RSS Feeds
Readers Sound Off
By Design News Staff -- Design News, July 18, 2005
Design News received many responses to Ken Russell's Calamities article, "The Case of the Collapsing Coliseum" (http://rbi.ims.ca/4394-534). Below are some letters and responses from Ken Russell.
Thankful and Lucky
On January 18, 1978 at 4:15 in the morning, after the University of Connecticut men's basketball team defeated the University of Massachusetts, the roof collapsed. I am thankful that I had left the coliseum 6 hours before. I doubt if fewer than a thousand people would have died. If you have not read the attached report (Engineering News Record, http://rbi.ims.ca/4394-536) please do so.
Harry Peterson, Design Newsreader
Ken Russell's response: Thanks for the interesting report. All four of the failures discussed were of interest to me. It now appears that buckling was the chief suspect all along, and a metallurgist (me) was hired by the architect's attorneys in an act of desperation. The report gives little credence to the opinion of a consultant who blamed the failure on a defective scoreboard weld. I felt all along that a number of screw-ups, none fatal by itself, reinforced each other, causing the failure.
Now for a piddly point. You refer to a basketball game. Both the Martin report and The Boston Globe refer to a hockey game. Could you search your memory? In any case, you were lucky!
Bad CAD
I had heard from two different sources that the reason for this failure was that the internal diagonal bracing of the spans directed the static and dynamic load to the center of the span instead of outward toward the upright supports. One person that told me this was a CAD trainer and cited the importance of understanding the basics of engineering when using CAD, offering the Hartford Coliseum as an example. He said that while everything was sized correctly a simple "mirroring" error by a CAD operator had placed the diagonal bracing in the wrong direction. In the review of details, the investigators overlooked, for some time, the direction of the bracing.
Richard E. Wronski, Springfield, MA
Ken Russell's response: Ouch! That sort of screw up is one for the books. It is now clear to me that the failure was buckling. The design was faulty, as you note. Add to that fragmented authority and the absence of engineering oversight. The architect requested a structural engineer but the construction manager refused one, to save money. The engineer's salary would have been a heck of a lot less than the price of a new roof!
Blind Justice
I think it is clear that the collapse was a buckling failure (Loomis,R.S., et al., "Torsional Buckling Study of Hartford Coliseum," Journal of the Structural Division No. 15124, American Society of Civil Engineers, Jan 1980). AISC has since made changes in the code in response to the collapse. The physical evidence and the calculations point to a buckling failure. I do not know what happened in court but a blind justice system can have problems with buckling theory.
Bob Loomis, Design Newsreader
Ken Russell's response: Thanks for the E-mail. Other readers sent in some references and it sure looks like the failure was in buckling. Both design and oversight were seriously deficient. Too often bunches of people have to die to get a common sense change through.
Ignoring the Human Element
I appreciate that nobody has complete information in cases like this, but you seem to be overlooking the likelihood that few and maybe only one fracture surface would show evidence of fatigue. After the first fracture shifts the loads around, all later fractures would be due to overloads. The problem would be due to a structure designed with insufficient redundancy and an unexpected fatigue mechanism that broke critical connections. Space frame structures are famous for low redundancy. Welds are famous for being subject to fabrication errors that make them vulnerable to fatigue. I would suspect that some motor-powered device like a fan or pump induced a resonance in a key part of the structure and led to the initial fracture, which then led to all the other failures. Usually you would find that workers in the building would have noticed that a particular system caused more noise than expected, and so was only run at night.
George Hawkins, Design Newsreader
Ken Russell's response: You have an interesting scenario, which brings in the often ignored human element. It is, of course, possible that there was cyclic loading and fatigue. But, I saw no part of it nor do I recall any sort of a source for such loading.
Talkback
DN's Resource Center Get Free Information, Made Easy
-
2008 Adhesives Sourcebook - Loctite Products for Design, Assembly, Manufacturing and Maintenance
Henkel Corporation | Design Guide
Request Item -
Solutions for the Appliance Industry
Henkel Corporation | Design Guide
Request Item
