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

[DeathSabre]

Hey Ferram.

Thanks for a great mod. makes flying planes so much harder. but I'm sure it'll get easier once i figure out all the settings and stuff for control surfaces and the like.

made a plane that achieved Mach I before the end of the runway which was funny. couldn't land it tho.

Thanks for replying.

I have build almost 20 planes since I got the mod, and all of the ones with more than about 5 pieces fabricated into a wing had an offset COL. I was experimenting and found that it is completely random. It can be going well for the first 3 pieces, but at the fourth the COL kicks right about a meter, and upon removing the fourth piece, it stays about 3/4 of a meter to the right.

I'm Having this Problem as well. My plane is much smaller. and the problem seems to occur more often with the long plane wings(can't remember what they called. i'm at work).

I also have the B9 pack installed. but my plane isn't using any of those parts(only Stock Parts at present). i find that removing the offending (if you can locate which one)wing and re-adding it will usually solve the problem. but it can go off-center again by adding a fuel line or something totally irrelevant in the SPH.

I've only encountered this glitch when i have wings touching or overlapping each other. but i only installed this mod 2 days ago.

I'll post my craft file when i get home for you. another thing if noticed with this glitch. in the SPH it looks like COL is off to the left. but when i fly the plane it pulls left and rolls left indicating the COL being offset too the right...

[taniwha]

Great mod: it's made my rocket launches a pleasure (once I figured out how to deal with ASAS, fins, etc). FAR + Deadly Reentry makes for deadly ascent, too . However, space planes are another matter, but I'm slowly figuring things out.

In the description for Cm, it says that so long as Cm is decreasing for increasing AoA, the plane will be stable, but what about the bumps that seem to be caused by non-square wing parts during stalls? Is it really the case that so long as Cm is negative for all AoA (or really, Cm * AoA < 0, for pitching the other way) the plane will be stable?

[DauntingFlyer]

I'm having trouble making this mod work. I tried making the rocket test with and without nose cone like in the video but the noseless rocket always wins by a long shot. The the plane parts in general I got it and they work but the nose cones seem unaffected by the mod. I checked what I installed and there's not one file that replaces the nose cones cfg. files. Am I missing something here?

[ferram4]

@DeathSabre: Okay, thanks for the info; if it is caused by wings overlapping or a silly interaction with struts / fuel lines, that should narrow things down. I think the main problem in the editor is just a visual bug; as for flying, try adding more struts to keep your wings in place and see if that helps. KSP has this weird issue where joints can flex unevenly under equal loads if the joints are oriented in different ways.

@taniwha: Any region where Cm * AoA < 0 indicates a stable AoA region for the plane; the situation you're describing basically indicates that the plane is stable until it starts to stall, at which point it becomes unstable. If you're certain that the plane won't be capable of stalling itself, then it shouldn't be a problem. If you want to make sure that it is stable during a stall make sure that your forward wing / canard has a higher aspect ratio than the rearward wing / tail; that will cause the forward surfaces to stall at a lower angle of attack than the rearward ones, making the plane stable. FYI, aspect ratio is equal to the wingspan / wing chord (distance from leading edge to trailing edge).

@DauntingFlyer: It's possible I screwed something up during the drag update; I'll check it out. The plugin will properly account for nosecones, the only part.cfgs distributed are for stock parts that need special overrides; nosecones don't require that.

[DauntingFlyer]

Thank you for the quick reply. Maybe I'm doing something wrong. Do I need to change any settings in the new tables that come up in the space center to make the nose cones work? Or are those just for the wing parts?

I tried messing with the .cfg files of the cones but to no avail.

[ferram4]

Don't bother changing anything in the nosecone cfgs; you shouldn't need to mess with anything to make it work properly.

[taniwha]

Thanks. So, so long as the bumps don't cross the 0 line, I'm ok?

Hmm, also, thanks for the starter on what to read (I already have wikipedia's aspect ratio page open )

[ferram4]

If you want some more in-depth stuff, I'd say to take a look at this: http://www.desktop.aero/appliedaero/preface/welcome.html

It contains math, but will tell you a lot more than wikipedia will.

As an update for everyone, the payload fairing bug has been fixed, as well as the attach node bug DauntingFlyer found. The next version should be out as soon as I track down the CoL display bug.

[taniwha]

Heh, yeah, just the table of contents seems to be extensive than wikipedia. Thank you.

[a.g.]

As an update for everyone, the payload fairing bug has been fixed, as well as the attach node bug DauntingFlyer found. The next version should be out as soon as I track down the CoL display bug.

I sent you a fix for air intake percent calculation and a couple of other small tweaks some time ago via pm - did you receive it?

[Laphtiya]

You won't need to reinstall the whole game.

Awesome thanks a bunch.

[ferram4]

@a.g.: Yep, I got it; haven't added it to my version yet.

[Bluegobln]

How does one prevent one's rockets from tumbling end over end. This mod is ruining the game for me, and I am about to remove it, which is sad. I've tried redistributing fuel. I've tried making craft more aerodynamic. I've tried balancing, massive clusters of RCS, winglets in all kinds of different positions. I've tried slow ascents, fast ascents, staging, non-staging, small craft, large craft, more thrust vectoring, less thrust vectoring. I've done everything I can think of to discern the source of this issue, and it seems to me the issue is the mod does not work properly.

Real rockets do not tumble end over end like this in EVERY SINGLE CIRCUMSTANCE AND SCENARIO.

The only thing I have not tried is putting my rocket engines up top and on the sides and TOWING my payloads into orbit. This is just plain ridiculous.

If you can explain it, I'm happy to learn it, but as I see it there are things severely wrong here.

[camlost]

How does one prevent one's rockets from tumbling end over end. This mod is ruining the game for me, and I am about to remove it, which is sad. I've tried redistributing fuel. I've tried making craft more aerodynamic. I've tried balancing, massive clusters of RCS, winglets in all kinds of different positions. I've tried slow ascents, fast ascents, staging, non-staging, small craft, large craft, more thrust vectoring, less thrust vectoring. I've done everything I can think of to discern the source of this issue, and it seems to me the issue is the mod does not work properly.

Real rockets do not tumble end over end like this in EVERY SINGLE CIRCUMSTANCE AND SCENARIO.

The only thing I have not tried is putting my rocket engines up top and on the sides and TOWING my payloads into orbit. This is just plain ridiculous.

If you can explain it, I'm happy to learn it, but as I see it there are things severely wrong here.

If you can elucidate your problem everyone's happy to help. Did you try the ferram's example rocketes in the pack?

[Taverius]

How does one prevent one's rockets from tumbling end over end. This mod is ruining the game for me, and I am about to remove it, which is sad. I've tried redistributing fuel. I've tried making craft more aerodynamic. I've tried balancing, massive clusters of RCS, winglets in all kinds of different positions. I've tried slow ascents, fast ascents, staging, non-staging, small craft, large craft, more thrust vectoring, less thrust vectoring. I've done everything I can think of to discern the source of this issue, and it seems to me the issue is the mod does not work properly.

Real rockets do not tumble end over end like this in EVERY SINGLE CIRCUMSTANCE AND SCENARIO.

The only thing I have not tried is putting my rocket engines up top and on the sides and TOWING my payloads into orbit. This is just plain ridiculous.

If you can explain it, I'm happy to learn it, but as I see it there are things severely wrong here.

Ehm ... rockets are unstable. They tend to have all the weight at the back, but the CoL/CoD further 'up'/'forward' than the CoM. Which means yes, they do want to tumble. All the time.

Things you can do:

• Winglets at the back. Mind that too many winglets will make it too stable and hard to turn.
  
• Start the gravity turn EARLY! This is not the stock drag model, start turning right as you're clear of the launch tower, generally. Like real rockets.
  
• Do not go more than 5 degrees off from your velocity vector. Geeeeeeentle turns.
  
• If you're using Mechjeb ascent, make sure you have a gentle turn slope. A hard turn means you will soon be pointy-side down and will not go to space today.
  
• Do not asparagus. Asparagus only works in stock model. Unless you're lifting very heavy loads a single stack is bettar.
  

Main thing to keep in mind is you have to start using the CoL/CoM indicators in VAB too, and ideally learn to use at least the static analysis tab of the FAR panel.

If you do keep that in mind as you (re)design, you can go to space quick and easy for way less dV than stock.

This is my main satellite delivery vehicle.

See how much like a real rocket it looks like? That's a good guideline in FAR.

[Bluegobln]

What if your payload is non-circular, something odd shaped, or can't be made to fit in a faring? Are you boned? Try to put farings around it? Must everything I build fit in a faring or be made aerodynamic?

Asparagus could be one of my problems, I usually use some form of additional liquid tanks and engines with decouplers around the main, but I don't see how that might cause the tumbling by itself. Doesn't an array of engines with good vectoring typically INCREASE your control and stability? *confused as hell by that* I can see how it could be a problem of drag though. Will extra engines mounted with aerodynamic surfaces flush against the main stack work though?

If you can elucidate your problem everyone's happy to help.

I thought it was clear from the first sentence? All of my rockets tumble no matter what I do about it. The only successful launches I have are some kind of insane combination of skill and luck while threading a fine needle. :|

Thanks for the tips.

[Camacha]

What if your payload is non-circular, something odd shaped, or can't be made to fit in a faring? Are you boned? Try to put farings around it? Must everything I build fit in a faring or be made aerodynamic?

Depends on whether the thing will fly without fairing You will have to get creative to get weird loads into space, as is true in real life. You could build foldable vehicles, launch them in multiple parts and assemble them in space or come up with other solutions

What if your payload is non-circular, something odd shaped, or can't be made to fit in a faring? Are you boned? Try to put farings around it? Must everything I build fit in a faring or be made aerodynamic?

I thought it was clear from the first sentence? All of my rockets tumble no matter what I do about it. The only successful launches I have are some kind of insane combination of skill and luck while threading a fine needle. :|

Well, people are able to launch their rockets with FAR, so they must be doing something different than 'no matter what I do about it'. Without you being a little more precise what it is you are doing and what you are not doing, it is very hard to identify the problem and give helpful tips.

About the vectoring; it might help steering your rocket around, but moving everything too much around could be just as bad or worse as too little.

Edited May 18, 2013 by Camacha

[NonWonderDog]

If your payload can't fit in a fairing, build it in orbit out of pieces that can. Like a real space program. Read up on what gravity turn actually means and try to get to orbit at a maximum of 1-2 degrees angle of attack the whole way to orbit after pitchover at 500 meters or so.

Or you can always build a tractor rocket. build a cradle for your draggy payload and stick rockets around the side of it so that the center of mass is in front of the center of pressure. It takes a lot more impulse, but it works.

Here's a UFO with a frameshift drive lifted to orbit using that principle. It's made almost entirely of stalled wing sections, so the drag is insane, but the lifter was completely stable. (Sorry, ferram4 -- this probably wasn't what you had in mind with this mod. I can post more EELV replicas as penance if you want.)

[ferram4]

@bluegoblin:

What if your payload is non-circular, something odd shaped, or can't be made to fit in a faring? Are you boned? Try to put farings around it? Must everything I build fit in a faring or be made aerodynamic?

Allow me to be very mean and ask an obvious question: you install an aerodynamics mod and don't expect to have to make your launch vehicles aerodynamic?

Ribbing aside, un-aerodynamic payloads on top of your rocket will make the vehicle unstable; your current design sounds like you're trying to fly a dart into orbit backwards. If you don't want to make your payload aerodynamic you'll need to throw a lot of fins at the base of your rocket to keep in pointing in the right direction. While adding more vectoring rockets might help with the control aspect of your vehicle, they won't help with the stability aspect; the two, while related, are different issues: you can have a controllable rocket or plane that is unstable (constant course corrections to maintain control) or a rocket or plane that is stable, but uncontrollable (think a plane with the CoL way behind the CoM; no amount of control surfaces will get its nose up before it crashes into the ground).

I'd suggest trying Taverius's ideas if you haven't already, but I'd also add:

Keep your TWR below ~1.5 at launch; this will help keep your speed down in the lower atmosphere, making it easier to maintain control and making it easier to keep delicate payloads / launch vehicles together.

If you wouldn't mind posting a picture of your rocket and a quick summary of how you launch to orbit I (or another FAR enthusiast) should be able to critique your design / ascent profile.

@NonWonderDog: My inner aerodynamicist is twitching. O_o

Edited May 18, 2013 by ferram4

[Taverius]

What if your payload is non-circular, something odd shaped, or can't be made to fit in a faring? Are you boned? Try to put farings around it? Must everything I build fit in a faring or be made aerodynamic?

Its going to help, a lot. Just like real rockets, not having an aerodynamic shape around your randomly-shaped payload leads to bad things. Also just like real rockets, keeping the fairing as small as possible is best.

KW has fairings that fit some pretty huge stuff, if that's not enough there's always the 5m fairings in NovaPunch, and lets not forget Fairing Factory.

Asparagus could be one of my problems, I usually use some form of additional liquid tanks and engines with decouplers around the main, but I don't see how that might cause the tumbling by itself. Doesn't an array of engines with good vectoring typically INCREASE your control and stability? *confused as hell by that* I can see how it could be a problem of drag though. Will extra engines mounted with aerodynamic surfaces flush against the main stack work though?

Sure, except for the part where you move the centre of mass even further back and even further away from the centre of lift/drag (engines are heavy and dense) thus making your rocket even more unstable. The same if you have shorter outrigger boosters on the sides, whether they be liquid or solid ... they're going to be denser than the payload up top, shifting the CoM down.

If you go into this building rockets like it was stock KSP bad things happen.

For one thing, you don't need neeearly as much engine or fuel to get to orbit. Since drag is sort of realistic, it eats way less dV.

Also keep in mind transonic effects are present here. Its better not to pass the sound barrier too low down ... I know ferram4 likes to keep his first stages TWR very low, I think below 1.5, and inch into the tropopause. I've found if I go above 2 bad things start to happen as you pass Max Q in the lower atmosphere. That's the point of maximum drag, and if your rocket is not perfectly stable (like most rockets are) that's the point at which its subject to the strongest destabilizing forces, and if you do that in the densest part of the atmosphere on top its triple-minus bad D:

Also, I can't stress it too much, geeentle turns. More than 5 degrees off from your velocity and its bye bye. If you want to drive it like a racecar on a track you're gonna have to winglet that baby down low like some buck-rogers era design.

[Van Disaster]

What if your payload is non-circular, something odd shaped, or can't be made to fit in a faring? Are you boned? Try to put farings around it? Must everything I build fit in a faring or be made aerodynamic?

This went up ok:

I appear to have either deleted the launcher or it got trashed, if I find it I'll edit this post with it. It won't be anything too unconventional, I suspect: FAR doesn't really encourage that.

My biggest launchers are usually three stacks of 3.5m KW parts, arranged ala Falcon Heavy. I tend to cluster four of the biggest 2.5m KW engines for the centre stack instead of using one of the 3.5m ones though.

[taniwha]

Bluegoblin: FAR's lower delta-v to orbit does not come cheap. While FAR still allows flying brick outhouses (so long as they're built with due care), they won't get the delta-v gains that FAR provides to well designed ships. Also, flying ships with FAR takes a bit more skill (cf various comments about gravity turns). It took me a week to work out how to get my rockets into equatorial orbits (ASAS + untweaked control surfaces (rv-8 is it?) + FAR = a bucking bronco that winds up in a near polar orbit (for me)), and my ships were stable (CoM ahead of/above CoM).

Make your payload streamlined, or get it hidden behind something that is (ie, use fairings. I'm not sure if FAR supports wind shadows caused by non-connected parts being behind other parts). If you can't streamline it, at least keep it balanced. Randomly protruding bits makes for random aerodynamic forces. not fun.

For any stage that will be operating in atmosphere (below about 45km for Kerbin for these purposes, though different ships might have different limits), get that CoL below/behind the CoM (turn on the indicators). If the CoL is not below the CoM, slap tail fins on the stage until it is. Don't put the CoL too far behind the CoM, or the stabilizing forces will overpower your control forces. Set the deflection angles of any control surfaces to about 3 degrees or less (first page in the FAR box in the VAB. Nothing showing = no control surfaces so effectively 0).

The above means that if you do onion or asparagus (the latter not recommended due to the distribution becoming uneven), put fins on the inner stages as well as the outer stages. Keep that CoL behind the CoM.

It's generally a good idea to keep ASAS off during ascent, even with tweaked control surfaces, at least until you're above =45km (beyond that, the air's too thin to cause control problems, though it's still thick enough to drop your apoapsis if you cut your engines. It's also too thin for your fins to be effective for control :/). ASAS will happily throw your ship way outside the recommended 5 degree angle of attack (because its control algorithm doesn't correct properly).

Keep an eye on your surface speed. You do not want to be going much faster than 300m/s when you hit 10km (if, like me, you use deadly reentry too, you will burn). Up to around 400m/s seems to be ok for 10km-15km. About 600m/s for up to about 18km. 800m/s for about 20km. 1000m/s seems to be ok above 22km. 1500m/s is safe nearing 30km. Above 33km, it doesn't seem to matter. Rule of thumb: if you see reentry effects (above 10km) or hypersonic effects (below 10km, the white streaks streaming off your ship), you're probably going too fast. Thin, dull red streamers seem to be safe for the most part (though some parts (eg, Kerbal Attachment System's detachable connector) have much lower temperature tolerances, despite being behind a fairing), but if you see any yellow, consider it a failure: you're already dead. What, you don't use Deadly reentry? Well, consider this: in the not too distant future, everyone will be because heating effects are planned. NOTE: the above numbers are off the top of my head and are only approximate. I'm too busy watching my space plane's pitch, AoA, and vertical and surface velocity to take accurate notes. Again: try to avoid hypersonic/reentry effects on the way up: they mean you're going too fast for that altitude (if nothing else, you're wasting delta-v that should be saved for higher altitudes when you're going nearly horizontal).

One accurate figure I can give: ~1100m/s at 17-18km produces near-complete destruction of my planes.

Keeping the TWR of your lower stages fairly low will allow for worry free ascents. Just keep that angle of attack low by starting your gravity turn as early as you can (eg, 200m (yes, 0.2km)). If you use mechjeb: turn start of 0.2 and 40% seems to work well (I've never had any problems when my TWR wasn't too high). I sometimes have trouble keeping my lifter core out of orbit .

tl;dr: if it looks like a brick outhouse, it will fly like a brick outhouse. Make it look like a rocket and be happy.

[edit] I forgot to mention that if you use 3x symmetry for your fins, your ship will rotate such that one fin points down once you start your gravity turn. This is because once you have some attack angle, the CoL will be off-center and your ship wants to hang the CoM below the CoL (thus many real planes have non-level wings to get free roll stability: put the CoL higher than the CoM and the plane will naturally want to right itself).

Edited May 18, 2013 by taniwha

[ANWRocketMan]

What modifications need to be made to mod parts concerning their stock drag values etc.?(not wings, just fuel tanks and the like).

[Van Disaster]

None, iirc - tanks behind nosecones shouldn't have any real drag until you start going sideways anyway. Not sure how nosecones work though, they don't appear to have any special config. I guess it calculates AoA of every surface? I'm fairly sure I've seen FAR give parts some default values, but I can't remember where I saw that.

That was enough to get to 205km orbit to throw some fuel in the booster stage, it seems. I guess being a spaceframe rover it wasn't *that* draggy.

[ANWRocketMan]

I found that 3000m.s^-2 delta-V is enough for a 100x100km orbit if you have high enough TWR and an ascent path that allows a high TWR. If you can get some good delta-V in your last stage you can then push your orbit up much easier.