[1.3.1] Ferram Aerospace Research: v0.15.9.1 "Liepmann" 4/2/18

[ferram4]

If your rockets were only slightly stable, they might be unstable due to a change I made in the attach node drag code. Since the drag of pretty much everything has been lessened this update, the need for artificial stabilization at the back of rockets was removed, and a large number of designs that I have been able to launch don't have stability issues, and I'm using stock and KW for these tests. This was partially motivated by the fact that pancake rockets were made much more stable than they should have been with that setup, so I changed things to punish flat, wide rockets even further.

All of that said, without pictures I don't know what you're doing wrong. All of my launches have been perfectly fine, so I need more information (and pictures are great for that) so I can explain what you're doing wrong or to figure out if it's a bug.

[Arrowmaster]

Something is broken when I try to use fairings. I've tried both KW and pFairings. In the VAB all the parts inside the fairing correctly show isShielded: true but once on the pad everything says false. I did two test launches with some cargo that wasnt very aerodynamic. Using MJ (terminal velocity limiter off) to get an 80km orbit. Both launches used 4365m/s +-5 of dV even though I ejected the fairings before launch on one and around 45km on the other. I first started looking into this when I noticed my terminal velocity was under 200m/s until past 3km. Its usually way higher than that.

[ferram4]

Are you using the most recent version of FAR? What other mods are you using? Can you post a copy of the output_log.txt?

It sounds like FAR is breaking and isn't actually applying its drag model and you're stuck with the stock model for some reason.

[Arrowmaster]

Let me restart KSP and get you a cleaner log. "NullReferenceException: Object reference not set to an instance of an object" is being spammed into the log constantly during launch so your theory sounds right. If it matters I'm on Linux but v0.13x1 worked fine back in 0.23.0.

[ferram4]

Are you using v0.13 (not v0.13x1)? Anything before the new official release might have issues.

Can you see if there's a version of the log floating around that has a stack trace in it? I know where that log goes for windows, but I don't know where it goes on Linux (if a stack trace even gets sent to a log in Linux). Without the trace I can't really do anything with it, because anything could be nulling; it might not even be FAR causing the issues.

[Drew Kerman]

I have been thinking of setting FAR up so that the Flight Data window (in flight) and the Editor Analysis stuff can dump information to a text file, would that work? I was thinking more for debugging purposes, but I think that would work for what you're thinking of.

This would be awesome, because hopefully the guy working on the Ascent Komputron could take the FAR data and use it as input for his model, so instead of predicting a rocket performance based on the .craft file we have the rocket run a scripted test flight (with kOS or something - or just straight up with SAS on if that's all that's needed) and the Komputron can use that data to extrapolate good ascent profiles. He said he's considered doing this, but I don't know what data he would like to see

[Arrowmaster]

Yes I'm using v0.13 now. KSP.log is where I found "NullReferenceException: Object reference not set to an instance of an object". But I just pulled this from the unity log in ~/.config/unity3d/Squad/Kerbal Space Program/Player.log

NullReferenceException: Object reference not set to an instance of an object
  at ferram4.FARBasicDragModel.DragModel (Vector3 local_velocity, Single M) [0x00000] in <filename unknown>:0 
  at ferram4.FARBasicDragModel.RunDragCalculation (Vector3 velocity, Single MachNumber, Single rho) [0x00000] in <filename unknown>:0 
  at ferram4.FARBasicDragModel.FixedUpdate () [0x00000] in <filename unknown>:0 

[ferram4]

Alright, so what other mods are you using, and what did you do to cause the issue? What kind of vehicle did you use to cause this (pictures please)? That is the earliest exception to occur in the log, correct (other exceptions, no matter what they're caused by, can cause issues downstream)?

[sidfu]

ferram im having issues to but think i might know what it is. i took and tested all the fairings they work fine but if just 1 part inside the fairing fails to have a far modle and it gets a error the fairings lose any far modle they had and other parts of the ship do also.

[LOG 21:45:05.909] KW Rocketry SA-1 LFT: FAR drag model added; Size: (1.3, 1.3, 1.3), LD (1.3, 1.3, 0.0), UD (1.3, 1.3, 0.0)

Surface area: 5.086428

Fineness Ratio: 1.036193

TaperRatio: 0.9999996

Cross Sectional Area: 1.227193

Cross Sectional Tapered Area: 9.362676E-07

Major-minor axis ratio: 0.9999984

Centroid: (0.0, 0.0, 0.0)
[EXC 21:45:05.918] NullReferenceException: Object reference not set to an instance of an object
[LOG 21:45:06.014] ATV solar panel: FAR drag model added; Size: (0.2, 1.4, 1.0), LD (0.1, 1.0, 0.0), UD (0.2, 1.0, 0.0)

Surface area: 2.566386

Fineness Ratio: 2.482228

TaperRatio: 1.047861

Cross Sectional Area: 0.1269366

Cross Sectional Tapered Area: 0.02357304

Major-minor axis ratio: 0.1616187

Centroid: (0.1, -0.6, 0.0)

this is the error i got ima try to narrow it down to a specific part maybe

[ferram4]

I need it from the output_log.txt; I need to have the stack trace or I can't figure out what actually went wrong, only that something went wrong.

[Arrowmaster]

I tracked the issue down to the AIES Galaxy-VR2 engine. I was using as the last stage engine of my payload because it was the smallest engine I had unlocked in career but when I moved the ship to a sandbox save and swaped it for stock Rockomax 48-7S it fixed everything. My terminal velocity was even at 700m/s a few hundred meters up.

AIES Galaxy-VR2: Choosing up axis for FAR drag model.

(Filename: /BuildAgent/work/d3d49558e4d408f4/artifacts/LinuxStandalonePlayerGenerated/UnityEngineDebug.cpp Line: 53)

NullReferenceException: Object reference not set to an instance of an object
  at ferram4.FARGeoUtil.PartMaxBoundaries (.Part p, Vector3 worldOffset, Matrix4x4 partUpMatrix, UnityEngine.Transform[] ModelTransforms) [0x00000] in <filename unknown>:0 
  at ferram4.FARGeoUtil.CalcBodyGeometryFromMesh (.Part p) [0x00000] in <filename unknown>:0 
  at ferram4.FARAeroUtil.SetBasicDragModuleProperties (.Part p) [0x00000] in <filename unknown>:0 
  at ferram4.FARAeroUtil.AddBasicDragModule (.Part p) [0x00000] in <filename unknown>:0 
  at ferram4.FARGlobalControlEditorObject.FindPartsWithoutFARModel (System.Collections.Generic.List`1 editorShip) [0x00000] in <filename unknown>:0 
  at ferram4.FARGlobalControlEditorObject.LateUpdate () [0x00000] in <filename unknown>:0 

(Filename:  Line: -1)

AIES Galaxy-VR2: Choosing up axis for FAR drag model.

(Filename: /BuildAgent/work/d3d49558e4d408f4/artifacts/LinuxStandalonePlayerGenerated/UnityEngineDebug.cpp Line: 53)

NullReferenceException: Object reference not set to an instance of an object
  at ferram4.FARGeoUtil.PartMaxBoundaries (.Part p, Vector3 worldOffset, Matrix4x4 partUpMatrix, UnityEngine.Transform[] ModelTransforms) [0x00000] in <filename unknown>:0 
  at ferram4.FARGeoUtil.CalcBodyGeometryFromMesh (.Part p) [0x00000] in <filename unknown>:0 
  at ferram4.FARAeroUtil.SetBasicDragModuleProperties (.Part p) [0x00000] in <filename unknown>:0 
  at ferram4.FARAeroUtil.AddBasicDragModule (.Part p) [0x00000] in <filename unknown>:0 
  at ferram4.FARGlobalControlFlightObject.FindPartsWithoutFARModel (.Vessel v) [0x00000] in <filename unknown>:0 
  at EventData`1[Vessel].Fire (.Vessel data) [0x00000] in <filename unknown>:0 
  at Vessel.GoOffRails () [0x00000] in <filename unknown>:0 
  at OrbitPhysicsManager.LateUpdate () [0x00000] in <filename unknown>:0 

(Filename:  Line: -1)

[ferram4]

Cool, I found the bug. Turned out that it was due to the Galaxy-VR2 having a ModuleJettison defined (for engine fairings and such) but that doesn't actually point to an existing transform, so FAR choked on it because it didn't expect a messed up config.

I'll do some more testing, make sure that I didn't break anything else, see if I can fix any other bugs and try to get the fix out either tonight or tomorrow morning. As a stopgap, you can go and remove the ModuleJettison from the Galaxy-VR2 config and that should fix it.

[sidfu]

this is the error i get it comes from the fasa redstone launch clamps remove the clamps its fine. other launch clamps work fine also. but here is the kicker when u put them on any design in sandbox they work fine.

(Filename: C:/BuildAgent/work/d3d49558e4d408f4/artifacts/StandalonePlayerGenerated/UnityEngineDebug.cpp Line: 53)

NullReferenceException: Object reference not set to an instance of an object
  at ferram4.FARGeoUtil.PartMaxBoundaries (.Part p, Vector3 worldOffset, Matrix4x4 partUpMatrix, UnityEngine.Transform[] ModelTransforms) [0x00000] in <filename unknown>:0 

  at ferram4.FARGeoUtil.CalcBodyGeometryFromMesh (.Part p) [0x00000] in <filename unknown>:0 

  at ferram4.FARAeroUtil.SetBasicDragModuleProperties (.Part p) [0x00000] in <filename unknown>:0 

  at ferram4.FARAeroUtil.AddBasicDragModule (.Part p) [0x00000] in <filename unknown>:0 

  at ferram4.FARGlobalControlEditorObject.FindPartsWithoutFARModel (System.Collections.Generic.List`1 editorShip) [0x00000] in <filename unknown>:0 

  at ferram4.FARGlobalControlEditorObject.LateUpdate () [0x00000] in <filename unknown>:0 

found something else out if u put a wing on decouplers to get rid of in space far gets a error. well atleast that decoupler did and it caused all wings to make no drag or lift

i give up got everythign working then on reentry started throwing million errors

Edited April 4, 2014 by sidfu

[Tidus Klein]

Ha ferram I think I found a bug in B9, The small landing legs produce a ton of drag even when retracted.

SCREENY

I looked in the cfg file for it but thats not it so I believe it my be FAR's doing,

other mods..

toolbar

AJE

B9

Distant object

engineer

JSI/roster prop

hot-rockets

Pfarings

Real chute

and alarm clock.

hope this helps.....I would give a log but I dont think its that type of error

EDIT wait I think It was because the gear was turned backwards, because the front gear produces almost no drag. let me test it.

EDIT2 nope.... HMM It must be because its on the wing.

let me try something stupid

YUP its the wings?...odd

Javascript is disabled. View full album

and yes I landed it....ON ONLY 2 WHEELS

Edited April 4, 2014 by Tidus Klein

[smunisto]

These cd, cl, ct things on the right-click menus of parts. Can they be removed? They generate some clutter.

[rottielover]

Request: Someone with some YouTube skills please work with Ferram to create the "how to" video of rocket design and launch under FAR?

Reason: All the best tips and tricks are spread out over 300 pages of forum post "wall of text". With the release of .13, I too had some designs that were no longer working, but based on ferram's notes I kinda expected to have some issues. But I'm still not 100% sure what I'm actually doing right and wrong.

Bonus points for explaining the flight data dialogue box, like what is "Q", why it's important, and what it's trying to tell me about my crappy rocket design.

I suspect this might be a series in itself.

[Green Skull]

These cd, cl, ct things on the right-click menus of parts. Can they be removed? They generate some clutter.

Although useful in some instances I do have to agree somewhat because it does help to make some of my part menus so long I have to change camera angle and scroll out to read Maybe make it toggle in the GUI?

Request: Someone with some YouTube skills please work with Ferram to create the "how to" video of rocket design and launch under FAR?

Reason: All the best tips and tricks are spread out over 300 pages of forum post "wall of text". With the release of .13, I too had some designs that were no longer working, but based on ferram's notes I kinda expected to have some issues. But I'm still not 100% sure what I'm actually doing right and wrong.

Bonus points for explaining the flight data dialogue box, like what is "Q", why it's important, and what it's trying to tell me about my crappy rocket design.

I suspect this might be a series in itself.

Scott Manley you need to do this and you also need some custom suits!

[rottielover]

Scott Manley you need to do this and you also need some custom suits!

Agree that Scott would be an awesome narrator for this. I've noticed that a lot of people (you tubers streamers posters) seem to consider FAR as "that mod for spaceplanes" (as I've heard on streams). IMO that's a dangerous misconception. FAR is an aerodynamics re-work that affects rockets and planes (anything that flys). (Even that is an oversimplification).

IMHO this thread would calm somewhat if people had a really good resource to learn how to work with and use what FAR is telling them.

[Renegrade]

I've noticed that a lot of people (you tubers streamers posters) seem to consider FAR as "that mod for spaceplanes" (as I've heard on streams).

It's funny, I actually prefer it for rockets, myself. Stock aero makes it feel like adding mass to a rocket is completely useless, cancelled out by drag -- it's not, obviously, but the drag penalty for mass in vanilla is obnoxiously high and makes it feel that way.

Not only does FAR make slim, aerodynamic rockets more efficient, but if you're a bit crazy (like me), you can go for a launch TWR of 4+ and get to orbit in only a fraction of the normal time with only a small loss in efficiency. It's a wild, crazy, hard to control ride, and totally DRE-incompatible, but it's fast and fun. Stock aero would eat all the extra energy and just laugh.

On the other hand, I find it makes planes very twitchy and sensitive, and have high landing speeds, which makes my Sir Wingsalot science plane not particularly well suited for landings on unprepared surfaces.

[rottielover]

It's funny, I actually prefer it for rockets, myself. Stock aero makes it feel like adding mass to a rocket is completely useless, cancelled out by drag -- it's not, obviously, but the drag penalty for mass in vanilla is obnoxiously high and makes it feel that way.

Not only does FAR make slim, aerodynamic rockets more efficient, but if you're a bit crazy (like me), you can go for a launch TWR of 4+ and get to orbit in only a fraction of the normal time with only a small loss in efficiency. It's a wild, crazy, hard to control ride, and totally DRE-incompatible, but it's fast and fun. Stock aero would eat all the extra energy and just laugh.

On the other hand, I find it makes planes very twitchy and sensitive, and have high landing speeds, which makes my Sir Wingsalot science plane not particularly well suited for landings on unprepared surfaces.

I think you've just illustrated my point. It is possible to have a completely stable aircraft using FAR and the SPH/VAB tools window gives you the graphs and numbers needed to design such aircraft (ask Dirt_Merchant). However, for my to sit and explain in a wall of text would do neither of us much good.

See a while back Mr, Dirt _Merchant and I had a lengthy Skype call to go over FAR's resources. I've had years of experience with flight sims, and once dirt explained what each number was and how to read the graphs, I started making planes that could fly well. But not everyone has the benefit of getting a personal 1-1 session to explain how to use it.

Hence I see a need for a tutorial video series, and I'm talking more in-depth than what Scott did in the 2nd video linked in the main post.

I think you could very well split it into a rockets series and a plane series.

I would volunteer or even do it myself, but I have a horrible voice that no one wants to listen too and I have way too much RL crud going on than anyone should be forced to deal with.

[Renegrade]

Hence I see a need for a tutorial video series, and I'm talking more in-depth than what Scott did in the 2nd video linked in the main post.

That would definitely be helpful. Especially for those stability derivatives.. I tend to design FAR planes like I do stock ones, just with the two (three with RCS Build Aid) indicators, and only a little peek at the yellow line in the graphs.. and then test flights. Also a detailed overview of Cd/reference area would be good (ex. does the graph's Cd line incorporate reference area? Cd is kinda meaningless without the reference area, no?)

Still haven't been able to replicate the 20m/sec takeoff version of Sir Wingsalot though. Doesn't even get close without the tier 8 parts, and at that point, the Science Plane is useless or will be useless shortly.

[Camacha]

Ferram, it's really ironic that it took breaking rockets when node size is wrong to finally get modders to fix their node sizes...just as you're working on an update that makes that no longer necessary for FAR.

My first thought when I heard about the new requirements for node sizes was that I am glad that you guys finally catch a break. I know that the random use of those has been a pain until now, so hopefully that is mitigated a bit.

Whatever the case, I really appreciate your and Ferram's great efforts and expertise.

[White Owl]

I'd do it, but I still don't know how to use most of the information FAR gives us.

Make it look like a real-life airplane. Put the center of mass a little forward of the center of lift. Place your fuel tanks directly on the center of mass, so the CoM won't move as fuel is consumed. Place your elevators far enough from both CoM and CoL that you'll have plenty of leverage. Avoid canards unless you really know what you're doing, 'cause those get crazy unstable if they're just a little bit wrong. If you're going fast, give it swept wings. If it's unstable, sweep 'em some more.

All that's simple enough. But how to use the charts and graphs with all that really important looking data... pfft, I ain't got no idea!

[Lexif]

These cd, cl, ct things on the right-click menus of parts. Can they be removed? They generate some clutter.

Yes, I agree, I wanted to report this, too. At least I should be able to switch them off, maybe? Also, I think they were not in the last full release, right? (Didn't try the experimental.)

[CaptainKipard]

Does putting an inverted cone on the bottom of a radially attached stack also help with aerodynamics?