Direct-Drive Linear Actuator with Lead Screws?

Hi all; I've been reading this forum for a while and am just now starting to really plan out my simulator project. With that said, I've seen a fair number of people building their own actuators but none of them are done in the way I'd have thought. I'm not at all saying they are bad designs, there are a lot of really clever ideas and great implementations. I'm just surprised that no one is doing a direct-drive approach with a lead screw which seems the simplest approach. Have I just missed seeing these?

In specific, if you take a motor like this one:
https://www.motioncontrolproducts.co.uk/products/8/24/zyt90-155-12v-traction-motor-04-nm/
I've seen people talking about it so I figured I'd use it as an example though it seems a little pricey to me. That is 0.4 Newton Meters @ 1500 RPM according to the site. To get away with using online-tools as much as possible, we need to do some conversion. Let's turn that into ounce-inches here:
http://www.unitconversion.org/energy/inch-ounces-to-newton-meters-conversion.html

So that motor has 56.644 ounce-inches available. Now we need to turn that into linear motion. My initial inclination would be a lead screw and so we go to Roton's website and find they have lots of them. I've heard lots of people mention 200mm/s is a good starting point for speed. So by looking at this site:
http://www.roton.com/Engineering_Data.aspx?line=Hi-Lead
and then using this calculator:
http://www.cncroutersource.com/linear-speed-calculator.html
we find that a lead of .333 @ 1500 RPM will give us a velocity of 500 inches/minute or ~211mm/s. That would be a 1/2" x .333 lead screw.

Now how much moving power does that give us with that motor? Use this calculator here:
http://www.cncroutersource.com/linear-force-from-torque-calculator.html
Plug in the .333 lead, 56 oz/in, and the efficiency listed by Roton for that screw (60%) and you end up at ~40 lbs of linear moving force.

What I like about the linear actuator using a lead screw is that it doesn't have any pulleys or any other reduction than the screw/nut itself. Direct driving a lead screw is very easy and dead reliable. No belts/chains, etc. It's also going to be extremely quiet. The price is reasonable as well. While that specific screw choice isn't terrible cheap, you can see here what everything would cost:
http://www.roton.com/Mating_Components.aspx?family=7060912
1 foot of the screw is $12. The flange nut you would need is $20 dollars. You then use a straight shaft coupler between the motor and the lead screw:
http://www.ebay.com/itm/11-mm-x-12-...411?pt=LH_DefaultDomain_0&hash=item27daf626c3
That's not quite the right size, this motor would need 11mm to 12.7mm but you can see that this company sells em for about 10 bucks. You can use the other side of the lead screw to attach a good rotary encoder for position feedback:
http://www.ebay.com/itm/600P-R-Phot...173?pt=LH_DefaultDomain_0&hash=item258e0ceebd
Anyways, you get the idea.

Now taking that further to a more capable motor but still pretty cheap to drive, you could do:
http://tncscooters.com/index.php?route=product/product&path=52_55&product_id=56
24v @ 450 watts @ 2600 RPM - getting ounce-inches for that is a little tricker and requires this equation:
power in watts = torque in ounce/inches * speed in RPM * 0.000739
So that motor puts out 234 ounce/inches at a reasonable RPM. So using the calculators before we find that 1/2" x .2 lead is a good mix for that motor. You can get 215+ lifting pounds at 220 mm/s of speed. This dirt cheap driver appears capable of running that motor as well:
http://www.ebay.com/itm/Semiconduct...242?pt=LH_DefaultDomain_0&hash=item5b03f468fa

Anyways, am I crazy with this? Is there a reason very few, if any, people are using lead screws for their actuators in simulators? Anyways, I have a working linear actuator design that can be build on a drill press and a chop (or even hand saw) that should work nicely for our application if I can understand why people are not using this more frequently. Thanks much for any feed back!

-Mike

I have also been wondering the same thing as to why no one has built an actuator like you have described. I am interested in learning more or trying to tackle something along the same lines too. Let us know how you progress on this.

I have been investigating the same thing for some time, but as I have a working sim other jobs get in the way. A gear driven arm is just simpler/cheaper then a lead screw setup.

I looked at ball screws as well after I bought the parts to build the Ian BFF actuators, of which @eaorobbie has the parts now.
Actuators are off the "to do" list here.
I am looking at building my own pedals next. Then a mod to a current FFb wheel base.

Unfortunately, I think you've described in great detail why they aren't more popular. $20 here, $30 there, well, it adds up, especially if you have more than 1 DOF.
You still need the heim joints, so in effect, you are replacing a few dollar lever or a $15 chain/gear with a much more expensive system (comparatively).
The rotary encoder needs to be high speed and continuous, you'll need hard stops and sensors, it'll need re-calibration.

On the extreme positive side, it'll probably give better feeling movement, as lever or chain systems don't have linear movement, but rather based on sin().
That's a big plus, but it can be compensated for in software, or even completely ignored (we humans aren't the best judges of motion when it's happening to US).

One disadvantage is that you have to be very careful to avoid backlash and slop, either of which would have a very negative impact on your sim. Most gear boxes are very nicely designed to avoid that, and a lever obviously adds none. With a ballscrew/nut, it in effect becomes the gearing, so make sure it's up to snuff.

I would love to go the leadscrew route, it's just out of my price range.

with DIY actuators, allow $400 each Id say per unit to build, but they would be stronger than most on the market that come in at double the cost