Understanding hydraulics and cylinder drift

[willie59]

And I would be curious as to what is causing aerated oil in the first place, that's not normal.

 

[maytag]

Check out the shaft seals on the pumps-I've had them go bad and suck air in to aerate the oil. The pump does not have to leak oil to suck air in but have seen them leak oil when not running.

Good luck in your endeavor,
Maytag

 

[mrtutz]

Thanks for the replies gents.

Apologies for the delay in responding. I was convinced that the only way for air to be entering the system would be though somewhere on the suction lines to one of the 2 pumps. I made a point of checking the hose fittings and clamp bolts on the suction lines first thing Thursday and got a nip on most of them (they weren't too loose however I reckon I got a turn or 2 on most of them).

Worked the machine for 20 mins or so then set it back up in the same position and started testing for drift. Test results were much better (4.5mm in 3 hours) so we are thinking we may have sussed it. We then operated the machine for a good 2 hours to get it warm and retested. The drift was back and worse than before!

I'm still thinking it's an air leak on the suction line (maybe one that opens up when the oil gets warm). Each suction line has an SAE fitting with an O ring on it at either end held on with 2 half moon clamps, and a ball valve by the tank end. I have known air to get in the suction line through a leak in
the ball valve spindle before now. Maybe some sort of smoke test near the suction lines will show a result?

I'll keep you posted

 

[tctractors]

The return to tank situation might well be worth checking out, it needs to be away from the oil feed tap and if possible piped low into the tank to limit frothing, HV 46 oil is a very stable bit of fluid to use Winter-Summer, the only thing that will compress in the oil is air, oil will not compress it acts as a solid media in the Law of Hydros (Greek Stuff) it is usual to find on Boom cylinders an amount of lift in hight when left in 1 position due to Sunlight heating up the cylinders, are your pumps load sensing V.P.S.P or just a direct pump set up like a bent axis piston pumps or gear pumps??? do both pumps draw from the same tank or is there 2 tanks with a ballance line?

 

[mrtutz]

Hi TC

Thanks for responding. Tank is massive for the application (450 ish L). Machine is using standard 32 grade hydraulic oil. 2 Rexroth piston pumps (pressure compensated). as soon as Hi rpm is selected the compensators move the swash plates one way which brings system pressure up to 200 bar. Idle pressure is approx. 20 bar. (adjusted on the pumps).

Tank has a baffle plate in the centre with the suction lines on the opposite diagonal to the return line. We have had the access plate off the top of the tank and sometimes at full flow you can see small bubbles coming back into the tank. However it's on the opposite diagonal to the suction ports! One tank for the machine, 2 main pumps and one small electric pump in case of main engine failure (main pumps PTO driven, back to back)

Can't see any froth in the tank, just what I'd call small bubbles in the oil (like what you'd see in flat lager or cider).

Regards

Phil

 

[theironoracle]

i am glad this just came to the top of "shop talk" i love this stuff thanks atco for this great post even though i am 2 years late..............theironoracle

 

[willie59]

You're quite welcome theironoracle, but one is never too late to noodle out how things work. In most cases, understanding theory of operation plays a big role troubleshooting problems.

 

[blitz138]

Hi TC

Thanks for responding. Tank is massive for the application (450 ish L). Machine is using standard 32 grade hydraulic oil. 2 Rexroth piston pumps (pressure compensated). as soon as Hi rpm is selected the compensators move the swash plates one way which brings system pressure up to 200 bar. Idle pressure is approx. 20 bar. (adjusted on the pumps).

Tank has a baffle plate in the centre with the suction lines on the opposite diagonal to the return line. We have had the access plate off the top of the tank and sometimes at full flow you can see small bubbles coming back into the tank. However it's on the opposite diagonal to the suction ports! One tank for the machine, 2 main pumps and one small electric pump in case of main engine failure (main pumps PTO driven, back to back)

Can't see any froth in the tank, just what I'd call small bubbles in the oil (like what you'd see in flat lager or cider).

Regards

Phil

Great explanation Willie!

Phil I if its just bubbles like a flat beer I doubt its enough aeration to cause the type of drift your describing.

 

[tctractors]

Phil, if the oil return to tank is good the only logical thought is that air is entering via the line from Tank to pump inlet, in the U.K. we tend to only use the Pressure Compensated set up to adjust the Delta P of the pump/s with the Load Sensing system being the prefered method to lift (activate) the swash of the pumps, on the Cranes that I install I also set into the ECU a warm up speed , 4 mins at 1200 rpm, then low idle speed around 700-750 rpm and 2 other speeds upto max rpm oil flow checked not max engine speed asI find using Load Sensing does not need top engine design speed, the boom hoist cylinders are usualy way up to the task of holding things up, I have more trouble with the extension side (Recycling Vave) on the stuff I work on, as your crane has a 2 pump through drive it would possibly have 2 oil feeds from the tank?? are both pumps making Froth? is it possible to swap around pipe work on pumps?

tctractors

 

[nutwood]

What an excellent discussion to discover. One question I have, and I qualify this by saying that I haven't done the sums, but, in the circumstance where we assume the piston seals are cactus and we're assuming that the load is being supported by the difference in displacement between the above piston space, with rod, and the below piston space, without rod; surely there's potential for fairly severe amplification of pressure? It seems to me to be a form of leverage and I can imagine that in many circumstances this could be enough to open a pressure relief valve and allow the load to drop.
The point I'm making is that whilst theory says that dodgy piston seals can't allow the load to drop (without squirting oil out the rod seals!), surely they can cause excessive pressure in other areas which will result in the same thing?

 

[willie59]

What an excellent discussion to discover. One question I have, and I qualify this by saying that I haven't done the sums, but, in the circumstance where we assume the piston seals are cactus and we're assuming that the load is being supported by the difference in displacement between the above piston space, with rod, and the below piston space, without rod; surely there's potential for fairly severe amplification of pressure? It seems to me to be a form of leverage and I can imagine that in many circumstances this could be enough to open a pressure relief valve and allow the load to drop.
The point I'm making is that whilst theory says that dodgy piston seals can't allow the load to drop (without squirting oil out the rod seals!), surely they can cause excessive pressure in other areas which will result in the same thing?

Good point nutwood. And as you stated, one would actually have to work out the math, but pretty much anything is possible. The reasoning for me posting this info was not to imply "this is the way it is and no way else". On the contrary, it was a response to so often when an individual is experiencing cylinder drift that they automatically consider it to be failed piston seals. That was my main point, that this is not always the case, and I attempted to present an argument that supported such. Yes, things can get way more complex than the situations that I described, which is all the more reason that one should do a little troubleshooting before they pull a cylinder apart because it's leaking down.

 

[lantraxco]

Nutwood, you are exactly right, when a cylinder changes from a piston type cylinder to a displacement type cylinder due to the piston seals leaking pressures can increase by orders of magnitude. If a cylinder has a 4" piston, the area of the piston is roughly 12 square inches. A 12,000 pound load would create a 1,000 psi pressure on the piston side of the cylinder. If the same cylinder has a 2" rod, then the displacement area of the rod is roughly 3 square inches, if the piston seals are bad then the same 12,000 pound load would induce about 4,000 psi on BOTH cylinder ports, piston and rod end. If there are no relief valves opening to tank, then damage can occur to the cylinder tube, rod packing, hoses, steel lines, or even the control valve. Note that a crossport relief block won't help at all because the pressures are equal on both lines due to the piston seal leakage.

Sorry, just felt like preachin' for a minute I guess, lol

 

[nutwood]

Nutwood, you are exactly right, when a cylinder changes from a piston type cylinder to a displacement type cylinder due to the piston seals leaking pressures can increase by orders of magnitude. If a cylinder has a 4" piston, the area of the piston is roughly 12 square inches. A 12,000 pound load would create a 1,000 psi pressure on the piston side of the cylinder. If the same cylinder has a 2" rod, then the displacement area of the rod is roughly 3 square inches, if the piston seals are bad then the same 12,000 pound load would induce about 4,000 psi on BOTH cylinder ports, piston and rod end. If there are no relief valves opening to tank, then damage can occur to the cylinder tube, rod packing, hoses, steel lines, or even the control valve. Note that a crossport relief block won't help at all because the pressures are equal on both lines due to the piston seal leakage.

Sorry, just felt like preachin' for a minute I guess, lol

Not preaching, good stuff! Thinking about this, and considering that few seals are absolutely 100% tight, I'd suggest that this potential pressure amplification is a really good reason not to leave a load on a hydraulic cylinder that's not been designed for continuous load holding.
I've always considered it bad practice but I've never considered that, combined with worn piston seals, you could be seriously stressing other parts of the system. Fascinating stuff!

 

[oceanobob]

When I tried to explain some of this stuff the other day, the guy politely told me 'all you need to do is remove and re-install the cylinder the other direction and that will show you it doesnt matter'....

No need to comment, stuck this in here for the pure amusement.

 

[theironoracle]

A customer the other day was telling me about a cracked boom on his 3000hr 324d cat excavator. I have seen plenty of cracked booms but this was the most premature I had ever heard about. I know the conditions this machine has worked in since new and it has only been standard use. Because of this thread I got to thinking what could have caused this? My conclusion that perhaps one of the cylinders had bad piston seals. So I explained to them over the phone to lift the bucket off the ground and see if it settles then stops (equalizing pressures in all four fluid zones of cylinders). Sure enough they said it dropped slowly then stopped! Next I instructed them to safely isolate each cylinder and determine which one allowed the drift and which one had no drift. They identified the bad cylinder and repaired. Thanks to all who post high quality info on HEF, I love the theory of mechanics.........TIO

 

[willie59]

Nutwood, you are exactly right, when a cylinder changes from a piston type cylinder to a displacement type cylinder due to the piston seals leaking pressures can increase by orders of magnitude.

Right on Don. :yup

I'd suggest that this potential pressure amplification is a really good reason not to leave a load on a hydraulic cylinder that's not been designed for continuous load holding.

Which is why crane and aerial lift cylinder designs have to account for this and be designed strong enough. With cranes (that support loads over persons/property) and aerial lifts (that support the persons themselves over people and property) you simply can't have failures like this. Result would be major damage and/or people dead.

When I tried to explain some of this stuff the other day, the guy politely told me 'all you need to do is remove and re-install the cylinder the other direction and that will show you it doesnt matter'....

No need to comment, stuck this in here for the pure amusement.

Exactly, never argue with an idiot...they will defeat you with superior ammunition.

Thanks to all who post high quality info on HEF, I love the theory of mechanics.........TIO

It is some awesome info that folks have posted here TIO.

 

[Labparamour]

Started the thread thinking, "huh?"
Then, reading more of the posts, and thinking about it all it made sense: the rod changing the volume of the barrel. I think the single-acting forklift example brought it together for me. Thanks for the post!

 

[td25c]

Great thread willie59. I have one piece of equipment that uses the displacement cylinder princible. The old Shaver post driver,mine has about a 2 inch cylinder bore using a 3/4 inch dia rod. Instead of a piston & seals it has a finned guide at the bottom of the rod that lets the oil pass by it quickly allowing the H beam hammer to drop fast.So it is using the 3/4 inch displacement to lift the 300 lb hammer. www.shavermfg.com/standard-post-drivers

 

[lantraxco]

Iron Oracle:

Great bit of diagnostics there, the twin boom cylinder situation never occurred to me, the racking must have been intense with one cylinder carrying the load!

 

[pajibson]

I found this presentation of the very thing you are discussing on this thread.

http://www.hydraulicsupermarket.com/flash/cyl_drift_video.html

it might help with the more "visual" learners.