Need a book? Engineering books recommendations...

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

RE: Lag screws in withdrawal from end grain

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Although I don't have knowledge of the intent of the codes, I have seen many
failures of connections - especially at patio trellises which were added as
extensions to eaves by laging a 2x ledger to the end of the truss or rafter
tails. This was especially true in Northridge.
Rather than debating the intent of the code, you might consider a few
options. These are more costly, but should be able to be structurally
1. You can epoxy in place a threaded sleeve and double the embedment depth
into the wood. Epoxy installion (either of the sleeve or of a threaded rod)
will be uneffected by shrinkage or changes in the wood since the connection
is made by adhesion of the fibers to the rods. The use of the threaded
sleeve allows for installation of a threaded rod.

2. I needed a hidden connection to secure a Portola to a heavy timber beam
and to the column below (through a wood Corbel). I could not use a
mechanical connector since the style of the home was old rustic and
primative Southern Arizona (rough Santa Fe). The main purpose was to replace
a column cap for possible uplift. The beam was already laterally supported
by the residence shear-walls (actually it was my home). I had the framer
drill a hole through the corbel and 16" deep beam. He used a 24" long bit
and drilled 18" into the end of the column. The rod used was a SAE1018
(A307) 5/8" diameter threaded rod. I had him Epoxy the rod into the hole
which was slightly less than the width of the screw. My purpose was not use
solely an epoxy type connection, but the allow the threads to engage, and
let the epoxy lock them to the wood fiber.

3. A third type of connection would be to saw cut a vertical groove into the
10x10 and insert a steel plate which can then be bolted through the side of
the 10x10. You can conceal the nut and bolt head and finish with a wood plug
sanded smooth to the face of the beam.

4. If you must use a wood screw type connection (or lag bolt) you may wish
to strap across the 10x10 and the bolted member in order to compensate for
the reduced tension created in the end grain connection.

5. You did not describe what was being being connected to the beam. I assume
that it a wood ledger or joist that is parallel to the end of the beam. Make
sure you have an oversized washer (minimum 2" square x 3/16") to reduce the
possiblity of failure perpendicular to grain.

Finaly, if you are doing what I think you might, you are trying to create a
discontinuity strap connection but either don't want to have an exposed
strap or don't have access to the top or side of the beam. If the above does
not help, fax me a copy of the condition and let me see what I can suggest.


-----Original Message-----
From: NRoselund(--nospam--at) [mailto:NRoselund(--nospam--at)]
Sent: Monday, September 07, 1998 9:18 AM
To: seaint(--nospam--at)
Subject: Lag screws in withdrawal from end grain

The 1994 Edition of the UBC says "lag screws should not be loaded in
withdrawal from end grain.  When this condition cannot be avoided, the
tabulated nominal withdrawal value, W shall be multiplied by the end grain
factor, Ceg=0.75."

The 1991 edition has a similar statement.  I don't find the 'should not'
statement in the 1991 NDS.  Neither Standard allows wood screws to be used
end-grain withdrawal.

"Should not" seems to be advisory, not mandatory.  How serious is the
advisory?  The required modification factor does not seem a very restrictive
way of dealing with a bad idea; is Ceg=0.75 a realistic adjustment?

I have a retrofit condition in which use of lag screws inserted into the end
of an existing 10x10 will make a difference between a quick and simple
installation and a costly, complicated, and possibly unsightly installation;
does this qualify as "Cannot be avoided"?  A Ceg=0.3, or even less, is
workable for the purposes of this particular Detail.

Does anyone have any background on the intent behind the uncharacteristicly
advisory wording of these Code provisions, or insights on how to handle an
allowable factor for something that should not be done?

Nels Roselund
Structural Engineer