EFFECT OF CAPILLARITY ON THE BEHAVIOUR OF TEST PLATES ON SANDS

Abstract

Capillarity is known to introduce apparent cohesion in sands resulting in a change in the engineering properties such as, shear strength and compressibility. As a result, the bearing capacity of a test plate resting on fine sand/silty fine sand beds is increased and the settlement of the plate, for a given load intensity, is reduced signi ficantly. However, no method is available to estimate, quantitatively, the effect of capillarity on bearing capacity and settlement of plates/ footings. The literature provides some evidence to show that plate load tests often get conducted on capillary beds either unintentionally or otherwise. In the absence of knowledge on the quantitative effect of capillarity, the results of a load test conducted on a capillary bed may lead to unsafe predictions of bearing capacity and settlement of prototype foundations . In a natural deposit, water is held in soil voids up to a certain height above the water table due to capillary action. This height, known as the height of capillarity, is reported to vary over a wide range, viz. between lm and 10 m in the case of silty fine sands, the type of soil generally encountered in this part of the country (Indo- Gangetic Plain). A more precise information on the height of capillarity is of importance to identify situations affected by capillary action. In view of above, the following investigations were planned. 1. Experimental investigation on the height of capillarity in silty fine sands . 2. Experimental and analytical investigations on the effect of capi llarity on bearing capacity and settlement of test plates. (iii) Capillary equilibrium in soils is achieved in two ways, viz. (i) by the rise of water from water table into the voids of dry soil mass and (ii) by retention of water in the soil voids when a soil mass, initially submerged, is drained to the water table. A suitable test set-up for simulating the above two processes has been designed and fabricated. The set-up consists of a PVC pipe column (150 mm dia and 3.0 m long) connected to a rectangular water tank at the bottom. The soil to be tested is placed inside the pipe column. The rectangular tank is a closed water tight container in which water can be maintained under pressure to enable raising and lowering of water level in the PVC pipe column. Using the set-up, the height of capillarity has been obtained for sands containing different percentages of fines (minus 75 micron). For all the test situations, the height of capillarity, hc, has been computed using the equation (h = C/e D,„, where C is a constant, e is void ratio, and D,„ is the effective grain size) repor ted in the literature. These results have been discussed and a possible range of height of capillarity in silty fine sands has been brought out. In order to study the effect of capillarity on bearing capacity and settlement, a number of load tests on 300 mm x 300 mm plates were carried out on prepared capillary sand beds. The test bed was prepared in a large testing tank of size 1.48 m x 1.48 m x 1.50 m. The tank is provided with two side tanks (0.50 m x 0.50 m x 1.50 m) which facilitate raising, lowering or maintaining a desired water level in the testing tank. The tank was filled with dry sand by rainfall method and the sand was saturated by raising water level in the testing tank to the top of the sand bed. After complete saturation, the water level was lowered and maintained at a desired elevation by manipulating the (iv) water level in the side tanks. Thus a capillary zone was created between the water table and the top of the sand bed. The test plate was then placed at the centre of the test bed and loaded to failure. The loading was carried out by jacking the plate against a load frame built around the testing tank. The tests have been carried out on fine sands and silty fine sands varying the thickness of capillary zone. Load tests have also been carried out on dry and submerged sand beds for comparison. The results of all the load tests conducted have been presented in the form of load settlement curves. The results have been discussed to bring out (i) the increase in bearing capacity due to capillarity, (ii) the effect of thickness of capillary zone on the magnitude of bearing capacity and settlement and (iii) the increase in settlement of a loaded test plate due to rise of water table. An attempt has been made to quantify, analytically, the effect of capillarity on bearing capacity of plates/footings. A failure wedge similar to that assumed in Terzaghi's analysis has been considered. An expression for bearing capacity has been derived considering the equilibrium of the failure wedge under the action of various forces inclu ding those introduced due to capillary action. The analysis has been programmed and the bearing capacity factors as applicable to the capillary situation have been obtained. Based on the analysis, the ultimate bearing capacity of test plates on capillary sand beds has been obtained and compared with the results of load tests conducted in the laboratory. The results of a few field load tests which appear to have been carried out on a capillary bed and reported in the literature have been compiled and compared with those obtained using the present analytical approach. It has been noted that the bearing capacity values obtained analytically (v) and experimentally agree reasonably well suggesting that the increase in bearing capacity due to capillarity can be quantified by the present analytical solution. A foundation is often proportioned such that its settlement does not exceed the permissible settlement when water table rises to the base or above the base of foundation. Therefore, one is interested in the settlement of a test plate resting on a submerged bed. However, it is not always possible to conduct a load test at the position of water table as it might necessitate large excavation and quite often the load test gets conducted on a capillary zone. Therefore, an approach to predict the settlement of a plate resting on submerged bed from the results of load tests conducted on capillary bed is needed. Assuming that the settlement is inversely proportional to the confining pressure, an expression for the ratio of settlement of a plate on submerged bed, Ssub' to that of the Plate on capillary bed, S , has been developed. P The ratio, ssut)/scap» has also been obtained from the results of load tests on 300 mm x 300 mm plate conducted in the laboratory. The values obtained analytically have been found to agree well with those obtained experimentally . Based on the investigations carried out the following significant conclusions have been drawn : 1. The height of capillarity in silty fine sands, the type of soil generally encountered in Indo-Gangetic Plain, can be of the order of 1 m to 2 m. (vi) 2. The ultimate bearing capacity of a plate resting on capillary bed, depending upon the thickness of capillary zone, can be three to six times higher than of a plate resting on a submerged bed. 3. In the case of a plate resting on a capillary bed and loaded to a given pressure intensity, the rise of water table to the base of the plate increases the settlement by many times, the order of increase being dependent upon the thickness of capillary zone. For example, the increase is about four times when the thickness of capillary zone is about 1.0 m. 4. The analytical equation for the bearing capacity derived in this investigation closely predicts the bearing capacity of plates resting on capillary beds. 5. An approach has been suggested for the estimation of settlement of shallow foundations from the results of load tests conducted on capillary bed. 6.. On the basis of the analytical and experimental evidence presented in this thesis, it is now possible to use the results of a load test conducted on a capillary bed in a meaningful manner for pro portioning of foundations.