• PDF / 492,214 Bytes
• 7 Pages / 595.276 x 790.866 pts Page_size
• 98 Downloads / 303 Views

DOWNLOAD

REPORT

---

TECHNICAL ARTICLE—PEER-REVIEWED

Classical Fatigue Design Techniques as a Failure Analysis Tool Dean Harris Æ T. Jur

Submitted: 7 November 2008 / Accepted: 18 November 2008 / Published online: 6 January 2009 Ó ASM International 2008

Abstract The classical method for designing against high-cycle fatigue fracture is based primarily on statistical models derived from laboratory experimental data. This paper considers a number of actual fatigue failures where the analyses of the failures, in part, made use of classical high-cycle fatigue resistance design methodology as an analytical tool. This paper uses failure analyses to demonstrate that the long-taught classical methodology is useful and accurate as both a design and an analysis tool. The usefulness and accuracy of the method is verified in that it is shown to have predicted actual failures, given known materials, manufacturing histories, and service operating conditions. Example analyses include: a fatiguecracked roll from a paper-making machine, a fractured anvil on a steam powered forge, and a fractured shaft on a helical ribbon dryer. Keywords

Fatigue
Fracture
High-Cycle Fatigue

Introduction Classical fatigue methods have typically been taught at the introductory level in undergraduate mechanical design classes and in metallurgical/materials classes and have also been expanded upon in graduate level design and failure analysis classes. The mathematics are relatively simple, the methodology is straightforward, and a wealth of design data exists in the published literature. However, the proof of a design or a design method is how well it models the

D. Harris
T. Jur (&) Engineering Design & Testing Corp., Columbia, SC, USA e-mail: [email protected]

actual in-service behavior of manufactured parts. As consultants in failure analysis for a combined five decades, the authors have had many opportunities to examine actual fatigue fracture situations and to analyze the fractures in accordance with classical fatigue design methods. In every case in which sufficient data were available to conduct a load/life fatigue design analysis, the classical methods predicted the fracture. This paper presents an overview of classical fatigue design methods and then discusses the application of those methods to three fatigue fracture situations, demonstrating that the classical methods predict the fractures that actually occurred.

Classical Fatigue Design The use of classical fatigue design techniques in the analysis of an existing fatigue fracture provides conservative estimates because the design techniques are virtually always conservative. To the engineer designing a permanent part or structure, there is no difference between crack initiation and final fracture. The initiation of a fatigue crack most often constitutes failure of the design. Therefore, when used in analysis of an existing fatigue fracture, the values produced are either a stress magnitude that represents the minimum cyclic stress amplitude or cycle life at which fatigue fracture will initiate. The