cmh-issc-lessons (Раздаточные материалы), страница 2

The bridge was closed totraffic, and Professor Farquharson went toobserve what was happening.On the bridge was a logging truck (ref. 14), a car,its owner (a newspaper reporter), and his dog;the driver and passenger of the logging truckhad escaped to safety. Farquharson joined thereporter on the heaving deck. Together theytried to get the dog out of the car. As thebridge's motion became increasingly violent, thetwo men gave up trying to rescue the dog.Instead, they concentrated on rescuingthemselves.The last few minutes of the bridge's demise wascaptured on film. The resulting footage hasprobably been seen by just about everyengineering student in the last 50 years.

On thefilm, Farquharson and the reporter can be seentrying to make their way to safety. The professorhad an easier time of it, because he walked alongthe centerline of the bridge, which was nearlymotionless. The reporter struggled along theedges of the roadway, which was heavingviolently. Both made it; the only casualty of theeventual collapse was the dog.Condron was an advisory engineer for theReconstruction Finance Corporation.Hisapproval of the bridge design was a necessarypart of the approval of the loan application tohelp finance construction.

As he studied theplans, Condron became concerned about thenarrow width of the bridge relative to the lengthof its main span. He developed a table (table 1)to compare the ratio of span to width in theproposed design to that of recently completedsuspension bridges (ref. 15).Table 1 - Span to Width RatiosWhen the amplitude of the undulations in thebridge reached twenty-five feet, the suspenderropes starting tearing, and the deck broke,sending the car and truck into the water. Within30 minutes, the rest of the deck fell into PugetSound, leaving only the towers remaining.These towers had been bent out of shape by theviolent motion; they were dismantled before areplacement bridge was built.Investigation: The Federal Works Agencyappointed three engineers to investigate thefailure: Theodore von Kármán, the director of theDaniel Guggenheim Aeronautical Laboratory atthe California Institute of Technology; Glenn B.Woodruff, the engineer of design for the SanFrancisco-Oakland Bay Bridge; and OthmarAmmann, a world-renown bridge designer.

Theyissued their report less than five months after thecollapse occurred.This report exonerated the bridge designers andengineers saying that “the Tacoma NarrowsBridge was well designed and built to resistsafely all static forces, including wind, usuallyconsidered in the design of similar structures. …It was not realized that the aerodynamic forceswhich had proven disastrous in the past to muchlighter and shorter flexible suspension bridgeswould affect a structure of such magnitude as theTacoma Narrows Bridge” (ref.

15). That is, theengineers had followed the current state of theart. They had used the accepted techniques fortaking wind effects into account. As mentionedearlier, these techniques had been developed byMoisseiff himself. It just so happened that thesetechniques turned out to be flawed.The report did record, however, that oneparticular engineer had raised concerns about thedesign before the bridge was built. Theodore L.BridgeSpan(ft) Width(ft)Delaware River1,75089Ambassador1,85059.5Whitestone2,30074San Francisco Bay 2,30066Geo.

Washington 3,500106Golden Gate4,20090Tacoma Narrows 2,80039Ratio1:19.71:31.11:311:351:331:46.71:72This table showed the proposed TacomaNarrows bridge to be significantly more slenderthan any other existing suspension bridge. ToCondron, this seemed to be going far beyondcurrent experience.Engineer Condron was sufficiently concernedthat he continued to investigate. After hearingthat the University of California at Berkeley hadconducted some tests on models of suspensionbridges, he visited with Professor R. E. Davis inBerkeley. According to Condron, ProfessorDavis “felt reasonably confident that the lateraldeflections of the Tacoma Narrows Bridge asdesigned and determined by Mr. Moisseiffwould be in no way objectionable to users of thebridge” (ref. 15).Condron found additional support for deflectiontheory in the written discussion thataccompanied the article in which Moisseiff andLienhard published their theory.

The discussioncited the University of California studies asconfirming the accuracy of deflection theory.One discussant went so far as to say thatMoisseiff and Liendard’s analysis was sufficient“to silence all arguments for unnecessary floorwidths” (ref. 16).What was lacking in thisdiscussion and in the Berkeley tests, and whatwould eventually lead to the downfall ofdeflection theory at Tacoma Narrows, was therecognition that accounting for lateraldeflections alone was not enough: verticaldeflections mattered, too.Because he was the only one who seemed tohave doubts about the bridge design, andbecause the deflection theory of Moisseiff andLienhard had widespread support among bridgeengineers, Condron ultimately acquiesced. Hewrote in his final report: “In view of Mr.Moiseiff's recognized ability and reputation, andthe many expressions of approval … of hismethods of analyses of stresses and deflectionsin the designs of long span suspension bridges,… I feel we may rely upon his own determinationof stresses and deflections” (ref.

15).His support was not unqualified, however. In hisfinal report he also recommended consideringwidening the bridge to 52 feet. Had this beendone, the width-to-span ratio would have been1:53.8. The bridge would still have been thenarrowest in existence, but less radically so thanit turned out to be. According to Petroski (ref.12), “had Condron's recommendation beenfollowed, it is very possible that the TacomaNarrows Bridge would have been stiffenedenough that, even had it exhibited some degreeof flexibility in the wind, that might have beenwithin tolerable limits and thus subsequentlycorrectable, as it was to be in other contemporarybridges.”As we know, Condron's recommendation wasnot followed, and the bridge collapsed.

Sometime was to pass before the actual cause of thecollapse would be determined (the report fromAmmann, Woodruff, and von Kármán left thematter vague). The details of the cause are notimportant for this paper. What is important is torealize that the theory on which the bridge wasdesigned was flawed because it did not take intoaccount everything that needed to be taken intoaccount.

In particular, the dynamic effects ofwind load on the bridge were ignored. Relianceon the flawedtheory was a significantcontributor to the failure.Relevant Lessons: Some of the lessons of theTacoma Narrows failure are specific tosuspension bridge building. One such lesson,for example, was the need for aerodynamictesting; this testing became a standard procedurein suspension bridge structural analysis in everybridge built afterwards.

In addition to suchspecific lessons, there are at least four lessonswith application beyond bridge building. Theselessons are explained in the rest of this section.Lesson 1: Relying heavily on theory, withoutadequate confirming data, is unwise.At the time of the design of the Tacoma NarrowsBridge,Moisseiff and most other bridgeengineers believed that the accuracy ofdeflection theory had been adequatelyconfirmed. As mentioned earlier, the results oftests on scale models at the University ofCalifornia had shown close agreement with thetheory’s predictions for lateral deflections. Also,several bridges had been designed using thetheory, and they were still standing.As Theodore Condron had suspected, neitherthe scale model tests, nor the existing bridgestruly provided adequate confirming data. Thescale model tests were inadequate confirmationbecause they did not produce any data aboutvertical deflections.

The existing bridges wereinadequate confirmation because none of themwere nearly as narrow and shallow as the TacomaNarrows Bridge.The first real test of the accuracy of deflectiontheory occurred above the waters of the PugetSound. When this test failed, the inaccuracy ofthe theory became apparent.Over time,problems occurred in other bridges that had beendesigned using deflection theory. Many of themwere eventually modified to employ additionalmeans of stiffening.Lesson 2: Going well beyond existing experienceis unwise.Although many previous suspension bridgeshad been built, including two with longer spans,the Tacoma Narrows Bridge was unique. Astable 1 showed, the span to width ratio of thebridge was 54% greater than that of anycontemporary bridge.The Tacoma NarrowsBridge was not a simple extrapolation fromexisting experience; it was a radical departurefrom that experience.Even with deflection theory seeming to justifysuch a departure, the most prudent action wouldhave been to make small, incremental steps innarrowing bridge deck widths.