At 10:45 AM 5/25/2004 -0600, you wrote:
John Carmack,
I was wandering if you folks had any concerns about the deep
enclosure around the jet vanes? It seems to me it may cause
some problems with the gas dynamics for controlling the thrust
vectoring. It seems that as the vanes deflect a significant amount
the gases are going to impinge on the side of the box and basically
have the steering effects basically canceled out. Also it would
seem that at extreme vane deflection the back pressure behind and
to the side of the vane and the box could be the cause of the
stress failures of the shaft and vane that might not of other wise
have happened in such a constricted volume.
The vanes should only tip +/- 15 degrees or so, and we aren't fully
expanded in the nozzle, so the jet isn't as wide as it looks.
Have you considered instead of the shielding box, instead covering
the bearing and servo motor with a low profile as possible high
temperature flexible thermal insulation blanket, perhaps made of
fiberglass or nominex? Also perhaps the servos could be moved
further up and away from the exhaust backwash by coupling them
the the vane shaft by a universal joint and thus moving the servos
up and way, say up to 60 degrees.
Blankets would tend to get soaked in propellant during the warmup, and any
edges that can start flapping near a supersonic exhaust will fray
rapidly. We like having solid metal surfaces down there.
We want to avoid adding linkages if possible.
I know that the system is now an experimental steeping stone to
your next vehicle, and thus your trying to keep it as simple as
possible. The landing gear looks to have a very narrow foot print
and only a modest amount of sideways motion on touchdown
would result in the vehicle toppling over. Have you considered for
at least the first free flights putting 'training' struts out on the
landing structure?
We have the GPS position hold working now, so we expect to set down with
almost no sideways velocity. Also, the CG is quite low, so it isn't as
tippy as it looks.
A wider stance would be nice, but I am not going to do any kind of
deployable landing gear if I can at all avoid it. We could spread the
base by about a foot each way before we ran into problems with
transportation, but then we would lose the option of just fairing the
entire base in for better aerodynamics.
I know its probably going to be a while but when you plan on having
the vehicle going to higher speeds and altitude, have you
considered putting a angled down facing camera with a RF link?
The idea is that if the vehicle encounters control problems or
unexpected drift and possibly has to land out of direct view of the
ground control station, such as having to land it on the other side of
a hill or tree. This would make possible last minute adjustment to
avoid undesirable landing locations that the auto land code would
have no way of knowing about. Also along those lines any
thoughts about some sort of close range terrain sensors such as a
IR or ultra sonic ranger. I know you have a GPS system but it
can't take into account the odd little topography issues if you have
to land further away from your takeoff point.
Actually, we may be doing higher and faster flights next weekend. As soon
as we prove out the basic jet vane control, everything is set for boosted
hops. We can't go very high with a 15 second burn time limit, but it will
still do some accelerating and deceleration.
If the vehicle is landing out of sight, we have had a failure of a nature
that likely means it isn't going to "land" at all.
Ultrasonic rangefinders don't work at all next to a rocket engine. We
tried that very early on in our development process. We used a laser
rangefinder on our manned lander, which worked ok, but I have concerns
about kicked up dust confusing it. There is a decent little radar
rangefinder from Roke that I have been considering buying as a backup for
the GPS, but I haven't bit the bullet on it yet. It is only 10hz update,
so you can't get a decent velocity signal from it, but it should be able
to coast with the inertial system.
John Carmack
_______________________________________________
ERPS-list mailing list
[EMAIL PROTECTED]
http://lists.erps.org/mailman/listinfo/erps-list
