Oct 17, 2006 - Key words: masonry, FRP, strengthening, confinement, shear test, strength, ... new materials as fibre reinforced plastics give opportunity for ...
CONFERENCE AND BROKERAGE EVENT CONSTRUCTION RESEARCH NEEDS FOR THE HERITAGE PROTECTION Cavtat, Croatia, 14 – 17 October 2006
STRENGTHENING OF CLAY-BRICK MASONRY WITH CARBON FIBRE REINFORCED PLASTIC STRIPS Samo Gostič* & Roko Žarnić* *Building and Civil Engineering Research Institute ZRMK Dimičeva 12, 1000 Ljubljana, Slovenia Key words: masonry, FRP, strengthening, confinement, shear test, strength, stiffness, ductility, energy dissipation. Abstract: Shear failure of slender masonry walls between the openings is a commonly observed phenomena that in many cases led to the colapse of entire buildings during earthquakes. Strengthenig of slender clay masonry walls is demanding because it is accompaigned simultaneously with increase of their stiffness what is not favourable. The new materials as fibre reinforced plastics give opportunity for development of innovative strengthening techniques and their application to heritage buildings. In this paper the innovative approach to strengthening of clay masonry slender walls is presented. The innovation is in the placement of narrow CFRP strips arround the brick rows in critical regions of slender walls. Altogether sixteen masonry walls were tested. The new approach to strenghtening was compared to the oftenly used one (diagonally cross layed CFRP strips). Specimens were tested by horizontal cyclic load under the different intensities of constant vertical load. The advantage of the horizontally applied confinement was clearly demonstrated in terms of very significant increase of ductility and energy dissipation withouth increase of stiffness. The most important is that the new innovative strengthening approach favourably influenced the mechanism of slender wall behaviour which can prevent the entire structure collapse during an earthquake.
1. INTRODUCTION There is an average of 152 earthquakes of magnitude grater than 6 on the Richter scale in the World every year. Two well defined seismic belts, the circum-Pacific and the Mediterranean-Himalayan belts, are subject to the most frequent earthquake shocks. The latter also includes the region of Slovenia. There were at least 10 earthquakes that caused material damage in the last decade. Statistics from past earthquakes clearly pointed the collapse of unreinforce masonry as the major cause of human casualties (URM). Sudden brittle failure makes the URM one of the most dangerous constructional elements. Unfortunatelly a large ammount of heritage building worldwide are constructed from URM. Several conventional strengthening methods were developed in the past. Many of them are still used in Slovenia. The most common are: changing of weak mortar in joints, jacketing of URM walls with reinforced concrete and binding of walls with steel ties. Each of these mentioned methods has its own advantages and dissadvantages. They all are disruptive to residents, realization takes a lot of time, and some of the methods significantly change seismic characteristics of a building. That is why new methods are being more often used in practice. One of the first studies of effect of strengthening masonry wall by fibres was done by Croci et.al. (1987). Schwegler was the first to propose the use of carbon fiber laminates. Triantafillou's researches include wide spectra of composite strengthening. The experimental work focused on masonry, led to proposal of equations for masonry strengthened with FRP. Tumialan (2001) demonstrated the use of FRP bars embedded into the horizontal joints. Valluzzi (2002) performed series of diagonal tests on differently reinforced walls. Double sided, diagonal strengthening with GFRP showed to be the most efficient method. In the present experimental work, the in-plane shear test was performed on series of clay masonry walls. Investigation was focused on diagonal and horizontal strengthening with carbon-fiber-reinforced polymer. Both sides of the masonry were strengthened.
2. EXPERIMENTAL PROGRAM 2.1 Test specimen Sixteen reduced size walls have been made on reinforced concrete base footing (150×20×20 cm). Wall specimens’ height was 126 cm and their width was 106 cm. They were made of new, solid clay brick (250×120×65 mm). Mortar used to build wall specimens was a mixture of cement, lime and sand (Dmax = 4mm) in a ratio of 1 to 2 to 6. All specimens were cured for at least 1 year. Four walls were left unreinforced. Others were strengthened with unidirectional carbon fibers in different configurations; 6 diagonally, 3 horizontally and 3 horizontally with wide strips on left/right side. To avoid the influence of eccentricity strips were applied on both sides. To apply CFRP to the wall, we used wet layup technique. The panels were cleaned with abrasion, before epoxy primer could have been applied. Epoxy adhesive, combined with filler was applied to bond the CFRP on the surface of the wall. Finally the top coat of adhesive was applied to ensure saturation of the Fibers. The material was provided and strengthening carried out by the company I.A.R. S.r.l. from Rovigo in Italy.
(A) Diagonally
(B) Horizontally and vertically
(C) Horizontally
Figure 1. Configuration of CFRP reinforcement
Strengthening of the specimen
Vertical load [kN] 200
300
URM (A)
Z12
Z11
400
500
Z02, Z05, Z17
Z10
Z16, Z18, Z06
Z14
(B)
Z13, Z15, Z08
(C)
Z03, Z07, Z09
Table 1: Test matrix with designation of the specimen Majority of walls were tested under a compression load of 400 kN (aprox. 1/4 of compressive strength). Under that load we have tested 3 specimens from each strengthening configuration group. Diagonally strengthened walls were also tested at 3 different levels of vertical load (200, 300, 500 kN). 2.2 Test Setup Test specimens were tested as the shear cantilevers in the test frame shown in Fig. 2. A combination of vertical compression (loads from building above) and in-plane shear load was applied to every specimen. Free end was at the bottom of the test frame, where vertical and also horizontal load is transmitted into the panel.
Figure 2. Masonry wall during testing (wall is inserted up-side-down)
Concrete cubes, which were layed on the end of the 5m meter long lever, loaded the specimen with vertical loads up to 500 kN. Load was transmitted through momentum joint and a trolley, which enabled cantilever boundary conditions (free rotation and free horizontal displacement of the wall specimen at the free end). Principle of scales held vertical load constant during the experiment (measurements showed alteration of less than 0.25%). The shear load was applied to the wall by a horizontal hydraulic jack, which was controlled by computer software. 2.3 Loading sequence Available equipment enabled us displacement controlled experiment. Controlled displacement was at the free end of the wall specimen. First part of a loading protocol was same for all walls (Fig. 3). After the final displacement was reached, further protocol was programmed with regard to the damage and expected bearing capacity of a particular specimen. Each loading sequence had 3 cycles with the same amplitude and velocity. 30
24.0
Displacement [mm]
20 10 0.5 1.0
1.5 2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
0 -10 -20 -30 0
300
600
900 1200 1500 1800 2100 2400 2700 3000 3300 3600
Time [sec]
Figure 2. Loading protocol 2.4 Instrumentation The instrumentation used consisted of strain gauges to measure fiber reinforcement strains, linear variable differential transducers (LVDTs) to measure displacements and deformeters. Horizontal displacements of the tested walls were measured using 5 LVDTs placed at intervals along the height of the wall. Masonry strains were also measured using LVDTs attached diagonally and vertically along the full-size of the wall. Shear loading was provided by servo-hydraulic actuator of 250 kN capacity (two way acting). Actuator was fixed to the supporting frame, which transferred the shear load to the RC plate of the laboratory.
3. EXPERIMENTAL RESULTS 3.1 Characterization of materials Mechanical properties of masonry depend on variety of factors. Masonry is nonhomogeneous composite structural material, consisting of masonry unit, mortar and strengthening (in our case). We also have to bear in mind the influence of workmanship and the mortar to brick interface conditions including friction. Table 2 gives the results of the various materials tests carried out to support experimental program.
Item Brick Mortar CFRP Masonry
Stack bonded masonry Triplets
Property Compressive strength Tensile strength Compressive strength Flexural strength Tensile strength Young's modulus Compressive strength Young's modulus ν Gc
Result 32 MPa 5.34 MPa 6.77 MPa 2.08 MPa 3400 MPa 230 GPa 12.40 MPa 5.74 MPa 0.12 2.29 GPa
Method EN 772-1:2000 EN 1015-11 EN 1015-11 ASTM D 3039/D 3039M ASTM D 3039/D 3039M EN 1052-1:1998 EN 1052-1:1998 EN 1052-1:1998 EN 1052-1:1998
Bond strength
0.31 MPa
prEN 1052-5
Cohesion Dry friction coefficient
0.7 MPa 0.67
prEN 1052-3:2001 prEN 1052-3:2001
Table 2. The main properties of masonry components 3.2 Failure modes of unstrengthened and strengthened masonry walls Z02 (Vvertical=400kN)
Z10 (Fvertical=500kN)
(a)
(b)
Z08 (Fvertical=400kN)
Z07 (Fvertical=400kN)
(d)
(e)
(c)
(f)
Figure 4: Crack patterns and failures of unreinforced and strengthened masonry walls; (a) and (b) URM, (d) – type B strengthening, (e) – type C strengthening, (f) failure of type (C)
Figure 4 shows the crack pattern of the URM walls. In almost all cases, except for the wall loaded with vertical load of 500 kN, masonry panels failed by propagonation of diagonal crack. Failure mode was a combination of shear and flexural mode. Tensile cracks caused splitting of the bricks. First cracks occurred in corners of the panel, because of the rocking of the wall. Behaviour of strengthened panels was strongly related to configuration of CFRP reinforcement. The predominant mode of failure in type (B) and (C) strengthening was flexural mode, which resulted in local failure of the wall toe. Shear cracks started to develop at approx. 70 % of maximum displacement. Cracks propagation was efficiently obstructed by the CFRP reinforcement, which resulted in appearance of many new minor cracks. Diagonally strengthened walls showed flexural and diagonal crack development. Toe crushing was main cause of failure at this type of strengthening.
4. DISCUSSION OF TEST RESULTS 4.1 Lateral load and displacement Average lateral load-displacement hysteresis envelopes for vertical load of 400 kN, obtained from measurements, are shown in Figure 5. Quantitative results for different types of strengthening are summarized in Table 3. Displacements (de-elastic displacement, duultimate displacement) are in millimetres. Shear resistance Hmax is in kN. Average values for 3 URM walls were taken as reference values for comparison. As can be seen from Table 3 ultimate displacements (and not max shear force) gained the biggest increase. The biggest increase (aprrox. 113%) was in type (C) strengthening (horizontal confinement). In some cases ultimate displacement increased for more than 200% (maximum displacement for wall Z15 was 30.94 mm, and for wall Z09 displacement was 29.65 mm). Improvement in ductility was 57%, where walls were strengthened with horizontal CFRP strips. Maximal shear resistance of the wall increased from 10% to 38% in case of wall Z15. An average increment of 21% showed type (C) as the most efficient strengthening method. Stiffness of the wall was not changed with the applied strengthening. A drop of stiffness is a logical result of equation, where maximum shear resistance is considered. This, of course, leads to lower stiffness of strengthened walls.
URM (a) (b) (c)
de 2.44 2.63 3.17 3.21
du 9.76 10.78 18.66 20.82
Ke 37.53 35.28 32.88 34.28
Hmax 100.33 102.32 117.77 121.40
duct 4.13 4.21 5.84 6.50
Table 3: Experimental test results - average values (Fvertical=400 kN) Strengthening the walls with (B) and (C) configuration of CFRP strips also increased the average ultimate elastic displacement from 2.4 mm to 3.2 mm (an increase of 32%). The least improvement was in cases where diagonal strengthening was used. Almost no or little increase in the meaning of measured quantities was noticed (Fig. 6). This was due to the nature of the URM wall, where 90% to 95% of ultimate load was reached, before diagonal cracks occurred. Diagonal application of CFRP strips improves only the the shear strength of masonry. Therefore the failure is caused by extinction of compressive strength in the
areas of wall toes. In the cases of the diagonal strengthening the effect of vertical load intensity was clearly seen. In the case of lower compressive loads the higher ultimate displacements were achieved and rocking failure mode was predominant. However, rocking was observed at higher horizontal displacement at all axial load levels and toe crushing was observed in cases of high compressive load (400 kN and 500 kN). Effect of CFRP 150
50
Increase [%]
125
Force [kN]
60
Z02 Z16 Z13 Z09
100 75
40 30 20 10
50
0
25 0 0
5
10
15
20
25
30
del duct Hmax
35
(a)
(b)
(c)
Type of Strengthening
Displacement [mm]
Fig. 5. Comparison of Load-Displacement Hysteretic envelopes for different types of strengthening 4.2
Fig. 6. Comparison of Load-Displacement Hysteretic envelopes for different vertical loads
Energy Dissipation
For the purpose of comparing dissipation capacity of tested walls, input energy and dissipated hysteretic energy were calculated. Input energy has been defined as work of the actuator, which was needed to deform the wall, and dissipated energy has been calculated as the area of hysteretic loop in one cycle of loading. Figure 7 shows the amount of dissipated energy in each cycle of loading. As can be seen each loading sequence consisted of three loading cycles (with the same amplitude). Usually peak value of the sequence was reached in the first cycle. That is when new cracks were formed. Next two cycles had the same amplitude and less energy was needed to deform the wall as only friction in crack was to be overcome. Formation of new cracks in the first cycle also caused slight change in stiffness, which resulted in less dissipated energy in every next cycle of the same loading sequence. This can be seen in Figure 8 as a drop of ratio between dissipated and input energy.
0.8
/E
inp
750
diss
diss
500
E
E
Z17 Z16 Z08 Z09
0.9
[Nm]
1000
[Nm]
1
Z17 Z16 Z08 Z09
1250
0.7 0.6 0.5 0.4 0.3 0.2
250
0.1
0
5
10
15
20
25
30
35
40
45
Cycle
Figure 7. Dissipated Energy during testing
0 0
5
10
15
20
25
30
35
40
45
Cycle
Figure 8 Ratio between dissipated and input energy
The type (B) and type (C) strengthened walls dissipated less energy than URM walls in the range of small displacements (elastic region). The initial and also post-yield stiffness was nearly the same for all specimens. From the comparison of the first 20 cycles of loading appears that approximately the same amount of input energy was needed to deform all specimens, but on the other hand less energy was dissipated. At the ultimate displacement of URM, hysteretic loops of type (B) and (C) walls were rather closed and majority of cracks were still to develop. The real advantage of horizontal CFRP can be judged from the ammount of cumulative energy at ultimate displacements. The most energy was dissipated by type (C) strengthened walls. Average dissipated energy of three walls was more than 300% higher than energy dissipated by unreinforced walls. That indicated the advantage of CFRP reinforcement. Comparison of only type (B) and (C) strengthening shows higher difference in the meaning of dissipated energy than in the meaning of other measured quantities.
5. CONCLUSION Nature of URM walls is brittle. To improve seismic resistance, present research was carried out. Sixteen walls with different reinforcement configuration and different vertical load were tested. Failure mechanism of strengthened walls differed from URM. It was changed from brittle shear failure to more ductile toe crushing. The primary mode of failure of diagonally strengthened wall was attributed to the exceedence of the compressive strength at the wall toes, which resulted in a localized compression failure. This failure led to the loss of stiffness and wall became unable to resist any further lateral load. Effectiveness of diagonal strengthening is limited by the wall itself. Dimensions, vertical load and compressive strength have to be considered as a prevailing factor as they determine the failure mechanism of the wall. In some cases compressive strength can be first exhausted in the wall toes, because of the flexural behaviour. The only possibility to achieve the most significant results under toe crushing (which is a compression failure) is confinement of a compressive zone. Considerable strength increases were achieved by high percentage of reinforcement near the highly stressed zones.
REFERENCES [1] Triantafillou, Thanasis C., 1998, "Strengthening of masonry structures using epoxybonded FRP laminates", Journal of Composites for Construction 2 (2) May, 96-104, ASCE [2] Croci, G., D'Ayala, D., D'Asdia, P., Palombini, F., 1987, "Analysis on shear walls reinforced with Fibers.", IABSE Symp. On Safety and Quality Assurance of Civ. Engrg. Struct., Int. Assoc. For Bridge and Struct., Lisbon, Portugal [3] Schwegler, G., 1994, "Masonry construction strengthened with fiber composites in seismically endangered zones.", Proc., 10th European Conf. On Earthquake Engrg.,A. A. Balkema, Rotterdam [4] Valluzi M.R., Tinazzi D., Modena C., 2002, "Shear behavior of masonry panels strengthened by FRP laminates" Construction and Building materials, 16, 409-416 [5] Tumialan, J. G., Nanni, A., 2001, "In-Plane and Out-of-Plane Behaviour of Masonry Walls Strengthened with FRP Systems", Report No. CIES 01-24, May, Center for infrastructure Engineering Studies
