fabricated 24" flange 1

[Jay_]

hello all,

I am required to design a 24" flange according to ASME sec 8 div 1 appendix 2 under an operating pressure of 16 bars.
I am designing it because of the lack of this flange in our markets, especially that it's SS316.
Anyways i am going through the code and i realize that there is no way to design an optimized flange according to the code. the tangential stress is just way too high.
And oh i am assuming that the flange to be loose slip on type without a hub.
So i decided to perform some FEA analysis, and the result came a quiet shock to me.
all of the forces were calculated according to the code and put in FEA analysis, and the result that the von-mises stresses were as high as 146 MPa for a 20 mm thickness flange.
But i didn't stop here, i added stiffeners on the back of the flange welded to the 24 inch pipe, and the stress were as high as 80 MPa.
As per the code the tangential stress was 1380 MPa.
Did anyone before encounter such a confusion?
Regards,

Detailing is a hobby,

 

[r6155]

20 mm flange thickness is very thin.

Regards

 

[SnTMan]

What is an "optimized" flange? I've never seen an Apx 2 flange that the stresses could not be brought to acceptable values by simply making the flange thicker. 16 bar is not an especially high pressure. You are doing something wrong.

Regards,

Mike



The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[Jay_]

Dears,
I checked my manual calculations in PvElite and it did give the same results,
And i do know that 20mm is very thin. But i’m looking for an inexpensive ways to design the flange.
I am using ASME code just to calculate the forces on the flange due to operation and seating conditions.
I am taking into consideration the stiffening method of the flange face, and by that i mean adding one stiffener (trapeze form) to the back of the flange in between two consecutive bolts (28 Nos). It just make sense to use 20mm stainless steel flange stiffened by some cheap 15mm carbon steel plates. This detail is not so included in the code and i’m having troubles going through the calculations.
Do you think it’s worth a try? Or should i just stick to the conservative results of the code?
It can save lots of money.
Any advice; please.
Regards,

Detailing is a hobby,

 

[r6155]

You continue to design incorrectly. Consult a pressure vessel engineer.

Regards

 

[SnTMan]

Do whatever you want.

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[MJCronin]

You are not going to be designing any bridges in the future, are you ???

MJCronin
Sr. Process Engineer

 

[JStephen]

I would think that applying FEA to a flange could lead to highly variable results, depending on the assumptions made and the amount of detail incorporated into the analysis. I would not put a lot of faith in the results unless it was done by somebody that routinely did that kind of work.
If adding stiffeners, you may run into distortion of the flange face from welding, making it harder to bolt up.
If this is a non-vessel application, you might look into AWWA C207 flanges which are plate-type flanges considerably thinner than 150# flanges. Typically used for waterworks applications, never for steam or petrochemical service. Also, normally flat-face with full-face gaskets.

 

[Jay_]

thank you JStephen for your helpful input,
it's a pressure vessel at a high pressure and at 95C.
and the flange i'm aiming to work myself around is simply a manhole like opening for the heat exchanger tubes.
i know how to calculate the stresses as per app 2, but i don't know how to consider the stiffeners plates in the calculation, i have never done this before and the owner wants to go with a cheap solution which i don't really approve on.
in my my perspective the stiffeners plates are playing a role in increasing the thickness of the flange depending on their dimensions. but i can't and i don't want to assume that it's increasing the flange thickness by x amount. and for the distortion, we always machine the face of the flange, in this case i assume it will need at least 5 mm.
any inputs?

Detailing is a hobby,

 

[SnTMan]

...but i don't know how to consider the stiffeners plates in the calculation,...

You DON'T because you CAN'T under Apx 2. You are into U-2(g) territory (look it up).

A complicated welded assembly is rarely less expensive than a simple non-welded part. Cost it out. Don't forget fabricating all the parts, welding, clean-up, NDE (non-destructive examination).

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[BJI]

For pressure equipment work, stiffeners on flanges should not be used. If you are trying to come up with an economical design why would you treat it as a loose ring, this will always require a thicker flange and it wouldn't be surprising that the tangential stresses are so high at 20mm in that size (CL150 hubbed lap flange is 48mm thick). This is one of the easiest formulas to verify, so I would expect that you have over-restrained your FEA model and/or applied the loads incorrectly (which also means all subsequent analysis is incorrect). Design it as an integral flange and play around with the proportions to get the most economical design, remove stiffeners from the equation. You could maybe look to use a c/s lap flange with a stainless steel stub end but I doubt it would be much cheaper than fabricating out of stainless, this is mostly beneficial for higher end materials. I would be surprised if a grade 316 weld neck was not market available.

 

[r6155]

May be the stiffeners interfere with the torque wrench.
Machining after weld the stiffeners shall produce more distortion of the flange. PWHT is necessary to avoid distortion after machining.
Try with standard carbon steel SORF with weld overlay or clad SS316 in gasket surface. Same concept with blind flange.

Regards

 

[LittleInch]

More smaller bolts instead?

Have you looked at ASME B 16.47 type B flanges? a 26" is about 40mm thick, but 1.5 times the number of bolts of a type A (36) and only 31" OD for a 26" pipe

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