Experimental study on PP-band mesh seismic retrofitting for low earthquake resistant arch shaped roof masonry houses

Transcription

1 From the SelectedWorks of Navaratnarajah Sathiparan October 15, 2009 Experimental study on PP-band mesh seismic retrofitting for low earthquake resistant arch shaped roof masonry houses Navaratnarajah Sathiparan Available at:

2 EXPERIMENTAL STUDY ON PP-BAND MESH SEISMIC RETROFITTING FOR LOW EARTHQUAKE RESISTANT ARCH SHAPED ROOF MASONRY HOUSES NAVARATNARAJAH SATHIPARAN Post Doctoral Fellow, The University of Tokyo, Japan PAOLA MAYORCA Project Research Associate, ICUS, Institute of Industrial Science, The University of Tokyo, Japan KIMIRO MEGURO Director/ Professor, ICUS, Institute of Industrial Science, The University of Tokyo, Japan ABSTRACT Unreinforced masonry structure is one of the most popularly used constructions. It is also unfortunately the most vulnerable to the earthquakes. It would collapse within a few seconds during earthquake movement, and does become a major cause of human fatalities. Therefore, retrofitting of low earthquake-resistant masonry structures is the key issue for earthquake disaster mitigation to reduce the casualties significantly. When we propose the retrofitting in developing countries, retrofitting method should respond to the structural demand on strength and deformability as well as to availability of material with low cost including manufacturing and delivery, practicability of construction method and durability in each region. Considering these issues, a technically feasible and economically affordable PP-band (polypropylene bands, which is commonly utilized for packing) retrofitting technique has been developed and many different aspects have been studied by Meguro Laboratory, Institute of Industrial Science, The University of Tokyo. Unreinforced masonry structures with masonry arch shaped roofs are generally characterized by weak, brittle materials, weak element connections and excessive weight. The main points of weakness of the traditional masonry arch roofs that contributed to poor seismic response can be inability of the roof to act as a diaphragm and heavy weight of the roof. But ability of PP-band mesh kept the structure integral during the shaking will help to overcome this issues. Therefore, to evaluate the beneficial effects of the PP-band mesh retrofitting method, shaking table tests were carried out on arch shaped roof masonry structure with and without retrofitting. From tests results, a scaled model with PP-band mesh retrofitting is able to withstand larger and more repeatable shaking than that without PP

3 October 2009, Incheon, Korea band retrofitting, which all verified to reconfirm high earthquake resistant performance. Therefore proposed method can be one of the optimum solutions for promoting safer building construction in developing countries and can contribute earthquake disaster reduction in future. 1. INTRODUCTION Unreinforced masonry structures with masonry arch roofs are generally characterized by weak, brittle materials, weak element connections and excessive weight. The main points of weakness of the traditional masonry arch roofs that contributed to poor seismic response can be summarized as follows: Inability of the roof to act as a diaphragm: Masonry arch roofs are incapable of diaphragm action. This is due to their curved geometries, the load-carrying mechanisms and the weak and brittle materials. The loadcarrying mechanism of these types of roofs is primarily in compression. The non-homogeneous masonry of the roofs is unable to carry tensile or flexural loads. As a result, they are not capable of restraining the top of their supporting walls during ground shaking, nor are they capable of transferring excessive horizontal inertia forces. Furthermore, they induce a pre-earthquake static horizontal force at the top of the walls as they transfer their compressive load to the walls. In the shared load-bearing walls, the thrust from the two adjoining arches cancel each other out. However, in end arches this force causes an unbalanced outward thrust on the wall. Heavy weight of the roof: Perhaps the most important seismic weakness of the masonry arch is their excessive weight. The masonry arch roofs are, by nature heavy, as a minimum roof thickness is required to enable the successful transfer of the gravity load in an arch action. 2. EXPERIMENTAL PROGRAM 2.1 Description of the specimens For shaking table experiment, three models were built in the reduced scale of 1:4 using the unburnt brick as masonry units and cement lime and sand (1:2.8:8.5) mixture as mortar with cement/water ratio of Attention was paid to make the models as true replica of adobe masonry buildings in developing countries in terms of masonry strength even though the construction materials used were those available in Japan. Buildings dimensions were 933mmx933mmx720mm box shape with 380 mm height arch shape roof as shown in Figure 1. Wall thickness is 50mm. The sizes of door and window in opposite walls were 243x485mm 2 and 325x245mm 2 respectively. New Technologies for Urban Safety of Mega Cities in Asia

4 Figure 1: Model dimension (mm) Table 1 shows the shaking table testing programs. The specimens were named according to the following convention: B-R-P-S in which; B is Brick type used for construction of the model A: unburned brick; R is Roof type - AR: arch roof P is Retrofitted condition NR: non-retrofitted RE: retrofitted; S is Condition of tie bar TB: with tie bar (10 mm bar with 160x160mm 2 timber plate at end) TX: without tie bar Table 1: Summary of shaking table test Case No. Specimen Name PP-band Retrofitting Tie bar 1 A-AR-NR-TB X O 2 A-AR-RE-TX O X 3 A-AR-RE-TB O O Totally two out three models were retrofitted with PP-band mesh after construction. The cross-section of the band used was 6x0.24mm 2 and the Experimental Study on PP-band Mesh Seismic Retrofitting for Low Earthquake Resistant Arch Shapes Roof Masonry Houses

5 October 2009, Incheon, Korea mesh pitch was 40mm. The mechanical properties of masonry in terms of compressive, shear and bond strength were given in Table 2. Table 2: Mechanical properties of masonry specimens (in MPa) Specimen Diagonal Compression Shear Bond shear strength strength strength Strength A-AR-NR-TB A-AR-RE-TX A-AR-RE-TB Construction and retrofitting procedure All specimens have two part; box shape house and arch roof. Box shape house consisted of 18 rows of 44 bricks in each layer except openings and arch roof consisted of 29 rows of 11½ bricks in each layer. Box shape house construction process takes place in two days, first 11 rows in first day and remaining rows construct in following day. Arch roof construction procedure take place after 7 days from initial construction of box shape house. Remaining two side walls constructed take place 7 days after arch roof construction. In case of retrofitted specimen initially boxed shaped was constructed and retrofitted. Then arch framework place above box house model and PP-band mesh cover that arch frame. Arch roof construction procedure take place after 7 days from initial construction of box shape house. After 4 days of curing arch framework was removed. PP-band mesh come from inside the arch roof and box house model were welded by ultrasonic welding machine. Remaining two side walls constructed 7 days after arch roof construction. And finally PP-band mesh come from outside the arch roof and outside box house model were welded and retrofitted by wire connectors. Construction and retrofitted procedure of house model was shown in Figure Input motion Simple easy-to-use sinusoidal motions of frequencies ranging from 2Hz to 35 Hz and amplitudes ranging from 0.05g to 1.4g were applied to obtain the dynamic response of both retrofitted and non-retrofitted structures. This simple input motion was applied because of its adequacy for later use in the numerical modeling. Figure 3 shows the typical shape of the applied sinusoidal wave. Loading was started with a sweep motion of amplitude 0.05g with all frequencies of 2Hz to 35Hz for identifying the dynamic properties of the models. The numbers in Table 2 indicate the run numbers. General trend of loading was from high frequency to low frequency and from lower amplitude to higher amplitude. Higher frequencies motions were skipped towards the end of the runs. New Technologies for Urban Safety of Mega Cities in Asia

6 Figure 2: Retrofitting Procedure Figure 3: Typical Shape of Input Sinusoidal Motion Experimental Study on PP-band Mesh Seismic Retrofitting for Low Earthquake Resistant Arch Shapes Roof Masonry Houses

7 October 2009, Incheon, Korea Table 2: Loading Sequence Amplitude Frequency 2Hz 5Hz 10Hz 15Hz 20Hz 25Hz 30Hz 35Hz 1.4g g g g g g g g g sweep 01,02 3. CRACK PATTERN AND ENERGY DISSIPATION MECHANISM For specimen A-AR-NR-TB up to run 21, no major crack was observed in this model. Major cracks were observed closer to openings from run 22. After that, cracks widened with each successive run. At run 30, cracks appear three corner of the window opening and propagate towards the corner of the wall. Also crack observed left half circle of the east wall-arch interaction and this crack propagate full circle of east wall-arch interaction with successive runs. At run 36, wall-arch interaction in both east and west wall was totally cracked. Even it totally cracked, presence of tie bar, arch roof prevented from splitting outward. Also large amount cracks were observed in roof part. Particularly, long horizontal crack at initial roof part of walls in direction shaking was observed. At run 39, top part of the east wall (part, above the window opening) was totally separated from the specimen and it was fallen from specimen. At run 41, top part of the west wall (above the door opening) was totally separated from the specimen and it s fallen from specimen. As a result arch part only supported by two walls, which are in the direction of shaking and due to walls subjected to out-ofplane load; arch part was bursts outwards in shaking direction. This finally led to the structure collapse at run 42. For retrofitted specimen A-AR-RE-TX up to Run 21, no major crack was observed in this model. Major cracks were observed closer to openings from Run 22. After that, cracks widened with each successive run. At run 33, wall-arch interaction in both east and west wall was totally cracked. Long horizontal cracks were observed in top of the arch roof. Also large amount cracks were observed in roof part. Particularly, long horizontal crack at initial roof part of walls in direction shaking was observed. At run 45, although at the end of this run almost most of the mortar joints were cracked, the specimen did not lose stability. Although PP-band mesh kept the structure integral during the shaking, it allowed the sliding of the bricks along these cracks to some extent. At the final stage of the test, run 54, with 50mm base displacement, 33 times more than the input displacement applied in run 42 and 7 times more velocity, virtually all the brick joints were cracked and the building had substantial permanent deformations. New Technologies for Urban Safety of Mega Cities in Asia

8 However, building did not loose the overall integrity as well as stability and collapse was prevented in such a high intensity of shaking. Another important point to mention that this retrofitted model has sustained 12 more runs with higher input energy before this run. In smaller input motions in case of retrofitted specimen A-AR-RE- TB, amount of crack relatively fewer than that case of specimen A-AR-RE- TB. But in higher input motion, in both models, there are not much difference observed in dynamic performance. Like specimen A-AR-RE-TX, run 54 was the last run for specimen A-AR-RE-TB. 4. FAILURE BEHAVIOR AND PERFORMANCE EVALUATION The performances of the models were assessed based on the damage level of the buildings at different levels of shaking. Performances were evaluated in reference to five levels of performances: light structural damage, moderate structural damage, heavy structural damage, partially collapse and collapse. Table 3: Damage categories Category Damage extension D0: No damage No damage to structure D1: Light Hair line cracks in very few walls. The structure structural damage resistance capacity has not been reduce noticeably. D2: Moderate structural damage D3: Heavy structural damage D4: Partially collapse Small cracks in masonry walls, falling of plaster block. The structure resistance capacity is partially reduced. Large and deep cracks in masonry walls. Some bricks are fall down. Failure in connection between two walls. Serious failure of walls. Partial structural failure of roofs. The building is in dangerous condition D5: Collapse Total or near collapse The Japan Meteorological Agency seismic intensity scale (JMA) is a measure used in Japan to indicate the strength of earthquakes. The JMA scale was colored according to the following convention: Index JMA ~4 JMA 5- JMA 5+ JMA 6- JMA 6+ JMA Performance evaluation based on JMA scale Table 4 shows the performances of non retrofitted model A-AR-NR-TB with different JMA intensities. Partial collapse of the non-retrofitted building was occurred at the 39 th run at intensity JMA~4 and total collapse at the 42 nd run at intensity JMA~4. But it should be noted that the model was already cracked in different loadings as discussed in previous section. Experimental Study on PP-band Mesh Seismic Retrofitting for Low Earthquake Resistant Arch Shapes Roof Masonry Houses

9 October 2009, Incheon, Korea Table 4: Performance of A-AR-NR-TB model Acceleration Frequency (Hz) (g) D4 D3 D2 D2 D2 0.6 D5 D4 D3 D2 D2 D1 0.4 D4 D3 D2 D2 D2 D1 0.2 D1 D1 D1 D1 D0 D0 D0 0.1 D0 D0 D0 D0 D0 D0 D0 D D0 D0 D0 D0 D0 D0 D0 D0 Table 5 shows the performances of non retrofitted model A-AR-RE- TX with different JMA intensities. The A-AR-RE-TX performed at heavy structural damage at 42 nd run at which A-AR-NR-TB was collapsed. Heavy structural damage level of performance was maintained until 52 nd run. Test was stopped after the 54 th run due to limitation of the shaking table capacity. It should be noted again that this model survived 13 more shakings in which many runs were with higher intensities than A-AR-NR-TB was collapsed before reaching to the final stage at the 54 th run. Table 5: Performance of A-AR-RE-TX model Acceleration Frequency (Hz) (g) D3 1.2 D3 1.0 D3 0.8 D4,D4 D3 D3 D3 D3 D2 D2 D2 0.6 D3 D3 D3 D3 D2 D2 D2 D2 0.4 D3 D3 D3 D3 D2 D2 D2 D2 0.2 D3 D2 D1 D1 D1 D0 D0 D0 0.1 D0 D0 D0 D0 D0 D0 D0 D D0 D0 D0 D0 D0 D0 D0 D0 Table 6 shows the performances of retrofitted model A-AR-RE-TB with different JMA intensities. The A-AR-RE-TB performed at moderate structural damage at 42 nd run at which A-AR-NR-TB was collapsed. In the same run specimen A-AR-RE-TX was performed at heavy structural damage level. In the 47 th run, another JMA 5+ intensity shaking, the A-AR- RE-TB got the heavy structural damage level, which is crushing, extensive cracking, and damage around openings. Test was stopped after the 54 th run due to limitation of the shaking table capacity. It should be noted again that this model survived 13 more shakings in which many runs were with higher intensities than A-AR-NR-TB was collapsed before reaching to the final stage at the 54 th run. New Technologies for Urban Safety of Mega Cities in Asia

10 Table 6: Performance of A-AR-RE-TB model Acceleration Frequency (Hz) (g) D3 1.2 D3 1.0 D3 0.8 D3,D3 D3 D2 D2 D2 D2 D1 D1 0.6 D3 D2 D2 D2 D2 D2 D1 D1 0.4 D3 D2 D2 D2 D2 D2 D1 D1 0.2 D2 D1 D1 D1 D1 D0 D0 D0 0.1 D0 D0 D0 D0 D0 D0 D0 D D0 D0 D0 D0 D0 D0 D0 D0 4.2 Performance evaluation based on Arias intensity scale The Arias intensity was initially defined (Arias, 1970) as t π 2 Ia = a ( t) dt (1) 2g 0 and was called scalar intensity. It is directly quantifiable through the acceleration record a(t), integrating it over the total duration of the earthquake. The arias intensity is claimed to be measure of the total seismic energy absorbed by the ground. Figure 4 shows the performance level of each specimen against dynamic motion. Figure 5 shows the specimen capacity against dynamic motion for each specimen. From results, it s clearly show that; retrofitted model damage level performance at least 5 times better than that of nonretrofitted model. Figure 4: Performance evaluation based on arias intensity Experimental Study on PP-band Mesh Seismic Retrofitting for Low Earthquake Resistant Arch Shapes Roof Masonry Houses

11 October 2009, Incheon, Korea Figure 54: Specimen capacity against shaking 6. CONCLUSION This paper introduced the shaking table test program on arch shaped roof masonry houses. Form test result showed that; For specimen A-AR-NR-TB, after run 36; due to presence of tie bar, even wall-arch interaction in both east and west wall was totally cracked; arch roof prevented from splitting outward. It totally failed at run 42. AR-RE-TX & A-AR-RE-TB were withstands base displacements 50 times larger and velocities 10 times higher than the A-AR-NR-TB. It should be noted again that these models survived 13 more shakings in which many runs were with higher intensities than A-AR-NR-TB was collapsed before reaching to the final stage at the 54 th run. Considering overall performance of the both specimens, PP-band can effectively increase the seismic capacity of masonry houses and therefore reduce the number of casualties in the coming earthquakes. REFERENCE Mahmoud R. Maheria. Performance of Building Roofs in the 2003 Bam, Iran Earthquake, Earthquake Spectra, Volume 21, No. S1, pages S411 S424, December 2005 T.Fujieda, P.Mayorca, N.Sathiparan and K.Meguro, Experimental Study on the Behavior of PP-band Mesh Retrofitted Masonry Houses using Miniature Models, Bulletin of Earthquake Resistant Structure Research Centre, March 2008, No.41, Pilot Studies for Knowledge Assistance for Verification and Promotion on a New Seismic Retrofitting method for existing masonry Houses by Polypropylene Band Mesh (The Islamic Republic of IRAN), Final Report, JBIC, ICUS, OYO International Corporation, July New Technologies for Urban Safety of Mega Cities in Asia