Which Ballscrew pitch / Dia / Length ?

T

I thank you for your support once again.
It's almost unbelievable to think that 4 motors so small (mm63x74) can have the strength to move a 180kg sim fast but the data say this.

Hi Gigi
You use 20% from base speed wattage with 2G, but the specs are with extremly good forced air cooling.
I cant thell how mush you need to cool.
Cooling costs money to and if you look at aliexpress with bldc motor with encoder ,you have a industryal motor with encoder for less . And he will have a far longer motor lifetime
400/500watt will do
Regards Ad

Hello,

I am building a 6DOF platform with linear actuators.
I have 80ST-M04025 servos (1kw, 4Nm at 2500rpm).
I predict that the platform + load will be at maximum 150kg all together (more likely less than that).
Platform will be used mostly for rally racing (RBR, DR1, DR2...) and occasionally some other racing (F1, AC...).

I want the actuators range about 500mm.

So I am looking at two possible ball screw types here, 1610 and 1616.
In other words max. actuator speeds of 416mm/s and 666mm/s.

If i put these numbers in excel sheet made by @Dirty it seems that the power is sufficient for both types.

What I would like to know: is the increase in actuator speed with 1616 actually beneficial for the mentioned types of racing or is it better to go with 1610 and lower speed with more headroom for acceleration?

If anyone has any opinion based on experience I would really appreciate it.

Thanks,
Pawks

My advice is fit a belt drive at the end of the actuator and you can adjust the speed by changing the pulley size and rpm

I ended up using gearmotors and levers and i am very happy with the performance, the speed is in excess of 1000 mm/s but with less movement. It gives instantanious repsonse and high fidelity of sensation. You can feel all the bumps and kerbs etc

Lots of linear actuators I have seen are a bit lazy and the experience becomes a bit damped trying to keep up with motion

Hey

The setup you're describing is exactly what I have (80ST-M04025, SFU1610, 417mm/s speed, 580mm stroke, 150Kg moving mass). I don't have any experience with racing sims, but a lot with flightsims. It might still be helpful for you:

I don't think I have ever needed more than ~20% of the available torque, even in peak demand. 99.9% of the time I operate at around 10% torque. However I've had some occasions where I needed all of the available 417mm/s speed.

Now, on my current actuators I could change the belt ratio for a few bucks, but I am already deep into building the next rig and one of the changes will be a slight shift away from torque and towards speed.

If I were in your position (which is in fact very similar to mine) I'd go for the SFU1616 and a 1:1 belt ratio, which will give you...


An added benefit would be a much quieter operation. Let me put it this way:
"You will not hear the difference of the motors having to apply 37.5% more torque, but you will certainly hear the difference of the ballscrews spinning 37.5% slower!"

I am even considering using 2525 ballscrews with a 2/3 belt ratio to reduce ballscrew RPM further. It' s those last 10% of RPM that actually create 50% of the noise. ( That's a guess, not a measurement!)

Cheers...

Hi.

Thank you for very informative reply...
I haven't even thought of the noise...

After hearing this, 1616 is definitely the minimum for me...now I'm even considering the next option, which is 2020... and it checks out fine...



Still, in case the power would be insufficient I can reduce the ratio on the belt drive as you mention.

Thanks,
Pawks

Bump!

In my quest to DIY a traction loss simulator, I have resurrected this thread to ask a question.

I am highly interested in pairing an 80ST-02430 with a 1616 ballscrew (or even possibly a 2020, like @Pawks has discussed above). My understanding is that the torque required for the 1616 and 2020 ballscrews (3.33N*m and 3.12N*m, respectively) exceed the rated torque of the 80ST-02430 (2.40N*m). Should I be concerned that the motor will prematurely fail, or am I just being silly?

If my worries are indeed unfounded, could I then pair the 80ST with a ballscrew of a larger pitch, as long as I stay below the instantaneous torque peak of 4.00N*m? Since this actuator will not have to work against gravity, I can play with a tighter margin, right?


Thank you for the spreadsheet, the thread and the discussion that has followed. It has certainly given me a lot of ideas!

Hi

If the actuator is not carrying any vertical load, but only pushes the rear of the sim sideways, then I'd say an 80ST-02430 is way enough. On a 1616 ballscrew it would generate around ~950N (95Kgs/200lbs) of lateral force. Probably overkill already.
I made the spreadsheet for 6DOF Stewart-type rigs, but you can still plug in the data of your actuator (tabs 1-4) and see the results. The True-False flags will be meaningless though

Dirty

Hello Dirty. Thank you for your insight; you've provided the clarity I was missing, haha.

Take care!