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

[Hyomoto]

I know the assists aren't 'official' flight aids, but I figure it's worth bringing up anyways. I love using them like an autopilot, set a course (especially if it happens to be quite some ways away) and turn on the wing leveler. Then I just manage my pitch and altitude and enjoy the flight. However, there is an issue with where it comes to rest on a roll. I can usually counteract it by adding some trim opposite of it, but I find it odd the leveler doesn't do that on it's own. Especially if I level it out manually, then turn it on and it decides to come to rest with a list. I think the problem is planes with a tendency to roll but if I can trim it out, it seems like the leveler would too. You can highly exacerbate the issue by warping, in which I think you can see your wings bend more noticeably and that must contribute but I don't know if it is always the case at 1x speed. It doesn't always seem to roll the same way or even the same number of degrees even during the same flight.

I also would like to ask about the pitch aid. What exactly does it do? The wing leveler explains that it tries to bring roll to zero degrees, but pitch just says, 'stops the plane from pitching'. If you set a pitch and turn it on, it doesn't seem to try and hold that pitch, nor does it appear to level out to the horizon. So what is the pitch assist supposed to be doing?

Edited November 8, 2014 by Hyomoto

[noobsrtoast]

this mod is amazing, at first I was struggling with it but after I started to pick up how it worked it makes the game so much easier to play, launches are faster and take less fuel, planes that should fly actually fly, you can actually make big things that glide well, and when you get into a stall you can actually recover from it

[Gfurst]

@ ferram4, Van Disaster, Wanderfound: I'm well aware of how to lose speed, in fact I'm actually a pilot, not professional one, but still. And well aware too that even 100m/s is quite fast, 360km/h, 194nm/h, I understand that at this speed its pretty much a tube bullet cutting through the air. One common misunderstanding mos people is not aware in aviation, during more stable situations, we control height by the throttle, control speed by pitch attitude. For example if coming for a landing , you're too low, you throttle up and slow down the descent rate, if coming too slow you pitch down to immediately gain speed. I'm pretty sure this apply to most planes too, even airliners though I've never piloted them.

Another cool way of loosing airspeed while quickly descending, is gliding sideways while pointing down (or up to lose more airspeed), then you somewhat control your roll to maintain direction, this is pretty cool and works with FAR. I'm pretty sure you never see planes going S, crazy sideway climbs or whatever to lose airspeed. But again, this is for most general aviation planes, jet fighters and acrobatic can do that stuff, supersonic planes too must have quite different behavior.

Anyway I find it weird that even on low altitudes its very hard to slow down naturally, I literally just throttle all the way down, could cut off engines, and it would barely lose speed and altitude.

Below is a shot of my current working plane, its a pretty good design and flies wonderfully. Now I'm almost able to come down for final approach at about 60m/s, which is close to the stable control limit, but still too fast, often having too use the drag parachutes.

Another point too, flaps, while they drastically increase lift, and allow for lower angle of attack, should also drastically increase drag. For example on the shot, I'm only 70m/s and still I think an angle of attack as little as 1°.

Maybe I'm just misreading these drag values, afterall we can't compare KSP to regular planes.

Sorry for the bad shot, I thought I had another one from the profile.

On another totally different note, ferram, Do you take into consideration that mach speed depends on atmosphere density? What is that FRAC density value? Just really curious.

[Van Disaster]

@Gfurst: 95% sure that is also how you fly an airliner on an approach ( also they're occasionally known to forward slip, although I don't really want to be onboard when an airline pilot has to resort to that! that is a fun thing to do in a sim, a little more of something else in a glider ). I took a 30t plane near supersonic at 1000m and throttling off practically threw the pilots out the screen, but I agree that sometimes it does seem hard to lose speed when you're slower. I think a lot of it is that we have reasonably real-world weights and drag for aircraft that are probably somewhat undersized.

Flaps do seem to be a bit less draggy than I was expecting, but again see above I suppose. What I'm having a few problems with is stalling spoilers ( and also that on my install the stall tinting appears to have an occasional memory leak :S ).

FRAC density is just current density/sea level density I think.

Edited November 8, 2014 by Van Disaster

[HoneyFox]

I know the assists aren't 'official' flight aids, but I figure it's worth bringing up anyways. I love using them like an autopilot, set a course (especially if it happens to be quite some ways away) and turn on the wing leveler. Then I just manage my pitch and altitude and enjoy the flight. However, there is an issue with where it comes to rest on a roll. I can usually counteract it by adding some trim opposite of it, but I find it odd the leveler doesn't do that on it's own. Especially if I level it out manually, then turn it on and it decides to come to rest with a list. I think the problem is planes with a tendency to roll but if I can trim it out, it seems like the leveler would too. You can highly exacerbate the issue by warping, in which I think you can see your wings bend more noticeably and that must contribute but I don't know if it is always the case at 1x speed. It doesn't always seem to roll the same way or even the same number of degrees even during the same flight.

I also would like to ask about the pitch aid. What exactly does it do? The wing leveler explains that it tries to bring roll to zero degrees, but pitch just says, 'stops the plane from pitching'. If you set a pitch and turn it on, it doesn't seem to try and hold that pitch, nor does it appear to level out to the horizon. So what is the pitch assist supposed to be doing?

Leveler is a PD controller for 0/180 bank angle, with no Integrate portion it wont be able to handle your case perfectly. It will generate roll command base on your current bank angle and your roll rate, but if your aircraft is not balanced, it will finally stabilize itself at a bank angle at which the roll torque generated by P portion roll command just equals to your aircraft's imbalance roll torque.

Adding Integrate portion (which is not implemented in current FAR) and it will gradually 'trim' your aircraft, though it might take some time.

Pitch damper is a D controller for AoA. it will calculate AoA rate and generate a pitch command base on it and it only works when you are not pulling your stick. It is just a damper so asking for pitch angle stabilizing is a bit too much for it.

[camlost]

Hi ferram.

I've just sent a pull request to your git repo. See if you feel interested in those ideas.

This definitely needs some attention. Tweak the parameters well and many marvelous things can be done. For example, unstable aircraft

[camlost]

afterall we can't compare KSP to regular planes.

Do you have the exact weight, length, wingspan and engine profile of a regular plane? If not, they certainly don't compare to anything.

Do you take into consideration that mach speed depends on atmosphere density?

Yes!

[Hyomoto]

Pitch damper is a D controller for AoA. it will calculate AoA rate and generate a pitch command base on it and it only works when you are not pulling your stick. It is just a damper so asking for pitch angle stabilizing is a bit too much for it.

So, does that include trim? Either way, I'm guessing it's like the yaw dampener then and the only time I'd really see it 'work' is if my aircraft were oscillating on its lateral axis?

[DaMichel]

This definitely needs some attention. Tweak the parameters well and many marvelous things can be done. For example, unstable aircraft

Yeah this looks interesting indeed. I always wanted the tail plane to help with roll but the large control authority needed for pitch won't let me do this effectively.

[HoneyFox]

So, does that include trim? Either way, I'm guessing it's like the yaw dampener then and the only time I'd really see it 'work' is if my aircraft were oscillating on its lateral axis?

Nope... it doesn't include trim. Another case that you might see pitch damper working is, after you have pulled your stick and get a big AoA, you suddenly release the stick and the AoA starts to reduce, with pitch damper enabled, this reducing process will be slower and milder.

[ABZB]

I am getting errors of the following kind:

FAR Error: Aerodynamic force = NaN AC Loc = 5.26257513287772 AoA = 0

MAC = 1.388 B_2 = 1.311 sweepAngle = 6.12303176911189E-17

MidChordSweep = 27.6 MidChordSweepSideways = 0.111888451949879

 at R8winglet

Many, many lines of that kind of thing - is there some mod/Modulemanager edit that might have caused this? it was working fine a few days ago...

[ferram4]

@HoneyFox: I'll look into merging it. I'll have to see how it works out with the refactor I'm doing right now.

@ABZB: I'll need a mod list, a complete output_log.txt and full reproduction steps.

[LethalDose]

OMG I just learned wings' strength/weight is tweakable in FAR. The insane wing mass since 0.25 has been screwing up the thrust-mass alignment of my high-wing designs, and now I can fix it!

Seriously, though, how long has this genius feature been implemented?

[camlost]

@Ferram,

Is your new refactor code able to recognize part shielding as well?

[ferram4]

Considering I'm the one implementing part shielding, it would be pretty poor if it didn't.

[Wanderfound]

OMG I just learned wings' strength/weight is tweakable in FAR. The insane wing mass since 0.25 has been screwing up the thrust-mass alignment of my high-wing designs, and now I can fix it!

Seriously, though, how long has this genius feature been implemented?

Since about a week after 25 dropped. Tailfin weight balancing and the like allows all sorts of tuning trickery.

[HoneyFox]

Yeah this looks interesting indeed. I always wanted the tail plane to help with roll but the large control authority needed for pitch won't let me do this effectively.

With these new tweakables, I usually set rudder's control coefficients to Pitch 0%, Yaw 100% and Roll -30%. Here Roll -30% works as a simple Aileron-Rudder-Interaction mechanism to reduce sideslip angle when doing a roll maneuver.

Edited November 10, 2014 by HoneyFox

[HoneyFox]

@HoneyFox: I'll look into merging it. I'll have to see how it works out with the refactor I'm doing right now.

One note, the way I calculate the aircraft's current AoA for the new AoA% controll coefficient is not optimal at all. Every control surface calculates that value itself, although the results are all the same no matter which control surface does the calcuation. If possible, I would like to get FARControlSys of the entire aircraft and read its AoA value directly but I don't know if there's a simple way to do it.

Though normally there won't be too many control surfaces on an aircraft and the calculation of AoA is not too expensive.

EDIT:

I didn't add AoA% control response logic in the "Static" Sweep AoA/Mach function of the FAR Editor GUI yet. But guess that it's not very hard to implement it.

Edited November 10, 2014 by HoneyFox

[camlost]

@Ferram,

Is your new refactor code able to recognize part shielding as well?

What I mean is that people using multiple parts to build a convex, and assume the parts in the middle would be shielded?

[ferram4]

That is not going to happen. Using wings to create a fuselage is unlikely to get any support at any time.

[camlost]

That is not going to happen. Using wings to create a fuselage is unlikely to get any support at any time.

I hope so.

[Mekan1k]

That new code of yours- the one that generates a lifting value from the mesh- is there a chance it can be used to WRITE the FAR values into parts that don't have a FAR value?

That way, you build a wing that works in stock, and then add it to the gamedate folder, then activate KSP to have FAR write the value into KSP.

[ferram4]

That's not how it works. That's not how it works at all. All it's doing is getting the geometry from the mesh itself. So far, it doesn't do anything beyond that.

I'm not even sure what you want from it from that post.

[Mekan1k]

That's not how it works. That's not how it works at all. All it's doing is getting the geometry from the mesh itself. So far, it doesn't do anything beyond that.

I'm not even sure what you want from it from that post.

What I was thinking was that the code would check to see if the part had FAR values attributed to it already.

If so, then it would use these values instead of the mesh, to decrease required runtime.

If not, FAR would get the geometry from the mesh, and isolate the equivalent values that it uses in determining the lift, Center of Lift, Center of Drag, etc, and write the values into the CFG for the part. That way, if you use the part again, it could find the values without mapping the mesh.

Just my 2cents.

Any idea when FAR with the mesh-mapping thing will be available?

[endl]

i want to build an SSTO VTOL with interplanetary engines. any advice? would it make sense to go with sabres/nukes/none air breathing vtol? or is there some better approach to this scenario?

i would like to reach duna.