do stiffeners resist torsion in w-shape

huangyhg

do stiffeners resist torsion in w-shape?
is adding transverse stiffeners to a wide flange beam a viable option to resiste torsion?  we have had this discussion in every company i worked for, whether adding stiffeners at so often along the length of the member will increase the torsional properties of the beam.  if this is the case, can you point me to a reference on how to size and space the stiffeners.  i have already looked at design guide 9.

stiffeners cannot resist significant torsion.  the beam must either be designed for the torsion, or diagonal braces from the bottom flange up to the deck must be added.
if there is a deck welded to the top flange of the beam, a small amount of torsion could be taken out of the beam by the deck--in this case, the stiffeners will help.
daveatkins
stiffners do nothing for torsion as daveatkins states.  adding side plates to wf beams to create a tube-type section will add significantly to torsional strength and stiffness.  but vertical web stiffners don't help.
in certain circumstances, adding stiffeners can be used to brace the compression flange of a plate girder.  this is covered in the ssrc guide to stability design criteria for metal structures",  under "plate girder with elastically braced compression flange."  this is not the same as adding torsional stiffness to a beam, but it's close.  
bracing the compression flange with stiffeners is not torsional bracing.  compression flange bracing is required in many situations but these stiffeners do not satisfy torsion stiffness requirements.  good luck.
the only thing the stiffeners might do is help reduce the warping of the wide flange section, but i know of no way to quantify this, and there is certainly nothing in any code that i know of.  aside from that, i agree with the above that the stiffeners described do not appreciably help torsion resistance.
what jmiec may be suggesting is using the stiffners as a sort of cantilever arm that extends out to the compression flange.  this works as long as there is a sort of moment connection between the end of the stiffner and some outside bracing element.
an example:  you have a sky walk that is comprised of two 36" deep wide flanges spaced, say, 8 feet apart.  the deck of the skywalk is a steel channel, metal deck, concrete slab assembly with the channels spanning from bottom chord of wf to bottom chord....i.e. the deck is set down low and the wf beam compression flanges are about 3 feet above the deck....with no lateral bracing.
you can use vertical web stiffners, rigidly attached to the cross channels such that you have an effective, rigid, "u" shape comprised of vertical stiffner, horizontal channel, and then vertical stiffner.
thus, the stiffners can laterally resist the lateral buckling of the top flanges because of their rigid connection to the deck channels.
if you treat the torsionally loaded wide-flange as two parallel rectangular sections (the flanges), each loaded laterally (torsion/h), and then analyze each rectangle as a   
i believe this compression flange bracing concept was used extensively in railway bridges.  some engineers may think of this as adding torsional stiffness to an i beam.  i brought it up to answer twinnell's original question.
adding stiffeners like that helps with lateral-torsional buckling, not resistance to an externally applied twisting moment.  they aren't the same thing.
it may decrease the warping but i don't think stiffeners are an economical option to take out torsion, and you are not even taking it out.  and i have no idea how to figure out how much of a decrease there may be.