Meadville Space Center
Welcome, Guest. Please login or register.
May 18, 2013, 02:14:09 AM

Login with username, password and session length
Search:     Advanced search
20010 Posts in 1929 Topics by 2250 Members
Latest Member: mikegk
* Home Help Search Login Register
+  Meadville Space Center
|-+  Project Apollo - NASSP
| |-+  Project Apollo - NASSP News & Discussion (Moderators: movieman, Tschachim, Swatch, lassombra)
| | |-+  LVDC++ Software
« previous next »
Pages: 1 ... 5 6 [7] 8 9 ... 18 Print
Author Topic: LVDC++ Software  (Read 11765 times)
meik84
Project Team Member
Sr. Member
****
Posts: 299



View Profile
« Reply #90 on: October 29, 2012, 11:42:49 AM »

It's obviously July, 1, 1968, 00:00 UT; all other AGC routines use that time.
I've tried those rotations; seems to work, my velocity vecor at GRR and those of AP 8 postflight trajectory agree. The only odd thing is that my X-position is good, but my Y- and Z- Positions are almost 0, whereas postflight trajectory calls for +17 km in Y and -5.5 km in Z. Confused
Logged
eddievhfan1984
Full Member
***
Posts: 231



View Profile
« Reply #91 on: October 29, 2012, 07:42:31 PM »

Well, it IS only a simulation. Is it possible you've coded better than a team of IBM's finest? LOL In any case, Orbiter is not a perfect model of the universe, and we don't know how close to the accuracy (or inaccuracy) of the LVDC coding we can get, as I don't the Smithsonian is gonna let us break into an IU and do a core dump. Tongue
Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 831


View Profile
« Reply #92 on: October 29, 2012, 11:23:14 PM »

I actually asked them about that.
Logged
eddievhfan1984
Full Member
***
Posts: 231



View Profile
« Reply #93 on: October 30, 2012, 12:04:00 AM »

*thunderous applause* Good show, sir! What did they say?
Logged
kerlix
Full Member
***
Posts: 83


View Profile Email
« Reply #94 on: October 30, 2012, 03:17:16 AM »

Sorry to say, I tried going that (the "core dump") route several years ago. Basically, I was told I was crazy and it was never gonna happen. The response I got swayed somewhere between "You're a whack job if you think we'd let you touch that" and "We're not letting you anywhere close to anything that could be remotely considered missile guidance software because you might be a terrorist". Sadly, this is 21st century America. Everyone is suspect. Imbeciles (IMO).

However, I wish you the best of luck. Perhaps you can compose a better request or speak to someone less paranoid.
Logged
meik84
Project Team Member
Sr. Member
****
Posts: 299



View Profile
« Reply #95 on: October 30, 2012, 05:43:18 AM »

Quote
and we don't know how close to the accuracy (or inaccuracy) of the LVDC coding we can get
3 sigma boundaries were:
+/- 1.38 m/s for V_t
+/- 0.01 ° for Inc
+/- 0.00032 for Ecc
In reality, those boundaries were never violated. AP 8 hit its orbit with -0.01 m/s V_t, +0.006° Inc and 0.00001 Ecc (!).
I am able to stay inside of the boundaries for V_t and Inc, only Ecc is the problem.
Quote
as I don't the Smithsonian is gonna let us break into an IU and do a core dump.
Oh, I wouldn't need that. One of the infamous "equation defining documents" would be enough. Unfortunately, there's none in the net. But I'm sure that at least one copy must have survived -almost every evaluation doc relates to it, whichs prompts that it was a widely spread document.
Quote
"We're not letting you anywhere close to anything that could be remotely considered missile guidance software because you might be a terrorist"
Idiotic. Every student of aerospace technology learns at least the basic laws of rocket guidance, both atmospheric and space; if he wants to learn more, he just has to ask the resp. professor. The only thing he'd need more for a guidance program (which would be better then our LVDC) are 2 or 3 computer geeks and a 24 h pizza service. Software is not the problem in rocket science; it's metallurgy, statics and kymatology.
Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 831


View Profile
« Reply #96 on: October 30, 2012, 07:45:54 AM »

What did they say?
Sadly, their IU didn't have an LVDC. They have a different LVDC that they got from another institution but it was stored badly and was basically destroyed - it was badly corroded, and many of the modules were missing. They don't believe the core was ever loaded because the log sheets that came with it indicated it was strictly non-flight. In any event, it's not in readable condition. They may restore it to display condition and display it near the IU, but it wouldn't be able to be powered up.
Logged
eddievhfan1984
Full Member
***
Posts: 231



View Profile
« Reply #97 on: October 30, 2012, 03:25:35 PM »

I second 24-hour pizza. LOL So we're not necessarily trying to match historical orbital parameters so much as we're just trying to achieve historical guidance accuracy? And how is the LVDC supposed to maneuver in TLI; will we still have to program a flight path via LTMFD/IMFD, or will the 9th-order ephemeris polynomial in the AGC/LVDC take care of that?
Logged
meik84
Project Team Member
Sr. Member
****
Posts: 299



View Profile
« Reply #98 on: October 30, 2012, 03:57:04 PM »

Quote
So we're not necessarily trying to match historical orbital parameters so much as we're just trying to achieve historical guidance accuracy?
Doing the one is achieving the other. LVDC always aimed for a 100 x 100 nm orbit. So the better we get, the more historical accurate we get. I've pressed down my ecc to 0.0009 by the way, and I still believe there are some screws to turn...
Quote
And how is the LVDC supposed to maneuver in TLI; will we still have to program a flight path via LTMFD/IMFD, or will the 9th-order ephemeris polynomial in the AGC/LVDC take care of that?
Uuuuuuh. That will come once I get this whole thing running in orbit and transplanting it to the SIB for AP7. There's still so much to do before I think about that; starts with making it able to save the scenario.
Logged
meik84
Project Team Member
Sr. Member
****
Posts: 299



View Profile
« Reply #99 on: November 04, 2012, 08:56:21 AM »

Well, people...
I've uploaded what I would call my 'provisional results' regarding LVDC++ to CVS. The stupid ecc problem remains and I believe that the code really needs some 'clean up'. Nevertheless -it works, so I don't see any reason to keep it back.
I've attached a little .scn for testing. Starts some seconds before GRR, all systems up and running. For those who want to let Checklist MFD take care of the systems: the standard checklist applies to a SIB launch; checklist MFD will throw some switches too early (the TWR JETT sw especially). However, LVDC++ doesn't bother about that, just like it should.
Oh, and one thing: I've changed the keyboard <-> DSKY assignments a little.
Keypad Numbers,+,- are DSKY numbers,+,-
Keypad Enter is ENTR
Keypad Decimal is CLR (I really missed that key! Wink)
Keypad / is VERB
Keypad * is NOUN
End is PRO (as it ends a display)
Home is KEYREL (as it 'goes home' from a display you've called)
Pg Up is RSET
Note that you've got to press Shift with the resp. key and that Num Lock has to be OFF for the keys on the Numpad.
If somebody's bothered by this, you may scream. I found this assignment most convenient.

* Apollo 8 - LVDC.scn (84.24 KB - downloaded 18 times.)
Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 831


View Profile
« Reply #100 on: November 04, 2012, 11:22:00 AM »

Worked as expected here. Good job!

I probably can't help with the eccentricity because I still don't understand the math, but I can do scenario save/load for you if you want.

Logged
Tschachim
Project Apollo - NASSP
Administrator
Hero Member
****
Posts: 3687


nassp.sf.net


View Profile WWW
« Reply #101 on: November 04, 2012, 01:12:37 PM »

Yeah, great job, a very smooth ride!  Thumbs Up

I'm at a complete loss with the code, so I fear I can't help much, too. I know that the S-IVB cutoff needs to be very precise in order to get the orbit right, but if the cutoff velocity is fine, it might be a steering problem, I don't know. Stupid question, how is this "vis-viva energy" thingy defined? Another question as the code is very hard to read, the cutoff inclination is very good, 32.51°. I'm very happy about that, the old autopilot wasn't able to do that. But I found this line in the code:

// Let's cheat a little. (Apollo 11)
Inclination = 32.5574;

... so this setting isn't used then?

Thanks!
Tschachim

PS: You didn't commit the "FDAI reset" change, did you?
Logged

meik84
Project Team Member
Sr. Member
****
Posts: 299



View Profile
« Reply #102 on: November 04, 2012, 01:39:42 PM »

Quote
Stupid question, how is this "vis-viva energy" thingy defined?
Uuuh. As far as I understand, this is some sort of energy which a rocket in the resp. orbit has. Don't ask me more about that, please. That's astronomy, and I'm just a guy who believes to understand an old piece of computer tech...
Quote
// Let's cheat a little. (Apollo 11)
Inclination = 32.5574;

... so this setting isn't used then?
Optimum Inc and other parameters of the target orbit are time dependend. The real LVDC computed them by itself when it received the GRR interrupt from the ground. We lack some parameters for this equations, but we have the parameters of AP 8's orbit. So I 'cheated' a little and inserted those values directly.
Quote
PS: You didn't commit the "FDAI reset" change, did you?
When the error needles go back to 0 shortly before SIC/SII staging, I have.
Logged
Tschachim
Project Apollo - NASSP
Administrator
Hero Member
****
Posts: 3687


nassp.sf.net


View Profile WWW
« Reply #103 on: November 04, 2012, 02:22:11 PM »

As far as I understand, this is some sort of energy which a rocket in the resp. orbit has.

Ah, it's googlable, really just the total enery in orbit: http://en.wikipedia.org/wiki/Vis-viva_equation

Optimum Inc and other parameters of the target orbit are time dependend. The real LVDC computed them by itself when it received the GRR interrupt from the ground. We lack some parameters for this equations, but we have the parameters of AP 8's orbit. So I 'cheated' a little and inserted those values directly.

I'm not sure if I understand you correctly, you entered 32.5574 and the LVDC++ magically achieved the correct inclination of 32.51?  Confused

When the error needles go back to 0 shortly before SIC/SII staging, I have.

Well, you didn't commit CSMcomputer.cpp, so I assume not, or did you do it in another way somewhere in the LVDC++ code?

Cheers
Tschachim
Logged

meik84
Project Team Member
Sr. Member
****
Posts: 299



View Profile
« Reply #104 on: November 04, 2012, 02:50:04 PM »

Quote
I'm not sure if I understand you correctly, you entered 32.5574 and the LVDC++ magically achieved the correct inclination of 32.51?
That's no magic. LVDC uses Inc and some other precomputed (or, in this case, preset) values in its computations. It's (theoretically) possible to feed it with any orbit you want. Of course, those values have to be reasonable, e.g. you can't set Azimuth to 180° and Inc to 20°. But apart from that, there are no limits.
And no, that's not 'correct'. As I said before, the real LVDC was a lot more precise.
Quote
Well, you didn't commit CSMcomputer.cpp, so I assume not, or did you do it in another way somewhere in the LVDC++ code?
Ah, I see. CVS wants me to 'solve the conflicts' for CSMcomputer.cpp. What conflicts? Confused
Logged
Pages: 1 ... 5 6 [7] 8 9 ... 18 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!