertical component of fibre pull-out forces along the inclined cracks, Vc is the shear force across the compressive zone and Vd is the transverse force induced in the longitudinal rebars by dowel action (Fig. 9). The combined effect following Narayanan and Darwish [17] and considering the fibre factor F, is expressed as:

    qd þ 0:4  s  F  bw  d V u ¼ e  0:24  f sp þ B  a

ð4Þ

where e is the arch action factor, e ¼ 2:8 da for a/d < 2.8, f sp is the split-cylinder tensile strength of concrete, B is a dimensional constant, q is the flexural reinforcement ratio, a/d is the shear spandepth ratio, s is the fibre-matrix interfacial bond stress, F is the fibre

Fig. 9. Forces in the shear span at inclined crack in a beam with steel fibres.

Table 8 Comparison of ultimate shear force. Method of calculation

Ultimate shear force Vcalc (kN)

Vexp/Vcalc

Narayanan and Darwish

a/d = 2.5 91

a/d = 2.3 103

a/d = 2.5 1.32

a/d = 2.3 1.15

Model Code 2010

82

82

1.03

1.32

factor, bw is the smallest width of the cross-section in the tensile area, and d is the effective depth of the cross-section. According to Model Code 2010 [18] the shear resistance can be calculated according to the following equation:

"  #   13 f V u ¼ k  100  q1  1 þ 7:5  F:tu  f c  0:18 þ 0:15rcp  bw  d f ct ð5Þ where k is the strength reduction factor, q1 is the reinforcement ratio for longitudinal reinforcement, f F:tu is the ultimate residual strength (post-cracking strength for ultimate crack opening), f ct is the value of the tensile strength for the concrete without fibres, f c is the cylinder compressive strength of concrete, rcp is the average stress acting on the concrete, due to loading or prestressing actions, bw is the smallest width of the cross-section in the tensile area and d is effective depth of the cross section. A comparison between the analytical and experimental shear capacity is given in Table 8. Fairly accurate, yet conservative results are obtained. Comparing beam types F and FD a difference in average load of 55 kN is observed. The capacity of the single diagonal shear link (VD) equals:

V D ¼ Asw  f y  sin a

ð6Þ

with, Asw the area of diagonal reinforcement, f y the yielding strength of the diagonal reinforcement and a the inclination angle of the diagonal rebar. Introducing web openings in the beam, the predictive model should be further adopted to calculate the shear capacity of a beam without stirrups, with web opening, diagonal shear link and dispersed steel fibres (beam type FOD). Adopting a simplified approach, the shear resistance can be calculated as follows: ( )    1=3 h  hg f V Rd ¼ þ 0:15  rcp  0:15  k  100  q1  1 þ 7:5  Ftu  f c h f ct X   bw  d þ Asw;i  f y  sin a

ð7Þ where hg is the height of the web opening, Asw;i is the area of the shear link, f yd is the design value of tension yield stress of nonprestressing reinforcement and a is the angle that is made between the shear link and the horizontal (Fig. 10). In the case of the beams

Fig. 10. Beam with web opening and shear links.

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268 Table 9 Shear strength prediction for beams type FOD. Beam type

Vexp (kN)

Vcalc (kN)

Vexp/Vcalc

FO_25.1 FO_25.2 FO_23.1 FO_23.2

70 70 75 85

50

1.40 1.40 1.50 1.70

FOD_25.1 FOD_25.2 FOD_23.1 FOD_23.2

133 125 154 154

85

1.56 1.45 1.81 1.81

in this test program a single diagonal shear link has been used PA ¼ Asw;1 ). ( sw;i A comparison between the analytical and experimental shear capacity of the beams type FO and FOD is given in Table 9. Conservative predictions are obtained.

265

openings and a single diagonal shear link (type FOD) the resistance was about 80% of the beams with single diagonal shear link (type FD). The use of the fibres and diagonal reinforcement adds strength and ductility requirements.  From the analytical verification it appeared that the shear resistance calculated after Narayanan and Darwish and Model Code 2010, gave conservative predictions for the beams considered in this test programme. Considering that analytical verification has been done without considering characteristic values and safety coefficients, we can conclude that Model Code 2010 will gives save confidence in predicting the shear capacity of the developed SFR-UHPC I-shaped concrete beams without stirrups and with disperse fibre reinforcement.  The simplified equation proposed in this study for beams with web opening and single diagonal shear link (type FOD) gives very conservative predictions of the shear strength. Further research work is needed to establish a more accurate prediction.

5. Conclusion

Acknowledgements

A total of ten I-shaped beams were tested to investigate the shear behaviour of SFR-UHPC beams, including the influence of web openings, single diagonal shear links and stirrups on the mechanical behaviour of these elements. From the results of this study the following conclusions can be drawn:

This research is supported by the project ‘‘Improvement of the doctoral studies quality in engineering science for development of the knowledge based society-QDOC” contract No. POSDRU/107/1.5/S/78534, project co-funded by the European Social Fund through the Sectorial Operational Program Human Resources 2007–2013. The author would like to acknowledge the support of CON-a company for financing and producing the test elements.

 All the tested beams, except those with both stirrups and steel fibres, showed a shear tension failure through the web. Both web shear cracks and flexural shear cracks were observed. When the shear cracks opened the sound of fibres being pulled out and failing was observed. Besides distribution of shear forces over the fibre reinforced shear crack, additional shear was carried by the concrete in the compression zone and by the dowel action.  The conducted experimental work confirms the feasibility of SFR-UHPC for I-shaped reinforced concrete beams and replacing classical stirrups by fibre reinforced concrete, with or without the combination of a single diagonal rebar as additional shear link.  It was observed that all the beams with single diagonal shear link collapsed at a more than 50% higher load, compared to the equivalent beams with only longitudinal reinforcement. As all beams have the same amount of fibres, the higher failure load relates to the additional single 45° shear link.  In general, beams with higher shear span to depth ratio failed at a lower shear capacity. For the conducted tests, test results at a/d = 2.5 were 12–16% lower than those of a/d = 2.3, and was attributed to a direct shear arching effect for the shorter shear span;  The shear resistance of the beams with web openings was about 62% of the reference beams (type F), while for beams with web

Appendix A. Appendix A.1. Beams type F In Fig. 11 (left), the load-deflection behaviour of the 2 tested beams type F with a/d = 2.5 is shown. Shear cracks appeared at about 60 kN and 80 kN, resulting in a shear failure of 149 kN and 93 kN, for beams F_25.1 and F_25.2 respectively. The large variation in failure load (60%) between these identical beams is remarkable. This was not observed for the same beams with a/d = 2.3 (variation in failure load 8% – Fig. 11, right) or with other beam types. As no anomalies were observed during testing of beams F_25.1 and F_25.2, the large variation in failure load is assumed to relate to an unexpected variability in effective fibre content and distribution between these 2 specimens. The shear failure of the beams type F was rather explosive, with the beam end being separated from the element in a sudden way. A.2. Beams type FD In Fig. 12 (left), the load-deflection behaviour of the beams type FD tested with a/d = 2.5 is shown. Due to the additional single

Fig. 11. Deflection under loading point for F beams with a/d = 2.5 and a/d = 2.3.

266

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268

Fig. 12. Deflection under loading point for FD beams with a/d = 2.5.

Fig. 13. Deflection under loading point for FO beams with a/d = 2.5 and a/d = 2.3.

Fig. 14. Deflection under loading point for FOD beams with a/d = 2.5 and a/d = 2.3.

diagonal shear link, and considering the average of the 2 tested beams, these elements failed in shear at a 32% higher load than the equivalent beams without web opening (F_25). Shear cracks appeared at 60 kN, which is similar to beams type F. In Fig. 12 (right) the load–deflection behaviour of beams type FD with a/ d = 2.3 is shown. The first shear crack appeared at about 80 kN and the ultimate shear force was 206 kN and 172 kN, for beams FD_23.1 and FD_23.2 respectively. These beams failed at a 59% higher load than beams type F tested at a/d = 2.3. The cracks observed were denser compared to the rest of the elements. At the ultimate state, the diagonal shear links started to yield with increasing shear crack opening and finally shear failure. The failure behaviour was more ductile compared to the beams without single diagonal shear link.

failure appeared at about 42% lower load level than for the equivalent reference beams type F tested at a/d = 2.5. In Fig. 13 (right) the load–deflection behaviour of beams type FO with a/d = 2.3 is shown. The first shear cracks appeared at about 70 kN, resulting in an average shear failure of 80 kN for beams FO_23.1 and FO_23.2, at a 33% lower load than for the beams F_23.1 and F_23.2.

A.3. Beams type FO In Fig. 13 (left), the load-deflection diagram is presented of beams type FO tested with a/d = 2.5. Shear cracks appeared at about 50 kN, resulting in a shear failure of 70 kN, for both beams FO_25.1 and FO_25.2. For these beams with web opening, the

Fig. 15. Deflection under loading point for FS beams with a/d = 2.5.

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268

Fig. 16. Crack pattern for tested beams.

267

268

R. Zagon et al. / Construction and Building Materials 124 (2016) 258–268

The critical shear crack leading to failure is located through the web opening. A.4. Beams type FOD In Fig. 14. (left), the load-deflection behaviour of the 2 tested beams type FOD with a/d = 2.5 is shown. Shear cracks appeared at about 50 kN, resulting in a shear failure of 133 kN and 125 kN, for beams FOD_25.1 and FOD_25.2, respectively. Due to the single diagonal shear link, these beams failed in shear at an 84% higher load than beams type FO tested at the same shear span. The capacity of the beams with respect to FD is 81%, yet exceeds the capacity of the reference beams F (no web opening and no shear link) with 6%. In Fig. 14 (right) the load–deflection diagram of beams type FOD with a/d = 2.3 are presented. The first shear cracks appeared at about 60 kN and the failure was at 154 kN. The ultimate load was 92% higher than the beams type FO tested at a/d = 2.3. Similar trends are observed between a/d 2.5 and 2.3 respectively, though for the latter the additional single shear link seems to have a somewhat more pronounced effect. This is because the higher direct shear load transfer to the support, for the shorted shear span a/ d = 2.3. The failure was more ductile comparative to the beams without diagonal reinforcement. All the failures happened through the web openings. A.5. Beams type FS In Fig. 15, the load-deflection behaviour of the 2 tested beams type FS with a/d = 2.5 is shown. The tests for these elements were stopped when reaching a deflection of at least 8 mm under the load. At that point there was extensive flexural cracking, the maximum crack width in flexure was over 3 mm and there was no sign of the development of a critical shear crack. The failure didn’t appear as the full stroke of the hydraulic jack was reached, though with the longitudinal steel yielding, a flexural failure was likely to occur. Because a flexural failure was obtained for a/d = 2.5, tests for a/d = 2.3 were not further performed.

References [1] R. Narayanan, I. Darwish, Use of steel fibers as shear reinforcement, ACI Struct. J. 84 (3) (1987). [2] P. De Pauw et al., Replacement of shear reinforcement by steel fibres in pretensioned concrete beams, in: International FIB Symposium 2008-Tailor Made Concrete Structures: New Solutions for Our Society, CRC Press-Taylor & Francis, 2008. [3] J. Gustafsson, K. Noghabai, Steel Fibers as Shear Reinforcement in High Strength Concrete Beams, 22, Nordic Concrete Research-Publications, 1999, pp. 35–52. [4] J. Hegger, G. Bertram, Shear carrying capacity of ultra-high performance concrete beams, in: Walraven, Stoelhorst (Eds.), Institute of Structural Concrete ta RWTH Aachen, Germany-Tailor Made Concrete, Taylor & Francis Group, 2008. [5] M. Mansur, W. Alwis, Reinforced fibre concrete deep beams with web openings, Int. J. Cem. Compos. Lightweight Concrete 6 (4) (1984) 263–271. [6] Y.L. Voo, W.K. Poon, S.J. Foster, Shear strength of steel fiber-reinforced ultrahigh-performance concrete beams without stirrups, J. Struct. Eng. (2010). [7] J.C. Walraven, High performance fiber reinforced concrete: progress in knowledge and design codes, Mater. Struct. 42 (9) (2009) 1247–1260. [8] K.-K. Choi et al., Shear strength of slender reinforced concrete beams without web reinforcement: a model using fuzzy set theory, Eng. Struct. 31 (3) (2009) 768–777. [9] A.F.d.G. Civil, Interim Recommendations for Ultra High Performance FibreReinforced Concrete, 2002. [10] M.A. Mansur, K.H. Tan, S.L. Lee, Collapse loads of RC beams with large openings, ASCE J. Struct. Eng. (1984) 17. [11] I. Markovic´, High-Performance Hybrid-Fibre Concrete: Development and Utilisation, IOS Press, 2006. [12] T. Soetens, S. Matthys, Shear design of full-scale prestressed SFRC girders, in: 2nd Joint ACI-fib FRC Workshop, 2014. Montreal, Canada. [13] J. Hegger, G. Bertram, Shear carrying capacity of ultra-high performance concrete beams, in: Walraven, Stoelhorst (Eds.), Institute of Structural Concrete ta RWTH Aachen, Germany-Tailor Made Concrete, Taylor & Francis Group, 2008, pp. 341–347. [14] R. Zagon, S. Matthys, Z. Kiss, Shear tests on SFR-UHPC beams with or without web opening, in: Proceedings 7th Int. Conf. Fibre Concrete 2013 – Technology, Design, Application, 2013, 2013. Prague, Czech Republic. [15] RILEM, Technical Recommendations for the Testing and Use of Construction Materials, 1990. [16] C. Cucchiara, L. La Mendola, M. Papia, Effectiveness of stirrups and steel fibres as shear reinforcement, Cem. Concr. Compos. 26 (7) (2004) 777–786. [17] R. Narayanan, I.Y.S. Darwish, Use of steel fibers as shear reinforcement, ACI Struct. J. (1987) 12. [18] fib, Model Code 2010, Final draft, in: International Federation for Structural Concrete, 2012. Laussane, Switzerland.