Investigation of reinforced concrete wall shear strength

Reinforced concrete shear walls have been widely used in buildings located in seismic regions due to their high rigidİty and ductility against wind loads and earthquake. Majority of thebuildings with reinforced shear walls that were not constructed based on recent seismic codes (e.g. Turkish Seismic Code 2007, ACI 318-14, ASCE 7, EuroCode 8 and Japanese Standard for Seismic Evaluation of Existing Reinforced Concrete Buildings 2001) haveproblems such asinadequate reinforcement and detailing, poor material quality. Post-earthquake observations have shown that such buildings are more likely to experience a greater degree of damage or even collapse because of considerable influences of insufficient reinforcementdetailing and inadequate material quality. Rehabilitation and retrofit of existing buildings has been vital and commonly used to minimize the risk of possible damage/collapse or potential post earthquake damages of such buildings.For better rehabilitation, it is necessary to understand the behavior of reinforced concreteshear walls. Analytical models that are representative ofexpected behavior of existing building stockcan lead an effective rehabilitation. Wall shear strength is one of the most important features that play an important role inidentifying actual response of shear walls. Therefore, analytical models should estimate shear strength as accurate as possible.Shear strength equations given in modern seismic codes are considered while calculating wall shear strength for both designing a new shear wall structure and investigation of existing structures. Wall shear strength equations provided byACI 318-14 and TSC- 2007 are relatively simple and includefewernumber of parameters, whereasthat of Japanese Standard for Seismic Evaluation of Existing Reinforced Concrete Buildings is more complicated and includes more number of parameters. According to TSC-2007 and ACI 318-14, nominal shear strength is directly proportional to horizontal web reinforcementmaterial properties of concrete, whereas ACI 318-14 equation considers influence of aspect ratio on shear strength as well. Compressive concrete strength was mentioned in TSC-2007 and ACI 318-14.Experimental studies have shown that various other parameters that were not mentioned in shear strength equations provided by recent seismic codes can affect wall shear strength. These studies focused on influence of axial load, vertical web reinforcement, and boundary confinement. Studies have also shown that experimental shear strengthgives much higher value than calculated shear strength that may cause non-economical rehabilitations and over-conservative designs. For this reason, alternative equations with more detailed expressions for shear strength were needed. This thesis aims to assess modern seismic code provisions and to investigate reliability and accuracy of shear strength equations given in modern seismic codes.Alternative shear strength equations were alse developed, using a detailed wall test database consisting of a large number of shear wall tests (a total of 172) conducted by various researchers around the world. The database included 35 shear- controlled wall specimens that failed bysliding shear, diagonal tension failure or web crushing,64 flexure-controlled wall specimens that failed by rebar buckling, flexural cracks or concrete spalling and 73 transition wall specimens that contain both failure modes,respectly. Previous research have shown that shear walls have different behaviour and responses depending on their properties, especially shear span ratio or aspect ratio. Specimens in the database were classified based on their reported failure modes and statistical studies were carried out. Detailed linear regression analyses were carried out to investigate influence of the key parameters and were conducted to derive best-fit empirical equaions for shear strength for each failure mode. Correlation coefficient ( ) shows the linear relationship between two variables (0 to 1.0). It becomes 1.0 for a parameter that correlates with shear stressexcellently and 0 for a parameter that is not corraleted with shear stress.Parameters that showed higher correlation with stress were compressive strength of concrete ( ) and horizontal web reinforcement ratio ( ) for the shear controlled walls, however they are not well-correlated with shear strenght for transition and flexure-controlled walls. For other walls (transition and flexure-controlled walls),both vertical web reinforcement ratio ( ) and horizontal web reinforcement ratio ( ) and showed higher correlation with shear stress. Mean values of the ratio of experimental strength to the estimated strength according to ACI 318-14 were, • Shear-controlled : 1.11 • Transition : 0.79 • Flexure-controlled : 0.66 Mean values of the ratio of experimental strength to the estimated strength according to TSC-2007 were, • Shear-controlled : 1.07 • Transition : 0.66 • Flexure-controlled : 0.56 Mean values of the ratio of experimental strength to the estimated strength according to JSC-2001 were, • Shear-controlled : 1.03 • Transition : 1.08 • Flexure-controlled : 0.91 Results of the analyses to determine reliability of estimated shear strength showed that equations provided by ACI 318, Turkish Seismic Code, and Japanese Seismic Code 2001 underestimate the shear strength of shear-controlled walls by 11%, 7%, and 3%, respectively. It is also notedthat equation provided by Turkish Seismic Code 2007 is not appropriate for non-rectangular shear walls. To obtain equations that calculate the shear strength closer to accurate, multi-linear regression analyses were conducted and alternative equations for shear strength were derived for the shear-controlled walls. Different combinations of various key parameters were tried until the best correlation between the equation and the test results wasobtained. These equations were found to capture expected response (shear strength) reasonably close to accurate, based on mean values of the test results. During this process, results of single parameter regression analysis were helpful to prioritize the parameters to be used in the equation. Equationsthat are physically the most meaningful with a correlation coefficient as high as possible were selected. These equationsaim to help the profession better assessbehavior and failure of shear walls and achieve more economical and reliable seismic rehabilitation.