Risa Connection Block Shear Welded Connection 1

[LeonhardEuler]

Hello,

I was designing a tension brace in Risa connection and noticed that it does not check block shear of the gusset plate for bolted nor welded angles. For bolted I understood, because I was only using one row of bolts so bolt bearing/tearout should govern and there won't be any block shear of connection; however, I don't understand how the program can get around calculating block shear for a welded angle. The weld will clearly be attempting to pull out the parent metal in a rectangle equal to the size of the welded area. The way I see it the gusset plate should be checked for tensile yield and tensile rupture along the whitmore section and block shear along the weld line. Is this not correct?

 

[canwesteng]

Block shear is only for the case of bolts tearing blocks off the end of a member. For welded connections you would check shear lag.

 

[JoshPlumSE]

I'm looking at the results for RISAConnection for a brace connection and I see the following:

Bolted or Welded:
1) Gusset plate tensile rupture (basedon whitmore section)
2) Brace Tensile Rupture (of the full brace cross section w/ U factors)

Bolted
3) brace block shear
4) Bolt bearing (which is limited by bolt tear out)


So, are you saying that the gusset plate tensile rupture based on the whitmore section isn't quite right? That it should be replaced by a block shear check instead? That's actually a pretty interesting point. My guess is that looking only at the whitmore section rupture (and not including the shear portion of block shear) is more conservative. But, the geometry of the tear out section could be slightly different.

If you ask the RISA guys to add this check to the program be sure to explain this to them with two pictures. One showing them the whitmore section failure and the other showing the similar block shear failure. Best to really explain your questions when you send things over there now....

 

[LeonhardEuler]

canwesteng I disagree. 16.1-444 of the 15th ed spec and commentary explicitly states that block shear occurs in gusset plates with force transmitted through a welded angle. Also from memory I believe shear lag for longitudinal welds is 1.0. Shear lag applies mostly to bolted connections.

Josh plum I was thinking there should be three gusset plate checks yield and rupture of the whitmore section and block shear around the toe of the weld

 

[canwesteng]

Shear lag is not 1.0 for longinutidinal welds. You might be thinking of transverse welds - the U factor is 1, but the net area is reduced. I see that the commentary points out to check block shear for welded conncetions to gusset plates, but I'm struggling to see how that can govern over weld strength for a reasonably detailed gusset plate.

 

[JoshPlumSE]

I agree that there should probably be a block shear check as well in the program. I just wonder if it will ever control. Because, at least how I understand them, the two failures are related.

Not the best picture because it shows a bolted connection and we're talking welded. But, it's the best I could quickly copy and paste.

The Block shear failure method I'm thinking of is tension rupture on the Ant area in the picture (perpendicular to applied force) and shear failure in the areas parallel to the force. Looking at it from this perspective, block shear is a tensile failure of the base material which is partially resisted by shear yielding along the sides of the block.

Now, the tension rupture on the whitmore section is on a wider area than the Ant shown. So, the geometry is a little different. I just don't know that it's a large enough difference that block shear will end up controlling. Also, there is another failure method (tension yielding combined with shear fracture) that should get checked as part of block shear as well.

 

[LeonhardEuler]

I would tend to agree with your sentiment JoshPlum, but I have designed a welded brace that had a lower block shear strength at the weld toe than the tensile rupture strength at the width of the whitmore section. The whitmore section allows for much greater area than the block shear failure path does.

I feel that we can't assume on something that is not very clear on this. Welds should always have the base metal checked, because it is possible for the base metal to fail before the weld in some connections.

 

[LeonhardEuler]

Also, there is another failure method (tension yielding combined with shear fracture) that should get checked as part of block shear as well.

Is this hiding in section J somewhere?

 

[JoshPlumSE]

Nothing hidden.... It's just another side of the block shear equations. But, a side of it that I was ignoring in my discussion.

Since I was so focused on comparing the tension FRACTURE of the Whitmore section to block shear equations considering tension fracture (along with shear yielding), I wanted to point out that I realized this was ignoring the alternate block shear behavior which checks tension YIELDING (along with and shear fracture).

 

[LeonhardEuler]

JoshPlum understood. Thought there was an additional provision I wasn't aware of