beam web buckling

huangyhg

beam web buckling
i am wondering if it is possible in bending to have the web of beam deform (buckle) without the flanges showing any signs of deformation.  if the web is slender, and the flanges are really stiff, can this occur?  
supposively if this beam is being operated correctly, there will be no concentrated shear loads, and large distributed loads will cause large bending moments.  the web is buckled almost the entire height, but i don't see how this could be from bending. i think this kind of failure is due to shear, but all the shear checks from asd involve concentrated forces (crippling, yielding, compression buckling, etc).
on an aluminum beam - i have seen the web buckle into an almost perfect "s" but it was caused by distributed loads much higher (4 to 5x)than anyone could have expected and the beam was well braced laterally and had large bearing areas.
in aircraft structures, the typical beam is designed with intentional web buckling.  we call it diagonal tension.
i've even worked on a wing spar that buckles at limit load.
i don't know if "without the flanges showing any signs of deformation" is correct, but it's certainly possible to have web local buckling if the web is slender enough.  in plate girder design, some call it bend buckling.
what you're describing almost sounds like sidesway web buckling.  are the two flanges moving transversely relative to each other?
unless it's mainly buckling at the high shear areas only, i wouldn't suspect shear buckling, but it's hard to know from your description.
the two flanges should not be moving transversely relative to each other.  the web is actually a bulkhead, and the flanges are actually the deck and bottom of ship.  the flange width was taken as an effective width.  the bulkhead is very high, about 7' with a thickness of less than a half of an inch.  there are lots of stiffeners running down the plate, spaced about every two feet.  this is transverse to the ship.  also, there are longitidunal bulkheads running along the ship.
none of the stiffeners are buckled, none of the longitudinal bulkheads are buckled.  since the top and bottom flanges are essentially continuously supported, i did not think sidesway buckling could be an issue.  the buckle looks as if the web has bowed out with the largest part being at its center.
another reason i thought it must due to shear, is because of where the web is buckled.  it is buckled in the tension zone.  
i am just not very familiar with the slenderness aspect, and maybe something may be occuring that i am not thinking about.  i've looked into the local buckling affects in asd such as web crippling, web yielding, and compression buckling of the web.  however, these all involve concentrated loads, and the buckles are occuring in areas where there should be none.
thanks for your replies
it has been years (decades, now that i think about it) but i seem to recall web buckling was part of a plate girder design process.  web stiffeners are added near supports, where the shear exceeds the capacity of the web, and other sources of concentrated loads.  if you want to understand how a steel (or aluminum) section acts, design a plate girder.  you really get into the behaviour of sections.   
don phillips
not sure if i completely understand your problem, as i don't know much about how ships are built.
could your transverse bulkheads be acting as compression struts to brace the sides of the ship?  are the buckles forming locally between the stiffeners or globally across the ship?  when you say the web is buckled in the tension zone, how did you come to that conclusion?
the shape of the bottom of the ship may have something to do with the bulkhead buckling.  as the shape changes due to different loading conditions, it may be distorting the thin bulkheads.
as kwan pointed out in aircraft structure beam shear webs are not typically designed to be buckle resistant up to ultimate loading.
read the introduction of nasa tn-2661 to understand the concept of "pure shear resistant webs" and "diagonal tension" web design. it sounds as though your case may be "intermediate diagonal tension" web design.  
web buckling is caused by shear forces, not bending.  the pattern of the buckling aligns itself with the principle stresses, often near 45 degrees inclination at a support.  typical civil engineering codes have provisions for avoiding this by choosing a proper web thickness and adding stiffeners if necessary.  aircraft design is another matter where you have to pay dearly for each additional pound.  we civils avoid it because it is anticipated to cause concern in the public.  a web can safely be designed that would still buckle but we just don't go there.  good luck.
large aircraft design has generally moved on now from using diagonal tension. although you allow the webs to create the field, you still generally have to add weight on the boundaries to compensate, f&dt people tend to complain also. large aircraft are generally designed to be non buckling at ultimate now. but i think i'm in agreement about the idt being the cause of the buckling.