Yield strength vs Collapse pressure

[metboss]

Hi all,

We have ordered Side Pocket Mandrels (Quantity: 50 No’s) with monogram & testing to API 1G1 V1 F2.
As per API 19G1, for F2 grade, 5% of ordered quantity shall undergo external pressure testing at 5,000 psi.
In order meet F2 requirement, our manufacturer carried out this test on 3 No’s (5%) of Mandrels and one of the mandrels was found to be collapsed at 4,000 psi during the test and consequently, this mandrel was scrapped.
Vendor carried out the root cause investigation and found that yield strength for raw material (API 5CT L-80, 80,000 psi yield strength grade) is found to be less than the minimum.
i.e. 72 ksi (tested & measured value) below the minimum required limit 80 ksi (API 5CT).

After this, remaining all 24 Mandrels were re-heat treated to achieve 80 ksi yield strength and subsequently, again 3 No’s (5%) of Mandrels were externally pressure tested at 5,000 psi and successfully passed.

My questions are :

1)Can we ask vendor to perform 100% external pressure test on all 24 No’s as a penalty & to ensure integrity of all mandrels?

OR 5% ( 3No’s) would be suffice to assure the integrity of remaining quantity ?? It has to be noted that API 19G1 does not call for any penalty in case of failure if any of external pressure testing.

2)Does yield strength under value (below 80 ksi) really attribute to external pressure testing failure (collapse) ?

3)What are factors to be considered to mitigate such failures in future ?


Regards

 

[MJCronin]

Could you post a picture of these devices ?

Based on your question, I believe that you may have posted in the wrong place. Consider posting in either the Piping or Pressure vessel forums

MJCronin
Sr. Process Engineer

 

[metengr]

1)Can we ask vendor to perform 100% external pressure test on all 24 No’s as a penalty & to ensure integrity of all mandrels?

Sure you can if you are willing to pay for it. This should have been spelled out in the PO or contract documents. If not, your expense for your piece of mind.

OR 5% ( 3No’s) would be suffice to assure the integrity of remaining quantity ?? It has to be noted that API 19G1 does not call for any penalty in case of failure if any of external pressure testing.

The vendor re-heat treated the 24 remaining, tested 3 (5%) and passed. I don't see what more you can do without added expense. Did you verify the re-heat treatment, charts or data?

2)Does yield strength under value (below 80 ksi) really attribute to external pressure testing failure (collapse) ?

Possibly, it depends on design margin.

3)What are factors to be considered to mitigate such failures in future ?

Get the heat treatment correct.

 

[redpicker]

A few comments.

As far as asking them to pressure test all 24, sure you can ask. Whether or not they comply with your request is another matter.

From an order quantity of 50, they tested 3; one failed and now they have 24? I'm a bit confused.

Yes, plastic collapse is dependent on the material yield strength. There are a lot of other variables involved (OD, wall thickness, concentricity of OD to ID, just to name a few), so there is generally a lot of "cushion" built into the required collapse testing pressures. You can ask your vendor to show in their analysis where a 10% reduction in the yield strength would result in collapse. In general, APIs collapse equations are pretty conservative.

You can ask your vendor to show in their root cause analysis what in the heat treatment was the cause of the low yield strength and how they avoided this with the re-heat treatment. The assumption is that they originally believed the heat treatment to be correct. Why do they now believe the re-heat treatment to be correct?

The only way to be sure every piece is OK is to test every piece. It usually isn't practical to test every piece. The purpose of testing a sample is just to demonstrate the lot meets the requirements, and consistency within the lot is dependent on process control. Does your vendor make enough of these to have a consistent process and proper controls in place? Are they going to stand behind their product if you have problems running it? If they are a large supplier, these answers are probably "yes" and they will have taken steps to insure what they deliver meets the requirements. If they are small, maybe not. Can they demonstrate that their process control is adequate to insure consistency?

Finally,

If they are not willing to pay to pressure test the whole lot, are you willing to pay for it? If not, it must not be that important. If it is that important, perhaps you can work a deal where they would be willing to discount future purchases for the amount of the testing required to satisfy you of the quality of their product.

 

[mcguire]

Check the concentricity. It has the same influence as yield strength.

Michael McGuire

 

[racookpe1978]

Look very closely at the piece that failed: Porosity, machining tolerances (eccentricity as noted above), or too deep or shallow? Too much diameter? Bad gasket face or gasket depth - if they did not say exactly where and why the piece failed, and cannot show the position of the failure and why, then you don't know whether the assumed better metallurgy in the re-heated pieces is the real solution.

 

[metboss]

Thanks to all of you for your valuable advice.

I'll convey my comments to our vendor.

Regards