Developing artificial neural network models to predict allowable bearing capacity and elastic settlement of shallow foun

  • PDF / 2,809,226 Bytes
  • 11 Pages / 595.276 x 790.866 pts Page_size
  • 80 Downloads / 210 Views

DOWNLOAD

REPORT


ORIGINAL PAPER

Developing artificial neural network models to predict allowable bearing capacity and elastic settlement of shallow foundation in Sharjah, United Arab Emirates Maher Omar 1 & Khaled Hamad 1 & Mey Al Suwaidi 1 & Abdallah Shanableh 1 Received: 14 December 2016 / Accepted: 10 August 2018 # Saudi Society for Geosciences 2018

Abstract This research proposes the use of artificial neural network to predict the allowable bearing capacity and elastic settlement of shallow foundation on granular soils in Sharjah, United Arab Emirates. Data obtained from existing soil reports of 600 boreholes were used to train and validate the model. Three parameters (footing width, effective unit weight, and SPT blow count) are considered to have the most significant impact on the magnitude of allowable bearing capacity and elastic settlement of shallow foundations, and thus were used as the model inputs. Throughout the study, depth of footing was limited to 1.5 m below existing ground level and water table depth taken at the level of the footing. Performance comparison of the developed models (in terms of coefficient of determination, root mean square error, and mean absolute error) revealed that the developed artificial neural network models could be effectively used for predicting the allowable bearing capacity and elastic settlement. As such, the developed models can be used at the preliminary stage of estimating the allowable bearing capacity and settlements of shallow foundations on granular soils, instead of the conventional methods. Keywords Shallow foundations . Allowable bearing capacity . Elastic settlement . Artificial neural network . Granular soil . Sharjah

Introduction Allowable bearing capacity and settlement are the two main criteria that control the design practices of shallow foundations so that safety and serviceability requirements are realized. While assessing the soil, geotechnical engineers face a great amount of uncertainty exhibited in terms of soil variability, time effects, construction effects, human error and error in

* Maher Omar [email protected] Khaled Hamad [email protected] Mey Al Suwaidi [email protected] Abdallah Shanableh [email protected] 1

Department of Civil and Environmental Engineering, University of Sharjah, Sharjah, United Arab Emirates

soil boring, sampling, and laboratory testing. All these factors make artificial neural network (ANN) techniques more preferable as they are practical and less expensive than the conventional models. Though the ANN modeling approach offers a number of advantages, its main one is the ability to detect complex nonlinear relationships between the dependent and independent variables without having a priori assumption on the form or relationship between these variables. Furthermore, ANN could use raw input data without the need to manipulate or preprocess such data resulting in considerable time saving in any project. These advantages make ANN technique more practical and possibly less expensive than the conventional models. ANN mo