Hello, i'm working on a simple "exoskeleton" where i would help the moving of a human elbow (i will attach a picture for an example (EduExo)).
For the start my project would just work in a simple way: the upper arm would be in a vertical stationary position and the forearm would move according to this at its limitations (i will attach a second picture). A structure would hold the forearm and the motor would rotate this structure at the elbow. The motor would be somewhere on the upper arm. The used forearm's weight and other dimensions are known.

I had a hard time to decide the type of control for this project. In the first testing round i tried a dc motor with gearbox (with wormgear). The motor was placed at the elbow. A potentiometer was placed on the motor's shaft, so i can measure the rotation, because the position of the forearm (referenced to the ground) affects the needed torque that the motor has to put up. A loadcell was built in the forearm holding structure, so i could determine load limits (up and down directions) where the motor has to rotate with a determined torque in one of the directions.
This didn't seem like to work in real life, because at moderately "higher" speeds the angular momentum affects the measured loads in a bad way, also at rotation the motor also puts a "load" on the other end of the loadcell which ruins the measure and also the loadcell "wears out" pretty quickly in this situation (the measured load values flows away in one of the directions compared to the real values, thus creating an untrustworthy measuring device) .

One of my friends had an other idea. I have to use a strong dc motor with gearbox (not with wormgear). The situation is almost the same (stationary vertical upper arm, a potentiometer or other device will be used to measure the position of the forearm, the forearm's weight is known). But you can regulate the intensity wherewith the motor helps the rotation with a device (for example a linear potentiometer or something other) placed in your hand.

So for example: if the forearm is 90° to the upper arm (forearm is parallel with the ground) let's say Forearm Torque will be 5Nm torque at the elbow to keep the forearm in this position (this torque will be dependent on the forearm's and the upper arm's angle, but at 90° it will be the largest). We need to help the human to move this forearm, so with the device in the hand we could adjust the value of X. X will be between 0...1 and X times Forearm Torque will be the torque applied by the motor. SO if X = 0.8, then the person needs to exert just 0.2 torque to keep its forearm up compared to normal conditions (without exoskeleton). If X=0, then the person has to do all the work and if X=0,999 then the exoskeleton does almost all the work.
The motor maybe would help in just lifting up the arm (not in down direction), if the person wants to move its forearm down, the motor maybe would shut down, and the person would have to work against the gearbox (it is without wormgear, so the person could do it) with also the help of the gravitation.

Sorry for the long building up. My question is would this usage "hurt" the motor? In my interpretation this would be a kind of constant stalling on the motor. It could not rotate as much as it would want in most of the cases (for example if the person just wants to keep its arm in a stationary position with the help of the exoskeleton) and it could burn down in a long run (if i'm not wrong).
Is there any kind of dc motor that can stall better? Is ther any control method that could help the motor? I dont know yet if a brushed or brushless dc motor could be the better option for this situation.
I don't want to use a servo or a stepper motor for this project, because in my interpretation i have to work on the torque regulation and i can achieve it with just dc motor (servos and steppers are for position or speed regulation, correct me if i'm wrong).

I was thinking about current limitation, so if the motor has to stall for a long time (maybe 10 minutes) then it wouldn't peak through that dangerous limit and wouldn't kill itself with heating up in a long run. But i don't want to put up limitations, because it is already hard for me to find an usable strong dc motor with gearing that doesn't cost too much money (it is a diy project).

I would gladly love to hear your answers or your any other tips and corrections.

 

You can use a servo or stepper motor in a non feedback positioning mode, DC brushed or BLDC can be used.
The DC servo type motor has some superior characteristics over the run of the mill type usually, high pole count and skewed rotor lamination's etc.
Most high end motor manuf. post a couple of torque characteristics, one is the continuous stall torque, this is the degree of current it can take in a continuous mode, the other is the momentary peak torque where the motor can only be operated for a very short period of time before damage occurs.
So in effect, taking advantage of these characteristics would entail a controller capable of monitoring the motor current.
Simple controllers such as made by KB etc, have settings for current limit which does not allow the drive to go over a limit you can set by pot etc.
Just some basics.
Max.

 

It would be nice if a 45,000:1 gearhead motor could be found, or 3,000 RPM to 90 deg. movement in 1 sec. A very non scientific
test. Using a small Power Electric Distribution motor No. 21267. , 80 RPM @ 6 V, 110 mA, I attached a vice grip type plier, 12 oz. 7.5 in long,
attached 4.5 V & noted vertical PRM close to 60. With no power plier would gradually fall from horizontal to about 30 deg. At 500 mA position would be held. No heat rise noted. The Electronic Gold mine also has a 6-24 V , about 80:1 reduction with dual ball bearings.

 

To lower forearm from vertical with only gravity will be impossible as there is no torque available; from horizontal with any decent gear reduction again impossible with out reversing the motor. I was looking at a 270:1 reduction motor, The Surplus Center, Item No- 5-1796, 8 in. lb., 25 RPM,24 V, 60 mA. & I believe that you were asking for about 5 X this torque?

 

A stalled motor is a dead short. You have to be careful of that.