^^ Edit: 1. I must be able to control the *stroke length*.

okay, sorry I posted this thread a little prematurely.  I believe I have solved this problem.  I will just slow the piston down with a spring (duh! I tried to make it more complicated than it had to be).

Still not quite sure how to make the stroke length continuously variable. Are there computer controlled pressure control valves?  (Sorry, I know next to nothing about hydraulics)

If you want to vary your stroke length to follow desired pressures, you are best buying a pressure transducer (or a number of pressure switches?) to feedback to your control system. You could then use the signal to de-energise the control valve to neutral (assuming you have a neutral), stopping the cylinder motion... I would also install a solenoid operated relief unloading valve in the circuit so that you can unload the pump on the desired signal.

Hope that makes sense.

I have doubts that you can achieve 40Hz frequency response with a hydraulic piston. This means that a complete cycle is 25 msec. The control valves ON-OFF response time should be very fast, just few msec (the smaller the valve response time the shorter the time delay). To my opinion you can hope to achieve this only if the piston stroke at 40Hz is close to zero distance. This is all depends on the oil pressure and the piston diameter too.

"I have doubts that you can achieve 40Hz frequency response with a hydraulic piston."
Agreed,  first do the physics so you know the velocities, accelerations and forces required BEFORE you make this a hydraulic problem.

Peter,
This is already a PROBLEM?

A spring in a cylinder?
What does pressure have to do with position? (pressure transducers)
A 2-way 2-position valve takes about 25-30ms to shift, not a few ms?
No mention of stroke, bore or rod diameter?

You always blame the hydraulic guys for bad circuits, look at who are making the posts?

Westerndynamics.com

Gkranz, what can I say, but at least 20vtmk1 is asking for hydraulic advice.  I think he needs help with the mechanics and physics too.

I have fatigue testing customers that measure the sinewaves with amplitudes measured in microns.  They just stretch metal so the stretched metal acts as the spring.  20Hz is a challenge even with a well desgined system.  It is the same old problem.  How does one get energy into and out of the cylinder quickly.  20vtmk1,hint, start with a high quility servo valve.  It will be expensive but if you don't have that kind of money you will not be able to do the project.
This will be expensive.

Thanks for all the GOOD advice.  I am simply trying to consider all aspects involved in this project.  Since I am new to hydraulics I do not know they limitations, and am having a hard time finding resources applying to the physics of hydraulics.

I have found at least one source for high speed valves with response times of 200 microseconds (not allowing for stroke distance).

Anyways, thanks a lot.

Oh yeah, sorry.  

Bore: 5.8 mm
Max Stroke: 10.5 mm

OK 20vtmk1, just what are you trying to do?

If this is for fatigue testing application, then you will need a high speed servo valve, a plc controller for motion profile, a cylinder that is seal-less and friction-less special made.

Like Peter said, better get some one that has done these applications before, not to mention the hydraulic system needed to support this cycle rate.

Quote - "quite enthusiastic about a project I have going".

Better screw your hat down tight, you picked one of the hardest circuits to design for a beginner.  99% of the hydraulic guys cannot do these circuits right, whatever made you raise your hand and say "I'll do it".

Westerndynamics.com

Gkranz, you are exactly right.  
We are going to be testing valve springs for internal combustion engines.
It is to make sure the springs operate to the rpm and lift at which we rate them.

HAHA! I'm going to have fun with this one then.  It was more like "yeah sure, I'll give it a shot".  I like to try new things, and in this case it is related to a possible business venture.

Reconsider a cam on a shaft.

Now there is a great idea, give it some thought.

Westerndynamics.com

That is a good idea, thanks!  In fact, you've all been really helpful, I appreciate it.  It definately does a lot of good to take a step back and consider all the problems, instead of jumping the gun.

problem solved, and peter and gary agree!! it's a sign of a very good week to come  : )

kcj

No, the problem has just moved.

The acceleration rate that must be achieved is 331 meter/second^2.  This is over 33g!
This is for a 10.5 mm stroke at 40 Hz.

We never were told the mass but it is probably only that of the actuator and spring.

One thing I watch for is the frequency at which the system is supposed to move at.  Every time you double the frequency the acceleration rate is quadrupled and the power required is cubed.   It makes no difference whether one is using an electric or hydraulic actuator.  The dynamics are the same.

What about a variable speed PM motor and a balanced excenter? Not a cam, where you will have a problem keeping the roller from leaving ground on downward movements. Some flywheel may be needed.

Gunnar Englund
www.gke.org