High strength metals for external pressure vessel

[TeejT]

I have to design a corrosion resistant cylindrical pressure vessel that withstands high external pressures. I am limited to a certain wall thickness which, according to testing already performed, is insufficiently strong if the tube is made out of stainless steel. There is some machining required in reaming the ID of the tube to tolerance to accept an O-ring plug. What material options do I have to improve the pressure rating? Inconel, titanium? Any other suggestions? Again, the ends of the tube need to be machined to an ID tolerance spec and roughness spec to accept the O-ring plugs, so the material should be relatively easily machinable, or at least not impossibly difficult to machine. Thanks!

 

[Metalguy]

What kind of corrosive fluid is involved, temp, and size?

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[TeejT]

Fluid is a brine solution with petroleum derivatives present, maximum temperature 200 deg C, rough size ~ 2" diameter, 2 ft long. Thanks!

 

[Metalguy]

Sounds like a job for titanium or one of its alloys. It is machinable with a rigid tooling setup and should be readily available.

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[rb1957]

when you say "external pressure" do you mean that the tank is a vaccuum ? which sounds odd, maybe the tank is at high pressure, so the tank is in tension ??

i don't think Ti will do better than high strength steel.

what about a thin Ti inner sleeve wrapped with graphite ??

 

[zekeman]

How high is the pressure? And what is the internal pressure?

 

[TeejT]

The tube internal pressure is 1 atm.
The external pressure is 10,000+ psi.

rb1957,
Why wouldn't Ti outperform steel? The tensile/compressive strength is several times higher.

 

[rb1957]

"Why wouldn't Ti outperform steel? The tensile/compressive strength is several times higher." ... i guess you're refering to the more common steels, <125ksi ftu. I'm used to 200ksi steels, ti ftu = 160ksi. your application will put the pv into compression, which can be hard to stabilize, how thin is the wall ?

looking at things, p = 10000psi (give or take), R = 1" ...
stress = pR/t = 10000/t
125000 = 10000/0.08 ... sounds pretty thin, but onlu 1"radius ... maybe (on the road, haven't got my books for buckling a cyclinder from pressure)

 

[MikeHalloran]

I got some counterintuitive results.
For an external pressure of 10,000psi, and an OD of 2.00",
a perfect tube will resist collapse with a wall as follows:
316 .105"
Ti6Al4V .120"
for the same conditions,
an imperfect tube with a radial deviation of .005"
will resist collapse with a wall as follows:
316 .336"
Ti6Al4V .127"

I could be holding my tongue wrong...



Mike Halloran
Pembroke Pines, FL, USA

 

[sreid]

How about 440C Stainless Steel, heat treated and passivated?

 

[cloa]

When you say external pressure, is that just along the sides of the pressure vessel?

 

[TeejT]

MikeHalloran,
How did you reach those results?
And how is the 316 SS tube stronger than Titanium as would be implied by the thinner wall requirement to withstand 10,000 psi?
We're just talking von Mises stress (equivalent tensile stress) in the material created by the external pressure, and the tensile yield stress for Ti is much higher than SS, about quadruple (130 vs 30 ksi). Correct me if I'm wrong...

cloa,
I mean external pressure as if the tube was submerged in water.

rb1957,
I'm referring to stainless actually.
And stress = pr/t is valid for INTERNAL pressure vessels as far as I know...

 

[Canadieng]

One thing that should be kept in mind, is that the dominant failure mechanism due to external pressure is buckling - which is more dependent on Modulus of Elasticity (& Poisson's ratio) than it is on tensile strength (one could argue of a relationship between them though)

 

[MikeHalloran]

I use a spreadsheet based on these:

ref 1: "Tubing Limits for Burst and Collapse", Tech Note, CTES, L.C., Conroe TX

http://www.ctes.com

ref 2: "Effect of Initial Eccentricity on Collapse Pressure of Circular Beam Tubes", S. Yadav
Fermi National Accelerator Laboratory, Batavia IL

http://www.flab.gov

Ti6Al4V properties from

http://titanex.com/tigrades/titaniu...rade-1-2-3-4-5-7-9-11-12-19-beta-c.php?sel=11

I did a goal seek for wall thickness based on 10,000 psi as the critical collapse pressure, for circular and oval tubes.

The spreadsheet has given results that correlate well with an unfortunate incident in a former employer's shop a while ago, and that also correlate with results reported here in another discussion.

I'd rather not make it public, not least because I don't have a lot of confidence in the numbers I used for Ti6Al4V's properties; yield stress, tensile modulus and shear modulus. If anybody has trustworthy numbers for those, please speak up.



Mike Halloran
Pembroke Pines, FL, USA

 

[MikeHalloran]

Now I've got a Matweb page open for annealed Ti6Al4V.
The listed values for shear modulus and elastic modulus yield a computed Poisson' ratio of 0.29, but Matweb lists Poisson's ratio as 0.342. The buckling calculations are highly sensitive to the ratio. So, what to believe?



Mike Halloran
Pembroke Pines, FL, USA

 

[Canadieng]

MMPDS-04 lists Poisson's ratio as 0.31 for Ti-6Al-4V (regardless of condition or form)

 

[MikeHalloran]

MMPDS-01 lists Poisson's ratio as 0.31, but gives elastic and shear moduli that make it compute to 0.36.



Mike Halloran
Pembroke Pines, FL, USA

 

[Canadieng]

hmmm ... perhaps the most reliable method would be to go through the ASME BPVC external pressure calcs/charts and compare different materials.

 

[MikeHalloran]

I think the calculations suggest that a theoretical perfect tube of 316 will be a little stronger than a similar Ti tube because its moduli are higher, but a realizable imperfect tube will be weaker because of the lower tensile yield, all dimensions being equal.

For example,
2 2 in od of tube
0.120 0.120 in wall thickness
316L Ti-6Al-4V
2.70E+07 1.70E+07 psi Elastic modulus
1.06E+07 6.39E+06 psi Shear modulus
0.27 0.33 Poisson's ratio
2.70E+04 1.32E+05 psi yield point for tube material

15179.3 9922.5 psi critical buckling pressure for a _perfect_ tube


0.005 0.005 in radial deviation from circle at worst point
2668.73 8667.60 psi critical pressure for collapse of imperfect tube

(Numbers pasted from a spreadsheet and edited; I hope they line up for you.)



Mike Halloran
Pembroke Pines, FL, USA

 

[flash3780]

How about 17-4PH SS? It may rust a bit, though. Inconel 625 is a good material as well (quite tough). As was mentioned previously, buckling is a likely failure mode, so stiffness is a prime concern. If buckling turns out not to be an issue, Inconel 718 has a higher yield strength than 625.