Formula for Elbow under Internal Pressure? 3

[Christine74]

What is the equation to calculate the required thickness of a 90 degree elbow under internal pressure?

Thanks,

-Christine

 

[jte]

Christine-

See API-579 A.5.5 (pgs A-31, 32) and Fig A.14 (pg A-53). ASME B16.9 part 2.2 refers to B16.49 for bends. I don't have B16.49 so I don't know if it provides the same formulas as API-579.

One problem is that many manufacturers opt out of the tmin calc's by proof testing.

See also thread378-122201.

jt

 

[JohnBreen]

Christine,

From B31.3, 2004, paragraph 304.2.1:
Based upon the classic torus equation

The minimum required thicknes of a bend after bending in its finished form is calculated by:

t-min = (PD) / 2[(SEW/I) + PY]

where at the intrados (inside bend radius)

I = [4(R1/D) - 1] / [4(R1/D) - 2]

and at the extrados (outside the bend radius):

I = [4(R1/D) + 1] / [4(R1/D) + 2]


and at the sidewall of the bend

I = 1.0

and;

Y is the bend angle (e.g., 90 degrees)

R1 is the centerline radius

D is the pipe ouside diameter

P is the internal pressure

You must decide what corrosion allowance is needed and of course you may want to add thickness for groving, erosion, threading, etc.

If you are using a standard B16.9 welding elbow you only need to specify the schedule of the matching pipe.

Regards, John.

 

[davidribeiro]

Why cant one just base the calc on a straight pipe under int press? Just a ridiculous question from my side

 

[jte]

david-

Because it is not a straight pipe under internal pressure! If you have access to a FEA package, model an elbow and apply internal pressure. You'll see a nice distribution of stresses, not a consistent pr/t stress.

jt

 

[jamesl]

jt,

David was not totally wrong. According to UG-44, if the elbow was made to meet ASME B16.9 or B16.11, it could be calculated as a straight seamless pipe.

 

[jte]

jamesl-

Depending on how you interpret "calculate" I might agree with you: If by "calculate" you mean "determine a pipe schedule for straight pipe and order an elbow of the same schedule" then I'll agree. If you mean "determine the required wall thickness of an elbow" then I'll disagree. This would be true for either VIII-1 or B31 applications.

This is generally not an issue which comes up in new construction, but in evaluating the fitness for service of existing piping (and less frequently, vessel nozzles).

jt

 

[ijzer]

Just a remark on the JohnBreen reference to ASME B31.3 :
The Y in the formula is not the bend angle but a material/temperature depending coefficient to be taken from table 304.1.1

 

[jamesl]

jt,
I mean the first one.