STUDY OF PRESTRESSED MASONRY RETAINING WALLS

Abstract

Although masonry has been a traditional material of construction in India for centuries, research effort to establish masonry construction on scientific principles has been minimal and only recent. Elsewhere too, masonry of late has been receiving increasing attention of researchers because of ever accelerating costs of cement and steel. One important ;lication of masonry in India is its possible use in the construction c nedium high retaining walls which form an integral part of the road net*- -k currently being constructed to open up far flung and remote areas fr~ -"-heir speedy development. The present work relates to finding possibility of the use of prestressed masonry to effect economy in the construction of masonry retaining walls. It is known that masonry is weak in tension and in structures like retaining walls which are subjected to significant lateral loads, it has to be of massive section to eliminate tensile stresses. This entails expensive construction. Ordinary reinforced masonry is now-a-days an accepted practice to carry flexural loads. Prestressing of masonry is only a recent innovation which considerably enhances the flexural capacity of masonry thereby resul ting in thinner sections and effecting economy in construction cost. There are several aspects which require considerable research effort in order to make prestressed masonry acceptable as a viable scientific mode of construction. The present study has been confined to investigation of the following areas : (i) Study of the creep characteristics of brickwork under sustained stress due to prestress; (iv) (ii) development of a suitable prestressing system requiring only a margi nally higher level of skill as is normally available for masonry cons truction . (iii) evolution of structurally efficient cross-sectional shapes for prestressed masonry retaining walls; and (iv) study of economic viability of prestressed masonry retaining walls. Experimental investigations have been carried out to study the creep characteristics of masonry. A significant finding of these investigations is that creep rate decreases rapidly after about 30 days of sustained loading and it becomes almost constant after 100 days of such loading. Further, it has been found that creep also depends upon the initial stress level (S. ) in masonry. As a result , Lenczner's model for creep ratio has been suitably modified to include the effect of stress level on creep. In addition, it has been possible to propose an expression to compute loss of prestress due to creep for any duration upto 100 days of sustained loading. Because of several factors affecting the quality of brickwork, the permissible stress in masonry is generally low. This allows lower levels of prestress in masonry compared to concrete construction. The magnitude of prestressing force required is therefore comparatively much smaller than for concrete. Considerable economy in prestressing operations has been achieved by developing prestressing technique which employs ordinary low capacity mechanical jacks and a steel stressing girder. The proposed method is very simple and the prestressing operations can be carried out even by unskilled labour. To evolve structurally efficient cross-sectional shapes, five prototype walls in the following four cross-sectional shapes have been constructed and tested : (v) (i) Diaphragm wall having cellular box shapes; (ii) Open box shape; (iii) Zigzag configuration; and (iv) A solid wall. The prototypes have been prestressed and tested under linearly varying lateral load to simulate earth pressure. Exact simulation of linearly varying lateral load is not possible in the laboratory especially because the load has also to be constantly increased during testing. The linear variation of the lateral load has been simulated in the present investigation by applying it in a stepped manner by four jacks each transferring its load to a vertical girder simply supported on two horizontal wooden rafters suitably spaced. The linearly varying load has thus been simulated through eight suitable concentrated loads applied laterally on the wall height. In case of the zigzag wall, special internal trusses have bee- impro vised to apply the load normal on the leaves of the wall. All walls have been tested only upto the appearence of the first crack. For theoretical prediction, finite element analysis has beer, carried out. Discretization of a masonry structure by individual brick ar.d mortar elements is too cumbersome and computationally beyond the capacity of most medium sized computers. A super element which is an assemblage of adequate number of bricks and horizontal and vertical mortar joints has beer, devised for finite element analysis in the present study. To determine the character istics of these super-elements, some equivalent masonry properties need to be determined to include the effect of brick-mortar joints. A finite element (vi) procedure used to determine these properties shows that the theoretically determined values are at much variance with the experimentally determined values. Suitable corrections to the theoretical values have been suggested to obviate the need for model tests for determining masonry properties for use in the formulation of super-element. Two-dimensional elemental program (TDEP) and flat shell elemental program (FESP1) have been used for the analysis of the walls as plane strain and shell type problems and a comparison has been made as regards the suitability of the two approaches for analysis of masonry retaining walls . The efficiency of the cross-sectional shapes has been ascertained from the following two criteria : (i) Factor of safety against first crack load; (ii) Material consumption. The economics of the prestressed masonry retaining walls has been studied by comparing their cost per metre width vis-a-vis the cost of masonry gravity type wall and the reinforced concrete wall of similar heights and subjected to same lateral loads. Some of the main conclusions drawn as a result of the present study are summarised below. Concept of super-element for the analysis of prototype masonry struc tures is valid yielding reasonably accurate results. Masonry properties needed in the formulation of the super-element can be predicted reasonably accurately by the proposed equations based on experimental investigations. (vii) Diaphragm and open box types of prestressed masonry walls exhibit effect of shear wall-cross wall interaction so that the ordinary flexural beam theory cannot be applied to correctly predict the behaviour of such walls. The effect of shear wall-cross wall interaction is more pronounced in the case of open box types of walls. The diaphragm walls have been analysed both as two-dimensional plane strain problems and treating them as flat shell structures. The results of the analysis by the two finite element formulations are almost similar. Masonry walls can thus be analysed as two-dimensional plane strain problems without much loss of accuracy. Diaphragm and open box types of prestressed masonry retaining walls are structurally more efficient as determined by their capability to resist lateral loads before the onset of cracking. Diaphragm type of prestressed masonry retaining wall is the most economical amongst the shapes analysed. As compared to reinforced concrete walls of similar height, it is nearly 40% cheaper. It is finally concluded that prestressed masonry retaining walls are structurally quite sound and construction of such walls can be safely re commended for the wall heights upto 4-5 metres.