Thick wall / high pressure pipe wall thickness calculations

[mariolucas75]

Dear All,

For pipe thickness calculations:

API 574 says for metallic pipe for which t ≥ D/6 or P/SE > 0.385 requires special consideration.
What is this special consideration? Where shall i find it ?

ASME 31.3 Para 304 Straight pipe under internal pressure says:
For t ≥ D/6 or for P/SE > 0.385, calculation of pressure design thickness for straight pipe requires special consideration of factors such as theory of failure, effects of fatigue, and thermal stress.

So, ultimately where is this formula available for thick wall / higher pressure (t ≥ D/6 or for P/SE > 0.385) pipe wall calculations ?

Thank you

 

[mariolucas75]

OP,
ASME B31.3 don't provide Code equations when t ≥ D/6 or P/SE > 0.385.
You must be aware that at higher wall thickness, the stress across the wall thickness is no more a membrane stress (equal/average) but rather becomes exponential. ASME B31.3 Chapter K is based on High Pressure Piping definition that is based on a ASME B16.5 Class 2500 P/T rating.
Because the Piping Codes are drawn from ASME BPVC, a better response to your query can be found in Mandatory Appendix 1 Para 1-2 (Cylindrical Shells) of ASME Sec VIII Div1 that follows the rule in your query.
Both equations in Chapter K and Sec VIII Div 1 MA 1 are in exponential form.
You have already done the thickness calc per Chapter K. Can you also do the calc using Div 1 MA 1? It will be interesting to see the result.

Dear GD2,

Formula from ASME Div VIII Mandatory appendix 1, Para 1-2 for Cylindrical shells is the same as B31.3 Chapter XI, K304.1.2 Straight Pipe Under Internal Pressure and that gives me the thinner value for thickness rather than "conventional" thickness formula from 304.1.2 Straight Pipe Under Internal Pressure. Kindly see please attached file ...

 

[goutam_freelance]

With refined calculations using Matlab, the following results are found:

Allowable stresses for B31.1 and VIII-1 are the same.
VIII -1 MA1 gives a slightly lower thickness than B31.1
B31.3 gives the lowest thickness as the allowable stress is more.

 

[LittleInch]

Well 55mm would seem to cover most of the options.

Comprehensive check there so write it up and use that.

That is some gun barrel type pipe though - 420 bar! And 370C. Hope it isn't very long.

 

[mariolucas75]

My question (that i struggle with) still is - within B31.3:
Between B31.3 Chapter IX ("High pressure" thickness formula 34 a) and B31.3 Chapter II Par 304.1.2 Straight Pipe Under Internal Pressure ("Not High Pressure" thickness formula 3a) .... as shown in PDF in my previous submission?

Why B31.3 high pressure formula gives thinner wall thickness for pipe rather that B31.3 "Not high pressure" formula ??
Or what do I do wrong and can not get ??

 

[GD2]

OP,
That is always the case. The thickness equations in B31.3 Chapter IX and Div 1 MA1 are more accurate in terms of stress distribution across the wall.
The answer really lies in the stress tables (example Table A1-M Vs Table K1-M). The basic allowable stress, S, for the A106 Gr B pipe is 20 ksi Vs 28 ksi. The Code is allowing you to use a higher allowable at the cost of doing more material/welding examinations, stricter NDE acceptance limits. If you note, a fatigue analysis is also required in Chapter IX.
In other words, more analysis allows more stress allowable.
The same is true when you perform Sec VIII Div 1 vs Div 2. Div-2 wall thickness will always be lesser than Div 1.

Hope it explains your query.

 

[mariolucas75]

GD2

Thank you very much some light is shed now onto this....
May I approach with another question ... how / whereto use Y=(d+2c)/(D+d+2c) defined in B31.3 Para 304.1 Straight Pipe, definition for Y when t>D/6?
Or this Y in this case only to used in pressure determination ? Since knowing D and d in fact you know the thickness ?

 

[goutam_freelance]

@mariolucas75

This is an iterative solution that can easily be implemented in MATLAB. See the script for my example above (implemented as per B31.1 formula)

>> D0=355.6;% OD
SE=114.45;% Allowable stress
P=42;% Design Pressure
c=1.6;% corrosion allowance
% tt the variable for the calculated thickness

sigma1=@(tt) P-2*SE*(tt-c)./(D0-2*yy(tt).*(tt-c));% Formula for allowable pressure in B31.1
yy=@(tt) (D0-2*(tt-1.6))./(D0+D0-2*(tt-1.6));%yy is Y
tt=fzero(sigma1,50)% solves the equation to give 0 in RHS

tt =

58.4062

So the solution is t=58.4062 mm

@LittleInch

Thank you for the suggestion. Presently I am looking for derivation of exponential thickness formula in VIII-1 MC1. Shall see the possible use after this.

 

[GD2]

GD2

Thank you very much some light is shed now onto this....
May I approach with another question ... how / whereto use Y=(d+2c)/(D+d+2c) defined in B31.3 Para 304.1 Straight Pipe, definition for Y when t>D/6?
Or this Y in this case only to used in pressure determination ? Since knowing D and d in fact you know the thickness ?

OP,
when t>D/6, we are essentially talking about thick wall pipe. Piping, even pressure vessel Codes always draws a line between thin and thick wall definitions. The reason for it is the stress distribution across the wall.
ASME B31.3 Equation (3a) is an empirical approximation of more accurate and complex Lame's equation.
Hoop Stress distribution is typically higher at ID than OD.
The Y factor (which is temperature based) is given to take care of the stress distribution at creep (look at the temperatures in Table 304.1.1) where the stress distribution becomes almost uniform across the wall. This factor increases with temp to decrease the calculated wall thickness at the constant allowable stress, S.

The Lame equation (3a) works well for a range of D/t ratio but increasingly becomes conservative when D/t<6 (thick wall). This is when the value Y=(d+2c)/(D+d+2c) is used with the same Lame's equation in (3a) matching the Lame's equation for heavy wall pipe (it will be a different equation than (3a)).

Most of the time, you will use Y value as 0.4 as the design temp will be below the creep range.