ENGINEERING ETHICS: THREE CASE STUDIES

1 ENGINEERING ETHICS: THREE CASE STUDIES (Subtitle: EVERYTHING SHOULD BE AS SIMPLE AS POSSIBLE, BUT NOT SIMPLER)Victor Singer, (retired)Dan Schiffbauer, : Canons or Codes of Ethics Noteworthy Historical Background for Each Case In Point The Tacoma Narrows First Suspension Bridge: failed 1940 The Silver Bridge (US Hwy 35 Ohio River Crossing): failed 1967 The Shuttle Challenger: failed 1986 Questions To Ponder For Each Case In Point: Are There Relevant Ethical Issues Or Just 20-20 Hindsight? When, If At All, Did Which Ethical Issue Come Into Play? If Errors Occurred, Were They Errors Of Judgment, Omission or Commission? Or Murphy s Law Events - - Accidents In An Imperfectly Understood World? Who Should Have Done What, That He Didn't Do?CANONS OR CODES OF ETHICS NSPE, ASCE, ASME, AAES: Engineers shall hold paramount the safety, health and welfare of the public .. (Fundamental Canon) AIChE: Members shall hold paramount the safety, health and welfare of the public.

2 IEEE: We, the members of the IEEE .. , do hereby .. agree to accept responsibility in making decisions consistent with the safety, health and welfare of the public .. AIAA: The AIAA member will have proper regard for the safety, health and welfare of the public in the performance of his professional duties. AIA: Members should .. thoughtfully consider the social and environmental impact of their professional activities. ABA Model Code of Professional Responsibility: lawyer should assist in maintaining the integrity and competence of the legal lawyer should assist the legal profession in its duty to make legal counsel lawyer should assist in preventing the unauthorized practice of lawyer should preserve the confidences and secrets of a lawyer should exercise independent professional judgment on behalf of a lawyer should represent a client lawyer should represent a client zealously within the bounds of the lawyer should assist in improving the legal lawyer should avoid even the appearance of professional impropriety.

3 CASE IN POINT: TACOMA NARROWS FIRST SUSPENSION BRIDGE Highway 16 Puget Sound Crossing between Tacoma WA and Kitsap Peninsula opened to traffic 7/1/1940; collapsed 11/7/1940; 6-minute video of collapse at : (1940)VON KARMAN VORTEX STREET (observed from space): Wind blowing above the sea is disturbed by an island, creating a "vortex street." Clouds serve as markers, allowing us to see the vortices in the flow downstream of the Karman Vortex Street in the atmosphere: SeaWiFS image courtesy of Orbimage, Inc., SeaWiFS Project (Code ) and Distributed Active Archive Center (Code 902) at NASA's Goddard Space Flight Center, Greenbelt, MDTACOMA NARROWS BRIDGE: before, during and after failure 3 hrs transverse flexural (up and down) vibration, amplitude ft, 35 mph wind Wind rose to 42 mph, a midspan hanger cable broke (slipped), and the torsional mode started. Frequency Hz, 28 ft double amplitude Frequency associated with von Karman vortices at 42 mph wind: ~ 1 Hz Federal Works Agency's Failure Commission concluded: Vortex shedding unlikely important because of the frequency difference.

4 Violins teach that excitation with RANDOM frequency content produces responses at resonant frequencies AND AT HARMONICS unless the vibrating system provides substantial internal damping. Currently popular attribution: AEROELASTIC FLUTTER: jargon for stimulation of multiple degrees of freedom CONCURRENTLY, A MURPHY S LAW PHENOMENA from: Damper (SCE 1928)PROTECTION AGAINST MURPHY S LAW EVENTS In general, interpolate within rather than extrapolate beyond prior experience. Extrapolate in small steps, only where risk and benefits are balanced. Tacoma Narrows Bridge was a substantial extrapolation into the unknown. Much lower flexural rigidity and torsional rigidity than prior art, to reduce cost. Eminence of the designer - - Leon Moisseiff - - played role in suppressing criticism. Exception - - David B. Steinman, also an eminent suspension bridge designer, criticized the design at a 1938 meeting of the ASCE Structural Division.

5 Othmar Ammann (Failure Commission member) wrote: "The failure has given us invaluable information.. [that] every new structure that projects into new fields of magnitude involves new problems .. [for] which neither theory nor experience furnish an adequate guide. It is then that we must rely on judgment. If errors or failures occur, we must accept them as a price for progress."QUESTIONS TO PONDER Are There Relevant Ethical Issues Or Just 20-20 Hindsight? When, If At All, Did Which Ethical Issue Come Into Play? If Errors Occurred, Were They Errors Of Judgment, Omission or Commission? Or Murphy s Law Events - - Accidents In An Imperfectly Understood World? Who Should Have Done What, That He Didn't Do?SECOND CASE IN POINT: THE SILVER BRIDGE US Highway 35 Ohio River Crossing, Point Pleasant WV to Kanauga Ohio. completed 1928, collapsed 12/15/1967 center span 700 ft, side spans 380 ft each, deck width 28 ftfrom: BridgeSILVER BRIDGE 46 dead, 2 missing (never found)HERCILIOLUZ BRIDGE Florianopolis, Brazil (26 deg S latitude) center span 1113 ft.

6 Opened 1926, closed 1982, re-opened (pedestrians) 1988, closed 1991 Designers: David B. Steinman and Holton D. Robinson Using eyebar chain as upper chord of stiffening truss reduced weight and cost. American Bridge Company proposed using their "special" 1060 carbon steel eyebarsheat-treated to over 2X strength of the 1020 ASTM A-7 in common use. Steinman dissociated his firm from further connection with the project. SILVER BRIDGE: concept & proportions copied from Florianopolis For main span 2/3 as long, chain tension force was about half as large, allowing chains with only two heat-treated 1060 steel eyebars rather than four as on Florianopolis. If one outer eyebar breaks, pin cannot sustain unsymmetrical load; bridge comes down Eyebars 12 x 2 , Eye dia 28 in. Pin dia 11 inches Heat Treat: ST 875 degC, WQ, temper 625 degC 2 hrs. Average eyebar properties: Ftu 120 ksi, Fty 81 ksi, El >20%, R/A >45%, CVN 2 to 3 ftlbs at 32 degF, KIc ksi root inch Other steel shapes and plate (1020 carbon steel, ASTM A7-24) Ftu ksi, Fty ksi, El , R/A > 50%, CVN > 15 ft lb at 50 degF, KIc ~100 ksi root inch Eyebar allowable stress 50 ksi (67% of min required elastic limit)Ignorance is Truly Bliss Only one recovered eyebar had a broken head Fracture Surface, lower side: corner crack inch long (parallel to 2 inch thickness), and inch deep radially.

7 Technology enabling quantifying the relationship between flaw size and failure load didn't yet exist in the 1920's when Silver Bridge was designed. 1920 s era brittle fracture provisions for steels for buildings and bridges commonly reflected a 15 ft lb minimum CVN (Charpy Vee Notch Impact Test) Eyebar steel, at 2X the strength, provided less than 20% as much fracture energy Steinman s view (evidently): Stronger eyebars either too much extrapolation from prior experience or too little fracture energy for Florianopolis, as far from equator as Miami and much warmer than BRIDGE: post-mortem Found original stress analysis reasonably accurate Evaluated actual loading at time of failure to estimate actual load on failed eyebar. Conducted extensive material testing program, including modern fracture toughness tests and related crack growth rate quantifications, and scale model testing program. Final Highway Accident Report NTSB Report SS-H-2 (adopted 12-16-1970): Fig 36: approximate positions of all vehicles on the bridge at time of failure.

8 Westbound lane 1/3 empty, eastbound lane more than half empty. Page 98: "the probability of 100 percent of design stress is quite remote." Failed eyebar head obviously loaded to actual capability when it failed. The state-of-the-art circa 1920 s lacked recognition of potential premature - - flaw-induced - - failure in brittle materials at low temperatures. Fracture cause: a flaw grew over 40 years to critical size by stress corrosion and corrosion fatigue. Which was dominant is unknown. No indication main member deterioration by rusting to a point of inadequacy under intended or actual TO PONDER Are There Relevant Ethical Issues Or Just 20-20 Hindsight? When, If At All, Did Which Ethical Issue Come Into Play? If Errors Occurred, Were They Errors Of Judgment, Omission or Commission? Or Murphy s Law Events - - Accidents In An Imperfectly Understood World? Who Should Have Done What, That He Didn't Do?THIRD CASE IN POINT: SHUTTLE CHALLENGERWhat SHOULD happen: SPACE SHUTTLE DISCOVERY at STS-120 LaunchWhat DID happen: SPACE SHUTTLE CHALLENGER at SPS-25 Launch Jan 28, 1986 BACKGROUND: Space Shuttle Program Overview Vehicle Architecture: Shuttle's "Stages In Parallel" architecture was a substantial departure from Apollo and Mercury "Stages In Series" architecture.

9 Re-Usable Orbiter objective made Crew Capsule Jettison feature of Apollo and Mercury programs more difficult than before. NASA Decision: NO CREW CAPSULE ESCAPE SYSTEM FOR SHUTTLEAPOLLO flight vehicleBACKGROUND Space Shuttle Program Overview cont d. Funding Strategy: Seek Congressional approval by earmarking at least one component supplier contract for each US Senatorial and Rep District in the country. Delaware got space suits, Wisconsin got SRM steel cases, Minnesota got phenolic resin impregnated carbon cloth, Virginia got rayon for white woven cloth (carbon cloth precursor), Nevada got ammonium perchlorate oxidizer for solid propellant, etc. Congressional direction to NASA: YOU TELL US that revenue from commercial flights will cover the program development costs. Interpretation: "You tell us" more important than whether or not it's true. Responses: 1) Flight schedule large enough to recover program development costs; 2) Intense program-long focus on costs.

10 Original launch schedule built from "modest" start to SIXTY flights/year; peak launch rates actually achieved: 9 in 1985 and 8 in 1997. All ground support systems designed AND BUILT to support 60 flights/year schedule. SUBTLE DEMONSTRATION OF NASA SHOP CULTURE Projected 60 flights per year schedule is one flight every six days. Time in orbital flight much longer than time during ascent. Potential for damage by collision with space junk (natural or otherwise) during ascent or orbital flight, precluding normal return, should be among design considerations. Obvious solution: Standing policy NOT to fly Flight "X" until Flight "X + 1" is within six days of flight-ready. No such policy was in effect when the Shuttle Columbia failed re-entry (2/2003). After Columbia, it was implemented for only one of the several additional flights. Credible reason: In the year such a policy becomes effective, implementation requires paying for one more total vehicle build-up than the number of flights.