NTRS - NASA Technical Reports Server

Back to Results
Global/local methods for probabilistic structural analysisA probabilistic global/local method is proposed to reduce the computational requirements of probabilistic structural analysis. A coarser global model is used for most of the computations with a local more refined model used only at key probabilistic conditions. The global model is used to establish the cumulative distribution function (cdf) and the Most Probable Point (MPP). The local model then uses the predicted MPP to adjust the cdf value. The global/local method is used within the advanced mean value probabilistic algorithm. The local model can be more refined with respect to the g1obal model in terms of finer mesh, smaller time step, tighter tolerances, etc. and can be used with linear or nonlinear models. The basis for this approach is described in terms of the correlation between the global and local models which can be estimated from the global and local MPPs. A numerical example is presented using the NESSUS probabilistic structural analysis program with the finite element method used for the structural modeling. The results clearly indicate a significant computer savings with minimal loss in accuracy.
Document ID
Acquisition Source
Legacy CDMS
Document Type
Conference Paper
Millwater, H. R.
(NASA Lewis Research Center Cleveland, OH, United States)
Wu, Y.-T.
(Southwest Research Inst. San Antonio, TX, United States)
Date Acquired
August 16, 2013
Publication Date
January 1, 1993
Publication Information
Publication: In: AIAA(ASME)ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, 34th and AIAA/ASME Adaptive Structures Forum, La Jolla, CA, Apr. 19-22, 1993, Technical Papers. Pt. 2 (A93-33876 1
Publisher: American Institute of Aeronautics and Astronautics
Subject Category
Structural Mechanics
Report/Patent Number
AIAA PAPER 93-1378
Accession Number
Funding Number(s)
Distribution Limits

Available Downloads

There are no available downloads for this record.
No Preview Available