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

[Karretch]

That's really all that's required for a basic plane. Because examples are helpful, have a look at http://forum.kerbalspaceprogram.com/threads/90747-Kerbodyne-SSTO-Division-Omnibus-Thread?p=1961865&viewfull=1#post1961865 Pull that one apart and see if you can "reverse engineer" it.

That visual really helped. Also ksc123's explanations helped with the way I take info in. Still having trouble with constant pitch bouncing w/ sas on and nose diving with it off

Edited June 13, 2015 by Karretch

[kcs123]

That visual really helped. Also ksc123's explanations helped with the way I take info in. Still having trouble with constant pitch bouncing w/ sas on and nose diving with it off

It is kind of "normal" to have pitch bouncing effect. I (hopefully) will try to add more info in my thread about it. It is already written by ferram in FAR wikipedia about dynamic stability. I have already prepared some screenshots with FAR graphs to cover up that area too, but real life issues kicked in, so I didn't have time to edit those pictures and upload.

Stock SAS and some other mods, such as pilot assitant try to conpensate for that bouncing. Unfortunately, on some craft designs it leads to even more bouncing effect.

So, if your craft is more-less stable design wise, sometimes it is better to fly without SAS. Use ALT or whatever mod key is if you use linux or MAC and WASD keys to adjust trim control instead of SAS.

PID controler mods could help too because you can fine tune how SAS will behave (authority, damper). Also, thing that missing in my thread, on FAR graph there is pitch input.

Set it to "1", and run static analysis. FAR graph will going to show where yelow line cross X axis when you pitching up plane trough pilot input.

You should adjust authority on control surfaces in such way that when you have maximum pitching, you never reach AoA area where lines on FAR graph splits.

Meaning, you will not be able to push plane too hard where craft is in dangerous stall AoA area, no sudden changes of lift when craft stall and bouncing effect will be less noticable.

Try my Sporty Mk1 plane as example, all stock parts, no mods required for it, altough it is not optimized for wave drag effects it is nice stable plane to fly and simple to build.

[Nerd1000]

Yeah, but I'm playing a hardish career, I've only unlocked the mk 1 fuselage, and I can't afford to keep crashing and killing pilots

It is possible to fly aerodynamically at over 20km, even at subsonic speeds, with AJE installed (I've not tested, but you may get engine flameouts with stock engines). To pull it off you'll probably end up building something that has a lot in common with the Lockheed U-2: really, really long high aspect ratio wings and a high-ish thrust engine. The only issue is that because this vehicle shares many design traits with the U-2 it will also fly like a U-2, which means that the range of speeds you can fly at at high altitude will be very small- if you go too slow you'll either stall or enter the 'second regime' where drag is greater than thrust and increasing angle of attack will cause you to descend. If you go too fast you'll encounter transonic effects and probably go out of control (the cold air at high altitude has a low speed of sound and your long thin wings will cause you to have a low critical mach number, so transonic effects will be significant at speeds not far above your stall/second regime speed).

[TrainEngie]

I have a question regarding the cross section graph

Anybody, do you have any tips on making the yellow curvature-cross-section line on the cross section graph smoother?

Also, what does a smooth yellow curvature-cross-section line look like on the graph?

The most successful design I have in 1.0.2 FAR is the E-2 Hawkeye with FantomWorks, KAX, BD landing gear, yatta yatta. I'm too tired to think: It's midnight in Cali, US.

[entaran]

I don't know if this has been reported or not, but I'm encountering some SPH weirdness with COL calculations. Basically, the blue bubble doesn't update at all until the part is placed (as in it doesn't move as you move the part around to help you place it) and it makes no sense once it is place, move a wing forward and CoL moves back. Move Back, CoL jumps to middle or BELOW the airplane etc. It's generally in line with the wings or other parts but once you start adding multiple lifting surfaces it gets really weird, i've had the CoL showing in the "floor" of the SPH just before.

This is on a completely fresh install of 1.0.2 with only the contents of FerramAero download from Kerbalstuff and no other mods at all.

I made a test install because I started encountering this issue on my main save since updating to the Ferri release and I wanted to try to isolate what was causing it.

If there's any other information you would like, please just let me know (and possible a short idea of how to get it if it's not obvious)

[Kitspace]

Also a related thing which could possibly be an issue.

The issue with jumping graphs when replacing parts appears fixed. But still even the most minor changes in the wing position results in strange abrupt changes in stalling angles. If this is intended like this why is it so?

Also why does change in mass like fuel load affect pitch of the yellow graph even if the center of mass does not move?

[steve_v]

Anybody, do you have any tips on making the yellow curvature-cross-section line on the cross section graph smoother?

Also, what does a smooth yellow curvature-cross-section line look like on the graph?

Current experiments (modded, CTT career, hence odd part choices )) indicate sticking *stuff* (in this case an extended antenna & mk2 chines from Mk.2 Expansion) on the nose has interesting possibilities:

Loss of these parts in flight is

.

Edited June 14, 2015 by steve_v

[softweir]

I don't know if this has been reported or not, ...

It's been much discussed. Sarbian has threatened to remove the CoL bubble, because even when it works it is such a poor approximation of reality that it is useless. (It doesn't give any idea of how the CoL moves as you change airspeed or angle-of-attack or altitude.) Instead, he recommends we use the graphs.

[steve_v]

I'm finding the new blue line quite useful - If I'm reading it right that is Polishing up the nose shifts the peaks rearward, improving supersonic stability. Confirmed in flight . What does it actually represent?

- - - Updated - - -

While the COL marker being inaccurate is known, I'm certainly not seeing such extreme behavior here.

[entaran]

Using graphs is fine, except not everyone is an aerospace engineer. I mean, yeah we could use them and will to get past the majority of the issues, but still at least having a basic marker to help you approximate the design of the plane and THEN using the graphs to fine tune is probably a more "what you see is what you get" approach. There's a fine line between simulator and real life in some instances and being forced to check 4-8 graph results every time you move a wingtip is a bit rough.

The CoL marker should show you balance, I agree it has no use in super and transonic or aoa calc's etc, but it can tell you the basic tendency of your plane (pitch up, down, neutral, lawn dart, etc).

[steve_v]

True, but thanks to those graphs, most of what I build flies first time .

Rough out the airframe (I vote the COL marker stays, with it's flaws if need be), then focus on one issue at a time using the graphs.

When you find instabilities in flight, go back to the editor and simulate the flight conditions and control inputs that caused it.

Result: fewer rapid unplanned dis-assemblies.

[Max.10.07]

Why does it seem to not like TAC LS? Ill test more to see if it's some other mod with TAC LS and FAR.

[Naten]

I know that leading edge flaps work to increase curvature of the wing a bit, but do leading edge slots work? I mean, do slotted flaps and control surfaces, as well as fixed leading edge extensions with a gap between them and the wing, work currently? Thanks.

[JaredTheDragon]

Body lift - All parts lift: a fast enough brick will fly, if not that well.

Is this not evidence against the angle of attack theory of lift on a wing? Many planes (if not all) can fly upside-down, which negates angle of attack as the cause of lift on a wing.

Waxing theoretical, here. FAR is still awesome and the actual cause of lift need not be introduced, since the up-vector is the same regardless of the cause.

[tetryds]

Is this not evidence against the angle of attack theory of lift on a wing? Many planes (if not all) can fly upside-down, which negates angle of attack as the cause of lift on a wing.

Waxing theoretical, here. FAR is still awesome and the actual cause of lift need not be introduced, since the up-vector is the same regardless of the cause.

Lol, I never thought I would see this talk here.

No, that is not how it works.

Also, good luck trying to fly an airliner upside down.

[ferram4]

Umm... I think you don't understand what angle of attack is; it is the angle between the wing reference line and the incoming airflow. Lift due to angle of attack is the only reason (some) planes can fly upside-down. Also, in good news, it seems like mjn33 on github managed to find what might be one of the remaining sources of voxelization and aero breaking. So good news on that front.

[salajander]

I'm running a dev build downloaded last night.

Had my first encounter with the no-drag bug. I don't have reproduction steps (going to work on that now), unfortunately, so I expect this data may be useless, but perhaps something here may help in any case.

It was an early career first-orbit. I did a de-orbit burn, and time-accelerated through to the atmosphere interface, then re-time-accelerated again until I realize there was no drag.

I took a screenshot, and did a quicksave, and let the capsule go around to space again. When I reloaded the quicksave, the drag was still gone. Same a third time. Then I restarted KSP, and tried the quicksave again .... and of course the drag was there.

I've uploaded a screenshot and a quicksave while it was happening. There's also a log file, but that was after I restarted, and the bug had gone away. All available here.

Plug-ins enabled at the time:

ContractConfigurator

ContractPacks

FerramAerospaceResearch

JSI

KerbalEngineer

KSP-AVC

ModularFlightIntegrator

ModuleManager.2.6.3.dll

NavyFish

RemoteTech

Squad

ThunderAerospace

TriggerTech

WaypointManager

Edited June 14, 2015 by salajander

[Gryphon]

Also, good luck trying to fly an airliner upside down.

Well...that has been done, for certain with the

.

- - - Updated - - -

As a side note, Tex Johnston does describe how he did the maneuver, so that's something to try with your big KSP plane...

[tetryds]

Well...that has been done, for certain with the

.

- - - Updated - - -

As a side note, Tex Johnston does describe how he did the maneuver, so that's something to try with your big KSP plane...

It did not require lift being generated by the bottom of the wing as it was a 1g maneuver, a croscrew pitching up.

Sorry, I will not explain how lift is generated here, but yes it has to do with the angle of attack.

Airplanes made to fly upside down have no camber on the wings thus must have their nose pointed up to generate lift, the top and bottom are the same.

Wing cambering introduces a built-in angle of attack, that greatly enhances the lift proprieties at a given mach speed and altitude, at the cost of a bit of drag (but both L/D and L/W increase) and heavily penalizing upside down flight.

So, do not bring this discussion here, instead go check out the aerodynamic reason for the engines and weapons to be placed under the wings rather than above them.

Edit: FAR lift is generated just as it should be.

Edited June 14, 2015 by tetryds

[Jovus]

Nitpick that actually supports your point: some planes are designed with a negative camber.

[ferram4]

No no no! Wing camber does not produce built-in angle of attack. Instead, it shifts the angle of attack of 0-lift by deflecting air at the trailing edge downwards from where it was originally going, creating a lift force through the same mechanism that an increase in AoA does, but it is not built-in angle of attack. That's just making things more confusing for no good reason.

[jrandom]

No no no! Wing camber does not produce built-in angle of attack. Instead, it shifts the angle of attack of 0-lift by deflecting air at the trailing edge downwards from where it was originally going, creating a lift force through the same mechanism that an increase in AoA does, but it is not built-in angle of attack. That's just making things more confusing for no good reason.

Several of my aerodynamics books point this out emphatically in chapter 1. Seems there's been a ton of misinformation on camber circulating out there for decades and it just won't go away.

[tetryds]

Several of my aerodynamics books point this out emphatically in chapter 1. Seems there's been a ton of misinformation on camber circulating out there for decades and it just won't go away.

Indeed, time for cleaning up my sources, esp. the subsonic ones.

Also, my bad for disregarding saying the "works like if".

[salajander]

I don't have reproduction steps (going to work on that now)

A few hours later, no luck reproducing it. If I come up with anything solid, I'll definitely report.

[Wanderfound]

I have a question regarding the cross section graph

Anybody, do you have any tips on making the yellow curvature-cross-section line on the cross section graph smoother?

Also, what does a smooth yellow curvature-cross-section line look like on the graph?

The most successful design I have in 1.0.2 FAR is the E-2 Hawkeye with FantomWorks, KAX, BD landing gear, yatta yatta. I'm too tired to think: It's midnight in Cali, US.

This is a fairly slick (but not incredibly so, it's built for high-g aerobatics rather than pure speed) small Mk1 ship:

This is a tolerably slick (but still quite draggy, because big) large Mk3 ship:

In each case, you need to pay attention to the black horizontal lines; they provide scale. The yellow line for a super-slick ship may still look quite bumpy, because it's zoomed in a long way. Conversely, the line can be misleadingly smooth on a big, draggy ship.

In either case, though, the way to smooth out the line is to avoid sudden changes in cross-sectional area as you move from the nose to the tail. But because the distribution of that area in the cross-section is irrelevant, you can move from wide, skinny wings to tucked-in fuselage bumps without causing a lot of drag.

You can see a few tricks above:

1) The nosespike on the Mk1 ship. If you can extend the length over which an unavoidable cross-sectional transition occurs, you reduce the drag.

2) Smoothing the transition to the wings. On the Mk1 ship, this is done with landing gear, intakes and wing sweep. On the Mk3 ship, the canards and intake nacelles do the job (and, again, wing sweep).

3) Smoothing the post-wing transition. On the Mk1, the landing gear serve this purpose. On the Mk3, fuselage bulges, airbrakes and the nuke smooth the way into the tail empennage.

4) Vertical and horizontal stabilisers placed to jointly ease the stern drag. Small adjustments on these can have a large influence on drag; fiddle with your tail assembly until you've got it minimised.

5) To state the obvious, that many people seem to miss: you need to do all of this with your landing gear up​.