Lateral pressures on rigid retaining walls: A neural network approach

Yildiz E., ÖZYAZICIOĞLU M. H., Yener Ozkan M.

Gazi University Journal of Science, vol.23, no.2, pp.201-210, 2010 (Scopus) identifier

  • Publication Type: Article / Article
  • Volume: 23 Issue: 2
  • Publication Date: 2010
  • Journal Name: Gazi University Journal of Science
  • Journal Indexes: Scopus
  • Page Numbers: pp.201-210
  • Keywords: Artificial neural networks, Hardening soil model, Lateral earth pressure, Strip load
  • Middle East Technical University Affiliated: Yes

Abstract

An artificial neural network solution (in closed form expression) is established for the total lateral thrust and its point of application on rigid retaining walls due to finite surface strip loads. The model accounts for soil nonlinearity and dilatancy. Data necessary for the model is produced through finite element analyses. The solution relates the total lateral thrust and its point of application to six parameters, including the strength parameters of the soil, wall height and the position as well as the extent of the surface load. The effects of each input parameter on the response are summarized and the results are compared with the linear elastic solution.