Effective length of Equal Angle bracing (welded) 3

[mte12]

Has anyone seen a reference for the effective length of welded bracing, such as used on trusses.
Guess it would be less than 1.0 but need a reference.

Shown is a cross section. All members are angles.
The bracing ion question is coloured green.

 

[TLHS]

Sorry, is that a built up box section with angles on the corners and plate connecting them?

 

[le99]

Similar to truss members, you couldn't simply assume full fixity at the joints, so "k" should be conservatively taken as 1.0.

 

[dik]

yup... 1 is the number, I'd use. Need a little X-bracing, every so often, to keep it from collapsing.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Do you feel any better?

-Dik

 

[mte12]

Similar per photo, 3D truss with all angle members.

I think Le, is actually less than 1 if it's welded but difficulty in quantifying. Similar, it there was more than one row of bolts for a bolted connection.
Case I'm looking at needs refinement, that's why I'm asking.

 

[le99]

Look like this is pipe support. The majority of the load will be taken by the chord members in the longitudinal direction. Le for chord members, as well as the diagonal bar, will be the distance between panel joints. In the plane shown on your sketch, you need to make sure the presence of intermittent X bracings that are adequate to prevent the in-plane distortion, or you may consider adding stiffener plates at the corners. Note the latter approach is much more expansive to construct, but it may be justifiable to use a shorter Le, measured from the center of weld at one end to the opposite end.

 

[mte12]

Hi, the above is a sample photo only.
The truss is existing.
I was going to use the actual length "L" from inner edge of welds (refining as much as I can), and was hoping for a reference which justified "Le" less than 1.0. I'm sure ot depends on rotational stiffness.

 

[jayrod12]

I'd need to check my Canadian code, but I feel there's a justification somewhere in there for using 0.9 for K for web members in certain circumstances.

 

[dik]

This is from a publication by Alcan titled "Strength of Aluminum" I had the entire publication scanned, but cannot find it. You may have to tilt your monitor 90 degrees... The publication is from 40 years back.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Do you feel any better?

-Dik

 

[mte12]

Thanks, this is a really good start.
If you look at the middle line for A-K, B-K, D-G, D-H and F-N, it allows a reduction for Kx and Kz.

I'm guessing Kz applies as this is a natural behaviour (about one of the principal axis). Or was this for bending? Brace length may affect to a degree.
So for 1 bolt, Kz = 0.8 and for 2 bolts, Kz = 0.7.

Do you agree with using Kz?

 

[dik]

I generally check...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Do you feel any better?

-Dik

 

[Accumulated]

I normally only assume them at 1.0 but see below.

Take a look at BS EN 1993-1-1 BB.1, it allows a reduction in the Lcr value.
For buckling in the plane of the truss beam: The buckling length can be taken as equal to 90% of the system length (Node to Node distance) IF the truss members are connected at each end with at least two bolts or by welding.
For buckling out of plane of the truss beam: The buckling length should be taken as equal to system length.
For buckling in the plane of the chord member: The buckling length can be taken as equal to 90% of the system length (Node to Node distance).
For buckling in the plane of the chord member: The buckling length should be taken as equal to system length.

 

[jayrod12]

In the Canadian code I stand corrected, for trusses there's no such reduction in length. However for steel joists the web members can use a K=0.9.

 

[mte12]

Thanks to all for your input.
Will have to make an assessment of the plane of buckling, somewhere in between principal and geometric plane.
Forgot to add screenshot previously.

 

[hoshang]

For buckling [highlight #EF2929]in the plane of the chord member[/highlight]: The buckling length can be taken as equal to 90% of the system length (Node to Node distance).
For buckling [highlight #EF2929]in the plane of the chord member[/highlight]: The buckling length should be taken as equal to system length.

I Can't catch this.

 

[le99]

@hoshang,

I think there are missing words "weak" and "strong" in front of "plane".