THE STUDY OF PRESTRESSED MASONRY FLEXURAL ELEMENTS

Abstract

The present research work has been carried out to investigate the complete behaviour of primary post-tensioned masonry flexural elements such as beams and one-way slabs both analytically and experimentally. The study would help in a better understanding of the behaviour of such members in pre and post cracking phases and lead to development of a rational basis for their analysis and design. The theoretical study consists of development of a plane-stress non-linear finite element model capable of predicting the complete response of the prestressed masonry planer structures and validation of the model on the basis of some experimetally tested beams and slab specimens. The experimental investigations consist of two parts, the first one relates to the determination of structural properties of masonry and prestressing steel used in the construction of presently tested prototype specimens, whereas the second part relates to the tests on a number of post-tensioned full scale beams and one-way slab specimens. Masonry control specimens of different shapes and sizes with varying brick arrangements have been tested to determine their strength and stress-strain characteristics in axial compression and in flexural tension as well as to determine the control specimen which correlates closely the masonry stress state in the test specimens. The stress-strain relationships of the two materials, (ii) thus obtained, have been idealised in the proposed mathematical model for use in the theoretical prediction of the behaviour of the prototype test specimens. As the second part of the experimental study, a total of 21 prestressed masonry flexural elements consisting of 16 rectangular beams, 2 double T-beams and 3 one-way slab specimens have been cast to full scale and tested upto failure. For rectangular beam specimens, three different type of cross section with different depth/width ratios of about 1.5, 2.0 and 3.3, suitable for small and .^ medium span beams, have been selected for investigation. The other parameters chosen for study of such members are the masonry strength (i.e. normal and medium strength masonry), the shape of prestressing cable profile (i.e. straight and parabolic) and the bond conditions between the prestressing cables and the surrounding masonry or grout. To facilitate the placement of prestressing cables, plain concrete cores have been provided in which separate P.V.C. ducts for each wire have been embeded. Reinforced concrete end blocks have also been designed to transfer the prestressing force safely and economically to masonry, thereby obviating the need for using the conventional steel end plates. The deflections of the beams at suitable points and strains in high tensile steel wire and on the masonry surface have been measured during the progress of the tests. Similarly in the series of tests conducted on one-way slabs and double T-beam specimens the only parameter chosen for the investigation has been the masonry strength and the bonding condition of the prestressing wires respectively. The other test (iii) conditions and instrumentation details for these tests have been similar to those of rectangular beams. All the prototype beam and slab specimens, tested experimentally, have been analysed theoretically using the presently developed finite element model. Using the constitutive relationships of the materials and suitably assumed failure criteria based on the experimental results, the theoretical values of the deflections and strains in steel and masonry have been predicted at different load levels. The experimentally obtained results and the predicted theoretical values at selected points have been compared at each load level. In almost all cases, the predicted and the experimental values are found to be in good agreement. The ultimate loads could also be predicted to a reasonable degree of accuracy. However, the predicted first cracking load obtained by theoretical analysis has been found to be always less than the load at the first visible cracking, obtained during the tests. The effective flange width of the double T-beam specimens could be predicted after comparing deflections and top fibre strains in masonry at mid-span section, obtained experimentally, with those obtained theoretically using different values of the effective flange width. Based upon the theoretical and experimental studies undertaken, the following conclusions have been drawn : (i) Experimental investigations clearly establish the feasibility of the use of prestressed masonry to cover large spans. Prestressing of masonry in flexure provides a crack-free (iv) construction within working loads. Also such constructions give enough warning before failure, (ii) The theoretical model using finite element technique has been develoed to predict the complete response of the planer prestresssed masonry structures. The proposed formulation incorporates effects of non-linearities caused by materials' non-linear stress-strain behaviour, cracking and crushing of masonry and yielding of steel. The employment of masonry super-elements makes such an analysis economically viable propositions for prediction of the behaviour at macro level, (iii) The proposed non-linear finite element model is suitable for predicting behaviour of both bonded and unbonded structures as it takes into account the appropriate bond-slip characteristics of prestressing strands with respect to the surrounding masonry/grout. (Iv) The evaluation of realistic material properties of masonry for the brick-mortar arrangement in actual construction is important for use in theoretical analysis. Realising this, three different type of control specimens have been evolved and tested for determining the characteristics of the compressive zone of masonry in flexure. The results have shown a wide variation in these characteristics as determined from the three different control specimens. Proposed control specimens best correlate the stress state of masonry in the test specimens and recommended for use. t (v) Based upon the limited expperimental and theoretical study of the double T-beams, an expression for the width of the contributing flange has been proposed. This would help in (v) simplified design of the T-beam slab contruction in masonry. Finally some suggestions have also been given for the purpose of further research in this area.