Shear lag in welded double angle compression member 1

[Baffled Engineer]

Hello,

I am checking a double angle welded connection to a gusset plate. The double angle is a compression web member of a girder truss. Does anyone know if I have to consider shear lag effects when checking the capacity of the member at the connection? Is shear lag critical for tension members only?

Thank you

 

[dik]

I don't normally... how long is the weld?

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

-Dik

 

[Baffled Engineer]

I don't normally... how long is the weld?

2-L2x2x1/4" welded to a gusset of 5/16" thick with 4" long welds on both the toe and the heel of the angles.

 

[DrZoidberWoop]

AISC 15th Ed updated Case 4 in Table D3.1. You could adapt it for this.

In case you don't have the 15th Ed:

Shear Lag U=[3(L^2)/[3(L^2)+(w^2)]]*[1-(x_bar/L)]

L= (L_weld_@_toe + L_weld_@_heel)/2 = 4 inches in your case.
w= distance btw welds= 2 inches in your case.
x_bar= self explanatory = 0.586"

If you really wanted to use it, U=0.788 (please check my math) or perhaps Case 2, depending on the number you get.

Side note: Company where I work almost always checks for tension limit states, even if members are designated as compression only. It's a good CYA practice.

 

[dik]

Thanks, Dr... x_bar?

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

-Dik

 

[KootK]

Does anyone know if I have to consider shear lag effects when checking the capacity of the member at the connection? Is shear lag critical for tension members only?

I would say that:

1) Shear lag would apply to your cross sectional yielding limit state in compression.

2) Shear lag would not apply to your buckling limit state in compression.

Since buckling will govern in many applications, the impact of shear lag is reduced.

 

[Baffled Engineer]

AISC 15th Ed updated Case 4 in Table D3.1. You could adapt it for this.

In case you don't have the 15th Ed:

Shear Lag U=[3(L^2)/[3(L^2)+(w^2)]]*[1-(x_bar/L)]

L= (L_weld_@_toe + L_weld_@_heel)/2 = 4 inches in your case.
w= distance btw welds= 2 inches in your case.
x_bar= self explanatory = 0.586"

If you really wanted to use it, U=0.788 (please check my math).

Side note: Company where I work almost always checks for tension limit states, even if members are designated as compression only. It's a good CYA practice.

Thanks. I use the Canadian Code CSA S16 and they specifically state "When a tension load is transmitted by welds, the effective net area, shall be computed as the sum of the effective net areas of the elements...", which would imply from my understanding shear lag is applicable (or critical) to tension members only. I don't have much CYA wiggle room since the girder truss I'm working on are from the 1960s and they don't have much structural capacity to spare.

Would it be fair to say, a tension member can fail at the ends where the net effective section is critical as affected by shear lag and net area of the section, but for compression members, if the welds at the ends are sufficiently strong, then the failure would just limited to in-place, out-of-plane, or torsional-flexural buckling?

How would I include the shear lag in a compression resistance check? Would it be the gross area multiplied by the shear lag factor? Would it affect the weld capacity? And does it affect the buckling / compressive resistance of the member itself?

Thanks

 

[Baffled Engineer]

I would say that:

1) Shear lag would apply to your cross sectional yielding limit state in compression.

2) Shear lag would not apply to your buckling limit state in compression.

Since buckling will govern in many applications, the impact of shear lag is reduced.

Thanks! That answers my questions. The shear lag for compression members would just then affect the cross sectional check at the ends, and will not affect the buckling capacity of the member as a whole.

 

[DrZoidberWoop]

dik

x_bar = the "x" with the bar on top, symbolizing the centroid distance from the outer face of the leg.

_
x

 

[dik]

I don't know what x bar is... based on your formula I have:

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

-Dik

 

[dik]

The centroidal distance from the gusset to the L centroid... like 0.586 for an L2x2... thanks... Something like (I do these up like this so I can cut and paste them into other programs)

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

-Dik

 

[JoshPlumSE]

Shear lag is certainly a real effect. But, it's more important for tension loads than for compression.

Other than that clarification, I'd go with what KootK was saying.... that you can check compression yielding with shear lag. But, that probably won't control.