Gear Boxes, the way of the future, or a dead end?

[Gargamel]

I've been trying to stack rotors, to increase the total RPM's being put out.   But.... I'm running into some weird glitches.  Probably user error though.

Anybody had any luck stacking rotors?

I had thermometers on each rotor and they seemed to be increasing in speed with each iteration of rotor.   But once I placed a heavy load on the rotors, it all went.... sideways... couldn't make heads or tails out of it....

[Pthigrivi]

1 hour ago, Gargamel said:

I've been trying to stack rotors, to increase the total RPM's being put out.   But.... I'm running into some weird glitches.  Probably user error though.

Anybody had any luck stacking rotors?

I had thermometers on each rotor and they seemed to be increasing in speed with each iteration of rotor.   But once I placed a heavy load on the rotors, it all went.... sideways... couldn't make heads or tails out of it....

Someone mentioned unity has an RPM limit. Maybe that’s it? Someone who knows more have thoughts?

[Gargamel]

Sorry, got distracted during my previous post, had to handle some forum stuff and I hit submit.    The thermometers are there for visual references on each rotor, not to measure their temp :D.   So with just  a small load on them, they seemed to work as intended.   But with a heavy load.... perhaps I'm just not familiar with how they work yet, or they're bugged, but they didn't work like I though they would, and I'm not even sure what it was doing enough to describe it.....

[Kerbart]

2 hours ago, Pthigrivi said:

Someone mentioned unity has an RPM limit. Maybe that’s it? Someone who knows more have thoughts?

That's what I'm thinking. Squad mentioned that 500 RPM is kinda like "light speed" which is why they limited their max RPM to 480. The glitches you're seeing are probably the "why"

[Gargamel]

12 minutes ago, Kerbart said:

That's what I'm thinking. Squad mentioned that 500 RPM is kinda like "light speed" which is why they limited their max RPM to 480. The glitches you're seeing are probably the "why"

Hmmm.... but that limit should relate to each individual rotor?   So If I stack 2 of them, with each doing 480 rpm..... hmmm...

[RocketSquid]

Two seems to be the limit. After that, the one on the end can’t seem to stay on the same axis as the rest—it gets pulled out by centrifugal forces.

[Kerbart]

12 hours ago, Gargamel said:

Hmmm.... but that limit should relate to each individual rotor?   So If I stack 2 of them, with each doing 480 rpm..... hmmm...

Without knowing what causes that limit it’s hard to tell. Perhaps, inside the Unity Physics model, angular velocity is stored but it causes some kind of overflow effect above 500 rpm?

Gets out slide ruler

500 RPM is about 52 rad/s (note to self: 10 RPM is roughly 1 rad/sec. Interesting but makes sense. Good rule of thumb to remember) which in itself doesn’t seem to be a magical number... but who knows?

Anyway if the issue is angular velocity per part, than stacking the rotors will not really solve things.

However it seems to make this nice humming noise that sounds exactly like a Kraken invocation so by all means, continue your experiments and don’t forget to post videos!

[MechBFP]

The problem with stacking rotors is that the top of a rotor has torque one direction, which create torque in the opposite direction at the bottom of the rotor. 

As such two rotors stacked on top of each other will need to be configured so that the middle between the two rotors has torque going in the same direction and the spinny part on both are connected together. The very top of the top rotor (the stationary part) will also need to be physically connected to the bottom of the bottom rotor (the other stationary part) so that those pieces  always move together. 

In theory that should double the torque in the middle if you then launch clamp the stationary parts of the rotors so they don’t spin.  

However the RPM will stay the same. 

Edited June 3, 2019 by MechBFP

[Kerbart]

4 minutes ago, MechBFP said:

The problem with stacking rotors is that the top of a rotor has torque one direction, which create torque in the opposite direction at the bottom of the rotor. 

As such two rotors stacked on top of each other will need to be configured so that the middle between the two rotors has torque going in the same direction and the spinny part on both are connected together. The very top of the top rotor (the stationary part) will also need to be physically connected to the bottom of the bottom rotor (the other stationary part) so that those pieces  always move together. 

In theory that should double the torque in the middle if you then launch clamp the stationary parts of the rotors so they don’t spin. 

I think you’re falling for a classic physics trap.

Imagine a rope with two donkeys, one on each end. They’re both pulling with 1000 N on the rope. What is the stress the rope us under? It’s easy to think 2000 N because that’s on both ends, right? Wrong! Replace one donkey with a wall that the rope is tied to, and it’s obvious that the rope endures 1000 N of tension.

The same is going on when stacking two rotors, except that it’s torque instead of linear force. But the concept remains the same. Stacking two rotors is not going to make them twice as strong together. It does double the rotation speed though (well, in a perfect world at least).

[MechBFP]

3 minutes ago, Kerbart said:

I think you’re falling for a classic physics trap.

Imagine a rope with two donkeys, one on each end. They’re both pulling with 1000 N on the rope. What is the stress the rope us under? It’s easy to think 2000 N because that’s on both ends, right? Wrong! Replace one donkey with a wall that the rope is tied to, and it’s obvious that the rope endures 1000 N of tension.

The same is going on when stacking two rotors, except that it’s torque instead of linear force. But the concept remains the same. Stacking two rotors is not going to make them twice as strong together. It does double the rotation speed though (well, in a perfect world at least).

That is incorrect. The torque indeed doubles when you connect two motors together (assuming no gear changes)

[Kerbart]

7 minutes ago, MechBFP said:

That is incorrect. The torque indeed doubles when you connect two motors together (assuming no gear changes)

When you connect them, so they are driving the same axle: yes

When you stack them, so one drive the other who drives the axle: no.

 

Assume you have two motors, each delivering 1 kW.

The power output is torque x angular velocity (basic physics). We can agree that the speed on the axle is twice of what each motor delivers. And that the power output is twice of what each of them delivers (barring any losses, bla bla bla...)

If we have twice the power, with twice the RPM, what do you think happened to the torque?

[MechBFP]

22 minutes ago, Kerbart said:

When you connect them, so they are driving the same axle: yes

When you stack them, so one drive the other who drives the axle: no.

 

Assume you have two motors, each delivering 1 kW.

The power output is torque x angular velocity (basic physics). We can agree that the speed on the axle is twice of what each motor delivers. And that the power output is twice of what each of them delivers (barring any losses, bla bla bla...)

If we have twice the power, with twice the RPM, what do you think happened to the torque?

Indeed. I wasn’t talking about it that way though as isn’t possible to connect them that way in KSP. 

In KSP stacking a rotor on a rotor is like putting an entire engine on the shaft of another engine, lol. 

[Kerbart]

2 minutes ago, MechBFP said:

Indeed. I wasn’t talking about it that way though as isn’t possible to connect them that way in KSP. 

In KSP stacking a rotor on a rotor is like putting an entire engine on the shaft of another engine, lol. 

Aaah. Ok, that clarifies it

 

[khadgarion]

It'd be neat if you could somehow make a planetary gear set

[Yargnit]

13 hours ago, Kerbart said:

Without knowing what causes that limit it’s hard to tell. Perhaps, inside the Unity Physics model, angular velocity is stored but it causes some kind of overflow effect above 500 rpm?

Gets out slide ruler

500 RPM is about 52 rad/s (note to self: 10 RPM is roughly 1 rad/sec. Interesting but makes sense. Good rule of thumb to remember) which in itself doesn’t seem to be a magical number... but who knows?

Anyway if the issue is angular velocity per part, than stacking the rotors will not really solve things.

However it seems to make this nice humming noise that sounds exactly like a Kraken invocation so by all means, continue your experiments and don’t forget to post videos!

FYI The limit before Unity starts going haywire is 50 rad/s (just over 477rpm)

The reason for this is because Kerbal runs physics at 50 'frames' per second, so this is the point where Unity crosses 1 rad/physics tick. As you cross this line Unity's ability to accurately calculate where the parts are in relation to each other rapidly breaks down. You can see this even approaching 51rad/s, and anything above that is almost certain to cause explosions. The limit of 460 for rotors is just a safer buffer again over-speeding causing it to cross this threshold.  

[Kerbart]

4 hours ago, Yargnit said:

FYI The limit before Unity starts going haywire is 50 rad/s (just over 477rpm)

Thank you, @Yargnit ! That clears things up!

Edited June 4, 2019 by Kerbart