Flange design with external loading 2

[Spoonful]

Hi All,

One problem we have here is we have a vessel designed to 10.2Mpa at 38 degree C, fitted with CL600 A105 WN flange. You may notice this vessel is designed at max flange design pressure as per B16.5.

Now the problem is that client want to increase the design temp to 60 degree C. Where according to B16.5 CL600 A105 flange is only rated at 9.3 Mpa.

I have done my calculation as per AS1210 for flanges(very similar calculation to ASME appendix 2 calculation), with the assumption of no external load, surprisingly this CL600 A105 flange is ok to use at 10.2Mpa at 60 degree.

The question is, there are external loading on the nozzles,
Fx, Fy, Fz and Mx, My, Mz.

I have done some search, suggesting convert external load to an equivalent pressure:

Pe=4*F/Pi*G^2+16*M/Pi*G^3

for F to be used in the equation, should I just use axial tensile load Fy?

Where shear force Fx and Fz, at flange face won't have impact on the calculation result?

for M, I am confused which value to use, I think torsional moment My will have no impact on this case, and should use M=Mx+Mz, as they are both bending moment.

Please correct me if I am wrong at any point.

Thanks in advance

Regards

Spoonful.

 

[roca]

Are these nozzles or body flanges?
If nozzles how is the connecting piping / flange being rated or designed for this new design condition?
Why not use 900# flanges?

 

[Spoonful]

Hi Roca,

This is CL600 flanges, The problem is this vessel will going onto a skid, where on the skid other components are designed by piping code which allow 10.2Mpa @60 degree C for CL600 flanges. If change the flange on this vessel to CL900, all other connecting flanges on the skid all need to be changed. The skid is already fabricated. So change flange will be the last option.

Regards
Spoonful

 

[roca]

Sorry, I'm a bit confused (maybe I am missing something). What is the piping code being used?

 

[C2it]

Spoonful,

You are correct in assuming that it is normal to exclude shear forces and torsional moment when calculating external loads in flanged joints. Also, it is commonly the case that B16.5 flanges are only moderately stressed at their pressure limit, although the bolt up stresses depend heavily upon the gasket parameters used.

Applying what we usually call the 'Kellogg' equation .....
Consider the axial force only if it is tensile, and use the net longitudinal moment (srss, not simply added) to get the equivalent pressure.

If you want a better authorised method of dealing with external loads directly, use ASME VIII Div. 2 (>=2007 Ed.)

 

[Spoonful]

roca,
the piping code used would be AS4041(from memory, need to check), which part are you confused?

C2it,
Could you given some reference on the Srss? or ref which part of ASME ＶＩＩＩ Div 2 I should look into?
I had a quick in section 4.16 of ASME VIII Div 2, the method used is almost the same as method given in Div 1 and AS1210, and have no mention of the external load as well. Perhaps I haven't read the Div 2 careful enough. Any hints?

Thanks

Spoonful

 

[C2it]

Check out ASME VIII Div 2(2007) section 4.16. The method is similar to Div 1 App 2, but it's an improvement.

Equation 4.16.16 onwards refers ME, absolute value of the external net-section bending moment.

My reference to 'srss' is an abbreviation. I meant don't add longitudinal axis moments directly as you inferred, use Square Root Sum of Squares.

If your pipe axis is Y then ME=(Mx^2+Mz^2)^0.5

 

[roca]

I am confused by the fact that the piping code (e.g. ASME B31.3) refers to B16.5 for flanges which in turn implies that the flange-temp ratings in B16.5 have to be used - same as for the vessel. So how can the piping code allow a higher value? Am I missing some clause from the piping code which I am not aware of? My background is not piping.

 

[Spoonful]

Roca,
piping code AS2885.1 states below:

3.4.3 Strength de-rating
Carbon steel and carbon manganese steel flanges and valves complying with nominated
Standards may be used without derating at design temperatures not exceeding 120°C.
NOTES:
1 Reference ASME B31.3, ASME VIII, and MSS SP44 – At temperatures up to 120°C flange
designs are based on (a constant) ultimate tensile strength resulting in no strength derating
requirement.
2 The temperature limit for flanged valves applies only to the flanges. Assurance should be
sought from the valve manufacturer that the valve body and seals are suitable for the required
service conditions.
3 The adoption of a higher design temperature for flanges requires that the pipeline and the
piping each satisfy the stress limits required by the design standard.
4 This permission does not currently apply to vessels designed in accordance with AS 1210
(e.g. filter vessels). AS 1210 currently requires strict compliance with the temperature
derating requirements of B16.5 flanges – although it does permit the use of MSS SP44


Where I don't really understand why would they allow this. It doesn't make sense to me.

 

[roca]

AS2885 is for pipelines - why not post the question regarding not derating flange allowable pressure due to temperature in that forum? I have not come across this before.
With respect to the above moments / forces I agree with the comments from C2it above.

 

[roca]

http://pipelinesoz.wordpress.com/2010/10/06/un-de-rating-flanges/

I think the verifier will have the last word on this.

You should be happy if you have to change all the flanges to 900# as it will be a costly client mistake..!

 

[Spoonful]

c2it,


According to AS1210 it won't allow me to use ASME calculation method, I did had a look into Div 2 methods, it did included external load, but it still calls a equivalent net bending moment ME,

Could you give some reference on ME=SRSS of stress?

and also could you give some reference to support of ignorance of shear force and torsional moment?

Thanks

Spoonful

 

[roca]

Pe = 16M/πG^3 + 4F/πG^2
Where
• Pe = Equivalent pressure N/mm2
• F = Axial external force in tension N
• M = External bending moment Nmm
• G = Diameter at location of gasket load reaction mm
‘M’ is the resultant of the longitudinal and circumferential moments acting on the flange = Square Root (Mc^2 + Ml^2)
Check out AS1210-2010, equation 3.21.6.4.1(1) which spells it out as well:


Torsion and shear forces are ignored

 

[Spoonful]

roca,

would you have any reference for

'M' is the resultant of the longitudinal and circumferential moments acting on the flange = Square Root (Mc^2 + Ml^2)

and

Torsion and shear forces are ignored

Thanks

Spoonful

 

[EngAddict]

Spoonful,

The addition of the moments is basic math theory, you will not need to look to the codes for a reference of how to combine them. Look at 3.21.6.4.1 for the equivalent pressure formula in AS1210, then look at 3.21.6.2 for the notation as to what force and moment mean (in this case). You will find that the shear forces and the torsion are not considered. This is based on the longitudinal stress check (which you are essentially performing) that doesn't consider torsion and shear.

Now refer to 3.21.1, down the bottom of the page, it gives a comparison to 50% of the flange rating for equivalent pressure.

AS4041 also states:
3.24.4.1 "For flanged joints with high longitudinal stresses (greater than 75% of design strength),
designers should consider including these loadings with pressure loads for rating
assessment."
Therefore if you have longitudinal stress below 75% of design allowable i.e. from external loading, then you don't need to consider this in the flange rating.

I am not familiar with the other piping code but it doesn't matter anyway when you are designing a vessel, up to the first flange face you must comply with AS1210.

 

[Spoonful]

EngAddict,

Thank you for your comment, AS1210 3.21.6.2 did specified only axial tensile load to be used,

and in regarding with "Meo= resultant external moment acting on flange for operating condition"

C2it suggest I should use Meo=(Mx^2+Mz^2)^0.5(y is pipe axis), I am trying to understand why is the resultant of 2 moments acting at 90 degrees apart is the square root of the sum of the square.

Section 3.21.6.6,
Longitudinal hub stress, radial flanges stress and Tangential stress is calculated by Mo and other factors, where all other factors are only geometric factor. Therefore the only force/pressure input to work out these stress is by Mo. And Mo is calculated based on total pressure P(design pressure + equivalent pressure by external load)

Therefore the Longitudinal hub stress, radial flanges stress and Tangential stress calculated based on P input, and we have ignored Shear force and torsional moment to work out P, it making sense for Longitudinal hub stress have nothing to do with shear force and torsional moment, but how about radial flanges stress and Tangential stress, are they still have nothing to do with shear stress and torsional moment?

Spoonful.