UWA Logo What's On at UWA
   UWA HomeProspective Students  | Current Students  | Staff  | Alumni  | Visitors  | About  |     Search UWA    for      
 

SEMINAR: Evaluating soil parameters using numerical optimization - Civil and Resource Engineering Seminar

* Login to add events... *
Today's date is Thursday, March 28, 2024
Evaluating soil parameters using numerical optimization - Civil and Resource Engineering Seminar Other events...
Abstract: Evaluating parameters for complex constitutive models can be a laborious and time consuming task, particularly where data from a number of different tests and test types are available.

This paper explores the use of mathematical optimisation techniques to derive parameters for a given constitutive model based on the stress-strain/load-displacement response measured in both laboratory and in-situ tests.

An example is presented involving triaxial, plate-load and pressuremeter tests conducted at different elevations in a clay deposit for which the strength and stiffness increase (non-linearly) with depth.

Finite element models representing each test are used to calculate the constitutive model’s response with a given set of parameters. Individual norms, which measure differences between the actual and calculated load-displacement response, are then calculated for each test.

Individual norms for each test are combined to form a global norm, or objective function, for all tests. Direct search methods are then employed to vary the constitutive model parameters with the aim of minimising the objective function.

Methods for forming both individual norms and objective functions are discussed and the performance of a commercially available search method is assessed.

It is concluded that mathematical optimisation can assist in the practical application of constitutive models, as it promotes a systematic treatment of data and removes much of the subjectivity involved in parameter selection.

It also provides a rational basis for extrapolating from the known test data to predict the unknown field response. This is because precisely the same assumptions are made when interpreting soil test data as those that are made when undertaking calculations to predict the performance of real structures (i.e. the assumption in the constitutive model).

Speaker(s) Dr James Doherty
Location Room 1.51,1st floor, Civil and Mechanical Engineering building, Fairway Entrance 3
Contact Dr James Doherty <[email protected]> : 3991
Start Fri, 25 Sep 2009 15:45
End Fri, 25 Sep 2009 16:45
Submitted by Magdalena Matuszczyk <[email protected]>
Last Updated Fri, 04 Sep 2009 11:54
Included in the following Calendars:
Additional Information:
  • Locations of venues on the Crawley and Nedlands campuses are available via the Campus Maps website.
  • Download this event as: Text | iCalendar
  • Mail this event:

Top of Page
© 2001-2010  The University of Western Australia
Questions? Mail [email protected]