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

[DaMichel]

I noticed the same issue with the FAR Gui as elfindreams. (sorry, another one of those reports) I only build two planes in 0.24 yet .. so .. one of them has a cockpit with a symmetrical counterpart and this is the one which doesn't work (no FAR Gui in flight). I have a feeling that it is a symmetry issue, because the other, a "normal" plane, works well. Also, when this bug occurs, my Kerbal Flight Data plugin has trouble obtaining a reference to FARControlSys.activeControlSys. Apparently it is null and never set to anything else. Here is the craft file https://www.dropbox.com/s/cvqs827ebp4913a/Test%20Plane%20No6.craft

[elfindreams]

@Maramusine: That implies an improper install for previous versions of FAR. FAR has never been able to east away several km/s of dV during launch.

@Dinokin: Then you can certainly provide explicit reproduction steps to cause the issue, correct?

@lammatt: Nope, that's the way drag is in real life. Planes do need to be very careful to control their descent rate or they end up going too fast and being destroyed by aerodynamic forces / not being ale to land because they're going to fast. Unless you're gaining speed while descending at a rate of less than 5 m/s, there's no issue.

@AndreATGB: Yeah, I'm gonna switch away from TAS, but that error looks like it's outside what I can control, since it's inside Unity. Unless you can provide reliable reproduction steps there's nothing I can do.

@elfindreams: Based on the logs there is nothing clearly going wrong, although I see that you're using a very large number of mods and not all of them may be up-to-date for KSP 0.24. I'll need a lot more information to reproduce the issue, since you have not provided any reproduction steps whatsoever. Surely, all the ships that have this issue have something in common, right?

I wish... I just don't know what they could have in common, some have been probes, one was just the basic capusle and a big booster rocket... I will do more testing as soon as I am done chasing down another issue and post more details.

[klikkolee]

@Dinokin: If that is a craft file from before 0.24, you will need to follow the directions given at the bottom of the FAQ in the OP / in the past few pages to uncorrupt it. If you had read the past few pages, you would be aware that it is a known issue and that there is a known fix.

@dewin: Noted. I'll fix that when I get a chance, but given that the number of support requests hasn't died down enough for me to determine what is and isn't a valid bug, I'm not making another release and making things even more confusing for people.

@klikkolee: Search for "Stability Derivatives" and you'll find resources on it. Searching for "airplane stability and control" might also bring stuff up. A final thing to try is to see if you can find the USAF Stability and Control DATCOM, since that should also include information on that. Sorry I can't point you somewhere good, most of the data for that has come out of my textbooks.

do you remember what textbooks in particular they were? that would be quite helpful, since it is comparatively easy to find textbooks or their sources.

[ferram4]

@elfindreams: I think the latest hotfix might have fixed it. Try that out.

@DaMichel: Confirmed, fixed, and pushed.

@Sabazi: There will be a new NEAR version incoming soon. Keep in mind that the PartModule classes are identical in FAR and NEAR to reduce the likelihood of bugs creeping in due to mismatched names as well as providing support for mods that already have explicit support for FAR.

@klikkolee: The ones I know off-hand are Flight Stability And Automatic Control, 2e, Robert C. Nelson, ISBN-13: 978-0-07-046273-1 and Aerodynamics, Aeronautics, and Flight Mechanics, 2e, Barnes W. McCormick, ISBN 0-471-57506-2.

So FAR v0.14.0.2 is up to deal with some more nullref issues I came across and fix displaying shielding from cargo bays and payload fairings in the editor.

[lammatt]

@Maramusine: That implies an improper install for previous versions of FAR. FAR has never been able to east away several km/s of dV during launch.

@Dinokin: Then you can certainly provide explicit reproduction steps to cause the issue, correct?

@lammatt: Nope, that's the way drag is in real life. Planes do need to be very careful to control their descent rate or they end up going too fast and being destroyed by aerodynamic forces / not being ale to land because they're going to fast. Unless you're gaining speed while descending at a rate of less than 5 m/s, there's no issue.

@AndreATGB: Yeah, I'm gonna switch away from TAS, but that error looks like it's outside what I can control, since it's inside Unity. Unless you can provide reliable reproduction steps there's nothing I can do.

@elfindreams: Based on the logs there is nothing clearly going wrong, although I see that you're using a very large number of mods and not all of them may be up-to-date for KSP 0.24. I'll need a lot more information to reproduce the issue, since you have not provided any reproduction steps whatsoever. Surely, all the ships that have this issue have something in common, right?

not exactly

i am talking about it is gaining speed like when you released v0.13

when the plane gains speed like crazy until it breaks because of structural failure.

i dont think real world works that way.

in the real world you cant gain total internal energy, can you?

[Hodo]

not exactly

i am talking about it is gaining speed like when you released v0.13

when the plane gains speed like crazy until it breaks because of structural failure.

i dont think real world works that way.

in the real world you cant gain total internal energy, can you?

I am running 64KSP and I can't say I have seen this yet. I have two installs my RO install which is still .23.5 and my main install which is .24. And the Main install has the current version of FAR and I haven't noticed this problem yet. I have had a few test aircraft out and about since it came out. Can you tell me the exact profile of what you are doing to get this bug so I can see if it is a difference in flight profiles? And do you have a picture of the craft? Is there anything clipped into the craft?

[elfindreams]

@elfindreams: I think the latest hotfix might have fixed it. Try that out.

Nope, but I have good news, I have it narrowed down to the smallest of differences causing the problem I could conceivably do.

I have two craft files out at: https://c4990ae854af0f39d9dc5e95454cc71275334fd5.googledrive.com/host/0B8a1K8EUgZmVeV8zUHNDdzQ4X3M/

One named FARout: Docking Port ROOT is simply a clampotron Jr with a command capsule underneath it. The FAR menu does not come up for this "craft".

The one named FARout: Command Pod ROOT has the same capsule with a clampotron Jr on top of it (I started with the command pod on this one, started with the docking port on the other) and the FAR menu comes up.

I stripped out my GameData to just the Squad and NASAmission directories, logged in realized I couldn't find the FAR button because I had stripped my GameData out to just the Squad and NASAmission directories, added FAR and was able to reproduce it with just FAR and Stock.

[Entropius]

I think it would be nice if somebody came up with a more comprehensive tutorial on the FAR analysis tools. Thus far I've been getting away with saying "Does it LOOK aerodynamic?", but frankly I'd prefer a more qualitative technique, and I imagine the FAR analysis tools in the build-editor are exactly for that. But they're not very self-explanitory. I've seen the videos on the first page, but they're not answering my questions.

For example:

I know Cl is lift, Cd is drag, and L/D is some sort of comparison of lift-to-drag, but are there supposed to be units? If so, what are they? Does it mean anything particularly special if one line is above or below another?

I know cM (yellow line) should be negative to be stable, but what does cM stand for? What is it beyond the fact it influences my stability? How to I raise or lower that line?

When I type into the textfield "Mach/AoA:" what is that parameter setting up exactly? What does "Sweep AoA" or "Sweep Mach" mean? When I first saw these, I initially thought "maybe when I type in 25 into the textfield it shows me a graph of my plane's performance with a 25º AoA", but then I saw the graph's X-axis is already Angle of Attack, so that doesn't seem to make sense.

And then there's the Stability Derivatives and Simulation panels, which I comprehend exactly nothing of and could use an explanation of, from top to bottom.

Is there a detailed guide for all this stuff and I'm just overlooking it?

[NathanKell]

Have you checked the wiki on the git repo?

[klikkolee]

@klikkolee: The ones I know off-hand are Flight Stability And Automatic Control, 2e, Robert C. Nelson, ISBN-13: 978-0-07-046273-1 and Aerodynamics, Aeronautics, and Flight Mechanics, 2e, Barnes W. McCormick, ISBN 0-471-57506-2.

thank you!

[Hodo]

Ferram4, is 14.0.2 compatable with KSP.23.5?

[nickexists]

Just came back after a long break from KSP, so i'm totally new to the whole aerodynamic failures feature. It appears that engines are about 3 times more powerful than stock when aerodynamic failures is enabled. Is this intentional?

[Kalloran]

I know Cl is lift, Cd is drag, and L/D is some sort of comparison of lift-to-drag, but are there supposed to be units? If so, what are they? Does it mean anything particularly special if one line is above or below another?

Cl is the coefficient of lift, and Cd is the coefficient of drag, unitless parameters that characterize the lift and drag properties of the aircraft at a specific AoA. Lift itself, at a specific AoA is a function of the coefficient of lift, the dynamic pressure, and either the chord length of the airfoil, for section lift, l or the planform area of the lifting body for finite wing lift L. Realize that finite wing lift is also affected by Oswald's efficiency number which takes into account how efficient the lift profile of the lifting body is: finite wings produce less lift at the wingtips due to the effects of wingtip vortices. L/D is the lift to drag ratio, or how much more lift the body is producing relative to drag. Again, there are no units.

I'm assuming that the span efficiency number (oswald's efficiency number) is constant for all aircraft in FAR since it is usually experimentally derived (or derived via CFD). For that matter, I'm not sure what lifting shape profile he is using for lift, drag, and moment curves or if he has integrated different curves for each wing surface.

I know cM (yellow line) should be negative to be stable, but what does cM stand for? What is it beyond the fact it influences my stability? How to I raise or lower that line?

cM is the static coefficient of longitudinal stability, also known as one of the static stability derivatives. It is the slope of the line defined by the aerodynamic moment about the aircraft center of gravity as a function of AoA at a trim airspeed. Since standard convention dictates that a nose-down moment is negative, cM must be negative for the aircraft to be statically stable, or, initially, want to return to its unperturbed state after some deviation from trim. It is also why you want your center of lift behind your center of gravity for longitudinal stability. All aerodynamic forces act through the aerodynamic center of the aircraft, while the motion of the aircraft is centered about the center of gravity. Imagine an aircraft flying straight and level (Lift = Weight, Thrust = Drag) at some trim AoA, a. Now imagine that the nose deviates up, briefly increasing the angle of attack. An increase in angle of attack with no decrease in speed will increase lift. Since the lift is acting behind the center of gravity the increased lift will initially create a nose down moment, pushing the aircraft back toward its trim a. Now imagine that same aircraft with the center of lift in front of the center of gravity. As the nose deviates upward, the aircraft generates more lift. The increased lift causes a nose UP moment which further exacerbates the deviation, generates more lift, and pushes the nose up further. This is why your airplane flips on its back. It is trying to establish a stable condition. Also, this is a grossly simplified explanation, since I didn't mention the balance in the inherent aerodynamic moment of the aircraft, but I think you get the gist.

When I type into the textfield "Mach/AoA:" what is that parameter setting up exactly? What does "Sweep AoA" or "Sweep Mach" mean? When I first saw these, I initially thought "maybe when I type in 25 into the textfield it shows me a graph of my plane's performance with a 25º AoA", but then I saw the graph's X-axis is already Angle of Attack, so that doesn't seem to make sense.

By "sweep," it will generate a range of results for a range of Mach and AoA. It's "sweeping" through the Mach/AoA range.

And then there's the Stability Derivatives and Simulation panels, which I comprehend exactly nothing of and could use an explanation of, from top to bottom.

Is there a detailed guide for all this stuff and I'm just overlooking it?

Do you have 3 semesters available? I can recommend a number of good books.....

Edited July 20, 2014 by Kalloran

[ferram4]

@elfindreams: I'll see if I can reproduce it. Probably some variation of the issues I've been tracking down.

@Entropius: Wiki has information, as does the help function on the GUI, which is the big button with a "?" on it. You should click it.

@Hodo: Nope. Compatibility broke due to support of the new AppLauncher system.

@nickexists: Engine thrust is not affected in any way by the aerodynamic failure setting in FAR. Further, FAR does not change any engine behavior besides to slightly nerf the jet engines, though no where as much as they should be for realistic behavior. Other than that, all engines are left with the thrust and efficiency they have in stock KSP. If you would prefer more realistic jet engine behavior, you'll have to get AJE, since that is outside the scope of an aerodynamics mod.

[Hodo]

@Hodo: Nope. Compatibility broke due to support of the new AppLauncher system.

Ok good to know, won't install on my RO install then.

[Entropius]

@Kalloran: Thanks.

@NathanKell& ferram4: No I hadn't seen that.

That said I still didn't see an answer to my biggest question, neither in the Wiki nor the in-game "?" button.

I get that the graph is generating results across a range of AoA values when I click "Sweep AoA", but I still don't see an answer as to how the numeric textfield "Mach/AoA:" is influencing the results on the AoA graph. That parameter doesn't appear to be mentioned in the wiki:

Is that textfield something I should be playing with on AoA-graphs? If so, how?

EDIT: NEVERMIND, I get it now. I kept reading the "Mach/AoA" literally as "Mach divided by AoA", rather than "The mach of the AoA chart or the AoA of your mach-chart".

Edited July 20, 2014 by Entropius
Because I'm dumb.

[DaMichel]

Thanks Ferram, the lastest FAR version fixed the GUI issue. However this was with the version from the source repository, *not* the one you packaged with 0.14.0.2 which seemed to be a little bit older.

@elfindreams: try this https://github.com/ferram4/Ferram-Aerospace-Research/blob/master/GameData/FerramAerospaceResearch/Plugins/FerramAerospaceResearch.dll?raw=true It did fix the problems for me.

Beside that, i noticed a strange interaction with the (currently broken) Improved Chase Cam plugin: While moving the view with the mouse it makes FAR data temporarily unavailable, i.e. speed display revers to stock and Kerbal Flight Data apparently cannot get an instance of FarControlSys. Very strange but whatever ...

[IronStar]

I have two problems with FAR after 0.24( i use 14.0.2 FAR)- first is that when craft exceeds 560 m\s camera begins to behave wierdly- at this speed craft moves forward while camera remains on the same spot, and it gets worse when craft exceeds 700 ms and so on

I ve also noticed that this happens faster(on 360 m\s) if TWR is lower

It looks like this

http://i.imgur.com/VuwjcmM.png

http://i.imgur.com/7uYPM0l.png

Changing camera settings or reseting it does not help

And this stops if i delete FAR

Second is that FAR doesnt work in sandbox- its ok in career mode(except that camera issue) but in sandbox- FAR is on the toolbar and GUI appears on a Space center screen but it doesnt appear on a launchpad and in flight- FAR is on a toolbar but GUI is not there and craft behaves like stock KSP in sandbox(and no camera issue).

Edited July 20, 2014 by IronStar

[Bluelogic32]

How do I build Rockets using this mod? I can always get my rockets to pitch but sometimes they don't turn with gravity, sometimes they turn the wrong way, sometimes they do turn with gravity but when I jettison the first stage the rocket will just start spining out of controll. What is happening? I've looked for guids on how to build rockets with far but I can't find anything, all the guides I've found are just for planes. I found something to do with adding wings to my rockets, but I found they make no difference and I don't want to add wings on my rockets. I use KW for rocket parts andd engines, I don't know if that makes any difference.

[Aedile]

Will this work without the stupid mod-statistic stuff?

[softweir]

How do I build Rockets using this mod?

It sounds to me as if you may have excessive thrust to weight. On the pad you need to keep the thrust down to between 1.2 and 1.4 times weight, and have each stage start at roughly the same ratio. Keep an eye on the time to Apoapsis: it should stick at around 30 to 45 seconds. If it goes above that, throttle down a bit. If you are using MechJeb, go for an early start to the gravity turn, and keep it very low and flat.

To get to LKO you need about 3,5000m/s of dV.

If you were using the stock aerodynamics then this would result in failure because you'd lose far too much dV fighting the aerodynamic soup. With FAR, you can use much lighter engines, so the apparent inefficiency of a low, flat gravity turn is compensated for by having to loft much less engine mass and having much less drag.

[Bluelogic32]

It sounds to me as if you may have excessive thrust to weight. On the pad you need to keep the thrust down to between 1.2 and 1.4 times weight, and have each stage start at roughly the same ratio. Keep an eye on the time to Apoapsis: it should stick at around 30 to 45 seconds. If it goes above that, throttle down a bit. If you are using MechJeb, go for an early start to the gravity turn, and keep it very low and flat.

To get to LKO you need about 3,5000m/s of dV.

If you were using the stock aerodynamics then this would result in failure because you'd lose far too much dV fighting the aerodynamic soup. With FAR, you can use much lighter engines, so the apparent inefficiency of a low, flat gravity turn is compensated for by having to loft much less engine mass and having much less drag.

Thank you so much, I got into orbit. My TTW was around 6.4

[Kamuchi]

I am seeing alot of "isShielded: False" again with 14.0.2 with x64

[Northstar1989]

@Ferram

I was just wondering a little about how FAR calculates lift in relation to wing size/shape.

Namely, I was curious if FAR accurately accounted for the fact that thicker wings generate more lift, though at the expense of lift/drag ratio.

I'm sure you're already familiar with this, but for anyone else reading this, there's a mod called Procedural Dynamics that lets you design custom wings. One of the things you can customize is wing thickness.

In stock aerodynamics, lift seems to just be proportional to mass (the way drag is), so thicker Procedural Dynamics wings only generate more lift because of their higher mass.

But in a realistic aerodynamics system like FAR, my expectations are a little higher, of course. In real life, thicker wings generate more lift because they part more air, and thus are able to move greater volumes of air above/below the wing in differing pressure-ratios (using the model of lift where lift is due to differences in pressure above/below the wing)

Of course, this is mainly of use in subsonic flight- since lift/drag ratio, rather than raw lift potential, becomes the dominant factor in supersonic flight. However in subsonic flight, thicker-winged aircraft tend to be able to lift much heavier payloads and fly at much slower speeds than those with thinner wings, due to their superior lift.

I was wondering how well this was simulated in FAR (if at all). This is because I would like to build some large-winged subsonic propeller-aircraft for exploration around Kerbin using Procedural Dynamics, and I should be able to attain a higher altitude ceiling and lower liftoff speed (due to lower wing-load) for the same craft (if FAR does this part of the aerodynamics right) if I build the same craft with thicker wings of otherwise the same dimensions...

By the way, there's also some stuff that goes on with root-to-tip thickness ratio that I don't fully understand: basically, aircraft with wings that taper more towards the tip have superior performance to those with wings that are relatively flat in terms of thickness. But I'll leave this to somebody with more experience with aerodynamic modeling like you to figure out, if you haven't already incorporated it into FAR...

Regards,

Northstar

[ferram4]

@Da Michel: I'll make sure that I didn't package the wrong version. That would be bad.

@IronStar: The camera issue is not a FAR bug. That is due to another one of the mods you have installed and you should check those first. I suspect a similar issue might be causing the issue between sandbox and career, unless you can replicate the issues using the exact same craft file for both.

In any case, not providing an output_log.txt means that nothing can be done about your bug report.

@Aedile: Yes, that's why ModStatistics is set up as a separate dll rather than being integrated into the main dll like InstallChecker and CompatibilityChecker are. You would disable it far more reliably by simply setting it to disabled in the config file, but I understand why you'd rather go with that method. It will not affect how FAR performs, but you will have to remove the dll every time you update FAR or another mod that packages it.

@Kamuchi: That's very odd. Try the version that DaMichel linked from github.

@Northstar: In practice, in real life there is little difference between the thickness of the wing and how much lift you get from it (assuming the same angle of attack and camber) until you get near stall. At that point, thinner wings will have much more sudden stalls than thicker wings (the difference being leading edge versus trailing edge stall). In general, anything with a thickness greater than ~12% chord will have its peak lift dropped severely by early onset of trialing edge stall while anything thinner than ~12% chord will have its lift dropped severely by leading edge stall at much lower AoAs.

FAR doesn't bother with this, since it's kinf od complicated to figure out and would make procedural wings a de facto required mod. All wings are set up as 5% thick, with whatever shape is ideal for either subsonic or supersonic flight.