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

[Master Tao]

Oops. Glad you both caught that. I'll fix it.

[Temeter]

Well, it depends on your design.

It's hard to say "this is what you have to do", in a way it will always work.

Normally you would want to increase the angle of attack of the wings, but, as I said, it can be or not a good idea.

Send a pic of your plane and it will be much easier to help.

Make the pics in the SPH, from top and side, with CoM/CoL/CoT indicators on.

However, most likely suspects:

* Too much drag.

* Insufficient pitch authority.

Drag increases and shifts backwards once you've gone supersonic; this tends to pull the nose down. Conventional control surfaces also lose much of their effectiveness. It tends to be worse on large planes, because they have more drag and lack the agility or TWR to overcome it.

All-moving surfaces (AV-R8's, canards etc.) near the nose may help, as may a general streamlining overhaul. As a last-ditch brute-force measure, a few Vernors under the nose may also come in handy.

So, i had some issue with the usual random crashs, but i've made a number of pics. Mind this is a test vessel, my first big arm spaceship. It's rough around the edges and I actually never really expected it to reach space but to randomly fail or prove unusuable like stock planes somtimes tend to do. That's also why the tanks run with half their fuel capacity (see last picture), and it doesn't use the Sabre intakes of B9 i didn't yet unlock. Inside are 42 tons of freight in the middle of the two storage's. The plane is generally nicely strutted and nothing moves by alone, the wings are 2 b9-style procedural wings and two B9 wingtips. Vernors for full articulation already built in.

Now i have trouble figuring out why that thing caused issues, though. Before doing the spaceflight, I added the big winglets and the small radial gimbal engines to get some more control. The inner produceral control surfaces clip somewhat into the plane when moving down, but that should afaik not be an issue when pulling upwards?

Javascript is disabled. View full album

Edited August 21, 2014 by Temeter

[Hodo]

So, i had some issue with some usual random crashs, but i have some pics. Mind that this is a test vessel, my first big arm spaceship. So it's rough around the edges. I actually never expected it to reach space (which it surprisingly did). That's also why the tanks also just run with half their fuel (see last picture), inside 42 tons of freight are in the middle of the two storage's.

Now i have trouble figuring out why that thing caused issues, though. Before doing the spaceflight, I added the big winglets and the small radial gimbal engines to get some more control. The inner produceral control surfaces clip somewhat into the plane when moving down, but that should afaik not be an issue when pulling upwards?

http://imgur.com/a/AUoye

I can see some minor issues that would cause some problems and some other issues I would fix for purely looks.

First it looks like you are not strutted enough, the body may flex from the side tanks causing some issues. The other issue I see is the control surfaces, I am not sure exactly how you have them setup but you may have to many winglets set to many things and you lack a real vertical stabilizer.

For aesthetics I would dump the quad coupler of intakes and replace them with a single SABRE-M cone intake, that is about the same intake area as the 4 RAM intakes you have there.

Also how is your fuel drainage, in otherwords what is your dry CoM vs your wet/loaded CoM?

[Temeter]

Thank you, i'll work of the individual points:

The body doesn't flex, it's actually surprisingly stable. I was careful about that stuff because of my experience with stock space-planes.

I also forgot to mention something, which might be part of my issue:

I changed the control law of some cotrol surfaces. The inner surfaces on the wings are set to pitch only, the outer ones to roll and the middle winglet at the back to yaw. Maybe it's just a silly mistake of the pitch surface's being to small?

The 6 other winglets have been added after the first flight(s) for more control. I assume you are talking about a 'real' plane-tail with elevators+rudder? Never really thought about it, since i was kinda used to elevon's. Might certainly be a good idea.

I didn't yet unlock the sabre-s+m cones (very close, though), but i'm planning to replace all intakes by sabre-stuff. Just looks so much cooler and a bit more efficient on the right angle.

Fuel drainage is neglible, at least for my issue. The CoM actually wanders behind the CoL with empty tanks, which would be pretty bad normally, but the air-powered flight sees basically no shift of the CoM. The 2 big 2.5m tanks are the only source of fuel, so the shift is very low and only appears after some time of rocket burn.

Funnily enough, i actually managed to land that thing on my second try, despite a screwed up dry CoM/CoL. Ofc still something which needs rework.

edit: Now i feel a bit silly. There is actually a part called stabilisator. Seems like i confused winglets with stabilisator. Yep, that's the kind of part i usually want for the plane. I'll switch them around and add 2 horizontal stabilisator's to the vertical one, just for test. Also made the elevon's bigger, just recently found out you can change their length.

edit3: Still the same issue, and weight distribution also doesn't change the plane trying to dive down. No change with high wing position and wheels zeroing out imbalance.

Edited August 22, 2014 by Temeter

[Renegrade]

I'm starting to think I should just strip out all the jet engine behavior and tell everyone, "You want jet engines? Run with AJE, welcome to proper jets and aerodynamics. Want magic thrust things that pretend to be jets? Stick to stock."

Well one intermediate thing that might help would be to put the engine-changing parts into their own little .cfg and just tell people who whine of nerfs to delete that file.

It could be like.. 'FAREngineNerf.cfg' -- they might even figure it out on their own then

(I happen to like the 50% setting. It has a similar-ish feel to AJE, if you don't get into #lolintake territory)

Pretty sure you mean *= 2.0 there. What you posted will cut the thrust in half *again*.

And speaking of #lolintake - didn't you mention something about an SJE at some point, Nathan?

(I'd take a stab at it myself but I only know the vaguest of details about jet engines --something about air going in one end and heat differences and thermodynamics and catching on fire and exploding -- and nothing about KSP modding)

[Wanderfound]

I'm starting to think I should just strip out all the jet engine behavior and tell everyone, "You want jet engines? Run with AJE, welcome to proper jets and aerodynamics. Want magic thrust things that pretend to be jets? Stick to stock."

The main reason I'm not using AJE is that I like it when other people fly my designs. If I swap to AJE, I'll massively narrow my potential customer base.

[NathanKell]

All SJE would do would be to make thrust be dependent on static pressure. That should be easy enough to make I think, a partmodule that OnStart grabs the max thrust of the ModuleEngine/FX it's bound to, and then sets it, OnFixedUpdate, to stored_value * vessel.staticpressure

[camlost]

All SJE would do would be to make thrust be dependent on static pressure. That should be easy enough to make I think, a partmodule that OnStart grabs the max thrust of the ModuleEngine/FX it's bound to, and then sets it, OnFixedUpdate, to stored_value * vessel.staticpressure

That's the easy part to do I guess. The hard part is Isp calculation. Isp curve should be dependent of speed not altitude, and as a matter of fact, most jet engine parts have poorly-defined Isp values.

The goal of SJE(BJE) should be 'work out of the box' without extra config. So what's in my mind is to read the velocity curve and guess the type of the engine, then choose from a set of pre-defined typical Isp curves and apply it.

I also plan on propellers, the general idea is to scale horsepower to pre-defined thrust (say 100kN=2000HP)

[Mister Spock]

I think I understand most of the fields in the FAR interface, but the "simulation" has me stumped. What do the values "init u", "init q", "init w", and "init theta" mean? Are they expressed in meters/second, or some other unit? Suppose I want to simulate how the aircraft behaves at Mach 1 with an angle of attack of 5 degrees; what values would I put into each of the four fields? And what would the shape of the resulting curves indicate? Is there a tutorial somewhere? I have similar questions about what kind of values I might input to run the lateral simulation, and what the output would tell me. I've read and re-read the included help notes, and while they do tell me what u, q, w, theta etc mean, they don't give me a sense of how to use or interpret the simulation tool.

Incidentally, I searched on "init" and I found one post, by dgriletz, asking these and other questions. Perhaps someone answered, but I can't find it. His post is here: http://forum.kerbalspaceprogram.com/threads/20451-0-24-2-Ferram-Aerospace-Research-v0-14-1-1-7-25-14?p=850936&highlight=init#post850936

[ferram4]

Those are the initial deviations of those variables (the same ones as used for the stability derivatives) from the steady state value. U is forward velocity in m/s, q is pitching rate in rad /s, w is downward velocity in m/s and theta is pitch angle above the horizon in rad. For the lateral sim, beta is sideslip angle in rad, phi is roll angle in rad, r is yaw rate in rad/s and p is roll rate in rad/s.

For simulating how an aircraft behaves at Mach 1 with an angle of attack of 5 degrees, you would not use that panel. You would instead want to go to the stability derivative tab, set Mach number equal to 1, and then vary temperature and altitude until you find that the steady-state AoA is 5 degrees. Then, if you want to see how it responds to a disturbance at that condition, you can move to the simulation tab to see how those variables change when disturbed from steady state. The main purpose of the simulation tab is just to see if the plane's oscillations die out in a suitable manner / if there are any blatant instabilities in the system.

[Hodo]

I'm starting to think I should just strip out all the jet engine behavior and tell everyone, "You want jet engines? Run with AJE, welcome to proper jets and aerodynamics. Want magic thrust things that pretend to be jets? Stick to stock."

This is one of those actions that would only make me mad at this point. I am fine with the 50% jet nerf, if people can't handle that tell them to get NEAR.

Really if you take it out I will adapt, again, if you leave it in fine. But at this point I think I and many others would be happier if you just stuck to one or the other and kept fine tuning the existing code to add more depth and realistic simulation for us.

Thank you Ferram4 for your work on one of my MUST HAVE plugins and one of 2 that I can't do without in KSP. Without FAR and DRE I would have quit KSP 11 months ago 30days after I bought it.

[Mister Spock]

Ah, OK; that helps a lot. I read keptin's tutorial, and he talks about the difference between static and dynamic stability (or terms like that, I think). So I'd start with the stability derivative tab (and my results there are encouraging, as they are when I fly straight and level at various altitudes and speeds). Then I'd switch to the dynamic-stability test (the simulation) and pop in one or more numbers (like a change in forward velocity in m/s), and see if the resulting curves dwindle over time, or instead whether they get worse over time. If they dwindle, then I've got what I want -- an aircraft that recovers from a disturbance. If they get worse, then I have an aircraft that might pitch over or spin or whatnot. Er, is that the right idea?

Off to give it a go. Thanks for a wonderful mod. Until I tried FAR, I hadn't bothered with a joystick; now I don't leave the tarmac without one.

[NathanKell]

camlost: for props you'd want to also use velocity curves, as otherwise things get *very* weird.

[tetryds]

So, i had some issue with the usual random crashs, but i've made a number of pics. Mind this is a test vessel, my first big arm spaceship. It's rough around the edges and I actually never really expected it to reach space but to randomly fail or prove unusuable like stock planes somtimes tend to do. That's also why the tanks run with half their fuel capacity (see last picture), and it doesn't use the Sabre intakes of B9 i didn't yet unlock. Inside are 42 tons of freight in the middle of the two storage's. The plane is generally nicely strutted and nothing moves by alone, the wings are 2 b9-style procedural wings and two B9 wingtips. Vernors for full articulation already built in.

Here are some things you can try.

I cannot guarantee that it will work, but these changes will at least make it respond better:

But above 20km altitude it's hard to say, but one problem can also be that you have your COT above your COM, and not enough air to hold the nose up.

If these problems persist with the changes above, try decreasing a bit the max thrust of the upper engines on the back.

Edit: when applying those changes, try to do them on the order I mention, then it's easier to see what's going on with the indicators and graphics.

And aparently all big planes will need some pitch up: http://i.imgur.com/LNVNujS.png

Edited August 22, 2014 by tetryds

[Temeter]

I had already did some of the stuff, like the removal of the useless middle stabilisator and modification of the rear tail stab's. Following the instructions, I now changed the outer control surface pitch, increased front stab size, and angled the outer wingtip. Also remove a number of b9 wheel's, so the CoM is now better balanced.

The result was quite promising, the plane's lacking aerodynamic still have been somewhat of an issue (and lol, turns out that thing needs some dedicated roll surfaces), but it was much better controllable, if a bit instable. Especially the big front stabs are a big improvement. Had to move the tanks a bit more to the front to balance the CoM, though.

I've had some suspicions about the wing-shape for a while, and your picture also pointed me in that direction. So i decided to give them a quick rework, making them wider instead of a delta-design, which also left some place for bigger control surfaces. Much more fragile, but also delivering more lift. Shown in the second picture.

The only remaning issue was the lack of powerful SAS. I missed that the plane still run on a single S2-SAS, since this was originally more of a mashup instead of a serious space-plane. Added a HL-SAS, but deactivated it for the test-flight, to get a better feel for the new wing-design.

In conclusion, the old wings - and original control's - probably just sucked above 20k and mach 4. The new design and improved control allowed me to bring that thing into space without using RCS for the first time. The HL-SAS should further improve the flight. Otherwise my piloting skills also need work, since i'm still relatively new to ARM.

Thanks for the help, i've learned quite a bit working on bringing that brick into brick. Maybe i'll even get to assemble that duna ship now that, the original reason for building the plane (and reason for the unimaginative name).^^

Pics of the reworked original and new wing design:

Javascript is disabled. View full album

Edited August 22, 2014 by Temeter

[tetryds]

Great that you figured out increasing wingspan makes it better on your own.

After a certain altitude your control sufraces won't help much, you may want to add some SAS modules.

Be careful though, when flying at low altitudes using SAS modules can be harmful, if you have too many.

The plane looks a lot better now.

Are you still running on half tanks? I believe that you can try adding more fuel.

On extreme cases you can use some engines controlled by action groups for extra control.

Or just bring some RCS thrusters.

What is the paylod it can bring to orbit?

Edited August 22, 2014 by tetryds

[Mister Spock]

Forgive me for another pretty basic question. At supersonic speeds, I sometimes see white shimmering around my aircraft. I'm using FAR, Deadly Reentry, and KSPI, among other mods. Is FAR generating the shimmering graphic, or is that a Deadly Reentry thing? And is the shimmering a warning to me that I'm experiencing instability, or verging on the edge of aerodynamic failure?

Sometimes the shimmering turns red; I assume that's Deadly Re-entry doing its thing.

Thanks in advance.

[ferram4]

@Temeter: You probably don't want any dihedral on that wing, since highly swept wings have a strong enough dihedral effect on their own. You'll simply make the plane very roll-happy at high AoA during reentry.

@Mister Spock: The "shimmering" is the stock "Mach" effects; they're supposed to be the type of cloud that shows up near Mach 1, but instead they're simply dependent on density and velocity for some reason. The same system is used for the reentry effects, and between whatever conditions are used to separate them they'll be "reddish" because it's trying to blend the effects.

In stock, the switchover is purely based on pressure. With DRE, it's hijacked and is dependent on velocity instead.

[Wanderfound]

Is there any way to quickly toggle on and off the effects of FAR on the CoM/CoL indicators while in the SPH?

It would be handy when designing to be able to quickly check that the aerodynamic balance is acceptable for both FAR and stock aero.

[ferram4]

No, there isn't. Because of the way that is coded for the stock winglet behavior, I can't just change the properties in the SPH; it would require you to back out and reload the entire GameDatabase to make the necessary changes to parts due to the fact that the stock lifting behavior is based on a legacy system of handling part behaviors.

[tetryds]

@Wanderfound: just open two KSPs and import the ship save to the stock one.

Very simple, and fast.

[Wanderfound]

My laptop struggles with one KSP, unfortunately.

[tetryds]

My laptop struggles with one KSP, unfortunately.

But it's only a matter of RAM, KSP freezes on the background if you use fullscreen.

Even with "simulate on background" checked. (it's broken)

[zwarag]

Is it possible to install this plugin on Linux?

[Mister Spock]

@Ferram4: Thanks for the explanation of the shimmering graphics.

I'm still learning to fly in FAR. I can get spaceplanes to orbit in stock, and I can get rockets to orbit easily in either stock or FAR, but with FAR flying a spaceplane into orbit is a more interesting challenge. Today I'm planning to try some of the aircraft included with FAR. Of these, are there any simple "trainers" that are good for practicing getting to orbit? Comparatively simple aircraft that can reach orbit? I'm using FAR, Deadly Re-entry, and KSPI (although I've changed rapier.cfg to rapier.txt in order to combat overheating of the RAPIER engine).

I've been trying with the spaceplane "SP+Explorer" that's included with SpacePlane Plus, but that aircraft has some instability supersonic, and I have trouble climbing with it past 20,000 meters. Not coincidentally, it generates one or two "red" results in FAR's build interface. Can anyone recommend a simple, stable "trainer" craft to practice flying into orbit with my combination of mods? Maybe one of Wanderfound's fleet? Thanks in advance.

Edited August 23, 2014 by Mister Spock