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

[StevenRS11]

Hey ferram, two things. Is FAR feature complete to you, other than smallish refinements and fixes? It seems like it addresses just about every game relevant aerodynamic mechanic to me, except maybe boundary layer control for control surfaces. Or maybe it does, and I just dont build good enough planes where thats actually important.

The other thing- have you had any contact with squad about making this, or something related to it, the actual aerodynamic model in game? Other than very slightly reducing the feasabilty of some insane aircraft, it seems to make just about everything make more sense. It does kind of necessitate fairings of some sort, but who doesnt love the look of fairings splitting off into space, revealing a payload nestled inside?

[ferram4]

FAR is missing two gigantic features: induced velocities and local variations in dynamic pressure. Without induced velocities there are essentially no negative effects to canard designs and planes are slightly more statically pitch-stable than they should be, but are also less dynamically pitch-stable than they should be. Local variations in dynamic pressure would allow modeling deep stall for T-tail designs as well as better hypersonic modeling. Unfortunately, the only time I tried to model induced velocities in the current framework the lag went through the roof, so that basically necessitates a total recode to make things work better.

SQUAD hasn't gotten in touch with me, probably because they either aren't considering how they would overhaul aerodynamics yet or because they're intending to overhaul it to be something different than what FAR is or because they're confident of their ability to code an aerodynamic model without me. They'd also have to look into a complete rebalancing of stock parts, particularly part masses, if they were to switch over to something like FAR, since it's really difficult to make properly stable rockets with the over-mass engines. I'm open to coding something like that for them, but I know that the current priority is career mode, so there's probably no chance they'd even think about overhauling aerodynamics right now.

[Frederf]

One figure that I'm frequently desiring but not having is my terminal velocity or rather what the relationship between my current speed and it is. It would be great to know if I'm 80% TV or 120% TV during ascent to tune my ascent speed.

[Surefoot]

They'd also have to look into a complete rebalancing of stock parts, particularly part masses, if they were to switch over to something like FAR, since it's really difficult to make properly stable rockets with the over-mass engines.

Ha ! Talk about it My attempt to an Eve lander that can re-orbit does one (and only one) backflip through its ascent, ruining the already very limited DV (in regards to crazy Eve requirements). Trying to do an universal grand-tour SSTO lander (that refuels itself with Kethane on the ground), so far managed to make one work until Tylo, but Eve requires additional stages, and those are already quite crazy difficult with stock game, with FAR it's a real challenge Since every additional weight on the top stage means exploding masses for the bottom stage, and that universal-SSTO is around 83t (but has around 4000 DV by itself), lifting that off Eve surface (after landing successfully ! that means sound structural design) will require some creativity.

[ferram4]

@Frederf: Believe it or not, an estimated terminal velocity readout will be coming with v0.10.1, along with ballistic coefficient. That said, unless you're reaching ~400 m/s the second you leave the pad and increasing it up to orbital velocity around 20km, you're not anywhere near it. Trust me, trying to match your rocket's terminal velocity on the way up is not something you want to try doing with FAR.

[Keloooe_Pest]

Hey ferram, brilliant mod! One question though, my "Scout" plane wobbles on the roll axis uncontrollably with ASAS, is that what the dampers on the GUI are supposed to fix?

[ferram4]

SAS will override any of FAR's control systems, so if you have SAS on while you're using the built-in control systems it will be as if those systems aren't on. The reason that SAS is freaking out is because you have too many control surfaces capable of rolling your plane; reduce the amount of roll control that you have and the problem should go away.

[KerbMav]

As you are already working on an update again, can the automatic AG setting be removed in favor of UI buttons or made optional?

[Traches]

Hey ferram, thanks for all the great mods you make. You've added a lot to this game and to this community.

I've got a small request about formatting the numbers presented in the flight data window-- would it be too much trouble to put them in a more easily readable format? For example I don't need to know my pitch angle to 5 decimal places, but those extra digits make it a little harder to decipher the information at a glance. Units (where applicable) would help as well, and the biggest thing I'd like to see would be comma separators. Basically, it's a lot easier to read 123,456 m/s instead of 123456.78901 in a pinch.

[Frederf]

@Frederf: Believe it or not, an estimated terminal velocity readout will be coming with v0.10.1, along with ballistic coefficient. That said, unless you're reaching ~400 m/s the second you leave the pad and increasing it up to orbital velocity around 20km, you're not anywhere near it. Trust me, trying to match your rocket's terminal velocity on the way up is not something you want to try doing with FAR.

Huh sounds like I've been holding back on TWR needlessly to avoid severe drag (I see the white effects).

[Camacha]

I there a way to get a fixed number for a certain rocket in relation to drag/aerodynamic efficiency? I am trying to do a statistical analysis on the (lack of) performance of a kOS script and something like that would be very useful. I found a deviation that seems to be fairly constant and I would like to confirm whether that is due to aerodynamics or not.

[Tharios]

SAS will override any of FAR's control systems, so if you have SAS on while you're using the built-in control systems it will be as if those systems aren't on. The reason that SAS is freaking out is because you have too many control surfaces capable of rolling your plane; reduce the amount of roll control that you have and the problem should go away.

I'm having the same problem, but even with just two control surfaces.

That said, it's also happening in the stock game.

[123nick]

can you make a tool or utility or something that helps with winglets that are not compatible with FAR? like make a warning message popup if it detects non-far winglets or a fixing tool that makes them far winglets.

[ferram4]

@KerbMav: I'm going to admit that I don't quite understand what you're asking for. What do you mean by "AG setting?"

@Traches: What do you mean that you don't need to see down to seconds of an arc? I'll limit degree settings to 1 decimal place and see if it would be too bad to add comma separators.

@Frederf: Well, you have been avoiding extreme aerodynamic stresses and stability issues.

@Camacha: You can use the drag coefficient and reference area in the Flight Data readout; it'll change a little bit based on angle of attack and Mach number, but that's the best measure of aerodynamic efficiency. I have noticed that with RSS kOS doesn't seem to pitch down enough for surface velocity, but I think that might have to do with the control system itself.

@123nick: It's easy to tell if a winglet is compatible with FAR or not; most mods advertise their compatibility. Writing something to automatically get the proper wing parameters is something that I have planned, but it's difficult to make it work consistently and reliably.

[Camacha]

@Camacha: You can use the drag coefficient and reference area in the Flight Data readout; it'll change a little bit based on angle of attack and Mach number, but that's the best measure of aerodynamic efficiency.

Great, that is what I was looking for. The angle of attack will always be the same or very close and the mach number will remain low, so that is going to work out fine.

I have noticed that with RSS kOS doesn't seem to pitch down enough for surface velocity, but I think that might have to do with the control system itself.

What is RSS kOS?

[ExEvolution]

Can anyone help me solve this problem I've been having? Its all kind of explained in a video from my stream but basically the command pod for whatever reason was showing an unusual center of lift, making it want to flip so strongly during reentry that it couldn't be controlled.

http://www.twitch.tv/exevolution/c/3216708

[ferram4]

@Camacha: I was referring to the fact that my only uses of kOS were in conjunction with the Real Solar System mod, which means that if kOS' author made any assumptions about Kerbin being a certain size that might adversely affect the outcome.

@ExEvolution: First, you're using the old version of FAR; v0.10 makes pods much more stable during reentry.

Second, the reason that the CoL looks like that is because just showing the direction of "lift" or "drag" at any orientation is garbage for stability purposes, since a slight change in orientation could change the amount of lift or drag drastically. Instead, FAR makes the CoL display the direction and location of the additional force that would be added if you pitch up the vehicle just slightly; the stock CoL indicator is far more like a Center of Pressure indicator, which is not useful at all for calculating the stability of a vehicle.

[KerbMav]

@KerbMav: I'm going to admit that I don't quite understand what you're asking for. What do you mean by "AG setting?"

AG = Action Group

Some control surfaces add a "increase/decrease something" to AG 9 and 0 in the VAB. I guessed it would be FAR - but might be Pizza, B9 or Firespitter that adds this function to the fins - quite sure it is the first (career-wise) controllable stock winglet.

[ferram4]

Oh, that's for the flap settings; I have them set to that automatically. I'll simply change it so they don't get set to defaults then and anyone who wants to use flaps will have to manually set their action groups. I'll have to see if the action can be unavailable if the control surface isn't configured to be a flap.

[Camacha]

After not being able to play for a while and then being distracted by low-atmosphere technologies, I have just performed my first aerobreaking manoeuvre with FAR, DRE and AIES parts. Yay!

I tried to get the pod to fly a lifting trajectory by tilting it slightly upwards. Is that possible with the stock pod? G load did not exceed 3,5 g, ablative shield was used up for about 50%.

@Camacha: I was referring to the fact that my only uses of kOS were in conjunction with the Real Solar System mod, which means that if kOS' author made any assumptions about Kerbin being a certain size that might adversely affect the outcome.

Ah, thanks for that heads up, I hadn't thought of that yet. Currently I am using it with a normal sized Kerbal system. I found that the problem with my script is not related to aerodynamics and that the aerodynamics have very little impact in general on what I am trying to do. I will need to investigate further.

Edited November 21, 2013 by Camacha

[Frederf]

It is harder with stock SAS as it doesn't allow independent stabilization in yaw pitch and roll. Ideally you want roll stabilization only and use asymmetry in CG-CL to fly the craft passively. Lift vector is controlled by roll. One of the early DRE versions had the Mk1 pod set up this way by default but the aero was more of a hack.

All in all, yes, you can fly the Mk1 pod in FAR. Watch your L/D readout with regards to pitch. I find it has L/Dmax at around 25 degrees AoA. The trick is getting it to stay there as pod torque is weak and even if it was strong it drains batteries fast. I believe ballast radially attached would have to be far too massive to achieve L/Dmax so that leaves bolting on an aerodynamic appendage to shift the life vector.

I figure a Squad small control surface, programmed as a flap in FAR, could cause a Mk1 pod to orient closer to L/Dmax AoA naturally. I should really test this notion myself.

[Camacha]

I was flying the Mk1-2, the three Kerbal pod. Holding an attitude above prograde would work only if you locked it there and kept it there. As soon as you let it turn to prograde, that is where the ship is going to point due to the massive airflow. Power seemed to hold up though.

I never used the FAR compensatory mechanisms, I have never felt I really need them.

[KerbMav]

Huh? I should disable SAS when using the FAR support thingies?

[a.g.]

The way SAS is currently implemented, it only takes into account the actual user input, and overrides any input added by code that is before it in the callback list.

[KerbMav]

So, activate all FAR helpers (in atmosphere only?), SAS off (until in space?) and my control surfaces actually do what they are supposed to do?

Also: Is it possible that TWR should be kept below 2.0 while graviturning under 30.000m? Seemed to prevent the summersaulting my rockets liked to do lately.

Edited November 22, 2013 by KerbMav