Return to index: [Subject] [Thread] [Date] [Author]

Re: wall anchorage to metal deck flexible diaphragm

[Subject Prev][Subject Next][Thread Prev][Thread Next] Gerard,

Here are some of my thoughts, you are on the right track, design similar to
the wood diaphragm and develop the subdiaphragms and cross ties.  I have
added some comments below after your initial comments.

Michael Cochran S.E.

In a message dated 6/27/2001 3:15:32 PM Pacific Daylight Time,
GMadden(--nospam--at) writes:

Searched the archives on this topic and found some discussion and was
wondering if anyone had more opinions or reference they could site one way
or the other.

When anchoring CMU and Concrete walls to an unfilled metal deck flexible
diaphragm there is some debate as to the effectiveness of the deck and its
connections to behave as intended.

Here are the issues/conditions

1) When the deck flutes are parallel to the wall being anchored, it is very
weak and would crumple and expand like an accordion. Some mentioned they
design the anchors at the joists or purlins every  4-8ft on center -
avoiding the use of the deck as a subdiaphragm. However, if an angle was
welded to the underside of the deck to stiffen it, I think you could rely
on the deck to behave as a subdiaphragm similar to wood as long as the
angle drags into the building far enough so it develops the diaphragm shear
demand from the wall anchorage. The drag should terminate with a
perpendicular steel member to take the chord out of the sub diaphragm

Your approach seem correct to me, we have used this approach on some seismic
retrofits of tilt-ups with metal deck.  You might be able to use screws
instead of welds depending upon gage of deck.  You can probably use the every
bar joist in one direct for the cross ties and possibly the girders in the
other direction.  If the girder spacing is to much (in excess of 24 feet) you
can probably use continuous spliced HSS 4x2 flat (or similar size) in between
the girders for the cross tie.  There should be enough room since the bar
joist sit up on top of the girders.

2) When the deck flutes are perpendicular to the wall being anchored, it is
much more stiff and is good in resisting wall separation (tension at the
weld). When the wall moves into the diaphragm (compression) the buckling
stiffness of the deck should be investigated and struts should be provided
if necessary.

I would still add some sort of wall tie at 8'-0" o.c. other than just using
the metal deck.  When the wall pulls away from the deck, the puddle welds see
shear in two directions simultaneously; shear perpendicular to the wall as
the wall pulls away, and shear parallel to the wall due to flexure of the
diaphragm spanning between the cross walls.  Think of the diaphragm as a deep
glu-lam beam in bending where the top is in compression and the bottom is in
tension and the glue (puddle welds) is trying to keep the individual lams
(metal deck and ledger) from sliding past each other.

3) The wall anchorage provisions still apply for metal deck diaphragm even
though the provisions were primarily developed due to cross-grain
bending/tension failures of wood ledgers. My understanding is that equally
valid is the idea that local failures should be avoided to ensure the
global system and its assumptions function properly (i.e. roof support for
out of plane wall design is treated as a roller support in a beam analogy")

I agree, you want to ensure the global system behavior

4) The puddle welds are adequate to transfer both tension and compression
loads at the ledger.

 As describe in item #2, you need to account for all the shears acting, and
it is probably better to add wall anchorages back into the diaphragm even
when the metal deck is perpendicular to the wall.  If using just the metal
deck, need to check the eccentricity of the connection to the steel angle
ledger since the wall pull force is at the horizontal leg is above the anchor
bolt and how far does the vertical leg extend below the anchor bolt to resist
the compression portion of the bending moment induced as the wall attempts to
pull away.  

If anyone can point me to a solid reference or share their thoughts on these
matters, I'd appreciate it. We are involved in a large prison project with
many buildings and would like to be as efficient as possible in regard to
this issue. Most of my wall anchorage experience has been with wood
diaphragms, so I am tending to think that little is different here in terms
of the need for subdiaphragms and load path continuity. Project is in Zone
4 97 UBC - 1 story CMU and Tilt Up.

thanks in advance,

Gerard Madden, P.E.
Civil Engineer

Middlebrook + Louie, Structural Engineers
71 Stevenson Street, Suite 2100
San Francisco, CA 94105
Tel: 415.546.4900
Fax: 415.974.3680
Email: gmadden(--nospam--at)