weld designation on back to back channels 5

[shaneelliss]

I want to weld two steel channels back to back to hold them together for gravity loading. I don't think there is really a strength requirement but I want to keep them together. I was thinking of using a stitch weld (say 3-6) along the seam at top and bottom, but I am not sure what type of weld to designate. Is this a square groove weld symbol?

 

[holdpt]

Different channels have different radii (rolled heels). Our shop will use the AWS D1.1 flare groove that states differing weld sizing for various radii and welding processes. (Pg. 85, 2010)

 

[shaneelliss]

I am not sure I understand your meaning holdpt. There is no radius at the back edge. I want to put these back to back and weld along the interface of where the webs meet at the top and bottom. It is like a butt weld but doesn't need to go through the full material thickness because it is a 10" channel. But I am not sure how to call out this weld.

 

[holdpt]

Inconsistencies of the radii have caused us problems for other reasons than welding. If there is no radius to speak of, then my opinion is you can only refer to the square groove.

 

[shaneelliss]

So is the image of the weld callout shown correctly?

 

[dik]

I usually spec it as a butt weld... two parallel lines on the welding symbol.

Dik

 

[gtaw]

As shown in the preceding sketch, the weld is a square groove, welded from both sides (two arrows), and the required joint penetration is complete. I'm not sure that is what was intended.

I suggest AWS D1.1 be reviewed to determine the minimum weld size required based on metallurgical considerations. The back of the channels can be beveled to ensure the depth of the groove is sufficient to provide the required joint penetration dictated by AWS D1.1. The actual joint penetration is a function of the thickness of the base metal at the location where the welds are placed.

Best regards - Al

 

[shaneelliss]

I don't really need it or want it to be a CJP. I just want them connected with some weld. It is not required for strength. What is the best way to show that? Should I leave it as shown in the sketch, but arbitrarily add a 1/4" size to it, or what?

 

[gtaw]

Replace the square groove symbol with a V-groove. To the left of the "V" symbol place the numbers (5/16) indicating the required joint penetration is 5/16 inch.

If you are the designer, not the fabricator, you can simply place "(5/16)" under the reference line indicating the required weld size (joint penetration). The fabricator would be responsible for the selection and detailing of the actual partial joint penetration groove.

Best regards - Al

 

[shaneelliss]

Thanks gtaw. I am the designer but obviously not a weld expert. In school the only thing we ever talked about was fillet welds. Just trying to figure out the rest as I go along.

But to make sure I am clear, are you saying that there is there no way to weld this without a bevel? I can't just set them back-to-back and weld along the top and bottom without beveling or gouging the channel first?

 

[gtaw]

As a ruler of thumb (which means it doesn't really work) is to assume 0.001 inch of joint penetration per ampere. According to the "rule", you should achieve 1/8 inch of penetration if you weld with 125 amps. The truth is, you will not get that much. Some form of preparation ensures, or at least help ensure, the weld size is what the designer expects it to be. One would rather have more weld than necessary rather than find out there wasn't enough to hold the parts together.

Best regards - Al

 

[3DDave]

You could put straps across the top and bottom or just a continuous strip and use fillet welds.

The reason for putting in a groove is that you can be certain that the parts are actually welded together. Without weld prep you can get anything from zero penetration to an unknown amount and they will both look the same.

The original weld concept is what I'd call a hope-weld, as in you hope it is welded.

Don't hope. Be certain.

 

[MintJulep]

You really need to rethink this.

You're putting welds of unknown penetration and thus unknown strength at the highest stress location of the beam.

And intermittent to boot. If there is any cyclic loading this is a certain recipe for fatigue cracks.

 

[tbuelna]

If all you wish to do is keep the two channels aligned under 1g static loading, then a double bevel stitch weld along both sides will be sufficient. The bevel depth probably only needs to be about 1/4", and 1" of weld per foot should provide more than enough strength. The only other consideration is whether the weld profile can be crowned or must be flush.

If the channels are long, you might also want to provide instructions to the welder as to the sequence the welds should be made. With long parts, the weld sequence should start at the center and alternate side-to-side and end-to-end, in order to minimize distortion.

 

[gtaw]

I don't know your application, but you should consider reviewing the requirements of AWS D1.1 Structural Welding Code/Steel.

Best regards - Al

 

[BillPSU]

Why don't you just specify an I-beam and stop welding all together?

Bill

 

[dhengr]

Shaneelliss:
If your loading condition distributes the load to both channels equally, no weld would be required, except that you want the two channels to brace each other and act in unison. You have to use some engineering judgement in determining what connecting weld is required. If your load is applied to only one channel, and you expect it to transmit half that load to the second channel, then the weld must transmit that shear flow to the second channel. In the first case, I think I would grind or nibble a .125 or .1875" chamfer on both heel corners of one channel and fill it (concave) with weld, paying some attention to sequencing the welding to control final camber. If there are concentrated loads or real load transfer needed then some weld design is needed.

 

[shaneelliss]

Thanks all for the comments. I am doubling a channel because one channel is already in place but it is not enough, by about half. So I am simply adding an identical channel next to it. Some steel grating is resting on the channels and theoretically each channel will take half the load and they don't need to be connected at all. As dhengr points out, I just want to connect them to brace each other and to act mostly in unison. I think the take-away I am getting here is that I can't really just have them do a surface weld at the seam, that I must have them create at least a one sided bevel for even the small welds I would like to see.

I just looked up AWS D1.1 and the cost for that shocked me a bit. And I thought the AISC steel construction manual was expensive.

Thanks again.

 

[Duwe6]

You have it exactly:
A) Without some bevel, 'face' welds have too little strength to use in a design load.

B) AWS D1.1 is a Code Book and thus quite expensive, just like the rest of 'em.

 

[MintJulep]

Punch some holes in the web of your new channel and ring weld or plug weld to the web of the in-place channel.