Your S10 would have to be attached in such a way that torsional twist in your main W14 beams would be resisted by flexural stiffness in the S10 - i.e. - a moment connection of some kind.
If you ONLY bolted the S10 to the W14 at the bottom flange, you wouldn't have that "moment" connection as the W14 could twist away from the S10 (W14 web would bend away or the flange would twist somewhat near the stiffener plate.
I would be inclined to do what you suggest, only add some kind of connection between the top of the S10 and the stiffner - just a small tab plate would do.
Usually, when you need to keep the top flange clear of bracing but need the brace, you see the vertical stiffeners (like you suggest) and the cross/brace member (your S10) attached to the vertical length of the stiffener by either bolts or welds.
That is exactly what I had done, my intermediate stiffener 'leaks over' the edge of the bottom flange of the W14x30 and is welded to the top of the S10.
Like this, I hope it shows up the way I made it
|||||||||||||||||||||||||||||| ||||\ <-- welded to top of S10
JAE,
Thanks for the clarification. I would like to read Yura's research on this subject. However, I am still curious about weron4u's original question: i.e. how far below the top flange is standard lateral bracing still effective? As previously stated, Yura has determined that bracing at the neutral axis is not effective, so is there a range within the region between the top flange and neutral axis that can be used? Could you use a straight-line interpolation with respect to the effectiveness of bracing? Perhaps an imperical rule of thumb should be established, say a factor of safety of 60% of the distance from the top flange to the neutral axis? The force on the lateral brace would be increased proportionally based on the straight-line interpolation.
I haven't seen any research or articles where this issue has been developed. My own practice would be to attached to the top flange as much as possible, and if there is some need to set a joist or other "pinned" type connection to the side of the beam, I'd use a vertical stiffener plate to better tie the top compression flange into the framing member in a more rigid fashion.
I have to admit that I have set a web-type joist seat near the top (upper half) of beams before and assumed full translational bracing, but this is rare.
In the seminars and material I have, Yura doesn't speak about some limit or rule of thumb here - his examples deal with either top flange translational braces and mid-height rotational braces.
something I've done is calculate the brace force required at the top flance and then checked the web in bending between the top flange and point of bracing to see if it is adequate (it has never been yet)
