Hydraulic pressure test. Pressurization speed 5 bar/s 2

[raton]

Hi there.
I am new here and this is my first post.
I have received a pressure test procedure and the pressurization and depressurization speed are 5 bar/s and 10 bar/s.
I am not sure if these values come from any standard or are the good engineering practices for pressure test speed.
Could you please help me?

Thank you very much
Jaime (Raton)

 

[MintJulep]

Not until you ask an answerable question.

 

[raton]

Sorry for that.
My question is related to the pressurization speed in a pressure test @ 315 bar for a sea water valve.

The procedure says that the increase in pressure should be 5 bar per second so from 0 bar to 315 bar it will take 315/5= 63 seconds.

I would like to know if this speed to increase the pressure (5 bar/s) is a normalized value or if this is normal practice.

The same applies for depressurizing the valve from 315 bar to 0 bar.

Thank you very much and sorry again.

Jamie

 

[zdas04]

A 0.50 caliber bullet masses 620 gm and travels at 887 m/s, it takes 9E-7 seconds to traverse a piece of 0.8 mm steel and it has a momentum of 550 N-s.

If you threw a 60 kg mass at 9 m/s you would get the same momentum, but instead of penetrating the steel it would just dent the hell out it.

Same momentum, about the same total force, much longer duration. Duration really matters.

Same thing happens with pressurisation/depressurization. Going from 0 to 315 bar in 32 seconds is a really good way to turn a fabrication into scrap metal. If you give the elastic properties of the metal enough time to equilibrate, you'll get a test that means something. If you don't, then places of localized stresses will reach a peak stress and not be able to transfer the stresses to adjacent pipe and can just come apart.

Whoever wrote that you could go up at 10 bar/s and down at 5 bar/s is making a silly and very dangerous mistake. My guess is that he took a U.S. procedure that said you should go up at 10 psi/s (which is common, if not very conservative) and down at 5 psi/s and "converted" it to metric by replacing psi with bar. Just for reference, 10 psi/s would get you to 315 bar in about 8 minutes. That is REALLY fast. For a liquid test I usually use 50 psi/min (3.4 bar/min) with a 15 minute soak every 1,000 psi (69 bar). To get to 315 bar, I would take 90 minutes of active pressurization plus 15 minutes soak time every 69 bar (total 2.5 hours, not 32 seconds).

David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

"It is always a poor idea to ask your Bridge Club for medical advice or a collection of geek engineers for legal advice"

 

[zeusfaber]

David,

Not a criticism as you're making lots of sense - but I think it would be useful to see another version of your last paragraph critiqueing a proposal that's "up at 5 and down at 10", instead of the other way round.

A.

 

[micalbrch]

I read David's answer with a lot of interest and everything he wrote makes sense for me. But what is the purpose of Jaime's 315 bar valve? I think when we know that we might understand why these requirements are specified.

 

[zdas04]

I took the 315 bar requirement to mean that that was his hydro pressure.

Another shot at the last paragraph:

Whoever wrote that you could go up at 5 bar/s and down at 10 bar/s is making a silly and very dangerous mistake. My guess is that he took a U.S. procedure that said you should go up at 5 psi/s (which I haven't seen before, but is just barely within the realm of possibility) and down at 10 psi/s and "converted" it to metric by replacing psi with bar. Just for reference, 5 psi/s would get you to 315 bar in about 16 minutes. That is really fast. For a liquid test I usually use 50 psi/min (3.4 bar/min) with a 15 minute soak every 1,000 psi (69 bar). To get to 315 bar, I would take 90 minutes of active pressurization plus 15 minutes soak time every 69 bar (total 2.5 hours, not 63 seconds).

Not many changes.

David

 

[ivymike]

well, seawater valves are unfamiliar to me, but I regularly run other valves all the way up to and back down from 190 bar in around 0.015 seconds (11400 bar/sec). I give them about 0.1 seconds before repeating the process. They're usually at about 400C when I do it, and nearly 100% of them will last over 20k-hrs in such service. They don't typically have a very thick cross-section.

 

[raton]

Hi All.
Thank you very much for your useful answers.

The pressure test is for a valve installed in a submarine.
This submarine will see a depth of about 550 m and the valve will be placed in the pressure hull. This valve will be responsible for the inlet and outlet of sea water from the cooling services on board the ship. This ball valve will have double restriction in one body, which means two ball valves in one body.

I have a quick question for Ivymike.
What kind of valve are you testing? seems that from 190 bar in around 0.015 seconds is very quick!!!!

Thanks in advance.

Jamie

 

[micalbrch]

Jamie, I assure you that you violate any confidentiality agreements your company signed when you post the diving depth of your client's submarine here. Be careful!

 

[raton]

Thanks Micalbrch
No worries, this depth is not a sensitive value... it is just an aproximate one.
Basically it is the normal depth for most of the submarines in the world.

Cheers

Jaime

 

[ivymike]

inlet and exhaust valves on diesel engines (and the nearby components as well) see pressure rise rates as quoted above, and higher. The "test" times listed above were for operation at about 900rpm, but diesels with firing pressures that high (190 bar +) also operate at 1500, 1800, and maybe even higher speeds. The internals don't shatter (or at least don't shatter very often).

I just pulled up an arbitrary set of cylinder pressure traces, averaged the measured cyl pres between two cylinders (to make it more "typical"), and found that for a particular engine and rating, the maximum rate of cylinder pressure rise over a firing event at full load at 900rpm was 700kPa/deg, which translates to 37800 bar/sec. Average over 40 deg period during firing was 350kPa/deg. Fastest on the down side was -407 kPa/deg, with a -300 kPa/deg average rate over 40 deg.

So anyway, it's interesting to me that the non-engine components are so sensitive to rate of pressure rise. The components I'm used to are, but in a completely different rate regime and the main effects are due to inter-body behavior rather than intra-body behavior.

 

[zdas04]

ivymike,
I've never thought about the hellish environment that engine valves see continiously. One possible difference is that static tests are done at ambient temperature and engine valves operate at significantly elevated temperature. Cooler steel is far less elastic than warmer steel and normalizes stresse much slower. Whatever the reason, I've seen static tests fail from pressurizing the line too fast when I was sure the system would have passed with a much slower rate of pressurization.

David

 

[ivymike]

yes, the valves do spend most of their life at higher temps. Engines do, however, get started up on cold days and it would be a happy coincidence (in some cases) if pressure rise was as slow as at steady state operation. You can definitely see hundreds or thousands of firing cycles where all the components are below 32F (especially in a truck engine). You probably see a few hundred firing cycles before the in-cylinder components are really starting to get hot. The in-cylinder parts are made out of many different kinds of steel and iron, and most see some approximation of the abuse described above.

perhaps a difference between slow vs. fast pressurization is in the momentum of the fluid used to pressurize the line (it has to stop @ the valve, which may increase pressure at that end considerably)?

 

[dhengr]

Zdas:
I thought your original explanation was very well presented, even though someone wanted you to change a few numbers a little bit. And, I thought your bullet vs. 60kg mass was relevant and educational as related to your explanation and situation. Then, ivymike had to stick his nose in the discussion, and get things all complicated again. Now, I agree with you both and are confuseder than ever.

As long as the materials we use act properly within the range of their operating temps.; that is, maintain their yield and ultimate strengths and toughness and their modulus of elasticity and other mechanical properties; and we pay some considerable attention in our design and shaping and sizing of them so as to minimize stress raisers and stress concentrations; and then treat them as best the design will allow as regards fatigue and shock loading, we seem to come up with some pretty nifty machines. Don’t we? My questions are: can you actuate a ball valve fast enough so the OP’ers. test makes any sense; it may have to resist this pressure but the sub can’t dive fast enough to cause the pressurization rate; zdas is talking about an incompressible fluid test, and ivymike is talking about a highly compressible fluid problem, how does this enter the discussion; zdas assumed a fairly large complex fabrication or casting, while ivymike is talking about smaller, generally symmetrical, well refined parts, of special materials, how does this enter the discussion? You guys are a good deal more versed on fluid mechanics than I am after all these years. So, let’s you and him fight, I’ll watch and hold the bets.

My, my, what complicated things we engineers are asked to design and explain to someone who just wants to turn the valve on and not have it explode, or someone who just wants to get in his truck and drive away without warming it up.

 

[zeusfaber]

I'll just take a second to apologise here for not noticing the shift from bar/s to bar/min until David had already indulged me by reposting.

It's possible that the pressurisation rate for the hull valve is being driven by the valve being opened to admit pressurised water while the boat is already dived.

A.

 

[zdas04]

I think the big difference between pressurizing very slowly for a test vs. a crash dive giving you the same dP in a few seconds is "why would you do a test that fast, no one is shooting at you?". Of course in the real world, things happen at the speed that they happen. That doesn't mean that a hydro (which is designed to be significantly more pressure than the system will ever see in real life--either that or your MAWP is wrong) has to replicate that crash dive. The hydro is trying to prove strength, not toughness, elasticity, or hardness, just strength. I don't see any reason to approach limits of toughness, elasticity, or hardness in a strength test. Consequently, I change pressures slowly.

David

 

[ivymike]

hmm.. we do also do hydraulic pulsator testing of heads, blocks, liners, and pistons... typically at room temperature and to around 2x the intended firing pressure. The big reason to cycle quickly in those tests is that it would take a really long time to get 6E+6 cycles otherwise. The tests seem to take about 4-6 months to complete, so it's possible that we're going up & down in pressure as slowly as 2.5 sec/cycle (which would give about 76 bar/s each way)... but typically the tests will be slow to get started and something will cause a delay partway through, and the cycle involves periodic inspections, so I suspect the cycling rate is more like 1hz. The test is, of course, intended to evaluate high-cycle fatigue of the parts albeit at relatively low temperatures (where the materials in question are in general substantially stronger than they will be in hot operation).

To be honest, it had never occured to me before this thread that testing quickly was more likely to break the parts than testing slowly (for a given load range and number of cycles).

 

[ivymike]

hmm... maybe engine parts are designed for a much higher number of cycles at the max pressure? How many cycles would a seawater valve be designed for? Certainly not 30E+6, right?

 

[raton]

Hi All,
thank you very much again for your valuable posts.
Basically the question is answered, however I would like to support the final discussion with some standard or code regarding the speed to increase and decrease the pressure during the pressure test.

From the top of your heads, is there any standard that deals with this kind of topic?

Thank you very much in advance

Jaime