Unusual Canopy Connection 5

[dik]

I have an unusual canopy connection with W12x30 beams framing into an HSS 8x8x0.375 column. They connect at different levels. There is a continuation of the beam on the opposite side of the column. I'm planning to field weld 5/8" end plates on the beam ends to the face of the columns as shown. Does anyone have any concerns about this approach? Worked a better approach. Thanks...

 

[SWComposites]

Tolerances on beam length leading to gaps?

 

[dik]

Tolerances on beam length leading to gaps?

The BAR stock should be able to accommodate the gaps, and the bolted clip angle will hold the beam temporarily while welding the BAR material.

 

[HTURKAK]

This could be assumed semi rigid connection rather than rigid. I would prefer the W12 X30 sloped beam is continuous and the HSS column extends to the bottom of sloped beam with a cap plate .

 

[DrZoidberWoop]

You have a lot of options. Perhaps something like this? or a field-welded similar option?

Running the W12 continuous over the column is always good, unless it interferes with the perpendicular W12. Resolving the flange force at the top of an HSS is going to be difficult, unless the conx plate or beam is continuous.

Edit to tack on a shop welded alt for the cantilevers.

 

[XR250]

At 40 ft-k, the flange plates will bow the walls of the HSS.

 

[dik]

At 40 ft-k, the flange plates will bow the walls of the HSS.

I'm in the process of determining if the flange plates work, including the axial load. Thanks... I may be back to my end plate solution.

 

[EngDM]

At 40 ft-k, the flange plates will bow the walls of the HSS.

How do you figure? Did a quick check against J10-2, J10-4 and Eq 9-29 and it's max is like 40% utilization under those, especially with the flange plate being nearly as wide as the column.

 

[XR250]

How do you figure? Did a quick check against J10-2, J10-4 and Eq 9-29 and it's max is like 40% utilization under those, especially with the flange plate being nearly as wide as the column.

My gut. 40ft-k / 12" beam = 40k load on the weak axis of a 3/8" plate

 

[EngDM]

My gut. 40ft-k / 12" beam = 40k load on the weak axis of a 3/8" plate

Gotcha. IMO should be fine after running the numbers.

 

[DrZoidberWoop]

My gut. 40ft-k / 12" beam = 40k load on the weak axis of a 3/8" plate

Depends on whether it's LRFD or ASD. I'm getting the HSS limit states to squeak by with LRFD. ASD, without a doubt, will have major problems and the lower flange plate would have to be made wider and the HSS split into an upper and lower section. Perhaps the "f" in "Mf" is some indicator I'm not familiar with.

 

[EngDM]

Depends on whether it's LRFD or ASD. I'm getting the HSS limit states to squeak by with LRFD. ASD, without a doubt, will have major problems and the lower flange plate would have to be made wider and the HSS split into an upper and lower section. Perhaps the "f" in "Mf" is some indicator I'm not familiar with.

It means "factored". Dik practices in Canada, so we use LSD (limit states design, along with the psychedelic) which is similar to LRFD.

 

[XR250]

I am not all that familiar with the checks for this situtation. Are you just checking the tube wall itself or the fact that the deflection of the tube will stress the crap out of the welds that connect it to the plate as the plate will not match the deflected shape of the tube wall?

 

[EngDM]

I am not all that familiar with the checks for this situtation. Are you just checking the tube wall itself or the fact that the deflection of the tube will stress the crap out of the welds that connect it to the plate as the plate will not match the deflected shape of the tube wall?

While you're probably right about the true behaviour of the connection, it's not something that is checked under AISC 360-16. Not sure if there is a check in 360-22 for that.

IFAIK you check the connected face for yielding and puching, and you check the side walls for yielding and crippling due to the point load (basically web buckling but for HSS). I'd imagine the deflection that the column face experiences is quite small before it starts yielding, so if you size for yielding it's a non-issue.

The yielding check is also based on an effectice width of the flange plate, so it assumes you are applying the load over a smaller area then what is actually provided which might be another reduction linked to your concern baked directly into the equation.

 

[XR250]

It would be interesting for one the FEA guys here to run this case to see the true joint behavior. I think we will be suprised by the results.

 

[dik]

How do you figure? Did a quick check against J10-2, J10-4 and Eq 9-29 and it's max is like 40% utilization under those, especially with the flange plate being nearly as wide as the column. The other issue I have is the effect of the transverse plates on adjacent faces.

I haven't finished my SMath program yet, but that was one of the concerns I had. I'll be looking at AISC as part of the work. I still have to include axial loading and clip angle design.

 

[dik]

Thanks gentlemen, it works.

 

[XR250]

I still don't buy it. I'm curious how much testing has been done?

 

[JAE]

I'm late to the party here but the photo that XR250 just posted is exactly what I wondered about -

As a stiff connector plate is pulled away from the tube face, the "force follows stiffness" effect is immediate - the center portion of the tube wall is not stiff at all perpendicular to its plane but as you approach the tube corners it gets much stiffer. All of the stress in the connector plate will flow towards its sides, where the tube is stiffest.

The welds at the ends of the connector plate will be stressed much more than the center portion of welds. This could create a rupture of the welds at the edges (thus XR250's photo).

Now dik's calculations may in some way account for this stiffness effect but it seems that this would indeed be a more complex behavior than a simple formula might cover - i.e. an FEA model accounting for the actual, unique stiffnesses between plate and tube wall.

 

[dik]

Thanks gentlemen... I'll do a FEM on it. Just peaked my curiosity. The HSS is 3/8" wall thickness and the transverse plate is BAR 5/8" x 6 with 1/4" fillet weld all around (I'checked with the fabricator to confirm he has the clearance to weld).