ORIGINAL PAPER Estimation of Ultimate Bearing Capacity of Footings on Soft Clay from Plate Load Test Data Considering Variability Parbin Sultana 1 • Ashim Kanti Dey 1 Received: 23 January 2018 / Accepted: 2 May 2018 Ó Indian Geotechnical Society 2018 Abstract The ultimate bearing capacity of a foundation can be determined from the load–settlement response obtained from plate load tests. For clayey soil it is known that the ultimate bearing capacity of footing is equal to that of a test plate. However, the present study shows that it varies with the variation of type of loading, size and shape of test plate and soil consistency. For reaction type of loading, the load is observed as normally distributed ran- dom variable with an average co-efficient of variation of 0.018. For different shape and size of plates, the bearing pressure–settlement curves show a high degree of varia- tion. Subsequently the bearing pressures and settlements are normalized to dimensionless parameters and the nor- malized plot shows a reduction in variability. Further, the normalized curves are fitted to the hyperbolic model by nonlinear regression with non-constant variance. The model parameters, their standard errors, 95% confidence intervals and co-efficient of determination are quantified. The horizontal asymptote of the hyperbolic model has been used to estimate the ultimate bearing capacity. This method yields a value very close to other conventional bearing capacity methods. A sensitivity analysis shows the influ- ence of various parameters on the bearing pressure. Keywords Soft clay Á Uncertainty Á Normalized load–settlement curve Á Hyperbolic model Á Non linear regression Introduction Highly compressible inorganic clays (CH) are found at different parts of the world [1, 2] and they pose two major problems, namely high compressibility and low bearing capacity. Since availability of good soil is scarce now, constructions are being taken up in marshy land. Some ground improvement techniques are applied to these soils in case replacement is not possible. In order to obtain the improved bearing capacity of a soft clayey soil the original bearing capacity is required. It is very difficult to correctly predict the bearing capacity due to inconveniences in sampling operations [3]. Kempfert and Gebreselassie [4] discussed the various aspects of difficulties while dealing with the soft clays. During the sampling process with a thin walled tube sampler, soil de-structuring along with sig- nificant reduction in effective stress due to the release of in situ stresses affect the soil sample [5]. As a result, the sample does not represent actual field conditions. This uncertainty in sampling contributes to the uncertainties in laboratory test results resulting in an inaccurate bearing capacity determination. For this reason, in situ soil tests are better relied upon than laboratory tests for the soft clays. Two/three decades ago, plate load tests (PLT), field or laboratory, were extensively used for determination of in situ properties of clayey soils instead of small scale, non- representative laboratory tests [6]. Load tests were per- formed on soft clays to observe the lateral load–deflection of piles [7], uplift capacity of granular pile anchor [8], load carrying capacity of micropiles [9], bearing capacity of stone columns [10–12], bearing capacity of shallow foot- ings [13, 14] etc. The major disadvantage of the PLT was cost involvement, hence the researchers gradually shifted to more improved laboratory tests and numerical analyses. & Ashim Kanti Dey ashim_kanti@yahoo.co.in Parbin Sultana parbinsultana@rediffmail.com 1 Department of Civil Engineering, NIT Silchar, Silchar, Assam 788010, India 123 Indian Geotech J https://doi.org/10.1007/s40098-018-0311-9