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

[Rebelgamer]

@Noio

It's happening every time I warp for long periods in LKO(haven't tried other bodies yet). That was actually from when I originally started having the problem, now I'm producing it with just FAR installed on a clean steam install, no other mods.

Edit: Also occurs in low munar orbit(24k). 50x time warp for 2 kerbal days, same exception thrown.

Edit 2: Still occurs even if you slowly step down time warp, it doesn't occur until you go from 5x to 1x.

Edited November 13, 2014 by Rebelgamer

[ferram4]

I need full reproduction steps. I have not been able to reproduce in my install.

Worse, I don't think this is a bug in FAR; none of FAR's functions are causing the stack overflow. Everything is coming from KSP itself.

[Rebelgamer]

Ferram, here is the save file, https://www.dropbox.com/sh/ecu50fn80recdw9/AADSPFRNGt_rnYOrsnBdBoD8a?dl=0

Its a stock craft, no mods were installed when launched.

Reproduction on my end:

1: Launch craft into LKO(i was at 95km)

2: Time warp at 50x(max below 120km) for 2 kerbal days.

3: Stop time warp, either from 50-1x or 50-10-5-1x, exception gets thrown.

I'm at a loss myself, but I only get this when FAR is installed, I've tried over and over with stock ksp and it hasn't happened yet.

[Noio]

Okay, I've managed to reproduce it at will, thanks to Rebel's notes. Yay!

0.25 Win32, two mods installed: FAR 0.14.3.2, and HyperEdit, used to put the vessel in orbit in the first place.

You can find the save game quicksave here:

https://drive.google.com/file/d/0B-8cKOlqLStFMlNnbnVMN2Zad3c/view?usp=sharing

And the output_log here:

https://drive.google.com/file/d/0B-8cKOlqLStFTzlNMnpyM0Y2dWM/view?usp=sharing

To reproduce:

- Load the quicksave. After the quickload, you should be focused on a Kerbal X in orbit. Click the time-warp arrow icons in the upper left to the fastest warp you can get.

- Let it warp until T+30 days.

- Then, using the mouse on the time-warp gadget again, set speed to 1x.

Using this recipe, I was able to trigger the failure three times in a row.

If I uninstall FAR and repeat the same test, the error does not occur.

[camlost]

@ferram,

Been talking to some guys using HoneyFox's flight helper. We would feel really excited if downwash/vortex can be implemented. I remember you talked about it before, was it too expensive to do?

[praise the suuun]

im having trouble getting my experimental plane to change direction when i want it to. the design objective was to create a supersonic VTOL that flies horizontally. every time i try to level off i either get too hasty and enter side-slip stall or i reach horizontal flight but the prograde marker keeps moving down and i have to follow it otherwise i stall.

im using FAR v0.14.3.2, Pwings, KSP 0.24.2, coffee industries magic COL adjustment device. the COL needed drastic modification due to some kind of bug that places the center in strange locations i think the cause is the welded engine block at the rear i might have to take that piece to the drawing board. the coffee industries adjustment device seems to move the COL magically at will but i dont know if its the real spot or not because does not FAR calculate that in real time? isnt the hangar COL indicator a false reading when FAR is installed?

the important things you need to know to help me solve this puzzle:

• the biggest problem is the prograde marker doesnt follow the nose. do i need more lift?
  
• speed is absolutely capped out. i cant go above 2/3 throttle without the engine block crushing itself.
  
• i have to keep the impractical shape of the craft but the wings and stabilizers are ok to change as much as needed.
  
• its a VTOL horizontal flight concept.
  
• im using ksp32.exe to ensure FAR is active.
  
• i dont know if its because of my design or because of a bugged COL. i have a tonne of mods and i can look through them if you ask.
  
• the only time the hangar COL is in the COM is when its dead horizontal. all other angles are in crazy locations.
  
• launching straight up gives uncontrollable pitch up at around mach 0.4 so that might be the air speed needed to generate lift with current wings.
  

any thoughts or musings about general supersonic characteristics will be much appreciated. i can provide any information you request.

if i need to remove these images and just link the album let me know.

https://www.youtube.com/watch?v=yqZz0KWYTik&feature=youtu.be

Edited November 14, 2014 by praise the suuun

[Wanderfound]

any thoughts or musings about general supersonic characteristics will be much appreciated. i can provide any information you request.

With that many mods and part-hacks involved there are so many potential complicating factors that I wouldn't know where to start. Yes, the difficulty in maintainng level flight is related to lack of lift, but there are ways around that that don't require big wings.

To see a successful implementation of a similar concept (for reverse-engineering inspiration), have a play with http://forum.kerbalspaceprogram.com/threads/90747-Kerbodyne-SSTO-Division-Omnibus-Thread?p=1509571&viewfull=1#post1509571

Your video is set to private, BTW.

Edited November 14, 2014 by Wanderfound

[praise the suuun]

snip

your guide on the FAR graph analysis is really helpful! that section always confused the hell out of me so i can use that to tweak the wings without constant test flights. the example craft you linked is an SSTO i think while this is just for the atmosphere but id be interested to see how it performs upon reentry. both crafts are shaped so similar that i think you would experience a side-slip leading into a tumble stall. i cant try your craft because id have to download a fresh installation.

as for the birds nest of mods i think the only ones in play for the puzzle are pwings, FAR, ubiozur part welding. if anyone knows if the hangar COL indicator with FAR installed is false or not that would help so much. do P wings generate stable lift, with FAR, based on there shape-thickness-sweep? finally, ubiozur partwelding i can experiment with that crazy lifting engine block i imaged earlier with different crafts. that should give me an indication of its disturbance on lift.

ive also set the video to unlisted now. notice how the prograde doesnt follow the nose.

*edit* when FAR analyses the craft and produces graphs will it take readings as if it was flying straight up or horizontal?

Edited November 14, 2014 by praise the suuun

[Wanderfound]

your guide on the FAR graph analysis is really helpful! that section always confused the hell out of me so i can use that to tweak the wings without constant test flights. the example craft you linked is an SSTO i think while this is just for the atmosphere but id be interested to see how it performs upon reentry. both crafts are shaped so similar that i think you would experience a side-slip leading into a tumble stall. i cant try your craft because id have to download a fresh installation.

<snip>

*edit* when FAR analyses the craft and produces graphs will it take readings as if it was flying straight up or horizontal?

It has adequate yaw stability; flying straight is no problem at speed. If you are having yaw problems, though: bigger tailfins or a Vernor each side of the nose.

FAR analysis assumes that you're flying towards the door, AFAIK.

[praise the suuun]

does this look like a normal COL? what causes that?

https://www.youtube.com/watch?v=XkaNqL39R0Q&feature=youtu.be

[Wanderfound]

There are some things that mess up the CoL indicator, but they're cosmetic; they don't affect flight behaviour. I'm not sure if that's happening here.

Why are you angling it tail-first? Remember, the "wind" is coming from the door.

[praise the suuun]

.

for the COL indicator im getting a real reading only at a parallel angle. sometimes when i take a test flight out ill be hit with a real heavy pitch up at about 45deg and itll stall; even when the nose is dead center in the prograde. they may be unrelated and i might just not have enough lift so ill work on getting enough lift for the FAR analyser.

without someone elses knowledge of bugged out COL markers i wont be able to know the real location during level off.

Edited November 14, 2014 by praise the suuun

[Wanderfound]

Uncontrolled pitch-up is most likely caused by insufficient longitudinal stability, not insufficient lift. Run an AoA sweep at the speed at which you tend to lose control; if your Cm line inflects up, that's the point at which your craft is going to try to flip backwards.

If you post screenshots of the FAR analysis screens here, we may be able to identify the problem.

[tetryds]

for the COL indicator im getting a real reading only at a parallel angle. sometimes when i take a test flight out ill be hit with a real heavy pitch up at about 45deg and itll stall; even when the nose is dead center in the prograde. they may be unrelated and i might just not have enough lift so ill work on getting enough lift for the FAR analyser.

without someone elses knowledge of bugged out COL markers i wont be able to know the real location during level off.

First I want to know which version of FAR you are using, because the vector marker recently got removed.

Another thing is that the wind comes from the open dor, don't expect things to behave nicely if your plane is going backwards, haha.

And also, your wings are too far forward, for a plane with that many engines they should be way further back.

Don't put the COL adjustment device like that, never have wings with AoA higher than 15 degrees, it will mess up.

Edit:

Let me rephrase, your wings are too far forward, that is the main problem.

Move them closer to the engines, remove those "COL Fix" winglets and point your command pod towards the open dor of the hangar, perfectly on the horizontal (no inclination).

The FAR GUI has assists which simulate different angles of attack etc.

Edited November 14, 2014 by tetryds

[Max.10.07]

I like how the drag puts my plane retrograde from a stall even though most of the drag is behind the center of mass

[tetryds]

I like how the drag puts my plane retrograde from a stall even though most of the drag is behind the center of mass

Stalls are stalls, I bet you have canards.

Haha, maybe try to fly it backwards.

Also, did you check the COM shift?

[CDK]

FAR .25! YES!! Thanks! Good work!

[CaptainTightpants_64]

Not sure if this has been posted before, but apparently FAR really doesn't like the new wing parts that were added in 0.25. In addition to the CoL being higher than it should be, it gets shifted off-center if symmetry is enabled, towards the original part. This applies to the new triangular Structural Wing parts, and the new swept wing parts. Tested in Windows and Linux, 32 and 64-bit. http://i.imgur.com/05mbF2T.png

[praise the suuun]

It has adequate yaw stability; flying straight is no problem at speed. If you are having yaw problems, though: bigger tailfins or a Vernor each side of the nose.

FAR analysis assumes that you're flying towards the door, AFAIK.

First I want to know which version of FAR you are using, because the vector marker recently got removed.

Another thing is that the wind comes from the open dor, don't expect things to behave nicely if your plane is going backwards, haha.

And also, your wings are too far forward, for a plane with that many engines they should be way further back.

Don't put the COL adjustment device like that, never have wings with AoA higher than 15 degrees, it will mess up.

Edit:

Let me rephrase, your wings are too far forward, that is the main problem.

Move them closer to the engines, remove those "COL Fix" winglets and point your command pod towards the open dor of the hangar, perfectly on the horizontal (no inclination).

The FAR GUI has assists which simulate different angles of attack etc.

thank you so much for the advice it flys great this time. finally i got it to work after months of casual development! i have to admit it looks ALOT prettier after the redevelopment. i need a little bit more help analysing my graphs.

i removed some of the large intakes on the side to make room for a surface further back. i chose a shallower delta wing and wider surface area to make sure i dont stall while leveling off at subsonic. it can handle a larger aoa with the new vertical stabilizers . they have been angled ideally copied from thunderbird 1 but i am noticing a little bit of dutch roll with the verticals being angled like that. there are no ailerons what so ever; to reduce weight and moving parts. its relying purely on gimble. ive had to use the coffee industries magic COL adjuster because the wings are too far from the COM. it seems to give a real adjustment to COL as i tested with and without.

i tried to get some graph readouts but they seem to be giving odd readings so i think i got the input numbers mixed around. to reiterate im using FAR 0.14.3.2 and ksp 0.14.2.

sweep mach:

sweep aoa:

and successful test flight as a treat for you guys.

im curious why hotrockets is not giving the blue plumes. maybe it doesnt work in 32 while far does.

https://www.youtube.com/watch?v=lWF4fVmyNvc&feature=youtu.be

Edited November 16, 2014 by praise the suuun

[Fueniker S.]

Hey,

Just wondering, did Ferram change something in FAR that pretty much makes it impossible to make super-high-G-pulling planes? Because I'm just starting to remake some of my past planes, and they can't survive the forces, and yes, aero-failures is disabled.

[tetryds]

thank you so much for the advice it flys great this time. finally i got it work after months of casual development! i have to admit it looks ALOT prettier after the redevelopment. i need a little bit more help analysing my graphs.

i removed some of the large intakes on the side to make room for a surface further back. i chose a shallower delta wing and wider surface area to make sure i dont stall while leveling off at subsonic. it can handle a larger aoa with the new vertical stabilizers . they have been angled ideally copied from thunderbird 1 but i am noticing a little bit of dutch roll with the verticals being angled like that. there are no ailerons what so ever; to reduce weight and moving parts. its relying purely on gimble. ive had to use the coffee industries magic COL adjuster because the wings are too far from the COM. it seems to give a real adjustment to COL as i tested with and without.

That is because even now your wings are way too far forward.

Look, this is about what a supersonic airplane should look like:

http://upload.wikimedia.org/wikipedia/commons/3/3a/North_American_X-15_three_side_view.jpg

Small wings, bigger elevons, COL waaay far back behind the COM.

Your bigger elevons will pull it a bit forward and won't drain your stability for more maneuverability.

So you need to move your wings even more further back, they should end about where your engines are.

i tried to get some graph readouts but they seem to be giving odd readings so i think i got the input numbers mixed around. to reiterate im using FAR 0.14.3.2 and ksp 0.14.2.

sweep mach:

http://i103.photobucket.com/albums/m155/haveachillday/sideslip/sweepmach_zps7593129c.png

sweep aoa:

http://i103.photobucket.com/albums/m155/haveachillday/sideslip/sweepaoa_zpsf3f0571d.png

Just some tips when using those tools.

First is that you don't need to set lower AoA a value less than zero, if you notice there are two lines one of them (thick) means positive AoA and the other one (thinner) means negative AoA.

So if you do it like that it will just be mirrored.

Your drag is high up on the sky because you have a very wide body with tons of intakes and engines, those "COL fixers" also add an insane ammount of drag.

You don't need that many of either, given you want to fly supersonic.

On supersonic it's more important to have less drag than more thrust, so you can easily break past mach5 with a small plane: http://i.imgur.com/zTgAG2w.png

Thus your readings are fine

Ah and I think you meant KSP 0.24.2, haha

and successful test flight as a treat for you guys.

im curious why coolrockets is not giving the blue plumes. maybe it doesnt work in 32 while far does.

https://www.youtube.com/watch?v=lWF4fVmyNvc&feature=youtu.be

Interesting, with the gimbal alone, plus the COL shift due to SAS trimming you were able to keep it stable, but that is very dangerous.

I like your plane, it looks good, just needs a bit more fixing.

Now that it flies it will be a lot easier to do that.

If you want I could draw some changes that you can do in order to improve it a bit.

[Van Disaster]

There's no way of setting up yaw-control ( horizontal ) spoilers currently? also I noticed using rotational symmetry on control surfaces in the VAB does some slightly random assignments, might be easier not to attempt to assign a left/right in VAB mode...

[tetryds]

There's no way of setting up yaw-control ( horizontal ) spoilers currently? also I noticed using rotational symmetry on control surfaces in the VAB does some slightly random assignments, might be easier not to attempt to assign a left/right in VAB mode...

IIRC when you set them as spoilers they will deflect outwards, you just need to place them out of the plane of simmetry.

Please correct me if I am wrong.

[Hodo]

Hey,

Just wondering, did Ferram change something in FAR that pretty much makes it impossible to make super-high-G-pulling planes? Because I'm just starting to remake some of my past planes, and they can't survive the forces, and yes, aero-failures is disabled.

Yeah somethings changed.

And I am not sure what is happenning with your craft but I am still able to make aircraft that can pulled 10-15Gs subsonic and around 7-9Gs supersonic, with aero-failures ON.

[Van Disaster]

IIRC when you set them as spoilers they will deflect outwards, you just need to place them out of the plane of simmetry.

Please correct me if I am wrong.

Well technically they'd be horizontal surfaces set up for yaw only rather than FAR's idea of spoilers - but horizontal surfaces don't deflect when set up for yaw like you'd set up a vertical surface. You can't independently control spoilers as far as I can see, so you'd have to have a flap on one wing and a spoiler on the other, which would be horrendous to actually use