Continuous Weld Vs. Stitch Weld Calculation Help

[StrucPEng]

Hello All,

I am looking for a bit of confirmation on my process to make sure I am not missing anything (Please refer to attached detail). I am looking to reduce a continuous weld to a stitch weld to prevent warping during fabrication. Currently plate "PA" is welded with a continuous 1/2" fillet weld top and bottom for the full length of the plate. I would like to reduce it to the weld shown in the detail (1/2" fillet @ 3-8 w/ 4" cont. ea/end). The loads shown I have factored and are quite conservative. My process amounts to:

Calculating the required weld size and length based on AISC 14th Table 8-5 (Angle of 30 Degrees)
- D = 1/16
- L = 5.1"

Taking a ratio of these to what is originally specified
- (1/16) /(1/2) = 0.117
- 5.1"/43.5" = 0.117

This would mean I only need 11.7% of a continuous weld, but to be conservative would like to keep 37% of what I had (~ 3 x 11.7%) so 1/2" fillet 3-8 is what I decided on.

Is there anything I am missing or not considering here? I did not design the original weld so it threw me off to see how little I actually needed which made me question my work.

This is not a normal design for me so any confirmation that I am on the right track would be great. Thanks for the help!

Matt

 

[JAE]

Your download can't be viewed for some reason. Can you try to re-upload it using a pdf?

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[StrucPEng]

Sorry about that, this link should work.

 

[wannabeSE]

Have you considered using a 3/8” or 5/16” single pass fillet weld instead of the 1/2” which requires more than 1 pass?

 

[StrucPEng]

That is another route I could go also. Would you still expect to see distortion even with a single pass of that length? I would assume less than a 1/2" weld.

I would still like to know from a learning standpoint if my approach above is correct.

Thanks

 

[Bobby46]

Table 8-5 is for eccentrically loaded welds. It looks like your weld is not eccentrically loaded. I think you would just want to use equation J2-4 and J2-5.

In general you can substitute a larger stitched weld for a smaller continuous weld. I believe AISC has guidelines for minimum lengths and spacing of stitched welds that you will need to check (your dimensions seem reasonable). You will need to place the weld stitching so it is not eccentrically loaded.

Minimum fillet weld length should scale linearly with fillet weld size, assuming everything else stays constant.

 

[ldeem]

I just see your beam in the attachment so I can't see the calculations. What jumps out to me is how is the load being applied to the beam? From your drawing it looks to be applied out on the face of the plate. In that case you have a lot going on the transferring the load to the beam as well as torsional loading of the beam.

This is a strange arrangement and I would strongly encourage you to contact the EOR and find out what they were trying to accomplish with this design.

 

[dhengr]

M_S:
Instead of all the screwin around with tables and angles and percentages of existing welds, which you claim are way over strength, without any proof, why not just design the weld you really need? Given the great shortage of real design and engineering info. provided in the OP this looks like there may be problems other than a little weld distortion or weld sizing. Hanging 90kips longitudinal and 175kips vert. off of a .625” pl., at an eccentricity of 6”-7”, off of the flg. tips of a W12x79, with .735” thk. flgs. seems a little iffy to me, without a bunch more design info. and rationale.

Except for very basic welds and details, stitch welds are generally an inferior structural weld solution. They are difficult to make well and they are difficult to lay out and inspect. They are almost always over-welded because of this, since the welder does not want to be called out for too short or too small a weld, and the elder never really develops a rhythm. The ½” fillets are probably/usually 3 or 4 pass welds depending on the welder and weld process used, and all the starts and stops are difficult to make well and to inspect. They are certainly an inferior weld for fatigue conditions. Thus, the same amount of weld material can be applied in a continuous weld in less time and for a better result. Finally, the calcs. are basically the same since you are basically looking for kips per inch of weld over some length.

The weld distortion could be handled by making both welds at the same time and by fitting the WF and the side pl. together in a slightly deformed/cambered shape to counteract any sweep distortion which might occur. But, I suspect you will get as much distortion or deflection and rotation due to the eccentrically applied loads as you will due to well thought out weld sizing and sequencing.

 

[StrucPEng]

All,

Thank you for all the replies and your input, it has all been very helpful. I think my basic question in regards to stitch vs. continuous welding was answered. As for the other questions raised in the discussion I plan to go back and review my assumptions further. My gut is telling me to leave the weld as is and the additional points raised have given me greater confirmation of that feeling.

Thank you all again for your help!


Best,

Matt