NPT for cyclic service

[XL83NL]

Since a lot of the stuff our company builts has a small line size, a lot of the connections we specify for instruments, valves, etc. are threaded, mostly NPT. The following question comes up every now and then within our company.

Some of our projects deal with fatigue due to cyclic service (pressure/temperature cycles). A current project has cyclic service involved due to pressure and temperature cycling, from 0/ambient to max operating. I have been under the impression that cyclically loaded NPT threaded connections are (more) prone to failure due to stress raisers on the thread's tops, but I have no hard data or source to back this up. Straight threaded connections with a (metal) sealing ring, e.g. ISO 1179-1, are to my understanding more reliable. But again, no data to back this up. However, straight threaded joint with a sealing ring are found (more) in automotive, where cyclic behaviour is also of concern, to some parts. So that tells me something.

Are there any papers or data available that discuss fatigue life (due to pressure and/or temperature cycles) of NPT threaded joint? I have looked at the ASME digital collection, so far the best hit I could find was this, but Im hesitant to give it a try as it seems to relate fatigue to a different nature then our application.

 

[SwinnyGG]

Straight threads and sealing rings are used in automotive because they are an order of magnitude easier to design for consistent sealing behavior and component positions on a large scale. Fatigue and failure of the actual components are not a factor.

 

[XL83NL]

Thanks. Wasnt aware of that, I used to believe cycling of taper threads in automotive parts could results in damage during disassembly (cracks in thread, galling, damaged parts during disassebmly, etc.).
Nevertheless, lets not expand on that and stay with the original question I posted.

 

[SwinnyGG]

Well.. the threads on tapered fittings and the threads on straight threaded fittings with flanges and sealing rings are all cut the same way. I would expect fatigue behavior of the fittings to be similar regardless of the type of thread.

 

[EdStainless]

I don't think that mechanical fatigue is a real limit for NPT, but sealing fatigue sure is.
Being a truncated thread profile there is a guaranteed leak path.
And whatever temporary sealant that you use in assembly will age and become less effective.
In our plant for high purity or cyclic service we won't use NPT, and not that people are paying attention to compressed air leaks we will likely stop using them there also.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[XL83NL]

A metal (316) sealing as per the referenced ISO 1179-1 standard wont age.
@Ed, if you step away from NPT, then what do you use?

 

[SwinnyGG]

ISO 1179 describes straight thread ports with elastomer (note: not metal) sealing rings. It does not describe metal-on-metal tapered thread sealing connections.

 

[XL83NL]

Note I specified ISO 1179-1, which is for metal to metal sealing rings as well.
Also note we use ISO 1179-1 (perhaps with a few dimensions from that standard modified to our best practice) with straight threads, not tapered. My OP was also for straight threaded ports with a metal ring for sealing.

 

[TimSchrader2]

Hello

In hydraulics the straight thread or JIC with metal rings are prefered. When using the 1/4 turn rule the NPT fittings end out at one orientation (without redoing), the Straight Threaded can be positioned more easily then the NPT.
I can almost say for sure, but not fully that I have seen more leaks at NPT fittings. This may be due to the installer not fully following the 1/4 turn rule in order to get a desired orientation of rotation on the NPT fitting. Like lining up with a hose. Aslo, over time the hose may pull on the fitting, if not supported fully.
I do not have a paper to back this up.