Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Control of piston oscillation

  1. Nov 27, 2012 #1
    Hello everyone!!
    We have built the following model with Automation Studio. It's a simple electropneumatic system with a 4/3 valve and a piston. I would like to eliminate the oscillation of the piston with a PID controller. Does anyone know how to do it?
    Thanks
     

    Attached Files:

  2. jcsd
  3. Nov 27, 2012 #2

    Integral

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    What oscillation? You will need to provide more details before anyone can help you.
     
  4. Nov 27, 2012 #3
    The double acting cylinder has two air cells.
    Firstly we put voltage on the left solenoid of the valve. The air flows through the valve and goes into the left cell of the cylinder. Then the air pushes the stroke of the cylinder, so the stroke outspreads.
    When there is no voltage, the valve doesn't allow the air to flow into the cell, so the stroke stays outspread. It does so, because the air pressure in the left cell is equal to the pressure in the right cell. Naturally until the pressure difference is zero, the stroke does an oscillation. I would like that oscillation to last as little time as possilbe.
    Thanks
     
    Last edited: Nov 27, 2012
  5. Nov 27, 2012 #4

    dlgoff

    User Avatar
    Science Advisor
    Gold Member

    Do you have any idea as to what the oscillation frequencies are?

    FYI

    http://en.wikipedia.org/wiki/PID_controller
     
  6. Nov 27, 2012 #5

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    is this a real piece of equipment or a computer simulation?

    If a simulation, how much friction do you have in it?

    I would be surprised to see a real piston oscillate, unless velocities are quite high.
    If it's a real piece of equipment how fast do you need to move the piston?
    Adjusting air pressure to just what's needed will probably help. As will limitig rate of air inflow perhaps by longer tubes or an orifice.

    If you want to add a linear control element you'll have to decide what you're going to control - position, velocity or force, and write some equations of motion for the system.
     
  7. Nov 27, 2012 #6
    Never mind about the precise values of the three terms. I can't understand which quantity I will control exactly. What will I have as an input? The voltage on the solinoid of the valve? The way I see it is this: The voltage on the solinoid is 230 V for example, which means that the valve allows a certain amount of air to flow. If the voltage is 0 V, no air will flow. Now if the voltage is 150 V for example, what will happen? The valve will move more slowly, so that less air will flow? That means I can control the oscillation of the stroke by changing the value of the voltage. I may be completely wrong, though.

    Does anyone have any idea?


    jim hardy thanks! Do you know if I can do either of the things you suggested with a PID controller? I mean if I could adjust air pressure to just what's needed, or limit the rate of air inflow.
     
    Last edited: Nov 27, 2012
  8. Dec 2, 2012 #7

    Integral

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Control solenoids are on or off. There are no intermediate states.

    I have worked a LOT with pneumatics, if you limit the rate of air flow into the cylinder it can be jerky. To control the speed of extension you need to limit the exit speed of the air from the back side of the cylinder. They make needle/check valve combinations which allow free flow on the input stroke and throttle the air through the needle valve in the other direction.

    If you limit the inflow rate the pressure will build until system friction is overcome, it will then move until the internal pressure drops below the friction forces. It will then hesitate until the input pressure builds till it again overcomes friction. This is why you do not want to limit the input rate.

    What you are describing could be incorrect plumping, it should not occur in a normal operation.
     
  9. Dec 3, 2012 #8
    Integral, thanks a lot for your answer. In the meantime I found this, which is about a hydraulic system. Except for the fact that practically water can't be compressed, the circuit has the same function as the pneumatic one. I believe the PID controls the flow through the transducer. Is that possible? And how does it do it?
     

    Attached Files:

  10. Dec 3, 2012 #9

    Integral

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I have encounterd hydrolic pistons which used DI water for the driving fluid.

    You can get digital flow control, but it won't be cheap. Not sure why you would want to. Just put normal flow controls on the cylinder and you should be good.
     
  11. Dec 4, 2012 #10

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    A PID is a linear control element, meaning it's continuous not discrete. It usually employs a "transducer" to control flow or pressure.

    A transducer might be an electromagnet that's in a tug-of-war against a bellows with pressure inside, so that current through the electromagnet controls pressure in the bellows. Search on "force balance transducers" and see what tutorials you can find from control companies like Rosemount, Fisher, Conoflow, Foxboro etc.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Control of piston oscillation
Loading...