OHM ScramAttitude for Orbiter 2016 v1.1

[OrbitHangar]

Author: adswnj

MFD to control the SCRAM ascent profile for XR-series spacecraft. This MFD automatically controls the vertical acceleration to trend the Dynamic Pressue to a desired target, and then keep it there. 
 

V1.1 release      Nov 26th 2017

• Moved the tuning options into a ScramAttitude_VesselParams.cfg file
• Renamed Dump.csv to ScramAttitude_Dump.csv for consistency
• Tuned the autopilot response for DG-S, XR1, XR2, XR5
• New scenario files with no external dependencies apart from the respective vessels.

DOWNLOAD

 

[ADSWNJ]

Guys,
Another MFD from the ADSWNJ foundry...

This time I've been playing with precision ascent profiles for XR1, XR2, and XR5 spacecraft in the SCRAM ascent phase. This MFD allows you to do a fully hands-off control of the Dynamic Pressure through the ascent, or a semi-assisted control where you manually select the desired VAcc and the AP just does the elevator trim for you.

See the documentation in the Doc\ScramAttitude folder for full details.

For the technical members: this is an interesting feedback problem, controlling DP indirectly from Elevator Trim. I have implemented a fuzzy logic PD loop, based off a 1994 paper by Robert P Copeland and Kuldip S Rattan from Wright State University, based off an earlier 1976 paper by Macvicar and Whelan. Quite nice that 41 years later, Macvicar's and Whelan's fuzzy control matrix paper is still 100% valid, in a context they could never have imagined!

 

[jgrillo2002]

Nice. will there be a 2010 version for the other guys?

 

[Enjo]

Nice Andrew. This is something that wasn't automated before.
Have you tried if it works together with Launch MFD's "PEG-off" mode? That means DEF action enabled on the first page.

 

[turtle91]

Thats's a really great and helpfull MFD.
I performd a quick and dirty test usng the XR1 with the help of LaunchMFD and ScramAttitudeMFD:

-took off at KSC for a southbound launch to ISS
-I used XR1 pitch-hold AP to reach the desired heading and altitude of about 25 km
-disabled XR-1 AP
-enabled LaunchMFD's AP, but using azimuth-only correction (PTC (or DEF)button, to disable pitch AP)
-enabled ScramAttitudeMFD's full-auto mode using default settings
-enabled the SCRAM engines

At the end, I have had a small RINC error of 0.6 degrees and 5 percent SCRAM-fuel left.
I believe, if I would do more training, I could do better results. But not too bad for a firts test.
It might be better, to kill LaunchMFD's AP BEFORE running out of SCRAM-power btw.

Orbit circulation costed about 1000 m/s. (BurnMFD)
So all in all...a perfect MFD.

There might be other scenarious to use this MFD, i.e. doing high speed/high altitude point-to-pint flights (i.e speedruns) within earth's atmosphere.(using VACC fixed mode)

Many thanks for sharing + the good provided documention as usual.

And a 2010 re-compile would be perfect.....;-)

 

[ADSWNJ]

Thanks guys... always surprised to see peeps still wanting the O2010 versions, but sure, I'll recompile and release for you all!

Nice work Turtle ... good to know it worked as advertised.

@Szymon - I have not tried it with Launch MFD, but looks like Turtle confirms all is good.

Back working on Lagrange graphics with Brian now .

 

[boogabooga]

Thanks guys... always surprised to see peeps still wanting the O2010 versions, but sure, I'll recompile and release for you all!

Don't be. Only 2010 has workinging UMMU, for example.

 

[jgrillo2002]

Don't be. Only 2010 has workinging UMMU, for example.

Not to mention Orbitersound and DGIV and UCGO only work with 2010. Dan hasnt been online for a while. He did reply a month back but didnt reply again.

 

[ADSWNJ]

O2010 version posted. It's a quick recompile and very brief test, so let me know if any issues.

 

[boogabooga]

Thank you.

I really only use XR in Orbiter 2010. The XR series still has a dependence on not-fully-functional UMMU in Orbiter 2016, and I don't need a CTD risk. Also ground contact and sound issues in 2016.

 

[boogabooga]

I'm having some trouble with this.

First, why is there a dependency on Carl Sagan Space Center in the scenarios that come with this?

Second, I ran the XR2 scenario that comes with this, and I burnt up in the atmosphere in a matter of seconds. The autopilot does not appear to be adjusting the elevator trim nearly quickly enough.

 

[ADSWNJ]

First, why is there a dependency on Carl Sagan Space Center in the scenarios that come with this?

Been thinking about this all day . I have to admit that I test on my usual deployment with 20+ addons installed. You are encouraging me to do the right thing, and test my addons on a clean install before release! Noted ... will do from now on (and/or note any addon dependencies to use scenarios).

Second, I ran the XR2 scenario that comes with this, and I burnt up in the atmosphere in a matter of seconds. The autopilot does not appear to be adjusting the elevator trim nearly quickly enough.

That's interesting. Are you in a "bad" state prior to switching on the AP? I.e. nose temps above 1000C, Dynamic Pressure 50+ kPA, -50m/s VSpd? If so, you probably need to intervene more aggressively.

We can tune the response though - either to drive more VACC rate for the extreme situations, or to drive a faster elevator trim response. The latter can lead to wing stress though if we jerk the vessel up too fast. Interesting case though ... let me know if you can reproduce this at will.

 

[boogabooga]

That's interesting. Are you in a "bad" state prior to switching on the AP? I.e. nose temps above 1000C, Dynamic Pressure 50+ kPA, -50m/s VSpd? If so, you probably need to intervene more aggressively.

We can tune the response though - either to drive more VACC rate for the extreme situations, or to drive a faster elevator trim response. The latter can lead to wing stress though if we jerk the vessel up too fast. Interesting case though ... let me know if you can reproduce this at will.

It's the XR2 scenario that comes with this (with the aforementioned CSSC dependency),for Orbiter 2010. Are you sure the add-on works just as well in Orbiter 2010 as it does in Orbiter 2016?

 

[ADSWNJ]

It's the XR2 scenario that comes with this (with the aforementioned CSSC dependency),for Orbiter 2010. Are you sure the add-on works just as well in Orbiter 2010 as it does in Orbiter 2016?

Bug confirmed. Looks like the exact same scenario transferred from Orbiter 2016 behaves very differently in Orbiter 2010. You have about 8 seconds to catch it before you get a hull breach: you need to drop the SCRAM to idle immediately, and get the nose up to at least 100m/s VSpd, and then let the AP do its thing. Even with that, it looks slower to respond than the XR2 on Orbiter 2016.

I'll do some more work to try to figure it out, and I'll release a clean package based just on the vanilla O2010 install.

See ... you do a quick O2010 recompile and release, and see where it gets you Dammit!!

 

[boogabooga]

I don't call myself "Bug Crusher" for nothing. :lol:

If it helps you, I am noticing that the Act VACC is fluctuating wildly. Perhaps there is an issue because the concept of "surface" is different between Orbiter versions (terrain vs. not)?

 

[ADSWNJ]

Well, for that, we have the diag files and some graphing .

Bug crusher ... nice work sir!

 

[ADSWNJ]

V1.1 BoogaBooga release completed!

Details:

V1.1 release      Nov 26th 2017

• Moved the tuning options into a ScramAttitude_VesselParams.cfg file
• Renamed Dump.csv to ScramAttitude_Dump.csv for consistency
• Tuned the autopilot response for DG-S, XR1, XR2, XR5
• New Orbiter-version-specific scenario files with no external dependencies apart from the respective vessels and their dependencies (e.g. UMmu 3.0)

 

[turtle91]

Nice update !
Since you have tuned the MFD for the standard Deltaglider-S, I tried to use this MFD as a "helper" for LaunchMFD's "Direct-Ascent-AP".

I targeted ISS for a southbound launch using LaunchMFD's direct ascent logic/guidance.
My launch-timimng was a bit too early, so I ended up with a coasting-time of about 1800 seconds.

The enire coasting phase was controlled very stable by ScramAttitudeMFD (while Launch-MFD's AP took care about the heading.
(coasting at about 30 km altitude and velocity of about 2200 m/s)

About 10 seconds before Delta-T (LaunchMFD), I switched off SramAttitudeMFD.
So LaunchMFD took over and perfectly placed my DG-S 500 meters below the ISS.
There where only 11 m/s DV different at AP cutoff.

So this MFD seems to be a great help to have very successfull direct ascents.:thumbup:


Btw., just tested the XR2 using a standard ascent.
IMO much better reaction-time of rudder compared to previous version.
Now XR2 works perfect like the XR1 in V 1.0.
All tests done in Orbiter 2016.

And....I must agree with BoogaBooga, for the XR vessels Orbiter 2010 seems to be still the better choice, esp the ground-handling is a pain in Orbiter 2016.
I.e. ground-braking-phase much too long. If tuned the brakes via XR-config file, brakes are ok, but as soon as applied the vessel might looses controll. (esp the XR2)

 

[ADSWNJ]

Thanks for the thorough shake-down of the new version @turtle91. Nice to see it's working in different environments than my tests. The nice thing I saw with my tests was you can focus on manually rolling the vessel to optimize the RInc value, whilst the ScramAttitude adjusts the elevator trim to compensate for the varying lift component. (I assume the fully automated version is LaunchMFD on the azimuth and ScranAttitde on the elevation, as you documented.)