wtc, again

huangyhg

wtc, again
while watching the latest nova program concerning the whys and wherefore of the wtc collapse an interesting point came from one of the investigators/narrator. it concerns the failure mode of the bolts holding the floor joists to the beams. the description was going along very well as he explained the fire with resultant sagging of the floor joist. when he was describing the connection failure that caused the progressive collapse he noted the bolts were supposed to fail in shear. he then said that due to the sagging joist the bolts failed in tension. if i heard him right he then said this was the weakest direction. i think the weakest direction was a slip.
can anyone explain how the bolt at the joist to beam connection could ever be in tension?  even with the joist on the floor below the clip would also be bending and the bolt would still be in shear.
thanks,
here is an interesting site on the wtc.
i got the same impression that you did when i first heard mr. charles thornton make that statement. fortunately i have the progarm on tape and listened to that part again. he says that the heat of the fire caused the floor trusses to sag and "begin acting like a catenary". this of course would mean that the floor trusses are now actively "pulling" on there connections (instead of primarily "bearing" on them). the bolts at the connections were oriented so that this "tension" from the sagging floor trusses caused the bolts to shear.
slideruleera,
thanks.
i missed the statement about the catenary. that was one hell of a good bracket on the beam.  it done its job to the last as it was designed to do.  i don’t think their animation displayed this scenario to well or did i miss this as well.  
still have my k&e log log duplex decitrig and a set of log tables along with ibm punch cards.
also numerous types of k & e graph paper. also a 4 function 2 digit calculator (not operating) that i paid $125 usd in the 60s’and was the envy of the engineering group.
unclesyd - the discovery channel & history channel both had specials on the wtc, but the nova program seemed the best (from and engineering viewpoint).
i still have my pickett n4-es log log dual-base speed rule and still use it occasionally (for practice). remember friends who had your k&e model - a very good rule. my first electronic caluculator was a texas instrument sr-10 (sr stood for "slide rule"). would not do much more that the basic four functions however. sometimes would work parts of a certain problem on the pickett (trig, logs, etc.) and other parts on the ti (add, subtract, etc.).
best wishes