Modern Statistics, Data Analysis, and Specimen/Structural Reliability Modeling
October 16, 2011, 9 a.m. to 5 p.m.
Instructors: Jeffrey T. Fong, P.E., National Institute of Standards & Technology and Stephen W. Freiman, Freiman Consulting
Location: In conjunction with MS&T’11 hosted in Columbus, Ohio
ACerS Member – $620
Student – $250
Nonmember – $710
Course plus Membership – $740
This course will be divided into three parts.
Part 1, 9 to 11:50 a.m. will review key concepts in statistical data analysis as applied to the following four specific tasks:
- Task (A): A distributional fit of a small sample of specimen test data.
- Task (B): A battery of tests for goodness-of-fit of the assumed distribution of the specimen test data.
- Task (C): A confidence interval and a predictive interval estimate of the parameters of the fitted distribution of the specimen test data.
- Task (D): A tolerance limit methodology for extrapolating specimen test data information to the full-scale level of an engineering component.
A major component of this part is to introduce (1) DATAPLOT: A public-domain English-command-based statistical data analysis software package, and (2) a downloadable and free .pdf-formatted ebook entitled “Engineering Statistical Handbook.” With those two powerful tools, the review will examine key data analysis concepts and illustrate their applications by running DATAPLOT codes. Installation of the DATAPLOT software in a participant’s personal computer is optional, but highly encouraged with on-site hands-on assistance by the instructor.
Part 2, 1:30 to 3:20 p.m. will provide the fracture mechanics background and measurement protocols needed to assess the mechanical reliability of glasses and ceramics. A focal point will be a detailed discussion of a case study of the estimation of the time-to-failure of full-scale aircraft glass windows using three specific sets of specimen test data generated in a laboratory. After using DATAPLOT to execute Tasks (A), (B), and (C) for each set of data, we apply the error propagation theory to arrive at an estimation of the specimen time-to-failure. We then use the tolerance limit methodology to extrapolate the specimen-level information to the full-scale structural product level.
Part 3, 3:30 to 4:30 p.m. will be a live demonstration of the features, capabilities, and user-friendliness of the DATAPLOT code. The purpose of this part is to acquaint each participant with the power, versatility, and easy-to-learn feature of the DATAPLOT code such that each participant will be able to hit the ground running immediately after taking this course. Participants are encouraged to bring their own data for analysis.
A participant will develop the ability to use modern statistical analysis tools to analyze experimental data sets for predicting failure of brittle materials, including the ability to predict the component failure based on laboratory data. They will learn how to use a publically available software system without knowing the intricacies of advanced statistical calculations. It’s analogous to learning how to drive a car without understanding how a motor runs inside the car. Details on required laboratory test procedures needed to minimize experimental uncertainties will be provided. Specifically they will:
- Use an English-based statistical data analysis code to represent sample test data with critical tests to check whether the assumed distribution is statistically “correct.”
- Use the theory of error propagation to estimate the confidence interval of time to failure of a class of materials with a formula to work with.
- Use a large set of example codes to learn how to deal with a student’s own data sets.
- Students are encouraged to bring a laptop to be able to download programs from a CD ROM which will be supplied by the instructors.
Jeffrey Fong was educated at the University of Hong Kong (B.Sc., engineering), Columbia University (M.S., engineering mechanics), and Stanford University (Ph.D., applied mechanics and mathematics). He had ten years of design and construction experience in the electric power generation industry, and more than 40 years with the federal government as a researcher and consultant on mathematical modeling using “modern statistics.” Fong is a registered professional engineer in the State of New York (Lic. No. 039467-1), Fellow of the American Society of Mechanical Engineers (ASME), Fellow of ASTM International, and the recipient of an ASME Pressure Vessel and Piping Medal (1993). Since 2006, he has taught a graduate-level 3-credit course at Drexel University on “Finite Element Method Uncertainty Analysis,” and an on-line 3-hour short course at Stanford University on “Engineering Reliability and Uncertainty Estimation.”
Stephen Freiman graduated from the Georgia Institute of Technology with a B. ChE. and a M. S. in Metallurgy. After receiving a Ph.D. in Materials Science and Engineering from the University of Florida in 1968, Freiman worked at the IIT Research Institute and the Naval Research Laboratory. He joined NIST (then NBS) in 1978. From 1992 to 2002 Freiman served as Chief of the Ceramics Division at NIST, overseeing programs in ceramic processing and properties. Prior to his leaving NIST in 2006 to start a consulting business (Freiman Consulting Inc.), Freiman served for four years as Deputy Director of the Materials Science and Engineering Laboratory. Freiman has published over 200 scientific papers focusing primarily on the mechanical properties of brittle materials. He was the first Chairman of the ASTM Subcommittee addressing brittle fracture and a past Chair of the Steering Committee of the Versailles Project for Advanced Materials and Standards. Freiman served as Treasurer, and President of The American Ceramic Society, and is a Fellow and Distinguished Life Member of the Society.
ACerS reserves the right to cancel a course up to 4 weeks before the scheduled presentation date. Please contact ACerS customer service at 1-866-721-3322 or 1-240-646-7054 to confirm that the course is happening before purchasing your airline tickets. ACerS is not responsible for the purchase of non-refundable airline tickets or the cancellation/change fees associated with canceling a flight.