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

[0bj4ct7]

I'm sorry if this has been explained a lot (i couldn't seem to find it), but is there any fix for 0.22 SAS roll-spazzing the plane, or do we have to fly without it?

I've tried a lot of designs, stock, ferram and mine and all seemed to show some level of twitchy behaviour, while flying well (by my standards) without SAS.

EDIT: something like

but on a plane, which you can imagine is much more obnoxious. Edited November 12, 2013 by 0bj4ct7

[ferram4]

@AdmiralTigerclaw: Your vehicle needs more either more yaw stability or more struts. If it's the former, what's happening is that the vertical tail is becoming less effective as you go faster, allowing the plane to sideslip more; as it sideslips more the wings end up at slightly different sweep angles. Since greater sweep reduces lift, that causes a serious rolling tendency. Basically, slap a large vertical tail on it, see what happens.

@Obj4ct7: Use fewer control surfaces. There's no real way around the behavior without setting control surfaces to insta-deflect (which is something I really don't want to do, since it would make control surfaces stalling really bad). That said, most designs that have the problem don't need anywhere near the number of control surfaces they have.

[AdmiralTigerclaw]

@AdmiralTigerclaw: Your vehicle needs more either more yaw stability or more struts. If it's the former, what's happening is that the vertical tail is becoming less effective as you go faster, allowing the plane to sideslip more; as it sideslips more the wing slightly different sweep angles. Since greater sweep reduces lift, that causes a serious rolling tendency. Basically, slap a large vertical tail on it, see what happens.

@Obj4ct7: Use fewer control surfaces. There's no real way around the behavior without setting control surfaces to insta-deflect (which is something I really don't s end up atwant to do, since it would make control surfaces stalling really bad). That said, most designs that have the problem don't need anywhere near the number of control surfaces they have.

Ah, thanks. I'll try that when I get home tonight.

Also, would it not help Obj3ct7's problem to also turn down the deflection of surfaces and use the scaling option provided in flight assistance options? (One of my designs I actually improved performance and prevented it from exceeding a certain AoA by adjusting maximum deflection so that stick input all the way to max wouldn't exceed the rate of turn that kept it in stable flight.) I've noticed the roll wobble myself with several designs and that it would go away when I used stick pressure to override SAS. Figured it was just the computer being a bit too agressive at correcting itself. Turning down the deflection should also scale down the control surface tweaking.

[gunnyfreak]

wow, I thought it's just myself until I saw scott manley's video today... and read the comments

I just manual'ed it until I reach high atmos, will try scaling down deflection angle tonight (reducing isn't an option since the rocket already only has 3, having only 2 would be redundant...). I have already tried to turn deflection down to 10 though, maybe even less?

[redshift690]

wow, I thought it's just myself until I saw scott manley's video today... and read the comments

I just manual'ed it until I reach high atmos, will try scaling down deflection angle tonight (reducing isn't an option since the rocket already only has 3, having only 2 would be redundant...). I have already tried to turn deflection down to 10 though, maybe even less?

I'm not sure if FAR is conflicting with one of my other mods or what but I to am having the exact same issues as Scott in his latest video with odd phantom forces taking control of all my flights no matter how configure control surfaces. Though I am not having all that wobble and flexing like he was experiencing. I have built, and have launched many model rockets in rl, and so long as you have your stabilizer fins aliened properly that rocket will stay fling straight every launch even in moderate to strong winds. Sadly I had to uninstall FAR for the time being.. just to much frustration failing every launch.

[ferram4]

There were no phantom forces; the roll-wobbling was due to SAS being enabled and not properly handling the control surfaces. Scott's rocket was heading along the velocity vector every launch, he only truly lost control when he tried to bring it 15 degrees off of prograde and caused the fins to be overpowered by body lift. Basically, everything went as it should have.

Most problems that people have with launching rockets with FAR installed can be boiled down to:

1. Having a stable rocket, but making large enough control inputs that it flips out and becomes unstable.
   
2. Putting a tiny, ultra-light payload beneath a giant payload fairing so that it forces the CoL in front of the CoM.
   

@redshift690: Well, what kinds of rockets are you launching? Are you considering the mass of the payload at the top and making sure that it is large enough compared to the space it takes off? It is possible to make a rocket that is unstable, even if you put fins at the bottom.

[HoneyFox]

Most problems that people have with launching rockets with FAR installed can be boiled down to:

1. Having a stable rocket, but making large enough control inputs that it flips out and becomes unstable.
   

And these stabilizing fins may get into partial stall or at least lose some of their steering effect when AOA exceeds a certain value due to being in the "shadow" of the rocket body, which reduces the stability even more, right? I want to confirm if this is somehow simulated or not.

[ferram4]

Yep. Usually an uneven stall, which causes a massive rolling moment, and then inertial coupling between roll and the other axes causes it to drift further off. Fins are simply wings mounted at the bottom of a rocket; same principles apply.

[AdmiralTigerclaw]

Yeah, a lot of people try a normal KSP 'gravity roll' at 10 k. I find that a good ten degree 'kick' about three seconds after launch and letting it go from there works brilliantly. Just watch some NASA vids. Watch spacecraft almost immediately tilt as they clear the tower.

[Hodo]

Yeah, a lot of people try a normal KSP 'gravity roll' at 10 k. I find that a good ten degree 'kick' about three seconds after launch and letting it go from there works brilliantly. Just watch some NASA vids. Watch spacecraft almost immediately tilt as they clear the tower.

Pretty much what I have found also. That and throttle back, like the real rockets do to. I run usually 60-80% throttle till I complete my gravity turn then I kick it up to 90-100%.

[HoneyFox]

Pretty much what I have found also. That and throttle back, like the real rockets do to. I run usually 60-80% throttle till I complete my gravity turn then I kick it up to 90-100%.

I normally start my gravity turn by offsetting about 5 degrees when vertical velocity is approaching 60~80m/s, after that i control TWR to be at around 1.6~1.8 and keep pointing to surf-velocity prograde. Usually with some adjustment of throttle, i can reach 10km~12km when my pitch angle is around 45~50 degrees, and that is a sign of having a good ascend path for me.

If the pitch angle is out of that range, after the rocket has raised to upper atmosphere where the instability issue can be compensated by engine's TVC， i will pull up a bit with full throttle to extend the ascent if the pitch angle is too low, or pitch down to reduce some vertical velocity as well as reduce throttle if the pitch angle is too high.

[White Owl]

Yeah, get your initial angle off vertical well before the rocket hits mach 1, then it's very small corrections until you ascend through 30k-40k into the very thin air. Don't fight the aerodynamic forces until you're high enough that those forces are too weak to really matter anymore.

[CheesySquid279]

Hi! I am making a modpack of mods that improve the stock KSP experience! I was wondering if it would be possible to add your mod into the pack?

[ferram4]

@CheesySquid279: I'd prefer if you didn't, because then that gets into versioning problems. If I release a new version and you aren't up-to-date on making sure that the pack contains the most recent version, then people will continue to suffer from already-fixed issues, and then they'll all show up here to complain, and I'll have to tell them to download the most recent version from here. It's more hassle than it's worth, for everyone.

[GeneralIssue]

I keep having a problem where the 'FAR CAS' button in the assembly buildings disappears after reverting to hangar (eg after building something, flying it, then reverting to make changes). I think at one point I could delete the whole craft and the FARCAS button would be hiding underneath the Craft Name bar. I tried messing around with the window position in the config file, but so far no joy. Any solutions or things I'm overlooking?

[ferram4]

I made a change in the latest version that forcibly shifts the window down on load to prevent that button from getting hidden. You are using v0.10, correct?

[AdmiralTigerclaw]

Quick report on my aircraft I asked about yesterday. Was indeed insufficient vertical stability at speed. Adding a large fin smoothed everything out nicely.

Which leads me to another question.

Got any good information resources on aerospace design I can go study? Books, websites, or the like? Especially stuff concerning supersonic or hypersonic design. My google-fu for finding anything even remotely useful failed me in this regard.

[Pyromaniacal]

Hey, I have another question. I slapped on a single stock low end jet engine (the low altitude first one you get, my memory fades), and tried flying it. Is this small plane supposed to haul ass at 230m/s at 1km altitude, and glide forever? For wings, all it has is a pair of the wide, short stock wings, and the delta winglets with the control surface included. For fuselage, all it has is a stock liquid fuel fuselage piece, and 2 air intakes on the underside of the wings.

[GeneralIssue]

I made a change in the latest version that forcibly shifts the window down on load to prevent that button from getting hidden. You are using v0.10, correct?

Ah, I found it, it was floating over the parts list. Please excuse me as I go see an optometrist. lol

[ferram4]

@AdmiralTigerclaw: Well, to start off, I'd look at this digital textbook. It doesn't go too in-depth, but it does cover a lot of the stuff that goes on, and it is free. If you want crazy in-depth stuff, including supersonic flow, you should go here where you can find a link to a PDF download of the 1978 USAF Stability and Control DATCOM, which covers pretty much everything needed to get a back of the envelope calculation of how anything will behave in any type of flow. FAR doesn't model all of that stuff because it would melt your computer, but it is close.

@Pyromaniacal: You through an engine intended for a larger plane on a tiny plane; of course it's gonna haul ass. Most planes do actually glide quite well in real life; if anything, FAR over-estimates drag, so it should glide better in real life.

[English Mobster]

@Pyromaniacal: You through an engine intended for a larger plane on a tiny plane; of course it's gonna haul ass. Most planes do actually glide quite well in real life; if anything, FAR over-estimates drag, so it should glide better in real life.

I don't know about that. I just used made a rocket consisting of a probe body (Probodobodyne OKTO2), a regular ol' nose cone, and the large SRB (Rockomax BACC). Only mods are FAR and Deadly Re-entry. Rocket launched, spun wildly out of control, then when it ran out of fuel proceeded to gently rock back and forth as it very slowly glided to the ground. No aerodynamic surfaces other than the nose cone, so it should've fallen like a brick.

[ferram4]

Well, what speed qualifies as "slowly?" Keep in mind that each of the Space Shuttle's SRBs dropped to ~102 m/s just before the parachutes deployed, and those were much more massive (yet aerodynamic) than the BACC is. That is because you've got a giant empty cylinder (that makes quite a bit of drag) but nothing in it to keep in moving forward. Also keep in mind that 100 m/s is really fast; it's equivalent to 230 mph / 310 kph. It is by no means "slow."

Of course, if you haven't updated to v0.10, then it's due to the fact that body lift in earlier versions allowed rockets to be more stable in a lifting configuration than they should have.

[English Mobster]

I'm on the latest version. Once the fuel runs out, it is going roughly horizontal at about 400 m/s, losing altitude at about -10 m/s. It quickly slows down to about 150 m/s and begins to pitch up. As it pitches up, it slows down to about 40 m/s, at which point it runs out of momentum and stalls, allowing gravity to take over.

It accelerates from gravity to about -20 m/s vertical speed. Provided the fuel tank is still pointed horizontally (due to ASAS or what have you) it will start to pitch up again, slowing down to about 60 m/s and bleeding vertical speed. Rather than stalling again, it begins coasting roughly horizontally (akin to a glider), despite the fact that it doesn't have any means of generating lift. Its vertical speed is about -10 m/s, so it is going downward, but much more slowly than one would expect considering that it doesn't have any wings or propulsion. From here it handles very much like a glider -- you can roughly angle it in the direction you want to land and it'll head towards where you're pointing. You'll slowly pitch down and gain more vertical speed over time, but for the most part you can glide around until your electricity runs out.

He had a parachute deployed while I did not, but the behavior remained the same. Edited November 14, 2013 by English Mobster

[Senshi]

I just did recreate your described case 1:1 and launched it 5 times. No control inputs from my end. Depending on flameout altitude and speed, the resulting end orientation of the vessel differed greatly, which indeed had an important effect on the descent profile. If the vessel was near-vertical, it descended fairly quickly and after some nice mach effects during descent it impacted with around 100m/s.

If it was horizontal, it stayed horizontal and impacted with ~60m/s, but drifted much more horizontally. This behavior appears odd to me as well. I'd expect the vessel to be shifted to a vertical position during descent automatically due to the path of "least drag".

No audio in video:

EDIT: On 2 of 5 roughly the same of the above happend, in variations (one time its orientation was horizontal, but it slowly spiraled downwards instead of coasting in a single direction). The other times, the vessel stabilized in a vertical orientation, with aforementioned much higher speeds.

Edited November 15, 2013 by Senshi

[ferram4]

I'm gonna be honest, I think you're mistaken about using the latest version of FAR. I just tested it using v0.10 and while the SRB would tend to shift off prograde, and did produce quite a lot of body lift it slowed down to ~150 m/s after the booster ran out, afterwards maintaining a fifteen degree angle of attack as it gained vertical speed and plunged into the ocean. If it was behaving the same as in Scott's video, since he was using v0.9.7, and this was an issue fixed in v0.10, I have to assume that you were using v0.9.7 or that I built the craft wrong.

All of that said, don't sit there and say that an object without any wings has no means of generating lift. Even the most cylindrical of bodies will produce lift in an airflow if it's angled upwards enough; you can try this yourself if you stick your arm straight out the next time you're riding in a car. Also, if only wings made lift, the fine folks working for the USAF and Martin Marietta back in the 60's would have been incredibly foolish to build this beauty and drop it from a B-52.

Edit: Ok, I was looking at it flying forwards, which it does correctly; I never got it to flip around backwards. I see what the issue is; the SRB's attach nodes aren't set to be the correct size, so it's acting far more aerodynamic than it should be. You can quick-fix it by going and finding the config file and setting the last number on the nodes to "1" instead of "0", and it should work from there on.

Edited November 15, 2013 by ferram4