We all know that during operating condition, a pipe may undergo plastic deformation. Code takes it in account and therefore the allowable stress range may be greater than yield strength of pipe in hot condition. In reality the pipe can never be stressed beyond yield point and is plasticaly deformed beyond this limit.

Now, while calculating the nozzle loads, the software will actually assume the pipe to expand as per thermal expansion coefficient, calculate moments and stresses till considered temperature and accordingly give the loads.

The nozzle loads thus provided by software will not take any self springing in account.Although there will be stress relieving in the system, due to which the nozzle loads may be much less than value given by software.

In my view, if stresses exceed a certain value, software should calculate the nozzle loads directly based on the yield strength of material, as this is limit to which a material will be stressed.

Can you please correct me where I am wrong or suggest a way which gives us a more reasonable and less conservative estimation of the nozzle loads?

Equipment such as air coolers, pumps and compressors have very low allowable loads and related pipe stresses are tiny. Vessel nozzles in, for example, Norsok R-001 are limited to approximately 6000psi bending stress in equivalent size Sch40 pipe.

What sort of nozzles would you expect to be so heavily loaded that the pipe may exceed yield ?

Not true MoverZ. Barrel shaped nozzles can definitely take much higher loads than the pipe. I just did a 60" barrel shape nozzle that passed NozzlePro, but the pipe was at 130% overstressed in the sustained case.

Mr. Threeouts,

We are dealing with generalisations here, in particular, the suggestion that software such as Caesar should operate in a different way regarding nozzle loads. I do not agree with this.

There are always exceptions, but in general, the usual nozzle load restrictions equate to very low pipe stresses.

Like you I have worked with HP compressors where the barrel was 6in. thick or so. The allowable loads were comparitively very small because the limit was set by shaft alignment and pedestal / frame stiffness.

Just check out any set of nozzle loads from NEMA, API 610, API 650, API 661, Norsok etc. Always they will lead to low pipe stresses.

I doubt that even if the stresses on the adjacent bend(bend before the nozzle) exceeds yield, the stresses at the nozzle may still be less.

Similarly the stresses may exceed yield on bend before the adjscent bend. There might be an impact on the nozzle load due to self springing of these connected components.

If the stresses on the adjacent bend are relieved, so will be the loads at the nozzle.

Therefore it might be a good idea to consider yeild strength at these bends to calculate the Nozzle Loads.

If you have yield stresses in pipe anywhere near a pump, you have big problems. Adding a 'fix X' or 'fix Z' in a stress analysis model might reduce your numbers, but in reality, local stiffness of supports etc will ensure that you have very large and probably damaging loads at the machine nozzle. As one of our forum contributors points out, "All the world is a spring". Pipe supports and restraints included.

I resemble that remark

Hi Ed.