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

[NathanKell]

Camacha: Canards the KSP part (i.e. a control surface), not canards the type of forward-of-the-wing-stab-and-control-surface.

[Surefoot]

Canards (as a control surface) are not a sign of bad design, it's just a different approach and allow for high AoA control.

[Nemoricus]

That would be a relatively light vehicle (remember, it is empty) producing quite a bit of body lift due to its angle of attack. Yes, things other than wings can produce lift.

So this is something that you would see in real life as well? That seems quite bizarre...

[iornfence]

So this is something that you would see in real life as well? That seems quite bizarre...

I know pods like the apollo reentry vehicle had positive lift, so its not too uncommon.

[Nemoricus]

Oh, I know about the Apollo craft, sure, but what I'm curious about is whether you'd see an empty rocket behaving like this in real life.

[Sokar408]

Has anyone else experienced the "parachute craft glide effect"? Its has me so frustrating, as I thought this aimed to make the aerodynamics better. Craft with semi deployed chutes gliding in whatever direction you point towards needs fixing :/

Scott Manley experienced it aswell:

[AbeS]

@Sokar408 check this post by ferram:

TL;DR: Your vehicle is not stable for landing on an atmospheric body; the CoM is too close to the top and the CoL is too close to the back.

The main problem is that empty fuel tank on the bottom, which makes quite a bit of drag, particularly in that orientation. A cylinder one meter in length in cross flow (which is what you have going on there) has a drag coefficient of 1.2 based on its cross-sectional area. So let's do some math here.

I'm gonna assume you're at ~2 km for this calculation.

Drag Force = Dynamic Pressure * Drag Coefficient * Reference Area

Dynamic Pressure = 0.5 * Air Density * Velocity2

Air density is calculated from 1.225 * e- altitude / 5000m

This gives us an air density of 0.8211 kg / m3.

Dynamic pressure is then calculated to be:

Dynamic Pressure = 0.5 * 0.8211 * 602

Dynamic Pressure = 1478 Pa

Your reference area is the cross-section of that cylinder, which is:

Area = Pi * Diameter2 / 4;

Area = 1.227 m2

Drag coefficient of 1.2 for a 1 m cylinder, so extend that to a 2.5m cylinder, and we get:

Drag Force = 1478 Pa * 1.227 m2 * 1.2 * 2.5;

Drag Force = 5440 N

Now that's the drag that's being created by the fuel tank alone, and it's very far from your center of mass (keep in mind, your center of mass is about where the camera points if you don't move it). The experiments make similar amounts of drag, the landing legs don't make that much, but that's counterbalanced by their distance from the CoM. And if any of these were close to being edge on, you'd have to deal with the lift these cylinders create from being slightly off-axis.

Basically, the long story short is that those cylinders all make more drag and lift than you think, your CoM is higher than you think, and the pod torque is weaker than you think; your vehicle is in its proper stable orientation and it is not properly designed for landing on an atmospheric world. Add more weight at the bottom, lighten up the top, or put the parachute on a tower or something silly like that.

Go to the post and look for the vessel he is talking about, I bet it looks like your problem

[a.g.]

This is the same open attach node effect: basically, the open nozzle creates a lot of drag, which pushes the craft off-center; but the center of mass is very close to the engine, so the body of the craft acts to counteract this, and at a certain deviation it is stable with almost 1 L/D. Putting a nose cone on the engine completely eliminates this effect. If this is wrong, it might be a problem with how the drag from open nodes is computed.

[Camacha]

Camacha: Canards the KSP part (i.e. a control surface), not canards the type of forward-of-the-wing-stab-and-control-surface.

That makes sense.

Canards (as a control surface) are not a sign of bad design, it's just a different approach and allow for high AoA control.

Maybe I should rephrase that to that (the need for) canards in KSP generally indicate bad design - the extra wing in front of the main wing type of canards that is.

[AdmiralTigerclaw]

There are a couple of other things to consider as well in aircraft design.

Placement of landing gear in relation to the center of mass is one such issue to watch out for. Remember, the point of rotation (pivot) is the wheel. If the gear are too far behind the CoM, then it takes a very large of control surface action to get your craft to rotate for takeoff because all the weight is sitting out in front of a fulcrum like your arm outstretched carrying a 50 kg weight. This initial action may far exceed what the craft needs once its airborne. So the result of this action is that the craft refuses to pick itself up no matter how hard you pull on the stick, then as you hit a critical airspeed, control surface acquires enough force > the aircraft picks up > and lurches back before you can ease off the stick and you throw yourself into a stall almost off the runway. Putting canards on the front in this situation provides lift and control surface action out ahead of the center of mass to pick the nose up.

Of course, if your wheels are too far forward, you risk a tail strike. However, if your aircraft suffers chronic tail strike issues, then the problem may also be landing gear that are too short, or insufficient lift for your mass requiring a steep angle of attack (And the resultant engine smack.)

If you suffer roll stability issues, try checking how high your center of lift is. I've not skimmed much here to know if it's covered. So far I see CoL covered as far as how forward or back it is. But for getting the hang of things and producing a stable aircraft, try raising it. Note a lot a stable aircraft have wings that are high (Cessna Skyhawk, C-130 where the wings are on top), or a lot of aircraft have upward canted (V-shaped) wings. Raising the center of lift above the CoM helps. Of course, the mental image of a pendulum comes to mind, but that is incorrect. For more on that, just hit up Wikipedia and read up on relaxed stability.

Now, I have my own set of problems.

I have an aircraft that likes to fly SOOOOO much, that it refuses to come down without pitching a fit... and half its parts down the runway in a flaming ball of glory.

The problem?

It can't/won't SLOW DOWN.

Wonderful aircraft. Hits mach 4 at an altitude of 18 km reliably, handles extreme G-loads in hard, full thrust turns at 1500 meters without going to pieces at full fuel load, and

Now, here's the behavioral issue this thing has. No matter how I configure air brakes, no matter how deep I configure flaps, it refuses, in LEVEL FLIGHT, to come down from 120 m/s

I've configured this thing to open up like a pine cone and it doesn't seem to acknowledge the drag on any of these surfaces.

For Air Brake configurations I've tried the following combinations:

- 'Shoulder' air brakes (F-18E style)

---- Below critical deflection angle

---- Above critical deflection angle

---- Facing aft AND forwards

- Tail air brakes (dorsal and ventral positions)

----- Same as above

- Mid body side brakes.

----- Same as above

- Wingtip V-brakes (Two sets opening up and down at the same time...)

----- Same as above. (Produced lethal roll instability when open. Was not pretty. Not trying that again.)

I tried giving the flaps (large control surface towards the wing roots) very high deflections for settings hoping for deep-flaps drag effect for a nice, SLOW approach, but no such luck.

All forms of air brakes and flaps I've tried over the past... Let's see. Started at ten last night, stopped at six AM... eight hours of work yielded insignificant drag forces no matter how utterly absurd the surface became. I even watched the in-flight Cd readout and it barely flinched between 'clean' flight and braking/landing configuration.

My only success in this matter has been spoilers, by allowing me to dump lift and keep the nose up to slow down. But even then, I have to come in shallow, and I'm still coming in extremely hot to the KSP runway.

Now, I can't claim to know how FAR should be calculating drag, but I feel that something isn't working right with the drag calculations. Not when I can turn my aircraft into a pine-cone and it continues to fly like an arrow.

Inicidentally, I tried a parachute braking action after touchdown as well as using KAS anchors on winches to act as 'arrestor cables'. Let's just say that was... messy.

[BrickedKeyboard]

I'm trying to find out how much drag there is with different design variants of a basic rocket. Here's the rocket : I want to know how much those nosecones actually help (they are from procedural fairings mod, which claims to support FAR), and I want to know how much drag all those batteries are adding.

I did read the help documents, and I clicked static analysis, and told it to sweep the mach numbers. In theory, that should tell me how much friction there is at different speeds. In practice, no friction is plotted on the graph...

I would also like to find out if it matters if I place those batteries in-line with each other, such that the ones in the front should block the airflow to the ones in the rear, or scattered around.

[ferram4]

@AdmiralTigerclaw: And how much thrust are the engines making while you're doing this? They're still quite a bit overpowered compared to real life, so thrust should easily overpower any drag you create. What flight path angle are you taking when you do this? It doesn't really matter how much drag you try to make with it, if it's aerodynamic enough to fly it will easily pick up speed even in a moderate descent.

@BrickedKeyboard: That looks like it choked on figuring out the reference area; does the output_log.txt have any errors or exceptions in it?

As for some actual answers:

The procedural nose cones will reduce the drag compared to having a parallel setup without nose cones, but it will still make more drag than a purely vertical setup.

The battery placement won't affect the total drag, but it might affect stability, since battery placement will affect where drag is applied.

Making every stack equal heights will result in the vehicle being unstable due to all the drag at the top of it; shorter boosters will make our rocket less likely to flip out during flight.

Honestly, that doesn't look like a basic rocket. It looks like you took a small asparagus rocket and just slapped nose cones on the top thinking that would solve all the aerodynamic issues you have to deal with now.

[AdmiralTigerclaw]

I control-group tested it at Zero Newtons thrust dead-stick on a straight and level specifically to check braking actions; And at shallow approaches of five degrees. I understand why you have to ask of course. I'm coming out of nowhere and you're not sure if I know what I'm doing at all.

But yeah, I sanity-checked myself to the point of absurdity... And after eight (plus at this point) hours of constant testing everything from 'normal' braking designs to the 'pine cone' (Brakes... Brakes everywhere) layout and extremely deep flaps... The drag just isn't behaving as it should for such extreme configurations.

As an example, I tested the vehicle at engines idle (zero newtons thrust), flaps angled to 45 or even 50 degrees. What I would expect from a configuration like that is to have copious amounts of lift, but be forced to spool the engines up or stall in under a minute as the flaps would act essentially like giant half-wing sized air plows. (As per simulation experience with a number of crafts. Learjet, Citation, and Superbug.)

Instead, I ended up gliding for about twenty kilometers with the only slowdown being the gentle, level static-pressure related drop in speed as I approached sea level. In short, my vehicle was configured like a heavy business jet coming in for a full-flaps landing, and ended up gliding like a cross-country competition ultra-light. I also tested it light by dumping fuel and verify that it wasn't simply a mechanic of the inertia of a full fuel load.

That's all the information I have for you at this time.

[velusip]

208MiB of log file and growing fast.

When building in the VAB I'm getting lines like:

[LOG 02:18:59.443] M-1x1 Structural Panel Part Loc: (-0.6, 29.1, 0.1) Attach Loc: (-0.6, 29.1, 0.1) Dist: 0.0275
[LOG 02:18:59.443] M-1x1 Structural Panel Part Loc: (0.9, 27.9, -0.3) Attach Loc: (0.9, 27.9, -0.3) Dist: 0.0275
[LOG 02:18:59.443] M-1x1 Structural Panel Part Loc: (0.9, 27.9, -0.3) Attach Loc: (0.8, 27.9, -0.3) Dist: 0.0275
[LOG 02:18:59.443] M-1x1 Structural Panel Part Loc: (0.4, 29.1, 0.1) Attach Loc: (0.4, 29.0, 0.1) Dist: 0.0275
[LOG 02:18:59.443] M-1x1 Structural Panel Part Loc: (0.4, 29.1, 0.1) Attach Loc: (0.4, 29.1, 0.1) Dist: 0.0275
[LOG 02:18:59.444] M-Beam 650 I-Beam Part Loc: (0.3, 28.2, 0.0) Attach Loc: (-0.2, 28.2, 0.0) Dist: 0.5286794
[LOG 02:18:59.444] M-Beam 650 I-Beam Part Loc: (0.3, 28.2, 0.0) Attach Loc: (0.9, 28.2, 0.0) Dist: 0.5286794
[LOG 02:18:59.444] M-Beam 650 I-Beam Part Loc: (-0.6, 28.2, 0.0) Attach Loc: (-0.1, 28.2, 0.0) Dist: 0.5286793
[LOG 02:18:59.444] M-Beam 650 I-Beam Part Loc: (-0.6, 28.2, 0.0) Attach Loc: (-1.1, 28.2, 0.0) Dist: 0.5286793
[LOG 02:18:59.444] M-Beam 650 I-Beam Part Loc: (0.3, 28.2, 1.1) Attach Loc: (-0.2, 28.2, 1.1) Dist: 0.5286794

Of course Unity Engine is keeping it's own log with debug references, so that file is twice as large.

I'm guessing this has something to do with using both FAR and KJR? Maybe part detection when using SelectRoot (rebuilding the craft from another parent node)?

Edited November 8, 2013 by velusip
adding conjecturs

[a.g.]

@BrickedKeyboard: That looks like it choked on figuring out the reference area; does the output_log.txt have any errors or exceptions in it?

Note that the static analysis graphs in VAB are still computed as if the wind is coming from the door like in SPH. Also, the graphs ignore Cm and Cd when selecting the vertical scale, since they are intended for planes which tend to have more lift than drag.

After fixing that and some other discrepancies, I got the static graphs agree even better with the in-flight behavior for that gliding rockets effect. In the flight scene it's possible to see how lift and drag force centers are exactly opposite each other relative to CoM and thus precisely balance each other's torque.

It's not for me to say if the existence of this particular stable flight mode is realistic.

Javascript is disabled. View full album

[ferram4]

@AdmiralTigerclaw: You know what? Just post the craft and I'll take a look at it. No mods besides B9, correct? Worst case scenario I make some changes to increase drag.

@velusip: That's a FAR debug line that I forgot to shut off; it'll be removed in the next release.

@a.g.: It shouldn't be as dramatic as it is, but that should be a sort-of-stable configuration. The problem is the current version doesn't simulate the lift that should come from the air pressure on open attach nodes, only the drag; I'd originally had difficulty making command pods stable with that implemented, but I'm not sure exactly how I managed to fail at that, since the lift effect tends to bring the node to face directly into the flow. All that needs to happen now is for me to try and make sure that the potential and viscous lift of the rest of the body is correct and then, while this will still be possible, it won't be as... interesting.

[SFJackBauer]

Wait... how do I get to see the forces acting like the above shot from a.g.? My install doesn't show that button.

[AdmiralTigerclaw]

@ Ferram:

I think the gear is firespitter bomber gear (Only TALL landing gear I could get early in a 'traditional' tech tree campaign), but otherwise yeah, that craft should be all B9 and stock. I'll pull the blue lights off. Do you want the configuration in the pictures? Or do you want me to place it plain, or with a braking configuration?

[ferram4]

@SFJackBauer: a.g. has been reworking a lot of the flight GUI to fix errors and make it more useful. He's also worked to make the CoL indicator more accurate, and all of his changes will be in the next release.

@AdmiralTigerclaw: The configuration in the pictures should be sufficient. Doesn't B9 have larger landing gear though? Oh well, I'll figure it out.

[AdmiralTigerclaw]

Alright, here's a link to the craft file.

http://www./download/oosogkntch53fpo/Turbodash_X2_EXPORT.rar

I'll note in advance for you that you'll find the CoM element way far back. I forgot I was using a weighted brick (weighted bricks mod) on the nose (Really could use in-wing fuel tanks in KSP... I'll have to hunt and see if that's floating around). I pulled that brick off the craft so it didn't tag you by surprise. If you want to put it back, weighted bricks/counterweights or whatever the mod's called. Good, simple, useful little mod.

I also have new info that the actual 'air brake' parts from Tetragon Projects work effectively. (Unlocked them and stuck them on to test them... Worked quite well.)

As for your question about B9, I'm doing career mode, so the model I have in the picture is all I had unlocked when I built it. (Using treeloader and the 'traditional' tech tree that makes you start with only some rover parts and goes from there into aerospace. Haven't even hit rockets yet.)

[SFJackBauer]

Ferram and a.g. thats fantastic! Thanks for the work poured into this. Its epic!

[AbeS]

I have never experienced that problems you are talking about, and how come there are 187 pages of people talking about how wonderful this mod is, and just because you think it doesn't do what you would like it to, it is broken? Maybe your rockets are unstable or not aerodynamic at all.

Btw if the problem is wobbling you could download Kerbal Joint Reinforcement, which is another great mod by ferram.

But seriously, I don't see a reason to get upset without a reason! you could say this stuff and maybe ask about what you don't understand nicely

EDIT: Also, I can fly those planes that come with the mod without a problem, without using ASAS like you said.

Edited November 9, 2013 by AbeS

[ferram4]

@calkhi: Alright, here's what's happening:

If your rockets are wobbling apart, it means that you're overspeeding, your TWR is too high, and your rocket is being destroyed by aerodynamic forces, like it would in reality.

If your rockets are flipping out, it means that your rockets are producing more drag in front of the center of mass than at the back; you're trying to fly a lawn dart backwards, essentially; this also happens in reality.

If the indicator is stalling on your rockets, it means that you hit a high enough angle of attack that one of the fins / control surfaces you stuck on started to suffer flow separation, like in reality.

If you're pushing your rockets far off of prograde, they will suddenly suffer more severe aerodynamic forces, and possibly lose control, like in reality.

If you're pulling a plane up to an angle of attack greater than 25 degrees, it's going to stall the second you do that, and then probably crash, like in reality.

Honestly, your problems seem to be a combination of a failure to apply realistic aerodynamic principles to rocket and plane design as well as a large degree of pilot error. Honestly, it's to be expected, since stock KSP really punishes proper aerodynamic design and proper flight practices. You must unlearn what you have learned to fly realistic aerospace vehicles.

Sorry to hear you're willing to give up so quickly though.

[nater]

It seems a lot of the time it auto generates the config.xml file it corrupts or something, and the FAR GUI ceases to function, so if I delete config.xml it functions fine again until a while after another one is generated.

Edited November 9, 2013 by nater

[ferram4]

Is KSP not given proper write permissions in the directory it's in? If that's the case, then the config might not contain the proper values.

An alternative is that you're repeatedly placing the GUI somewhere where it can't visibly load; it's there, but it's moved off the screen. What exactly do you do that causes this?