Hydraulic servo valves--where to start? 1

[Tenaja]

I want to convert a pneumatic cylinder to hydraulic, for more positive control. I need to tightly control the cylinder pressure and speed, which may vary at different points on the stroke depending on the needs at the time.

It would be simple with a motor and screw, had it been designed that way in the beginning, but there is not room for a motor. A remote hydraulic pump is easy.

The system is controlled with a PLC. I'll be installing a position feedback system. I know pneumatics, but where do I start with hydraulic servo (proportional?) valves, and how does it work with PLC controlled pressure? Is the regulator separate? Who is a good supplier that can set me up? (We are talking 1", maybe 1.5" bore with small stroke.)

Thanks in advance!

 

[PNachtwey]

So what was wrong with the motor and screw?
What if you can control the speed and force of a pneumatic cylinder? It is possible you know but not with a PLC.
Are you really controlling pressure or applied force? You control pressure of a volume but across the piston you control force by using a load cell ( accurate ) or differential force by using Pa*Aa-Pb*Ab=netforce ( not so accurate due to unknown friction ).

PLCs are not be fast enough to control the pressure/force well.
How much mass? How far. How much time? How accurate? etc.
What are you really trying to do because odds are someone has done it before.


Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/

 

[Tenaja]

Unfortunately, this machine preceded me, and as I said, it was originally designed with a pneumatic cylinder, so there is no room for a motor that can apply 1,000 lbs of force. OTOH, a hydraulic system will be a relative a drop-in fit. At about 1,000 lbs of force, with 6" per second travel, that requires more motor than can fit physically.

The control I have is sufficient for an industrial servo motor, so I'm confident it should be sufficient for a hydraulic system with position feedback, unless hydraulic valves require faster timing that I'm not aware of.

I do not need the forces so accurate as to worry about unknown frictions. Known hydraulic pressure is close enough. I just need predictability and repeatability.

The current air cylinder engages an over-center cam at around half travel, then goes on to generate another movement. What I want is to be able to move up to the cam at high speed low pressure (to ensure a part is not jammed), lock the cam (higher pressure), then finish the move with a selectable pressure.

What is new to me are the valves and regulators required for hydraulic control. I don't know exactly what valves and regulators I need. I also don't know what signal they take (PWM power signal? 0-10v DC?)

With a 1.5" cylinder and a 6" stroke, and a 1-second stroke target, I need a realistic 2 gal/minute, but I don't know how pumps are rated. Is the flow rated at no load, or at max pressure, or some random place in between? I've looked at a number of pumps, and none have said. Also, what kind of over-spec do I need?

And, and I better off using a 2" cylinder at 300 psi & 4 gpm, or a 1" cylinder at 1200 psi and 1gpm? See, I know how to spec a pneumatic or motorized system, but not hydraulic. I just want to call a shop and tell them what I need, and have them quote me the needed valves and tell me what signals they require.

 

[BUGGAR]

The hydraulic power steering system of cars comes to mind. Not because they are relevant to your need, but because I learned a lot about hydraulic valves and controls from them (I'm not hydraulic). They are 100% duty cycle, silent and cheap. And, of course, my mention of car parts gets me a ride. Drop in a hydraulic power steering rack and hook up your previous pneumatics to the steering wheel?

 

[PNachtwey]

At about 1,000 lbs of force, with 6" per second travel,

That isn't much but it is a lot to ask of a pneumatic system.
Just about any small servo quality proportional valve will do.
What are you using for feed back? Usually hydraulic cylinders have the piston and rod drilled so a MDT rod can be stuck through the back of the cylinder however it is possible to mount external position feed back devices.

I do not need the forces so accurate as to worry about unknown frictions. Known hydraulic pressure is close enough. I just need predictability and repeatability.

You should use a load cell or a pressure sensor on both ends of the cylinder because the pressure on the cap side alone doesn't take into account the back pressure and force on the rod side. I have seen too many make this mistake.

The current air cylinder engages an over-center cam at around half travel, then goes on to generate another movement. What I want is to be able to move up to the cam at high speed low pressure (to ensure a part is not jammed), lock the cam (higher pressure), then finish the move with a selectable pressure.

This seems easy enough but again I think you should be thinking in terms of force not pressure.

What is new to me are the valves and regulators required for hydraulic control. I don't know exactly what valves and regulators I need. I also don't know what signal they take (PWM power signal? 0-10v DC?)

True servo valves use current like +/- 20ma to move a torque motor that controls hydraulic flow to move the spool. Servo solenoid valves should have an on board amplifier so that +/- 10 volts gets amplified internally to move a solenoid and spool. You want a servo valve ( expensive ) or servo solenoid valve with on board electronics ( OBE ). The key is to get a 0 overlap valve.

Servo quality valves are rated at 70 bar or about 1000 psi or half that per port so if your operating pressure is less than that you will get less than the rated flow. The flow is proportional to the square root of the pressure drop so if you use 500 PSI supply pressure then the flow will be about 70% of the rated flow.

Pumps are rated with a GPM, PSI and HP rating. You don't need a big pump.
HP = PSI*GPM/(1714*eff) efficiency is about 0.95. If the duty cycle is low a smaller pump can charge an accumulator during the dwell time.

Your local hydraulic shop should be able to do the calculations for you.
I would go with the 2 inch diameter cylinder. You have to remember there will be a pressure drop across the valve when pushing the cam so not all the supply pressure will be available.
A key equation for computing maximum speeds as a function of loads is the VCCM equation.
I have a good thread on the VCCM equation here with an example problem provided by a forum member.

http://forum.deltamotion.com/viewtopic.php?f=18&t=497

Notice that the VCCM equation uses a load force as one of the parameters not pressure.








Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/

 

[Tenaja]

Thanks for the notes, Peter. I've dealt with +/-10v and 0-20mA signals before, so that's easy. The plan was to install an external absolute encoder--probably something simple like a Gefran, because I don't need 1 mil accuracy. I'd welcome other thoughts.

The air cylinder has a 3.15" (60mm) piston, so half a ton is not unreasonable with shop air--it only takes 125 psi. But there is a delay while it fills, and that's one of the major things I'm trying to eliminate.

When you say "You want a servo valve ( expensive )", what is "expensive"? Industrial sites rarely list prices, so my only reference is ebay....and I don't know what I'm looking for. And most of the valves with "proportional" in the title are not proportional.

When you say "Just about any small servo quality proportional valve will do", who are the players with decent balance of quality and cost? I've googled, and found Parker. I think Rexroth also has some. And who has them for these small cylinders? For this size cylinder, what size valve would I need?


Why do you recommend the larger, 2" cylinder? The smaller one will require less flow, and one at 1.5" can still get the job done with less than 600 PSI. (So I'm thinking a cheaper pump.) Is lower pressure just easier to deal with with seals and hoses? Don't most pumps put out the higher pressure regardless? Or is that pressure only when there is a load? I'm not questioning your expertise, just trying to understand. With pneumatics, I tend to try to size the cylinder so it uses arund 50-80% the available air pressure. (This old system requires 100%, so the delay is even worse at max pressure.) My ignorant guess would have been that trying to digitally regulate down to 50-300 psi would be harder when 1800 psi is available...but I suppose a hydraulic system doesn't have full pressure until there is a load?

A quick glance shows your site has controls; do they have touch screen displays and ladder programming available?

Thanks again.

 

[ScottyUK]

Tenaja,

You may find this link helpful in explaining the basics of hydraulic servo valves.

http://www.moog.com/literature/ICD/Valves-Introduction.pdf

If cost is a big factor then Moog possibly aren't the company to approach. ;-)

 

[Tenaja]

If Moog is expensive, then who are the affordable choices?

Thanks for the link, btw, it was helpful. Seems like direct drive servo valves are fastest, and proportional and pilot are slowest. Is that about right?

Unfortunately, the place I was referred to by my local supplier only deals with industrial valves, like tractors use.

 

[PNachtwey]

Parker D1F+
BoschRexroth 4WRPHE
These are good valves.
Moog makes good valves too. In your application is is the quality of the spool that will make a big difference. You need a valve zero overlap spool
You need the size 6 versions.
If you use the VCCM equation you can calculate the size of the valve. Then you can shop and compare what each brand has in the size you want.


Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/

 

[Tenaja]

Thanks, Peter!

Are these things really $4k each? That is expensive.

 

[BUGGAR]

Forgive my repetition but off-roaders build steering systems that sound like they do similar to what you need, except for the control input. Off-road durability/100% duty cycle, mass produced costs. Come out to the desert and watch these things.

 

[Tenaja]

I have no doubt a hydraulic steering system will be plenty strong. I would be surprised if they have a suitably fast response time or sufficient precision though, because "real time" in a car controlled by human hands is vastly different than in a manufacturing machine controlled by a computer. Also the definitions of "a little play" are very different for the two applications.

If it was for my hobby, I'd probably be all over it. Since it is not, I don't have time to dink around with something that "could" work. Thanks, though.

 

[PNachtwey]

Are these things really $4k each? That is expensive.

Yes! The tolerances are very fine. Yes, you could probably buy a whole motor system for that but you don't have the room.
Pneumatics will not provide very good control unless you add a controller that costs as much as the valve.

Why do you recommend the larger, 2" cylinder?

It depends on what you are trying to do. The larger the diameter of the cylinder the stiffer and easier to control the open loop system will be. you will get relatively poor performance from a standard controller because any errors will decay as a function of exp(-zeta*omega*T)
where zeta is the damping factor and omega the natural frequency of the cylinder and load. There is NOTHING the one can due to change the gains to make the response faster. It depends solely on the hydraulic and mechanical design. Making the diameter bigger increases the natural frequency and that allows errors to decay faster.

If you need speed and cheap then a smaller diameter will do. If you need consistent precision then get a larger diameter cylinder.
There are formulas for calculating the natural frequency and the cylinder speed. Everything can be calculated, simulated and controlled but the quality of control depends on the hydraulic-mechanical design.




Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/

 

[Tenaja]

Is it reasonable to have electronic control over the hydraulic pressures as low as 15 psi? That's pretty close to what it would take to be able to get the 2" cylinder down to 40lbs max force.

Thanks again.

 

[PNachtwey]

Is it reasonable to have electronic control over the hydraulic pressures as low as 15 psi?

You keep thinking in terms of pressure instead of force.
Pressure is used to test vessels, hydro forming, and some injection molding machines but most applications require controlling force. There is pressure on both sides of the piston. The force on both sides of the pressure much be subtracted to get a net force. There is also friction.
At low forces the frictional force is significant. It is usually much greater than 15 lbs. If you need net forces that low you must use load cells.

That's pretty close to what it would take to be able to get the 2" cylinder down to 40lbs max force.

NOT SO!!! THINK ABOUT IT. When the actuator is stopped or at constant velocity the net force on the piston is 0!!!!! You forget there is force on the both sides of the piston. You could have 500 psi on the cap side and 1000 psi on the rod side if the ratio of the areas is 2 to 1.




Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/

 

[Tenaja]

I have a different early 90's machine (probably designed in the 70's) where it pressurizes both sides to balance and control pressures. I guess I presumed that since valving and regulation technology has improved so much that it was an "old fashioned" way of doing things. (Experience with this machine is why I figured the other one would be a good candidate for hydraulic upgrade.)

With a zero overlap spool, does the pressure not relieve once you open it to vent? How long does it take? Sure, I understand there is back-pressure while forcing the fluid out (faster = higher pressure), but is it significant the full stroke? It's not like the two ports are independently controlled; they are either filling or venting, right? (What is venting or exhausting called in hydraulics?)

I don't need 40 lbs of force right out the gate, but I do need it close just before the "half-travel" cam position is reached.

I guess it's a good thing I'm asking questions. You've given me more helpful info than the salesman has. Thank you.

 

[Compositepro]

Pressure difference between the two side of the piston are caused by external loads on the cylinder, not by a controller or valve (unless the piston has hit the ends of its travel). A zero overlap spool will leak a very small flow when centered, but this flow will be equal on both sides. This gives very exact control cylinder position, with no deadband. Both sides of the piston will always be pressurized. This is what rigidly holds the position of the cylinder. This can also be done with pneumatic cylinders. Hydraulics are just more powerful because of higher pressure, and more rigid because of the lower compressibility of oil.

 

[PNachtwey]

With a zero overlap spool, does the pressure not relieve once you open it to vent?

NO! See page 24/25 of this

http://deltamotion.com/peter/Mathcad/Mathcad - Hydraulic Cylinder.pdf

Look at how the pressures change on both sides of the cylinder.
Notice that the pressures are different extending and retracting.
Read the link about the VCCM equation on my forum with the link in a previous post. There are relatively easy formulas for calculating the pressure on the A and B side of the piston as a function of velocity and direction.
When the piston starts to extend the motion compresses oil on the rod side and pushes the oil out through the valve. This means there is always an opposing force except for the case of a few injection molding valves.

Just about everything can be calculated.

How long does it take?

Usually there is always some pressure/force on the opposing side with few exceptions. Your system will not be one of them.

I understand there is back-pressure while forcing the fluid out (faster = higher pressure), but is it significant the full stroke?

Yes, the faster the motion the more force/pressure it take to force out the oil.

I don't need 40 lbs of force right out the gate, but I do need it close just before the "half-travel" cam position is reached.

You need force to accelerate the load, force to over come friction and force to push the oil out the other side of the cylinder.
That can easily be more than 40 lbs.

I guess it's a good thing I'm asking questions. You've given me more helpful info than the salesman has. Thank you.

Salesmen don't know .... If they did they would be engineers. Even then most engineers can't do solve differential equations or even know how to write a differential equation for their systems. The number of people that can correctly design a hydraulic servo system is small.

Too many of my customers ignore putting pressure sensors on both the A and B sides of the cylinder. They wonder why their quality control is awful. If you are using pressure sensors you need to use 2. The other option, and better one, is a load cell to measure force but requires a nice environment. When I show them the right way their systems run better.








Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/