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

[taniwha]

Tavario's Pizza and Aerospace provides some nice wings and control surfaces... and further nurfs the jet engines (but gives ram jets, too).

[mhoram]

If you mean exploits, none that i am aware of.

You just have to know that it changes the game quite radically, as launching big payloads becomes a lot more difficult, but small payloads is easier as long as you keep a reasonable ascent profile. Also re-entry changes totally and becomes a death trap if your vehicle is not aerodynamically stable. In short, easy small stuff becomes easier, planes now behave as they should, rockets too, so for people aware of the rules of aerodynamics it's easier, for people just used to the stock game it's a lot more difficult.

Thanks for the insight.

I am aware of the fact that the physics are very different, so the FAR and Stock-only entries will be compared in different categories.

One consequence is payload fairings become a must have, there is no real way around them for many payloads, so that means installing another mod at least (proc fairings), also stock game lacks a lot of aerodynamic parts which are in turn provided by B9 for example (i suspect that B9 might become stock though..) or those procedural wings which are another must have with FAR.

At the moment I do not intend to allow payload fairings, because

- For the intention of the challenge the form of the payload is irrelevant - only the mass counts, so they may take any form you like. If you want to bring about 36 ton into orbit, you could take an orange tank and put a nose-cone on top - wouldn't this do the trick for FAR?

- I want to keep the challenge as simple as possible - meaning in this case as few AddOns as possible

- I have no experience with payload fairings yet

[Vrana]

One consequence is payload fairings become a must have, there is no real way around them for many payloads

Sure there is a way around it... Sacrificing ascent efficiency and saftey is one.

For example i recently lifted this monstrosity under FAR and DRE (stock Kerbin settings, not RSS) without fairings. The launch stage(s) was directly under the main tower, an NP 5m stack with 3.5m(or were they also 5m?) liquid fuel boosters which were suppoused to be an asparougus but due to instability i was uanble to stage the first tanks in low atmo.

Javascript is disabled. View full album

Good thing we were launching it into space and then setting a Mun transfer cause im preety sure my crew wanted to quit and forget about all about my space program after that launch. After about 200 m/s the thing started flailing like mad, any control input on my side before around 25km ment an immidiate loss of control and flip. This includes staging, i had to delay the planned staging until we cleared most of the atmosphere. Im not sure anymore but getting that out of atmo and circularizing around Kerbin probably cost me at least 6km dV.

So while its still possible to launch huge unrealistic constructions even without fairings it is certainly unadvisable.

[Frederf]

I really like the redo of the flight data section with units and sensible decimal places. If you want you can add the ° symbol in the attitude, AoA, and sideslip section to complete the motif.

[nothke]

Is there any extensive tutorial on how to make planes with FAR? What's important to care about and how to use the simulation curves. The moment I say, oh I get it! Something goes completely wrong.

But how does FAR work at all? It looks into collider shapes? Casts a camera from the front and looks how the face normals are pointed or...? This is important for me cause I am also making mods and I don't know how to make them be compatible with FAR

Edited November 24, 2013 by nothke

[ferram4]

@mhoram: Forbidding the use of payload fairings with FAR basically makes designing a properly stable payload above ~50 tonnes almost impossible; a payload fairing is much more aerodynamic than slapping radially-attached fuel tanks with nosecones on either end. You could allow payload fairings but require that they be jettisoned before figuring out the total mass of the payload, effectively balancing out whatever aerodynamic dV benefits that might come from them by reducing the payload fraction. Don't require them, but don't forbid them either.

@Frederf: I was meaning to do that, it just slipped my mind.

@nothke: The same basic principles apply, CoL behind CoM. The yellow moment curve is really the only one you want to seriously care about (except for very specific point designs, like a U2, where you would care about the behavior of the blue lift curve more). Just make sure that it slopes downward to be sure that your vehicle is stable. Make sure to check at different Mach numbers, since planes that are stable at subsonic velocities can be unstable at transonic (0.8 < M < 1.2) velocities, and can lawn-dart at supersonic velocities if not designed properly. Looking into flight dynamics and aerodynamic stability on Wikipedia should tell you all you need to know to be honest.

FAR simply checks all of the meshes and bases its performance on that, with the exception of wing parts, which have specifically defined parameters. The documentation for the wing parameters is in the readme.

[Surefoot]

For example i recently lifted this monstrosity under FAR and DRE (stock Kerbin settings, not RSS) without fairings.

That's entirely fine, as long as you were laughing maniacally all along. Otherwise i'd be a bit worried if i were you.

I manage to launch some pretty heavy stuff but it's always aerodynamic (or at least fits inside normally shaped fairings). Side effect is that most of my stuff looks quite neat (that screenshot folder is growing fast)... And can perform aerobraking too, usually it's just slapping some air brakes and the same stuff can aerobrake without flipping around since it was stable during ascent to begin with.

[DasBananenbrot]

When I use NovaPunch parachutes my vessel always flips completely over as soon as I gently turn it for gravity turn or even after a while on direct ascent with no control input it just flips. If I remove the parachute everything is fine again THis makes no sense to me as the rocket is one of your example vessels (so it should be aerodynamicallly stable) just with the NP parachute. If I remove FAR I can use the parachute again. So it seems somehow it is related to FAR.... Just for your info I also use stupid_chris real chute mod with mm file for NP which is edited correctyl for use with FAR but it may be the reason so I will ask there too. Any idea what is happening there?

[ferram4]

Obviously, what's happening is that the real chute is still applying drag through the stock model. FAR zeroes out everything that needs to be zeroed out for stock parachutes, but not for real chute. Try adding this to the end of the FerramAerospaceResearch.cfg file in the FerramAerospaceResearch folder:

@PART[*]:HAS[@MODULE[RealChuteModule]]
{
	@MODULE[RealChuteModule]
	{
		@stowedDrag = 0
	}
}

[AdmiralTigerclaw]

Tavario's Pizza and Aerospace provides some nice wings and control surfaces... and further nurfs the jet engines (but gives ram jets, too).

It's less that jet engines are nerfed, and more that they behave like they should.

Turbofans develop a LOT of power and use little fuel up to around supersonic before issues with the fan itself begins to cause performance to degrade until thrust and drag match and you can get nothing more out of it. But for powering subsonic heavy lifter... wooohooohooo... If speed isn't an issue, use turbofans. Raw power and economy for the win.

Turbojets (A basic jet engine IS a turbojet) are less powerful at low speed and suffer from lacking the full thrust they can perform. But they perform wonderfully from high sub-sonic to supersonic up to about mach 2.5.

After mach 2.5, the SR-71 shows us that a hybrid Turboramjet can take you all the way to hypersonic. Though with the J-57 engines, the turbomachinery was still running, so they had a limitation in the form of heat on the forward compressor fans from high velocity, which limited the blackbird's performance to a declassified Mach 3.34. Though between you and me and something a pilot once mentioned in an interview, I think they could hit mach 4 when the pucker factor was high... At the risk of a machine.

Ramjets start becoming useful around mach 2.5 to mach 3, and continue to be useful depending on a number of factors to around mach 6 or 7. After that, you about reach the limits of deflagration burning and you're getting too much drag just slowing the air down long enough to burn in the engine before it gets shot out the exhaust. (Effectively, the engine blows itself out like a candle.)

Once you reach that limit, the next point really is the Supersonic Combustion Ramjet Engine. It's usable from Ramjet starting velocities on up to at least mach 9 or 10 in proven tests. Though with the right mixtures, theories say you could get up as high as mach 22 before you reach a limit where the heat differential between intake air and exhaust air are so close together that you can't effectively heat the air to produce thrust anymore.

Problem of course, is that you have four types of engine with four barely overlapping regimes of flight. A turbofan's regime doesn't overlap say, a scramjet's regime. And stacking all four engine types adds a ton of weight to a vehicle. A turbo-ramjet hybrid is a good starting point, as is an afterburning turbofan... but getting all of that set up right can be a nightmare for engineering. Which is why we haven't done it yet. And of course, over with Scramjets, the limit right now isn't that we don't know how to build one or do the math for it... but that our materials science hasn't caught up with the engines. We're building working scramjets... They just disintegrate at Mach nine after 90 seconds of flight when the structural materials give out from flight stress.

But meh...

[DasBananenbrot]

Obviously, what's happening is that the real chute is still applying drag through the stock model. FAR zeroes out everything that needs to be zeroed out for stock parachutes, but not for real chute. Try adding this to the end of the FerramAerospaceResearch.cfg file in the FerramAerospaceResearch folder:

@PART[*]:HAS[@MODULE[RealChuteModule]]
{
	@MODULE[RealChuteModule]
	{
		@stowedDrag = 0
	}
}

thanks, will try it out

[DasBananenbrot]

Still got problems with said parachute from NovaPunch pack. But not with other Realchute parachutes. So it seems it isn't a RealChute problem. To be sure I will be recreating it without realchute at all. Will tell you tomorrow if I found the problem...

[LoSboccacc]

Fourth, don't have a crazy TWR. TWR over 1.6 is generally bad enough; if you're the type of person that starts with a TWR of 3 (yes, I have seen that in screenshots) you will have a very bad time.

ah!

goes supersonic at takeoff and flies up at 60° to reach an apoapsis of 100km.

and if you try anything briskly in flight, it disintegrates.

[taniwha]

It's less that jet engines are nerfed, and more that they behave like they should.

most interesting bits snipped

Indeed, I should have put quotes around the word, as it was meant relative to KSP's stock settings. I rather like working with TV's engines.

[nothke]

@nothke: The same basic principles apply, CoL behind CoM. The yellow moment curve is really the only one you want to seriously care about (except for very specific point designs, like a U2, where you would care about the behavior of the blue lift curve more). Just make sure that it slopes downward to be sure that your vehicle is stable. Make sure to check at different Mach numbers, since planes that are stable at subsonic velocities can be unstable at transonic (0.8 < M < 1.2) velocities, and can lawn-dart at supersonic velocities if not designed properly. Looking into flight dynamics and aerodynamic stability on Wikipedia should tell you all you need to know to be honest.

FAR simply checks all of the meshes and bases its performance on that, with the exception of wing parts, which have specifically defined parameters. The documentation for the wing parameters is in the readme.

Thanks, I managed to build a flyable plane =)

The only problem I have with FAR is the weird STALL readouts that are more like on and off actions than realistic stalls. Sometimes I fly right into the prograde on navball, but with "minor stall detected", I can barely change attitude, but I'm still flying exactly prograde until at one moment the plane suddenly gets out of the stall and now I can control it better. In other cases, I manage to get into stall in a perpedicular direction / 90 degree angle of attack (between prograde and retrograde pointers on navball) where the plane starts pitching into giant loops (cause of lift from the wings) and I can't do anything just wait for the stall to "turn off". So even in severe 90 degree stall I still get loads of lift from the wing even when at 90 degree.

And in other situation if I fly very slow like when approach to landing, the plane will start to fall like brick, but there will be "no stall detected" on the FAR display, although there is, obviously. (I solved slow lift with bigger wings).

Otherwise, it's much more enjoyable than stock aero model =)

Edited November 25, 2013 by nothke

[KerbMav]

I think the problem is the old lift inversion bug. Flip some aerodynamic parts upside down and it seems to screw up the lift calculation, but the UI will still make it look like everything's fine. This is easiest to see with the big delta wing; make a simple plane with two of those wings and it flies fine, but flip the wing upside-down and it's completely unstable, despite nothing changing with the center-of-lift arrow. In practice, it appears it's flipping the lift direction, and pulling the plane downward when it should still raise the craft. So I'm betting that the game thinks some of my wing components are upside-down and I'm losing part of the lift that would otherwise pull my nose up off the runway at 110m/s.

Really??

Do FARians have to deal with this though?

[DasBananenbrot]

Still got problems with said parachute from NovaPunch pack. But not with other Realchute parachutes. So it seems it isn't a RealChute problem. To be sure I will be recreating it without realchute at all. Will tell you tomorrow if I found the problem...

It seems I found the problem. The parachute (which is the Mk66 from NovaPunch, just so you know) has massive drag which could be explained by its form

http://imgur.com/haXn8aH

Well when I use a nosecone created with Procedural Fairings MOd everything works like a charm. No drag from the parachute and no crazy flip after start. But when I tried with docking ports on top which themselfs cause drag but not too much , the drag from the parachute fully kicked in and the rocket turned around after the smallest control input or after few seconds when reaching about 200 drag. I also tried it wiht the shielded Docking Port which is also used on your example craft (the FAR Artemis Transfer by the way)ut although it sits on top of the parachute, the parachute starts to reach drag of over 200 in a bout ten second which of course lead to disaster. It is good that I can resolve this problem with nosecone but I actually would like to use this parachute with the shielded Docking port. So I want to know, is this all intended and I have to accept it or is something wrong that it produces that much drag so fast even with ashielded docking port on top?

Edited November 25, 2013 by DasBananenbrot

[ferram4]

@nothke: Any situation where it seems like stalling is night and day is probably the control surfaces stalling, since they tend to stall more severely due to their greater angles of attack with respect to the airflow. If you managed to stall a plane and get it at 90 degrees with respect to the air, its main lifting surfaces will only be stalled if that's 90 degrees in pitch; if it's 90 degrees in yaw, but 0 degrees in pitch the wing will still make quite a bit of lift, but due to how the stall works it will be very uneven; the solution to this is a larger vertical tail.

The situation with the plane losing lift even though no stall is detected is not actually due to stalling; the plane is too stable and the elevators cannot pitch it up enough to produce the necessary lift. Your plane is basically lawn-darting.

@KerbMav: No, FAR users don't have to deal with that. Personally, I'm of the opinion that that bug has never existed and is simply the result of players blaming the mechanics for their own failures. Out of ignorance / pride.

@DasBananenbrot: Well, there's no picture there, so I'll have to investigate on my end. It might be that the NovaPunch parachute is messed up in some way.

[DasBananenbrot]

@DasBananenbrot: Well, there's no picture there, so I'll have to investigate on my end. It might be that the NovaPunch parachute is messed up in some way.

Ooops seems I messed up with the image .... Well, it seems the NovaPunch parachute is somehow messed up,if you have time maybe look into the all NovaPunch parachtues. I had some problems with the radial parachutes too. Will test later and iform if those are messed up too . I hope you may find something out

Edit: Updated the previous post with a link to the image, it should be working. At least it does for me

Update: I made a album with few pics.Just enlarge if you want to read the drag values

Javascript is disabled. View full album

As you can see only with the PF nosecone the parachute gets properly shielded and therfore no drag is produced. But if not shield it always rises fast, which make all the rockets with NovaPunch Rockets crash.... Edited November 25, 2013 by DasBananenbrot

[Roxette]

Is it feasible to have a function to be able to temporarily disable FAR during flight? It could make debugging ****ty rocket designs easier if problems can be isolated as aerodynamic in nature or not ?

[Senshi]

If you can't get a rocket to space, it's aerodynamics/FAR that causes it in ~90% of the cases. If it's merely a structural issue, you'll see explosions first, then flipping over. If it's FAR, you first get the flip, then the splitting/explosions.

[LORDPrometheus]

If you can't get a rocket to space, it's aerodynamics/FAR that causes it in ~90% of the cases. If it's merely a structural issue, you'll see explosions first, then flipping over. If it's FAR, you first get the flip, then the splitting/explosions.

Rockets only flip if your center of mass is to high this is a real world fact not an issue caused by FAR if your rockets fail due to flipping its a problem with the player not the mod

[awdAvenger]

I think I've found a bug with the Rockomax Brand Adapter 02.

When ever I use it as an adapter it seems to flip out the rocket a few seconds after launch. If i right-click it I can see the drag forces are pretty high for it, even though it should be shielded by other parts. See the linked screenshot for a test rocket that flips out 3-4 seconds after launch.

I am using deadly re-entry, but I don't think it has anything to do with it as the rocket only goes at ~60-70 m/s when it flips and when I remove the Rockomax adapter everything works as it should.

Steps to reproduce:

1. Build the rocket from the screenshot

2. Launch it with SAS enabled and at full throttle.

3. See it flip uncontrollably after a few seconds.

http://i.imgur.com/mXstVNB.jpg

[ANWRocketMan]

I think something is wrong... I had a failed launch and my command pod started flying! Pretty well too! I haven't upgraded to v0.11 yet, will test it in a moment.

It works normally again when I go retrograde. But it might have a little too much lift...

[ferram4]

@LORDPrometheus: No, rockets only flip if their center of mass is too low, which puts it behind the aerodynamic center. I really don't understand where this idea that the center of mass has to be low for rockets comes from; haven't any of you people played with darts before?

@awdAvenger: Sure, it should make a lot of drag. It's a sudden decrease in cross-sectional area; in the configuration that it's in, that part will cause an area of very, very low pressure to form behind it, which is what drag is primarily caused by for shapes like that. Frankly, for that design the short Rockomax adapter isn't the worst; it's the other adapters above it that are further from the CoM that will cause it to flip over.

Second "isShielded" is for parts in cargo bays / payload fairings. I should probably make that clearer.

@ANWRocketMan: Yeah, that's been fixed for such low speeds in v0.11. Granted, you can still get things like that to happen at higher speeds, but that's because at high speeds everything makes a decent amount of lift.