Bill Clawson wrote:
And if it gets lost you just reset the heads to zero and retry, all you lose is access speed as you have to retry the data access. At under 10 feet of altitude this would mean a bad landing for a rocket. One non linear element can be compensated for with a closed loop system, when you get to two non linear elements the best reaction is luck, next best is some hunting time for the system which if it gets to long in duration goes beyond the systems ability to recover. It would appear as wobbly movements of the machine with occasional loss of control.First off, you can always linearize the turning of the valve to engine thrust via a look up table. We do this in disk drives because the position error signal from the disk drives is often-times non-linear in nature. Ideally it would be linear, such that the amount off track is directly proportional to some position error measurement, but this is rarely the case these days due to the nature of disk drive read/write heads, etc.
But you can't have acks and nacks or have the machine force a line low to stop the system. Also the issue of scaling comes in where the more mass you try to move the slower your controls work (with a reasonable energy budget of course).
Also, you can take into account any time delays in
building up thrust for a given change of valve opening
via the overall servo loop. This is pretty common
stuff in the servo world, so I wouldn't be too worried
about that. We get this all the time when dealing
with electronics and physical systems that don't
always respond right away, due to large amounts of
inductance or what have you.
Yes and how far from the center of gravity is the moment arm, the more moment arm the less throttling to effect the same RATE of turn.Lastly, slop in the valve mechanics themselves can be compensated for, if not taken out immediately, by the overall effects of the servo loop.
I would think that in a vectored thrust stearing
configuration you would also want to make sure that
the thrust vector for each engine DOES NOT pass
through the CG of the vehicle, but is off of the CG
such that the change in thrust really does cause the
rocket to change where it is pointing. Otherwise,
there is no point to vectoring your thrust.
Lastly, you are going to have to solve for the effectsI will do more CAR (Coffee Aided Research) in the morning. I think that the complexity of solving these problems is greater than the complexity of adding attitude thruster rockets.
of gravity by increasing the overall thrust in
proportion to the cosine or sine of the angle you have
now oriented your rocket (I leave it up to the readers
to do the trig) and for the change in the total thrust
you have caused across all of the engines due to
reducing the thrust on one rocket. NOTE: The easiest
way to compensate for changing the thrust on one
rocket is to increase the thrust of the opposite
engine by the same amount. If this does not cover
your rocket configuration, then the overall math is
only slightly more difficult.
My order of suitability for manned flight;
1 Gimbaled main engine with attitude thrusters 2 Vanes on main engine with attitude thrusters 3 multiple independently gimbaled thrust engines 4 multiple independently vaned thrust engines 5 multiple variable thrust
Of course 1 and 2 are multiple rocket machine with the 90 degree placement of the attitude rockets allowing for their size to be greatly reduced. if rockets were shaped like pancakes this of course wouldn't be true as you could not get a good moment arm. Rockets are shaped more like pencils for aerodynamic reasons so you can get further away from the center of mass going fore and aft, hence the smaller thrusters needed.
_______________________________________________ ERPS-list mailing list [EMAIL PROTECTED] http://lists.erps.org/mailman/listinfo/erps-list
--
----<<<<<<<<<<<<<<<<<
........ Alex Fraser N3DER ......... ......... [EMAIL PROTECTED] ....... [~]_>^</\-[~]_>^</\-[~]_>^</\-[~]_>^<
_______________________________________________ ERPS-list mailing list [EMAIL PROTECTED] http://lists.erps.org/mailman/listinfo/erps-list
