Preliminarily size the beam spanning from column to column. Then run a frame analysis with knee braces for gravity loads and lateral. Please note that the knee braces will cause reverse bending in the beam and so if the bottom flange is not braced (which is most common) it will reduce the allowable stress. Check the sway against allowable movements.

Sometimes the knee brace connections are hard to make work - esp in wood

Okay, your question is straight-forward, but has a complex answer...

First, to just answer your question:  You may consider the knee braces to be effective supports for DEFLECTION and calculate deflections based on the span between the connection points of the knee braces.  This presumes your knee braces are relatively short compared to the span of your rafter/beam.

Next, for strength you should strongly consider jike's recommendation of a computer model, however working it out long hand is fine as well, but your structure is automatically indeterminate and not necessarily easy to work out.  Consider the brace as pin-connected to make your job easier;  Also it is near-impossible to make rigid connections for this situation in any material other than steel (not counting rare materials like Alluminium, where I really don't know).  Pay close attention to the shear introduced into the knee region above the brace in the column and behind the brace (towards the knee) in the rafter/beam.

Finally when designing the brace be sure to give due consideration to the way in which it will connect and any eccentricity inherant due to the brace shape.  By this I mean you need to consider the secondary effects of a non-symmetrical brace member like a sigle angle.

DO NOT use eccentric connections.  There are lots of papers and procedures around for eccentric connections, however most are based on flawed research out of the US.  Unfortunately the research was quite competently carried out, well written with the best of intensions and just happens to be wrong for most out-of-the-lab applications.  Because it was well written, seemed perfectly reasonable and greatly simplified the connection it has been widely reproduced.  That doesn't make it right, or undo the structural collapses (yes, plural) which have occured due to the use of the detail....  You can design an eccentric connection, but it is NOT easy and should come out VERY inefficient.  If the numbers don't drop like a rock, your design method is likely based on the incorrect procedure.

Good luck,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

youngstructural,
What do you mean by papers based on flawed research?  Can you reference some of these papers?  Structural collapse(s)?  Please provide some more detail.

The knee bracing has significant impact on the buckling length of the beam in the axis it is being supported. Your question though depends on whether it a sway frame, vertically loaded or axially loaded. In a sway frame you are into double-bending of the top beam with knee braces. In axially loaded compression members you are dealing with the buckling load. Knee braces are very economical and a powerful tool for saving material costs. Refineries ar replete with knee braces!   

Robert Mote
www.motagg.com

connect2:  Happily able to acomodate you...

The flawed research paper: "Eccentrically Connected Cleat Plates in Compression" by S. Kitipornchai,(Fellow ASCE), F.G.A. Al-Bermani, and N.R. Murray.  Journal of Structural Engineering, Vol 119, No.3, March 1993

Sources Referencing Paper / Condemning Design Procedure:

Include the design procedure:
- AISC (that's AUSTRALIAN Institute of Steel Construction) Hollow Structural Section Connection Design Guide

Against:
- CISC Technical Memorandum No. 5 (Discussion regarding use in lateral load frames).
- Design Procedure for Eccentrically Loaded Cleats in Compression, Dr. G. Charles Clifton, HERA Senior Structural Engineer, et al. July 2006

Examples of Structural Failures (most likely) due to the use of this detail to date mostly relate to snow loading on shed or Cold Form Steel structures; This is because where it is most frequently used (ie: lateral load resisting system) would require either ULS wind or a seismic event to detect the problem.

I have personally worked on structures which collapsed due to this cleat detail under snow load.  There was plenty of strength in the system otherwise.  Unfortunately it is a matter of client confidentiality and I cannot tell you what structures where.  I can tell you they were principly CFS structures on the South Island of New Zealand.

I should also add that my cautions are based on this thread (ie: Knee braces);  Eccentric cleat connections are perfectly fine in many applications, including the most common "X" bracing lateral frame.

Hope that helps/clarifies,
Regards,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

youngstructural:

That is correct, the Australian Steel Institute (formerly known as the Australian Institute of Steel Construction) published a warning in their technical journal in 2004 that the design method for eccentrically connected cleats based on this research should not be used.

This is not to criticize the work of Kitipornchai, who is a well respected academic worldwide (particularly in Brisbane, Australia).  

I have a great deal of respect for the Academics of our Profession, however the paper has lead to a measure of difficulties. I'm sure Kitipornchai is an able researcher, but the end result in this case was not positive.

My intent (and I had thought this clear) was to warn about the design of eccentric connections for knee braces, NOT to criticise any of the researchers involved.  It is terribly difficult to predict and extend laboratory conditions to the field, and these authors have my respect for their efforts. However that does not make their research or design procedures based on it correct.

Regards,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

Woohoo,

Its the Kiwis versus the Aussies here.

Save it for the rugby field guys - oops thats probably a taboo subject for both our countries.

Anyway,

TimmyH76,

I personally would design it 1 of 2 ways:

1. full analysis incorporating the knee brace.
2. treat it as a moment connection and take the moment as a push/pull at the knee brace and the column.

If the knee brace is very long then I would only do it by method 1.

Actually csd72, I'm a CanuckKiwi.... Canadian transplanted down to the NZ side of the world by seismic interest and a gorgeous local gal...  But you're right; Neither NZers more Aussies have much cause to be cheering the rugby at the moment! *smiles*

Cheers,

YS

B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...