Subsea Pressure Housing Design – Wall Thickness for Stress & Buckling

[thereddevil]

I’m designing a subsea sensor that will go to a max depth of 600 m (6 MPa/60 bar). In a simple model, it will be made of a pressure housing cylinder and two end caps, all grade 5 titanium. Some geometry is attached.

I’m looking for advice on how to calculate the thickness of the cylinder/pressure housing based on the pressure. I’ve done some research and found a few methods:
1. ASME Section VIII – Division 1 (UG-28)
2. EN 13445-3:2009 Part 3: Design (Section 8 Shells under external pressure)
3. ‘Designing Under Pressure’ in the Journal of Ocean Technology Vol. 3, No. 1, 2008 (Link here: https://www.thejot.net/article-preview/?show_artic...) with two theoretical results:
a. Von Mises (same as first part of EN 13445)
b. Windeburg and Trilling’s Buckling Pressure

Failure mode I can see is the combination of external pressure on the cylinder with the buckling load caused by the end caps pushing inwards.

Now I’ve run them through with the methods above and have varying outcomes for the pressure at yield:
1. 7.8 – 2.4 MPa for different grades of titanium (grade 5 not listed)
2. 11.5 – 10.5 MPa
3. a. 77.9 MPa, b. 46.5 MPa

As you can see it varies a lot and not all good with a 6 MPa limit – what methodology do I use? Are these even correct? The last method (Journal of Ocean Technology) is interesting in that it explores adding factors to account for manufacturing defects (circularity for example).

Thanks in advance and happy to provide more info.

 

[berkeman]

Some geometry is attached.

I'm no expert in pressure vessel design, but is there a reason you chose to use a cylindrical shape and no internal bracing? It would seem to be an inherently weak design. Something more spherical with internal bracing would seem to be a better starting point, no?

 

[thereddevil]

I'm no expert in pressure vessel design, but is there a reason you chose to use a cylindrical shape and no internal bracing? It would seem to be an inherently weak design. Something more spherical with internal bracing would seem to be a better starting point, no?

There will be ports on the top and electrical connectors, therefore, not suitable for spherical design.

Bracing is a possibility, but internal space is very limited and I'm not sure how much I could add here (if any) as well as not knowing how I could calculate stresses involved.

As a side note, I have recently run FEA simulations with positive results - safety factor of ~7. But the aim here is to understand the hand calcs and then try to get something to match with the FEA.

 

[JBA]

It has been some time since I was involved in ASME Section VIII Div 1 calculations but to my memory the pressure limits were based upon stress allowables deemed safe for each listed ASME approved alloy rather than Sy and would therefore include an ASME approved safety factor.

 

[berkeman]

There will be ports on the top and electrical connectors, therefore, not suitable for spherical design.

Sorry, I don't understand that objection. Could you please elaborate? Ports and electrical connections work for whatever surface they have to go through...

 

[Baluncore]

Failure mode I can see is the combination of external pressure on the cylinder with the buckling load caused by the end caps pushing inwards.

Why fight the pressure?

You might instead use a silicon rubber housing, filled with an incompressible fluid such as transformer oil.

Is the flexible electrical cable protected from external pressure and salt water ingress by a petroleum jelly filling that displaces air and so prevents the crush ?

 

[thereddevil]

It has been some time since I was involved in ASME Section VIII Div 1 calculations but to my memory the pressure limits were based upon stress allowables deemed safe for each listed ASME approved alloy rather than Sy and would therefore include an ASME approved safety factor.

Yes, I can see that since the process involves using their charts. Is my application definitely relevant for use with ASME Section VIII Div 1? I feel like it is, but it appears they've not written this spec for subsea applications specifically.

Sorry, I don't understand that objection. Could you please elaborate? Ports and electrical connections work for whatever surface they have to go through...

I don't really have an objection tbh, you raise a good point (spacing and size restrictions will become tricky)! I can see the advantages of a spherical design - this would reduce the axial/buckling load on the pressure housing, right (by angling the force inwards more)?

Why fight the pressure?

You might instead use a silicon rubber housing, filled with an incompressible fluid such as transformer oil.

Is the flexible electrical cable protected from external pressure and salt water ingress by a petroleum jelly filling that displaces air and so prevents the crush ?

Oil is a no go in this case due to the optics inside. Could you expand on your last sentence?

 

[thereddevil]

As a side note, there are companies that design subsea housings similar to what I'm aiming for:
- https://prevco.com/shop
- http://www.agoenvironmental.com/main-service/custom-pressure-cases
I would like to understand the theoretical side of defining the wall thickness before I proceed to FEA and prototyping.

 

[thereddevil]

Here is a Ø150mm x 590mm instrument:
- http://www.valeport.co.uk/Portals/0/Docs/Datasheets/Valeport-MIDAS-CTD-plus.pdf

Here is a Ø205mm x ~205mm instrument:
- https://www.sonardyne.com/product/lodestar-gyrocompass-motion-sensor/

Obviously I have no idea of the wall sections, but can't imagine they'll be crazy thick.

 

[Baluncore]

Is the flexible electrical cable protected from external pressure and salt water ingress by a petroleum jelly filling that displaces air and so prevents the crush ?

Could you expand on your last sentence?

What type of tether/conductor/fibre is connected to the pressure housing.
How is that protected from pressure and water?

 

[thereddevil]

What type of tether/conductor/fibre is connected to the pressure housing.
How is that protected from pressure and water?

Oh, it will be two of these: https://www.macartney.com/what-we-o...5-6-and-8-contacts-and-g2-2-3-and-4-contacts/
The 'MCBH?F' variety.

 

[JBA]

Yes, I can see that since the process involves using their charts. Is my application definitely relevant for use with ASME Section VIII Div 1? I feel like it is, but it appears they've not written this spec for subsea applications specifically.

ASME Section VII Div 1 is focused on pressure vessels focused on plant and process services where there is a risk of injury or death of personnel and/or the potential release of toxic chemicals; and, as a result, it is extremely conservative; and, that probably accounts for its pressure limit being much lower than your two alternative calculation methods.
Section VIII is an excellent guide for the design of any pressure vessel service; but, unless the performance of your device is critical for safety, then its pressure limits may be lower than required for your equipment application.

 

[thereddevil]

...unless the performance of your device is critical for safety, then its pressure limits may be lower than required for your equipment application.

Ahh OK, makes more sense now. Still struggling with what the correct theoretical should be, though!

 

[Lucas Vilas Boas]

Ahh OK, makes more sense now. Still struggling with what the correct theoretical should be, though!

I'm also designing a subsea pressure housing and having the same doubts. Have you concluded your project? Which theoretical did you use?