Welding Deformations - Split T 1

[Al83]

Hi All,
I do fabricate split tee fittings for hot tapping operations. They give me a big headache due to excessive deformations although heavily supported before welding starts.
I roll those in 2 x half circles. One of these halves has the branch and flange welded on top. They are all thinner than 1.5in in thickness hence PWHT is not applicable.
Any idea of how to get rid or reduce deformations for this application? Is heat treatment after welding can be considered as an option? Or you reckon PWHT will still be a good option for thin plates?
I have attached a photo of what a split tee is.
Any suggestions, help or references/books will be highly appreciated.

 

[SJones]

Stress relieve the two half shells after fabrication (i.e. before installation) and pay attention to the welding sequence, particularly division of the long seam deposition.

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/in/drstevejones

All answers are personal opinions only and are in no way connected with any employer.

 

[r6155]

Al83
You sent us a drawing. Where is the photo you mentioned?
Give us more information: material, dimensions, etc.

Regards

 

[Al83]

Thanks SJones,
r6155, Assume the split shell is ID 42", length (through the page) is 72", Branch is 36" ID and flange is 36" 300#. Split shell thickness is 0.5in. Branch is also made from rolled from 0.5in thick plate and profiled to suite . Assume plates are ASTM A516 Gr70. Please see attached photos

 

[r6155]

Well now.
Where is the excessive deformation for you?

My comments:
Excessive fillet weld on nozzle to shell
Excessive lifting lugs (4) and excessive fillet weld
It may be that the rectangular plate before rolling is out of measure the diagonals.

Regards

 

[Al83]

Hi r6155,
What you see is a fitting that consumed days of jacking to realign into that beautiful fitting you have in the photo. However, I was after a way to reduce/eliminate that effort.
Yes, I agree, the welding of nozzle to branch is excessive indeed. It is a client design unfortunately.
Please see the attached sketch for how/where the deformations occur. Please note that we use half-moon plates and heavy supports on the inside and outside. However, excessive deformations still happen.
Not sure if chucking the whole fitting in an oven will help (stress relief)

 

[r6155]

Al83
1)The procedure should have been:
First roll a complete cylinder and tack welds in the longitudinal joint, then cut the hole and weld the nozzle, and ZERO lifting lugs (lifting lugs are not necessary).
Finally cut the second longitudinal, prepare the bevel and tack welds. Prepare for transport in one piece.

2)How much is deformation?

Regards

 

[Al83]

Thanks r6155.
Do you think the hole should have been cut before welding? What is the reason for that? (I want to learn). I thought cutting the hole after will provide more support to the shell. Maybe I am missing something essential.If there is a reference book/code I will appreciate very much.
Sizes range between 2"×2" to 66"×66". Most of the time they are not equal. Different sizes (i.e. 36"×24") and thicknesses between 3/8"in to 2.5"

 

[r6155]

I don´t know a book for that.
In the photo there are several things that are not right. Now I comment that it is not correct to support the flange like this, also the flange shown on the left.
This makes me consider that they are not working well. That's how things happen.

Regards

 

[weldstan]

Do you weld the entire branch (pipe and flange) to the shell? Did you weld the branch to the shell at the 12:00 position as shown? If either don't. Weld the branch pipe to the shell at the 3:00 position. Use back step welding technique for root and second passes, then complete the weld. Finally weld the flange to the branch.

I don't know why you roll the shell in 2 halves with 2 long seams and not roll the shell completely and make one longitudinal seam.

 

[0707]

PWHT depends on material composition and thickness. from what you said 38.1mm thickness is much, and PWHT shall be required!

During welding operations, if they were in the field, you have to be careful and protect your welding zone with a provisory shelter ovoid air and rain.

After welding it is important to involve your peace into ceramic blankets or rook wool to avoid distortions and of course, before welding starts , your peace shall be correctly supported.

luis

 

[gtaw]

Maybe I'm grasping at straws, but the material of construction appears to be stainless steel. Perhaps an austenitic stainless like type 304 or perhaps type 316.
If that is the case, weld the branch pipe before cutting the hole. Use a water spray on the ID while welding the branch to the half that is placed on the run pipe. It will not eliminate distortion, but the water spray will limit the base metal that is heated to a temperature above 300 degrees F. A thermal gradient of about 220 degrees F is all that is needed to reach the yield point of austenitic stainless steel. Ambient temperature + about 220 degrees F is the temperature at which the base metal reaches yield, so all the base metal that is heated to a temperature of about 300 degree F can be assumed to have reached yield. The base metal that has yielded is the cause of your distortion.

In conjunction with controlling temperature of the base metal, use prebending and restrains while the branch is being welded.

Just a couple of thoughts.

Best regards - Al

 

[DekDee]

Are you performing the installation on site as well ?
If so a couple of suggestions.
Cut the hole for branch.
Weld root run only.
Complete fill and cap of branch once attached to run pipe on-site (one welder welding sleeve to run pipe and one welder welding the branch at the same time.)
One lifting lug only, not four.
Hope that helps.

 

[Al83]

Thanks gtaw,
Great advice. The stainless steel is only an example, but vast majority is Gr70 steel. We make a lot of those a year.
Any specific advice for carbon steel?
Pleas see see attached photo for CS.

 

[Al83]

Thank DekDee
Unfortunately we fabricate only no installation services.
Do you think cutting the hole first will help if branch to shell weld is to be done at our shop? If that is the case, why if you dont mind asking?

 

[gtaw]

It wouldn't be a good idea to spray the underside of the "run" pipe fabricated from carbon or low alloy steels with water because of the potential problem with developing an undesirable microstructure in the HAZ. You can minimize the thermal gradient by preheating the entire piece of "run" pipe. Reduce the number of weld beads needed to fill the groove or deposit the fillet welds. Weld the groove and fillets before cutting the hole. Add restraints and prebend the "run" pipe if possible.

I like DekDee's recommendations, but you aren't installing the encircling sleeves. Darn!


Best regards - Al

 

[Al83]

Thanks all,
I guess preheat and rolling in full encirclement before splitting the tee is the answer.
Time to start reading about how to preheat.
Thanks all, you r legends

 

[LittleInch]

What you have here is a more extreme version of branch welding pipes in situ.

By fabricationg a complete pipe with some fairly decent welds in the split are you can probably contain the springing out of the bore, but the axial bending is harder.

I once had welded up a thick pipe, but then large "weldolets" at maybe one size down, 12 x 10 or maybe 10 x 8 with about 4 in a row. The thing was visibly bent like a banana. Put me off large weldolets at this sort of size difference.

The only way they could get something decent was to amend the weld spec to reduce total heat input and pause between runs to allow it to cool down, but not so far that it cracked.

So it looks to me like a mixture of welding procedure modification to reduce heat input plus some experimentation with PWHT to releive the stresses but all while doing it with a round pipe, not half shells.

Of cource you still might get some springing when you cut away the two parts, but might be able to mechanically recover these before shipping.

Let us know how it goes.

Could you not try and make them like a tee instead of all that welding?
Or use integrally reinforced fittings?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[r6155]

Can you send us a complete drawing of bevel dimensions?
The WPS and welders are qualified?

Regards

 

[Al83]

Hi LittleInch,
Thanks for your reply. Unfortunately most of the times the OD of the branch is larger than a standard pipe, and mind you the shell is set to fit on a standard pipe so it has a larger OD too. So splitting a standard tee is not an option, it just has to be made from scratch.
But I totally agree that allowing interpass temperature to cool down between welds, rolling a full circle, and PWHT would improve things a lot.

r6155, i will upload drawing soon. Thanks a lot.