Motor selection for high-speed/torque and accurate positioning application

Hi all! First post here; I read through the "How To..." so hopefully I'm not missing any needed info.

Please note that I haven't started programming/building of my project yet; rather, I am in the conceptual design stage and would like to be better informed before the buying begins $$ Also, this is a bit of a lengthy post so to cut to the chase I'm trying to decide between 1) stepper motors and 2) brushless motors w/ encoders for my application (but I come to that question at the end as well).

I am attempting to make a system similar to NASA's ARGOS (Active Response Gravity Offload System). Googling "NASA ARGOS" will bring you here: NASA - JSC Engineering - Active Response Gravity Offload System.

It's essentially a large overhead crane that tracks the position of the payload (in ARGOS's case, a person) and applies a constant tension to the cable attached to their harness perpendicular to the X/Y plane (which is anchored approximately around their CG), reducing their downward acceleration and mimicking a micro-gravity environment.

From the website, ARGOS's capabilities are below:

  • Working volume: 41’ x 24’ x 25’ (LxWxH)
  • Z-axis speeds of 10ft/s w/ 300 lb payload; 4ft/s w/ 750 lb payload
  • X/Y-axis speeds of 10ft/s w/ any weight

I hope to build a system with the following capabilities (differences):

  • Working volume: 12’ x 8’ x 5’ (LxWxH)
  • Z-axis speeds of 10ft/s w/ 2 lb payload
  • X/Y-axis speeds of 10ft/s w/ 2 lb payload

My current concern is with the type of motor to use for the X/Y axis. Knowing the position/velocity/acceleration of the payload, I need to be able to move the X/Y motors to track that position, up to 10ft/s (a hopeful ballpark acceleration is 16 ft/s^2 or about 0.5G).

Attached is a rough layout sketch (Gravity_System_Layout.jpg) with what I'm thinking looking at the system top-down, including color-coded components. My payload will be a small RC car which weighs about 4 lbs. When I say 2 lbs payload above, what I really mean is 2 lbs offloaded; this is independent of the attached payload weight, since the tension in the cable attached to the RC car will ~2 lbs at all times.

Some quick details about the image:

  • Black - system structure/frame
  • Orange - rails for y-axis gantry movement
  • Red - gantry AND rails for x-axis offload rig movement
  • Green - offload rig (containing offload motor not shown, topic for later discussion)

Motor type will depend on a) needed rpm/torque (tradeoff game) and b) positioning accuracy.

a)
x-axis force math
To simplify the math, I'll be making a bunch of shortcuts. The offload weight is 2 lbs, but that is the force acting down through the offload rig. If we assume the offload rig weighs 1.5 lbs, and the coefficient of friction is u=0.3 (assume kinetic), our normal force of 3.5 lbs would give us a frictional force of 3.5[lbs] * 0.3 = 1.05[lbs] in the x direction. So, the x-axis motor would need to meet/exceed the velocity and acceleration requirements moving 1.5 lbs of mass and overcoming ~1 lbs of frictional force.

y-axis force math
With the same frictional constant and a gantry weight 3 lbs, we get the following:
gantry weight + offload rig weight = 3[lbs] + 1.5[lbs] = 4.5[lbs]
normal force = 4.5[lbs] + 2[lbs] = 6.5[lbs]
frictional force = 6.5[lbs] * 0.3 = 1.95[lbs]
So, the y-axis motors (collectively because there are 2) would need to meet/exceed the velocity and acceleration requirements moving 4.5 lbs of mass and overcoming ~2 lbs of frictional force.

b) Ideally, the positioning would have an accuracy of about 1/16" or less, but the ability to control the gantry within about 1/8" might be good enough.

The first thought I has was to use stepper motors because they are easy to control and have high holding torque. A setup I envisioned included the following components:

Nema 17 Bipolar Stepper

  • 42mm x 48 mm (Width x Depth)
  • 83.6 ozin (5.2 lbin) holding torque
  • 200 step (1.8 deg)
  • 2.0 A rated current
  • (datasheet attached)

OR Nema 23 Bipolar Stepper

  • 57mm x 56 mm (Width x Depth)
  • 178.5 ozin (11.2 lbin, ~1.3 Nm) holding torque
  • 200 step (1.8 deg)
  • 2.8 A rated current
  • (datasheet attached)

Arduino Uno CNC Shield V3 w/ A4988 Drivers

  • 12-36 V board operating voltage
  • Driver step modes: full, 1/2, 1/4, 1/8, 1/16
  • 2 A driver rated current
  • 35 V driver rated supply voltage
  • (Driver datasheet attached)
  • (CNC Shield image attached)
  • I apologize, I can't find anymore specific datasheets for the board

The thought was to use the NEMA 17 for the x-axis and two NEMA 23's for the y-axis, using 6mm wide GT2 timing belts and synchronous wheels. The main problem I'm facing is that stepper motors don't have a very high rpm limit compared to brushless motors, so their synchronous wheels would need to be pretty big to come close to the 10 ft/s velocity needed. To give an example, attached is a torque curve of NEMA 23, similar to the one I've described (the product number is: 23H2065-300-4A, NEMA 23 Rotary — DINGS' Motion USA). The NEMA 23 in the torque vs rpm plot was run at the following values:

  • 1.7 N*m holding torque
  • 3.0 A
  • 1.8 deg step angle

And the motor I have shown above has 1.3 N*m holding torque and would only be able to run at 2 A.

These steppers and the CNC Shield are an attractive option because they are super easy to interface with the Arduino Uno Rev 3, but do you think I'm going down a rabbit hole with trying to make stepper motors work to fit my needs? Are these components just not powerful enough? And should I look into brushless (w/ encoder and FOC controls) or other options?

I'd like to keep the total cost of motors and related components under $250 if possible which is why I was pushing for stepper motors. Maybe I should go for a CNC board/drivers rated for more amps, more powerful stepper (even larger NEMA 23) that is geared up and lower my requirement standards? (yes I'd like to meet the current ones but can put into easier reach if necessary)

Again, sorry for the long post. I didn't want to leave my project under-constrained for the reader's understanding!

NEMA_23_similar_Torque_curve.png

NEMA_23_similar_Torque_curve.png

NEMA_23_Datasheet.pdf (758 KB)

A4988_Datasheet.pdf (1.04 MB)

I ran out of attachment space in the original post.

NEMA_17_Datasheet.pdf (762 KB)

Based ONLY on the title of your thread, you will need an industrial sized servo and control unit and associated power supply. How about telling the one most important question to begin with?
Paul

Do you know the mechanism called "CoreXY"?

https://corexy.com/index.html

By skillfully using two belts and two motors, it provides the same movement as a cartesian machine.

As a merit, you can use a powerful one because the motor is completely fixed.
Furthermore, since the motor is not attached to the gantry, it is possible to reduce the weight and inertia.
That is, it is suitable for high-speed operation.

I think you need to consider if it's worth applying to this project.

Thank you both for your responses!

Paul -

I attempted to give a cut-to-the-chase question for the reader in the second paragraph to present my overarching question, but wasn't sure how to ask properly without giving more insight to it's application. Are you thinking of a large brushless servo motor similar to what FANUC sells? (datasheet and image attached)

I had done some research into large servos as an option (even saw a closed-loop NEMA 23 that claimed to be a hybrid stepper-servo) but was turned off pretty fast because of the cost. If I could make steppers or smaller/cheaper closed-loop brushless motors work that would be ideal, but I suppose I can't stretch the laws of physics if they don't stack up.

Chris -

I checked out the website and it looks directly applicable to what I'm trying to accomplish, thanks for shedding light on the idea. I'll do some research on it and see where it takes me.

I forgot to attach those FANUC files...

FANUC_image.png

FANUC_image.png

Yes, the FANUC type was what I had in mind. The controller will cost more.
Paul

If you want to simplify the maths, the first thing to do is use SI units throughout...

This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.