pvc pipe under external pressure 5

[hyposmurf]

Hi all
How deep in water could I sink a 20" schedule 20 pipe, assuming the internal pressure of pipe is the atmosphere??
Of course it have not to collapse or be deformed.

http://www.cadtutorforum.net

 

[JStephen]

Wow, lots of approaches above. My suggestion is whip out the ol' handy 5th edition of "Formulas For Stress and Strain" by Roark and Young, and look up the buckling equations for cylinders. Try a factor of safety of 2 or 3 depending on how critical it is.

I don't believe Timoshenko's "Theory of Plates and Shells" really goes into the elastic stability issues. Try his "Theory of Elastic Stability" for that. However, the only copy of that which I have seen (in a library) was badly outdated. And the general shell formulas that are given in "Theory of Plates and Shells" are given in a more condensed form in "Formulas for Stress and Strain".

The shell buckling formulas in "Formulas for Stress and Strain" are similar to those typically used for steel tanks, if that helps.

One thing to be aware of is that most of this buckling information that is available was derived for steel (think submarine research, in particular.) If your material properties vary markedly from steel, you may not get the accuracy you'd expect from an allowable-stress type equation. And buckling equations are prone to wide variations in the first place.

One other resource is the PVC pipe people. I know the AWWA standards for ductile iron pipe give charts or other design data for wall thickness based on burial depth. I suppose this information is also available for PVC. You may or may not find it in a useful form, though. The external pressures in buried piping are different from your application.

 

[TGS4]

MikeHalloran, it appears that once I remove the design margin on my calculation, then I end up with (roughly - within 20%) the same external pressure that you get. I completely agree with both you and stanier that imperfections must be considered - at least insofaras the fabrication tolerances of the PVC pipe. Furthermore, because we are talking hydrostatic pressure, the pressure difference betwen the top and bottom of the pipe is 0.722 psi - that's more than 10% of the calculated maximum external pressure, so it's obviously going to have an effect on the results.

One other thing that occured to me as I was discussing this with collegues- at room temperature, PVC creeps. So, now you're talking about creep buckling, a regime that I am a little unfamiliar with. Howeve, I know enough that I would be uncomfortable designing in that regime.

 

[vzeos]

The Plexco Engineering Manual covers external hydrostatic pressure on plastic pipe. This manual is apparently no longer available on line so I have temporarily posted the relevant pages.

http://mysite.verizon.net/vzeos8kl/sitebuildercontent/sitebuilderpictures/page75.jpg

http://mysite.verizon.net/vzeos8kl/sitebuildercontent/sitebuilderpictures/page76.jpg

http://mysite.verizon.net/vzeos8kl/...builderpictures/.pond/page77.jpg.w300h493.jpg

 

[unclesyd]

Here are the API formulas in a group.
One thing to add if you get over about 8O ft in the Gulf Mexico you will have to add a little temperature correction, I know I do or better did.

http://www.grantprideco.com/engineering/api/calculators.asp

 

[stanier]

I hope everyone noticed the section immediately under the Plexco equation re Ovality.

"Pipe Ovality Effects

Pipe deflection will reduce flattening resistance, and lower critical flattening pressure"

The equation for buckling quoted is a Timoshenko and can be found in the following reference.

1) Materials Selection Manual for sewers, pumping mains and manholes, Water
Research Centre Buckinghamshire: 1993.

2) Pipe Materials Selection Manual , Water Research Centre Buckinghamshire: 1988.

Buckling will be the controlling criteira in design rather than yield. This is mainly because the pipe will not be perfectly round. Buckling is not something that is easily predicted because of imperfections in shape. Such imperfections have little impact on properties such as "yield".

 

[TGS4]

stanier - you are right on w.r.t. variations from a "perfect" cylinder resulting in lower buckling pressures. That is why vessels/piping under external pressure have very tight out-of-roundness tolerances - and even then, fairly high design margins. If you want to lower the design margins, then you have to reduce the out-of-roundness.

Anybody have any ideas on the creep buckling that I mentioned earlier? Do you think that it woudl actually be a factor in PVC pipe?

 

[MikeHalloran]

I got interested in the subject after buckling the inside tube of an expensive stainless jacketed assembly while performing a leak test. Everyone wanted to blame the leak test guys for using shop air instead of regulated air. They clearly didn't; the calculated buckling pressure came out real close to the regulated leak test pressure.

Once the buckle was detected (you couldn't observe it directly because of elbows), everyone remembered the loud bang/ clang/ pop they heard at the time. No one had reported it or thought it unusual, because our shop rings with loud noises all day long.

It was my fault, for not making the calculation before the event happened.

I have since tried to further correlate the equations I used with some real world situation, e.g. took a boat apart looking for a buckle that wasn't there, at least not yet. The equations may be a little conservative, but are so wildly sensitive to wall thickness and ovality that it's hard to make correlations with any precision.




Mike Halloran
NOT speaking for
DeAngelo Marine Exhaust Inc.
Ft. Lauderdale, FL, USA

 

[Cockroach]

TGS4, I think you're off somewhere in left feild.

Nowhere in the original post from HypoSmurf is "span" or the mere mention of pipe support entertained. Properly, buckling, traverse loading, what have you, is beyond the scope of computation until additional information is required. Maybe the entire pipeline is fully supported and laying on the lake bed, i.e. no collapse. Then again, maybe a support will be placed every six feet. Who knows? So the solution is incomplete regarding buckling.

I know one thing for sure. This morning I stood on a piece of 20 diameter pipe, schedule 20 over in my buddy's yard. Granted the pipe was on the ground, just blocked from rolling away from me. I'm a little guy, 62 kg. The pipe was strong enough to hold me, even bouncing up and down on it moderately. I think your "129 inches" is ridiculus, that's only 4.66 psi. Oh yeah, you unnecessarily assumed 20 feet span between supports. Bad assumption.

But reviewing your earlier posting(s), was that not the point you where trying to make?


Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[TGS4]

I'll stay in left field, if that will keep me from a failure due to elastic buckling. Go ahead and change the support spacing - make it 6 ft, make it 2 ft for all I care. I doesn't change the fact that elastic buckling of a cylinder within a given tolerance of perfectly round is a governing design case.

Coackroach, did you not happen to read the postings by MikeHalloran and stanier? I think that you are the one who is comletely out to lunch on this one. An external pressure of 222 psi is competely ridiculous! If you were to introduce even slightest perturbation (you do remember that elastic stability is based on perturbation methods, don't you?) to the wall, in the form of out-of-roundness, or even a pressure gradient between the top of the pipe and the bottom of the pipe, the PVC pipe would collapse (elastically I might add) at an external pressure of less than even 1/10 of the 222psi that you predict.

Cockroach, unless you can come up with a calculation that completely refutes the methodology and results from MikeKalloran's 4 May 05 21:58 posting, I will simply reject your assessment on this item. Elastic stability does not seem to be your strong suit.

BTW - if you were paying attention, the 4.66 psi external pressure that I calculated was based on a design margin of 3.

 

[Cockroach]

Thank you UncleSyd, the Grant Prideco site is a good reference!

By now TGS4, you can see we are measuring depth on the order of hundreds of feet, not a few hundred inches. Using the yield collapse equation, 20 diameter pipe of size schedule 20:

let M = D/t = 20.00 in/0.375 in = 53.333
P = 2S[(M-1)/M^2] = 2(5218 psi)[52.333/53.333^2] = 192 psi
h'=P/(rho g) = [192 psi / (62.5 lbf/ft)] X (12 in/ft)^3 = 5308 in

This is the depth below the point where wall stress is zero, already deduced as 406 in.

H = 5308 + 406 in = 5714 in = 476 ft.

I had reported 232.4 psi in an earlier post for collapse rather than 192 psi because of a reference to yield stress of 6316 psi for rigid polyvinylchloride material in a state of compression. Plastic properties vary in the literature because of manufacturing methods, HypoSmurf would need to obtain a valid MTR with shipment of product in order to validate his computation.

This computation is completely independent of the assumption of span. So TGS4, I rest my case.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[TGS4]

Cockroach,

I think I have figured out where you went wrong - you (inocrrectly) assumed that the "collapse" equations refer to elastic instability buckling. They do not - they refer to failure due to the pipe/vessel reaching the yield stress.

As you will find in any basic theory of elasticity textbook, the Young's Modulus will be a factor in a buckling analysis. Check out Euler's formula - where for columnar buckling the critical load:
Prc = (pi^2 * E * I)/(L^2)

So, unless you can find an ELASTIC BUCKLING equation that refutes what I and MikeHalloran stated, I again reject your assessment on this item. Elastic stability still does not seem to be your strong suit.

General statement - "For external pressure, the governing failure mode will be elastic buckling" Ipsa res loquitor.

 

[Cockroach]

Yes, and the loading condition in this case?

The answer is one end free and the other end fixed with a load parallel to the centreline. Is this the case we are considering here? I don't think so!

Anyway, I think this subject has been beat up enough. Take care.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[ApC2Kp]

To cockroach,
The PVC pipe manufcturer Harvel has a duct product with ratings given for internal pressure and external pressure loading. The duct product has thinner wall than for your size 20", but it would be interesting to compare lab tests of their .219" wall product rating to some of the predicted pressures of the previous posts.

http://www.harvel.com

The Harvel product rating is only 24" wc, that is inches of water column and not psi. If you want to factor the Sch20 0.375" wall thickness with a 2X or 4X plus Harvel's 5X design factor, then it would be only approximately 17 psi.
The 192 psi applied to a one foot length of the 20" pipe would calculate to 192 x 20 x 12 = 46,080 lbs / ft length.
That is a bit more than your 62 kg. Could you picture a 20" ring of PVC supporting a 46,080 lb truck? No way.
The length of the pipe under water is subjected to the buoyancy of the empty pipe. If the pipe is vertical, then there is not a bending effect, but a column loading effect. If the pipe is horizontally oriented and restrained at the ends, then it would be like gravity case of weight causing bending - except bending would be in the upward direction.

 

[Cockroach]

ApC2Kp, your point on buoyancy is well taken. Yes, this should properly be taken into account. On the other hand, HypoSmurf's problem is extremely simplistic and void of information that would assist us with that calculation. Obviously we can't address the specific issue of column buckling without length.

Depending on the manufacturer, wall thickness is expected to vary. I used a commercial pipe size table to obtain the cross sectional geometry. As others have pointed out on numerous ocassions, ovalness need be a consideration. Again, HypoSmurf would need to address this, the calculation is extremely simplistic and would need considerable project development to bring to reality.

As usual, UncleSyd brings an excellent point to the table, thermal gradient with water depth. The material properties for the PVC pipe would be expected to change somewhat at water depths measured on the order of hundreds of feet. HypoSmurf would need to consider material selection, hence obtain a superior MTR from his supplier.

Your observation about weight supported by the pipe. Remember my 62 kg mass is not over the entire surface area of the pipe, it is localized over the contact area of my foot. My foot contacts the surface of the pipe, ideally as a tangent to a circle, that is, at a point. What I am getting at is the issue raised of localized stresses and deflections. The 20 diameter schedule 20 pipe composed of "rigid PVC" material is surprisingly strong.

It is far easier to over analyze this problem. I would suggest that HypoSmurf is only interested with the order of magnitude depth a cross section of 20 diameter schedule 20 pipe could handle given atmospheric bore pressure. That is it, all we know, everything else is a complete unknown.

If anyone goes to the field and starts laying underwater pipeline based on this discussion, well, I give participants of this forum credit, I think we're all smart fellows here. Wouldn't you agree?

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[hyposmurf]

Well , well , I had post my question , in what I touhgt it was the simple way to do it .
Guess a horizontal pipe lying on a flat lake , one or both end are conected to the atmosphere.
I never guess that this question would make a so interesting discussion.

Now I will change my question.
I need to do a chamber with a 40" lenght pipe, the same pipe section diameter and material.
Both end caps are flat plates , tick enough to not to be in cosideration , guess 4" or more what ever is need .

The caps have a small grove and an "o" ring to seal it .
The caps are have 4 threaded bar to join it with the pipe, of course 8 nuts do the final job.
Like a old way of air filter.
Or as the caps where blind flange and the pipe a 40 inch length gasket , all arrange with the belonging stud or bolts and nuts.

All the array is vertical stand.
The atmosphere is normal or about 28.79174 inch Hg
The temperature is ambient

Now the final question , how low I can make vacuum inside the chamber????

Hope it does make a new and nice discussion , as the previous.

http://www.cadtutorforum.net

 

[hyposmurf]

Hi all , for the same question I post on other fourm, I got this answer

External pressure conditions are a little more complicated than internal pressure. For one thing, the surface area that pressure is acting on is larger for a given vessel(or pipe). There's a pretty good discussion of this situation in Roark's but the following may be of some help:

q = (t/R)*(Sy/(1+(4Sy/E)(R/t)^2))

q= external pressure
t=thickness
R=Outer radius
E=Young's Modulus
Sy=yeild strength

What do you think about , for the given values,

t=thickness t 0.375
R=Outer radius R 10
E=Young's Modulus E 406106
Sy=yeild strength Sy 5218

I got

t/r 0.0375
r/t^2 711.1111111
4Sy/E 0.051395448
(1+(4Sy/E)(R/t)^2) 37.54787447
(Sy/(1+(4Sy/E)(R/t)^2)) 138.9692512
(t/R)*(Sy/(1+(4Sy/E)(R/t)^2)) 5.21134692

So the final answer is 5.21 PSI .

Any comments???

http://www.cadtutorforum.net

 

[pennpoint]

hyposmurf
How does your original post relate to these last two. Why did'nt you just say in the beginning you wanted to build a vacuum chamber.
At sea level you'll have 15psi external pressure.
Connect a vacuum pump to one of the flanges.
Do not use PVC pipe.

Regards
pennpoint

 

[CRG]

hyposmurf,

Download the following +GF+ handbook, Chapter 3:

http://www.us.piping.georgefischer.com/external/us_handbook/03.pdf

Turn to page 8, and do the math for answers to both of your questions: Both pressure and vacuum.

CRG