Shear flow ? 1

[slickdeals]

Folks,

This may be a really simple, stupid question....

I have a situation where I need to connect 2 single angles (toe-toe) to create a U shape.

I think the weld should just be to keep the two members together, preventing them from separating. A skip weld maybe.

A single angle in bending unrestrained)....a case of two drunks leaning on one another? How do you design this?

 

[desertfox]

hi slickdeals

Whats the orientation of the two angles once welded in relation to the load path.
If it was me I would choose a continuous weld.

desertfox

 

[slickdeals]

Orientation of the angles is such that the two legs are pointing upwards (legs in compression)

Imagine a U shape, comprised of 2 angles. As if a Channel is bending in weak axis.

Loading is downwards (connected legs - tension flange)

One of the things I will have to check is local buckling in the outstanding legs.

 

[slickdeals]

This U shape will be carrying blockwork.

 

[Ron]

I don't think shear flow is an issue in the weld design. I agree with DF...make it continuous, but watch for warping from weld heat.

There is apparently a reason that a channel can't be used here?

Be aware of a strong potential for rust staining on the exterior (assuming exposed)with this design...don't let that come back to bite you in the a$$.

 

[slickdeals]

Thanks fellas, is there a "design" to this or just minimal "feel good" welding?

 

[Ron]

Assuming steel, the weld just extends the material to complete the section (considering near CJP weld)....can consider as if the weld did not exist and you had a rolled shape.

 

[Lion06]

Why not consider a bent plate?

To answer the question directly, though, as everyone else has said, there is no shear flow. You're just joining the sections to make a single U.

One thing I have done in the past is have the edges prepped for a CJP for the entire length. I provide a skip weld and then the spaces in between the welds get filled in with something like bondo or a metal filler with SS chips if you are using SS. The CJP should be ground flush so that when the filler is used you can apply a coating for rustproofing and you never know it isn't a single, solid CJP.

 

[paddingtongreen]

If there are wind loads on the wall, there would be shear flow in the weld, in the form of complimentary shear stress and at the maximum location.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[JAE]

paddingtongreen is correct of course in that with a lateral wind (or an internal 5 psf nominal load) you would have shear flow in the welds and a biaxial bending condition in the "U".

The wall/masonry itself may take a good portion of the lateral force as it might be significantly more stiff than the "U".

I always hate to have welds in bending so if there is any chance that the vertical legs of the "U" would move relative to one another, I'd also suggest occasional vertical stiffeners to tie the two vertical legs together a bit.

 

[slickdeals]

Thanks for your responses.

In India, elevator shafts require hoistways to be separated by 2 hour fire rated walls. There will be 100mm thick AAC block separation wall.

The contractor wants to use this U shaped steel member that will be set on the slab formwork on one side and bear on the shear wall (either in a pocket) or directly if the wall above is not cast.

This will also enable to attach the elevator guide rails to the sides of the U shape. Why a channel is not being used is beyond me.

The way I am designing this is such that the steel member will act only as formwork and there will be a 8" wide x 8" deep concrete beam spanning the width of the elevator. That way, the steel member will not need to be fire-proofed.

 

[Ron]

For lateral load, the weld will be at the neutral axis; therefore, the shear flow offset would only be the weld width, which will likely be less than 1/4".

I agree with JAE that the wall will take up most of the lateral loading, so the shear flow would be negligible and perhaps nonexistent.

Not a fan of using fillers for weld gaps. They tend to hold moisture and promote corrosion. Just use a seal (cap) weld if you only do a skip CJP weld.

 

[ToadJones]

if you assume these to angles are loaded independently and loaded the same, the load on the individual angles is likely not through the shear center. The angles will want to twist when loaded.

 

[steellion]

Use a bent plate or channel. It will be much cheaper than two angles joined with a CJP. Also, if it is not a CJP, legs welded in that manner are reminiscent of the Hyatt Regency failure.

 

[Lion06]

Ron,
Have you seen a lot of problems come from that type of detail? We've done that detail after getting push back from fabricators about the amount of weld and the architect wants to see what looks like a continuous weld. When we do that, we typically coat the entire thing with tnemec paint.

If that's a bad detail, though, it's definitely good to know.

 

[desertfox]

Hi slickdeals

Look at this link and go to design of welds 3.3.5

http://nptel.iitm.ac.in/courses/IIT...nections/3_welded_and_welding_connections.pdf

It talks about intermittent butt welds and fillet welds and indicates area's where these should and should not be used.
Personally I don't know what sections your considering but I would be tempted to weld the angles together with a backing strip.

desertfox

 

[hokie66]

slickdeals,

I don't like continuous welds on angle legs, as the whole thing is likely to distort. And it is a moot point in your case, as the angles are just for temporary support. Intermittent welds would be fine.

Don't you have concrete or masonry walls at both ends of this beam? Why not precast it and eliminate the need for temporary support?

 

[desertfox]

slickdeals

If there is any fatigue in the joint then intermittent welding is a non-starter:-

http://books.google.co.uk/books?id=...ion arising from intermittent welding&f=false

 

[BAretired]

The shear center of the combined shape occurs below the weld between the horizontal legs. For gravity load, shear flow varies linearly from 0 at the top of the vertical legs to a maximum at the corners, then linearly to 0 at the weld location.

The weld carries a continuous moment directed parallel to the member. This is necessary in order to prevent the combined section acting as two separate angles as suggested by Toad Jones.

Under lateral load, the shape has a tendency to rotate because the load is applied above the shear center.


BA

 

[Ron]

Lion06...if you prepare the surfaces and coat with a good coating like Tnemec to a reasonable dry film thickness ( > 5 or 6 mils)it should be OK. At one time, the practice was prohibited by code (AWS D1.1). I haven't checked it in a long time, so you might want to look into that.