pneumatic circuitry sanity check

[subsearobot]

Howdy folks.
I have a fairly simple requirement. I need to test a linear actuator (stepper with a lead screw) against a fairly constant load in both directions (~20 lbs, but possibly *also* 200lbs; trying to assess the needs of my team).

Pneumatic were suggested by the PM, sounds overly complicated to me, but I'd like to play... I've done a fair amount of hydraulic design, and I'd like to get some experience with air too.

can this be done with pnumatics? see the sketch.

thanks!

 

[subsearobot]

more details:
linear actuator speed:
~1 in/s

will also be using precision position feedback, and likely a load cell.

the test is to assess the motor(s) for use in a precision positioning application, to better than .001" repeatability, against ~20# of seal friction.

 

[HPost]

If you want precision, use a hydraulic system.

Pneumatic system will give the basic kinematics, but the system "as drawn" will not be controllable to level you need.

Also, the directional control valve you have drawn will be very rough with either air or oil. If you want really precise control, you'll need a proportional control valve.

If the system will be highly transient, the gauge in the position will not be much good at all.

Cheers

HPost CEng MIMechE

 

[subsearobot]

thanks HPost, I am only trying to load the motor (linear actuator) with the piston, not control position (the motor will "back-drive" the piston). Essentially, we are simulating loading the motor with a simple dead-weight system, but it would be bidirectional.

it will not be very transient.

that being said, I do want a fairly constant, repeatable load through the full travel. travel will be about an inch.

 

[Compositepro]

Dead weigh techniques are used to calibrate force instruments. Use a weight and pulley. You also have to be aware of inertia effects with any technique that you use.

 

[imcjoek]

Pneumatics like you have drawn might give you an adjustable load in a static condition. But once you start back-driving the air cylinder, the ability of the regulator to exhaust will cause transients in the pressure.

You could use a closed loop E/P transducer to pressurize the cylinder.
For 20 lbs you could use a glass body cylinder so that it's own seal friction doesn't skew your results.

You have to define the resolution/repeatability you need from your variable load to say if any of this will work.

All that said: Weights and pulleys sound a lot simpler...

 

[subsearobot]

imcjoek, i agree. thx.
the PM is an very focused on the control side (read: an EE), and I like to play with valves and pistons.
but this application sounds like a complicated way to get worse results for more money. I am not on the state payroll...

cheers