Sunday, July 27, 2003
1 - 6 PM
Albuquerque Doubletree
Course
Description
This
short course will summarize the terminology, methodology, and procedures
for verification and validation in computational mechanics. The subject
material will be generally applicable to all areas of continuum mechanics,
but the emphasis will be on fluid dynamics and solid dynamics. Terminology
to be discussed includes: definitions of verification and validation
accepted in engineering, definitions of verification and validation
used in other fields, certification, accreditation, prediction, calibration,
software quality assurance, and uncertainty and error in computations
and experiments. Verification
assessment procedures for finite element, finite difference, and finite
volume methods will be summarized, such as, grid and time-step convergence,
iterative convergence, the method of manufactured solutions, and the
use of analytical and benchmark numerical solutions. Validation assessment
procedures will be summarized, specifically the use of hierarchical
experiments in validation (unit problems, benchmark cases, subsystem
cases, and
the complete system). An example will be given showing how a complex
engineering system with coupled physics is separated into each of
the hierarchical levels for validation. Six key characteristics will
be discussed for the design and execution of high quality validation
experiments. Quantification of comparisons of computation and experiment
is introduced through the use of validation metrics. The tutorial
will close with a discussion of how a code validation database is
related to uncertainty estimation in code predictions.
The tutorial
is appropriate for both computationalists and experimentalists working
in confidence assessment in computational mechanics. Managers who
are responsible for the reliability and accuracy of analyses in computational
mechanics should also attend. Most topics are discussed at the conceptual
level so that detailed knowledge of numerical methods or experimental
techniques are not required.
Registration
Fee $200
The registration
fee includes the following:
Copy of the text
book Verification and Validation in Computational Science
and Engineering (1998) by Patrick Roache
Table of Contents [an $85 value]
A copy
of Guide for the Verification and Validation of Computational
Fluid Dynamics Simulations
AIAA-G-077-1998 [a $25 value]
All of the viewgraphs
presented at the Short Course and
a refreshment break.
Note: This Short Course sold out at the 2001 Congress in
Dearborn, so register early. You may register for the Short Course
without registering for the Congress.
Registration for this
Short Course will be available in January 2003.
Instructor
Dr. William
Oberkampf has 32 years of research, development, and applications
experience in both computational and experimental fluid dynamics.
He has worked in the field of verification and validation for the
last 14 years and co-authored the first engineering standards document
on the subject: "Guide for the Verification and Validation of Computational
Fluid Dynamics Simulations" published by the American Institute of
Aeronautics and Astronautics. He is one of the founding members of
the new ASME Codes and Standards Committee for Verification and Validation
in Computational Solid Mechanics. He is presently a Distinguished
Member of the Technical Staff at Sandia National Laboratories.
Dr. William
Oberkampf
Validation and Uncertainty
Estimation Dept.
Sandia National Laboratories
Email: wloberk@sandia.gov
V&V Short Course Coordinator Len Schwer
707-837-0559
928-833-1130 (eFAX) Len@Schwer.net
This
short course will summarize the terminology, methodology, and procedures
for verification and validation in computational mechanics. The subject
material will be generally applicable to all areas of continuum mechanics,
but the emphasis will be on fluid dynamics and solid dynamics. Terminology
to be discussed includes: definitions of verification and validation
accepted in engineering, definitions of verification and validation
used in other fields, certification, accreditation, prediction, calibration,
software quality assurance, and uncertainty and error in computations
and experiments. Verification
assessment procedures for finite element, finite difference, and finite
volume methods will be summarized, such as, grid and time-step convergence,
iterative convergence, the method of manufactured solutions, and the
use of analytical and benchmark numerical solutions. Validation assessment
procedures will be summarized, specifically the use of hierarchical
experiments in validation (unit problems, benchmark cases, subsystem
cases, 
and
the complete system). An example will be given showing how a complex
engineering system with coupled physics is separated into each of
the hierarchical levels for validation. Six key characteristics will
be discussed for the design and execution of high quality validation
experiments. Quantification of comparisons of computation and experiment
is introduced through the use of validation metrics. The tutorial
will close with a discussion of how a code validation database is
related to uncertainty estimation in code predictions.