Accelerometer within a freely rotating sphere?

[andyfive]

Hi,

Please could someone explain how they think an accelerometer would work if positioned within the center of a freely rotating sphere (e.g a kicked football)? If using triple axis accelerometer and the ball was kicked from a standstill but with no spin, I would imagine that the accelerometer would give the linear acceleration in the X,Y and Z axis. Is this correct? What would then happen if the ball was kicked with an amount of spin? Would the accelerometer readings be meaningless due to the change in orientation of the sensor due to the spin?

Also, would it be possible to determine the spin rate/axis using some form of accelerometer configuration?

Any help would be most appreciated.

Many Thanks.

Andy.

 

[mfb]

With spin and velocity, you have 6 degrees of freedom, so you would need 6 accelerometers (or some which can measure multiple things at the same time). With only 3, the measured result would somehow depend on spin and velocity at the same time, which might give confusing values.

 

[andyfive]

With spin and velocity, you have 6 degrees of freedom, so you would need 6 accelerometers (or some which can measure multiple things at the same time). With only 3, the measured result would somehow depend on spin and velocity at the same time, which might give confusing values.

Thanks for your reply mfb. If I understand correctly, a 3 axis accelerometer would measure the translational acceleration in all 3 axis even though the ball is rotating but not the rotational velocity. How would a second 3 axis accelerometer measure different values to the first one? Would this need to be positioned/oriented differently to measure the rotational velocity?

Many Thanks.

 

[potatoecannon]

Yes, you would need 2 of them at a fixed distance apart within the object.

An easier solution is a 6dof sensor; using 3 accelerometers & 3 gyros to capture all 6 values.

 

[mfb]

If I understand correctly, a 3 axis accelerometer would measure the translational acceleration in all 3 axis even though the ball is rotating but not the rotational velocity.

This is right, but keep in mind that the orientation of the accelerometers would change during rotation, so a constant acceleration in one direction would give variable accelerations for the individual axes.