Henry Petroski

Computation of the Weight Function from a Stress Intensity Factor

A simple representation for the crack-face displacement is employed to compute a weight function solely from stress intensity factors for a reference loading configuration. Crack face displacements given by the representation are shown to be in good agreement with analytical results for cracked tensile strips, and stress intensity factors computed from the weight function agree well with those for edge cracks in half planes, radial cracks from circular holes, and radially cracked rings. The technique involves only simple quadrature and its efficacy is demonstrated by the example computations. The weight function for a corner crack in an LMFBR hexagonal sub-assembly duct is constructed from stress-intensity-factor results for the uniformly over-pressurized case, and it is shown how this may be used to determine the stress intensity factors.

Structural dynamics of piping with stable cracks: Some simple models

Abstract The effects of cracks on structural response are illustrated through simple analytical models for cracked elastic and rigid-perfectly plastic beams. Such models can capture-essential mechanical phenomena and enable generic problems to be studied in order to identify worst-case scenarios. Stable cracks are shown to alter not only the magnitude of response but also the modes of plastic deformation. The effects of boundary conditions, as well as crack location and size, are shown to be amenable to study via simple models.

WEIGHT FUNCTIONS FROM MULTIPLE REFERENCE STATES AND CRACK PROFILE DERIVATIVES

Abstract Two modifications for implementation of the weight function method are presented. It is proposed to assume the derivative of crack opening displacements instead of the crack profile in the form of a series, as is commonly done, and to compute the unknown coefficients in the series from one or more known stress intensity factors. The new approach eliminates differentiations and reduces the number of integrations as compared with the commonly used method. The new method is tested by exploring the influence of assumed reference load cases and combinations of reference cases on the accuracy of computed stress intensity factors for a variety of applied loads on an edge crack. Use of two known stress intensity factors is shown to give excellent accuracy of computed weight functions and stress intensity factors, well within 1% for most load cases considered.

The Success of Failure

It is always more pleasant to talk about success than failure. The boss con ducting performance reviews much prefers to tell staff members that they have done a good job than a poor one. It is less confrontational. The teacher-student relationship is not one of boss-employee, of course, but similar dynamics can be in play. Performance reviews naturally take the form of assigning grades. Time was, as every academic knows or suspects, when grades assigned were relative to the cohort known as a class. In col lege courses, only the very best-performing student or students got As, the majority of the class got Cs, and those who brought up the rear actually got Ds and Fs. Grading was never a pleasant experience, but awarding an hon est distribution of grades was considered part of the responsibility that came with the privilege of being a professor. Accepting an honest grade was part of the responsibility of being a student. The student who tried unduly to influence a teachers grade was called an apple polisher, or worse. Over the past decades the institution of grading has become somewhat inverted. Students now grade professors, through what are called teacher or course evaluations, and some professors appear to crave good evaluations as badly as premed students crave As. Some observers have accused the professoriat of kissing up to the students with more palatable syllabi, easier reading lists, more entertaining lectures, fun field trips, and higher grades. Over half the grades at some elite institutions are said now to be As. With the anticipation of so many good grades, students rank all of their woebe gone classes as better than average, and all professors as better than average teachers. Students feel good about their transcripts; faculty members feel good about their continuing appointments and promotions. Individual failure, whether in the classroom or in the workplace, is an almost extinct concept. Judging performance has become . . . well, judg

Stability of a crack in a cantilever beam undergoing large plastic deformation after impact

Abstract A simple model is employed to determine the dynamic response of a rigid-perfectly plastic cantilever beam with an attached tip mass and a crack, taking into account the weakening effect of the crack. The crack is assumed to be located at the base of the beam, and an initial velocity is imparted to the tip mass. The subsequent stability of the crack is considered by calculating the tearing modulus based on the J- integral associated with the deflecting beam. For the example of circumferential cracks in thin-walled piping, whose idealized geometry models some stress corrosion cracks found in service, radial propagation and instability are found to be more likely than circumferential. Once a crack penetrates the wall, however, stability in the circumferential direction is found to depend in a complex way upon loading and crack geometry.

Controllable states of rigid heat conductors

ZusammenfassungEs werden stationäre Temperaturfelder bestimmt, welche in allen isotropen homogenen starren Körpern ohne äussere Wärmezufuhr möglich sind. Wenn der Wärmestrom nur von der Temperatur und vom Temperaturgradienten abhängt, muss das Temperaturfeld gleichförmig sein. Hängt der Wärmestrom nur vom Temperaturgradienten ab, dann muss das Temperaturfeld schraubenförmig sein. Für den letzten Fall werden Beispiele gegeben.

Plastic response to impact of a simply supported beam with a stable crack

Abstract A rigid perfectly plastic model is developed to study the effects of a central crack on the plastic response of a beam subjected to a central transverse impact load. The governing differential equations are expressed in terms of non-dimensional parameters representing the crack size, beam to projectile mass ration, and energy input. The equations are solved numerically to obtain general solutions, and the effects of the parameters on the final permanent deformation are isolated and studied to identify the relative significance of each parameter.

Paconius and the pedestal for Apollo: A case study of error in conceptual design

A classic case history of an engineering failure related by Vitruvius is presented as a paradigm of human error in the design process. It is argued that a familiarity among designers with such case studies could be instrumental in reducing conceptual errors in the design process generally.

Failure as a unifying theme in design

Abstract The concept of failure, which plays a central role in successful design, is viewed as a unifying principle of the whole design process. This idea is elaborated upon and illustrated for the problem of designing a hypothetical bridge, which is presented as a paradigm. Brief case studies of some real bridges are employed to reinforce the argument, and the danger of designing by emulating success are cautioned against.

On detecting a crack by tapping a beam

Abstract A simple model for the cracked elastic beam is employed to simulate its response to a sharp tap at an arbitrary location. Explicit expressions for the Fourier components of the response of a simply-supported beam make it possible to explore various strategies for detecting the crack via signature analysis and to provide rational explanations for some characteristic responses of cracked beams.