Dish end repair 4

[ukmet]

We are cold forming ellipsoidal dish ends. The problem i am facing is that large dish ends which need to be formed with a joint in it, produces cracks in it after forming. The cracks majorly arise at knuckle radius at joint and adjacent to joint. The material is SA-516 Gr 70 normalized. What could be the possible reasons? And is it ok to repair the cracks by grinding and buildup especially when they are at the both side of dish ends? Any precautionary measures before repair?

 

[sharmass]

This is a very common material and you should not have any problem.
Ensure that the weld seam is flush ground and that there are no undercuts as they may be the starting points for cracks.
Also ensure that the speed of knuckle forming is not too fast

You can always repair the crack and RT it to ensure that the complete defect has been removed.

You should have given additional information like thickness, diameter and post forming heat treatment

 

[metengr]

For this Grade material and being normalized versus normalized and tempered, the cold forming may be exceeding the strain capacity in the weld region and in the knuckle radius region of the head. If this is the case, you need to form using heat versus cold forming to ensure adequate ductility.

 

[chicopee]

Knuckle is highly strained during forming, so I would definetly talk to the AI about the cracks, after-all we seem to be talking about PV's with dished heads.

 

[racookpe1978]

How many times do you roll the heads?

Are trying to make the complete dish in one pass? Two? four or more?

Between passes, or after the second pass and before the 3rd, did you re-anneal the head?

 

[ukmet]

We are not doing any post forming heat treatment. Thickness varies between 14mm to 22mm max.

Metengr: Can you please further elaborate?

Racook: By pass you mean flanging of knuckle?

As thickness is not much so we do not go for any heat treatment process

 

[metengr]

You may need to hot form instead of cold form, in other words heat the plate to the proper forming temperature. The material was normalized and this may have limitations on formability with what you are trying to accomplish.

 

[Duwe6]

By 'pass', the meaning is allowing the material to rest a bit, and then complete the forming. Sounds like you are forming in a single pass.

 

[racookpe1978]

22 mm thick?

Yes, your experience is showing that is too thick for a single pass (once-through the machine) roll forming for the elliptical head diameters you are trying to roll.

So stop doing that. It is not working, you HAVE to change your process!

 

[ukmet]

Thank you. Yes we are doing in a single pass. The process is that we first make the crown radius on a hydraulic press. After that the doomed dish is send for knuckle formation on flanging machine. Please guide on how many passed it should be done? And how much break to be given in individual passes. 22mm is max. thickness. We are also manufacturing 14mm, 16mm, 18mm depending on pressure and diameter requirements.

metengr: we do not have facility for hot forming.

 

[metengr]

Going back over your post, I believe you may need to first PWHT the weld joint prior to forming the dished head. After crown forming, stress relieve (either in your shop or send it out to a heat treater) the entire head per UCS-56 requirements. After this operation is completed you can cold form the knuckle. Have you verified the fiber elongation strains? Because after final forming, this may require thermal treatment of the completed dished head to rid of cold forming strains, per code.

 

[ukmet]

No sir I have not calculated the fiber elongation. I have seen its method in ASME Sec VIII Div.1 can you elaborate?

can you please elaborate the process for forming for our ease.

 

[GenB]

I do not do any forming on SE heads, but upon inspection, I have rejected several heads cold and hot formed to a major supplier-head mfr due to small cracks not noticed to the naked eye.
in this case Ukmet may have a bigger problem if he's mfg heads w/o proper knowledge,
specially if segmented head, which may require preheatbefore welding and PWHT before forming and hot forming.

 

[metengr]

ukmet;
The correct fiber elongation equation must be selected in ASME Section VIII, Div 1 after forming to verify that forming strains have not exceeded a critical value. You can use other methods like marks before and after forming to calculate the tensile fiber strain. In your case, I would suspect formation of the SE head will require thermal treatment because you would exceeded the 5% limit.

This treatment is necessary to avoid work hardened material that can be susceptible to stress corrosion cracking or service-related cracks, like low temperature creep cracking.