Servomotor actuated worm gear prosthetic knee

Servomotor actuated worm gear prosthetic knee

I'm an amputee, and not exactly an master engineer, not by a long shot, (I can't even remember how I passed physics back in college) but I'd like to try and take a stab at building my own prosthetic leg. I'm a welder, so fabricating a frame wouldn't be a problem at all.

My problem is a bit of a catch 22, I don't know what servo motor to use, as I don't know what sort of torque is required, and I don't know what torque is required because A. I don't know what worm gear I'm using, and B. I'm not exactly sure what forces are involved.

As it stands, my current prosthetic is essentially a swinging hinge. I can't modify this one, or I'd be voiding the warranty.

**broken link removed**

Those are some of the dimensions of my current one

**broken link removed**

**broken link removed**

Thats a 3d model of what I've decided would work best. I've considered nitinol wire working against a spring, but it was too complicated, and SMAs aren't advanced enough or cost effective for this sort of application.

The worm gear would produce the sort of braking I need without consuming any current for holding my weight up at a half-bent position.

As far as controlling this thing, I figure I'll build a battery bank (yes, i know this is going to be heavy) and hook up a rocker switch for forward/backward. At some point, I'd like to try and hook up a more elaborate swith, more of a joystick type deal so I could vary the speed forward/backward for smoother operation.

Anyone got any tips? What servo should I use, and from where? What about the worm gears?

A worm gear behaves like a gear with one tooth so the reduction ration is simply 1:the number of teeth on the cog being driven. Calculating the required torque should be simple, you know how much you weigh and the length of your calf to knee so it just a simple moment calculation.

Also bear in mind that worm gears aren't very efficient, I remember seeing figures between 50% and 70% mentioned somewhere so you'll need use a motor with nearly nearly double the torque you calculate.

interesting project and I am impressed that you aren't waiting for some one else to do it. I hope we can help you here.

the other thing to consider is the speed that you will need.

I wonder how big the motor has to be. My gut instinct is that it would be pretty big to have the needed torque and speed. I think you've got the right idea of placing the motor in the thigh. Also, I think the worm gear plus servo approach could be slow. I'd suggest you opt for directly driving the motor.

Just a thought (and I'm far from an expert), looking at the way a leg works, I wonder if having 2 bands that attach to the front and back of the top of the lower leg. The bands would be pulled by a pulley in the thigh. Pulling on the front band would cause the lower leg to swing forward, pulling the back would cause the lower leg to swing back (and up).

It's actually amazing how well the simple free swinging legs work - there was a young girl who used to be at our local primary school (she's late teens or so now, and a stunning looking girl by all accounts?). I'd always presumed she was born with the missing leg?, but apparently she was in a car crash as a young baby and lost it.

Anyway, I was watching sports day at school (my daughter is a couple of years younger) and I couldn't spot tell which of the runners had the false leg!, she was absolutely amazing on it.

She provided years of entertainment for the school, a common occurance was the entire school searching for her leg - she was forever losing it! - it got used for playing cricket and all sorts of stuff.

As if losing a leg as a baby wasn't bad enough, when she was 9 or 10 years old her mother dropped dead as well - totally unexpected, just dropped and never got back up. Some people have so much bad luck in their lives, yet still seem to be happy beautiful people!.

Is your goal to hold the position of the joint fixed or actually power the joint thru the various motions that you require?

Some people have so much bad luck in their lives, yet still seem to be happy beautiful people!.

Click to expand...

Yet others who are fit and healthy and have perfectly good parents with enought money get depressed or get into trouble.

Aw, thanks for the support & kind words guys

I don't let it keep me from doing things, but there are some things I could do better if the knee actually behaved like a knee, and not a swinging hinge. As it stands, it locks at heelstrike, and when most of my weight is over the toe, it unlocks, allowing it to kick up behind me, giving the illusion that I'm moving the knee.

The main factor with how well you walk with an above knee prosthesis is how long your residual limb (stump) is. I've got what prosthetists call a "drastically short" residual limb; if it was any shorter, they most likely wouldn't bother fitting me for prosthetics. So because I have such a short limb, it makes it difficult to exercise more precise control over it (thats a no brainer my insurance company doesn't understand ). At some point, I'd like to have osseointegration surgery, but thats a whole other story.

Right now, I'd just like build a proof of concept, even if the knee doesn't work quite as I'd like, I'd still have more experience with things I've been interested in exploring regardless. I do think this would work, but as some of you have said, it depends on the servo.

I chose the wormgear setup because if the braking action inherent to the design. From what I understand (and I could easily be wrong) the worm cannot be turned by the workwheel, only the other way around. If I choose a 20 tooth wormwheel, I'll have a speed reduction of 20:1. If I only need up to 150 degrees of rotation, a 1200 RPM servo would be reduced to 60RPM. I was aiming for 1 rotation per second, and seeing as its only rotating 150 degrees, this is probably plenty.

21.5" X 200lbs = 4300... pound inches? If I did that right, I'll need a hefty servo If I use 20:1 gearing for the wormgear setup,then I think I'd need a servo that can provide at least 215 lb-in of continuous torque, at 1200RPM. Taking into account what Hero999 said about the efficiency of wormgears, 430 lb-in of torque might be more like it, but thats going to be a gigantic servo...

why the choice of a servo? With a bit more electronics, you can have much tighter control over the motor. This will allow you more freedom to select the right sized motor and will give you more control over the speed as well. I assume you would want to vary the speed of the flex. It takes more work to control the speed of a servo.

By the way, have you given much thought to using some sort of sensing technology to determine what the user is doing (walking speed) and adapt the motor control to that?

Servos are just what I chose, if you know of something better please give me some advice on the matter, I'm not an electronics genius, thats why I came to this forum

As far as determining walking speed and all that, I'm not ready for anything close to building a microprocessor control system for the knee, I'm just looking to build a simple functioning unit that can hold my weight and provide stability while moving, if I can get the knee working, then I can worry about microprocessor controlled stuff.

You might be able to fudge a lot on the servo power; the hinge has to support your body weight while your foot is on the ground, but you won't be lifting your entire body weight with the motor.

You do have to design the algorithm so that when you have too much strain on the motor, it stalls gracefully (without smoke).

You might be able to begin prototyping with the window-crank motor out of an auto. I see them cheap at surplus places like All and Electronic Goldmine.