Is there a torque to mass forumla

[Alewx]

Aloa everyone, my latest spacecraft has something at about 2000t in orbit, and the torque from reaction wheels are barley able to keep it stable.

So my question is quite simple, is there a formula which enables you to calculate the toque you need compared to the mass of the craft?

[Luckless]

There are a lot of variables at play for that, mostly relating to how the craft is put together, how the mass is spread out, and where the sources of torque are located. You can stick all the torque controls you want on the end of a 2km long noodle, but it isn't going to get any more controllable. But if you rebuild the same craft into a smaller and tighter ball like structure with a bit of torque control spread evenly around the centre of mass, then it can be nicely responsive and well controlled.

Post up some screenshots or the craft files and people can probably give you better pointers and areas to discuss than doing so blindly.

[Alewx]

There are a lot of variables at play for that, mostly relating to how the craft is put together, how the mass is spread out, and where the sources of torque are located. You can stick all the torque controls you want on the end of a 2km long noodle, but it isn't going to get any more controllable. But if you rebuild the same craft into a smaller and tighter ball like structure with a bit of torque control spread evenly around the centre of mass, then it can be nicely responsive and well controlled.

Post up some screenshots or the craft files and people can probably give you better pointers and areas to discuss than doing so blindly.

It is a welded Craft so the source of torque is located at the CoM.

Javascript is disabled. View full album

[Wemb]

Yes, Luckless ninja's me - you can calculate this - but you need to know the mass distribution of your rocket, do some integration to work out Moment of Inertia to work out how much torque is needed to get it up to a desired rotational speed in a desired timescale. Not a trivial calculation. - As he said, put the Reaction Wheels near to the CoM to help you rotate it. You can put thrusters as the extremities to rotate a ship quickly - but that will possibly screw you up if you ever want to use them for lateral movement instead.

Wemb

[Nathair]

But if you rebuild the same craft into a smaller and tighter ball like structure with a bit of torque control spread evenly around the centre of mass, then it can be nicely responsive and well controlled.

It is a welded Craft so the source of torque is located at the CoM.

It does not matter where torque is applied.

Edited July 23, 2015 by Nathair

[Wemb]

It is a welded Craft so the source of torque is located at the CoM.

http://imgur.com/a/sVtI5

Okay, so this means your ship will have uniform denisty - but it doesn't say anything about the shape.. The torque needed to rotation a long thing object about it's longitudional axis (e.g. the roll of a rocket) is obviosily a lot less than that needed to turn it 180deg around. If you can think of your ship as vaguely spherical, then the MoI is

But you haven't really said what you're trying to calculate - -any- reaction wheel will rotate any size spaceship to whatever RPM you wish. Eventually.. What sort of RPM are you trying to get your ship to rotate at and how fast to you want it to accelerate to that RPM?

As said, non-trivial.

Wemb

[Alewx]

Okay, so this means your ship will have uniform denisty - but it doesn't say anything about the shape.. The torque needed to rotation a long thing object about it's longitudional axis (e.g. the roll of a rocket) is obviosily a lot less than that needed to turn it 180deg around. If you can think of your ship as vaguely spherical, then the MoI is

http://latex.codecogs.com/gif.latex?I%20%3D%20%5Cfrac%7B2%7D%7B5%7DMR%5E2

But you haven't really said what you're trying to calculate - -any- reaction wheel will rotate any size spaceship to whatever RPM you wish. Eventually.. What sort of RPM are you trying to get your ship to rotate at and how fast to you want it to accelerate to that RPM?

As said, non-trivial.

Wemb

Oh ok didn't know that it will be made to a rpm; well then a rpm of 0,5 would be awesome compared to the size and mass, the acceleration should not be too high.

But as you can see it's shape is more of a rocket than the of sphere.

Edited July 23, 2015 by Alewx

[Wemb]

Oh boy - okay - well, your ship is vaguely long and thin - the formula for I in that case is I= 1/12 ML^2 - though, in reality, yours is clearly not uniform and has more mass at the end of one end that the other does - but still, that'll roughly give you the I. The equivalent of Newton's second law for rotational systems is

Where alpha is the rotational acceleration in radians per second per second. (A radian is ~57 degrees, there are 2 pi radians in a circle). So to get your ship spinning at 1rpm from 0 rpm in 150 seconds you'd need to calculate for a value of alpha of

Your value of I, using the above forumla will be too low - since most of your mass is at the end of the ship, not distributed evenly.

I wouldn't also assume that KSP doesn't play tricks with the calculations for this kind of stuff either - if it's a problem for us to work out, KSP may also make some simplifications in it's modelling.

Wemb

[Alewx]

Oh boy - okay - well, your ship is vaguely long and thin - the formula for I in that case is I= 1/12 ML^2 - though, in reality, yours is clearly not uniform and has more mass at the end of one end that the other does - but still, that'll roughly give you the I. The equivalent of Newton's second law for rotational systems is

http://latex.codecogs.com/gif.latex?%5Ctau%20%3D%20I%20%5Calpha

Where alpha is the rotational acceleration in radians per second per second. (A radian is ~57 degrees, there are 2 pi radians in a circle). So to get your ship spinning at 1rpm from 0 rpm in 150 seconds you'd need to calculate for a value of alpha of

http://latex.codecogs.com/gif.latex?%5Cfrac%7B2%5Cpi%7D%7B150%7D

Your value of I, using the above forumla will be too low - since most of your mass is at the end of the ship, not distributed evenly.

I wouldn't also assume that KSP doesn't play tricks with the calculations for this kind of stuff either - if it's a problem for us to work out, KSP may also make some simplifications in it's modelling.

Wemb

Thanks, yeah not easy, I know but if it is easy it must be broken .

This might already help me. But mostly I was thinking too oversimplified, more of something like 20KN torque per 5 tons of mass or something, but sure that is something totally non measureable.

[Wemb]

Thanks, yeah not easy, I know but if it is easy it must be broken .

This might already help me. But mostly I was thinking too oversimplified, more of something like 20KN torque per 5 tons of mass or something, but sure that is something totally non measureable.

The trouble is that isn't a thing - you want 20kNM of torque per 5 tonnes of mass at some distance away from you to accelerate your rotation by x rotations per second.

Having said that, I had a second look at the picture you posted and am pretty confident that, given your ship's size and, assuming all those orange tanks are full, then the optimal number of Advanced Reaction Wheels you should fit, at the CoM, would be 'all of them'. Or, alternatively, just consider rotating the university around your ship instead. It won't be any less expensive, but the maths might be easier.

Wemb

[Alewx]

The trouble is that isn't a thing - you want 20kNM of torque per 5 tonnes of mass at some distance away from you to accelerate your rotation by x rotations per second.

Having said that, I had a second look at the picture you posted and am pretty confident that, given your ship's size and, assuming all those orange tanks are full, then the optimal number of Advanced Reaction Wheels you should fit, at the CoM, would be 'all of them'. Or, alternatively, just consider rotating the university around your ship instead. It won't be any less expensive, but the maths might be easier.

Wemb

that is a nice idea, I might try this out, just need to write to mod for it

No it is really extreme with 7000t wetmass, the orange tanks are not even the Jumbos but Fueltanks+, the torque I have in that section is somewhere at about 2000KN, so very likely not enough.