STATIC AND DYNAMIC BEHAVIOUR OF PREFABRICATED JOINTED PANELS

Abstract

In prefabricated jointed panel construction system, walls are assembled out of storey high and room size precast wall panels. The prefabricated wall panels are erected in position one by the side of other and the vertical joints are filled with in-situ concrete. Precast floor panels are then laid over the wall and the erection of next storey wall panels is completed. The horizontal joints at the floor level between the wall and floor panels are again filled with in-situ concrete. Rigidity of the joints is provided by shear keys and the transverse reinforcement projecting from the precast walls and floor panels. Conventionally, the analysis of such assembled wall panels is made on the assumption of rigid joint, which is far from correct. The infill concrete in the joints in reality acts as precracked plane and is thus much more deformable. The joint thus forms shearing surface. Hence, the assumption of such wall assembly as a monolithic wall is far from satisfactory and calls for methods of analysis incorporating the deformability characteristics of joint. Jointed panel construction system is basically a three dimensional problem due to interaction of the floor slabs and longitudinal walls with the shear wall. However, the three dimensional problem may be reduced to a two dimensional problem with some simplifying assumptions regarding the actions of the floor slabs and longitudinal walls. Longitudinal walls and floor slabs may be modelled with reasonable accuracy by suitable modification of the plane stress elements where the top or bottom surface nodal variables are expressed in incremental form across the thickness, thus eliminating the ill conditioning of slender plane stress elements due to the presence of large off diagonal terms associated with adjacent nodes across the thickness. Three different incremental ele ments have been proposed for the longitudinal walls and floor slabs and finite element formulations presented. Planar prefabricated shear wall (P.S. wall) has been discretized into finite elements by using plane stress elements for wall panels and interfacial elements for joints* A brief discussion on the elastic and elasto-plastic plane stress elements and linear and nonlinear interfacial elements alongwith various solu tion techniques for linear and nonlinear material beha viour and step by step time integration techniques.for dynamic load has been presented. Later, in similitude to shear continuum approach for shear wall, partially bonded beam element has been developed with inter-layer slip as extra degrees of freedom. The partially bonded concept reduces the two dimensional plane stress problem to a simple one dimensional beam problem, thus cutting down the cost and effort for analysis of P.S» wall considerably. The most important and yet very little known para meters in a jointed panel system are the shear capacity and shear stiffness characteristic of the joint. In absence of suitable test data for the Indian condition, 29 joint specimens have been tested in shear. The para meters studied are strength of infill concrete, reinforce ment and area of shear keys in the joint. Two formulae have been derived for predicting the failure load of joints based on the shear friction hypothesis and the more rational Mohr-Coulomb's failure criterion. Relation ships for shear stiffness of joint with slip deformation for different percentage of reinforcement and strength of infill concrete have also been derived. Also, suitable suggestions have been made for better design of such joints. Using the numerical models of plane stress and interfacial elements and the partially bonded beam elements, both linear and nonlinear studies of P-S. walls have been made showing close correlation of the results. From the linear elastic study of P.S. wall, increase in the flexibility due to low values of shear stiffness, more number of joints, low aspect ratio of F.S. wall, flexible foundation and deformable horizontal joints has been established. Based on the parametric study by partially bonded beam elements, some design curves have also been provided for symmetrical two and three panel P.S. walls. Also, the behaviour of P.S. wall under self weight and the effect of separation in the joint due to development of tension in the joint have been studied. Nonlinear elasto-plastic response of P.S. wall under monotonically increasing static load indicates that yielding first starts at the base of the windward wall panel and gradually spreads towards the leeward panel. It is at this stage that failure occurs. Large enhancement in the failure load due to precompress!on in the wall from self weight has also been established. Analysis of P.S. wall by partially bonded beam element, wherein nonlinearity has been assumed to be limited in the joint only, also predicts the failure of joint close to that obtained from elasto-plastic analysis. Free vibration frequency characteristics of P.S. walls have been studied and elongation of time periods for low shear stiffness of joints, more number of joints, soil structure interaction effect and deformable horizon tal joints have been observed. The importance of verti cal and rotary inertia has also been highlighted. Finally, the linear and nonlinear response of a 12 storeyed P.S. wall under two strong motion earthquakes has been studied by implicit time integration technique, which is most suitable for such jointed system. The step by step results have then been compared with those obtained from modal analysis by average response spectrum method as given in the I.S. Code and suitable recommendations have been made.