Nozzle Loads 3

[mechmike]

I have a situation where a vendor has submitted calculations for WRC 107 nozzle load calcs done on Compress (Codeware) software. The vessel is designed for full ANSI 150# flange rating. The WRC 107 loads pass, however under the "add a flange" feature in the nozzle design menu Compress has a check box to "consider external loadings (WRC-107) on the flange MAWP rating" The results from calculating this equivalent pressure derate the flange to the point where I need to upgrade to a ANSI 900# flange to meet my design pressure. The program does not indicate the methodology used or show the calculation it used to arrive at this deration. Even more confusing is that our own CEASAR II flange leakage calcs show that the 150# flange should be more than adequate under these design loads. Any ideas on how Compress calculates this or what may be wrong in the input to cause this discrepancy.

 

[jte]

mechmike-

I don't know the details of Compress or Caesar, but a frequently used formula for determining the equivalent pressure would be:

Peq=(4F/pi*G^3)+(16M/pi*G^3)

Where F is the axial load, lbs; M is moment in in*lb; G is the gasket reaction diameter in inches.

This formula is listed in my 1992 edition of Section III ND-3441.10 as formula (3) in addition to some other sources.

jt

 

[unclesyd]

First off I would check my inputs and then call Codeware about the problem and I do believe it's a problem if the input data is correct. You might expect a callout for a Class 300 with a more rigerous analysis but I would never expect a jump to Class 900.

I tried to call the other Sidney, our company's Compress guru, but he out rebuilding Cushman Motor Scooters of all things.
There are several people apparently from Codeware that reguarly post on the forums so maybe one will step forward.

 

[prex]

Make sure that your nozzle loads do not include the pressure end load when you calculate the flange, as this is of course taken care of in the flange calculation.
I would also not consider loads of thermal origin in the flange calculation, only mechanical loads (those that derive from piping self and sustained weights).
If find that the consideration of pressure end loads and the classification and combination of other loads is unclearly managed by softwares and codes dealing with nozzle loads (and by many client's specifications, that define the loads without mention of their origin).

prex

http://www.xcalcs.com

Online tools for structural design

 

[codeeng]

I suspect Compress derates B16.5 flange ratings for those additional piping loads. This is likely due in part to the inconsistency between Appendix II calcs and B16.5 pressure/temperature ratings (you'll often find an Appendix II calc will fail a B16.5 rating). So if you're max'd out with the flange rating and then apply just 1 pound load from your piping the design will fail. For this reason it is not practical have the vessel MAWP limited by the flange rating. Jumping from 150 to 900# is not the right answer either. Better to switch off the external load check box and let your Pipe Stress group check the loads at the flange using Caesar.

 

[pipesRus]

There is a similar formula to the one JTE mentioned for ASME NC, NC3658. ADLPIPE will automatically calculate the Peq value if you specify the value of G whilst defining the flange. You can then rate the flange for the internal pressure and the effects of the mechanical applied loads.

 

[TomBarsh]

COMPRESS uses the method described above by to determine the equivalent pressure due to external loads on the nozzle. This method is described in "Pressure Vessel Design Manual", Second Edition, by Dennis R. Moss. Appendix G, page 273, as well as other references. The formulas from ASME Section III-1 NC-3658.1 are essentially the same.

Moss presents the formula for equivalent pressure acting on the flange as:

Pe = 16*M/(pi*G^3) + 4*F/(pi*G^2) + P

As described by original poster, COMPRESS optionally reduces the pressure-temperature rating of the standard ASME-ANSI flanges by the equivalent pressure resulting from the external moment and force. COMPRESS determines the equivalent pressure from the external loads as:

Pe = 16*M/(pi*G^3) - 4*F/(pi*G^2)

Note that in the COMPRESS formulation:

'G' is taken as the average of the bore and the raised face (This point could be debated since G is a function of the gasket dimensions. But for standard flanges there is little point in complicating the issue by requiring the designer to enter their..likely unknown..gasket dimensions)
'M' is the vector resultant of the applied bending moments Mx and My
'F' is taken per the WRC-107 sign convention as positive into the shell; thus the term due to F (P load in WRC-107 nomenclature) is subtracted from the term due to bending moment. The positive inward load F acts in direction of the internal pressure.

COMPRESS conservatively limits the equivalent pressure to non-negative values so that the flange rating doesn't get increased over its basic pressur-temperature rating.

 

[TomBarsh]

How reasonable is the change from Class 150 to Class 900?

Assume the design is for 125 psi at 650° F. This maxes out an A-105 B16.5 Class 150 flange. Rating for A-105 B16.5 Class 900 flange is 1610 psi. The difference in ratings is 1610 - 125 = 1485 psi. How much external load does this equate to?

For a given size flange there are 3 load variables: Mx, My, and F. Thus there are any number of combinations of load/moments that result in the 1485 psi equivalent pressure.

Say this is an 8" flange and let G = 9.3". Then a possible combination is:
Mx = 228,431 lb-in (19,036 lb-ft)
My = 228,431 lb-in (19,036 lb-ft)
F = 38,072 lb

Those are indeed pretty husky loads, but nothing more than what I have seen specified on some vessels. How reasonable is it to ignore loads of this magnitude when selecting the flange rating?

A question came up recently at work that is typical: software complained when the designer specified 48,000 lb-in bending moment on a 2" nozzle with Class 150 flange. Pe for this moment is 10,616 psi versus the 125 psi rating for the flange. So...

A recent thread brought up the issue of how external loads may affect the flange performance, it provides some additional viewpoints: thread794-123212

 

[StressGuy]

4000 ft-lbs on a 2" nozzle? I'd say that system has some serious problems. Even those loads you reported for the 8" nozzle are higher than anything I've ever seen in that size.

I wouldn't be surprised if you have a hard time getting your nozzles to pass with those kinds of loads. I would have a serious talk with your pipe stress guys if those are the kinds of loads they're giving you. Something is definitely wrong there.

Edward L. Klein
Pipe Stress Engineer
Houston, Texas

"All the world is a Spring"

All opinions expressed here are my own and not my company's.

 

[arto]

Sometimes you are given forces & moments based on pipe yielding [because the specifier doesn't want to design the piping?] which result in humongous flanges if you treat them as real values using the Kellogg method[& high shell stresses].

Similar to the vessel/pump maker to specifying "zero forces & moments" on their equipment.



If they're going to be arbitrary, how about "realistic arbitrary" - something like Force = xx# per inch of line size, or Moment = XXXin# per line size squared?

 

[TomBarsh]

Arto, you have hit the nail on the head!

 

[pipesRus]

Arto, I agree. Where do you buy glass vessels/pumps that take zero forces and moments?

I have used a pipe yielding technigue for designing pipe supports before, but would be supprised to see it used to rate flanges.