Tumialan, G., P.C. Huang, A. Nanni, and P.

Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

Strengthening of Masonry Walls by FRP Structural Repointing

GUSTAVO TUMIALAN, Doctoral Student, University of Missouri-Rolla, USA, PEI-CHANG HUANG, General Manager, CO-FORCE, Taiwan, ANTONIO NANNI, Professor, University of Missouri-Rolla, USA, and PEDRO SILVA, Visiting Assistant Professor, University of Missouri-Rolla, USA

INTRODUCTION Unreinforced Masonry (URM) walls are prone to failure and collapse during a major seismic event representing one of the main causes of injuries and loss of human lives. In this context, the search of effective and affordable techniques to strengthen URM walls is an urgent need. Repointing is a traditional retrofitting technique, commonly used in the masonry industry, which consists in replacing missing mortar in the joints. The term structural is added to describe a strengthening method aimed at restoring the integrity and/or upgrading the capacity of walls. This is achieved by placing into the joints deformed Fiber Reinforced Polymer (FRP) rods, which are bonded to the masonry wall by a suitable paste. In addition to improving structural performance (Tinazzi, 2000) this technique offers other advantages, namely; the surface preparation is reduced and aesthetics is preserved. This paper consists of two parts. The first part deals with the study of the shear behavior of URM walls strengthened using a method denominated FRP Structural Repointing. The test results of an experimental program are presented. Four specimens were tested under diagonal loads. The results showed that by using the FRP Structural Repointing technique, the shear strength of URM masonry walls was significantly increased. The second part presents a methodology to estimate the shear capacity of walls strengthened by the aforementioned technique. It was shown that the suggested protocol provided adequate values. EXPERIMENTAL PROGRAM Materials A total of four masonry walls were manufactured for this experimental program, which were built with 6x8x16 in. (150x200x400 mm) concrete blocks in a stack bond pattern. The dimensions of the walls were 64 in. by 64 in (1.6 m. by 1.6 m.). All the walls were built by a qualified mason to not introduce additional variables, such as handwork and different mortar workability that may arise from the construction of the specimens. The mortar used was Type N. The compressive strength of masonry obtained from the testing of prisms was 2090 psi (14.4 MPa). The walls were strengthened with Glass FRP rods having a diameter of inches (6.25 mm.), a tensile strength of 120 ksi (827 MPa) and modulus of elasticity of 5900 ksi (40.7 MPa). The GFRP rods were embedded into an epoxy-based paste with the following mechanical properties: the compressive strength was 12.5 ksi (86.1 MPa), the tensile strength was 4 ksi (27.5 MPa), and the modulus of elasticity was 450 ksi (3.1 GPa). Test Specimens One URM wall, Wall 1 was the control specimen for this test series. Wall 2 was strengthened with GFRP rods at every horizontal joint. Wall 3 was strengthened with GFRP rods in a grid pattern, which means that the rods were placed in every vertical and horizontal joint. For Wall 4 the amount of reinforcement was similar to that of Wall 3 but the reinforcement was
1

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

distributed in the two sides of the wall. The horizontal reinforcement was installed in the front side, whereas the vertical reinforcement was placed in the back side of the wall. Table 1 illustrates the matrix used for this experimental program. Table 1. Matrix for Experimental Program Specimen Reinforcement Wall 1 Wall 2 Wall 3 Wall 4 None #2 GFRP Rods #2 GFRP Rods #2 GFRP Rods

Front Side None

Back Side None None None Every Vertical Joint

Every Horizontal Joint Every Horizontal and Vertical Joint Every Horizontal Joint

Strengthening Procedure Structural repointing offers advantages compared to the use of FRP laminates. The method itself is simpler since the surface preparation is reduced (sandblasting and puttying) is not required. In addition the aesthetic of masonry is preserved. Figure 1 illustrates the strengthening procedure.

(a) Grinding of Joints

(b) Masking of Masonry to avoid Staining

(c) Application of Epoxy-based Paste Figure 1. Strengthening by Structural Repointing

(d) Installation of GFRP Rods

The diameter size of the GFRP rods is limited by the thickness of the mortar joint, which usually cannot be larger than 3/8 inches (9.4 mm.). The GFRP rods were placed into the joints by using a technique known as tuck pointing, which consists of: (1) cutting out part of the mortar using a grinder, (2) filling the joints with a epoxy-based paste, (3) embedding the rods in the joint and (4) retooling. To ensure a proper bonding between the epoxy-based paste and masonry, it is recommendable to remove the dust by means of an air blower once the grinding of the mortar joints has been completed.
2

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

Test Setup The specimens were tested in a closed loop fashion. Two 30-ton-capacity hydraulic jacks activated by a manual pump were used to generate the load along the diagonal of the wall being tested. When loading, the force was applied to the wall by steel shoes placed at the top corner, and transmitted to similar steel shoes at the bottom corner through high strength steel rods. Figure 2 shows an overall view of the test setup.

(a) Overall View Figure 2. Test Setup

(b) Specimen being tested

The load was applied in cycles of loading and unloading, except in the control wall. An initial cycle for a low load was performed in every wall to verify that both the mechanical and electronic equipment were working properly. By applying the load by cycles, the stability of the system can be verified. The data acquired by the load cell and the Linear Variable Differential Transducers (LVDTs) were collected by a DAYTRONIC data acquisition system at a frequency of one point per second. A total of four LVDTs were used to collect displacements in the walls. A couple of LVDTs was placed on each side of the walls. One oriented along the line force and the other perpendicular to the line. The latter one was placed to register the crack opening. EXPERIMENTAL RESULTS Figure 3 illustrates the load vs. crack opening curves registered at the front and back sides for the four tested walls. It can be observed that at about 10 kips (44.5 kN) all the four test specimens experienced a reduction in their initial stiffness. Wall 1 maximum capacity was registered at about 26 kips (115.6 kN), and as expected the capacity of Wall 1 was sensitive to the weaker planes along the bed and head joints (see Figure 4a) with cracks developing only in these joints. The maximum loads in Walls 2, 3 and 4 was approximately the same, with an average load of 53 kips (235.7 kN). In these strengthened walls, the presence of the reinforcement, forced the formation of diagonal cracks running through the masonry units (see Figures 4b, 4c and 4d). Thus, the tensile forces in the rods bridging the diagonal crack increased the shear capacity of the walls.

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

Figure 3. Envelopes of Load vs. Crack Opening

(a) Wall 1

(b) Wall 2

(c) Wall 3 Figure 4. Specimens after Failure

(d) Wall 4

In Wall 1 the failure was brittle, typical of a dominated shear failure. In this wall some material come loose after the ultimate load was reached that could potentially fail due to any out-of-plane loading. In a real building, this fact could cause injuries or loss of human life during a seismic event. On the other hand, at the final state, in all the strengthened walls no loose material was observed. Wider cracks were mostly observed in the unstrengthened (back) side or where minimum amount of reinforcement was placed such as in Wall 4. It should be noted that these cracks were not visible until the peak load was reached. In addition the strengthened walls tilted to the direction of the strengthened face, which was more evident in Wall 3 (see Figure 5a), which
4

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

was strengthened with GFRP rods in the horizontal and vertical joints placed in one side of the wall.

Crack Propagation

Strengthened Side

Unstrengthened Side

(a) Strengthened Wall after Failure Figure 5. Mechanism of Failure

(b) Crack Propagation

Mechanism of Failure In the strengthened walls, the failure was produced by the loss of bonding between the epoxybased paste and the masonry units. Comparing the recorded crack widths in the front and back sides, the crack growth in the unstrengthened or less strengthened back side increased at a higher rate than the strengthened front side; as shown in Figure 3. The crack produced by debonding of the masonry units from the mortar in the back side, traveled through the wall thickness until debonding of the epoxy-based paste from the masonry units (see Figure 5b). At this point the wall fails because the tensile stresses are not longer transferred to the rods. It is important to highlight that the observed mechanism defers from that where debonding is observed along the three epoxy paste-masonry interfaces of the groove. It will be assumed for the estimation of the maximum shear strength that since neither debonding nor breaking of the GFRP rods were observed, the bond stresses estimated by a previous investigation (De Lorenzis, 2000) were reduced in half. In similar way, the strength developed in the GFRP rods is assumed to be half of the ultimate tensile strength. EVALUATION OF SHEAR STRENGTH OF STRENGTHENED WALLS The nominal shear strength of a masonry wall can be estimated as the sum of the shear contributions of the masonry and the steel shear reinforcement. The shear strength of a wall strengthened with FRP systems can be quantified by adding a third term to account for the contribution of the new reinforcement: (1) Vn = Vm + Vs + Vf Calculation of Vf Vf depends on the shear contribution of reinforcing rods developing their full tensile capacity and rods being debonded. Thus, two areas can be identified in a masonry panel (see Figure 6). The following assumptions are considered: Inclination angle of the shear cracks constant and equal to 45. Constant distribution of bond stresses along the FRP rods at ultimate.
5

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

The ultimate bond strength is reached in all the rods intersected by the crack at ultimate. The spacing between rods is the layer height.

Le Le

Bond Controlled Region

Rupture Controlled Region

Bond Controlled Region

Figure 6. Controlling areas to calculate Vf The bond behavior is dependent on the type of the rod; thereby, the assumption of constant bond stresses at ultimate may not result adequate. In this case, the value of the average bond strength would depend on the bonded length and could be computed from the local bond stressslip relationship of the given type of FRP rod. Le is defined as the length at which the rod breaks, and can be derived from Figure 7:
b ffu

Le Figure 7. Effective Length Le By equilibrium the force due to the bonding stress is equal to the force generated by the tensile stresses in the rod; thus: b A f = f fu A b
d2 b b ( d b Le ) = f fu 4 Then the effective length can be expressed as: df L e = b fu 4 b

(2)

Shear in Bond Controlled Region (Vb) The number of rods (rb) in the bond controlled region can be quantified as: L rb ' = e (3a) s rb = 2rb (3b) The value obtained using equation (3a) is rounded to the immediate inferior integer.
6

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

The shear force resisted by the FRP rods in this region can be calculated as: Vb = n d b b L t (4) where all these parameters are known: n = number of the strengthened sides of the wall (1 or 2) b = assumed bond stress (De Lorenzis, 2000) Lt = sum of the bonded lengths of all the rods crossed by the crack, calculated in the most unfavorable crack position (minimum total length). The value of Lt is calculated as: L t = rb s

(5)

Shear in Rupture Controlled Region (Vt) The number of rods (rt) in the rupture controlled region can be quantified as: rt = r rb The shear force resisted by the FRP rods in this region are calculated as: Vt = n rt Ai f fu

(6) (7)

Since long-term exposure to various types of environments may reduce the tensile properties of the FRP reinforcement, the material properties used in design equations should be reduced based on the environmental exposure condition by an appropriate environmental reduction factor CE (ACI-440, 2000). Thus: * f fu = C E f fu (8)
* where ffu is the guaranteed ultimate tensile strength of the FRP rod as reported by the manufacturer. Finally, the shear force resisted by the FRP rods in both regions can be estimated as: Vf = Vb + Vt (9)

EXAMPLE OF CALCULATION OF NOMINAL SHEAR STRENGTH The validation of the strengthening of Wall 2, strengthened with GFRP rods at every horizontal joint in one side is presented. Computation of Vm The net area of a concrete block with dimensions of 6x8x16 in. including the mortar in one of the head joints is: Acb = 55.5 in2 Then, the net area of the horizontal masonry section: A mv = 4 (55.5 in 2 ) = 222 in 2 The contribution of masonry to the shear strength is computed by the 1997 Unified Building Code (UBC, 1997) as follows: Vm = C d A mv f 'm = 1.2 (222 in 2 ) 2090 psi = 12180 lbs = 12.2 kips where according to UBC the nominal shear strength coefficient Cd is estimated as 1.2 Computation of VF As previously mentioned, the bond stresses are limited to half of those estimated by De Lorenzis. Following the same criterion, the strength in the rods is limited to half of the ultimate tensile strength, thus:
* ffu CEffu = 0.5[(0.8)(120 ksi)] = 48ksi

b 0.5(0.45 ksi ) = 0.225 ksi

7

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

CE is equal to 0.8 for GFRP rods in a closed space environment. Also, to be consistent with the approach previously described, so-called bond-controlled and rupture-controlled areas are considered, with the understanding that neither debonding nor breaking of the rods were observed. Calculation of Vb Determine Lt: The effective length can be calculated from equation 2 as follows: df (0.25in )(48 ksi ) L e = b fu = = 13.3in > spacing ' s' = 8 in 4 b 4(0.225ksi ) Thus, from Eq. (3a) the number of rods in the bond-controlled area is: L 13.3 rb ' = e = = 1.7 , then rb' = 1 s 8 From Eq. (3b): rb = 2rb = 2 (1) = 2 rods in the bond-controlled area. From Eq. (4), the shear force carried by the rods in this region is: Vb = n d b b L t = (1)()(0.25in )(0.225 ksi )(16in) = 2.8kips Calculation of Vt: The rods in the rupture-controlled area are the remaining rods; thus in Eq. (6): rt = r rb = 7 2 =5 rods From Eq. (7): Vt = nrr Ai f fu = (1)(5rods)(0.05in 2 )(48 ksi) = 12 kips The shear contribution of the GFRP rods estimated from Eq. (9) is: Vf = Vb + Vt = 2.8 kips + 12 kips = 14.8 kips Finally the capacity of the strengthened wall is: Vst = Vm + VF = 12.2 kips + 14.8 kips = 27.0 kips By comparing the horizontal component of the ultimate loads with the expected loads the following results are found: 1 Control Wall 1: Vm-exp. = Vmax sin 45 o = ( 26.1kips ) = 18.5kips 2 1 Strengthened Wall 2: Vst-exp. = Vmax sin 45o = (51.4kips ) = 36.3kips 2 Figure 6 illustrates the correlation between the expected and experimental shear strengths. It is observed that the presented methodology provides reasonable values. Even though it is recognized that the vertical steel increases the shear strength of internally reinforced walls, in design its contribution is commonly not considered. Similarly, and to the light of the results (Walls 3 and 4), when implementing a design protocol for walls strengthened by structural repointing the contribution of the vertical FRP reinforcement can be neglected.

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

Figure 6. Expected and Theoretical Shear Strengths Alternatively to the diagonal tension failure, a crack along a horizontal joint can be observed at a lower load level in infill masonry walls. The resulting horizontal crack divides the infill wall in two parts. This failure mechanism is commonly known as Knee Brace or Joint-Slip. Due to its premature characteristic and negative effects, this kind of failure should be avoided. A potential way to prevent it would be to place of vertical FRP reinforcement on the masonry infill, which would act as a dowel action. CONCLUSIONS The following conclusions can be drawn from this research program: Remarkable improvements of about 100% in wall capacity were registered. Walls strengthened with same amount of reinforcement, distributed in one or two faces, exhibited similar behavior. However, the contribution of the vertical reinforcement may be fully realized in larger walls were more vertical reinforcing rods bridge the diagonal crack. In contrast with URM Walls, strengthened walls were stable after failure. In a real building, this fact can avoid injuries or loss of human life due to catastrophic failures. It was demonstrated that the methodology for the computation of the nominal shear strength of strengthened walls by structural repointing provided a reasonable and conservative fit with the experimental data. ACKNOWLEDGEMENTS This research study was sponsored by the National Science Foundation Industry/University Cooperative Research Center at the University of MissouriRolla. The authors would like to acknowledge the support of the Rolla Technical Institute (RTI); especially, that of Mr. Harold Martin and Mr. Jason Cox. REFERENCES International Conference of Building Officials. Masonry Codes and Specifications, in UBC97, Chapter 21, ICBO, Whittier, CA, 1997. De Lorenzis, L. Strengthening of RC Structures with Near-Surface Mounted FRP Rods, MSc Thesis, Department of Civil Engineering, The University of Missouri Rolla, Rolla, Missouri, 175 pp., May 2000.
9

Tumialan, G., P.C. Huang, A. Nanni, and P. Silva, "Strengthening of Masonry Walls with Structural Repointing," Non-Metallic Reinforcement for Concrete Structures - FRPRCS-5, Cambridge, July 16-18, 2001.

D. Tinazzi, M. Arduini, C. Modena, and A. Nanni, Strengthening of Masonry Walls with FRP Rods and Laminates, Third International Conference on Advanced Composite Materials in Bridges and Structures, Ottawa, Canada, August 2000. American Concrete Institute (ACI), Committee 440, Guide for the Design and Construction of Concrete Reinforced with FRP Bars, (document under review).

10