ASME VIII-1 MAWP and MAP calculation for flanged cover API 661 air-cooled

[FPPE]

Hi,

I would like to understand if it is possibile to obtain the MAWP and the MAP for a flanged cover type air-cooled API 661, with ASME VIII-1 as design code.
Generally, when air-ccoled has the flanged cover instead of plugsheet, we perform the finite element analysis in accordance to ASME VIII-2 because (as per our know-how) it is non correct to design a flanged cover type air-cooled only by formulas. Or more specifically, since ASME VIII-1 has no formulas for rectangular flanges, the best way is to proceed with a FEA.

Now, considering that some of our customers require the hydrostatic test pressure calculation in accordance with UG-99(c), which requires the MAP calculation, how we can obtain the correct value of MAP (or even the MAWP) with a FEA? Theoretically it should be possible, but in practice we have to iterate the FEA increasing by few decimal the design pressure and thsn perform the stress linearization until that the air-cooled does not satisfy the equations.

How do you approach design of such a case?

Thanks

 

[TGS4]

If you are performing a linear-elastic FEA, then the result scale linearly. Should be very simple to do. However, I would follow the rules in 5.2.2.5, noting that the only categorization that is required in the pressure test condition are the limits listed in 4.1.6.2. And noting that you are performing this analysis with respect to a VIII-1 vessel, pay attention to the note in 46-3(d), which also refers back to 4.1.6.2.

 

[r6155]

See UG-34 for rectangular flange.

Regards

 

[FPPE]

r6155, I believe that UG-34 is not applicable to the case mentioned.

TGS4, thank you very much for the detailed answer.
With reference to 46-3(d):
"Evaluation of the stresses during the test condition of Division 2, 4.1.6.2 is not required. However, such calculations may form the basis of a calculated test pressure as described in UG-99(c) or UG-100(b)."

This literally says that the analysis under test conditions should not be performed, right (if the calculation code is Div. 1)?
This confused me, how come when I fill out the mechanical calculation report of an equipment in VIII-1 using VIII-2 for FEA the analysis under test conditions is not required?

In the specific case where calculation is required for UG-99(c), then it would become necessary to evaluate the stresses for 4.1.6.2, if I understand correctly.

However, the basis for UG-99(c), which is MAP, should still be obtained from the calculation under design conditions right? So from what I'm understanding, 4.1.6.2 would not be required anyway.

Going back to the "then the result scales linearly" does it mean that, taking the membrane stress as an example, if it turns out to be 90 percent over the limit imposed by the code, the MAP will be 100 percent, so equal to the limit itself given by the equations? But in an analysis where the only load is given by pressure, thus excluding loads on the nozzles and others.

 

[r6155]

More information: design pressure & temperature, material, number of tubes, ..etc.

Regards

 

[TGS4]

FPPE - there is no maximum pressure test limit, as noted in UG-99(d). Anything above and beyond that (for example, a limit imposed by an Owner's requirements) is contractual and not based on the Code. You would need to obtain guidance from those requirements.

I don't understand you differentiating MAP vs MAWP.

Nevertheless, if your calculations during design show an MAWP results in a value of Pm or Pm+Pb of a certain value (these are going to be hand-calcs, not from an FEA), then the value during the pressure test will be 1.3*LSR*(that stress value). And, if you are using Appendix 46, that's what you will get.

As far as what loads result in Pm and Pm+Pb stress categories - I'll leave that to your engineering judgement.

 

[r6155]

I don't understand you differentiating MAP vs MAWP

Please see the following, I guess this is what OP says when corrosion = 0
MAP= new and cold

UG-99 STANDARD HYDROSTATIC TEST
(c) A hydrostatic test based on a calculated pressure may be used by agreement between the user and the Manufacturer.........


3-2 DEFINITIONS OF TERMS
calculated test pressure: the requirements for determining the test pressure based on calculations are outlined in UG-99(c) for the hydrostatic test and in UG-100(b) for the pneumatic test. The basis for calculated test pressure in either of these paragraphs is the highest permissible internal pressure as determined by the design equations, for each element of the vessel using nominal thicknesses with corrosion allowances included and using the allowable stress values given in Section II, Part D, Subpart 1 for the temperature of the test.

Regards

 

[FPPE]

r6155, correct, this is exactly what I mean for MAP vs MAWP.

TGS4, I have always interpreted the UG-99(b) VS UG-99(c) formulas this way:

UG-99(b) classic formula: 1.3xLSRxMAWP(or design pressure. LSR calculated for each component, lower governs.
UG-99(c): 1.3xMAP(of each component)-hydrostatic head (of each component). Lower value governs among all components.

I try to explain my post better:

In a design by formula the MAP is obtained by simply making explicit the formulas for each element to find the pressure considering zero corrosion and nominal thicknesses at ambient temperature. That value will correspond to the MAP for each element as defined in 3-2. Obviously flanges and tubesheets will need an iterative calculation provided in the calculation software to find the correct value. It is unclear whether calculations of openings should also be taken into account when determining MAP.
The MAP found will come for each element then be used in the UG-99(c) formula, for example 10 elements will correspond to 10 equations, the lower value among the 10 will be the vessel test pressure.

Considering this, how can I determine MAP if a component (in this case the flanged cover header box) is designed via FEA?

 

[r6155]

@ FPPE
I insist
More information: design pressure & temperature, material, number of tubes, ..etc.

 

[FPPE]

r6155,

Design pressure: 3.45 barg
Design temperature: 149 °C
Header boxes material: SA-516 Gr. 60
No. 231 high finned tubes in SA-179

Why do you need these information?

 

[r6155]

@ FPPE
I need it to understand the situation about design, MANUFACTURING and INSPECTION.
Now, for 3.45 bar there is no need to get so complicated with the calculations

Regards

 

[FPPE]

I absolutely agree with you, but unfortunately our customer does not think the same.
Often there are not many ways to explain to customers things that really make sense, and you just do it to save time.

Regards

 

[TGS4]

FPPE you are asking an interesting question. But, it's not quite the right question to ask.

Take another look at how VIII-2 calculates the maximum-allowable test pressure. It's given in 4.1.6.2. Whether you use FEA or not (and check the wording in 5.2.2.5) you are going to use essentially the some approach as described in 4.1.6.2. In effect, you are going to back away from the specific details in the FEA and back out to hand-calc level work. In your geometry, that's not always easy. And, as such, you may end up needing to skip the elastic approach and go with a Limit Load or Elastic-Plastic approach.

Don't overthink this, though. And certainly don't be modifying any geometry to accommodate a pressure test.

 

[FPPE]

TGS4 thanks for the clarification, I will follow these paragraphs.

I had not paid attention to the sentence "In the case where the stress limit is exceeded, the selected test pressure shall be reduced, but no lower
than the minimum test pressure as established in Part 8".

Unfortunatly ExxonMobil GP 05-03-01 requires calculation in accordance to UG-99(c), which often lead to increased thicknesses because the equations are not satisfied under test conditions.

 

[r6155]

Don't overthink this, though. And certainly don't be modifying any geometry to accommodate a pressure test.

I don't agree, maybe you want to say another concept.
"Design for Inspection - Evaluating the Inspectability of Components in the Early Stages of Design".

Regards

 

[TGS4]

FPPE - I recommend checking with the company - I don’t think that the intent is what you think that it is.