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

[ferram4]

@Donziboy: What Master Tao said.

@ABZB: Running KSP x64? If so, inherent issue in x64, due to god only knows what. The stack trace never points anywhere valuable, and those traces don't even point to where things are breaking, because those exceptions are handled which means they cannot crash the game.

[Donziboy2]

Donziboy,

The only current fix is to download the most recent dev version from GitHub.

Thanks, downloaded the FAR Master from the github that seems to have fixed it.

[Shaun]

What's going on with the Basic Jet Engine? It spools up to about 85kN on the runway stationary. As soon as it picks up speed it spools down. The engine seems to have more power in a vertical climb, and very low thrust level or descending...

[Renegrade]

You don't need the oomph thanks to the lessened drag and the fact you're already travelling upwards more than fast enough to not need it. That and the further away you get from the surface, the less effect gravity has on you so a surface TWR of 0.9 might be 1 in space (I don't know the exact numbers). I could be wrong about all this though - this is just what I think is happening.

It's actually about 78%-ish, depending on your orbit (assuming 80km here). Also as your horizontal velocity increases, you're getting more centrifugal force cancelling out that gravity...

Which is only true because Kerbin is the size of a softball, but hey.

Actually it applies on Earth too. The Apollo orbit was 400km according to info I could scrape up, and that's about 88% gravity.

A to-scale orbit of 850km (13.33% of the planet's radius) would be ... 78%-ish. Perhaps a better thing to say here is that Kerbin's atmosphere is too tall..

With a bit of practice you can make an all-solid boost stage and save the throttleable engines for circularisation.

I agree. My FAR-enabled games all have pure-solid boost platforms that work fine. I prefer solids in fact (better cost, and the decreased flexibility results in engineering mattering more).

What's going on with the Basic Jet Engine? It spools up to about 85kN on the runway stationary. As soon as it picks up speed it spools down. The engine seems to have more power in a vertical climb, and very low thrust level or descending...

This isn't new, it's had an altered performance curve in FAR for as long as I can remember. The only change recently is that thrust has been reduced. Corrections to the newest versions of FAR have resulted in even less drag than before, so engines were wildly overpowered, and had thrust altered as a result. They were balanced for stock air anyhow, which is much thicker than FAR's.

[The Pink Ranger]

When I use the auto leveler it tries to level my plane upside down. Anyone else having this issue? The particular plane I was using has the center of lift slightly below the center of mass, could that be the reason? Flies straight as an arrow with no leveler or asas.

Edit: Okay I had my k value set to high. Seems to work fine at the default value.

Edited August 11, 2014 by The Pink Ranger

[ABZB]

@ABZB: Running KSP x64? If so, inherent issue in x64, due to god only knows what. The stack trace never points anywhere valuable, and those traces don't even point to where things are breaking, because those exceptions are handled which means they cannot crash the game.

[tetryds]

Is there any useful data that I might supply in the future that I didn't supply in this case?

Let me quote this for you:

ABZB,Please include this important information in any bug report:

• Your KSP version
  
• Your mods and versions
  
• Your Operating System and version
  
• Steps to cause the problem with as few mods as possible
  
• Cause the problem, then quit KSP and find your output log:
  
  • Windows (32 bit): KSP_win\KSP_Data\output_log.txt
    
  • Windows (64 bit): KSP_win64\KSP_x64_DATA\output_log.txt
    
  • Mac OS X: Open Console, find Unity on the left side, and click on Player.log. It's also located at ~/Library/Logs/Unity/Player.log.
    
  • Linux: ~/.config/unity3d/Squad/Kerbal\ Space\ Program/Player.log
    

[Galenmacil]

Hello everyone,

First of all, thanks Ferram4 for this awesome mod. Really nice work to improve vanilla aerodynamics!

I'd like to report a few issues I have with this mod. Playing on KSP 0.24.2 32 bit on Windows 7 64 bit with FAR 0.14.1.1 :

1 - The roll control animation is reversed for aileron located behind the center of mass, for example, on a delta wing. The craft turn and respond correctly but the animation is reversed. Is this related to KSP or FAR?

2 - It would be great to be able to adjust the control range from -85 to +85 degrees. Currently we are limited to -30.

3 - Sometimes the craft as a tendency to roll to the right. All controls are centered, SAS is disabled as well as all FAR control "helpers". Seems to be related somehow to spoilers/flaps settings or to some types of parts combinations. On a particular case, jettisoning the two external tanks (which where balanced: both filled and symmetrically positioned on each wing underside) would "magically" solve the problem. I know this is kind of vague but I was not able to clearly identify the culprit... Anyone else noticed a similar effect?

Keep up the development on FAR. You are a genius!

Thanks

Edited August 12, 2014 by Galenmacil

[NathanKell]

#3 appears to be a stock bug, due to how forces are applied to symmetric parts.

[ferram4]

@Galenmacil:

1. I cannot confirm. Testing on the stock FAR Thunderbird craft resulted in ailerons (behind the CoM) deflecting as expected: a roll left command resulted in the left aileron deflecting upward (to produce a downward force) while the right aileron deflected in the opposite direction.

2. 30 degrees already puts the control surface into the stalling range. Further, deflections for flaps and spoilers (which you might want to stall) already allow very large ranges, they simply use a different setting (to allow a surface to simultaneously be a flap / spoiler and control surface)

3. I've gotten a bunch of these recently, so I'll address how to handle it if you think it's a bug in FAR:

Produce the smallest craft that shows the behavior possible; if the wing needs to be made of 20 parts, it's too big.

Strut it to hell to remove as much flexing as possible.

Determine which of the parts is responsible for the issue (as in, removing these two symmetry-attached parts causes the issue to cease).

Provide the craft file and point to which of the parts is the problem one.

Without all of that, I will assume that it is a documented and unfixed stock issue regarding symmetrical vehicles flexing asymmetrically, which means that I can't fix it.

[tetryds]

Hello, about 3 I believe that it's caused by this:

The control surfaces aren't perfectly mirrored, from what I know they are just fliped around.

So it may be an issue with the shape or the attachment point.

That is why trying to takeoff using flaps will make you bend to the left, from what I have tested so far.

I would place both manually on a plane, to make sure they are aligned, but at moment I don't have time.

Could someone manually fix that difference on a model which has such issue and post the results?

(maybe there are some issues when using the surfaces as their movement can mismatch a bit then.

Edited August 13, 2014 by tetryds

[NoClass]

I would place both manually on a plane, to make sure they are aligned, but at moment I don't have time.

Here goes.

I thought at first that it might be a clipping problem. You'll notice that wings in tetryds's post are overlapping slightly, despite the fact that they were placed on a centerline using symmetry. It's funny, because I noticed that in recent KSP versions, it can be hard to get a bilaterally placed part to appear. You kinda have to massage it into position for the part to pop up on the other side.

Anyway, to eliminate clipping, I simply moved the wing down the side of the tank.

Delta Wings Placed Bilaterally near the top of a tank, with Small CS, Hosted by imgur.com

...same problem.

So I did as asked, individually placing wings on top, and wings slightly lower.

Swept Wings Placed Individually on the Top of a tank, with Small CS, Hosted by imgur.com

Swept Wings Placed Individually near the Top of a Tank, with Small CS,Hosted by imgur.com

I also tried bilaterally placing the wings, but individually placing the Small CSs, but to no avail. I'll spare you the pics. You'll also notice that I used both delta wings and swept wings. They all seem to just be flipped, and are not symmetrical in the horizontal plane.

*sigh* This is making me OCD.

Edited August 13, 2014 by NoClass

[BudgetHedgehog]

The control surfaces aren't perfectly mirrored, from what I know they are just fliped around.

Someone was working on a plugin that actually mirrored flaps and stuff, not just flipped them. That would solve this.. Can't remember for the life of me who, though :/

[tetryds]

@Noclass, yes, in order to do this you need to mouse over about 3 degrees to the right of the part placed on the center so both will show up.

I made it so the fact that the wing part is perfectly aligned and the control surfaces are not gets more evident, but that overlaping didn't seem to cause any issues on designs which used it.

If you want a tip you can try to use some part that goes perfect on symetry like the cubic octogonal strut and use it as a reference when positioning the control surfaces.

[Drew Kerman]

Someone was working on a plugin that actually mirrored flaps and stuff, not just flipped them. That would solve this.. Can't remember for the life of me who, though :/

sounds like something Snjo would be whipping up...

[Nori]

Did a search for this but can't seem to find anything.

Does FAR calculate drag for the so called massless parts (PhysicalSignificance=1)?

[NoClass]

If you want a tip you can try to use some part that goes perfect on symetry like the cubic octogonal strut and use it as a reference when positioning the control surfaces.

Hrmm, how would this help with vertical symmetry?

Actually, here's a deeper question: Are we sure that the control surfaces are actually misaligned? There may be a difference between where a part is drawn on the screen, and where it is calculated to be for the sake of game physics.

Also, if you play around with it, you'll notice that the round pieces of the control surfaces are perfectly aligned. I don't know if this is significant.

Mysteries!

[ferram4]

@tetryds & NoClass: I'm pretty sure that's just the mesh. If it's caused by the control surfaces though, flying without any should cause none of the rolling effects that people are reporting; does that happen?

@Nori: Yes, because the 3.75m decoupler is massless. I would prefer not to, but I can't afford to not calculate drag for something that size.

[Nori]

@Nori: Yes, because the 3.75m decoupler is massless. I would prefer not to, but I can't afford to not calculate drag for something that size.

Completely agree. Can understand massless for something like the ladder, but the 3.75 decoupler seems like a bug or something.

So on that note, definitely want to use fairings with things like the ladder and RCS ports then right?

[WaveFunctionP]

It has just come to my attention that the deployable array for kspi has trouble with FAR. Seems it is seen as deployed at all times. I'm currently maintaining a working version of KSPI, and I'd like to fix this, but I know absolutely nothing about why this would be. I don't mind making changes to KSPI, but I don't know what in particular would cause the issue with FAR.

[ferram4]

@Nori: You don't have to, and it might not add any real benefits. There are dV losses associated with bring the fairing along (because more mass).

@WaveFunctionP: Was fixed many versions ago. Current behavior will cause the array to make full drag when it has been completely deployed, less drag when it is folded up, and the transition occurs when the animation stops.

[undercoveryankee]

Completely agree. Can understand massless for something like the ladder, but the 3.75 decoupler seems like a bug or something.

So on that note, definitely want to use fairings with things like the ladder and RCS ports then right?

There seemed to be a general trend toward making parts physicsless in 0.23.5, probably for game-performance reasons. I suspect the 3.75m decoupler is physicsless because it was a new part and it was small enough in comparison to the tanks that there was no reason not to follow the trend, and the only reason the smaller stack decouplers still have physics is that Squad rarely changes working parts.

I've also wondered whether joints for whatever reason worked better with a physicsless decoupler than with a small physicsful decoupler between two big tanks.

[NathanKell]

The second part is correct. The reasoning is that Squad finally realized that things were too wobbly.

PhysX (indeed any physics engine) has issues when a very light object is attached to a heavy object, and has *very* big issues when a light object is sandwiched between two heavy objects. If you put a light decoupler between two very heavy 3.75m tanks, you're going to get wobble. Physicsless parts are essentially skipped for joints; the joints are created between the parts attached to the physicsless part (i.e. tank-tank, rather than tank-decoupler-tank). This greatly decreases wobble, at the cost of making KSP prone to crashing if you don't use the decoupler juuuuust right (see known issues thread).

[WaveFunctionP]

@Nori: You don't have to, and it might not add any real benefits. There are dV losses associated with bring the fairing along (because more mass).

@WaveFunctionP: Was fixed many versions ago. Current behavior will cause the array to make full drag when it has been completely deployed, less drag when it is folded up, and the transition occurs when the animation stops.

Thanks for the quick reply. Seems the person that mentioned it hasn't actually tried it in a while.

[Donziboy2]

Thanks for the quick reply. Seems the person that mentioned it hasn't actually tried it in a while.

Its still has a fair amount of drag but its not exactly small lol. I was doing reentries with them on craft after it was fixed and it was manageable compared to back in the day when you would get going a few ms and your craft would literally stop and you would end up side slipping and not going up.

If you launch them without fairings your asking for a fun ride anyway.

Sorry I missed a whole page of KSPI and did not see it lol.

Edited August 13, 2014 by Donziboy2