shear center again

huangyhg

shear center again
when you design a channel do you load it with the actual loading plus a torsional load which equals the actual loading multiplied by an eccentricity of the distance from the center of flange to the shear center?
i've designed many channels without considering this and no other engineers checking my work have mentioned it.  i just wanted to get some other opinions.
find a job or post a job opening
most of the uses i have seen for channels have involved stair stringers or exterior canopies where the main factor driving the size was architectural.  the sections were so large and the loads so amall relative to the section that torsion was never an issue.  
the only time i considered it was when i designed the catwalks for a performing arts center.  there i did consider the shear center in how i detailed the catwalks as the major framing beams were channels.
mike mccann
mmc engineering
like you, i rarely design channel sections including the effect of eccentric loading.
as is often the case with torsion, i deal with it through detailing rather than calculation.
consider two most common situations:
1) something frames into the side of your channel. even if it's just a shear tab connection, it probably offers significant torsional restraint to the channel.
2) something frames over top of your channel.  when the channel begins to rotate as a result of the torque, the point of loading will shift towards the heel of the channel flange, much closer to the shear center.  additionaly, if the framing passing above the channel is is welded or bolted to the channel flange, you've got some torsional restraint once again.
loads applied at a sections shear centre are pretty rare on aircraft. generally speaking, torsion is something to be avoided if possible by designing. however there are times when you just have to bite the bullet and include its effects. it depends upon what the analysis is and the confidance in the structure is. as ever it can become more of an art than a science at times it seems.
sometimes it can be a driving factor in the design.
i know most engineers ignore it, however i have seen one case where the torsion of a channel was enough to add additional deflection, dropping the natural frequency of the floor syste, and causing a serviceability issue due to walking excitation.
personally i just frame into back of the web side of the channel every opportunity i get.  i am also a fan of using 4-bolt connections at the ends to try give a measure of torsional rigidity to the end connections.
i would echo the previous comments:  avoid it whenever possible, detail it out when it's not, and if you really can't do anything else and (as an impromptu rule of thumb) if the load/capacity ratio is more than .8, include the effects.  and if you're loading from above and nothing imparts the (very minimal necessary, refer to adamp's post) don't forget to check the twisting deflection too...  you'd be surprised just how much a channel can twist!
cheers,
ys
b.eng (carleton)
working in new zealand, thinking of my snow covered home...
agree with approach of others in using channels.  use connections which take account of the torsion.  one further thing, if you use a channel on a sloping surface, say as a roof purlin or whatever, if you turn the toes up the slope, the gravity load applied to the top flange is then closer to the shear centre.  the natural tendency may be to turn them down to ensure drainage, but up is best structurally.
ive designed quite a few channels in my time but never even thought of the shear centre and torsion because of the eccentricity of load.  now perhaps i will think a little bit more about it.  purlins ive always had the habit of turning the toes up towards the slope and re  
from adam: "additionaly, if the framing passing above the channel is is welded or bolted to the channel flange, you've got some torsional restraint once again."
i believe this is why most folks get away with not considering the torsion.  (as was typed in the vaporized thread: thanks moderator for deleting useful info.)
as i typed before also, i find that the deflection is usually what causes the problem for open sections with torsion.  for example, consider what happens if a spandrel beam rotates and this translates into additional brick deflection.