EXPERIMENTAL STUDIES TO VERIFY THE EFFICACY OF EARTHQUAKE RESISTANT MEASURES IN MASONRY BUILDINGS

Abstract

It is a common experience that traditional masonry buildings collapse or seriously damaged even in moderate earthquakes resulting in heavy loss of human lives besides colossal economic losses. These killer buildings are usually built using the locally available materials such as mud, bricks, stones or blocks and laid in mud, lime or cement mortar employing local artisans who may not be familiar with the relatively newer techniques, which are expected to improve the seismic performance of such construction. Inspite of the provisions and recommendations of the various codes, it still remains a common story to find that these are rarely implemented in actual practice leave aside their strict compliance particularly in private housing. The masonry buildings have unquestionably proved to be the most vulnerable to earthquake forces and have suffered maximum damage. There is a lot of research conducted in the development of earthquake resistant and retrofitting measures in masonry buildings and these methods have been introduced in the code of practices. To reinforce the confidence level of the users and builders in various provisions of seismic codes, there was a need to test the conventional model as well as the model with earthquake resistant (ER) features together under the same intensity of shaking as the majority of population in India lives in low strength masonry houses. The majority of testing done so far for evaluating the seismic resistance is on isolated (one at a time) small scale masonry models. Very few tests have been conducted on conventional and earthquake resistant masonry models together under the same intensity of shaking. In the present study, effectiveness of various strengthening and hi retrofitting techniques for masonry buildings have been evaluated through shock table testing and compared with traditional building models. In the present study, the tests have been conducted on half-scale models of one storeyed brick, concrete block and stone masonry houses with and without seismic strengthening measures constructed simultaneously and tested under increasing intensity of shaking to eliminate the uncertainty of base motion parameter and the aberrations resulting from the same. The behaviour of each model including pattern of cracking, identification of weak zones, mode of failure and damage with increasing shaking have been studied and conclusions have been drawn. Total five sets of testing of masonry models have been done on shock table. In the first set, two half-scale brick masonry models of one room (size 2.03 m x 2.28 m x 1.50 m with wall thickness as 0.115 m) with and without seismic strengthening measures were simultaneously constructed on the shock table in 1:6 cement-sand mortar. It is worth mentioning here that for constructing half-scale models, special half size first class bricks (4.5" x 2.25" x 1.5") were manufactured in brick kiln, locally. The models were imparted shocks with increasing intensity. The peak base acceleration (PBA) varied from 0.53 g (shock-1) to 2.73 g (shock-6). In the conventional model, the inertia force transferred from roof to the walls led to the development of diagonal shear cracks starting from window openings which intensified with increasing intensity of shaking and finally collapsed after shock-6. The other model with earthquake resistant features survived these shocks because of presence of seismic band at lintel level and vertical reinforcements at the corners and jamb from foundation to the rooftop. IV « In the second set of testing, half-scale concrete masonry blocks of size 200mm x 100 mm x 100 mm were made in the laboratory without frogs and the surface was made rough using wire brush for proper bonding. Special blocks with semi-circular hole of 2" diameter were made to accommodate the vertical steel in the earthquake resistant model. The two models were simultaneously imparted shocks withPBAvarying from 1.89 g from shock-2 to 5.50 g in shock-5. The traditionally constructed model started showing signs of distress from the first shock itself and finally collapsed after shock- 5. On the other side, the earthquake resistant model remained intact without any damage even during shock-5 (PBA 5.50 g). In the third set of testing, random rubble stone masonry models of size 2.25 m x 2.70 m x 1.50 m with 0.225 m wall thickness keeping the inner dimension same as 1.80 m x 2.25 m in all the models were tested. The traditional model was constructed without any ER features with sloping roof consisting of burnt clay tiles. While the earthquake resistant model incorporated 'bond' stones, long corner stones, lintel, roof and gable bands, vertical corner and jamb steel and trussed roof with cross bracings covered with clay tiles. To avoid the falling of clay tiles during shaking, wire-mesh laid in concrete mix was used on all four sides and at the ridge level. The two models were simultaneously imparted progressively increasing shock loadings. From the fourth shock (PBA 1.76 g) onwards, the traditional model developed diagonal shear cracks in both the shear walls emanating from four corners of the windows openings. Vertical cracks also appeared at the corners of the walls. Unreinforced gable ends of the masonry walls were also damaged because of strutting action of the purlins imposing additional bending tensile forces and finally collapsed after sixth shock. While in the ER model, seismic bands helped in transferring out-of-plane forces to the shear walls. Vertical reinforcements tied all the four walls and the proper anchorage of the roof with walls helped in avoiding the separation of the roof from the walls and the model remained undamaged. In the fourth set of testing, a conventional random rubble stone masonry model was constructed and strengthened externally applying splints and bandages techniques using 18 gauge wire mesh. For 'through' (header) stones in the walls, 'S' shaped mild steel bars were inserted in the holes at appropriate locations and the holes were filled with Ml5 grade concrete mix to avoid the separation of the two wythes of the masonry walls. For holding the clay tiles in position, wire-mesh laid in concrete mix was also used in the externally strengthened model. The same ER model was tested with the strengthened model because practically no damage was observed to it in the previous test. In the first three shocks, some moderate damage occurred to externally strengthened model and it finally collapsed in fourth shock (PBA 1.98 g). After careful observation it was found that retrofitting did not work effectively because of its inappropriate implementation. During this testing also model with ER features was safe and no damage was observed to it. In the fifth and final set of testing, proper precautions were taken to strengthen the conventional model. As mentioned in the code, vertical reinforcement at the corners should start from the foundation level itself which was omitted in the previous case. Firstly, a rectangular piece of 15 cm x 15 cm and 3 mm thick mild steel plate was welded with the platform at all the four corners and near the openings and then the traditional model was fully constructed on the inner periphery of mild steel plate. For incorporating the strengthening measures, 15 cm wide, 18 gauge wire mesh was properly welded to the mild steel plate acting as vertical reinforcements at the corners and jamb near the openings and were nailed to the wall and plastered. Horizontal VI $ bandages at the lintel and roof levels around the walls and gable belts at the two ends were constructed. ER model was the same as in the previous two tests because practically no damage was observed to it. Progressively increasing shocks were imparted to both the models and it was observed that no visible damage was done to externally strengthened model upto shock-4 (PBA varying from 1.18 g to 2.22 g) while in the previous test, the externally retrofitted model collapsed during shock-4 itself (PBA 1.98 g). From fifth shock onwards, the strengthened model started showing signs of distress and severe damage occurred to it after shock-8. The peak base acceleration varied from 2.55 g to 6.59 g during fifth, sixth and seventh shocks and finally reduced to 2.55 g during shock-8 because the eighth shock was of lower intensity. This verified the fact that external strengthening measures were really effective if properly implemented. Though there was severe damage to it at a peak base acceleration of 6.59 g but it did not collapse. The same model with ER features was being used for the third time under gradually increasing intensity of shocks, so there was bound to be some damage to it because of repetitive shocks. Minor damage started in the ER model at the junction of roof and gable band from shock-4 and the clay tiles started slipping during shock-5. From 6th shock (PBA 4.94 g) onwards diagonal shear cracks started appearing in shear walls and slipping of roof from the walls propagated further. The burnt clay tiles were disturbed from their position. Chunk of stone masonry started falling from the both the shear walls after seventh and eighth shock. Finally, during the 9th shock the externally strengthened model collapsed and moderate damage occurred to the model with ER features. This test proved the efficacy of retrofitting measures applied externally and earthquake resistant measures applied in a building model.