Course 105:

Bayesian Uncertainty Quantification:

Theory, Computational Tools, and Applications
Sunday, February 1, 2015 | 8:00 a.m.6:00 p.m.

Course Description Course Fee

In simulations of complex physical systems, uncertainties arise The regular course fee for Bayesian Uncertainty
from imperfections in the mathematical models introduced to Quantification: Theory, Computational Tools,
represent the systems and their interactions with the environment. and Applications is $400, and the student
Such uncertainties lead to significant uncertainties in the course fee is $200. Course fee includes lunch,
predictions using simulations. Since such predictions form the course handout material, and refreshment
basis for making decisions, the knowledge of these uncertainties breaks. Lodging and additional food or
is very important. The course will present the Bayesian framework, materials are not included.
the associated computational tools, selected applications, along
with the main challenges for quantifying and propagating Instructors
uncertainties in complex structural dynamic simulations. Professor Costas Papadimitriou
University of Thessaly, Greece
Course Outline Papadimitriou (PhD CalTech) is Professor of
Bayesian uncertainty quantification and Structural Dynamics at the University of Thessaly
propagation in structural dynamics simulations (Greece). He holds the position of the Executive
Bayesian model parameter estimation / model updating Vice-President of the European Association
Bayesian model class selection of Structural Dynamics (EASD). He has over 25
Updating robust predictions and robust reliability years of experience in the areas of Bayesian
Structural health monitoring using Bayesian uncertainty quantification and propagation,
model selection and updating computational structural dynamics, finite
Bayesian computational tools element model validation, structural health
Asymptotic approximations monitoring and structural reliability. He has
Sampling techniques co-authored over 180 papers in journals and
Case studies conference proceedings and co-edited two
Full-scale seven-story shear wall special journal issues on the subject and the
building tested on shake table section on Structural Health Monitoring in the
Dowling Hall Footbridge Encyclopedia of Earthquake Engineering. He
Metsovo Bridge has organized more than twenty minisymposia
Small-scale laboratory vehicle model on the subject and has given invited/keynote
with nonlinear suspensions lectures in international conferences. He chairs
High performance computing for the Dynamics committee of ASCE-EMI and
Bayesian UQ of complex models the Identification, Model Updating and Inverse
Component mode synthesis Problems committee of EASD.
Surrogate techniques (kriging, polynomial chaos)
Parallel computing Professor Babak Moaveni
Demonstration on high fidelity linear/ Tufts University
nonlinear bridge models Dr. Moaveni is currently an Assistant Professor
Optimal experimental design at the Department of Civil and Environmental
Information entropy Engineering at Tufts University. He obtained
Optimal sensor placement his Ph.D. in structural Engineering at
Optimal excitation characteristics University of California, San Diego in 2007.
Dr. Moavenis main research interests include
Who Should Attend vibration-based damage identification and
Engineers, researchers and graduate students who deal with finite structural health monitoring of real-world
element model validation as well as uncertainty quantification and structures; dynamic field testing; finite
propagation in structural dynamics simulations using vibration element model updating; and uncertainty
measurements. quantification in structural dynamics. 9

Held in conjunction with IMAC-XXXIII organized by the Society for Experimental Mechanics, Inc.
More information can be found at http://sem.org
IMAC-XXXIII | February 25, 2015 | Rosen Plaza Hotel, 9700 International Drive, Orlando, FL 32819 USA | rosenplaza.com