Looking for a low-latency force-based actuator - do they exist?

[some bloke]

TL;DR Summary

I am brainstorming a project and need an actuator which can apply a controlled amount of force, which could change very rapidly. Pneumatics and hydraulics are too slow to respond for what I need.

Without going into too much detail, I have a project which I am considering my options for, and one design would have potential if I can find some sort of force-based linear actuator. I suspect that there might be some sort of magnetic-based system out there which operates on a changing frequency of electromagnetism to apply a variable force, but I don't know what to put in to search for it - If it exists, then it's got a name I don't know!

My problem right now is that pneumatics are too springy for my needs, and hydraulics are potentially too slow to move, so I'm hoping for a middle ground that would allow essentially a spring which can be programmed.

My backup plan might be something along the lines of a hydraulic pump with an actuated release valve (which recycles the fluid) which would essentially give a hydraulic spring, but my concerns are that it would not move fast enough for my needs (I will be researching this next!).

My desired result:
A linear actuator where I can tell it programmatically to push/pull with "X" Newtons of force, and where I can change "X" suddenly to recieve as immediate of a response as possible. The value of "X" will be calculated based on the position of the actuator, so if an external force pushes it to a certain position, then the force "X" will increase and stop the movement, ideally with as little observable ramping as possible (as though it hit a hard stop, or a soft one if the new "X" is not enough to stop the motion).

Does anyone know of an actuator I can research to help with this project?

 

[Baluncore]

A DC servo motor applies a force proportional to the motor current.
Maybe driving a ball screw.
We need some numbers.
What is the mass of the load?
How much time is there to respond?
Is there any hysteresis in the mechanical linkage?

 

[some bloke]

I'm afraid I don't know the exact values just yet, but my experience with DC servo motors was that they were fairly absolute. Would it damage one if the load on the servo was pushing it back against the direction it was pushing?

Load will probably be around 50-300kg (variable during operation)
I want the time as quick as possible, I won't pretend I know enough to quantify it right now. I hope I will know by the end!
I aim to have as little play in the mechanical joints (partly by having as few as possible) as possible so hopefully as close to no hysteresis as possible.

This is a DIY project and I will be learning as I go, so it may be a case of "Well, that broke, so it needs to cope with more load than that!" or "Well, that wasn't fast enough, back to the drawing board!".

 

[Baluncore]

I'm afraid I don't know the exact values just yet, but my experience with DC servo motors was that they were fairly absolute.

I do wonder what you might mean by "fairly absolute" ?

Would it damage one if the load on the servo was pushing it back against the direction it was pushing?

The motor simply applies a force, proportional to current, while the current remains within specification.
The speed of the motor is related to voltage, but for a static force, current, with sufficient voltage to satisfy ohms law, is the only concern.

 

[some bloke]

I was meaning that my experience with servo motors tends towards them driving to an absolute position and then doing their darndest to not let that move!

Chances are this is more to do with how they are used in my workplace rather than their full capabilities.

So if I have a ball-screw run off a DC motor, I can control the force it outputs using the current? And by using a relatively high voltage (based on the motor specs) then it will respond with a fast movement?

Effectifely then I would need a fixed voltage and a variable current in the circuit, with the current controlled by the program, which is controlled by the position.

That's definitely food for thought! Thanks!

 

[Baluncore]

So if I have a ball-screw run off a DC motor, I can control the force it outputs using the current?

Yes. That is true since a ball-screw can be driven by the load. A capstan and a rope, or a sprocket and a chain, could be controlled similarly.

And by using a relatively high voltage (based on the motor specs) then it will respond with a fast movement?

High voltage is not needed until the velocity rises, and the (back EMF) / "internal resistance", opposes and limits current flow.

Current flow would be controlled by modulating an H-bridge motor driver.

 

[jrmichler]

TL;DR Summary: I am brainstorming a project and need an actuator which can apply a controlled amount of force, which could change very rapidly. Pneumatics and hydraulics are too slow to respond for what I need.

A linear actuator where I can tell it programmatically to push/pull with "X" Newtons of force, and where I can change "X" suddenly to recieve as immediate of a response as possible.

I have experience with moving objects at high accelerations using servo motors, pneumatics, and hydraulics. A reasonably complete answer to your question would require me to write a book. But, if you can supply ALL of the following information, I can help:

1) Moving a mass of XXX
2) At a peak acceleration of YYY
3) At a peak velocity of ZZZ
4) For a distance of DDD

OR:
1) Moving a mass of XXX
2) A distance of YYY
3) In ZZZ microseconds / milliseconds / seconds / minutes
The acceleration, velocity, and distance can be in linear or angular units.

Alternatively:
1) Applying a force of XXX
2) Changing at a rate of YYY
3) Over a force range of XX1 to XX2
4) Over a total movement distance of ZZZ
The rate of change of the force is in units of Newtons per second / millisecond / microsecond as appropriate.