EFFECTS OF PLAN SHAPE ON WIND INDUCED RESPONSE OF TALL BUILDINGS

Abstract

The advances in the development of the high strength materials coupled with more advanced computations methods and design procedure led to new generation of buildings, which are slender and light. Two main problems which emerge for such medium to high rise buildings are: (i) The vulnerability of glazed cladding to both direct wind pressuresand flying debris in wind storms, (ii) serviceability problems arising from excessive motion/deflection near the top of the tall buildings and integrity of non-structural components. The evaluation of wind loads on buildings is carried out mainly by using relevant wind codes/standards and available literature. The codes/standards suggest the pressure coefficients and force coefficients for the standard plan shape buildings and for specific wind angles, but do not cover too elongated and uncommon plan shape buildings. The building location, its orientation, shape, size and overall geometry are often pre-determined from space availability and/or architectural considerations. Therefore, attempts have been made in the present study to find the wind pressure distributions and hence pressure coefficients on the building models having different plan shapes but same height and plan area. The mean response of the buildings is also evaluated using the experimentally obtained wind loads on the corresponding building models. The present study has been made in two phases namely wind tunnel experiments and response analysis. The aim of the experimental study is to carryout extensive wind tunnel testing of building models having different plan shape to evaluate the effectiveness of the building shape and wind incidence angles in changing the pressure distributions and peak suction. Recent studies on full-scale to model-scale comparisons have shown the importance of flow simulation and measurement technique for better results and improvements in the existing information/database. So, in this study, attempts have been made to establish the wind flow similar to open terrain with well scattered obstructions having heights generally between 1.5 to 10 m. The boundary layer in the wind tunnel is simulated by means of the roughness grid. The simulated wind matches the vertical variation of the mean wind speed and the turbulent intensity of the atmospheric wind. The wind tunnel used is a closed circuit, with continuous flow having a test section of 8.2 m length and a cross sectional dimensions of 1.15 m(width) x 0.85 m(height). The height (300 mm) and plan area (10000 mm2) ofall the models are kept same for the comparisons purpose. The first building model has the basic IV square cross-section (plan shape) of dimensions 100 mm x 100 mm. One of the dimensions of the model (depth) is subsequently increased by 25 mm up to 200 mm to make rectangular models. The width of the models is selected such that the plan area of models remains 10000 mm . Finally, the rectangular model of dimension 200 mm x 50 mm is split in to two rectangular parts and arranged in different configurations to form "L" and "T" plan shapes. Thus total nine numbers of building models having a plan shape of square (1), rectangular (4), L-shape (2) and T-shape (2) are tested in the wind tunnel to evaluate the wind pressure distribution. All the models are made using the Perspex sheet of 6 mm thickness at a geometrical scale of 1:300. All the models are instrumented with more than 160 numbers of pressuretaps to obtain a good distribution ofpressures on all the surfaces of building models. Fluctuating values of wind pressure are measured at all pressure points at wind velocity of 15 m/sec on different surfaces of building models. Mean, r.m.s., maximum and minimum pressure coefficients are evaluated from the measured fluctuating wind pressure records at all pressure point over an extended range of wind incidence angles namely 0° to 180° at an interval of 15°. The force coefficients are also evaluated using the load cell for different wind angles and these are compared with that obtained from the integration of the measured mean wind pressures on the different faces ofthe model. In the second phase of this study, the prototype rectangular buildings are analyzed using the wind loads obtained from the wind tunnel study for different wind incidence angles of 0° to 90° at an interval of 15°. The mean pressures obtained at different pressure points from rigid model testing are interpolated to obtain the wind pressures at each nodal point on the prototype buildings. The pressures on the buildings are evaluated according to the modeling law for the wind velocity of 58 m/sec (terrain category-II, IS-875) at the roof level of the building. The pressure at each nodal point is multiplied with the corresponding tributary area to obtain wind load at the node. Mean response of all prototype buildings having different side ratios are evaluated and their responses are compared. Apart from rectangular buildings, 'L' and 'T' plan shape buildings are also analysed under wind in the present study. 'L' and 'T' plan shape buildings are assumed to be made of two blocks (buildings) located in close proximity in a variety of geometric configurations and they are assumed to be separated by the construction joints. The dimensions ofthe blocks are assumed according to the dimensions of the corresponding models and geometrical scale. In L-shape buildings, block-1 is placed on upwind side at the edge of block-2, whereas block-1 is placed at the center of block-2 in case of T-shape buildings, Block-1 cause inference to block-2, resulting in different values of wind loads on block-2, as compared to isolated case. Evaluation of wind loads on buildings by structural designer is carried out mainly by using the codes and standards, whose specifications are generally based on wind tunnel tests performed on isolated structures in an open terrain. However, it has been shown by several researchers that when upstream building blocks another building, it increases or decreases the forces on the downstream building, depending mainly on the geometry and arrangement of these structures, their orientation with respect to the direction of flow and upstream terrain conditions. Therefore this effect, commonly known as interference, must be properly assessed by designers and planners. Thus in the present study, both the blocks of the L-shape and Tshape buildings are analyzed using the experimentally obtained wind loads for wind incidence angle of 0° to 90° at an interval of 15° and their mean responses are compared to assess the effectiveness of building shape in changing the mean responses of the buildings. Block-1 and block-2 of L-shape and T-shape buildings are also analyzed in an isolation conditions and their mean responses are evaluated. The mean responses of block-1 and block- 2 of L-shape and T-shape buildings are compared with similar buildings in isolated position and efforts are made to quantify the interference / shielding effects between the pairs of closely spaced buildings. This extensive study has yielded significant results. It is observed that the change in the plan-shape of buildings considerably affects the suction on side faces and leeward faces. From the variation of mean and r.m.s pressure coefficients on side faces of rectangular building models, it is observed that reattachment of flow occurs when side ratio of the model approaches to about 3, the final steady reattachment of flow take place at wind incidence angle of 0°. The experimental data for 'L' and 'T' shape building models show different wall pressure distributions from those expected for single and rectangular models. The wind pressure coefficients distribution on inner faces of 'L' and 'T' shape buildings largely depend on the dimensions ofthe buildings (re-entrant corner). It is observed from the analytical study that as the side ratio of the building increases, the maximum mean torque developed as a result of an uneven mean pressure distribution around the buildings walls also increases. This is created by flow separation points at corners vi around the building cross section at skew wind incidence angles. It is also observed that the responses of block-1 and block-2 of L-shape and T-shape buildings are significantly different than that of the isolated buildings. The arrangement of the buildings, their relative size, upstream terrain and the direction of the wind determine the extent of interaction between the pairs of buildings. The downstream building (block-2) of the T-shape building is subjected to significantly lower wind forces along X-axis as compared to the forces on corresponding block-2 of L-shape buildings. The downstream building (block-2) of T-shape buildings are, thus subjected to significantly higher torque as compared to the torque on corresponding block-2 of L-shape buildings and isolated building at skew wind incidence angles. The maximum mean torsional responses of block-2 of L-shape building and T-shape building increase by up to 35% and up to 180% respectively as compared to the maximum mean torsional responses of the similar block in isolated position. It is also noticed that presence of downstream building block-2 reduces the mean torque on block-1 of L-shape building up to 50%, and on block-1 of T-shape building up to 30%, with respect to the maximum mean torque developed in the isolated building. The downwindbuilding block-2 causesan increase in the mean displacement of block-1 of L and T-shape building along Z-axis by as much as 15%, withrespect to the maximum mean displacement of the isolated building. vn