UTARI Seminar is held the last Friday of each month at 12:00PM (noon). Each seminar highlights a different speaker who will discuss their latest research projects, cutting-edge technology or what is happening within certain technological industries. These industries include biomedical technologies or microsystems, assistive technologies, automation and intelligent systems, unmanned systems, advanced manufacturing and composite materials.
Speaker
Jeffrey T. Fong, P.E., Ph.D.
Physicist and Project Manager Applied & Computational Mathematics Division National Institute of Standards & Technology (NIST)
Topic
Abstract
In engineering, from design (by analysis or formulas), manufacturing, construction, testing, to operation and maintenance, decisions are made by a combination of facts, analysis, and experience. Because these decisions are made under the constraints of cost and time, uncertainty is always present due to sources such as measurement inaccuracies, material properties, sampling errors, boundary and initial conditions, numerical approximations, physical assumptions of models, and sometimes the lack of a validated model. To ensure that these decisions are “correctly” made with “quantifiable” confidence (“high”) and risk (“low”), we need to measure, understand, and manage uncertainty at all aspects of our engineering decision-making process. In this talk, we present six easy-to use tools to quantify uncertainty:
- ( Tool-1 ) A multiple-goodness-of-fit (MGF) tool to fit and rank a sample of data to 64 continuous distributions
- ( Tool-2 ) A linear least squares (LLSQ) fit tool with predictive limits.
- ( Tool-3 ) A multi-scale modeling (MSM) tool for estimating tolerance limits from predictive limits.
- ( Tool-4 ) A fractional factorial design of experiments (FFDEX) tool to rank the importance of factors and to estimate uncertainty using Tool-2 .
- ( Tool-5 ) An expert knowledge elicitation (EKE) tool to obtain an estimate of the median and upper/lower bounds of a quantity of interest.
- ( Tool-6 ) A logistic-function-based nonlinear least squares (NLLSQ) fit tool to obtain asymptotic solutions with predictive limits.
To illustrate their power and versatility, four examples of application of three or more of the tools listed above will be given as shown below:
- (Example-1) Finite Element Method Accuracy Assessment ( Tools 2, 4, 6 ).
- (Example-2) Nondestructive Examination (NDE) Uncertainty ( Tools 2, 3, 4, 5, ).
- (Example-3) Design of an aircraft glass window ( Tools 1, 2, 3, 6 ).
- (Example-4) A Multi-Scale Creep Rupture Time-Life Model ( Tools 2, 3, 6 ).
Bio
Dr. Fong was educated at the University of Hong Kong (B.Sc. Eng.), Columbia Univ. (M.S. Eng. Mech.), and Stanford (Ph.D., Appl. Mech. And Math.). He worked as a design engineer (Ebasco, 1955-63) and research assistant (Stanford, 1963-66) before joining NIST in 1966, and has been with NIST as a physicist and project manager for more than 50 years. A Fellow of ASME and ASTM, he has published more than 150 papers and edited or co-edited 17 conference proceedings on fatigue, fracture, creep, NDE, and uncertainty in modeling.
Date
05/04/18
Time
12pm (noon)-1pm
Location
7300 Jack Newell Boulevard South
Fort Worth, TX 76118-7115
817-272-5900
utari.uta.edu