Influence of multiple anchor arrangement in the behaviour of FRP-to-concrete anchored joints

Journal Pre-proofs Influence of multiple anchor arrangement in the behaviour of FRP-to-concrete anchored joints Ilsen Adriana Cortez Flores, Jaime Fernández Gómez, Paula Villanueva Llauradó PII: DOI: Reference:

S0263-8223(19)31804-5 https://doi.org/10.1016/j.compstruct.2019.111528 COST 111528

To appear in:

Composite Structures

Received Date: Revised Date: Accepted Date:

14 May 2019 26 September 2019 4 October 2019

Please cite this article as: Flores, I.A.C., Gómez, J.F., Llauradó, P.V., Influence of multiple anchor arrangement in the behaviour of FRP-to-concrete anchored joints, Composite Structures (2019), doi: https://doi.org/10.1016/ j.compstruct.2019.111528

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier Ltd.

Influence of multiple anchor arrangement in the behaviour of FRP-to-concrete anchored joints Cortez Flores, Ilsen Adriana a; Fernández Gómez, Jaime a; Villanueva Llauradó, Paula a a

School of Civil Engineering, Universidad Politécnica de Madrid (UPM). Prof Aranguren, s/n, 28040, Madrid, Spain

Abstract The effectiveness of fibre-reinforced polymers (FRP) external reinforcements is limited by premature delamination from concrete. In recent years, various anchoring systems were developed for FRP sheets. Among these, FRP anchors stand out for their effectiveness. When the unit strength of a connector is insufficient, multiple anchors may be employed. This paper presents an experimental campaign consisting of single shear tests on CFRP anchored narrow (100 mm width) and wide (200 mm width) sheets in concrete substrates. One anchor was installed in narrow sheets for comparison with wide sheets. Wide sheets were tested with one anchor and with two anchors transversally distributed to evaluate the influence of multiple anchors. Results from both literature and tests show that an optimal arrangement of one anchor in narrow FRP sheets or two transversally distributed anchors in wide sheets can be equivalent, which allows the anchored joints to develop more than twice the bond strength.

KEY WORDS Fibre-reinforced polymer FRP, concrete retrofitting, FRP anchors

1.

INTRODUCTION

The use of fibre-reinforced polymers (FRPs) to strengthen and repair existing reinforced concrete structures has increased in recent years, as these composites are broadly recognized as strengthening materials due to their high tensile strength, strength-to-weight ratio, corrosion resistance and ease of installation. One of the limitations of this technology is its premature delamination from the substrate, which disables the full use of the material properties. Different existing codes for designing FRP externally bonded reinforcements [1],[2],[3],[4] have considered this limitation and express it in terms of either the maximum strain or maximum adhesion strength. To prevent delamination of FRP laminates in FRP-strengthened concrete elements, several anchorage techniques have been developed. These include the use of transverse FRP wraps [5], U-shape FRP anchors ([5], [6]), and steel bolt anchors [7], among others. Anchors made from fibre ropes or rolled sheets (both

called after FRP or spike anchors by different authors) have demonstrated to considerably enhance the joint strength and even change the failure mode in different experimental campaigns ([8], [9], [10], [11], [12], [13], [14],[15],[16],[17],[18]). The design of the FRP anchors has to be adjusted to the needs of the reinforced element, so it is not possible to determine whether it is preferable to use one single anchor or multiple anchors. Multiple anchors will be required when the unit strength of the connector is insufficient to properly enhance the adherent strength of the reinforcement. Different studies have been reported on the use of multiple FRP anchors ([10],[11], [13],[14], [19]) in FRP reinforcements to delay debonding. The results from these authors showed that the influence of multiple anchors in terms of strength and ductility of the anchored joint mainly depend on distribution and alignment, transversal and longitudinal spacing. Three different arrangements can be found in the bibliography: longitudinal ([19],[20]), transversal ([10],[11],[14], [20]) and staggered ([7]). Results from the literature also reveal that specimens with a longitudinal FRP anchor arrangement develop a more ductile behaviour before failure, whereas the transversal FRP anchor arrangement is generally related to the development of higher strength in the FRP plate. Ozbakkaloglu et al. [20] observed that, in double-anchor specimens, FRP plates with longitudinal and transverse anchor configurations develop similar strengths; however, plates with a longitudinal anchor configuration develop higher maximum longitudinal strains compared to plates with a transverse anchor configuration. Niemitz et al.[11], based on their results, concluded that the force transferred from the FRP sheet to the FRP anchors is proportional to the anchor splay diameter and FRP sheet thickness. The width of the sheet is thence an important parameter to consider, as the anchor fan must cover it completely to achieve better performance of the joint. In relatively wide sheets, locating two anchors transversally is crucial, since it allows the anchors to work similarly to how they work in longitudinal distributions. The most accurate arrangement for wide sheets can be claimed to be that employed in Niemitz et al. [11] and in McGuirk and Breña [14], who located two anchors in the transversal direction of the reinforcement. An important scatter of the behaviour of multiple anchors can be found in the bibliography, which makes it difficult to determine the improvement when compared to a single anchor. Orton et al. [12] and Niemitz et al. [11] observed a nonlinear increase with the number of connectors and concluded that the strength increase is not just the addition of capacities of individual anchors. Orton et al. [12] also observed that the

presence of anchors improves the efficiency of material usage in CFRP retrofits; thus, reducing the amount of CFRP material that is necessary to achieve a given strength. Present research is based on some conclusions that were drawn from the literature and seeks to complement existing information on the influence of multiple anchor arrangement in the behaviour of FRP-to-concrete anchored joints. Narrow sheets (100 mm width) were installed with just one anchor for comparison with wide sheets (200 mm width). Wide sheets were tested with one anchor and with two anchors distributed transversally as in Niemitz [10] and McGuirk [13]), with the aim to evaluate the influence of multiple anchors on the behaviour of the anchored joint. 2.

EXPERIMENTAL PROCEDURE

Forty-two specimens were tested to compare at least two results for each configuration. Two different widths of the FRP sheet were studied—100 mm width (narrow) with a single anchor and 200 mm width sheets (wide) with both single and multiple anchors—to determine the influence of multiple anchors´ arrangement. Test matrix is collected in table 1. Table 1 Test matrix FRP width (mm)

FRP width/concrete width

Bonded length in front of the anchor (mm)

Number of specimens tested

100

0.370

-

3

200

0.741

-

3

310

3

260

3

210

3

310

3

260

3

210

3

310 260

3 3

210

3

310

3

260

3

210

3

260

3

210

3

Anchors´arrangement

Control series (C)

100

0.370

Anchor fan splayed onto the reinforcement (O) 200

100

0.741

0.370

Anchor fan splayed between two reinforcement plies (B) 200 Multiple anchors transversally distributed (M)

200

0.741

0.741

Figure 1 shows the top and side views of the specimen configuration. For all specimens, the embedded length of anchors was 100 mm, inserted into a 20 mm diameter perforation with a dowel angle of 135° and a bending radius of 35 mm. This configuration was found to be optimal for the fibre bundle employed in the experimental campaign, according to Villanueva et al. [15]. Different series were tested, in which the free section of the anchor (anchor fan) was splayed onto the reinforcement and between two reinforcement plies forming a 60° and 75° angles in the 100 mm and 200 mm width sheets, respectively. These fan angles were set to cover all the width of the FRP sheets, following the criteria set by Niemitz et al. [11], and not tested as an independent variable. P

50

50

P

LVDT1

60° 50 20

0

0

° 135

10

200

200

75°

50

20

10

LVDT1

° 135

LVDT2

100

410

410

270

LVDT2

200

270

Figure 1. Specimen configuration. Most authors in previous studies located the anchors close to the loaded end of the reinforcement. This follows the improvement observed when the anchors were included within the stress transfer zone of the reinforcement. In this study, three different spacings of the anchor from the unloaded end were studied. The anchors were installed on the unloaded end of the joint, at 50 mm and at 100 mm from it. Figure 2 shows all the specimens´ configurations tested. A preliminary research [21] was conducted by the authors to compare anchors with the anchor fans expanded between two reinforcement plies and onto the reinforcement, with one anchor installed in narrow sheets. As the best results for single anchors in narrow FRP sheets were obtained in specimens with the anchor fan expanded between two reinforcement plies, in this paper, the load-slip responses for the anchor fan between two plies of reinforcement together with ones of the unanchored (control) series are shown in the results section for a comparison with wide sheets. Only one series with the anchor fan expanded onto the reinforcement was tested to verify previous observation by the authors that this arrangement behaved worse than reinforcements with the anchor fan expanded between the two reinforcement plies, as it did in narrow FRP sheets. The rest of the series were

studied with the anchor fan expanded between two reinforcement plies and with the anchor located in the same positions with respect to the loaded end (210 mm, 260 mm and 310 mm). P

50

50

160

60°

100

LVDT2

60°

50

LVDT2

C-100

100

100

100

270

270

270

C-200

O-100-0

O-100-50

P

P

P

270

O-100-100

P

50

40

50

50

100

P

50

270

200

100

100

100

100

P

60°

LVDT1

260

260

260

LVDT1

P

210

LVDT1

P

50

50

50

P

75°

200

270

B-200-0

50

LVDT2

60°

160 100

210 LVDT2

60°

60°

60°

LVDT2

200

200

200

270

270

270

270

B-200-50

B-200-100

M-200-50

M-200-100

100

100

200

LVDT1

170 100

75°

50

260

210

75°

Figure 2. Anchor position with respect to the unloaded end The model developed by Chen & Teng [22] was employed for estimation of adherence strength as it considers the influence of FRP thickness and is broadly acknowledge due to its good fitting to experimental results.

2.1. Material properties The anchors were installed in concrete specimens with dimensions of 410×270×200 mm with a compressive strength of 26.90 MPa at 28 days, which were obtained according to standard UNE-EN 123903 [23]. Sika Wrap-230 C unidirectional carbon fibre fabrics with a nominal width of 300 mm were used as reinforcement. According to the manufacturer, the fibres have an elasticity modulus of 240 GPa, a tensile strength of 4800 MPa and a rupture strain of 2%. The product datasheet specifies plate thickness with one ply of 0.129 mm, with an elasticity modulus of 220 GPa and a tensile strength of 3200 MPa. The fabrics were cut to the required width of 100 mm and 200 mm for the narrow and wide series. Anchors were made

from carbon fibre ropes having a diameter of the impregnated connector of 10 mm, and according to the product data sheet, their fibre section is 27.3 mm2 with 1900 MPa tensile strength of the impregnated ropes. 2.2. Specimen preparation In terms of manufacture and installation, it is possible to distinguish between fresh installed anchors and pre-cured anchors [15]. The main difference between both systems is the moment in which the portion of the sheet that will become the embedded region is impregnated with respect to the moment of the anchor´s insertion. A precured installation technique was used for all anchors in this research, which controlled the impregnated length so that it did not exceed 2/3 of the embedded depth and, thus, would not affect the bending section, as was recommended by Villanueva et al. [15]. The concrete surface was prepared according to ACI 440.2R-08 [3] recommendations for bond critical applications to achieve an adequate roughness. The anchors were installed in holes with their edge treated with a drill bit. The cleaning of the hole was performed before and after the edge treatment using pressurized air, as recommended in the guidelines for adhesive anchors, such as EAD 330499-00-0601 part five [24]. 2.3. Instrumentation and loading The single shear test setup was chosen in this research, mainly because it is simple, repeatable and results in specimens that are relatively easily handled and assembled. For the tests, the concrete block was completely restrained by a steel holding frame. Two linear variable differential transformers (LVDTs) were installed as shown in Figure 3 to measure the relative displacement between the concrete and the composite in two different places, with one in front (LVDT1) and one behind the anchor (LVDT 2).

Figure 3. Test set up.

3.

EXPERIMENTAL RESULTS

The results that were obtained in the different experimental series are collected in the tables 2-4 and figures 4-13 below. In the series’ name, the first letter refers to the position of the anchor fan with respect to the reinforcement being C for control (unanchored specimens), O for specimens with the anchor fan expanded onto the reinforcement, and B for the specimens with the anchor fan expanded between two reinforcement plies. The first number that follows represents the reinforcement width, and the last number below is the position of the anchor with respect to the unloaded end. 3.1. Single anchors in narrow sheets In the following paragraphs, results of the present research are collected together with previous results by the authors [21] for narrow sheets with the anchor fan splayed between two reinforcement plies for comparison. In most of the specimens, it was not possible to register the slips measured by both LVDTs until complete failure of the joint because the progress of the debonding produced premature failure of some of them. Load-slip responses of the tested specimens were analysed according to the generic model by Zhang and Smith [17], with the load slip responses of all the specimens tested having three defined zones. The limits between the zones are established by the loads and slips that correspond to the beginning of plate debonding (Pdb;Sdb), the moment when the plate completely detaches from substrate (P máx1,Smáx1), and the maximum load and its corresponding slip achieved after complete detachment (P máx2,Smáx2). These zones were identified from the test data and collected in tables 2-4. The different zones are clearly visible in Figure 5, and changes from one zone to the other are less evident in the other series.

Figure 4. Load-slip responses 100 mm Control series.

Figure 6. Load-slip responses for B-100-50 series.

Figure 5. Load-slip responses for B-100-0 series.

Figure 7. Load-slip responses for B-100-100 series.

In the narrow reinforcement series, four different failure modes were observed. In the first mode, plate debonding (DB) was a cohesive failure that occurred at the FRP-to-concrete interface with a thin layer of concrete attached to the plate. The other three failure modes began with plate debonding, followed by either plate or anchor failure. The second failure mode observed was plate debonding followed by plate rupture (DPR). The third failure mode observed was plate debonding followed by anchor rupture at the bending zone (DAR), which was mainly due to stress concentration in that zone. Finally, the last failure mode was a considerably more ductile failure that involved plate debonding followed by anchor pull-out (DAP). Table 1 summarizes the loads and slips corresponding to the complete debonding of the FRP plate (Pmax1,Smax1) and the maximum loads and slips reached afterward (Pmax2,Smax2). The failure modes presented in the specimens that were tested and the strength increase with respect to the control series (Pmax/Pdb) for all the configurations are also included. In some series, it was possible to observe more than one failure mode in different specimens, as seen in Table 2.

Table 2 Single-anchored narrow FRP sheet results Pmax1 Smax1 kN Lvdt1 Lvdt2 C-100-1 21.52 1.15 0.92 C-100 C-100-2 18.68 2.30 0.50 C-100-3 22.61 2,05 0.37 B-100-0-1 29.41 0.83 B-100-0 B-100-0-2 37.63 0.64 1.52 B-100-0-3 26.38 1.29 B-100-50-2 41.73 1.52 B-100-50 B-100-50-3 43.22 2.95 0.80 B-100-50-4 39.49 1.78 0.61 B-100-100-1 45.05 0.94 0.48 B-100-100 B-100-100-2 48.14 0.20 0.41 B-100-100-3 40.18 1.03 0.38 SERIES

SPECIMEN

Pmax2 kN 39.01 40.94 44.4 54.76 54.52 56.50 53.64 54.98 52.91

Smax2 Lvdt1 Lvdt2 4.17 2.26 4.46 5.07 5.07 0.75 2.61 0.52 0.52 -

Pavg kN

Pmax/ Pdb

20.94

-

41.45

1.70

54.64

2.23

53.85

2.20

Failure Mode DB DB DB DAR DAP DAR DAP DAP DAR DAP DAP DAP

In the results shown in Table 2, it is also possible to observe that anchoring a narrow FRP sheet with a single spike anchor enhances the joint strength, thus reaching peak loads up to 223% of the bonded strength calculated with the Chen & Teng model [22]. Experimental results of the control series were higher than the values predicted using the model. The predicted values (18.11 and 28.28 kN for the narrow and wide FRP sheets respectively) were very similar to the lowest experimental result of the control series, both for 100 and 200 mm FRP width specimens. It was considered, that the model properly reproduces the bond behaviour, providing safe values that take into account possible installation deficiencies. The strength

increase in the series with one reinforcement ply was obtained using this lowest experimental result, while for all the other cases Chen & Teng´s model was applied. Locating the anchor at a distance bigger than the effective length from the loaded end produces an increase in the joint strength as long as the bonded length behind the connector is sufficient for the joint to develop its adhesion capacity. 3.2. Single anchors in wide sheets Load-slip responses of wide sheets (200 mm) with one single anchor, together with the control series for wide sheets, are collected in Figures 8, 9, 10 and 11, and the results are summarized in Table 3.

Figure 8. Load-slip responses 200 mm Control series.

Figure 10. Load-slip responses for B-200-50 series.

Figure 9. Load-slip responses for B-200-0 series.

Figure 11. Load-slip responses for B-200-100 series.

Table 3 Single-anchored wide FRP sheets results

SERIES C/200/-/

O/200/50/

B/200/0/

B/200/50/

B/200/100

SPECIMEN C/200/-/1 C/200/-/2 C/200/-/3 O/200/50/1 O/200/50/2 O/200/50/3 B/200/0/1 B/200/0/2 B/200/0/3 B/200/50/1 B/200/50/2 B/200/50/3 B/200/100/1 B/200/100/2 B/200/100/3

Pmax1 kN 32.97 48.56 35.03

Smax1 Lvdt1 Lvdt2 2.73 1.29 1.44 0.69 1.49 0.49 37.74 0.42 3.90 39.25 0.83 2.43 34.29 0.69 1.70 49.82 0.43 0.63 51.86 0.31 50.51 1.66 0.54 57.80 3.39 58.38 3.24 52.88 0.44 3.97 50.53 4.99 1.61 47.65 3.93 48.77 7.05 3.06

Pmax2 kN 43.24 48.32 44.89 63.82 57.89 62.37 67.73 66.72 53.79 62.16 64.25 58.09

Smax2 Lvdt1 Lvdt2 4.12 3.78 2.06 0.33 4.35 4.50 4.92 4.33 7.53 6.57 10.83 6.56

Pavg kN

Pmax/ Pdb

38.85

-

45.48

1.37

61.36

1.58

62.74

1.62

61.50

1.58

Failure Mode DB DB DB DAR DAR DAR DAR DAR DAR DAR DAR DAR DAP DAP DAR

From the results presented in Table 3, it is possible to see that in 200 mm (wide) FRP sheets anchored with one connector, the failure mode is directly related to failure of the anchor itself, being mainly due to anchor rupture at the bending zone (DAR), with the exception of the B-200-100 series. Another thing that can be observed is that wide FRP sheets present higher strengths compared to that of narrow FPR sheets. This fact does not necessary represent better behaviour, as the percentage increase in the maximum load achieved by the presence of the anchors compared to the unanchored specimens is lower compared to the results of single-anchored narrow FRP sheets. This leads us to affirm that there exists a maximum width of the reinforcement that a single connector can optimally anchor. 3.3. Multiple anchors in wide FRP sheets In the last series, two anchors that were distributed transversally along the width of the FRP sheet with the anchor fan expanded between two reinforcement plies were tested. The anchors were located at 50 mm and 100 mm from the unloaded end, based on the best results that were obtained in the previous series. The load-slip response of all the specimens from these series can be seen in Figures 12 and 13. It was evident that using two anchors distributed transversally in wide FRP sheets does not affect the slip of the plate when compared to the results that were obtained in single anchored wide sheets. The slips reached values that are similar to the ones of single anchored wide FRP sheets and presented the same response, with higher slips for anchors located closer to the loaded end.

Figure 12. Load-slip responses for M-200-50 series.

Fig. 13. Load-slip responses for M-200-100 series.

Table 4 summarizes the maximum loads with the corresponding slips, failure mode and strength increase that were observed in specimens with multiple anchors. Table 4 Multiple anchored wide FRP sheets results Pmax1 kN M/200/50/1 65.88 M/200/50/ M/200/50/2 57.31 M/200/50/3 60.37 M/200/100/1 80.89 M/200/100/ M/200/100/2 64.02 M/200/100/3 78.01 SERIES

SPECIMEN

Smax1 Lvdt1 Lvdt2 1.94 7.67 0.57 1.09 0.13 2.40 1.51 7.87 0.13 7.94 4.18

Pmax2 kN 97.19 86.40 98.28 102.27 95.23 105.36

Smax2 Pavg Lvdt1 Lvdt2 kN 3.72 9.88 93.96 7.23 1.75 8.40 100.95 11.45 -

Pmax/ Pdb 2.42

2.60

Failure Mode DAR DAR DAR DAR DAR DAR

All the specimens in this series failed by anchor rupture at the bending zone (DAR), with both anchors presenting almost simultaneous failure. The best strength corresponds to the M-200-100 series, which is contrary to the behaviour that was observed in all the single anchored series (narrow and wide FRP sheets), where the best results were obtained from specimens with the anchor located 50 mm from the unloaded end. In wide FRP sheets, the use of multiple anchors allowed the joint to resist peak loads that were at least 50% higher than the ones resisted by single anchored specimens. The peak loads reached by the specimens anchored with multiple anchors were between 242% and 260% of the ones that were reached by the control series. 4.

DISCUSSION 4.1. Campaign results

As seen in Figures 4-7 and Table 2 from the results section, the presence of FRP anchors with the fan expanded between two reinforcement plies increases the joint strength regardless of the anchor position with respect to the loaded end. It is also possible to observe that, besides increasing the strength of the joint, the presence of an anchor allows higher plate slips, which presents a more ductile failure. All wide FRP sheets anchored with a single connector presented a behaviour in terms of load-slip curves according to the generic model by Zhang and Smith [17]. However, in contrast with the narrow FRP sheets, in wide sheets, locating the anchor closer to the loaded end allowed for higher slips; this was directly related to the most frequent failure mode observed in the B-200-100 series, which was plate debonding followed by anchor pull-out (DAP). In the remaining series of wide FRP sheets anchored with a single connector, the only failure mode observed was debonding of the plate followed by anchor rupture in the bending zone (DAR), which is coherent with the smaller slips that were registered. However, as no measurements of strains along the FRP width were registered, no further conclusions can be drawn regarding the differences in slippage between narrow single anchored and wide double anchored FRP sheets. Figures 14 and 15 show the main differences that were observed in the experimental campaign presented herein between single and multiple anchors in both narrow and wide FRP sheets. Figure 14 shows the strength increase compared to the unanchored joint for single and multiple anchors, with the anchors located at 50 mm and 100 mm from the unloaded end. It shows a clear decrease in the single anchor’s contribution with an increase of the FRP width, which supported the previous affirmation that there exists a maximum width of reinforcement that a single connector can optimally anchor. This may be because the anchor contribution to the strength is reduced in the outer limits of their fans, as was observed by Niemitz [10],

since this behaviour is more evident in wide reinforcements. The use of multiple anchors (two for this case) in wide sheets reduces the influence area of the connector, thus improving its capacity until reaching the strength increase that is obtained with single anchors in narrow sheets. In Figure 14, it is possible to see evidence that the use of single anchors in narrow FRP sheets and multiple anchors in wide sheets allows the anchored joint to resist loads of more than twice the loads of the unanchored specimens. On the other hand, Figure 15 compares the joint strength of the control series (with the two widths considered) and that of the series with single and multiple anchors located at two different positions with respect to the unloaded end. It can be seen that the FRP sheet width and the number of anchors employed are directly related to the maximum resistance of the anchored joint. Wider FRP sheets anchored with multiple connectors resist higher loads, but at the same time, this configuration produces a reduction in the exploitation of the FRP sheet’s properties.

Figure 14. Resistance increase.

Figure 15. Strength comparison.

With respect to the load-slip responses, two main observations can be made. The first one is that using two anchors distributed transversally does not affect the slip of the plate when compared to the results that were obtained in single anchored wide sheets. The slips reached similar values, with the higher slips corresponding the series with the anchors located closer to the loaded end, which is contrary to the behaviour of single anchored narrow FRP sheets whereby the higher slips corresponded to the series with the anchors located at the farthest distance from the loaded end. The second observation is related to the maximum strength and the anchor position with respect to the loaded end. In all the single anchored series (narrow and wide FRP sheets), the highest strengths were reached when the anchor was located at a distance of 50 mm from the unloaded end, assuming that the drop in resistance of the series with the anchors located at 100 mm from the unloaded end was due to the small bonded length in front of the anchor that caused the anchor to be stressed before the development of the total adhesion strength of the FRP sheet. This assumption was confirmed with the results of multiple anchored FRP sheets whereby the highest strength corresponds to the series with the anchors located at 100 mm, as in this configuration, the bonded length in front the anchor fan was slightly lower than the effective

bonded length that was estimated with the Chen & Teng model [25], which allowed the FRP sheet to develop its adhesion strength properly before inducing stress in the connectors.

4.2. Literature results Different authors have studied the influence of multiple FRP anchors arrangement in the behaviour of FRPto-concrete anchored joints. Table 5 presents some of the main parameters that were considered in each campaign regarding the reinforcement: width (W FRP), bonded length (LFRP), thickness (tFRP), and the anchors: number, bonded length in front of the anchor (L anc), diameter (da), embedment depth (hemb) and fan length (lfan). The tensile strength of the reinforcement (P ult) was obtained for each campaign based on its width and thickness. This value, together with the maximum load resisted in each test (P test), allowed us to obtain the efficiency (Ptest/Pult) of materials´ properties usage for each configuration. Most of the authors tested an unanchored specimen to use its debonding strength (Pdb) as a base point from which to estimate the increase in resistance (Ptest/Pdb) due to the anchors’ presence. Table 5 Literature results of FRP reinforcements anchored with multiple connectors. FRP sheet Reference

Niemitz (2008)

Specimen

Anchor

WFRP LFRP tFRP Lanc da hemb lfan Nº (mm) (mm) (mm) (mm) (mm) (mm) (mm)

Pult (kN)

Ptest (kN)

Ptest/Pult

(Ptest/Pdb)*100 (%)

A-0-0-5-0

122

762

0.165

-

-

-

-

-

78.28

35.58

0.45

-

B-Y-2-5-4

122

762

0.165

2

127

13

51

51

78.28

55.24

0.71

155.3

B-X-2-5-4

122

381

0.165

2

127

13

51

51

78.28

60.54

0.77

170.1

C-Y-4-10-6

254

762

0.165

2

127

19

51

102

156.12

87.58

0.56

246.1

C-X-4-10-6

254

381

0.165

2

127

19

51

102

156.12

96.57

0.62

271.4

S-1-0a-24

125

610

-

-

-

-

-

109.60

43.37

0.40

-

F-1-0a-24

125

610

1.016 1.016

-

-

-

-

-

113.42

49.82

0.44

114.9

S-1-2a-24

125

610

1.016

2

125

13

51

32

109.60

77.84

0.71

179.5

F-1-2a-24

125

610

1.016

2

125

13

51

32

113.42

80.60

0.71

185.8

F-2-0a-24

125

610

-

-

-

-

226.85

69.83

0.31

-

125

610

2.032 2.032

2

F-2-2a-24

2

125

13

51

32

226.85 150.48

0.66

215.5

S-2-2a-24

125

610

2.032

2

125

13

51

32

219.20 116.98

0.53

167.5

BII-13-1.3-5

130

762

2

127

13

51

51

78.30

55.30

0.71

155.4

130

381

0.165 0.165

2

127

13

51

51

78.30

60.60

0.77

170.3

250

762

0.165

2

127

19

51

102

156.60

96.60

0.62

271.5

BIIS-25-1.9-10

250

381

0.165

2

127

19

51

102

156.60

87.60

0.56

246.2

A-1,2,3

100

270

-

-

-

-

307.28

38.60

0.13

-

100

270

0.668 0.668

-

B60R1-1,2,3

1

45

11

60

60

307.28

48.70

0.16

126.2

B60R2-1,2,3 Ozbakkaloglu B60R3-1,2,3 et al. (2016) D60R1R1-1,2

100

270

0.668

1

135

11

60

60

307.28

47.70

0.16

123.6

100

270

0.668

1

225

11

60

60

307.28

40.10

0.13

103.9

100

270

0.668

2

45

11

60

60

307.28

54.10

0.18

140.2

D60R2R2-1,2

100

270

0.668

2

135

11

60

60

307.28

56.00

0.18

145.1

D60R3R3-1,2

100

270

0.668

2

225

11

60

60

307.28

38.00

0.12

98.45

260

0.258

1

-

-

-

-

123.84

41.58

0.34

-

McGuirk (2011)

Niemitz et al. BIIS-13-1.3-5 (2012) BII-25-1.9-10

U

150

Villanueva (2017)

Current research.

O

150

260

0.258

1

210

10

100

123.84

54.18

0.44

130.3

I

150

260

0.258

1

210

10

100

123.84

58.65

0.47

141.1

R

150

260

0.258

1

210

10

100

123.84

51.65

0.42

124.2

T

150

260

0.258

2

210

10

100

123.84

51.55

0.42

124.0

C-100

100

260

0.129

-

-

-

-

-

82.56

24.45

0.30

-

B-100-50

100

260

0.258

1

210

10

100

100

82.56

54.64

0.66

223.5

B-100-100

100

260

0.258

1

160

10

100

100

82.56

53.85

0.65

220.0

C-200

200

260

0.129

-

-

-

-

-

165.12

38.85

0.23

-

O-200-50

200

260

0.129

1

210

10

100

167

165.12

45.48

0.55

137.9

B-200-50

200

260

0.258

1

210

10

100

167

165.12

62.74

0.38

161.5

B-200-100

200

260

0.258

1

160

10

100

167

165.12

61.50

0.37

158.3

M-200-50

200

260

0.258

2

210

10

100

167

165.12

93.96

0.57

241.8

M-200-100

200

260

0.258

2

160

10

100

167

165.12 100.95

0.61

259.8

It should be noted that the efficiency (P test/Pult) of materials´ properties usage does not necessarily lead to better usage of the anchors. In this sense, maximum strength increase of anchored joints with respect to adherence strength was found for two plies reinforcements (B series), although the increasing material need produces a reduction in the (Ptest/Pult) ratio. Influence of the FRP sheet width and multiple anchors As mentioned before, in the present research, FRP sheets with two different widths were studied (100 mm and 200 mm), locating the anchors in three different positions with respect to the loaded end. The results of the series with the anchor located at 50 mm from the unloaded end and with the anchor fan expanded between two reinforcement plies (B-100-50, B-200-50 and M-200-50) can be directly compared to the results of the “I” and “T” series of Villanueva´s study [18], since the geometric configuration and the materials (with the exception of concrete blocks’ properties, which have little relevance on these results) were the same. Figures 16 and 17 show the results of this comparison. In Figure 16, the relationship of the FRP width with the efficiency in the materials´ properties usage is presented. It can be observed that an increase in the width of the sheet produces a reduction in the efficiency when the reinforcement is anchored with a single connector. For the 150 mm width sheet, there are not considerable differences in the efficiency reached by single or multiple anchors. This can be due to the failure mode presented by the specimens of the “T” series (premature FRP plate rupture, probably due to asymmetry loading producing chained fibre rupture). However, for 200 mm width sheets, an important increase in efficiency can be observed when multiple anchors are used. It can also be seen that using two anchors in wide sheets produces an efficiency that is similar to the one that was obtained in single anchored narrow sheets.

Figure 16. Resistance increase.

Figure 17. Strength comparison.

Figure 17 shows the strength of the same series analysed in Figure 16. A clearly different trend can be observed, since the increase in the FRP width produces an increase in strength. As expected, the highest strengths correspond to the wider sheets anchored with multiple anchors. From these results, it is possible to affirm that there exists a width of FRP sheet from which the usage of multiple anchors is essential to

1.00

120

0.80

100

Ptest (kN)

Ptest/Putl

ensure a certain strength with proper efficiency.

0.60 0.40 0.20

80 60 40 20

0.00

0

0

50

100 150 200 250 300 350

0

50

FRP tensile strength (Pu) Niemitz (2008) Niemitz et al (2012) Villanueva (2017)

McGuirk (2011) Ozbakkaloglu et al.(2016) Cortez et al.

Figure 18. Efficiency vs. FRP tensile strength

100

150

200

250

300

FRP width (mm) Niemitz (2008) Niemitz et al.(2012) Villanueva (2017)

McGuirk (2011) Ozbakkaloglu et al.(2016) Cortez et al.

Figure 19. Anchored joint strength vs. FRP width

Figure 18 shows the relationship between the efficiency in materials properties usage and the tensile strength of the reinforcement for all the literature results presented in Table 4. This clearly shows a decrease in efficiency with an increase in the tensile strength of the reinforcement, which is related to the width of the FRP sheet, as well as to the number of plies employed. On the other hand, Figure 19 shows the increase in the anchored joint strength that was observed when increasing the FRP width. Based on these two figures, it is possible to affirm that there is a minimum and a maximum FRP width (between 100 mm and 200 mm, which limits the number of FRP plies to obtain a tensile strength of the

FRP sheet that is no greater than 225 kN) where the employment of anchors achieves its maximum efficiency with an increase in the joint strength and an adequate exploitation of the FRP sheet´s properties. 5.

CONCLUSIONS

Based on the results that were obtained in this campaign, the following conclusions can be drawn: -

FRP anchors enhance joint behaviour, with anchored joints reaching peak loads up to 223% of the ones of the control series for single anchors in narrow FRP sheets and 260% for multiple anchors in wide FRP sheets.

-

The average exploitation of the FRP sheet’s properties in unanchored FRP sheets is 30% for 100 mm width (narrow) sheets and 23% for 200 mm (wide) sheets. The increase in the efficiency of materials usage is 35% for the single anchored narrow FRP sheets, 14% for single anchored wide sheets and 35% for wide sheets anchored with two connectors distributed transversally in the width of the reinforcement.

-

For single anchored FRP sheets, there is a decrease in the contribution of anchors with the increase of the FRP width, which implies that there is a maximum FRP width that a single connector can optimally anchor. The effectiveness of multiple anchors depends on their optimal spacing, which defines the optimum sheet width that can be anchored with a single connector.

-

The width of FRP sheets and the number of anchors employed are directly related to the maximum strength of the anchored joint. The strength of the anchored joint can be increased with the width of the reinforcement and the number of anchors.

-

In terms of the efficiency of materials properties usage and strength increase, using one single connector in narrow FRP sheets can be equivalent to the use of multiple (two in this case) connectors in wide FRP sheets with an adequate configuration (anchor position with respect to the loaded end and anchor fan position with respect to the reinforcement).

-

Based on the literature results, and the results obtained in the present campaign, an optimum design of anchored joints can be realized by using FRP width sheets ranging between 100 mm and 200 mm to obtain strengths between 40 and 100 kN, with an efficiency in material usage of at least 50%, provided that no execution faults exist.

Acknowledgements

The authors wish to thank SIKA S.A.U. for Adriana Cortez fellowship for supplying materials and for assisting with the preparation of specimens. The authors also want to express their grateful acknowledgement to Betazul S.A. for providing supplementary materials for the tests and to the laboratory of construction materials of UPM for allowing the use of its facilities. Data availability statement The raw/processed data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study. References [1]

Concrete Society. TR55. Design guidance for strengthening concrete structures using fibre composite materials. 2012.

[2]

Fib. Bulletin 14: Externally bonded FRP reinforcement for RC structures. 2001.

[3]

ACI Committee 440. ACI 440.2R-17 Guide for the Design and Construction of Externally Bonded FRP Systems. 2017.

[4]

CNR-DT 200. Guide for the design and construction of externally bonded FRP systems for strengthening concrete structures - materials, RC and PC structures, masonry structures. 2004. doi:10.1061/40753(171)159.

[5]

Baggio DF. Effect of FRP Anchors on the FRP Rehabilitation of Shear Critical RC Beams and Flexure Critical RC Slabs. University of Waterloo, 2013.

[6]

Grelle S V., Sneed LH. Review of Anchorage Systems for Externally Bonded FRP Laminates. Int J Concr Struct Mater 2013;7:17–33. doi:10.1007/s40069-013-0029-0.

[7]

Anil Ö, Belgin AM. Anchorages effects on CFRP-to-concrete bond strength. J Reinf Plast Compos 2010;29:539–57. doi:10.1177/0731684408100259.

[8]

Ozdemir G. Mechanical Properties of Cfrp Anchorages. Middle East Technical University, 2005.

[9]

Eshwar N, Nanni A, Ibell TJ. Performance of two anchor systems of externally bonded fiberreinforced polymer laminates. ACI Mater J 2008;105:72–80.

[10]

Niemitz CW. Anchorage of carbon fiber reinforced polymers to reinforced concrete in shear applications. University of Massachusetts Amherst, 2008.

[11]

Niemitz CW, James R, Breña SF. Experimental Behavior of Carbon Fiber-Reinforced Polymer (CFRP) Sheets Attached to Concrete Surfaces Using CFRP Anchors. J Compos Constr 2010;14:185–94.

[12]

Orton SL, Jirsa JO, Bayrak O. Design Considerations of Carbon Fiber Anchors. J Compos Constr 2008;12:608–16. doi:10.1061/(ASCE)1090-0268(2008)12:6(608).

[13]

Mcguirk GN. Development of Anchorage System for Frp Strengthening. University of Massachusetts Amherst, 2011.

[14]

McGuirk GN, Breña SF. Development of anchorage system for FRP strengthening applications using integrated FRP composite anchors. 2012.

[15]

Llauradó PV, Fernández-Gómez J, González Ramos FJ. Influence of geometrical and installation parameters on performance of CFRP anchors. Compos Struct 2017;176:105–16. doi:10.1016/j.compstruct.2017.05.035.

[16]

Zhang H, Smith ST. Influence of plate length and anchor position on FRP-to-concrete joints anchored with FRP anchors. Compos Struct 2017;159:615–24. doi:10.1016/j.compstruct.2016.09.086.

[17]

Zhang HW, Smith ST, Kim SJ. Optimisation of carbon and glass FRP anchor design. Constr Build Mater 2012;32:1–12. doi:10.1016/j.conbuildmat.2010.11.100.

[18]

Villanueva Llauradó P. Influencia de las condiciones de ejecución en la resistencia de anclajes de fibra de carbono para refuerzos en estructuras de hormigón. Universidad Politécnica de Madrid, 2017.

[19]

Zhang HW, Smith ST. FRP-to-concrete joint assemblages anchored with multiple FRP anchors. Compos Struct 2012;94:403–14. doi:10.1016/j.compstruct.2011.07.025.

[20]

Ozbakkaloglu T, Fang C, Gholampour A. Influence of FRP anchor configuration on the behavior of FRP plates externally bonded on concrete members. Eng Struct 2017;133:133–50. doi:10.1016/j.engstruct.2016.12.005.

[21]

Cortez A, Fernández J, Villanueva P. Advances in FRP anchoring systems for externally bonded carbon fibre reinforcements. Int. Conf. Technol. Innov. Build., 2019, p. 106–8.

[22]

Yao J, Teng JG, Chen JF. Experimental study on FRP-to-concrete bonded joints. Compos Part B Eng 2005. doi:10.1016/j.compositesb.2004.06.001.

[23]

UNE-EN 12390-3:2009/AC:2011 Testing hardened concrete- Part 3: Compressive strength of test specimens 2011.

[24]

EAD 330499-00-0601 Metal anchors for Use in Concrete. Part 5: Bonded Anchors 2013.

[25]

Chen JF, Teng JG. Anchorage Strength Models for FRP and Steel Plates Bonded to Concrete. J

Struct Eng 2001. doi:10.1061/(ASCE)0733-9445(2001)127:7(784).