Abstract
The paper has a two-fold purpose: 1) to give a brief overview of recent works at Northwestern University, and 2) to present a new application of the latin hypercube sampling to Bayesian prediction of creep and shrinkage effects. First a probabilistic prediction model in which the creep and shrinkage structural effects are considered as functions of a certain number of random parameters is reviewed, and determination of the statistics of these effects by latin hypercube sampling of the random parameters is outlined. Subsequently it is shown how this sampling approach can be combined with the short-time data for a given particular concrete to predict long time effects. Results of a few typical numerical examples accompany the presentation.
Original language | English (US) |
---|---|
Pages | 331-344 |
Number of pages | 14 |
State | Published - 1985 |
ASJC Scopus subject areas
- General Engineering