I take it that you have a portalised HSS frame with knee braces, made from reinforcing bar, at the corners and that the rebar is trying to "punch through" the HSS "face", resulting in plastic deformation due to bending.
  
Surely a better knee brace would be another piece of HSS, of suitable geometry.  The axial force in the HSS brace would then be transferred to shear force in the sides on the HSS column and rafter, instead of putting the HSS "face" into bending.

This is not a knee brace.

These other engineers are bracing a standard HSS rectangular tube, HSS 18x6x3/16ths, for example. The round bar (#6, I think) goes through the tube, completely. Two holes are drilled, one in each side to accomplish this. The end result is a Frankenstein-esque pair of rod ends projecting symmetrically from the middle of the the long sides of the rectangle of the HSS tube. Instead of a column, imagine Frankenstein himself standing on the footing with the beam resting on his head. Frank's bolts are the reinforcing rod.

I could explain better with pen and paper!

Thanks, DD

It sounds like they would be replacing a thru-plate stiffener with a single rod of rebar.  Is there a connection at this location that could cause some local instability in the side face of the HSS?  I hope they are using weldable rebar, A706.  With an 18-inch face and 3/16-inch thickness, the face is pretty slender.  That may be all they are doing, trying to stiffen the long side against local buckling.  Sound logical or way off?

UcfSE,

You are right on.

I used "rebar" where I should have used "round bar". And, judging by the savvy of the two engineers who do this, the qualities of the bar have likely been considered.

Yes, there is a connection here. About 1 to 3 feet above where the rod-stiffener is placed, is the column-beam connection, which is typically 1/2" to 5/8ths plates welded to beam and to column, then bolted together.

The forces here range from 20 to 100 kips.

I use ASD. I am guessing I can use formulas for stiffeners to prevent web crippling. This is the closest approximation I know of -- but I am not a guru of the Green Book!

What formulae would you use?

-DD

To determine the stiffeners at the connection, say at the shear plate, you would use the AISC HSS Connections Manual.  This is only offered in LRFD.  Basically you'll be checking the shear plate for punching shear through the HSS face and the slenderness of the HSS face for excessive deformation.  If you fail one of those then you go to a through plate or some other form of reinforcement for the HSS face.  There are other failure modes and equations given for other types of connections as well in that manual.  There are also lots of tables for capacities of standard connections, very useful in my opinion.

Outstanding! This is what I needed. Thanks! -DD

This seems unconventional to me; I've never heard of this method to reinforce an HSS tube.

You're talking about the HSS face punching, and other out of plane effects.  What about the in plane effects, such as compression & bending etc.?  Are those not existant, or have they been omitted from the analysis?

You use code specified HSS face slenderness (and web crippling expressions for that matter) only for IN-PLANE compression of your HSS face, not for the out-of-plane effects you describe.

Please be careful with this, it sounds plain wrong.

tg

The HSS Connections Manual is developed specifically for connections and includes design equations for the failure modes I mentioned above.  Punching shear in the HSS face and excessive out-of-plane deformation are failure modes given for instance for the typical single-shear plate connection welded to the outside face.  There are other guidelines and design equations given for other shear conenctions, moment connections and also truss-type connections.

Trainguy,

Thank you for your concern. All cautions are appreciated. I expect I am okay with this, here's why:

The forces involved are an axial load and a moment coming from the column-beam connection above. There is no wall attached to the column, so only the 8" width of the column itself takes an out-of-plane load. This is small compared to the load at the haunch, coming from a roof.

There are at least two engineers doing this locally. Both design HSS buildings regularly and have been doing so for years.

And, it makes sense. If an evenly loaded beam deflects too much, an extra support midway is an effective solution. And that is what happens here...

punching is not the concern here. There is no shear tab or other connection to the column.

UcfSE,

I am ordering my HSS connections book TODAY. I also found a handy website, Bull Moose Tube, http://www.bullmoosetube.com/images/PDFs/bmt-connections-man-low.pdf
This page shows HSS connections and considerations, with graphical displays of failure types, including yield lines and shear punching.

Peace, DD