OT: Hydraulic Circuit Problem

[scarince]

This discussion has turned to the core of what I'm concerned about. My problem only went away when I bled air out of the line, but that could have allowed some debris to become dislodged from some part of the return circuit. When I say I bled the line, I really only burped it, so I have more work to do.

I'm always concerned when a problem goes away and I don't really know why. On top of that this is a production process currently making parts. The force output is a key process parameter that has a huge impact on product quality, and the defect can only be identified by putting it on a the car or by destructive evaluation. So I really want to find some debris or some other smoking gun.

I think I'll take the line apart as much as I can and flush them out and then bleed it really well as several of you have suggested.

If I find anything interesting I'll let you know!

Bill

 

[Mark Snodgrass]

The flow control valves on your return lines need differential pressure for them to work. When you have that air pocket in the return line between the cylinder and the flow control valve that air pocket is reduced in volume as its pressure rises. That pressure of the air bubble does not get transmitted to the fluid between it and the flow control valve, and thus the flow control valve will not open and allow the fluid to move. That is basically what an air lock is in a hydraulic line. If the flow control device was an orifice type it probably would have worked and the air bleed out eventually on its own.

 

[Mill_Control]

Ya, this is the part that worries me. I'm under pressure to explain exactly this point, and the fact that I can't has cast some doubt on my explanation for the root cause.

I'll get it bled out and write some logic to detect this before each cycle. That will hopefully cover me.

Doesn't really matter what caused the problem. A multitude of things could have caused it. Maybe even air though I'm inclined to agree with Peter on that possibility.

Did you repeat the increasing pressure until it moved test or just do that once then crack the line? If you didn't you don't know whether increased pressure broke something loose.

Any which ways toughening the code to detect the problem next time is what really matters. And what the people pressuring you would really be interested in if they had a lick of sense.

 

[scarince]

Did you repeat the increasing pressure until it moved test or just do that once then crack the line? If you didn't you don't know whether increased pressure broke something loose.

Yes I did do that, and the rod would extend at high pressure then it would retract. I did mention that there is a tool set that weighs about 650 lbs or so sitting on top of this cylinder and under normal circumstances it would take ~ 400 psi to lift it. But I observed that the once I got to around 1100 psi the cylinder was fully extended and as I lowered the pressure the cylinder would start to collapse right away at 10K psi, more at 9K psi, etc.

So if I had a solid obstruction, I think it's probably still in the line or in the valve. I'm going to keep an eye out for it when I flush the steel line.

 

[scarince]

The flow control valves on your return lines need differential pressure for them to work. When you have that air pocket in the return line between the cylinder and the flow control valve that air pocket is reduced in volume as its pressure rises. That pressure of the air bubble does not get transmitted to the fluid between it and the flow control valve, and thus the flow control valve will not open and allow the fluid to move. That is basically what an air lock is in a hydraulic line. If the flow control device was an orifice type it probably would have worked and the air bleed out eventually on its own.

Mark, I've never heard this explanation before. Can you help me understand *why* the pressure of the air bubble does not get transmitted to the fluid? The flow control is a Vickers DGMFN-3y. I *think* it works by metering the flow past a needle valve, but I'm not positive. Why won't the system pressure just push the air bubble through the flow control?

 

[milldrone]

Mark, I've never heard this explanation before. Can you help me understand *why* the pressure of the air bubble does not get transmitted to the fluid?

Why won't the system pressure just push the air bubble through the flow control?

I'm more suspicious of the two pilot to open checks right next to the flow control.

Depending on the particular failure mode they are capable of some very strange things. Like rupturing the cylinder tube from pressure intensification, the piston seals were leaking, and the pump was off!

 

[OdinIII]

You are right, this is simple, it just isn't simple to find the smoking gun. The problem, if we ignore the possibility of a mechanical issue, had to be excessive pressure in the other side of the cylinder. Air and Fluid should easily be pushed back to the tank by much less than 400 PSI unless there was a blockage in the return line. That narrows it down to the debris, flow control, valve, or pipe. Good luck.

This is one reason I wish it were easier to have working gauges on both sides of cylinders. It makes everything clearer when you can see the pressure that is applied/held.

Your experience with the "hydraulic guys" is similar to mine. If they can't articulate the how behind what they are saying then they are probably just repeating something they've heard at some point and don't really have a clue.

 

[osmanmom]

check pump wear-off and oil/quality condition

 

[osmanmom]

another question what the latest modification you have done

 

[Mark Snodgrass]

Scarince, your piping as you described it is acting like a manometer (an upside down 'U') with the air rising to the top. The hydraulic fluid pressure must overcome the buoyancy of the air in order to exert the pressure further down the line to the flow control valve. It has been over 20 years sine I studied the physics behind it, and I don't use it everyday to be able to explain it better than that. Look up airlock on Wikipedia, it shows an example of airlock in a gravity fed system.

On a further note, the flow control valve you are using is basically an adjustable spring pushing a ball against a valve seat. When the flow of fluid is from the spring side it acts as a check valve pushing the ball against the seat. When the flow is from the other direction the pressure must first overcome the spring pressure before the ball comes off its seat, this is how it provides speed control as adjusting your spring tension changes the amount of back-pressure on the cylinder.

Given these two facts explains why 400psi would not allow the cylinder to move. At 400 psi your cylinder exerts a force of 900 pounds, and your load is 650 pounds leaving 250 pounds of force acting against the fluid on the rod side of the piston which is about 150 psi on the fluid if it had no air in it. But since the air is in the line that energy from the fluid is used to compress the volume of the air which results in a lot less pressure (can be calculated with ideal gas law if volume of air was known).

I hope this explanation helps you out.

 

[OdinIII]

Given these two facts explains why 400psi would not allow the cylinder to move. At 400 psi your cylinder exerts a force of 900 pounds, and your load is 650 pounds leaving 250 pounds of force acting against the fluid on the rod side of the piston which is about 150 psi on the fluid if it had no air in it. But since the air is in the line that energy from the fluid is used to compress the volume of the air which results in a lot less pressure (can be calculated with ideal gas law if volume of air was known).

I hope this explanation helps you out.

Please don't take this the wrong way, I'm trying to learn.

For the piston to stop moving wouldn't there have to be no flow on the other side of the cylinder? Wouldn't this mean that the pressure has to be the same throughout the pipe or at least to the blockage?

It seems to me that there would be hundreds of PSI on that side of the piston that should only have to overcome the head of the fluid to flow back to the tank.

 

[Peter Nachtwey]

There is no way an air bubble is going to keep oil from flowing as long as the air bubble has some place to go. Oil will push air out of a upside down u pipe easily. It may not push all of it out but there is no way the air will keep the oil from flowing by.

 

[pal]

Scarince - In your post #19 you say that as you reduced pressure the cylinder moved down a corresponding distance . This can only be because air is trapped above the piston and the air expands as the pressure is reduced moving the piston . Check the Flow control valve between the top of the cylinder& the 3 position valve preferably out of the pipework . If this valve is OK ,then check the 3 position valve . Another posibility is the top hose is faulty and causing a blockage - very rare though .
Let us know how you get on !

 

[scarince]

I just figured something out......I'll post as soon as I get cleaned up!

 

[Mark Snodgrass]

Peter, the air in the pipe is compressible and the pressure applied to it will be converted to a reduction in volume until the air bubble pressure and the fluid pressure equalize at a lower pressure than would be present with the pipe filled with an incompressible fluid. Since there is a flow control valve that will only open when enough pressure is applied a large enough air bubble could cause the pressure to drop enough to prevent flow. The cylinder is only stroking 1" so it would not take even a very large bubble to reduce the pressure between the cylinder and the flow control valve enough to prevent the flow control from opening.

If there was no flow control at all then the air bubble would have no real effect and it would flow with the oil. The reason increasing the pressure worked is that it was able to compress the air bubble enough to have enough pressure to overcome the valves opening pressure. The air would not leave the piping though because the lines (per the drawing) are long and the volume of them is much greater that the volume displaced by the 1" stroke cylinder.