go_away

Author Topic: Punching Arm - Motor Selection/Project Feasibility  (Read 1277 times)

0 Members and 1 Guest are viewing this topic.

Offline RobotTemplarTopic starter

  • Beginner
  • *
  • Posts: 6
  • Helpful? 0
Punching Arm - Motor Selection/Project Feasibility
« on: February 23, 2011, 04:31:08 AM »
As part of a larger project, I plan to build a robot arm that is to be controlled in real time using flex sensors.

I want the arm to be able to be usable as an arm, so stepper motors seem like the way to go because of their ability to lock in place.  However, it is also necessary for the arm to be able to swing at as high a torque (and a reasonable, human-like rpm) as possible.  I have read that stepper motors are not well suited to this. 

Basically it has got to be able to punch things and swing a sword as well as a burly human could, but also to not have its joints being continually affected by gravity or the wind (for example, it should be able to hold a shield).  Is this feasible?   

Offline HyperNerd

  • Robot Overlord
  • ****
  • Posts: 139
  • Helpful? 5
  • N3rd1n80r
Re: Punching Arm - Motor Selection/Project Feasibility
« Reply #1 on: February 23, 2011, 05:52:44 AM »
Welcome to the forum!

The things you need to consider are how much the sword / shield weigh, as well as how compact the whole machine will need to be. If size is of no concern to you (ie the rear will be behind a display with only the arm showing), then this project could be possible with hydraulics / pneumatics, but fine movement of the arm would be difficult.

You are correct in saying that stepper motors are not suited to this - they lack the speed which you are looking for.

How lifelike does the movement have to be? A solid metal arm which only moves at the shoulder would be very easy to build, but would look like just that - a swinging piece of metal.

Overall I think this project could be done, but more planning definitely needs to be carried out.

I hope this helps,

 -HyperNerd

EDIT: Try asking over in the Mechanics / Construction board - they may be able to help you better over there.
« Last Edit: February 23, 2011, 05:55:04 AM by HyperNerd »
There are 10 types of people in the world - those who understand binary, and those who don't.

Offline RobotTemplarTopic starter

  • Beginner
  • *
  • Posts: 6
  • Helpful? 0
Re: Punching Arm - Motor Selection/Project Feasibility
« Reply #2 on: February 23, 2011, 07:51:03 AM »
EDIT:  OKAY!  I found a solution (probably), using tendons.  Thank you HyperNerd for your useful input (particularly the knowledge that stepper motors would definitely be unsuitable)!

(Edited out now extraneous information.)
« Last Edit: February 23, 2011, 10:33:15 AM by RobotTemplar »

Offline hoosier122

  • Full Member
  • ***
  • Posts: 84
  • Helpful? 0
Re: Punching Arm - Motor Selection/Project Feasibility
« Reply #3 on: February 23, 2011, 10:59:50 PM »
Link to artifical tendons?

Offline RobotTemplarTopic starter

  • Beginner
  • *
  • Posts: 6
  • Helpful? 0
Re: Punching Arm - Motor Selection/Project Feasibility
« Reply #4 on: February 24, 2011, 12:31:01 AM »
Link to artifical tendons?

By tendons I just mean twine or cable or whatnot, powered by a motor.  Then stiffness is achieved by "antagonistic actuation", where you have the two motors that control a joint pulling with equal strength in the opposite direction.  My plan now is just to use dc or servo motors for this, and braze on some additional parts so it can act as a spool of sorts.  Perhaps there are linear motors out there that would be more suited for this?  I have looked at solenoids, and they don't seems like they would have nearly enough power.  The only problem I have not solved (mechanically speaking), is how you would root the twine or cable to the "spool".  For a prototype this should not matter, but once you're dealing with higher torques duct tape or adhesive is not going to be enough.  I wonder, has anyone ever tried soldering twine to metal?  My intuition leads me to believe that soldering would just burn through the twine or otherwise structurally compromise it.

I was inspired in part by this:
http://spectrum.ieee.org/automaton/robotics/humanoids/dlr-super-robust-robot-hand

Though I have come across a number of other robots on the internet that are modeled on the same concept. 

Offline HDL_CinC_Dragon

  • Supreme Robot
  • *****
  • Posts: 1,261
  • Helpful? 5
Re: Punching Arm - Motor Selection/Project Feasibility
« Reply #5 on: February 24, 2011, 03:56:34 PM »
I would use fine aircraft cable or perhaps that wire thats often used for hanging pictures instead of twine. Also, instead of using two motors that are pulling against each other, which of course would just burn them both out, use one motor but have your tendon be a closed loop and just have it coil around the spool on the motor a few times. So when the motor spins one way the hand opens, when the motor spins the other way, the hand closes. Each finger would only need one motor.
United States Marine Corps
Infantry
Returns to society: 2014JAN11

Offline hoosier122

  • Full Member
  • ***
  • Posts: 84
  • Helpful? 0
Re: Punching Arm - Motor Selection/Project Feasibility
« Reply #6 on: February 25, 2011, 01:16:58 AM »
I dont have a link but I did read about artifical tendons which an electrical charge can be applied to various points to get it to contract.

Offline HyperNerd

  • Robot Overlord
  • ****
  • Posts: 139
  • Helpful? 5
  • N3rd1n80r
Re: Punching Arm - Motor Selection/Project Feasibility
« Reply #7 on: February 25, 2011, 11:40:57 AM »
Hoosier, do you mean SMA (smart memory alloy) wire? We got to play with some of that stuff at school :D *nerdgasm*

To my knowledge though it doesn't have the strength to do what this project needs.

 -HyperNerd
There are 10 types of people in the world - those who understand binary, and those who don't.

Offline RobotTemplarTopic starter

  • Beginner
  • *
  • Posts: 6
  • Helpful? 0
Re: Punching Arm - Motor Selection/Project Feasibility
« Reply #8 on: February 26, 2011, 07:20:28 AM »
So after going to all this trouble redesigning the robot to use the motor-cable actuation system, I learn that worm gears have the ability to lock in place because they are not reversible (info. from "Robot Builder's Bonanza")!  Since I was planning to use worm gears for my gear reduction all along, I never had anything to worry about!  I post this here so that any other would-be builders of humanoid robots may learn from my floundering and not waste time on this problem.

(However, I have decided to stay with the cable system because it seems more organic, though it is much messier and generally problematic.)

EDIT:  I should have just edited my first post to put this information in, sorry
« Last Edit: February 26, 2011, 07:23:38 AM by RobotTemplar »

 


Get Your Ad Here