cycling temperature affecting threaded joint

[Tuckabag]

We're in the process of designing a pipeline insertion probe that due to it's length and special coating requirements, must be manufactured in two halves.
Current line of thought is that we used a threaded joint with grafoil o-ring face seal. Thread profile is 1.25" 18TPI UNEF.
The process undergoes a regen cycle where the temperature cycles from +20deg C to +242deg C every 36 hours.
There is some concern that the parts will expand under this cycling temperature and lose sealing capability on the grafoil face seal. We intend to use a locking bolt to prevent the thread from physically unscrewing due to vibration, etc.

Pic of the proposed joint attached.

Can anyone give any feedback on if they think this (losing seal) may be a concern?
Thanks!

 

[MikeHalloran]

That's a sketch.
You will need a drawing, which has tolerances.

I have no experience with Grafoil o-rings. In fact I'd never heard of them. I'm wondering where the strain energy is stored.

The threads really should have relief at the proximal ends to make manufacture a little easier.

I don't see any wrenching features, so I'm curious about how you tighten the threads to crush the o-ring, or loosen them to get the halves apart.





Mike Halloran
Pembroke Pines, FL, USA

 

[Tuckabag]

Yeah, apologies for the drawing, it is still early stages. I'll get a revised drawing through showing tolerances as soon as it's completed. Was just looking for some general feedback in case I'd missed anything blantantly obvious to someone else!

Grafoil is essentially just pure Graphite.

http://graftechaet.com/Home.aspx

We purchase it in a custom spriral ring to drop into the o-ring groove. We would expect to replace the ring each time we disassembled the probe.

Thread reliefs will be included (and shown on revised dwg).

There will be spanner flats on the half with the female thread, the male thread half is welded to the probe shaft which in turn is welded to a flange so we can secure this via the flange.

Thanks.

 

[MikeHalloran]

The only criticism I have that's on point is that a face seal at the bottom of a threaded joint like that comes with a fair probability of the seal falling out while the thread is being run down. You can't see it happen and you may not feel it in the wrench. For that reason, I'd put a back taper on the OD of the o-ring groove so that the radial compression would tend to hold the ring in place.

Actually, I'd probably put a pilot on the ID of the o-ring groove so that it's not strictly a face seal, and I'd probably extend the OD diameter, so effectively the o-ring seal gets squeezed at assembly by some real shallow lead-in chamfers. ... if I had the axial room to make the joint that long.



Mike Halloran
Pembroke Pines, FL, USA

 

[Tuckabag]

Thanks Mike, I'll take that into consideration.
Will post back once the tolerances are complete.

 

[desertfox]

Hi

If you reduce the threaded length in the female component and make a counterbore with a smaller diameter than that of the thread, you could use the 'o' ring to seal on the counterbore internal diameter instead of using it as a face seal.
In addition you don't need to worry about the thread backing off so much if you seal radially.

 

[LittleInch]

As with a number of posts, I get the impresison here that we're only being given about 20% of the requird info.

I've been in pipelines a long time and never seen anything like this before (but hey theres's always something new). Your comments imply that there is some sort of pressure differnetial going on here - how much? What is the thickness / diam / lengths of the tubes? What other forces are on this joint - axial, bending, torque, vibration?

What is it's purpose?

Have you looked at other standard joint connectors such as tapered threads, push connectors etc?

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Tuckabag]

OK, to clarify.
We are designing/manufacturing sample probes for inserting into pipelines/vessels. These are for a mol sieve dehydration vessel.
The probe consists of a 0.5" XXS 316SS pipe that is cross drilled along it's length to allow gas to flow into the pipe and down to an analyser. Fitted around the outside of this pipe/probe is a S/S screen approx 38mm in diameter.
Overall length of these probes once both halves are assembled is 3,800mm. The probe is connected to the vessel via a 2" 1500# flange connection and is supported at the opposite end from the flange by a support bracket. The probe is also supported along its length by the media that fills the vessel.

The pressure differential that the seal has to hold is minimal, it is used to ensure that none of the sample stream captured is able to escape out of the joint between the two halves.

We had considered other thread options, NPT would be our preference however we needed to keep the joint as small as possible as the requested spacing between sample holes was 25mm. The other issue with NPT was that the coating that needs to be applied (Silconert2000) tends to make NPT threads very susceptible to galling. We opted for a parallel thread to try and avoid this.

Hope that helps clear up what we're dealing with. I'll post a revised drawing with tolerances once we get it completed.

 

[LittleInch]

OK, that makes a bit more sense. I would look at moving your O ring seal from the inner surface to the outer shoulder and reducing slightly the male plug length by two or three mm to ensure you get a tight fit on the outer side. With only 1mm difference at the moment between plug length and insert length, you will never be quite sure if you've come to the end of the travel with the outside first or the shoulder of the plug before it pressures the O ring.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way