something about angular and linear acceleration

[rdem]

I made a craft and send it to space to test the relation between angular and linear acceleration. Here's the picture.

(SAS was off and gravity was hacked)

It has an accelerometer at the center of mass (at least very close to it).  First of all I wanted to test if it detects angular acceleration. So I fired the two engines marked red in picture at same time. So it has only angular acceleration. The accelerometer was detecting zero g in this process.

Then I only fired the engine towards the center of mass (yellow). Accelerometer counted it as 0.240g.

Then I only fired the engine marked with blue. That engine must create both angular acceleration and linear acceleration in the direction of thrust. It happened as I tought and in this process accelerometer counted it as 0.240g same as with the yellow engine. [Note: if craft turns too fast the value on accelerometer will start to decrease. that is because when it turns too fast, force is constantly changing direction and kinda disables itself over the time. The accelerometer can’t be trusted in this situation.]

So in my opinion, if the accelerometer detects only linear acceleration and not the angular, it should have detected smaller g than 0.240 because not all of the force is used to cause linear acc. some of it used to turn the craft.

I also tried with huge masses added near center of mass (symmetrically) but result was same.

Prove me wrong or KSP’s physics engine or accelerometer.

[Gaarst]

I guess that KSP's accelerometer only detect acceleration experienced by the centre of mass. If your craft is symmetrical, then the natural centre of rotation is at the same point as the CoM and therefore it doesn't experience any acceleration even if the craft is rotating.

[rdem]

9 minutes ago, Gaarst said:

I guess that KSP's accelerometer only detect acceleration experienced by the centre of mass. If your craft is symmetrical, then the natural centre of rotation is at the same point as the CoM and therefore it doesn't experience any acceleration even if the craft is rotating.

that is not what i asked. i already tested it with the red engines that create only angular acc. accelerometer didn't detect any a in this process. so i came to the conclusion that a.m. doesn't count angular acceleration. when only blue engine generates thrust it also turns the craft unlike the yellow one. but this time a.m. detects acceleration. so it must have caused also a linear acceleration. 

Edited September 15, 2016 by rdem

[Steel]

Because the engine is still accelerating the craft at 0.24g (the net force on the craft is still the same regardless of where the engine is pointing), its just that because you aren't thrusting through the centre of mass that it causes the craft to spin as well

[rdem]

6 minutes ago, Steel said:

Because the engine is still accelerating the craft at 0.24g (the net force on the craft is still the same regardless of where the engine is pointing), its just that because you aren't thrusting through the centre of mass that it causes the craft to spin as well

spinning a craft also requires force. so if it turns the craft, not all of it is used to accelerate it. so it should have less linear acceleration.

[Steel]

7 minutes ago, rdem said:

spinning a craft also requires force. so if it turns the craft, not all of it is used to accelerate it. so it should have less linear acceleration.

Not true, the net force on the craft will accelerate it linearly, the net torque due to that force not acting through the centre of mass will cause it to rotate. 

[rdem]

5 minutes ago, Steel said:

Not true, the net force on the craft will accelerate it linearly, the net torque due to that force not acting through the centre of mass will cause it to rotate. 

but if i use the two engines towards opposite directions: one with less and one with more r (distance from com) it should detect 0 g. but it still detects some acceleration.

[Steel]

3 hours ago, rdem said:

but if i use the two engines towards opposite directions: one with less and one with more r (distance from com) it should detect 0 g. but it still detects some acceleration.

My guess (if its quite a small acceleration) would be that flex in the structure of the craft is causing the engines not to fire in exact opposite directions in this case.

[mikegarrison]

Real accels detect acceleration in only one axis. The behavior you expected would have been seen with a real accel. But KSP's accelerometer part is like a three-axis cluster of accels. Those can measure the components of the acceleration in all orthogonal axes and then resolve them into a single vector.

Edited September 15, 2016 by mikegarrison