Accurat static pull

[morzh]

I am trying to see if I could accurately apply a static pull to a cord to create tension. I should not be measuring the tension directly, only be sure that the pull I apply is the force in Newtons that I want. The range is from a few tens of Grams to 7 kG. Either in one range or in several sub-ranges (say, 80G - 500G, 500G - 2kG, 2kG - 7kG) or such. With say, accuracy of 1% of the value.
I do not specialize in drives so I do not know how accurate a, say, BLDC motor torque could be when applying current, and what range I would get.
Any ideas? Any that have been tried?

 

[alec_t]

Welcome to ETO!
In view of commutation and pole shape effects I don't think a BLDC motor would be a good torque source for a static tension with a 1% accuracy. You might do better with a controlled current source driving a solenoid, and have a load cell to provide feedback for the control. Even then, you're asking a lot for such a wide range of tension.

 

[morzh]

My whole point is designing the load cell out. No measurement.
I already have the system that works and sells, using steppers and a Load Cell.

 

[JimB]

My whole point is designing the load cell out. No measurement.

How can you know the tension without measuring it?
Varying elasticity of the thing which you are tensioning, varying length of components due to thermal expansion.

JimB

 

[dougy83]

Stretching a spring a certain distance will create a certain tension. If the spring is linear, F = kx, where F is the force, x is the displacement and k is the spring constant.

Some form of eddy-current clutch could also provide what you want. Even just spinning a ring of magnets near an aluminium plate will create torque.

All of these methods will require characterisation/calibration over full range of loads and temperature. So you'll have to use a strain gauge at the factory for each device, rather than including the gauge in the end product.

Do you have a picture/diagram of the device you're selling?

 

[shortbus=]

Aren't you guys talking about a dynamometer? https://www.explainthatstuff.com/how-dynamometers-work.html

 

[morzh]

How can you know the tension without measuring it?
Varying elasticity of the thing which you are tensioning, varying length of components due to thermal expansion.

JimB

Jim

There are two way to apply a known amount of force:

1. Know what you are applying. Here - use a drive, and calculate the known effort by the motor constant and the amount of current which can be pre-determined. Same as if I hook a spring with a known K and pull it a known distance. I do not then have to measure.
2. Apply an unknown force and then measure it with a, well, dynamometer. Or a Load cell.

 

[morzh]

Stretching a spring a certain distance will create a certain tension. If the spring is linear, F = kx, where F is the force, x is the displacement and k is the spring constant.

Some form of eddy-current clutch could also provide what you want. Even just spinning a ring of magnets near an aluminium plate will create torque.

All of these methods will require characterisation/calibration over full range of loads and temperature. So you'll have to use a strain gauge at the factory for each device, rather than including the gauge in the end product.

Do you have a picture/diagram of the device you're selling?

Dougy

I thought of Eddy clutch. It has to be calibrated clutch. I do not know if they exist; what I've seen while searching is 1. Large, 2. Uncalibrated, meaning that after having applied the force I still have to measure it.

What I am trying to do is simply this:

A string. One end is getting pulled. The force is to be known or measured. Roughly 4 inches between the two holders, one of which is movable. THe range of movement is about 1/2 inch or so. But the actual movement is small as the string is glass. Currently there is a load cell attached to the immovable holder and it measures the effort. The effort can range between say 80g to 7kg (0.8N to 70N). The measurement accuracy is pretty good - under 2g.
I want the Load Cell out.

 

[dougy83]

Ok. Because of the ½" travel, you need some way of detecting when the cable is taut.

If heading down the spring with linear actuator route, a second weak spring with end-stop inline with the main spring allows the start position to be found by moving the actuator and detecting when the end-stop engages. The weak spring at this known position will give a known initial tension, <80g.

End-stop engagement can be detected via simple electrical continuity. Likewise, a broken cable can be detected if the end-stop disengages.

 

[alec_t]

Hang bucket on string. Add known volume of water to bucket.