SAFETY FIRST: HYDRAULIC CYLINDER TEST TO FAIL 1

[BEMPE16524]

I just want to share with all how safety is very important.
We were instructed to service an old hydraulic cylinder from one Client XXX. The working pressure is 250 Bar and test to 1.5 times.
During extend, at around 320-350 Bar, BANG!!!
The gland pop out. The thread were sheared off. Luckily no incident.

Now, for the next test after re-pair, need to add more safety features. Thinking of covering the perimeter with solid counter weight.

 

[KENAT]

Are you testing it hydraulically (i.e. non compressible fluid) or pneumatically (i.e. compressible fluid = spring = lots of stored energy to shoot components across the lab)?

If Hydraulically (with method of avoiding bubbles/air pockets in the fluid) then I'm not sure I see the great safety concern, a big drip tray and an old shower curtain plus some sawdust should do it.

Or am I missing something - entirely possible I've only been involved a tiny amount of this early in my career.

(If pneumatic then I suggest you look at some of the features of bomb sheds and the like from ordnance depots, and remote control of the test.)

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[DAVIDSTECKER]

Years ago a company I was working for made a brass packing gland for a hydraulic cylinder, why I do not know.
The gland threads failed and it shot down the rod until it hit the nut that was fortunately on the rod.
The gland hit the nut so hard that it left an impression of the nut in the gland.
If it wasn't for that nut there is no telling what damage the gland could have done.

 

[KENAT]

You know I half considered the spring energy of the actual gland and other cylinder components, DAVIDSTECKER (assuming no bubbles in your fluid) I suppose your story illustrates there can still be danger even with non compressible fluid.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[BrianPetersen]

Even if you test it hydraulically, it's unlikely that all the air pockets can be purged out in a practical manner. Doing this inside a test cell which consists of a concrete-block wall and with the controls preferably outside, or at least not in the foreseeable line of fire, is not a terrible idea.

I have a customer with some engine test cells (dynamometers) ... each one is in its own room with the door closed while it's running.

I believe one engineering term is RUD ... rapid unscheduled disassembly.

 

[rconner]

I believe you are most fortunate this as best I can tell (likely) was not a pneumatic test. I guess there are at least a couple good/cheap lessons here (as it is apparently reported no "incident"):

When testing completed hydraulic cylinders (particularly unrestrained from the outside), make sure the "stops" have been adequately designed. Also, confirm that they have been installed by the assemblers!

 

[Compositepro]

I wonder if the cap on the rod end was even designed to hold the rod in at such high pressure. I would assume that it should be but I would not bet my life on it.

 

[chicopee]

From the second picture, a few threads are still visible. When the gland popped out were all the threads damaged ?

 

[BEMPE16524]

Further investigation discovered that:
1. the cylinder has been used many years in subsea condition.
2. The housing outer layer has some corrosion.
3. During opening the original gland, heat flame was used which could contribute to the material becoming soften (the housing). The gland was actually new because the original was melted during heat flaming.
4. Calculation from the service contractor shows that the original housing size is not able to take such pressure. I wonder what standard the manufacturer used? the reason why original housing was used is because of cost factor.

i think the client try to hide as much info as they can. This is a lesson learn for us to ensure all info must be gathered sufficiently from all (but not all) the opportunist client who always want to save their budget.

it was hydraulically tested

Yes the thread is still available but the housing thread seems badly damage.

 

[desertfox]

Hi

I was looking at some hydraulic cylinder specifications the other day and I remember it clearly stated that the cylinders were to be hydro tested to only 10% above the system relief valve.
From memory hydraulic cylinders are not classed like standard pressure vessels but come under some machinery directive which might explain why the original casing was unable to take the test pressure.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[MikeHalloran]

Since it was intended for use in the presence of external pressure, and presumably with no innocent people in vicinity at that time, I wonder if maybe the original designer intended to use a hydrotest pressure less than 1.5x working pressure. It would have passed at 1.2x .



Mike Halloran
Pembroke Pines, FL, USA

 

[BEMPE16524]

They showed me a previous test certificate of 1.5 times.

 

[chicopee]

Do you know the manufacturer and if so contact it? Probably an ANSI Standard which number I don't know but what I know that Enerpac E324 catalog indicate ANSI (ASME)B30.1 Standard and ISO 10100 standard for compliance.
Your reasoning about overheating the housing is probably correct, altho. perhaps the threads were somewhat corroded from the elements of the sea.
Next time, make sure that you cover yourself from a liability stand point by indicating to your customers bringing in old equipment that failures as such could happen and you would not be financially responsible.

 

[BEMPE16524]

yes, it's a good point you have on the last sentence.

 

[MikeHalloran]

Did you monitor the rod end pressure during the test?

If that chamber is not vented or cross-connected to the blind end, the piston works as an intensifier, and I would expect the gland to pop out.



Mike Halloran
Pembroke Pines, FL, USA

 

[Compositepro]

Mike, interesting point about the pressure being intensified. That is true, but the load on the cap would not be increased by this effect. Maybe the cylinder was expanded by the intensified pressure.

 

[MikeHalloran]

If the piston is not mechanically stopped on the cap, even by .001", then the force on the cap is equal to the rod area x the blind end pressure, plus the rod end cavity area (piston area less the area of the rod itself) x the intensified pressure.

Did the customer or the OEM provide a detailed test procedure?



Mike Halloran
Pembroke Pines, FL, USA

 

[Compositepro]

"the force on the cap is equal to the rod area x the blind end pressure, plus the rod end cavity area (piston area less the area of the rod itself) x the intensified pressure."

Yes, that is correct. But that load is also equal to the piston area times the blind end pressure. Hence, there is pressure intensification, but not force intensification. This assumes that the cap area is equal to the piston area.

The load on the cap is the same whether the piston is contacting the cap or not, as long as there is no external force on the rod.

 

[BEMPE16524]

Mike,

The client insisted to test it to 375Bar. They said the working pressure is 250Bar.
Unfortunately no test procedure has been supplied. I have a thought that the failure was also caused by the continuous flow from the pump which acted like an intensifier.
Where in normal operation, the hydraulic cylinder has the counterbalance valve to control the flow.
Yes we have the control valve fitted during testing (set to 250Bar) but because the required test pressure, we had to continuously pumped in the pressure until it reached 375 BAr.

 

[RolMec]

Insisting on 375 bar, no test procedure, gross underperformance on correct business behaviour...
Has this cylinder been run continuously on higher pressure than it was designed for? So, running it usually at 300 bar WP the client just wanted to see a margin of 25% at the test. Perhaps needed more force than a WP 250 would deliver, once the thing was under the waves?
Can you look for damages on the other parts, rear cap, bolts, eyes/bushings? Fatigue cracke, even? If you have the cylinder still there, you might be able to reverse-engineer the actual pressure it can withstand from the dimensions. Any hint of the OEM? Contact them, at the first try of fingerpointing or invoice-dodging by this jewel of a client you are confronted with.
Best of luck!
Regards

RSVP