NavList:

   

Reply

At 11:42 AM 2/1/02, Jared Sherman wrote:

>  will find it inextricably confused with the other transient accelerations
>that rough weather creates.>


>  I confess to having no experience designing inertial navigation systems,
> for aircraft, submarines, or flying cars. (They did promise us flying cars!)
>  I don't see why a device measuring acceleration due to gravity could not
> be used to filter out--and thus remove--the effects of rough weather. It
> would seem that if the "sea state gravity" effects can be registered on
> the accelerometer chips, they can be subtracted from the overall effects
> in the fluxgate sensors, producing a net fluxgate reading without the
> rough weather effects.
>  What am I not seeing here?


Couple of things, one of which you blame on Einstein.
A three axis fluxgate can evaluate the direction and magnitude of
the  geomagnetic field. But if you rotate the flux gate using the mag field
axis as the axis of rotation, the fluxgate can't tell any change has happened!
So there has to be some other reference direction. A north star sensor, a
horizon sensor - a geoelectric field sensor ( or rather more practically) a
gravity sensor.
And that's where Einstein comes in: Einstein showed that the gravity force
due to mass, and the inertial force due to acceleration can not be
distinguished (in a reasonably small
volume). You remember that gymbals are essentially pendulums, that slosh about.

Imagine an almost massless perfectly damped pendulum. It could not tell you
which way is down, if you shove it sidewards. It sees the gravity force and
its sees the inertial force as it accelerates sidewards.
So the next step in sphistication is a very overdamped pendulum of any kind.
After all the rocks and shoves, it will point on average down. But to
capture the fast transient movements, you need something that acts like a
mechanical gyro whose spin axis is fixed in space. And to take care of
rotations in the gyro's axis, you need at least one other, perpendicular to
it.This drifts in direction (in one pointing direction, that would be 15
deg sin latitude for reasons you know well.)
The gyro has to have a slow correction in the pendulum's pointing
direction, and then the  gyro can tell you about transient movements away
from a reference direction. The field direction from the flux gate,
processed through a spherical correction provided by the gyros angular
change w.r.t the heaving 3 axis fluxgate and you have the promise of a
solid mag bearing indication through the stiffest blow.   But gyros are
notoriously fickle. And a strap down device can be expensive.    fiber
optic cable to wind into a ring of hundreds of turns is not too
expensive.   And laser LEDS are not too expensive. But the market is
really, really small. At present.



Brian Whatcott
   Altus OK                      Eureka!

   

Reply