Ruberband for measuring torque

I have to specify a motor/gearhead combination for automating the steering column of an ATV. The first step I believe is to measure the torque
necessary to move the wheels with the vehicle stopped on asphalt. I figure that if I choose a motor that has that torque and a little margin, I won't have problems when the ATV is in movement.
Here's the method I'm using. I got 20 rubberbands together, hooked at the end of the ATV's handlebar and pulled (tangent to the steering column) as fast as I want the motor to pull it. I chose 20 rubberbands because if I used less, it would stretch to the limit, with 20 I still have plenty of room to stretch. In this experiment, the rubberbands went from 3 inches (no torque) to 6 inches (max. torque), so 3 inches of displacement. Then I used the same 20 rubberbands to lift a weight of 30 pounds, it stretched 9 inches (6 inches displacement). Can I assume that the elasticity of rubber is linear in this case? If so, then 3 inches of displacement would lift 15 pounds right?
The next thing I did was to calculate the force necessary to cause that 3 inches displacement multiplying 15 inches by g, and then multiplying by the handlebar distance from the center of the axle.
I ended up with a torque close to 25Nm, so I'm thinking about choosing a motor/gearhead with a 50Nm available torque. RPM should be close to 20rpm.
Does that make sense? I saw some nice motors from Harmonic Drive that would fit my application perfectly (small, very torquey, hollow shaft), for instance the FHA17C, but they are too damn expensive (around $2800) http://www.hdsi.net/products/rotaryactuator/fha.cfm
I need something around $100-$200 (motor+gearhead), am I being realistic?
Cheers
Padu
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
The force needed to stretch rubber isn't linear, especially over repeated tests. Think of a rubber balloon; they're easier to blow up if they're stretched first. I think after the first couple of tests your results will start to go way off. You're better off with fish weights or something with a metal spring. You'll find some articles and example test results of doing this type of thing with a good Google search.
To get what you want for the price you need, you may have to get a more routine motor and add some sub-gearing either on the steering yoke or the steering column. Look for a motor with an initial output speed of around 500-1000 RPM unloaded, so that it can be further geared down to around 50-80 RPM. That will achieve a 90 degree arc in a few hundred milliseconds; anything less is like a drunk behind the wheel. I think you want to be able to steer right/left in under 250ms. (Assuming a 90 degree arc for the steering, that's 1 RPS, or 60 RPM -- but remember, under load.) Go to Industrial Liquidators on Convoy for a selection of nice large metal gears perfect for this job.
Another thing to look at is pneumatics or even hydraulics. You can find pneumatic cylinders with throws of 10-12 inches surplus for under $20; C&H in Pasadena is a good local source. Add an air ballast and a pump for maybe $100-150. The trick is in making this proportional, which can be done using feedback circuits and PWM to the valves. Might be a nice project for one of your team mates... <g> In any case, good speed and nice torque with air/liquid.
-- Gordon
Padu wrote:

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
"Gordon McComb"

I'll repeat my tests with a more formal method. The rubber band was a very informal test just to get a ballpark torque number. Turns out that it may be way off, either because of my measurement method or wrong calculations as I found very difficult to fina a 50Nm (double of what I measured) motor.

Thanks for the hint, I'll take a look.

Well, I have very little knowledge on motors. Seems like I'll have to go through an intensive course this coming days. Luckily (or unluckily), there are lots of information on the internet mainly on the motor manufacturers. I even downloaded a software from Parker's that help you specify and choose your motor, but most of the parameters it requires are still way over my head.
Thanks
Padu
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.