Need a book? Engineering books recommendations...

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

Re: CMU

[Subject Prev][Subject Next][Thread Prev][Thread Next]
In a message dated 97-07-30 09:57:42 EDT, you write:

Questions from Bruce Resnick:

<< <<The out-of-plane connection we have successfully used to TJI's is a pair
of
 holdowns each side of the TJI web which has been reinforced with a solid web
 stiffener each side of the web.  The web stiffener is usually flush to the
 outside edge of the top and bottom chords.  The stiffener is attached with
 glue (APA approved glue or glue used by joist manufacturer) and screwed
 through the TJI web into the stiffener on the opposite side.  The screws
from
 both sides provide a clamping force to provide better bonding between the
web
 and web stiffeners.>>
 
 Does TrusJoist mind that you are taking your shear force through the glued
 connection between the web and the top flange of their joist?  Does it
create
 any reduction in the vertical load capacity of the joist?  Have you used
this
 in LA?
 
Response:

It has been a while since I spoke to TrusJoist about this.  The original idea
came about after some discussions with Standard Structures.  Since the glue
shear transfer between the joist web and flange is around 1100 pounds/foot
and the joist are typically longer than 10 feet in length I don't believe the
seismic shear force transfer will be an issue.  At an anchor spacing of 8
feet on center you are designing for approximately 2400 pounds (8x300#/ft)
 to 4000 pounds a connection depending upon wall thickness and height which
is less than 11,000 pounds (10'x1100#/ft) of shear transfer capacity.  Under
the real expected earthquake demand (1.2g to 1.5g) on the wall connection
(3.33 x .45w = 1.5g) the design force is 7990 to 13320 pounds  which is close
to the 11000 pound given in this example.  If the load is to large, we have
gone to using a parallam. You can also always specity in your general notes
that you want the truss manufacturer to verify the joist can handle the
seismic load combination as part of your  truss design criteria.

It is a good point that was made about the joist web splice being a lack of
continuity.  I would think that since the top and bottom chords are
continuous they could transfer the shear force beyond the web splice (3'x
1100#/ft x 2 chords = 6600 pounds). If the web splice occurs under the web
stiffeners, then the web stiffeners act as splice members helping to transfer
the load (glue and screws) over the web discontinuity.  I will investigate
this furthur, this may become a limiting factor for the joist axial capacity
depending upon if the web splice is a true discontinuity  

 In the case of roof trusses, they are designed for dead and live load,
 during the seismic event you only have the dead load present reducing the
bending demand upon the joist to help offset the axial load demand on the
truss.  I would think there would be no reduction in joist capacity unless
the joist is damaged during the seismic event.

I believe we have used this detail on a building in LA.  But most building's
we have used this detail on have not been that large, if they were that large
we typically ended up with a standard panelized roof system using glu-lams
and 4x's or Standard Structures Truss Purlins.

Question:

 <<When the joist is parallel to the wall, we use a tie rod threaded through
a
 hole in the TJI webs (reinforced with web stiffeners) with 2x blocking tight
 against each side of the tie rod .  The blocking is nailed to the plywood
 sheathing using 10d at 4"o.c and attached to the web siffeners with A35's.
  At the end of the tie rod we have a bearing plate which bears on a web
 stiffener attached to the web of the TJI and is big enough to overlap the 2x
 blocking each side of the tie rod on the opposite side of the joist.  The
 wall connection is a threaded anchor bolt and coupler to the tie rod   As
the
 wall tries to pull away the tie rod goes into tension, putting bearing on
the
 blocking which is then transferred shear through nailing to the diaphragm.>>
 
 Do you worry about shrinkage in the 2x blocking not providing a tight fit
 over time?  Do you worry about loading this connection in compression when
 the wall pushes on the roof since this is really a "tension only"
connection?
  Would such a load cause the blocks to fall out or separate slightly perhaps
 creating an impact load in the tension rod when the load reverses, similar
to
 what happens in typical rod X-bracing?  Assuming that the tie rod is at the
 center of the TJI's, do you worry that the eccentricity between the tie rods
 and the diaphragm will cause the joists to "roll"?  
 
Response:

Typically we use 2x8 or 2x10 blocking each side of the tie rod, we do not
expect any significant axial shrinkage of the blocking since they are only
spanning 24 inches usually between the joist, depending upon joist spacing.
Obviously, some gaps could occur over time, but we feel these will be rather
small and not impact the performance.  The design is as a tension brace only,
when the wall pushes against the diaphragm, we depend upon the diaphragm to
distribute the force to the cross walls.  Typically we block the first joist
space parallel to the wall to help prevent buckling of the plywood.  The
added blocking around the tie rods I think would help furthur stiffen the
diaphragm against the push force of the CMU wall.

Since the end of the tie rod only has a bearing plate against the side of the
joist, which is tightened down by a nut on the tie rod, a compression force
from the wall pushing on the tie rod would only push the plate away from the
joist web (lets say less than 1/64" if any movement at all) resulting in no
net compression on the tie rod. I don't think you would experience any
elongation of the tie rod similar as occurs in tie rod "x" bracing during a
seismic event.  Since we are using A35's for the block attachment to the web
stiffeners, I don't expect that we would lose any alignment of the blocking.
 

The tie rod elevation is typically 4"-5" below the top of the wall ledger,
not at centerline of the TJI.  This allows for the use of 2x8 and 2x10
blocking and the tie rod to miss the top chord of the TJI.  This eccentricity
is small and since we have the blocking over several joist spaces to
distribute the loading to the diaphragm by means of nailing, we do not feel
that this is causing an eccentricity on the joist that you would have to
worry about.

Question:

 Lastly, in this parallel condition, do you use a similar type of tie rod
 detail completely across the building to satisfy the 24' maximum crosstie
 spacing requirement, or do you have beams in those locations?

Response:
 
Typically we have used a beam for this location.  But in many strip malls the
TJI joist typically spans from perimeter wall to perimeter wall, so the joist
acts as your cross tie.  At a spacing of 8 feet on center this keeps the
axial loads realitively small.  If the member has to act as a diaphragm chord
or significant drag element, we use a beam and not a TJI. 

Hope this helps.

Michael Cochran
BCASE1356(--nospam--at)aol.com