Calculating Stress in Fittings 2

[DG99]

I was looking for a way to calculate stress in pipe fittings (such as Tee's, elbows, reducers). I have the diameter, wall thickness, grade and pressure and would like to calculate the stress to compare to SMYS. If I calculate the stress in the pipe and it is ok, and the fittings are the same diameter, wall thickness and grade as the pipe, can I assume that the fittings are ok also?

 

[pennpiper]

You need to post your question here:

http://www.coade.com/support_discussion.asp

 

[JohnBreen]

Hello DG99,

Oh that it would be so simple.

The first question that we have for you is "what is the pressure piping Code to which you are designing this piping system?"

The first thing you must do is determine if the piping and fittings are compliant with the Code in regard to internal pressure. That is usually done by calculating the minimum allowable PIPE wall thickness (this addresses the circumferential or "hoop" stress). In doing this calculation you must address the corrosion allowance, mill tolerance for the piping and any other issue that would cause reduction of wall thickness during the life of the system. Then you "round up" the calculated minimum wall thickness to the next standard wall thickness (schedule). Once you have specified the pipe schedule (thickness) you need only specify the fittings as having the same thickness schedule (this assumes that the fittings are made to a Code accepted Standard (e.g., ASME B16.9). If you are using bends that are other than Standard elbows, there are other factors to consider.

Next, you must address the sum of the longitudinal stresses due to sustained weight and internal pressure. This is usually done by modeling the entire piping system using a good piping stress analysis program. This analysis calculates the sum of the longitudinal bending stress added to the longitudinal pressure stress. The sum of these stresses must be less than the Code specified maximum allowable stress at temperature. The pipe stress analysis software will apply the stress intensification factors (SIF's) that are required at the fittings as it calculates the stresses and before it makes the comparison to the Code maximum allowable stress. This analysis will (for one thing) tell you if your system is adequately supported. This evaluation addresses primary stresses.

Next you must address the thermal expansion (displacement) stress range due to temperature excursions. The range of stress due to system going from its "installed temperature" to its minimum temperature is added to the range of stress due to system going from its "installed temperature" to its maximum temperature - this give you the TOTAL thermal (displacement) stress range. The total thermal (displacement) stress range must be less than the Code specified maximum allowable stress range. This is usually also done by modeling the entire piping system using a good piping stress analysis program. Again, the pipe stress analysis software will apply the stress intensification factors (SIF's) that are required at the fittings as it calculates the stresses and before it makes the comparison to the Code maximum allowable stress range. This evaluation addresses secondary stresses.

Do not overlook addressing the loadings (forces and moments) that will be transferred by the piping to the equipment that the piping is attached to.

The procedure described above is not a rigorous and exhaustive description of all the factors and issues that must be addressed. To provide such a description of piping stress analyses requires a book. An excellent book that you might want to seek out is "Process Piping - The Complete Guide to ASME B31.3", by Dr. Charles Becht, IV, PhD, PE. Go to:

http://catalog.asme.org/books/PrintBook/Process_Piping_Complete_Guide.cfm

Regards, John.

 

[JohnBreen]

Hello again DG99,

I neglected to mention in my posting above, third paragraph regarding pressure design (minimum wall thickness for PIPE) that the ASME B31 Pressure Piping Codes DO NOT provide rules (or equations) for calculating the stresses in fittings. These Codes require that you calculate the minimum required wall thickness for the piping. After including additional thickness for erosion, corrosion, grooving, mill tolerance, etc., you round the thickness up to the next standard pipe schedule. Then you specify the fittings to be the same schedule as the the "as specified" pipe. The thicknesses and geometries of the fittings are a function of "proof testing". The "proof test" assures that in a test the matching straight pipe (attached to the fitting) will fail due to pressure before the fitting will fail.

Regards, John.

 

[DG99]

Hello JohnBreen,

Thanks for your reply.

I guess I should have explained a bit better what I am trying to do. The piping system I am looking at is a metering manifold. The manifold is already designed and has been operational since the early 90's. There was an incident where the manifold was overpressured (exceeded the max allowable operating pressure). We've been asked by the regulator to do an analysis on the manifold to confirm fitness for service. I am working on the piping and fittings. Based on the equation stress = (pressure*diameter)/(2*wall thckness), the pipe did not come anywhere near the specified minimum yield streangth for the grade of pipe. What I was not sure about, is how do you prove the same for the elbows and tee's? Any suggestions?

Thanks.

 

[JohnBreen]

Hello again DG99,

Well, consider the fact that a hydrostatic pressure test done to B31.3 requirements will exceed the MAOP of the piping. No harm, no foul.

Again, there are no simple equations for calculating stresses in fittings because they are more complex in geometry than cylinders (straight pipe). If you feel that you need to know the stresses you will have to use FEA.

If it were my responsibility, I would give the header a good visual examination looking for gross distortions and if there are no visual distortions I would do a leak test to B31.3 requirements and if it passes that test I would put it back into service.

If the header were made of a less ductile material I might do UT shear wave examination of the welds (looking for any cracks (smaller than the original NDE flaw threshold) that might have propagated. An alternative would be to do a hydrostatic pressure test to B31.3 requirements with acoustic emissions transducers attached to "listen" for any propagation of cracks during the test. If it passes all that I would put it back into service.

Regards, John.