Proportional valves and dirt in oil. Good/No Good?

[Skogsgurra]

Hello. I am active in the electric/electronics field of the automation business and know precious little about hydraulics. So I may be using words that you do not fully understand. Please ask again if I do.

The problem: A press with hydraulic cylinders moving two tables up and down. Each table has its own hydraulic circuit with proportional valves that are controlled from a PLC with analogue output cards, +/- 10 V.

It is important that the lower table moves faster than the upper table or there will be a collision between the two when going down. The same thing applies when going up, but upper table has to be faster - of course.

Sometimes, the upper table doesn't move fast enough when going up and sometimes the lower table doesn't move fast enough when going down. Which results in collision and sometimes ruined products. But always reduced production since the machine has to be reset manually.

We have studied the machine for two days. We have recorded the control voltages for the valves and we have also recorded the feed-back signals from the valves (the spindle's actual position). Control voltage and feed-back are identical. A 7 V control voltage results in a 7 V feed-back and X V results in an X V feed-back. No ripple, no delay, no overshoot in the feed-back signals.

My question is: Is it time to get the hydraulics guys in and make them clean the system? They have taken oil samples and are showing codes like 13-4 (or something like that) and they say that there is no problem with the oil. What else could we do?

Good ideas needed.

Gunnar Englund

http://www.gke.org

 

[Ashereng]

Here is an observation - the problem is intermittent. "Sometimes" the tables collide. This means, more times, they do not.

Could this be a timing thing? Sometimes, when you send 2 signals to 2 channels in an output module through your controller, they do no get there at the same time.

Maybe try sending the signals to move in sequence. Send the signal to move the upper table say 0.01 seconds before sending the signal to the lower table when moving up. The opposit when moving down.

It's a thought.





"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[Skogsgurra]

A good thought. But the valve feed-back signals are recorded simultaneously and we can see that the outputs are well timed. The movement takes about three seconds and we can actually see how one table or the other slows down.

The valve feed-back is recorded with 12 bits resolution and 80 S/s, so we can say without any doubt that the electrical signals are as they shall be. That's why I was starting to wonder about dirt in the oil. Any thoughts there? Can particles in the oil upset the functioning of proportional valves? What does the code 13-4 mean?

Gunnar Englund

http://www.gke.org

 

[Ashereng]

Despite my field "petroleum", I don't know very much about hydraulic oil. Sorry.

I don't know the effect of dirt in oil on movement. However, I would think that the dirt's effect should affect both tables equally (ie slow them both down)?

If your table takes 3 seconds to move, is it possible to series the outputs to the table as a test?

I too have seen cases where the feedback seems to indicate all is well, but reality indicates something is amiss.

In any case, if a staggered start is not a problem, for your product, it may be worth a trial.

"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[budt]

You don't mention the manufacturer of the Proportional valve. Does it have a Spool position Transducer? Without a Spool position Transducer it is impossible to know if the constant input signal actually moved the Spool to the same position each cycle.

From your description, the slow movement is always while a cylinder is retractiing. If system maximum pressure is at or near the pressure required to move the cylinder and load while acting on the reduced rod end area youcould be losing flow to the lagging cylinder due to pressure drop even though the valve is shifted to the same position.

Is there an Accumulator in the circuit? If not I would be adding one. An Accumulator would not help a low pressure problem but will help actuator response especially if the pump is pressure compensated/variable volume.

Contaminated oil can make a Spool valve sluggish and/or reduce its movement from a constant signal but is usually a continuous problem and continues to get worse with time.


Bud Trinkel CFPE
HYDRA-PNEU CONSULTING, INC.
fluidpower1 @ hotmail.com

http://www.fluidpower1.us

 

[Skogsgurra]

Yes, as I mentioned, there is a valve position feed-back telling exactly where the valve is. The valves are Parker.

Interesting tip about the pressure. We have between 50 and 80 bars (ca 700 - 1100 PSI, I think), but I haven't checked piston areas or weight of tables. I think that the mechanical guys should do that.

Gunnar Englund

http://www.gke.org

 

[budt]

A stop-gap fix could be the addition of a limit switch on one of the tables that would be made before they contact each other and cause a problem while moving. The limit switch would signal the faster cylinder to slow or momentarily stop until there is ample clearnce between the tables.

At least production would not be interupted.


Bud Trinkel CFPE
HYDRA-PNEU CONSULTING, INC.
fluidpower1 @ hotmail.com

http://www.fluidpower1.us

 

[Skogsgurra]

Thanks,

My job is to somehow find the root cause for the problem. Not to redesign the machine. I know very little about hydraulics. That's why I asked here. The electronics and control part of the equipment is proven to be OK. I now have to convince the mechanics/hydraulics guys that they have to do a little more than say "It's an electronic problem" - so I have to get some facts and tips so I can meet them on their own ground.

Gunnar Englund

http://www.gke.org

 

[budt]

One thing about the Proportional valve that can cause a problem like you are having is the fact that any valve with a symetrical spool has the same opening for pump flow into the cylinder as exhaust flow from the cylinder for a given input electrical signal. That means the flow paths are the same but flow that is trying to pass to tank from the Cap End can be twice as much as the pump is sending to the Rod End when the cylinder is retracting. Oversize rods are the greatest problem.

So a set flow to a cylinder extending is trying to control speed by metering oil to it while the same set flow for a cylinder retracting is metering flow from it. This could be overcome in some cases by using different electrical inputs to allow for more flow from the Cylinder Cap End while retracting.

In your case the problem with differing flows is exaggerated by one cylinder having its load pushing on retract stroke and the other one with the load pulling on its extend stroke.

Is that explanation anywhere near clear?


Bud Trinkel CFPE
HYDRA-PNEU CONSULTING, INC.
fluidpower1 @ hotmail.com

http://www.fluidpower1.us

 

[Skogsgurra]

Yes, that makes a lot of sense. The cylinders are working from above and the upper table is quite heavy. That would mean that the piston part of the cylinder has to do the heavy work while the cap end does no work at all?

Would that imply that a slight increase in friction could be a possible cause for the problems? I mean, the cylinder is strong "the wrong way".

I think that there is a rather broad pressure reserve. Time to increase pressure perhaps?

Gunnar Englund

http://www.gke.org

 

[budt]

I would think an increase in friction would retard movement and make the loads act more like resistive loads and be a help in the over running upper table while making the lower table move slower.

Increasing pressure will overcome the flow resistance if the present working pressure is close to maximum system set pressure.


Bud Trinkel CFPE
HYDRA-PNEU CONSULTING, INC.
fluidpower1 @ hotmail.com

http://www.fluidpower1.us

 

[Ashereng]

Gunnar, I would think that a design issue such as this with the spool valve would tend to result in a consistent problem, not an intermittent one such as what you are observing.

As budt mentioned, if the pressure is near the limit of what is needed, the system may have difficulty overcoming momentum - sometimes, resulting in the intermittentness that you are seeing. At these conditions, the very little dirt in oil may be sufficient to cause an effect whereas normally, the effect is negligible?


"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[maytag]

Am I understanding this correctly-the upper table is slower to raise but it outruns the lower table when both tables are lowering- is this correct. You mentioned the upper table being heavier. No mention was made regarding cylinder sizes- are both cylinders the same size. I understand you to say there is feedback from the valve as far as spool position but no mention of feedback for actual cylinder position. I assume( maybe wrongly)that both circuits are common to one power unit. The fluid will always do the least amount of work first-you can relate to this with your electricity always taking the path of least resistance. If you had cylinder or table position feedback you could control the valves to maintain proper movement and position- with no actual feedback from the cylinders or tables I don't think you will ever get the results you want. Maytag

 

[Skogsgurra]

Getting somewhere now. Yes, same pump for both units. Do not know about cylinders etc. Not my table, so to speak.

If this is the problem, why does it work at all? Sometimes for hours before you get any problem.

Gunnar Englund

http://www.gke.org

 

[maytag]

For consistent operation some type of feedback will be needed-this will compensate for changes in load and temperature both which affect flow- basically with the Parker prop valves you are metering in both directions. We use a lot of temposonics in our cylinders and use some cable resolvers. I'm not up on any Parker valves but know from experience with the Rexroth line that not all prop valves and controllers are suited for fine motion control. Maytag PS To answer your first question, dirt (contamination of any kind) is bad for any hydraulic system component-generally servo valves being the least tolerant followed by high performance prop valves but I personally don't think this is the problem here.

 

[Skogsgurra]

Thanks. This is really very helpful.

Up till now, I have been thinking dirt mostly.

We found brass slivers in the filter container only a couple of weeks after it had been cleaned last time. They are about 100 microns wide and 500 - 1000 microns long. Have no idea from where they come. Some machining taking place inside the system? Or leftovers from the installation?

The hydraulics guys don't seem to be worried about it. Shouldn't they be?

Gunnar Englund

http://www.gke.org

 

[Skogsgurra]

maytag,

"Cable resolver" would that be the Celesco kind of stuff?

Gunnar Englund

http://www.gke.org

 

[maytag]

Sounds like you have the leftovers from a pump/pumps failure floating around in the system. And yes, I would be concerned if it was my equipment.
The "Celesco" you mentioned is similar to what we use- we use Gemco and Amtek I believe. The cable length is matched to the cylinder stroke length and work fairly well. Good luck, maytag

 

[Skogsgurra]

maytag,

There was actually a pump problem about a year ago. The pump was running dry and got destroyed. It was changed for a new one. The problems have been there ever since.

I guess you hit the nail when you said "Sounds like you have the leftovers from a pump/pumps failure floating around in the system".

Anyone else being of same opinion? Or other thinking?

Gunnar Englund

http://www.gke.org

 

[budt]

Should have asked up front if there was any out of the ordinary happening with the circuit before the problem started. Just never thought to, dumb.

Good information can usually narrow down the problem area/source. I always try to talk to the maintenance person on the job and often ask the machine operator what they observed just before the troubles started. They may not know hydraulics but they usually know their machine and can tell you when it sounds/acts/performs in an unusual way and can often point to an area where an unusual noise/jerk/wrong motion or other thing took place just before the problem started.

I would say the oil needs cleaned up and probably a complete drain and flush is required.


Bud Trinkel CFPE
HYDRA-PNEU CONSULTING, INC.
fluidpower1 @ hotmail.com

http://www.fluidpower1.us