Constitutive model update using observed field behavior

C. Marulanda, J. Butkovich, Y. M.A. Hashash, J. Ghaboussi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Numerical simulation of construction staging of deep urban excavations is commonly used to estimate induced ground deformations. A novel method to extract the constitutive behavior of soil directly from field measurements of excavation response is introduced. This method develops a neural network based constitutive model of the soil, given the lateral wall deformation and surface settlement profile measurements. This paper demonstrates that this approach is able to extract complex and realistic aspects of soil behavior. The soil model can continuously evolve using additional field information. New procedures are also presented for optimizing the extracted soil behavior used for training the neural network based soil model. These include data reduction and enrichment.

Original languageEnglish (US)
Title of host publicationNumerical Models in Geomechanics - 9th Proceedings of the International Symposium on Numerical Models in Geomechanics, NUMOG 2004
EditorsS. Pietruszczak, G.N. Pande
PublisherCRC Press/Balkema
Pages217-222
Number of pages6
ISBN (Print)9789058096364
StatePublished - 2004
Event9th International Symposium on Numerical Models in Geomechanics, NUMOG 2004 - Ottawa, Canada
Duration: Aug 25 2004Aug 27 2004

Publication series

NameNumerical Models in Geomechanics - 9th Proceedings of the International Symposium on Numerical Models in Geomechanics, NUMOG 2004

Other

Other9th International Symposium on Numerical Models in Geomechanics, NUMOG 2004
Country/TerritoryCanada
CityOttawa
Period8/25/048/27/04

ASJC Scopus subject areas

  • Soil Science
  • Numerical Analysis

Fingerprint

Dive into the research topics of 'Constitutive model update using observed field behavior'. Together they form a unique fingerprint.

Cite this