- Email: [email protected]
Strengthening with FRP bars of URM walls subject to out-of-plane loads
Construction and Building
MATERIALS
Construction and Building Materials 20 (2006) 101–110
www.elsevier.com/locate/conbuildmat
Strengthening with FRP bars of URM walls subject to out-of-plane loads Nestore Galati a
a,*
, Gustavo Tumialan b, Antonio Nanni
a
Department of Civil, Architectural and Environmental Engineering, 221 Engineering Research Lab, University of Missouri-Rolla, Rolla, MO 65409, USA b Simpson Gumpertz & Heger Inc., Waltham, MA 02453, USA Received 1 May 2005; accepted 30 June 2005 Available online 24 August 2005
Abstract This paper presents the results of an experimental program on the flexural behavior of un-reinforced masonry (URM) walls strengthened with near surface mounted (NSM) fiber reinforced polymers (FRP) bars. A total of 15 URM walls reinforced with NSM FRP bars were tested. The specimens were strengthened with different amounts of reinforcement to observe their improved performance and the mode of failure. The influence of the bar shape (i.e., circular vs. rectangular), FRP material, dimension of the groove and type of embedding material (i.e., epoxy or cementitious-based paste), were studied. Based on experimental evidence and on the data found in the literature, the paper provides criteria that can be used in the development of design guidelines. 2005 Elsevier Ltd. All rights reserved. Keywords: Fiber reinforced polymer (FRP); Near surface mounted (NSM) FRP bars; Bars shape; Epoxy- or cementitious-based paste; Out-of-Plane; Masonry; Design
1. Introduction Unreinforced masonry (URM) walls are prone to failure when subjected to overstress caused by out-ofplane and in-plane loads. Externally bonded FRP laminates have been successfully used to increase the flexural and/or the shear capacity of the strengthening of unreinforced masonry (URM) walls subjected to overstresses [1–3]. The use of near-surface-mounted (NSM) FRP bars is an attractive method for increasing flexural and shear strength of deficient RC members [4] and masonry walls and, in certain cases, can be more convenient than using FRP laminates (i.e., anchoring requirements, aesthetics requirements). A previous investigation has shown the effectiveness of FRP bars for increasing the flexural capacity of *
Corresponding author. Tel.: +1 573 341 6223; fax: +1 573 341 6215. E-mail address: [email protected] (N. Galati). 0950-0618/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2005.06.047
URM walls [5]. In that investigation, the FRP reinforcement was internally placed, this technique demanded the cutting of slots at the top course of the wall to place the bars, drilling of holes to pump grout and grouting. The successful use of near-surface-mounted (NSM) bars for improving the flexural capacity of RC members led to extending their potential use for the strengthening of URM walls. The use of NSM FRP bars is attractive since their application does not require any surface preparation work, except the cutting of the slots, and requires minimal installation time. A field application on flexural strengthening with NSM FRP bars of cracked URM walls in an educational facility in Kansas City – Missouri, showed the effectiveness and practicality of such a technique [6]. This paper presents an experimental program on 15 URM walls reinforced with FRP bars using the NSM technique where the wall is subjected to out-of-plane loads. The influence of five parameters was investigated: type and amount of FRP reinforcement, shape of the
102
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
Nomenclature Af bm c Ef Em ff ffu fm0 h/tm L
area of FRP reinforcement width of the masonry wall considered in the flexural analysis distance from extreme compression fiber to the neutral axis tensile modulus of elasticity of FRP modulus of elasticity of masonry stress level in the FRP reinforcement ultimate tensile strength of the FRP material as reported by the manufacturer compressive strength of masonry slenderness ratio (wall height-to-wall thickness) clear spacing based on length of masonry units
FRP bars (i.e., circular or rectangular), groove size and type of embedding material (i.e., epoxy- or cementitiouspaste). Also the effectiveness of the FRP reinforcement for masonry panels having a running or a stack pattern bond type was studied. In addition, based on experimental evidence and on experimental results available in the literature [7], this paper provides criteria that can be used in the development of design guidelines.
2. Experimental program 2.1. Test matrix As shown in Table 1, 15 masonry walls were manufactured for this experimental program. Three specimens were built with clay bricks. The remaining twelve were built with concrete blocks. The nominal dimen-
Mn sf tm b1
em e0m emu efu efe jm qf xf
nominal flexural capacity maximum clear spacing between FRP strips nominal thickness of masonry wall ratio of the depth of the equivalent rectangular stress block to the depth to the neutral axis compressive strain in masonry compressive strain in masonry associated to peak fm0 in a parabolic distribution ultimate compressive strain of masonry design rupture strain of FRP reinforcement effective strain in FRP reinforcement bond dependent coefficient ratio of FRP flexural reinforcement FRP reinforcement index
sions of these walls were 1.22 m (48 in.) by 0.61 m (24 in.); their overall thickness was 95 mm (3 34 in:) for clay specimens. In the case of the concrete masonry walls the thickness varied from 92 mm (3 58 in:) to 143 mm (5 34 in:) (see Fig. 1). The walls were constructed using a Type N mortar. All the joints were finished flush with the surface of the masonry unit. All specimens were allowed to cure for at least 28 days before testing. The specimens were strengthened, using the NSM technique, with 6.35 mm (14 in:) and 9.53 mm (38 in:) diameter ribbed GFRP bars and with 2 · 15 mm (0.080 · 0.60 in.) GFRP or with 2 · 16 mm (0.08 · 0.63 in.) CFRP rectangular bars. A two-part code was used to identify the specimens. The first part of the code identifies the parent material. Thus, the first two characters identify the type of masonry used ‘‘CO’’ for concrete masonry and ‘‘CL’’ for clay masonry. Since the specimens were constructed
Table 1 Test Matrix Specimen designation
Thickness of the specimen mm (in.)
Bar type
Groove dimension
Embedment material
CO1-GTE1 CO1-GTE2 CL1-GTE1 CL1-GTE2 CL2-CTE1
143 (5 34) 143 (5 34) 95 (3 34) 95 (3 34) 95 (3 34)
GFRP rectangular GFRP rectangular GFRP rectangular GFRP rectangular CFRP rectangular
Rectangular groove 17 · 3 (0.67 · 0.12)
Epoxy
CO2-GRE21 CO2-GRE22 CO2-GRE23
92 (3 58) 92 (3 58) 92 (3 58)
GFRP circular GFRP circular GFRP circular
Square groove 2.25 the diameter of the bar 14.3 mm (0.56 in.)
Epoxy
CO2-GRC31 CO2-GRC32 CO2-GRC33
92 (3 58) 92 (3 58) 92 (3 58)
GFRP circular GFRP circular GFRP circular
Square groove 2.25 the diameter of the bar 21.4 mm (0.84 in.)
Cementitious paste
CO2-GRE21-SJ CO2-GRE22-SJ CO2-GRE21-S CO2-GRE22-S
92 92 92 92
GFRP GFRP GFRP GFRP
Square groove 1.5 the diameter of the bar 9.5 mm (0.37 in.)
Epoxy
(3 58) (3 58) (3 58) (3 58)
circular circular circular circular
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
103
Fig. 1. Test specimens: (a) stack bond pattern; (b) running bond pattern.
using bricks or blocks coming from different stocks, the third number in the code is to identify the stock of the material. From Table 1 it can be observed that four different types of masonry were used: two for the clay walls (CL1 and CL2), and two for the concrete walls (CO1 and CO2). The second part of the code identifies the type and amount of FRP reinforcement. In particular, the first character is the type of FRP reinforcement used: ‘‘G’’ for GFRP and ‘‘C’’ for CFRP. The second character represents the cross-section of the bar used: ‘‘T’’ for rectangular bar and ‘‘R’’ for circular bar. The third character identifies the type of embedding material: ‘‘E’’ for epoxy paste and ‘‘C’’ for latex modified (cementitious grout). For the specimens using rectangular bars the last number represents the number of bars used for the strengthening. For the other specimens, the two numbers following the character ‘‘E’’ or ‘‘C’’ represent the diameter in eighths of an inch and the number of NSM Bars per specimen, respectively. The final character ‘‘S’’ or ‘‘SJ’’ indicates that the masonry panels were built with a stack pattern bond type with the reinforcement crossing the blocks
CO2-GRE21-SJ
CO2-GRE22-SJ
or placed in the vertical joints (see Fig. 2). Thus CO2-GRE21-SJ, refers to a concrete masonry panel, built with a stack pattern bond type, strengthened with one 6.35 mm (14 in:) diameter GFRP bar embedded in epoxy-based paste and placed in the vertical joints. 2.2. Material properties Tests were performed to characterize the mechanical properties of the materials used in this investigation. The average compressive strengths of the concrete and the clay masonry obtained from the testing of prisms (ASTM C1314) are presented in Table 2. A mortar Type N was used for the walls construction; standard mortar specimens were tested according to ASTM C109. An average value of 7.6 MPa (1100 psi) at an age of 28 days was found. Tensile tests were performed on FRP bars to determine their properties, which were related to fiber content. The average tensile strength, ultimate strain and modulus of elasticity obtained from the testing (ASTM D3039) are presented in Table 3. Details of the coupon
CO2-GRE21-S
CO2-GRE22-S
Fig. 2. Reinforcement scheme for specimens built with a stack bond pattern.
104
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
Table 2 Compressive strength of masonry walls Dimensions of masonry units mm (in.) 150 · 200 · 100 · 200 · 100 · 200 · 100 · 200 ·
00 00 00 400 (558 778 1534 ) 500 700 300 400 (38 78 154 ) 00 00 00 65 (334 212 778 ) 00 00 00 65 (334 212 778 )
Masonry type
Specimen code
Compressive strength MPa (ksi)
Concrete
CO1
11.4 (1.6)
Concrete
CO2
10.5 (1.5)
Clay
CL1
19.43 (2.8)
Clay
CL2
15.78 (2.3)
fabrication and testing procedure are shown elsewhere [8]. Splitting tensile tests (ASTM C496) were performed on the epoxy-based and on the cementitious-based embedding material used. The splitting tensile strength was found to be 3.58 MPa (0.518 ksi) after 7 days and 5.59 MPa (0.81 ksi) after 28 days in the case of latex modified cementitious paste, and 16.31 MPa (2.36 ksi) after 7 days and 18.54 MPa (2.7 ksi) after 28 days in the case of epoxy-based paste.
against a steel frame. Linear Variable Displacement Transducers (LVDTs) were positioned in the middle of the walls to measure the midspan deflection during the tests. The load was applied in cycles of loading and unloading. An initial cycle for a low load was performed in every wall to verify that both the mechanical and electronic equipment were working properly.
2.4. Test results 2.3. Test setup The masonry specimens were tested under fourpoints bending (see Fig. 3). Loads were applied by 50.8 · 609.6 · 12.7 mm (2 24 12 in:) steel plates to the external face of the wall. Their distance was 101.6 mm (4 in.) from the midspan. The loads were generated by means of a 12-ton hydraulic jack reacting
2.4.1. Modes of failure The walls exhibited three different modes of failure (see Fig. 4): (1) debonding of the FRP reinforcement from the masonry substrate; (2) flexural failure (i.e., crushing of the masonry in compression of rupture of the FRP in tension); and (3) shear failure of the masonry at the supports.
Table 3 Mechanical properties of FRP bars Bar type
Dimensions of the bar
Average maximum strain (%)
Average maximum stress MPa (ksi)
Average elastic modulus MPa (ksi)
#2 GFRP bar #3 GFRP bar GFRP rectangular bar CFRP rectangular bar
Nominal diameter 6.35 mm (0.250 in.) Nominal diameter 9.53 mm (0.375 in.) 2.06 mm · 15.21 mm (0.08100 · 0.59900 ) 2.06 mm · 15.21 mm (0.08100 · 0.59900 )
1.78 1.85 2.5 0.98
824.5 760.0 1101.7 1392.4
50 163 (7276) 40 800 (5920) 44 000 (6382) 142 740 (20 702)
Fig. 3. Test setup scheme.
(119.6) (110.0) (159.8) (201.9)
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
105
Fig. 4. Modes of failure: (a) FRP bar debonding (CO2-GRC32); (b) FRP debonding: formation of the 45 cracks at the head joints (CO2-GRC31); (c) splitting cracks in the embedding material (CO2-GRE22); (d) flexural-shear failure (CO2-GRC33).
FRP debonding. This was the most frequent mode of failure. Initial flexural cracks were primarily located at the mortar joints (see Fig. 4(a)). A cracking noise during the test revealed a progressive cracking of the embedding paste. Since the tensile stresses at the mortar joints were being taken by the FRP reinforcement, a redistribution of stresses occurred. As a consequence, cracks developed in the masonry units oriented at 45 (see Fig. 4(b)) or in the head mortar joints. Some of these cracks followed the embedding paste and the masonry interface causing debonding and subsequent wall failure. Due to the smoothness of the rectangular bars, some of the specimens reinforced with rectangular bars debond is a result of the bar sliding inside the epoxy. For specimens having a deep groove, debonding was caused by splitting of the embedding material (see Fig. 4(c)). Flexural failure. After developing flexural cracks primarily located at the mortar joints, a wall failed by either rupture of the FRP reinforcement or by the masonry crushing. FRP rupture occurred at midspan and was observed for the specimens CO2-GRE21 and CO2-GRE22. Shear failure. Cracking started with the development of fine vertical cracks at the maximum bending region. Thereafter, flexural-shear failure was observed at an orientation angle at approximately 45. In the flexural-
shear mode, shear forces transmitted over the crack caused a differential displacement in the shear plane, which resulted in FRP debonding (see Fig. 4(d)). 2.4.2. Discussion of results Fig. 5 shows the Moment vs. Deflection Curves for the six series. It can be observed that the strength and stiffness of the FRP strengthened walls increased dramatically when comparing them to a URM specimen. The nominal moments at cracking for the unstrengthened specimens were calculated considering the Masonry Standards Joint Committee recommendations (MSJC-02). From the graphs in Fig. 5 can be stated increments ranging 4–14 times the theoretic flexural capacity of the un-strengthened masonry wall. Since masonry possesses a significant amount of variability attributed to labor and materials, this range of values should be taken simply as a reference. Table 4 reports the test results. The experimental results have been expressed as a function of the amount of reinforcement, qf, defined as AreaFRP =ðWall width Wall thicknessÞ. For some of the specimens utilizing carbon or glass FRP rectangular bars, a higher ductility was observed when compared with the specimens reinforced with circular
106
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110 9
9 Theoretic Unreinforced
8
7
CO1-GTE1
7
CO1-GTE2
6
Midspan Moment (kN-m)
Midspan Moment (kN-m)
Theoretic Unreinforced
8
5 4 3 2
5 4 3
1
0 0
5
10
15 20 25 30 Midspan Deflection (mm)
35
40
45
0 0
b
9
5
10
15 20 25 Midspan Deflection (mm)
30
35
40
45
9
Theoretic Unreinforced CL2-CTE1
8
Theoretic Unstrengthened
8
7
Midspan Moment (kN-m)
Midspan Moment (kN-m)
CL1-GTE1
6
2
1
a
CL1-GTE2
6 5 4 3 2
CO2-GRE21
7
CO2-GRE22
6
CO2-GRE23
5 4 3 2 1
1
0 0 0
c
5
10
15 20 25 Midspan Deflection (mm)
30
35
40
45
0
10
15
20
25
30
35
40
45
Midspan Deflection (mm)
9
9
Theoretic Unstrengthened
Theoretic Unstrengthened
8
8
CO2-GRE21-S
7
CO2-GRE22-S
CO2-GRC31
7
CO2-GRC32
6
CO2-GRC33
Midspan Moment (kN-m)
Midspan Moment (kN-m)
5
d
5 4 3 2 1
CO2-GRE21-SJ
6
CO2-GRE22-SJ
5 4 3 2 1
0
0 0
e
5
10
15 20 25 30 Midspan Deflection (mm)
35
40
45
f
0
5
10
15 20 25 30 Midspan Deflection (mm)
35
40
45
Fig. 5. Moment versus midspan deflection for all the specimens: (a) Series CO1-GTE; (b) Series CL1-GTE; (c) Series CL2-CTE; (d) Series CO2GRE2; (e) Series CO2-GRC3; (f) Series CO2-GRE2x-S.
Table 4 Test results Specimen name
Amount of reinforcement qf (·105)
Ultimate load kN (kip)
Maximum bending moment kN m (k ft)
Type of failure
CO1-GTE1 CO1-GTE2 CL1-GTE1 CL1-GTE2 CL2-CTE1 CO2-GRE21 CO2-GRE22 CO2-GRE23 CO2-GRC31 CO2-GRC32 CO2-GRC33 CO2-GRE21-SJ CO2-GRE22-SJ CO2-GRE21-S CO2-GRE22-S
33 67 50 100 50 54 107 161 136 272 408 50 101 50 101
14.7 (3.3) 35.1 (7.9) 9.4 (2.1) 16.0 (3.6) 25.6 (5.8) 11.0 (2.5) 11.5 (2.6) 21.6 (4.9) 13.1 (2.9) 15.0 (3.4) 26.6 (6.0) 2.84 (0.64) 16.2 (3.64) 8.4 (1.9) 15.8 (3.5)
3.35 8.03 2.16 3.66 5.86 2.52 2.64 4.94 2.99 3.43 6.07 0.81 3.43 1.91 3.60
Debondinga Debondingb Debondingb Debondinga Debondinga Debondingb Debondingb Debonding Debonding Debonding Shear Debonding Debonding FRP rupture FRP rupture
a b
Debonding due to sliding of the bar inside the epoxy paste. Debonding due to splitting of the embedding material.
(2.47) (5.92) (1.59) (2.70) (4.32) (1.86) (1.95) (3.64) (2.20) (2.53) (4.48) (0.60) (2.53) (1.41) (2.66)
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
107
bars. In fact, for these specimens the failure was due to the sliding of the bars inside the groove. In these cases, after the failure, the wall could still carry load (because of the friction between the rectangular bar and the epoxy paste). An interesting observation can be underlined for specimens built with a stack pattern bond type. There is not a considerable reduction in the out-of-plane performance by placing the bar in the vertical joints or when it crosses the masonry blocks. The inadequate performance of wall CO2-GRE21-SJ is attributed to construction problems. In fact, this wall had a big mortar joint and therefore the epoxy bordered on the mortar and not on the concrete blocks. Therefore it is advisable, as a construction detail, to make grooves bordering the concrete block surfaces when bars are placed along the joints.
theoretical flexural capacity of an FRP strengthened masonry wall was determined based on strain compatibility, internal force equilibrium and the controlling mode of failure. Thus, the theoretical flexural capacities were estimated based on the assumption that no premature failure was to be observed (i.e., either rupture of the laminate or crushing of the masonry would govern the wall behavior). For simplicity and complementary to the flexural analysis of RC members, a parabolic distribution was used for compressive stresses in the computation of the flexural capacity of the strengthened walls. The stress block parameters associated with such a parabolic distribution are given as: 8 2 > < cb1 ¼ eem0 13 eem0 ; m m ð1Þ 2 > : cb 1 1 b ¼ 2 em0 1 em0 . 1 2 1 3 e 4 e
2.4.3. Previous results Turco [7] conducted an experimental investigation on URM walls reinforced with FRP NSM bars. The materials used, the test setup and the dimensions of the masonry panels were the same as the ones described in this paper. In particular a ‘‘Type CO2’’ masonry specimen was used. The differences between the two sets of investigations were in the reinforcement details. Table 5 summarizes the aforementioned results in terms of ultimate bending moment and mode of the failure. It was observed that, in the case of strengthening with NSM FRP bar, the latex modified cementitious pasteGFRP bar system exhibited a better performance when the size of the groove was approximately 2.25 times the diameter of the GFRP reinforcement. On the contrary, a groove of 1.5 times the diameter of the bar was enough when epoxy paste is used.
According to MSJC, the maximum usable strain emu was considered to be 0.0035 mm/mm (in./in.) for clay masonry and 0.0025 mm/mm (in./in.) for concrete masonry [9]. The tensile strength of the masonry was neglected. The theoretical shear capacity was estimated according MSJC [9] where a shear strength of 13 fm0 is recommended for URM in a running bond that is not grouted solid. The net cross-section was used for the computation of the shear capacity. The reinforcement index, xf, expressed as qf Ef =fm0 ðh=tm Þ, is an index that intends to capture the key parameters that influence the flexural capacity. These include the FRP flexural reinforcement ratio, qf, the FRP tensile modulus of elasticity, Ef, the masonry compressive strength, fm0 , and the slenderness ratio h/tm. This index is intended to represent the ratio of axial stiffness (cross-sectional area · modulus of elasticity) between the FRP and the masonry (AfEf/bmtmEm) but, since the modulus of elasticity of masonry Em is directly proportional to fm0 , the latter can replace Em. The inclusion of the slenderness ratio h/tm has been identified as influential in the out-of-plane behavior of masonry walls. h/tm accounts for the ability of the masonry wall behavior to be controlled by flexural capacity rather
2.5. Basis for design approach Table 6 presents the experimental and theoretical results used as a database for the development of a design approach for the FRP strengthening of URM walls. The
m
m
Table 5 Previous test results [7] Bar type number of bars
Amount of reinforcement qf (·105)
Groove dimension
Embedment material
Maximum bending moment kN m (k ft)
Mode of failure
2#2 3#2 1#3 2#3 3#3 2#2 3#2 1#3 2#3
110 170 120 240 360 100 170 120 240
Square groove 1.5 the diameter of the bar 9.5 mm (0.38 in.) Square groove 1.5 the diameter of the bar 14.3 mm (0.56 in.)
Epoxy Epoxy Epoxy Epoxy Epoxy Cementitious Cementitious Cementitious Cementitious
1.68 2.27 1.56 3.93 5.57 2.07 2.93 0.94 1.64
Debonding Debonding Debonding Shear Shear Debonding Debonding Debonding Debonding
a
GFRP GFRP GFRP GFRP GFRP GFRP GFRP GFRP GFRP
bar bar bar bar bar bar bar bar bar
Value too low for practical applications.
Square groove 2.25 the diameter of the bar 14.3 mm (0.56 in.) Square groove 1.5 the diameter of the bar 14.3 mm (0.56 in.)
paste paste paste paste
(1.24) (1.67) (1.15) (2.90) (4.11) (1.53) (2.16) (0.69)a (1.21)a
108
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
Table 6 Comparison between experimental and theoretical results Source
qf (·105)
Flexure kN m (k ft) Mexp.
Mth.
CO1-GTE1 CO1-GTE2 CL1-GTE1 CL1-GTE2 CL2-CTE1 CO2-GRE21 CO2-GRE22 CO2-GRE23 CO2-GRC31 CO2-GRC32 CO2-GRC33 CO2-GRE21-SJ CO2-GRE22-SJ CO2-GRE21-S CO2-GRE22-S Turco [7] Turco [7] Turco [7] Turco [7] Turco [7] Turco [7] Turco [7] Turco [7] Turco [7]
33 67 50 100 50 54 107 161 136 272 408 50 101 50 101 101 161 136 272 408 101 161 136 272
3.35 8.03 2.16 3.66 5.86 2.52 2.64 4.94 2.99 3.43 6.07 0.81 3.43 1.91 3.60 1.68 2.27 1.56 3.93 5.57 2.07 2.93 0.94 1.64
4.59 9.02 2.87 5.63 3.00 2.10 3.83 4.57 3.87 5.20 6.07 2.20 3.77 2.17 4.05 3.83 4.57 3.87 5.20 6.07 3.83 4.57 3.87 5.20
a
(2.47) (5.92) (1.59) (2.70) (4.32) (1.86) (1.95) (3.64) (2.20) (2.53) (4.48) (0.60) (2.53) (1.41) (2.66) (1.24) (1.67) (1.15) (2.90) (4.11) (1.53) (2.16) (0.69)a (1.21)a
Shear kN (kip) Vexp. (3.38) (6.65) (2.12) (4.15) (2.21) (1.55) (2.83) (3.37) (2.86) (3.83) (4.48) (1.62) (2.78) (1.60) (2.99) (2.83) (3.37) (2.86) (3.83) (4.48) (2.83) (3.37) (2.86) (3.83)
7.32 17.56 4.72 8.01 5.52 5.52 5.78 10.79 6.52 7.50 13.29 1.78 7.50 4.18 7.89 3.68 4.95 3.41 8.60 12.19 4.54 6.41 2.05 3.59
Vth. (1.65) (3.95) (1.06) (1.80) (1.24) (1.24) (1.30) (2.43) (1.47) (1.69) (2.99) (0.40) (1.69) (0.94) (1.77) (0.83) (1.11) (0.77) (1.93) (2.74) (1.02) (1.44) (0.46) (0.81)
24.02 24.02 21.18 21.18 19.20 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99
(5.4) (5.4) (4.76) (4.76) (4.32) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37)
Value too low for practical applications.
than shear capacity. h/tm and the required out-of-plane load to cause failure are inversely proportional; thus, as the slenderness ratio decreases, the out-of-plane load becomes larger. Fig. 6 shows the relationship between the experimental–theoretical flexural capacity ratio (Mexperimental/ Mtheoretical), and xf, for all of the specimens included in Table 6. The ratio Mexperimental/Mtheoretical represents the effectiveness of the FRP reinforcement. The ratio Mexperimental/Mtheoretical for the specimens failing in shear was computed based on the bending moment associated with the shear capacity. For design purposes, rather than attempting to predict bond failure, the strain in the FRP reinforcement can be limited. In this context, since the flexural capacity is dependant on the strain developed in the FRP bar, it is reasonable to express the effective strain in the bar, efe, as the product jmefu, where jm is the bond dependent coefficient and efu is the design rupture strain of FRP. Thus, the effective strain in the FRP bar, efe, is limited by the strain controlled by debonding: efe 6 jm efu FRP rectangular bars: jm ¼ 0.65; FRP circular bars having a groove 1:5 the diameter of the bar: jm ¼ 0.35; FRP circular bars having a groove 2:25 the diameter of the bar: jm ¼ 0.55.
These limits are valid for the case of walls not subjected to sustained load. In walls under sustained load such as retaining or basement walls, creep rupture considerations need to be taken into account. Thus, for the case of GFRP, jm could be 0.2 (ACI 440.2R-02 [10]). The moment capacity of the strengthened FRP wall is obtained from the equilibrium of the internal forces in the cross-section; thus: bc M n ¼ Af ff tm 1 ; ð2aÞ 2 ðcfm0 Þðb1 cÞbm ¼ Af ff .
ð2bÞ
Since the product jmefu represents the effective strain in the laminate, the stress in the FRP, ff, can be written as ff ¼ efe Ef ¼ ðjm efu ÞEf .
ð2cÞ
Replacing Eqs. (2b) and (2c) in (2a), and multiplying both terms by the factors bm tm fm0 and h/tm, one obtains: Mn qf Ef b1 c ¼ ðj e Þ 1 ð2dÞ m fu 2 bm t2m fm0 ðh=tm Þ fm0 ðh=tm Þ Finally, making xf explicit on the right end side of Eq. (2d) one obtains Mn xf ðjm efu Þ ðh=tm Þ ¼ x ðj e Þ 1 . f m fu 2 c bm t2m fm0 ðh=tm Þ ð2eÞ The moment capacity is then the minimum of either that obtained from Eq. (2e) or the value calculated for the
1 0.8
κmm= 0.65 0.65
0.6 0.4 0.2
Debonding
0 0
0.5
1
1.5
1
κ m= 0.65 0.65
0.8
m
0.6 0.4 0.2
Debonding 0.5
1
0.6
0.35 κm m= 0.35
0.4 0.2 0 0.5
1
1.5
2
ρf E f f m' (h / t) 1.2 1 0.8 0.6
κ mm= 0.55 0.55
0.4 0.2
Debonding Shear
0 0
2
1.5
ρf E f f m' (h / t)
c
0.8
0
0 0
Debonding Flexural Shear
1
b
1.2
109
1.2
2
ρf E f f m' (h / t)
a
M experimental / M theoretical
M experimental / M theoretical
1.2
M experimental / M theoretical
M experimental / M theoretical
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
d
0.5
1
1.5
2
ρf E f f m' (h / t)
Fig. 6. Mexperimental/Mtheoretical versus xf: (a) GFRP and CFRP rectangular bars – epoxy system; (b) GFRP circular bars – epoxy system (dimension of the groove 1.5 the diameter of the bar); (c) GFRP circular bars – epoxy system (dimension of the groove 2.25 the diameter of the bar); (d) GFRP circular bars – latex modified cementitious paste (dimension of the groove 2.25 the diameter of the bar).
theoretical shear capacity. The validation of such a protocol is demonstrated in Fig. 7 which shows the ratio between the experimental and nominal moment capacities as a function of xf. Since such a determined nominal capacity does not contain any safety factor, therefore Fig. 7 demonstrates the safety of the approach. The factor xf is related to the failure mode: at low values of xf correspond debonding of the FRP strengthening while at high values of it, shear failure occurs. From this point of view, it is reasonable to conclude that any safety factor, adopted from a particular design code, can be taken as a function of xf. 1.80
Mexperimental / Mn
1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0
0.5
1
1.5
ωf
2
Fig. 7. Ratio between experimental and nominal moments.
2.5
3. Conclusion The following conclusions can be drawn from this experimental program: 1. Flexural strengthening with FRP systems has been proven to remarkably increase the flexural capacity (from 2 to 14 times), strength and pseudo-ductility of URM walls. 2. The test results identified three basic modes of failure. One, shear failure, related to the parent material (i.e., masonry); and the other two associated with the reinforcing material, debonding and flexural failure (i.e., rupture of FRP or crushing of the masonry). For large amounts of reinforcement, shear failure was observed to be the controlling mode. For other reinforcement ratios, either FRP rupture or debonding was observed, the latter being the most common. 3. In the case of strengthening with FRP rectangular bars, the sliding of the bar in the epoxy caused an increase in the ductility. 4. Based on experimental data obtained in the present and other investigations, it is recommended to consider the maximum usable strain in the FRP Bars as 0.65efu for NSM FRP Rectangular Bars utilizing Epoxy Paste and 0.55efu for NSM FRP Circular Bars having a groove at least 2.25 times the diameter of the bar and utilizing either Epoxy or Cementitious Paste.
110
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
The maximum usable strain in the FRP Bars must be limited to 0.35efu for NSM FRP Circular Bars having a groove at least 1.5 times the FRP bar diameter and using Epoxy Paste. 5. The proposed design algorithm seems to be suitable to determine the moment capacity of URM walls strengthened with FRP bars. However, the validation of the protocol is based on the findings on two limited experimental campaigns; more tests are needed in order to determine more statistically relevant results. 6. The findings reported in this paper apply only to the materials, configurations and geometries tested and much more work needs to be done before general design recommendations can be provided.
Acknowledgments The support of the National Science Foundation Industry/University Cooperative Research Center at the University of Missouri-Rolla. The authors also acknowledge the support of the Rolla Technical Institute (RTI).
References [1] Schwegler G, Kelterborn P. Earthquake resistance of masonry structures strengthened with fiber composites. In: Proceedings of the 11th world conference on earthquake engineering, Acapulco, Mexico, CD-ROM; 1996. 6 pp.
[2] Hamilton HR, Dolan CW. Flexural capacity of glass FRP strengthened concrete masonry walls. J Comp Const ASCE 2001;5(3):170–8. [3] Tumialan JG, Morbin A, Nanni A, Modena C. Shear strengthening of masonry walls with FRP composites. COMPOSITES 2001 Conv. and Trade Show, Composites Fabricators Assoc., Tampa, FL, October 3–6, CD-ROM; 2001. 6 pp. [4] De Lorenzis L, Nanni A, La Tegola A. Flexural and shear strengthening of reinforced concrete structures with near surface mounted FRP rods. In: Proceedings of third international conference on advanced composite materials in bridges and structures, Ottawa, Canada; 2002. p. 521–8. [5] Hamid AA. Strengthening of hollow block masonry basement walls with plastic reinforcing bars. Masonry Soc J 1996. [6] Tumialan JG, Galati N, Nanni A, Tyler D. Flexural strengthening of masonry walls in q high school using FRP bars. In: ACI, Special Publication 215-12, Boston, September 26–October 1; 2003. p. 413–28. [7] Turco V, Galati N, Tumialan JG, Nanni A. Flexural strengthening of URM walls with FRP systems. In: Proceedings of the 6th international symposium on fibre-reinforced polymer (FRP) reinforcement for concrete structures (FRPRCS-6), Singapore; 2003. [8] Secondin S. Masonry reinforced with FRP systems. Tesi di Laurea, Universita` degli Studi di Padova, Facolta` di Ingegneria, Padua, Italy; March 2003. [9] Masonry Standards Joint Committee. Building code requirements for masonry structures. ACI-530/ASCE 5/TMS 402. American Concrete Institute, American Society of Civil Engineers, and The Masonry Society, Detroit, New York, and Boulder; 2002. [10] American Concrete Institute (ACI). Committee 440. Guide for the design and construction of externally bonded FRP systems for strengthening concrete structures; October 2002.
MATERIALS
Construction and Building Materials 20 (2006) 101–110
www.elsevier.com/locate/conbuildmat
Strengthening with FRP bars of URM walls subject to out-of-plane loads Nestore Galati a
a,*
, Gustavo Tumialan b, Antonio Nanni
a
Department of Civil, Architectural and Environmental Engineering, 221 Engineering Research Lab, University of Missouri-Rolla, Rolla, MO 65409, USA b Simpson Gumpertz & Heger Inc., Waltham, MA 02453, USA Received 1 May 2005; accepted 30 June 2005 Available online 24 August 2005
Abstract This paper presents the results of an experimental program on the flexural behavior of un-reinforced masonry (URM) walls strengthened with near surface mounted (NSM) fiber reinforced polymers (FRP) bars. A total of 15 URM walls reinforced with NSM FRP bars were tested. The specimens were strengthened with different amounts of reinforcement to observe their improved performance and the mode of failure. The influence of the bar shape (i.e., circular vs. rectangular), FRP material, dimension of the groove and type of embedding material (i.e., epoxy or cementitious-based paste), were studied. Based on experimental evidence and on the data found in the literature, the paper provides criteria that can be used in the development of design guidelines. 2005 Elsevier Ltd. All rights reserved. Keywords: Fiber reinforced polymer (FRP); Near surface mounted (NSM) FRP bars; Bars shape; Epoxy- or cementitious-based paste; Out-of-Plane; Masonry; Design
1. Introduction Unreinforced masonry (URM) walls are prone to failure when subjected to overstress caused by out-ofplane and in-plane loads. Externally bonded FRP laminates have been successfully used to increase the flexural and/or the shear capacity of the strengthening of unreinforced masonry (URM) walls subjected to overstresses [1–3]. The use of near-surface-mounted (NSM) FRP bars is an attractive method for increasing flexural and shear strength of deficient RC members [4] and masonry walls and, in certain cases, can be more convenient than using FRP laminates (i.e., anchoring requirements, aesthetics requirements). A previous investigation has shown the effectiveness of FRP bars for increasing the flexural capacity of *
Corresponding author. Tel.: +1 573 341 6223; fax: +1 573 341 6215. E-mail address: [email protected] (N. Galati). 0950-0618/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2005.06.047
URM walls [5]. In that investigation, the FRP reinforcement was internally placed, this technique demanded the cutting of slots at the top course of the wall to place the bars, drilling of holes to pump grout and grouting. The successful use of near-surface-mounted (NSM) bars for improving the flexural capacity of RC members led to extending their potential use for the strengthening of URM walls. The use of NSM FRP bars is attractive since their application does not require any surface preparation work, except the cutting of the slots, and requires minimal installation time. A field application on flexural strengthening with NSM FRP bars of cracked URM walls in an educational facility in Kansas City – Missouri, showed the effectiveness and practicality of such a technique [6]. This paper presents an experimental program on 15 URM walls reinforced with FRP bars using the NSM technique where the wall is subjected to out-of-plane loads. The influence of five parameters was investigated: type and amount of FRP reinforcement, shape of the
102
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
Nomenclature Af bm c Ef Em ff ffu fm0 h/tm L
area of FRP reinforcement width of the masonry wall considered in the flexural analysis distance from extreme compression fiber to the neutral axis tensile modulus of elasticity of FRP modulus of elasticity of masonry stress level in the FRP reinforcement ultimate tensile strength of the FRP material as reported by the manufacturer compressive strength of masonry slenderness ratio (wall height-to-wall thickness) clear spacing based on length of masonry units
FRP bars (i.e., circular or rectangular), groove size and type of embedding material (i.e., epoxy- or cementitiouspaste). Also the effectiveness of the FRP reinforcement for masonry panels having a running or a stack pattern bond type was studied. In addition, based on experimental evidence and on experimental results available in the literature [7], this paper provides criteria that can be used in the development of design guidelines.
2. Experimental program 2.1. Test matrix As shown in Table 1, 15 masonry walls were manufactured for this experimental program. Three specimens were built with clay bricks. The remaining twelve were built with concrete blocks. The nominal dimen-
Mn sf tm b1
em e0m emu efu efe jm qf xf
nominal flexural capacity maximum clear spacing between FRP strips nominal thickness of masonry wall ratio of the depth of the equivalent rectangular stress block to the depth to the neutral axis compressive strain in masonry compressive strain in masonry associated to peak fm0 in a parabolic distribution ultimate compressive strain of masonry design rupture strain of FRP reinforcement effective strain in FRP reinforcement bond dependent coefficient ratio of FRP flexural reinforcement FRP reinforcement index
sions of these walls were 1.22 m (48 in.) by 0.61 m (24 in.); their overall thickness was 95 mm (3 34 in:) for clay specimens. In the case of the concrete masonry walls the thickness varied from 92 mm (3 58 in:) to 143 mm (5 34 in:) (see Fig. 1). The walls were constructed using a Type N mortar. All the joints were finished flush with the surface of the masonry unit. All specimens were allowed to cure for at least 28 days before testing. The specimens were strengthened, using the NSM technique, with 6.35 mm (14 in:) and 9.53 mm (38 in:) diameter ribbed GFRP bars and with 2 · 15 mm (0.080 · 0.60 in.) GFRP or with 2 · 16 mm (0.08 · 0.63 in.) CFRP rectangular bars. A two-part code was used to identify the specimens. The first part of the code identifies the parent material. Thus, the first two characters identify the type of masonry used ‘‘CO’’ for concrete masonry and ‘‘CL’’ for clay masonry. Since the specimens were constructed
Table 1 Test Matrix Specimen designation
Thickness of the specimen mm (in.)
Bar type
Groove dimension
Embedment material
CO1-GTE1 CO1-GTE2 CL1-GTE1 CL1-GTE2 CL2-CTE1
143 (5 34) 143 (5 34) 95 (3 34) 95 (3 34) 95 (3 34)
GFRP rectangular GFRP rectangular GFRP rectangular GFRP rectangular CFRP rectangular
Rectangular groove 17 · 3 (0.67 · 0.12)
Epoxy
CO2-GRE21 CO2-GRE22 CO2-GRE23
92 (3 58) 92 (3 58) 92 (3 58)
GFRP circular GFRP circular GFRP circular
Square groove 2.25 the diameter of the bar 14.3 mm (0.56 in.)
Epoxy
CO2-GRC31 CO2-GRC32 CO2-GRC33
92 (3 58) 92 (3 58) 92 (3 58)
GFRP circular GFRP circular GFRP circular
Square groove 2.25 the diameter of the bar 21.4 mm (0.84 in.)
Cementitious paste
CO2-GRE21-SJ CO2-GRE22-SJ CO2-GRE21-S CO2-GRE22-S
92 92 92 92
GFRP GFRP GFRP GFRP
Square groove 1.5 the diameter of the bar 9.5 mm (0.37 in.)
Epoxy
(3 58) (3 58) (3 58) (3 58)
circular circular circular circular
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
103
Fig. 1. Test specimens: (a) stack bond pattern; (b) running bond pattern.
using bricks or blocks coming from different stocks, the third number in the code is to identify the stock of the material. From Table 1 it can be observed that four different types of masonry were used: two for the clay walls (CL1 and CL2), and two for the concrete walls (CO1 and CO2). The second part of the code identifies the type and amount of FRP reinforcement. In particular, the first character is the type of FRP reinforcement used: ‘‘G’’ for GFRP and ‘‘C’’ for CFRP. The second character represents the cross-section of the bar used: ‘‘T’’ for rectangular bar and ‘‘R’’ for circular bar. The third character identifies the type of embedding material: ‘‘E’’ for epoxy paste and ‘‘C’’ for latex modified (cementitious grout). For the specimens using rectangular bars the last number represents the number of bars used for the strengthening. For the other specimens, the two numbers following the character ‘‘E’’ or ‘‘C’’ represent the diameter in eighths of an inch and the number of NSM Bars per specimen, respectively. The final character ‘‘S’’ or ‘‘SJ’’ indicates that the masonry panels were built with a stack pattern bond type with the reinforcement crossing the blocks
CO2-GRE21-SJ
CO2-GRE22-SJ
or placed in the vertical joints (see Fig. 2). Thus CO2-GRE21-SJ, refers to a concrete masonry panel, built with a stack pattern bond type, strengthened with one 6.35 mm (14 in:) diameter GFRP bar embedded in epoxy-based paste and placed in the vertical joints. 2.2. Material properties Tests were performed to characterize the mechanical properties of the materials used in this investigation. The average compressive strengths of the concrete and the clay masonry obtained from the testing of prisms (ASTM C1314) are presented in Table 2. A mortar Type N was used for the walls construction; standard mortar specimens were tested according to ASTM C109. An average value of 7.6 MPa (1100 psi) at an age of 28 days was found. Tensile tests were performed on FRP bars to determine their properties, which were related to fiber content. The average tensile strength, ultimate strain and modulus of elasticity obtained from the testing (ASTM D3039) are presented in Table 3. Details of the coupon
CO2-GRE21-S
CO2-GRE22-S
Fig. 2. Reinforcement scheme for specimens built with a stack bond pattern.
104
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
Table 2 Compressive strength of masonry walls Dimensions of masonry units mm (in.) 150 · 200 · 100 · 200 · 100 · 200 · 100 · 200 ·
00 00 00 400 (558 778 1534 ) 500 700 300 400 (38 78 154 ) 00 00 00 65 (334 212 778 ) 00 00 00 65 (334 212 778 )
Masonry type
Specimen code
Compressive strength MPa (ksi)
Concrete
CO1
11.4 (1.6)
Concrete
CO2
10.5 (1.5)
Clay
CL1
19.43 (2.8)
Clay
CL2
15.78 (2.3)
fabrication and testing procedure are shown elsewhere [8]. Splitting tensile tests (ASTM C496) were performed on the epoxy-based and on the cementitious-based embedding material used. The splitting tensile strength was found to be 3.58 MPa (0.518 ksi) after 7 days and 5.59 MPa (0.81 ksi) after 28 days in the case of latex modified cementitious paste, and 16.31 MPa (2.36 ksi) after 7 days and 18.54 MPa (2.7 ksi) after 28 days in the case of epoxy-based paste.
against a steel frame. Linear Variable Displacement Transducers (LVDTs) were positioned in the middle of the walls to measure the midspan deflection during the tests. The load was applied in cycles of loading and unloading. An initial cycle for a low load was performed in every wall to verify that both the mechanical and electronic equipment were working properly.
2.4. Test results 2.3. Test setup The masonry specimens were tested under fourpoints bending (see Fig. 3). Loads were applied by 50.8 · 609.6 · 12.7 mm (2 24 12 in:) steel plates to the external face of the wall. Their distance was 101.6 mm (4 in.) from the midspan. The loads were generated by means of a 12-ton hydraulic jack reacting
2.4.1. Modes of failure The walls exhibited three different modes of failure (see Fig. 4): (1) debonding of the FRP reinforcement from the masonry substrate; (2) flexural failure (i.e., crushing of the masonry in compression of rupture of the FRP in tension); and (3) shear failure of the masonry at the supports.
Table 3 Mechanical properties of FRP bars Bar type
Dimensions of the bar
Average maximum strain (%)
Average maximum stress MPa (ksi)
Average elastic modulus MPa (ksi)
#2 GFRP bar #3 GFRP bar GFRP rectangular bar CFRP rectangular bar
Nominal diameter 6.35 mm (0.250 in.) Nominal diameter 9.53 mm (0.375 in.) 2.06 mm · 15.21 mm (0.08100 · 0.59900 ) 2.06 mm · 15.21 mm (0.08100 · 0.59900 )
1.78 1.85 2.5 0.98
824.5 760.0 1101.7 1392.4
50 163 (7276) 40 800 (5920) 44 000 (6382) 142 740 (20 702)
Fig. 3. Test setup scheme.
(119.6) (110.0) (159.8) (201.9)
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
105
Fig. 4. Modes of failure: (a) FRP bar debonding (CO2-GRC32); (b) FRP debonding: formation of the 45 cracks at the head joints (CO2-GRC31); (c) splitting cracks in the embedding material (CO2-GRE22); (d) flexural-shear failure (CO2-GRC33).
FRP debonding. This was the most frequent mode of failure. Initial flexural cracks were primarily located at the mortar joints (see Fig. 4(a)). A cracking noise during the test revealed a progressive cracking of the embedding paste. Since the tensile stresses at the mortar joints were being taken by the FRP reinforcement, a redistribution of stresses occurred. As a consequence, cracks developed in the masonry units oriented at 45 (see Fig. 4(b)) or in the head mortar joints. Some of these cracks followed the embedding paste and the masonry interface causing debonding and subsequent wall failure. Due to the smoothness of the rectangular bars, some of the specimens reinforced with rectangular bars debond is a result of the bar sliding inside the epoxy. For specimens having a deep groove, debonding was caused by splitting of the embedding material (see Fig. 4(c)). Flexural failure. After developing flexural cracks primarily located at the mortar joints, a wall failed by either rupture of the FRP reinforcement or by the masonry crushing. FRP rupture occurred at midspan and was observed for the specimens CO2-GRE21 and CO2-GRE22. Shear failure. Cracking started with the development of fine vertical cracks at the maximum bending region. Thereafter, flexural-shear failure was observed at an orientation angle at approximately 45. In the flexural-
shear mode, shear forces transmitted over the crack caused a differential displacement in the shear plane, which resulted in FRP debonding (see Fig. 4(d)). 2.4.2. Discussion of results Fig. 5 shows the Moment vs. Deflection Curves for the six series. It can be observed that the strength and stiffness of the FRP strengthened walls increased dramatically when comparing them to a URM specimen. The nominal moments at cracking for the unstrengthened specimens were calculated considering the Masonry Standards Joint Committee recommendations (MSJC-02). From the graphs in Fig. 5 can be stated increments ranging 4–14 times the theoretic flexural capacity of the un-strengthened masonry wall. Since masonry possesses a significant amount of variability attributed to labor and materials, this range of values should be taken simply as a reference. Table 4 reports the test results. The experimental results have been expressed as a function of the amount of reinforcement, qf, defined as AreaFRP =ðWall width Wall thicknessÞ. For some of the specimens utilizing carbon or glass FRP rectangular bars, a higher ductility was observed when compared with the specimens reinforced with circular
106
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110 9
9 Theoretic Unreinforced
8
7
CO1-GTE1
7
CO1-GTE2
6
Midspan Moment (kN-m)
Midspan Moment (kN-m)
Theoretic Unreinforced
8
5 4 3 2
5 4 3
1
0 0
5
10
15 20 25 30 Midspan Deflection (mm)
35
40
45
0 0
b
9
5
10
15 20 25 Midspan Deflection (mm)
30
35
40
45
9
Theoretic Unreinforced CL2-CTE1
8
Theoretic Unstrengthened
8
7
Midspan Moment (kN-m)
Midspan Moment (kN-m)
CL1-GTE1
6
2
1
a
CL1-GTE2
6 5 4 3 2
CO2-GRE21
7
CO2-GRE22
6
CO2-GRE23
5 4 3 2 1
1
0 0 0
c
5
10
15 20 25 Midspan Deflection (mm)
30
35
40
45
0
10
15
20
25
30
35
40
45
Midspan Deflection (mm)
9
9
Theoretic Unstrengthened
Theoretic Unstrengthened
8
8
CO2-GRE21-S
7
CO2-GRE22-S
CO2-GRC31
7
CO2-GRC32
6
CO2-GRC33
Midspan Moment (kN-m)
Midspan Moment (kN-m)
5
d
5 4 3 2 1
CO2-GRE21-SJ
6
CO2-GRE22-SJ
5 4 3 2 1
0
0 0
e
5
10
15 20 25 30 Midspan Deflection (mm)
35
40
45
f
0
5
10
15 20 25 30 Midspan Deflection (mm)
35
40
45
Fig. 5. Moment versus midspan deflection for all the specimens: (a) Series CO1-GTE; (b) Series CL1-GTE; (c) Series CL2-CTE; (d) Series CO2GRE2; (e) Series CO2-GRC3; (f) Series CO2-GRE2x-S.
Table 4 Test results Specimen name
Amount of reinforcement qf (·105)
Ultimate load kN (kip)
Maximum bending moment kN m (k ft)
Type of failure
CO1-GTE1 CO1-GTE2 CL1-GTE1 CL1-GTE2 CL2-CTE1 CO2-GRE21 CO2-GRE22 CO2-GRE23 CO2-GRC31 CO2-GRC32 CO2-GRC33 CO2-GRE21-SJ CO2-GRE22-SJ CO2-GRE21-S CO2-GRE22-S
33 67 50 100 50 54 107 161 136 272 408 50 101 50 101
14.7 (3.3) 35.1 (7.9) 9.4 (2.1) 16.0 (3.6) 25.6 (5.8) 11.0 (2.5) 11.5 (2.6) 21.6 (4.9) 13.1 (2.9) 15.0 (3.4) 26.6 (6.0) 2.84 (0.64) 16.2 (3.64) 8.4 (1.9) 15.8 (3.5)
3.35 8.03 2.16 3.66 5.86 2.52 2.64 4.94 2.99 3.43 6.07 0.81 3.43 1.91 3.60
Debondinga Debondingb Debondingb Debondinga Debondinga Debondingb Debondingb Debonding Debonding Debonding Shear Debonding Debonding FRP rupture FRP rupture
a b
Debonding due to sliding of the bar inside the epoxy paste. Debonding due to splitting of the embedding material.
(2.47) (5.92) (1.59) (2.70) (4.32) (1.86) (1.95) (3.64) (2.20) (2.53) (4.48) (0.60) (2.53) (1.41) (2.66)
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
107
bars. In fact, for these specimens the failure was due to the sliding of the bars inside the groove. In these cases, after the failure, the wall could still carry load (because of the friction between the rectangular bar and the epoxy paste). An interesting observation can be underlined for specimens built with a stack pattern bond type. There is not a considerable reduction in the out-of-plane performance by placing the bar in the vertical joints or when it crosses the masonry blocks. The inadequate performance of wall CO2-GRE21-SJ is attributed to construction problems. In fact, this wall had a big mortar joint and therefore the epoxy bordered on the mortar and not on the concrete blocks. Therefore it is advisable, as a construction detail, to make grooves bordering the concrete block surfaces when bars are placed along the joints.
theoretical flexural capacity of an FRP strengthened masonry wall was determined based on strain compatibility, internal force equilibrium and the controlling mode of failure. Thus, the theoretical flexural capacities were estimated based on the assumption that no premature failure was to be observed (i.e., either rupture of the laminate or crushing of the masonry would govern the wall behavior). For simplicity and complementary to the flexural analysis of RC members, a parabolic distribution was used for compressive stresses in the computation of the flexural capacity of the strengthened walls. The stress block parameters associated with such a parabolic distribution are given as: 8 2 > < cb1 ¼ eem0 13 eem0 ; m m ð1Þ 2 > : cb 1 1 b ¼ 2 em0 1 em0 . 1 2 1 3 e 4 e
2.4.3. Previous results Turco [7] conducted an experimental investigation on URM walls reinforced with FRP NSM bars. The materials used, the test setup and the dimensions of the masonry panels were the same as the ones described in this paper. In particular a ‘‘Type CO2’’ masonry specimen was used. The differences between the two sets of investigations were in the reinforcement details. Table 5 summarizes the aforementioned results in terms of ultimate bending moment and mode of the failure. It was observed that, in the case of strengthening with NSM FRP bar, the latex modified cementitious pasteGFRP bar system exhibited a better performance when the size of the groove was approximately 2.25 times the diameter of the GFRP reinforcement. On the contrary, a groove of 1.5 times the diameter of the bar was enough when epoxy paste is used.
According to MSJC, the maximum usable strain emu was considered to be 0.0035 mm/mm (in./in.) for clay masonry and 0.0025 mm/mm (in./in.) for concrete masonry [9]. The tensile strength of the masonry was neglected. The theoretical shear capacity was estimated according MSJC [9] where a shear strength of 13 fm0 is recommended for URM in a running bond that is not grouted solid. The net cross-section was used for the computation of the shear capacity. The reinforcement index, xf, expressed as qf Ef =fm0 ðh=tm Þ, is an index that intends to capture the key parameters that influence the flexural capacity. These include the FRP flexural reinforcement ratio, qf, the FRP tensile modulus of elasticity, Ef, the masonry compressive strength, fm0 , and the slenderness ratio h/tm. This index is intended to represent the ratio of axial stiffness (cross-sectional area · modulus of elasticity) between the FRP and the masonry (AfEf/bmtmEm) but, since the modulus of elasticity of masonry Em is directly proportional to fm0 , the latter can replace Em. The inclusion of the slenderness ratio h/tm has been identified as influential in the out-of-plane behavior of masonry walls. h/tm accounts for the ability of the masonry wall behavior to be controlled by flexural capacity rather
2.5. Basis for design approach Table 6 presents the experimental and theoretical results used as a database for the development of a design approach for the FRP strengthening of URM walls. The
m
m
Table 5 Previous test results [7] Bar type number of bars
Amount of reinforcement qf (·105)
Groove dimension
Embedment material
Maximum bending moment kN m (k ft)
Mode of failure
2#2 3#2 1#3 2#3 3#3 2#2 3#2 1#3 2#3
110 170 120 240 360 100 170 120 240
Square groove 1.5 the diameter of the bar 9.5 mm (0.38 in.) Square groove 1.5 the diameter of the bar 14.3 mm (0.56 in.)
Epoxy Epoxy Epoxy Epoxy Epoxy Cementitious Cementitious Cementitious Cementitious
1.68 2.27 1.56 3.93 5.57 2.07 2.93 0.94 1.64
Debonding Debonding Debonding Shear Shear Debonding Debonding Debonding Debonding
a
GFRP GFRP GFRP GFRP GFRP GFRP GFRP GFRP GFRP
bar bar bar bar bar bar bar bar bar
Value too low for practical applications.
Square groove 2.25 the diameter of the bar 14.3 mm (0.56 in.) Square groove 1.5 the diameter of the bar 14.3 mm (0.56 in.)
paste paste paste paste
(1.24) (1.67) (1.15) (2.90) (4.11) (1.53) (2.16) (0.69)a (1.21)a
108
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
Table 6 Comparison between experimental and theoretical results Source
qf (·105)
Flexure kN m (k ft) Mexp.
Mth.
CO1-GTE1 CO1-GTE2 CL1-GTE1 CL1-GTE2 CL2-CTE1 CO2-GRE21 CO2-GRE22 CO2-GRE23 CO2-GRC31 CO2-GRC32 CO2-GRC33 CO2-GRE21-SJ CO2-GRE22-SJ CO2-GRE21-S CO2-GRE22-S Turco [7] Turco [7] Turco [7] Turco [7] Turco [7] Turco [7] Turco [7] Turco [7] Turco [7]
33 67 50 100 50 54 107 161 136 272 408 50 101 50 101 101 161 136 272 408 101 161 136 272
3.35 8.03 2.16 3.66 5.86 2.52 2.64 4.94 2.99 3.43 6.07 0.81 3.43 1.91 3.60 1.68 2.27 1.56 3.93 5.57 2.07 2.93 0.94 1.64
4.59 9.02 2.87 5.63 3.00 2.10 3.83 4.57 3.87 5.20 6.07 2.20 3.77 2.17 4.05 3.83 4.57 3.87 5.20 6.07 3.83 4.57 3.87 5.20
a
(2.47) (5.92) (1.59) (2.70) (4.32) (1.86) (1.95) (3.64) (2.20) (2.53) (4.48) (0.60) (2.53) (1.41) (2.66) (1.24) (1.67) (1.15) (2.90) (4.11) (1.53) (2.16) (0.69)a (1.21)a
Shear kN (kip) Vexp. (3.38) (6.65) (2.12) (4.15) (2.21) (1.55) (2.83) (3.37) (2.86) (3.83) (4.48) (1.62) (2.78) (1.60) (2.99) (2.83) (3.37) (2.86) (3.83) (4.48) (2.83) (3.37) (2.86) (3.83)
7.32 17.56 4.72 8.01 5.52 5.52 5.78 10.79 6.52 7.50 13.29 1.78 7.50 4.18 7.89 3.68 4.95 3.41 8.60 12.19 4.54 6.41 2.05 3.59
Vth. (1.65) (3.95) (1.06) (1.80) (1.24) (1.24) (1.30) (2.43) (1.47) (1.69) (2.99) (0.40) (1.69) (0.94) (1.77) (0.83) (1.11) (0.77) (1.93) (2.74) (1.02) (1.44) (0.46) (0.81)
24.02 24.02 21.18 21.18 19.20 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99 14.99
(5.4) (5.4) (4.76) (4.76) (4.32) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37) (3.37)
Value too low for practical applications.
than shear capacity. h/tm and the required out-of-plane load to cause failure are inversely proportional; thus, as the slenderness ratio decreases, the out-of-plane load becomes larger. Fig. 6 shows the relationship between the experimental–theoretical flexural capacity ratio (Mexperimental/ Mtheoretical), and xf, for all of the specimens included in Table 6. The ratio Mexperimental/Mtheoretical represents the effectiveness of the FRP reinforcement. The ratio Mexperimental/Mtheoretical for the specimens failing in shear was computed based on the bending moment associated with the shear capacity. For design purposes, rather than attempting to predict bond failure, the strain in the FRP reinforcement can be limited. In this context, since the flexural capacity is dependant on the strain developed in the FRP bar, it is reasonable to express the effective strain in the bar, efe, as the product jmefu, where jm is the bond dependent coefficient and efu is the design rupture strain of FRP. Thus, the effective strain in the FRP bar, efe, is limited by the strain controlled by debonding: efe 6 jm efu FRP rectangular bars: jm ¼ 0.65; FRP circular bars having a groove 1:5 the diameter of the bar: jm ¼ 0.35; FRP circular bars having a groove 2:25 the diameter of the bar: jm ¼ 0.55.
These limits are valid for the case of walls not subjected to sustained load. In walls under sustained load such as retaining or basement walls, creep rupture considerations need to be taken into account. Thus, for the case of GFRP, jm could be 0.2 (ACI 440.2R-02 [10]). The moment capacity of the strengthened FRP wall is obtained from the equilibrium of the internal forces in the cross-section; thus: bc M n ¼ Af ff tm 1 ; ð2aÞ 2 ðcfm0 Þðb1 cÞbm ¼ Af ff .
ð2bÞ
Since the product jmefu represents the effective strain in the laminate, the stress in the FRP, ff, can be written as ff ¼ efe Ef ¼ ðjm efu ÞEf .
ð2cÞ
Replacing Eqs. (2b) and (2c) in (2a), and multiplying both terms by the factors bm tm fm0 and h/tm, one obtains: Mn qf Ef b1 c ¼ ðj e Þ 1 ð2dÞ m fu 2 bm t2m fm0 ðh=tm Þ fm0 ðh=tm Þ Finally, making xf explicit on the right end side of Eq. (2d) one obtains Mn xf ðjm efu Þ ðh=tm Þ ¼ x ðj e Þ 1 . f m fu 2 c bm t2m fm0 ðh=tm Þ ð2eÞ The moment capacity is then the minimum of either that obtained from Eq. (2e) or the value calculated for the
1 0.8
κmm= 0.65 0.65
0.6 0.4 0.2
Debonding
0 0
0.5
1
1.5
1
κ m= 0.65 0.65
0.8
m
0.6 0.4 0.2
Debonding 0.5
1
0.6
0.35 κm m= 0.35
0.4 0.2 0 0.5
1
1.5
2
ρf E f f m' (h / t) 1.2 1 0.8 0.6
κ mm= 0.55 0.55
0.4 0.2
Debonding Shear
0 0
2
1.5
ρf E f f m' (h / t)
c
0.8
0
0 0
Debonding Flexural Shear
1
b
1.2
109
1.2
2
ρf E f f m' (h / t)
a
M experimental / M theoretical
M experimental / M theoretical
1.2
M experimental / M theoretical
M experimental / M theoretical
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
d
0.5
1
1.5
2
ρf E f f m' (h / t)
Fig. 6. Mexperimental/Mtheoretical versus xf: (a) GFRP and CFRP rectangular bars – epoxy system; (b) GFRP circular bars – epoxy system (dimension of the groove 1.5 the diameter of the bar); (c) GFRP circular bars – epoxy system (dimension of the groove 2.25 the diameter of the bar); (d) GFRP circular bars – latex modified cementitious paste (dimension of the groove 2.25 the diameter of the bar).
theoretical shear capacity. The validation of such a protocol is demonstrated in Fig. 7 which shows the ratio between the experimental and nominal moment capacities as a function of xf. Since such a determined nominal capacity does not contain any safety factor, therefore Fig. 7 demonstrates the safety of the approach. The factor xf is related to the failure mode: at low values of xf correspond debonding of the FRP strengthening while at high values of it, shear failure occurs. From this point of view, it is reasonable to conclude that any safety factor, adopted from a particular design code, can be taken as a function of xf. 1.80
Mexperimental / Mn
1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0
0.5
1
1.5
ωf
2
Fig. 7. Ratio between experimental and nominal moments.
2.5
3. Conclusion The following conclusions can be drawn from this experimental program: 1. Flexural strengthening with FRP systems has been proven to remarkably increase the flexural capacity (from 2 to 14 times), strength and pseudo-ductility of URM walls. 2. The test results identified three basic modes of failure. One, shear failure, related to the parent material (i.e., masonry); and the other two associated with the reinforcing material, debonding and flexural failure (i.e., rupture of FRP or crushing of the masonry). For large amounts of reinforcement, shear failure was observed to be the controlling mode. For other reinforcement ratios, either FRP rupture or debonding was observed, the latter being the most common. 3. In the case of strengthening with FRP rectangular bars, the sliding of the bar in the epoxy caused an increase in the ductility. 4. Based on experimental data obtained in the present and other investigations, it is recommended to consider the maximum usable strain in the FRP Bars as 0.65efu for NSM FRP Rectangular Bars utilizing Epoxy Paste and 0.55efu for NSM FRP Circular Bars having a groove at least 2.25 times the diameter of the bar and utilizing either Epoxy or Cementitious Paste.
110
N. Galati et al. / Construction and Building Materials 20 (2006) 101–110
The maximum usable strain in the FRP Bars must be limited to 0.35efu for NSM FRP Circular Bars having a groove at least 1.5 times the FRP bar diameter and using Epoxy Paste. 5. The proposed design algorithm seems to be suitable to determine the moment capacity of URM walls strengthened with FRP bars. However, the validation of the protocol is based on the findings on two limited experimental campaigns; more tests are needed in order to determine more statistically relevant results. 6. The findings reported in this paper apply only to the materials, configurations and geometries tested and much more work needs to be done before general design recommendations can be provided.
Acknowledgments The support of the National Science Foundation Industry/University Cooperative Research Center at the University of Missouri-Rolla. The authors also acknowledge the support of the Rolla Technical Institute (RTI).
References [1] Schwegler G, Kelterborn P. Earthquake resistance of masonry structures strengthened with fiber composites. In: Proceedings of the 11th world conference on earthquake engineering, Acapulco, Mexico, CD-ROM; 1996. 6 pp.
[2] Hamilton HR, Dolan CW. Flexural capacity of glass FRP strengthened concrete masonry walls. J Comp Const ASCE 2001;5(3):170–8. [3] Tumialan JG, Morbin A, Nanni A, Modena C. Shear strengthening of masonry walls with FRP composites. COMPOSITES 2001 Conv. and Trade Show, Composites Fabricators Assoc., Tampa, FL, October 3–6, CD-ROM; 2001. 6 pp. [4] De Lorenzis L, Nanni A, La Tegola A. Flexural and shear strengthening of reinforced concrete structures with near surface mounted FRP rods. In: Proceedings of third international conference on advanced composite materials in bridges and structures, Ottawa, Canada; 2002. p. 521–8. [5] Hamid AA. Strengthening of hollow block masonry basement walls with plastic reinforcing bars. Masonry Soc J 1996. [6] Tumialan JG, Galati N, Nanni A, Tyler D. Flexural strengthening of masonry walls in q high school using FRP bars. In: ACI, Special Publication 215-12, Boston, September 26–October 1; 2003. p. 413–28. [7] Turco V, Galati N, Tumialan JG, Nanni A. Flexural strengthening of URM walls with FRP systems. In: Proceedings of the 6th international symposium on fibre-reinforced polymer (FRP) reinforcement for concrete structures (FRPRCS-6), Singapore; 2003. [8] Secondin S. Masonry reinforced with FRP systems. Tesi di Laurea, Universita` degli Studi di Padova, Facolta` di Ingegneria, Padua, Italy; March 2003. [9] Masonry Standards Joint Committee. Building code requirements for masonry structures. ACI-530/ASCE 5/TMS 402. American Concrete Institute, American Society of Civil Engineers, and The Masonry Society, Detroit, New York, and Boulder; 2002. [10] American Concrete Institute (ACI). Committee 440. Guide for the design and construction of externally bonded FRP systems for strengthening concrete structures; October 2002.