Meadville Space Center
Welcome, Guest. Please login or register.
July 04, 2020, 01:45:57 PM

Login with username, password and session length
Search:     Advanced search
Project Apollo - NASSP 6.4.3 released!
http://nassp.sf.net
25068 Posts in 2094 Topics by 2266 Members
Latest Member: twa517
* Home Help Search Login Register
+  Meadville Space Center
|-+  Project Apollo - NASSP
| |-+  Project Apollo - NASSP Development
| | |-+  Programming (Moderators: movieman, dseagrav, Swatch, lassombra)
| | | |-+  AGC reentry autopilot report
« previous next »
Pages: [1] 2 Print
Author Topic: AGC reentry autopilot report  (Read 5427 times)
chode
Project Team Member
Full Member
****
Posts: 153


View Profile
« on: April 11, 2005, 10:24:20 PM »

I've been working on an AGC reentry autopilot with some success.

I found documentation on the guidance equations and have implemented a re-entry guidance based on these that seems to work pretty well.

One feature I want to support is for the pilot to just enter "Verb 37 Noun 61 Enter", proceed (PRO) through all of the displays, and the autopilot should bring them down, as long as they are in a good "entry corridor".

Here's how I think it should work (please feel free to comment or correct any errors):


Pilot: Verb 37 Noun 61 Enter


--------------Start P61:------------------------------

display: Flashing V06N61

     XXX.XX deg (Splashdown Latitude, + is North)
     XXX.XX deg (Splashdown Longitude, + is East)
     -00001 (RollAttitudeCode, -1 = headsdown)

--- These values are calculated based on the entry point of the current trajectory
--- and assume a nominal reentry profile.

Pilot: Enter new splashdown coordinates (using V21, V22 e.g.) or PRO to accept

display: Flashing V06N60

     XXX.XX Gs  (estimated maximum G force)
     XXXXX  m/s (predicted inertial velocity at 400k feet)
     XXX.XX deg (entry angle at 400k feet, should be 6.5 degrees for lunar return)

Pilot: PRO

display: Flashing V16N63

     XXXX.X km (estimated range, 0.05Gs to splashdown)
     XXXXX  m/s (predicted inertial velocity at 0.05Gs)
     XX XX  min, sec (time to 0.05Gs)

Pilot: PRO (advances to P62)

-------------Start P62:-------------------------------

display: Flashing V50N25

+00041 (Checklist Code 41: maneuver to CM/SM sep attitude and sep)

Pilot: PRO

display: Flashing V16N61

     XXX.XX deg (Splashdown Latitude, + is North)
     XXX.XX deg (Splashdown Longitude, + is East)
     -00001 (RollAttitudeCode, -1 = heads-down)

--- These values are either the originally calculated values, or values entered by the user
--- in P61. This is another opportunity to change the values.

Pilot: Enter new splashdown coordinates (using V21, V22 e.g.) or PRO to accept
(PRO advances automatically to P63)


-------------Start P63:-------------------------------
--- P63 autopilots the CM to a proper reentry attitude

display: V16N64

     XXX.XX g (G force)
     XXXXX  m/s (Inertial Velocity)
     -XXXX.X km (range to splashdown)

(when g force is greater than 0.05G, automatically starts P64 for lunar trajectory, or P67 for orbital)


-------------Start P64:-------------------------------
--- P64 autopilots the initial (constant drag) part of the entry.
--- Skipped for orbital entry.

display: V16N74

     XXX.XX deg (commanded bank angle)
     XXXXX  m/s (Inertial Velocity)
     -XXXX.X km (range to splashdown)

(P67 is automatically started when velocity is low enough)


-------------Start P67:-------------------------------
--- P67 autopilots the final phase of the entry.

display: V16N66

     XXX.XX deg (commanded bank angle)
     XXXX.X km (est. cross-range error)
     XXXX.X km (est. down-range error)

--- When V < 300m/s, autopilot is dis-engaged, display changes:

display: Flashing V16N67

     XXXX.X km (range from target)
     XXX.XX deg (present Latitude, + is North)
     XXX.XX deg (present Longitude, + is East)

___________________________________________END
Logged
movieman
Moderator
Hero Member
****
Posts: 1710



View Profile
« Reply #1 on: April 12, 2005, 05:41:56 AM »

Looks good. AFAIR there's also another section in the code for the re-entry program where it will add a 'skip' phase to the reentry if the angle or velocity are out of safe range... so you come in, brake a bit, skip out and come back down on a safer trajectory.

Presumably it will also need to check whether the splashdown site is achievable and raise some error alarm if it's not.
Logged
chode
Project Team Member
Full Member
****
Posts: 153


View Profile
« Reply #2 on: April 13, 2005, 03:08:13 AM »

Thanks for the feedback movieman.

So far, the only part of my autopilot that actually uses the real guidance equations is the final phase. So, I can make pinpoint landings if the entry angle and velocity are "nominal", and I don't change the predicted splashdown point much.

But, the real autopilot was more robust than that. As you have pointed out, there were "skip-out" programs (P65 and 66) that permitted quite a large departure from the "nominal entry", particularly in the ability to greatly extend the range from the entry point to spashdown. I've seen graphics that imply that the splashdown point could be extended by more than 2000 km with this technique.

I'm not sure, but I've found no evidence that P65 or P66 were ever used in a real mission. This may be just because there was never a reason to move the splashdown point by more than a hundred kms in the last day before reentry.

In any case, I've taken a look at the flowcharts and equations for the rest of the reentry guidance (before the final phase), and I think I have enough information to implement a reentry autopilot that is very true to the real one in every way.
Logged
movieman
Moderator
Hero Member
****
Posts: 1710



View Profile
« Reply #3 on: April 13, 2005, 05:16:18 AM »

Quote
I'm not sure, but I've found no evidence that P65 or P66 were ever used in a real mission.


As far as I know, they never were: they were more of an emergency thing, in any normal mission they'd set up the the TEI burn to get them to the correct entry position and velocity.

But I suspect that Orbiter pilots are far more likely to need them than NASA astronauts Happy.
Logged
chode
Project Team Member
Full Member
****
Posts: 153


View Profile
« Reply #4 on: April 16, 2005, 11:19:45 PM »

I did some more digging and I found out that P65 and P66 were indeed used in the Apollo 6 mission. There were a number of things that went wrong on that mission, and they came in at a shallower angle and with less velocity than originally planned.

Also, in doing some simulated autopilot skip-outs, I can see why they probably didn't want to use it. If you don't have very good lateral alignment with your target at skip-out, you can end up way to the left or right by the time you come back into the atmosphere, too far to steer back into alignment during the final phase.
Logged
movieman
Moderator
Hero Member
****
Posts: 1710



View Profile
« Reply #5 on: April 19, 2005, 10:29:32 AM »

That would make sense: the current simulation of Apollo 6 is a bit cheap-and-cheerful, skipping many of the failures that happened on the real flight.
Logged
chode
Project Team Member
Full Member
****
Posts: 153


View Profile
« Reply #6 on: April 22, 2005, 06:37:11 PM »

I've got it mostly working,  even the skip-out programs P65 and P66.

I'm not sure what I should do if the pilot enters new splashdown coordinates that are outside the "footprint". My idea right now is to pick the spot at the edge of the footprint that is closest to what was entered. Or maybe I should just light up the op err light. Any ideas?
Logged
movieman
Moderator
Hero Member
****
Posts: 1710



View Profile
« Reply #7 on: April 22, 2005, 06:47:53 PM »

Best thing would be to poke through the source code for the real programs and see what that would do Happy.
Logged
chode
Project Team Member
Full Member
****
Posts: 153


View Profile
« Reply #8 on: April 25, 2005, 12:12:33 AM »

Well,  it looks like the AGC just accepted any entered coordinates, assuming that any newly entered landing coordinates would be fully "checked out" by the ground control.

I think that we have a major departure from "reality" here. In Orbiter, we do not have the entire Houston Apollo ground team feeding us the best coordinates, so instead, I'm just looking for a "simple" solution that will get the spacecraft down to Earth.
Logged
movieman
Moderator
Hero Member
****
Posts: 1710



View Profile
« Reply #9 on: July 12, 2005, 07:49:45 AM »

Did this ever get anywhere? With the new LEM AGC programs this is one of the last major AGC software upgrades we need Happy.
Logged
chode
Project Team Member
Full Member
****
Posts: 153


View Profile
« Reply #10 on: July 12, 2005, 08:16:01 PM »

Things got real busy for me just as I was trying to finish it up.

I should have more time in the next two weeks, and will try to submit something soon.
Logged
movieman
Moderator
Hero Member
****
Posts: 1710



View Profile
« Reply #11 on: July 13, 2005, 04:49:41 AM »

Yay! Thanks!

Edit: I guess this means we'll soon need some aircraft carriers for orbinauts to try to land near Happy.
Logged
movieman
Moderator
Hero Member
****
Posts: 1710



View Profile
« Reply #12 on: August 21, 2005, 12:35:42 PM »

Just a reminder Happy.
Logged
Christophe
Project Team Member
Hero Member
****
Posts: 1072


View Profile Email
« Reply #13 on: August 28, 2005, 09:01:49 AM »

Very Happy As usual many thanks to all guys for your great job.
Chode, I would just want to point out a slight problem I discovered a few days ago when training for TEI, TransEarthCoast and reentry: everytime I wanted to enter atmosphere with nominal angle(6.5) with the standard Apollo procedure, that means "full lift" (heads down with 0 roll), I experienced about 7 or 8 G's during the first part of reentry. According to reals reentry data available in "Apollo by the numbers"doc, average deceleration was around 6 and an half G's. After many trys I found that, in Orbiter, the right angle for reentry to match the propers values(not only G forces but also range from target) is 6.4 instead of 6.5. I think the reason is due to the orbiters simulation of the high earth atmosphere that is not totally accurate. So be carefull with that because it can make the development of the autopilot very hard. I suggest that in orbiter we can use a corridor of 6.4 wich is finally very close to the real one and match an "Apollo style" reentry.

With best regards.
Logged
Coussini
Project Team Member
Hero Member
****
Posts: 699


View Profile
« Reply #14 on: September 12, 2005, 11:40:42 AM »

Quote from: chode
I've been working on an AGC reentry autopilot with some success.

.....

Pilot: Verb 37 Noun 61 Enter


--------------Start P61:------------------------------

display: Flashing V06N61

     XXX.XX deg (Splashdown Latitude, + is North)
     XXX.XX deg (Splashdown Longitude, + is East)
     -00001 (RollAttitudeCode, -1 = headsdown)
....


Do you have a scenario to test this  Confused:  
Is it ready to use  Confused:

Thanks a lot

 Happy
Logged

Have a nice day
Pages: [1] 2 Print 
« previous next »
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.10 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!