Need a book? Engineering books recommendations...

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

RE: Wood Truss Reinforcement

[Subject Prev][Subject Next][Thread Prev][Thread Next]
At 12:23 AM 11/18/99 -0600, you wrote:

>Mr. Greenlaw, 
>I have worried about the parallel to grain splits you described.  On the
other hand, what's your opinion about the possibility of staggered bolts
actually causing this splitting as the moisture content drops and the wood
shrinks?
>
>Ed Fasula, E.I.T.
>
>>... I always design a means of transversely clamping or "stitch bolting"
the wood tension
>>members so that they cannot split apart and release their bolts down the
split, parallel
>>to grain, as I have seen happen. This clamping is analogous to lateral
confinement ties in
>>concrete members...
>>Charles O. Greenlaw SE 
-----------
        Good question. Staggered bolts, split rings, etc. (ie, two or more
lines of fasteners) are a separate, independent source of splitting,
whenever the connecting side members do not shrink at the same rate. This
restraint -of- shrinkage effect occurs even in the absence of design loads
on the connection, and in the absence of any change of geometry of the
structure that introduces "secondary stress" moments into supposedly "pin"
connections.

        As such, a staggered or wide pattern of fasteners is definitely a
possible cause of splitting, all by itself.

        Yet any single, round connector in wood, loaded parallel to grain,
is going to have a measure of outward component to its bearing pressure on
the hole in the wood. This outward wedging component acts to split the wood
parallel to grain. Quick, what is the allowable tension stress in wood
across the grain? There ain't any, not in modern times. Wood is unreliable
in tension across the grain (except for fruit-tree firewood I try to split.)

        Perhaps the reason bolt holes in wood are to be slightly oversize is
to allow the wood to shrink laterally onto the bolt without causing tension
from an interference fit, and to reduce the outward component of bearing
stress on the bolt hole. (These reasons are my own speculation; I've never
seen or heard any commentary on this, but would like to.)  

        Now imagine the wood member already split along the bolt diameter.
There is still some outward component from bolt bearing that acts to widen
the split, and the wider the split the worst the outward, splitting
component: A vicious-circle failure mechanism.

        The truss connection that my earlier reply described was with
65-inch long 1/4"x5" steel side plates connecting 6x8 bottom chords at
third-points of 98-ft span bowstring trusses set 20-ft apart. Chord tension
was 23k at DL; 45k at TL. Seven- 1" diam horizontal bolts in double shear in
each butting 6x8; all bolts in one line without stagger. All of this met
code and good practice of the day.

        Trouble came from compressive buckling, between web connections, of
the curved top chords, which were a stack of six-2x6 planks, laminated with
nails, not glue. The pieces were not working compositely very well. (Their
as- buckled shape would take your breath away.) Resulting truss sag tried to
bend the 6x8 bottom chords, and the side plates fought back at chord joints
with a prying action on the long line of bolts. That plus direct stress
split the 6x8's along the bolt line. Splits were 3/8-inch wide at the bolt
holes, and the bolts had advanced into the split.

        What prevented collapse was a single 1/2-inch vertical bolt with
ordina ry cut washers through each 6x8 near the 1" bolt closest to the 6x8's
butt end. These stitch bolts were about 3/8-in loose, from seasoning
shrinkage of the 6x8, at unsplit members. The split chords spread open until
restrained by the single stitch bolt, and moved no farther, so the seven 1"
bolts still held their 23k at dead load. Whew!

Hope this is responsive to your question

Charles O. Greenlaw, SE