Need a book? Engineering books recommendations...

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

RE: Lateral Brace Deflection

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Ed

My reading of S&J says that  Kreqd = 2*Kideal does not require an
additional factor of safety.  The multiplier of 2 is a function of the
initial crookedness and does not vary based on the loading.  An exception
would be if other loads increased the initial crookedness of the member.

Kreqd is the translational stiffness at the point of bracing and may be
provided by members acting in tension or bending.  In other words if you
imposed a unit translational displacement at the point of concern the
resistance should exceed Kreqd.  You should consider not only the bending
stiffness of the beam and the diagonal brace but also any other
flexibilities in the bridge structure that would have a significant impact
on this term.

Assuming that the transverse beams contribute to the bracing of  chords for
each of the trusses I would be inclined to calculate the stiffness by
imposing a unit displacement inword for both truss chords at the same time.

If there were other transverse loads such as wind on the truss chord or
vehical loads on the transverse beam that produced deflections in the chord
consistent with the buckling shape you are trying to suppress, then a
stiffer structure would be required.  Note that these loads could be
thought of as increasing the initial crookedness of the member and would
not need to be considered when considering the availible stifness.

Hopefully you will be able to show that the transverse deflection of the
truss chord is not significantly influenced by other loads and that the
overall flexibilities of the bridge do not significantly influence the
stiffness of your brace.  If this is not the case or if it becomes too
involved you might want to consider performing an eigenvalue buckling
analysis.


Mark Gilligan


---------------------------------------------------------------------------
-----
Message text written by INTERNET:seaint(--nospam--at)seaint.org
> 
Mark,
   >Salmon & Johnson has a good treatment of point bracing for
   >beams.  You want
   >to make the brace stiff enough to inhibit undesirable
   >buckling shapes.  The
   >strength of the brace is then calculated as a function of
   >the crookedness
   >of the brace.
They mention members braced by transverse members, but they only address
axially loaded braces...

By their method,
The required lateral brace capacity, Q, is:
Q = 0.004*Kideal*L
Kideal = 4*Pcr/L  (with beta=4), Pcr = 156k (service) and L = 120" so,
Kideal = 5.2 k/in and
Q = 2.5k (conservative compared to 0.002*P, in this case)

2.5k is not hard to accommodate, but Kreqd:
Kreqd = 2*Kideal*FS  (FS = 2.12 for AASHTO) so,
Kreqd = 22 k/in

As I understand the derivation, Kreqd would simply be directly related to
the bending stiffness of the beam in this case, rather than the axial
stiffness of a brace for the case addressed in the book.  Given the
geometry, 1" vertical deflection of the beam = 1.7" lateral deflection of
the top chord.  Therefore, the beam's stiffness must be at least 22*1.7 =
37.5 k/in!

With a 6' cantilever, this is quite the beam.  If this is correct, I have
some serious re-thinking to do!

Ed
<