The beam is neither moving up (the
dead load holds it down at any time), nor is shifting in the floor plane due to
the diaphragm nailing.
What would be the purpose of the (4)
bolts? Among other things, the bolt holes will take away from the capacity
of the supporting beam at the location of the maximum
Zipping of the weld definitely came
to mind; however, the absence of the "fourth" weld, appears to make this problem
worse, not better. Besides, "zipping" is only a matter of stress,
right? Even considering the weld notching, the stresses in the weld and in
the plate are quite low, and welding will be done in a controlled environment of a shop.
Steve Gordin, SE
----- Original Message -----
Sent: Thursday, June 02, 2005 9:26
Subject: RE: Residential Steel
Seems like an unusual
detail for steel beams, but sounds similar to a joist header detail I remember
in the Vulcraft steel joist manual. I would be concerned about notching
from the all-around weld and from the weld ?unzipping? at the edges of the
plate if you don?t weld all-around. I?d recommend connecting the plate
to the top flange of the supported beam with (4) 5/8? dia. A307 Bolts.
Field weld to the top flange of the supporting beam with 3? fillet weld on 3
edges of the plate.
From: S. Gordin
Sent: Thursday, June 02, 2005 10:06
Residential Steel Detail
morning, fellow engineers.
experience, the use of structural steel in residential construction -
particularly, in remodeling, is always somewhat tricky because of the
different tolerances - demanding for steel and quite relaxed for wood.
Field welding, burning, cutting, etc. are as frequent as they are
designing the intensive remodel of a 3-story residential house, I came
up with this detail. Imagine two identical wide-flange floor
beams connecting at a "T." The reactions are low for steel (say,
6K). The standard detail is to weld a tab to one beam and to
cope/bolt the other. Due to the above considerations, I thought
of shop-welding 12"L x 8"W x 3/4" plate flat to the top flange
of the supported beam (5" overlap,
all-around weld). The plate then simply bears (1" max. gap, 6" overlap)
on the top flange of the supporting
beam, transferring the reaction to its
beams are nailed to the floor diaphragm through the bolted nailers (except at
the plate location), and "are not going anywhere."
Construction-wise, the connection appears perfectly "flexible."
Structurally, the deflections are negligible, the stresses in the plate and
beams are moderate, the weld is at about 30% of the allowable
still uneasy with the detail. What do you think?