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

[ferram4]

Yeah, that all looks correct.

The problem is the fins are too small and the rockets are too short for how slow those rockets are going. No hope of making those stable. Vehicles in previous FAR versions didn't handle body lift properly and so were unnaturally stable; perhaps the drag on the front of this is somewhat off, in which case it's less stable than it should be, but most of the instability is due to the lift at the nose which is acting as expected.

The absurdly low TWR for a sounding rocket doesn't help. For something that shape you'd be looking for an initial TWR somewhere > 10 so that you fly in the direction you're pointed rather than five degrees to the side. Then you have a hope of it being stable, but it needs more thrust. Look at how the real-life WAC Corporal and Aerobee rockets were set up and flew, then you'll have an idea of what you need to do.

[zomos]

Still my planes using FAR and OPT explode. Even when I disabled the Structural failures due to aerodynamic stress. The plane choice was to totally disintegrate when I turned on the OPT Engine. Here the log please:

https://www.dropbox.com/s/65797etjd5eungf/output_log_ksp_opt.txt?dl=0

extreme amount of NullReferenceException: Object reference not set to an instance of an object it says about FAR. The lack of OPT Engine and everything works fine... The null exceptions of mumech are from thermalnuclear engines of other mod. Any clues ?

[p1t1o]

If you're throwing on the smallest, most blunt fairing you can, yes, the payload might be more streamlined, especially if it's a typical KSP payload, since then what you're putting in a fairing doesn't really need the fairing. If you're making a longer, more smoothed out fairing, especially one that isn't significantly wider than the vehicle, then you won't get more drag. If your fairing's radius is twice that of the rest of the rocket (read: you made the stock fairing as wide as you could), you're talking about taking up 4x the area of the rest of the rocket cross-section, which should be a sign that your fairing design is silly and will produce a lot of drag at transonic and supersonic speeds.

Finally, stability is not a sign that the fairing is poor at it's job. Its job is to reduce drag and to protect the payload; the engines have enough control authority to keep the rocket on course with thrust vectoring. Extra stability is nice, but a properly designed fairing will probably produce more body lift than the payload will and make the rocket less stable.

I guess I'll have to add some extra code to make unstreamlined payloads really come apart at high speeds then just to make things absolutely sure.

Pictures are helpful. Pictures tell us if you're doing something reasonable and that there's a bug or if you're doing something really silly.

Well my fairings are usually wider than the fuselage, but not often twice the radius. Perhaps it makes sense that fairings be quite draggy, given that they are large, but surely a slightly larger smooth cone is much cleaner than folded up bits of space gear and a probe body? And the body lift makes sense as well, but thrust vectoring is never enough, it seems, all my rockets need a good old helping of fins. Is KSP-FAR a more kinda "finny" universe down in the maths?

[blowfish]

Well my fairings are usually wider than the fuselage, but not often twice the radius. Perhaps it makes sense that fairings be quite draggy, given that they are large, but surely a slightly larger smooth cone is much cleaner than folded up bits of space gear and a probe body? And the body lift makes sense as well, but thrust vectoring is never enough, it seems, all my rockets need a good old helping of fins. Is KSP-FAR a more kinda "finny" universe down in the maths?

As ferram4 said, pictures would really help in this situation.

[Drew Kerman]

Yeah, that all looks correct.

The problem is the fins are too small and the rockets are too short for how slow those rockets are going. No hope of making those stable. Vehicles in previous FAR versions didn't handle body lift properly and so were unnaturally stable; perhaps the drag on the front of this is somewhat off, in which case it's less stable than it should be, but most of the instability is due to the lift at the nose which is acting as expected.

The absurdly low TWR for a sounding rocket doesn't help. For something that shape you'd be looking for an initial TWR somewhere > 10 so that you fly in the direction you're pointed rather than five degrees to the side. Then you have a hope of it being stable, but it needs more thrust. Look at how the real-life WAC Corporal and Aerobee rockets were set up and flew, then you'll have an idea of what you need to do.

Ahh, I've seen sounding rockets launched from NASA Wallops so I knew they can go fast, but I started in 0.90 with only having experience using larger vehicles and I actually got a smaller one-stage USI sounding rocket on a nice flight through the lower atmo with the low TWR, so I carried that over to 1.0.4. I will up the thrust.

honestly I feel a bit silly now thinking I shouldn't be trying to launch these like I see IRL when dealing with FAR

[ferram4]

@zomos: And as I already stated, I cannot reproduce any issues with just FAR and OPT. This means that any bug is either due to 1) different parts causing the issue that you will not repeat, or 2) another mod interfering, in which case you need to find out and talk to them instead. From all my tests, FAR and OPT play perfectly fine, though their wings lack any config data and so they will not produce wing lift at all. But that's not the issue you're talking about, and so is irrelevant.

@p1t1o: FAR is as "finny" a universe as real life once everything else is taken into account. If you have the same amount of thrust vectoring and stability (note: real life rockets are often unstable) as real life, you don't need fins.

Pictures dammit, pictures.

[cantab]

@p1t1o: FAR is as "finny" a universe as real life once everything else is taken into account. If you have the same amount of thrust vectoring and stability (note: real life rockets are often unstable) as real life, you don't need fins.

The small scale of Kerbin impacts here I feel. The payload in its fairing is going to have high drag-to-mass compared to the lifter stages which will reduce stability. Real rockets have small payloads on top of big launchers, KSP rockets have much bigger payloads compared to the launchers. (Though I do wonder if real hydrolox upper stages also have high drag-to-mass.) If you replicated a real rocket in KSP I would expect it to launch nicely - and take the payload to Duna and back!

[NathanKell]

cantab: They do, but they're generally of the same diameter as the rest of the rocket and therefore don't cause a bulge. Even so, they're dwarfed by the lower stage in most cases (Saturn IB was something of an exception here, with the S-IVB nearly as long as the S-IB).

[cantab]

Have you thought of putting in some sort of indicator that the Flight Assistance tools are active? More than once I've ended up chasing "phantom torque" on spacecraft that turned out to be because I'd left FAR's Wing Leveller on at the end of my previous playing session and the setting is remembered. With the FAR GUI closed there's no indication I've done that. Maybe a simple highlight or effect on the FAR icon in the toolbar would help.

[Darren9]

I seem to have a problem with B9-PWing after updating to the latast FAR, if I attach one to the end of a stack it becomes hollow, has anyone else noticed or have I got other issues?

[Crzyrndm]

I seem to have a problem with B9-PWing after updating to the latast FAR, if I attach one to the end of a stack it becomes hollow, has anyone else noticed or have I got other issues?

@ferram4

If it helps, the voxels aren't being created on the faces that have a MeshCollider attached (The wings use the mesh to create the voxels since the collider is incomplete)

@PART[B9_Aero_Wing_Procedural*]:AFTER[FerramAerospaceResearch]
{
    @MODULE[GeometryPartModule]
    {
        %forceUseMeshes = True
    }
}

[ferram4]

@cantab: Already is an indicator; the various buttons are different colors when you push them down to activate them. If you decide to hide the interface, the results of that are on you; there really isn't a good way to change the icon on the toolbar for both stock and Blizzy's toolbar, so that's not a viable option.

@WildLynx: Nope, not adding it. If it's not on by default, it won't get any use. If it's on by default, people will be ticked about everything shutting off when they don't want it to. Finally, I'll never use it, so why should I bother?

@Darren9: Is this reproducible in older versions of FAR, or was it introduced by the latest version?

@Crzyrndm: So for pWings, it needs to voxelize both colliders and meshes?

Dammit, right after I made sure that when meshes and colliders are voxelized separately to deal with the AJ10 Lack made for RO. So the solution right now is to either break compatibility with that, continue to break compatibility with thick pWings, or add some extra convoluted thing to handle both that will just make this even MORE complicated.

[p1t1o]

@p1t1o: FAR is as "finny" a universe as real life once everything else is taken into account. If you have the same amount of thrust vectoring and stability (note: real life rockets are often unstable) as real life, you don't need fins.

Pictures dammit, pictures.

Recently I have been learning that you can fly fairly unstable rockets quite effectively, I still need fins, but Im actually happier with the fairings now, with your explanation.

Its still a bit of a downer that you can operate just as effectively without them, but hey-ho

Another question:

Leading Edge Root Extensions: If I apply them to my craft, are the high-AoA benefits (vortex generation) simulated? Or are they treated as simply a bit of extra wing?

Thanks!

[Lack]

Dammit, right after I made sure that when meshes and colliders are voxelized separately to deal with the AJ10 Lack made for RO. So the solution right now is to either break compatibility with that, continue to break compatibility with thick pWings, or add some extra convoluted thing to handle both that will just make this even MORE complicated.

If it's something I can fix on my end with the AJ10, just let me know and I'll fiddle around with it.

[Crzyrndm]

@Crzyrndm: So for pWings, it needs to voxelize both colliders and meshes?

There is a mesh on the end panels/frames identical to all the other surfaces, just the end pieces (the ignored sections in the above picture) are the only parts that also have colliders (used for part attachment) hence why I was suggesting that as a possible cause of the issue.

On further investigation, this doesn't appear to be the issue at hand. With colliders removed the end frames are still being ignored.

[Darren9]

@Darren9: Is this reproducible in older versions of FAR, or was it introduced by the latest version?

Just the latest version.

[NathanKell]

Lack: per ferram's tests you already fixed it, looks like--you did rebuild the AJ10 collider in 22.x, right?

[Lack]

Lack: per ferram's tests you already fixed it, looks like--you did rebuild the AJ10 collider in 22.x, right?

Yes, gave it a new collider once you mentioned it. So now the non-convex form fairly closely matches the the shape of the engine for both variants (non-convex pic for aj10mid, bulges a fair bit at the top to account for the pipes and the likes), with it obviously simplifying a bit once you've set it to convex. Before it had the same problem as the cargo-bay and the old Bonanza cockpit.

[NathanKell]

Then you've already done your part and needn't worry a bit!

[ZobrAA]

Exception Detector says something wrong with FARGUI (FAR v0.15.5.1):

[blowfish]

Exception Detector says something wrong with FARGUI (FAR v0.15.5.1):

http://i.imgur.com/NAc4iN1.jpg

Post your output log - only way to see the full stack trace.

[ZobrAA]

Post your output log - only way to see the full stack trace.

Ok, once again here is Bug on FerramAerospaceResearch.FARGUI.FARFlightGUI.FlightGUI.FixedUpdate : output_log.txt

And this time with dead freeze on the atmosphere border...

[ferram4]

@ZobrAA: I need reproduction steps and a full mod list of only the mods that need to be installed to create the issue. I cannot reproduce it at all with FAR alone, and the current code makes it impossible for an exception to occur where it says an exception is occurring unless something reaches in to what FAR is doing to set something null. There are only 4 objects that can be null in that function, and all 4 are created and given references at the same time, and all 4 are reset to null at the same time within FAR itself; this points to something else interfering somehow. How, I have no idea, but I need a lot more info than just a log.

[ZobrAA]

this points to something else interfering somehow. How, I have no idea, but I need a lot more info than just a log.

Yeah, definitely. But more info will are problematic - I have about 203 mods installed :/

If I find out which of them are interfere, I'll make a report.

PS: Dead freeze when crossing atmosphere sometimes appear even without FAR so false alarm at that one.

[123nick]

Yeah, definitely. But more info will are problematic - I have about 203 mods installed :/

If I find out which of them are interfere, I'll make a report.

PS: Dead freeze when crossing atmosphere sometimes appear even without FAR so false alarm at that one.

do you have a list of the mods? just wondering. on another note, how would one go about installing the full realchutes along with FAR? just remove the realchutes lite that comes packaged with FAR, or keep it?