I am a retired structural engineer and have viewed the structural design drawings and details for the failed area at the David L. Lawrence Convention Center, the subject of "Engineers Slam Failed Joint Detail". The drawing set was dated 12/24/01.
The failure area on the second floor was one of six identical 60-ft-square framed areas for truck loading docks. The only difference was that the failure had an expansion joint on its western edge. In addition, the floor slab consisting of 30-ft-long precast concrete double Ts, spanning in a south-to-north direction, had a downward slope of 2 ft toward the truck dock at the north end. The support framing for the double Ts consisted of two W36x280 steel beams with side plates at the south and north ends of the area and a W40x397 steel beam with side plates in the center of the area. The beam (W36) with the connection failure was at the south end of this area.
The article reported that the failed connection detail was heavily loaded. According to the drawings, the two W36 beams had a design end reaction of 181 kips and the center W40 beam had a design end reaction of 349 kips.
On the design floor plans, there are no sections cut through the expansion joint in the failure area indicating what connection detail to use for this joint. There is, however, a design drawing (S1435) showing a suggested steel detail for the expansion joint at the fourth floor. This detail is very similar to the failed connection, except that the supported end reaction at the fourth floor is 97 kips in one instance, and the remainder are 68 kips. It would appear that the fabricator took this detail and used it for the truck-dock area connections on the second floor.
Two things are apparent from the small pictures in the article. For the connection to have any chance of working, bearing bolts with threads excluded from the connected steel thickness would have to be used. You can clearly see that this was not the case, as there are abrasion marks on the threaded end of the bolt similar to marks on the unthreaded shank of the bolt under the head. Bolt threads digging into the steel angle would restrict movement. Also, the bolt should have been longer to permit double nutting, which would have allowed movement in the connection and at the same time ensured that the bolt could not slide out of the connection. It appears that the slotted angle is heavily rusted. Again, in a sliding bolt-type connection, all connected surfaces should have been painted to reduce friction and any possible movement restriction. For this connection to work, appropriate notes must be placed on the erection drawings to prevent erectors from using the wrong-length bolts and also from overtightening them.
Finally, the question has been raised again about connection design responsibility. Is it the design engineer in all instances or can this responsibility be shifted to the fabricator?
Green Retrofit Chicago Conference
Join your peers in Chicago on November 17, 2009 for insightful presentations that will showcase the best opportunities for green design and construction in Chicago.
