NavList:

   

Reply

George Huxtable wrote:
>
> Three axes,
> fixed in space could be- Toward the North celestial pole, with Dec = 90:
> toward Aries (in the Earth's equatorial plane), with Dec = 0,
> Right-Ascension = 0: and (also in that equatorial plane) 90 degrees from
> Aries, with Dec = 0, R.A. = 90. These axes remain (nearly) unchanging, in
> space.

That's correct, and such an orientation might be used in space
applications. I'm not sure; all my experience has been with aircraft
INS. As far as I have ever seen (except for strapdown systems), these
keep the gyro and accelerometer platform horizontal. In the horizontal
plane the platform may be oriented to true north. Other systems don't
attempt to stabilize the platform about the vertical axis. Instead, they
keep book on the "wander angle": the difference between true north and
the platform azimuth. Then there's the Litton Carousel INS, which keeps
the platform horizontal but constantly rotating at 1 rpm.

In all these systems, the initial azimuth orientation is precisely
established during INS alignment, which is necessary every time the
system is turned on. Alignment is almost completely automatic; the
hardest part for the operator is entering latitude, longitude, and
altitude. At military bases the parking spots have geodetic survey marks
embedded in the concrete to facilitate this. (Some aircraft with
integrated avionics have an "align to GPS" option to eliminate manual
coordinate entry if the GPS is running.)

After the alignment is started there isn't anything to do but monitor
its progress and wait. This can take 20 minutes or more. A full
alignment on a B-1 bomber takes 90 minutes, though you can terminate
early and accept reduced accuracy.

Some military aircraft systems let you "cock" the INS by doing a ground
alignment, then shutting down. If you don't move the plane, the system
remembers its orientation and location, making INS available for a
scramble take-off without a lengthy (possibly fatal) delay for
alignment. The stored alignment isn't as good as a normal alignment, though.


> Although inertial navigation systems always show some drift, as a result of
> having to integrate (twice) the accelerations, which always have some
> zero-error, I doubt (from my inexpert perspective) whether any of that drift
> derives from the pseudo-gyros, which sense the orientation.

Actually, I've read that gyros have always been the limiting
factor in INS performance. The B-1 navigator's manual says, "The
navigation accuracy of the INS inertial platform is most dependent on
gyro performance. Accelerometer accuracy is less critical unless major
malfunctions occur."

The B-52 SPN/GEANS INS had the most interesting gyro I've encountered.
The rotor was a hollow beryllium sphere roughly the size of a golf ball.
It rotated at about 650 revolutions per second, suspended by
electrostatic force in a spherical vacuum chamber. The gap between
sphere and wall was on the order of .002 inch, as I recall. Spin-up
torque was applied only at startup. After that the ball coasted in
vacuum with almost zero friction. To power down, the system braked the
sphere, then shut off the suspension voltage (the electrodes were on the
inner wall of the chamber). A power interruption would crash the sphere,
so a "rotor support battery" ensured backup power long enough for a
proper shutdown.

Since there was no contact between the rotor and its envelope, an
optical sensor observed a mark on the ball's "north pole" to keep the
INS platform aligned with the spin axis.

A couple of these spheres are depicted here:
http://periodictable.com/Elements/004/index.html

--
I block messages that contain attachments or HTML.

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to NavList@fer3.com
To , send email to NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---

   

Reply