Problem - Hydraulic Cylinders vibrate when stopped.

[AndreyG]

Please help me to figure out hydraulic problem – I am not hydraulics expert. I got large ship steering system: tiller operated with two cylinders, controlled by directional valve. When the valve closes the rudder vibrates instead of stopping abruptly - makes several minute oscillations back-and-forth around the stop point which results in strong vibration propagating all over the ship. My questions:

Is it normal for hydraulic cylinder to vibrate when stopped?
What can cause such vibrations?
How to get rid of vibrations?

The directional valve used in the system - ATOS DPHI 3714 /FC 53 – two stage solenoid valve. Following some recommendations I installed between the stages Flow Restrictor Valve and adjusting controls in the Flow Restrictor Valve I can eliminate the vibration but this adjustment is so sensitive that I can not make it work continuously (I believe the adjustment is dependent on oil temperature).

Other system parameters: flow ~ 2 x 50 gallons per minute, working pressure 250 – 1500 psi depending on
Rudder load.

[Doug Hanson]

What do you have for a control system? This valve is just a discrete on/off three position valve. Is your control system open loop, or did you try to include a feed back arrangement. If it is the case that you need a closed loop feedback system, you need a proportional or servo valve and someone to properly design and commision it. If it is just an open loop control with a rudder right, rudder left switch, check your linkages for excessive play. If the volume of oil in the hydraulic lines between your control valve and your cylinders are greater than the oil volume of your cylinders, you will need to bleed the air out of them. Also if you can't stroke the cylinders completely from stop to stop you may have a hard time getting the air out. Are your cylinders new or rebuilt?

[Doug Hanson]

How are you piloting and draining this two stage valve? External drain internal pilot, or internal pilot with external drain and back pressure valve?

[Doug Hanson]

Your could try getting the rudder moving, then pull both solenoid connectors off, do the cylinders stop solid, or shudder. if they stop solid then you have an electrical issue.

[AndreyG]

Hi Doug,

Thank you for your comments - you seem to know the matters.

Answering your questions:

>> What do you have for a control system?

I have simple control loop and can operate valves manually - the vibrations are the same.

>> Check your linkages ...
Linkages are OK, valve driving voltage is OK. All Electrical is OK (this is my specialty)

>> If the volume of oil in the hydraulic lines between your control valve and your cylinders are greater than the oil volume of your cylinders ...

Volume of oil in the pipes is less then oil volume in the cylinders, however there is substantial length of piping. Is it bad?

>> ... if you can't stroke the cylinders completely from stop to stop ...

I can't stroke the cylinders completely from stop to stop, there are mechanical limits to Rudder motion.

>> How are you piloting and draining this two stage valve? External drain internal pilot, or internal pilot with external drain and back pressure valve?

This is tough question - I do not have complete schematic. The directional valve is mounted on manifold and thin pipe goes from manifold back directly into the tank. This is in addition to main return path which goes into the tank through check valve and filter.

What is practical way to bleed the cylinders? The system was operated for long time (weeks) and does not improve.

Could the air get into the piping somewhere through the flanges if there is no oil leaks? If so how to test for that?
(Unlikely because there are two identical pumpsets and both behave the same way.)

[Doug Hanson]

Yes long line lenths are bad. What is the bore diameter, rod diameter, stroke length, diameter of tubing and length of tubing from the cylinder to the control valve?

It sounds like your control valve is set up for internal pilot, external drain with the check valve in the main return line providing the pilot source, double check the spring setting in this check valve, should be at least 85 psi, preferably higher.

Is this a new system, or retrofit? Has it ever worked properly?

[AndreyG]

Doug, you wrote:

>> this check valve, should be at least 85 psi, preferably higher.

what happens if I remove the check valve entirely?

Observation: Oil Pressure Gauge shows 250 psi when the cylinder is engaged. When the valve shifts on the Pressure Gauge oscillates: jumps to 500 psi, then drops to 100 psi and then only settles to 250... maybe even couple of oscillations successively decreasing. Is it how Pressure Gauges normally work? Or this is how pressure changes when valve opens?

[Doug Hanson]

If you take the check valve right out, there will be no pilot pressure available to the pilot valve to shift the main stage of the control valve.
It looks in the picture like you may be able to un-pin the rod eye from the rudder and get the cylinder to stroke fully.
The reason I requested all of the dimensions earlier is to come up with a natural frequency of the system, I would be making an educated guess as to the mass however.
The dissolved air within oil, which may be as high as 15 to 20 percent by volume lowers the bulk modulus of the oil (the inverse of compressability), enhancing the spring effect. Dissolved air (low bulk modulus), large and long tubing lengths and large cylinder volumes create a low natural frequency, not a very rigid system. Entrained air is even worse.
So even taking the control system out of the picture, when the valve shifts to block the cylinders, the inertia of the cylinders and rudder, along with external forces on the rudder will cause the system to resonate at it's natural freqency until it decades to a stop, unless external forces on the rudder keep changing in which case it will continue to resonate.
Anyone feel free to jump in if I have the terminalogy or description wrong, my mind is stuck in logger mode at the moment.

[AndreyG]

Found the Dimensions:

The cylinder is 10" bore 5" rod and 42" stroke
Diameter of Tubing: 3" I guess
Length of tubing
There is a network of tubing actually.
Please see attached picture
The length of the pipes - 10 - 12 ft

>> If you take the check valve right out, there will be no pilot pressure available to the pilot
Interesting. When I do that, idle pressure drops from 200psi to 100 psi and the system still work.
When I take check valve out I connect output of the manifold to filter directly and pressure on the filter
(there is separate gauge) is 7- 10 psi only.

[AndreyG]

Will increasing Check Valve Cracking pressure
(from 65psi I currently have to 100, 150 psi)
eliminate my oscillations?