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

[FireFaced]

-snip-

Yeah sounds about right. I'd have thought the smooth surface of the fairing exterior would be more aerodynamic subsonic. Is this realistic?

Maybe a conic fairing would be more aerodynamic? I'm not sure, though.

[kcs123]

The first graph shows a less stable design than the second one since the slope of the yellow cm curve is more positive.

Both are unstable, the second one moreso

Yeah sounds about right. I'd have thought the smooth surface of the fairing exterior would be more aerodynamic subsonic. Is this realistic?

All rockets are unstable, have tendeny to pitch up more as AoA increases, that is normal behaviour for rockets.

But you have missed main point here. Slope of yelow line by itself does not represents stability of your rocket.

Most important thing is where yelow line crosses X axis of graph.

So, conclusion based on readings from your pictures is that rocket without fairings is quite more unstable than other one. As soon as you start engine it will start to backflip slowly until it falls into unrecoverable spin.

Second design with fairings, however will fly directly where you have pointed your rocket. It will fly straight until it runs out of fuel.

If you don't belive me, try to fly both rockets. On both rockets turn off SAS and dont correct steering at all from your input. Just add full throtle and fire up engines.

Do the same with other rocket and report here what has happened.

[Sput42]

it is?

Hmm. I always thought non military, small single engine jet aircraft landed at about that speed.

Assuming neutral conditions. No head, tail, or crosswinds.

now fighter jets might typically land at higher speeds? Like an F16? Maybe because it's a little heavier.

So what speed would you say is typical for landing if not 90-100 mph? 150? 180?

So something tiny like a Learjet 45, while technically not a single-engine craft, has a final approach speed of around 140 kn (160 mph). I was hard-pressed to find anything jet-powered that approaches much more slowly. An F-16 comes in at around the same speed; even large commercial airliners tend to come in at around 130-150 kts. A standard glider plane approaches at maybe 70 kts (80 mph).

Spoilers/flaps are very much required to bring you down from that speed at any reasonable distance. Spoilers are not so much for "braking" per se; they're for killing lift so you can flare (pitch up without gaining height), which then slows you down due to increased drag. Flaps, on the other hand, alter your aerodynamics in a way that both drag and angle of attack increase, while stalling speed decreases; this means that you can flare much harder, thus slow down quickly.

Modern, small jet planes like Learjet or F-16 typically have a glide ratio of about 16:1. This means with engines shutoff, such a plane manages to glide for 800 m from an approach height of 50 m without slowing down if not using spoilers or flaps. You can probably imagine what happens if you try to slow down solely by pitching up (without going into a climb) - you run out of runway pretty fast.

There's virtually no reason for the KSC runway being long enough to land a plane without spoilers with realistic aerodynamics.

[Abdu]

Are you sure this is to do with FAR? You can't warp at 1000x in atmosphere, and FAR shouldn't be affecting anything in space!

i have no other mods installed

[Bakase]

i have no other mods installed

Well, it could be a stock issue then. If it's in space, FAR isn't doing anything.

Try reproducing it without FAR.

[Solar71]

It might depend on what craft you're using, but I have absolutely no trouble slowing down. If you're having particular trouble with an aircraft, post pictures. It's probably a problem with your design or piloting, rather than the aerodynamic simulation. Could be that your craft is unrealistically heavy.

If, by chance, there is something going wrong with the simulation, you've helped fix something. But until you show that an aircraft exhibits truly unrealistic behaviour, backing this up with data, don't assume the maths behind the program is at fault.

This is an issue in Ferri. Fixed in the current devbuild.

What in hell have you done? If you want support with that, you're going to need to post exactly what modifications you made... Even so, I don't think anybody's really going to bother helping you with that. It looks to me like you've totally screwed things up in the cfg and changed something you're not meant to -- the voxels look wrong as well.

Please just relax...

I deleted FAR... Then I re-installed it. No modifications yet.

Im telling you that FAR has bugs.

Its not from my changing the atmospheric viscosity (which I hadent even done yet).

What im telling you is this........ Sometimes FAR just does this.... I might leave KSP, come back in, and its working again....

Then sometimes IN THE MIDDLE OF A FLIGHT the physics will just SHUT OFF...

example: I take off, physics are perfect. I eject my boosters, physics are perfect. I eject my 2nd stage. Physics are gone. 100% gone.

I click the FAR button to see my physics... Where it normaly shows drag of the craft, it will display 0.0 Kn

ZERO DRAG!!!

But other times. It shows the drag PERFECTLY during the entire flight. very nice and happy physics. Sometimes not.

Just look at the pic I posted... It also does this when making Airplanes. Sometimes the lines are straight, sometimes they're squiggle and sometimes there are NO physics.....

Sorry for getting all pissed off, I know you're trying to help, this is just getting so tiring.

Edited June 11, 2015 by Solar71

[blowfish]

Zero drag was an issue others were having as well. Are you using the latest deve build?

[Solar71]

Zero drag was an issue others were having as well. Are you using the latest deve build?

You know TBH im not sure if im using the latest.

Someone gave me a link about 3-4 days ago, to a version of FAR that is called FerramAerospaceResearchMaster

The part "Master" seems to mean it was not a normal release. So maybe it was a special release? im not sure.

Let me just say. Before I got the "Master" version of FAR, the physics would just shut off MUCH more often.

Since I got the "Master" Far, its been Much much better. Maybe only once per 5-6 flights will the physics just shut off now.

Before it seems as if it was every 2 flights that I would lose the physics. So its much better now, then before. But it still happens regularly.

Also just to let you know, certain stock parts make the physics turn off more often, for some reason.

One part that comes to mind is the "escape rocket". Sometimes I use it and BAM physics are gone. About 50% of the time.

Thanks

[ferram4]

Then provide reproduction steps for breaking that. I've barely been able to reproduce anything related to that issue, and no one can provide reliable steps that work 100% of the time for me (or at all, in most cases). I can't fix what I can't reproduce.

And if your reproduction steps are, "just launch something, and it breaks" then you really need to ask yourself if you've stated everything you did to cause the bug.

[blowfish]

Someone gave me a link about 3-4 days ago, to a version of FAR that is called FerramAerospaceResearchMaster

The part "Master" seems to mean it was not a normal release. So maybe it was a special release? im not sure.

The Github repository is updated frequently, every time ferram4 decides to commit a code change. Actual versioned releases come from the same codebase but are less frequent. Of course anything that isn't a versioned release isn't guaranteed to be stable or 100% tested, but if there's a critical issue it fixes then that may be worth it.

[mossman]

So, conclusion based on readings from your pictures is that rocket without fairings is quite more unstable than other one. As soon as you start engine it will start to backflip slowly until it falls into unrecoverable spin.

Second design with fairings, however will fly directly where you have pointed your rocket. It will fly straight until it runs out of fuel.

If you don't belive me, try to fly both rockets. On both rockets turn off SAS and dont correct steering at all from your input. Just add full throtle and fire up engines.

Do the same with other rocket and report here what has happened.

By 'Stable' I mean that the moment applied from aero forces will return the craft to 0 AOA. That's what a cm graph with negative slope means. The craft in the image flips over both with and without fairings, and the effect is slower without fairings.

Second design with fairings, however will fly directly where you have pointed your rocket.

This is incorrect.

[ferram4]

To nip this argument in the bud, mossman is right, the fairingless rocket is more stable, and the reason is that the fairingless rocket likely has a greater change in cross-sectional area further away from the CoM, which will move a large amount of body lift further ahead of the CoM, making it less stable. This seems about right for the rather bulbous fairing shape involved; a sleeker one will probably result in less instability. In any case, that's not really the point of fairings: the point of them is to protect the payload, not make the rocket more stable. If that happens, that's a wonderful addition, but in practice, the rocket can control itself and the main goal is to prevent parts of the payload from being left behind.

[Kitspace]

Wait is the no drag bug finally actually under control?

That would be so extremely nice.

[mossman]

Thank you for the concise and informative answer Ferram.

[ferram4]

Kitspace, I have no idea. I get conflicting reports and no reproduction steps at all, so for all the rare situations that I can cause it, it doesn't happen. Whether that means it's actually fixed or not, who knows, I've got nothing to work with besides guesswork.

[Solar71]

Then provide reproduction steps for breaking that. I've barely been able to reproduce anything related to that issue, and no one can provide reliable steps that work 100% of the time for me (or at all, in most cases). I can't fix what I can't reproduce.

And if your reproduction steps are, "just launch something, and it breaks" then you really need to ask yourself if you've stated everything you did to cause the bug.

Ok i'll try...

If you launch a rocket. (emergency rocket+mk1+kerbal+some fuel+4 small winglets+raliable rocket engine), then once you get above 10,000m

Goto the next stage (hit spacebar)

In that stage, have a de-coupling, and an emergency rocket.

Not 2 separate stages. Decouple, then emergency rocket. But the same stage.

Usually this will trigger the problem more often.

You hit space bar, de-couple and emergency rocket fire off at the same time. Then as the emergency rocket+mk1 come down they

often times, have zero drag.

PS: One other thing I noticed... In the FAR debug screen, the first TAB all the way to the left. You have a Preset settings or Custom settings.

When I use the custom settings I put Mach 1 drag factor 1.0 (usually this defaults to 0.7)

This seems to make the problem happen more often... (if I use a Preset setting like the one im using now, FULL drag with STRICT rules)

the issue seems to happen less often.

hope this helps.

[blowfish]

Ok i'll try...

If you launch a rocket. (emergency rocket+mk1+kerbal+some fuel+4 small winglets+raliable rocket engine), then once you get above 10,000m

Goto the next stage (hit spacebar)

In that stage, have a de-coupling, and an emergency rocket.

Not 2 separate stages. Decouple, then emergency rocket. But the same stage.

Usually this will trigger the problem more often.

I think you can do a bit better than that. Provide a craft file (since it sounds like you're only using stock parts), make sure it is reproducible on a clean save (or provide your save file only if it is not), and describe the ascent path you are using - are you just going straight up or using a particular path to orbit?

[Solar71]

I think you can do a bit better than that. Provide a craft file (since it sounds like you're only using stock parts), make sure it is reproducible on a clean save (or provide your save file only if it is not), and describe the ascent path you are using - are you just going straight up or using a particular path to orbit?

Please allow Ferram to speak for himself. ok?

Ferram didn't ask for a save file.

If Ferram needs a save file or a ship file I could give him whatever he needs.

But he should speak for himself.

Don't you think?

[steve_v]

This is either something that is now fixed in the dev build, or it's yet another rare variant. So precise reproduction steps are needed, the more detail the better, but keep it simple. It's been said before (and that gif is awesome) but this thread moves fast.

If you want to help fix it, see the github issue. Is this the same bug? Are you running a dev build with commit _bd6d274?

Since I can't reproduce this bug anymore (post fix, above) who else is still seeing it?

Edited June 12, 2015 by steve_v

[Wanderfound]

Please allow Ferram to speak for himself. ok?

Ferram didn't ask for a save file.

If Ferram needs a save file or a ship file I could give him whatever he needs.

But he should speak for himself.

Don't you think?

It is accepted practice here for the established users to do what they can to answer basic questions, so that Ferram doesn't have to spend all of his time doing so. And this is a basic question.

When you're filing a bug report, you need to provide reproduction steps structured like the method section of a scientific paper. In other words, steps that are so clear and unambiguous that a mindless robot could reproduce them. Not "do the usual stuff and it'll probably happen", but "set up an install with no unnecessary mods, take this .craft file, do these exact steps in this exact order, and it will produce the outcome shown in this log file 100% of the time".

As described in the permanently stickied post at http://forum.kerbalspaceprogram.com/threads/92229-How-To-Get-Support-%28READ-FIRST%29

[ferram4]

Solar, people like Wanderfound, tetryds, blowfish, and a few others make my life easier by acting on my behalf getting the stuff necessary to actually fix bugs and by providing common solutions so I can actually fight with the code. Even if they had no history of doing that, being obstructive when people suggest ways to provide more information to help does not make for easier bugfixing, especially since it means that I have to stop coding to address this. >_>

[FourGreenFields]

Ran into a bug with the Mk 2 cargo bay, using the latest released verion. Seems the bottom half of the bay didnt get any voxels.

Example craft file (with Mk 3 cargo bay too, which isnt bugged): https://www.dropbox.com/s/jnsitqjwq8nfacq/Bugged%20Plane.craft?dl=0

What I hope to be the output log or whatever you needed too: https://www.dropbox.com/s/tr1n05zko0bmww2/output_log.txt?dl=0

[Solar71]

Solar, people like Wanderfound, tetryds, blowfish, and a few others make my life easier by acting on my behalf getting the stuff necessary to actually fix bugs and by providing common solutions so I can actually fight with the code. Even if they had no history of doing that, being obstructive when people suggest ways to provide more information to help does not make for easier bugfixing, especially since it means that I have to stop coding to address this. >_>

All you had to do was ask.

In in your post, you said something like, " I need to know what you did to recreate the problem"

thats why I explained it to you.

I thought you wanted to know what I did verbally. i don't remember you asking for any type of file.

in fact. You didn't.

but it's ok now. I fixed FAR. The same way you fix bugs. I just deleted it, and called fixed.

Good luck dude. Tell FAR to say hi to the dodo bird for me when he gets there. Lol

[BudgetHedgehog]

Ran into a bug with the Mk 2 cargo bay, using the latest released verion.

Known issue, fixed in the dev build.

[Red Iron Crown]

I removed a couple of off-topic posts, please restrict your discussion to the mod itself rather than discussing the people posting in the thread.