Torque and reactions wheels : uneven ships

[Warzouz]

I always build ship even. Only (not so) "physic-less" parts are non even. For example, as antenna aren't physic-less, I always stick 2 of them...

Then KER always shows 0 thrust torque-offset.

But recently I'm working with moving space stations and I noticed that my lander was nearly the same weight as my generic lander. So I sticked the side ways. that gave me 1° of torque offset.

When I fired the engines with SAS... nothing append : the ship flew straight. The engines don't have gimbal, so the course correction is managed by the reactions wheels (2 big ones and probe cores)

(click to enlarge)

The questions :

- How can we evaluate when it's necessary to balance parts ?

- How can we evaluate the reaction wheels torque correction needed ?

- Do reaction wheel can compensate better at low thrust ?

- What experimentation can we do to check that.

[MaverickSawyer]

I mock up the design in the VAB and use RCS Build Aid to evaluate the torque generated by offset masses.

[InterCity]

Deactivate the SAS on your landers/side modules. Leave only one on the main branch running. Shouldn't be a big deal.

[STARKEngineer]

It is only necessary to balance parts when the torque produced by the offset CoT compared to the CoM is greater than the Torque available to your reaction wheels.

Kerbal Engineer can give you a readout of engine torque on your spacecraft in the Vessel Tab when an appropriate engineer chip is included, a lvl 3 Engineer is on board or your tracking Station is at level three.

Reaction Wheel Torques are easily findable on the wiki (remember that most Pods include Reaction Wheels.)

If the amount of Torque KER gives is less than the Torque of the Wheel, then SAS will hold course. If the offset engine torque is greater than that of your reaction wheel, then your ship will be unable to hold course.

Due to lever arms and such, an engine running at low thrust will produce less torque from an offset CoM than an engine running at high thrust at the same point.

Several experiments to do with reaction wheel placement have been done over the years, just google for reaction wheel placement KSP. As for offset CoT/CoM torque, I am unsure as to where to find examples but mostly proven in the Math.

[bzimac]

Amount of torque generated by offset thrust depends on two things. First is distance from center of mass to the line along the thrust vector. Second is amount of thrust.

What helps here is that your ship has a lot of mass (at least with full fuel tanks), so center of mass doesn't move very far. Also, nuclear engines don't have a lot of thrust.

In case reaction wheels were not able to control the ship, you could try to balance it by transfering fuel between landers, or by limiting thrust of the engines.

[Warzouz]

Thank you for you replies.

SAS can handle torque offset. I didn't noticed first, but it also show a thrust torque which depends on the engine providing the thrust.

I did some testing, even increasing torque by moving fuel around, messing my save file. SAS is capable of handeling the torque. BUT, the heading is still sliding because SAS don't keep a clear heading. It works better when targeting prograde, radial or normal. Event a node will keep it straight.

That's for reaction wheels.

Now for Gimbal : engines (but not LVN on my test vessel) have gimbal. My ship would be stable is the torque offset angle is lower than the gimbal angle ?