Ferram Aerospace made the game 20x harder...

[AlternNocturn]

...and I'm certain it's my fault. Why do my rockets and planes flip out with the slightest deviation from the Prograde marker? I've heard it's meant to make the game easier but it really doesn't. Spaceplanes are impossible, mod crew pods point the wrong way on reentry, my rocket can only fly inefficient trajectories because anything else means the rocket flips out. And don't get me started on aerodynamic stress failures, which I've turned off.

What am I doing wrong? I'm totally open to suggestion because this game has stopped being fun, but I still love making minimalistic rockets that only work with FAR. Ergo, you can see the predicament I'm in. Help!

Edited May 30, 2014 by AlternNocturn

[LaytheAerospace]

Can you post a picture of one of your planes, with center of mass, lift and thrust enabled? The graph from FAR would be great too.

My guess is you're getting killed by drag forward of your center of mass, but I'll need to look at the ship to be sure.

[AlternNocturn]

Can you post a picture of one of your planes, with center of mass, lift and thrust enabled?

Sure. I'll post pics of a regular plane I made (A high-altitude drone) and a Dreamchaser-like spaceplane. Neither of them play well in the atmosphere.

This is an unmanned mini-shuttle inspired by the Dreamchaser and the x-37b spaceplanes. Flat spins almost immediately on reentry.

My high-altitude unmanned drone. I didn't think the bit at the top would cause the vehicle to be so aerodynamically unstable, but I guess I was wrong.

Edited May 30, 2014 by AlternNocturn

[LaytheAerospace]

For your X-37B, your problem is the center of mass being rear of the center of lift. The plane will try to reorient itself to fall center of mass first, so the moment you touch the stick on reentry it's going to try to turn ass first. It also looks like it could use a lot more wing area, unless it's lighter than it looks.

For your drone, the top bit looks like it's going to cause a pretty strong upward pitching moment. Spinning out of control is most likely because the center of mass and lift are right on top of each other. The closer they are, the more your control inputs are amplified.

If you put the CL ball just behind the CM ball, you'll get very good control without it being overly difficult to keep from going nuts on you. I also recommend getting to know the FAR editor window. It's pretty daunting at first, so I recommend hitting the help button first thing. You can also mouse over a lot of stuff for explanations.

It's a major adjustment to start using FAR, but once you get it down, it's very intuitive and tuning your ship's performance is easy.

[G'th]

If your rockets are flipping out, its because they're either far too top heavy, don't have enough lift and torque control to maintain a particular AOA, or are being flown at an AOA that it can't handle period. Its likely a mixture of all three. Solution is to build heavier as you go lower, not the other way around (in my experience, CM should be ~75% up the height of your first stage). And then ensure that whatever you're using to control your rocket while still flying in the atmosphere can control it. Bigger wings and stronger ASAS/SAS/Reaction Wheels. For planes its the same idea (outside of the mass issues, thats different for planes).

If you try going from a 5* pitch to a 90* pitch without doing it gradually, you're going to have a bad time as your plane/rocket ends up stalling which will lock out all the control until you bring your pitch angle back down closer to your prograde marker. The solution there is to keep your pitch angle close to your prograde marker. Depending on your craft there is an maximum AOA (IE, the difference between your pitch and prograde marker) that you can go to without stalling the craft. You can check that using FAR's tools to see where the craft begins to stall.

With the crew pod issues, thats up to the mod authors to fix if the issue is what I think it is. However, if the pods have some RCS on them, you can mount some thrusters on to it to help compensate.

And I'd recommend turning the aerodynamic failures back on. It'll teach you to be more efficient. (Hint: High TWR + large AOA = Too much stress = Failure) Before I got used to FAR's flight model, my rockets always fell apart in flight because they'd ascend too fast and I'd pitch way too hard. But now that I fly properly and design my rockets such that the TWR doesn't get too high, my rockets are more stable and capable than ever.

[Taki117]

Both your aircraft need a serious redesign. with your first "Dreamchaser" You should move the CoL behind the CoM, the further the better. Also, prior to reentry you need to shift the fuel as far forward as you can (Assuming it flies fine in atmo) With your second design as you burn fuel your CoM is going to move back and up, resulting in an unstable aircraft, it also needs a rethink.

Generally speaking if it looks like it will fly in real life it should fly with FAR. Design it first, fly it a few times, and then tweak the design.

Also, have a look at thisSometimes it does good to get back to basics.

[AlternNocturn]

Thanks, your information was really helpful. I didn't think the CoL had to be behind the CoM with FAR, I thought that only applied to stock KSP aerodynamics. I see now that isn't the case.

And thank you G'th, this will be helpful in tweaking my rocket fleet. Actually, I should have been more clear, the rockets donb't always flip out. They only do that when I go too far past the prograde marker, which means I can only fly inefficiently. But I'll make sure to follow your advice. I'll give this another shot before giving up, if I decide to.

Edited May 30, 2014 by AlternNocturn

[phoenix_ca]

http://forum.kerbalspaceprogram.com/threads/52080-Basic-Aircraft-Design-Explained-Simply-With-Pictures

https://github.com/ferram4/Ferram-Aerospace-Research/wiki

FAR doesn't make the overall game easier. It makes aerodynamics more realistic, which is actually a whole hell of a lot harder (fluid dynamics is easily more difficult than orbital mechanics, at least when talking about simple 2-body situations). The "more easy" part of it is that since aerodynamic drag is modelled so much better, you can get into orbit around Kerbin (and any other planet with atmosphere) with less ∆V than stock, and it actually makes serial staging and pointy rockets an advantageous thing.

Edited May 30, 2014 by phoenix_ca

[Drew Kerman]

You should move the CoL behind the CoM, the further the better.

Not entirely true. For example, with the CoL so far behind CoM I couldn't get my nose up on takeoff at all. So yes, the further behind the better but this also depends on the design of the aircraft (If this aircraft had stronger pitch authority it might be able to overcome the CoM being that far forward) and too far back will not allow you to pitch up properly.

[AlternNocturn]

Not entirely true. For example, with the CoL so far behind CoM I couldn't get my nose up on takeoff at all. So yes, the further behind the better but this also depends on the design of the aircraft (If this aircraft had stronger pitch authority it might be able to overcome the CoM being that far forward) and too far back will not allow you to pitch up properly.

This. While redesigning my FlameChaser, I moved the CoL a bit behind the CoM as you guys told me. However,I knew it was going to be too unstable once the fuel emptied (the forward tank empties before the rear one does) so on reentry I pushed all the fuel forward making the CoM far ahead of the CoL. This made reentry completely stable; the vehicle dove nose-first into the airstream as expected. But for landing, the vehicle does not have very much pitch authority given its nature, so I had to moved most of the fuel backward to get the CoM as close to default position as possible to give me more control in atmo. Positioning those two vectors is really critical to mission success. I know it's obvious, but it never really occurred to me at the time.

So yeah, the Flame Chaser works now, and I only crashed because I didn't bother to target KSC and I banked too hard to the right to avoid a hill on final descent. Otherwise it flew like a dream. FAR isn't really that much harder than stock KSP when you know what you're doing. Thanks again, guys.

Edited May 30, 2014 by AlternNocturn

[Streetwind]

Actually, I should have been more clear, the rockets donb't always flip out. They only do that when I go too far past the prograde marker, which means I can only fly inefficiently.

It's more the other way around, in actuality. KSP's stock aerodynamic model requires you to fly inefficiently.

From an efficiency standpoint, you are wasting fuel whenever you are not thrusting directly prograde. This includes engines gimbaling to affect control input. A real life rocket will use engine gimbals exactly once: shortly after liftoff, for one singular, tiny nudge to the side. Afterwards, the rocket flies entirely uncontrolled; it simply tilts over slowly and gradually by itself, pulled into the curved path towards the vertical by gravity and stabilized by its aerodynamic profile. That's why it is called a gravity turn, incidentally: because you don't actually steer. Gravity does the steering for you.

In contrast to that, stock KSP's atmosphere punishes you for trying to pitch over normally by applying excessive lateral drag until the air thins out sufficiently. As a result, stock KSP rockets go straight up for 5-10 km, and then aggressively pitch over through manual control input. This is a very inefficient launch profile, and one that (as you can see in FAR) isn't even aerodynamically valid under proper physics.

With FAR, this excessive early drag isn't present, and there's nothing keeping you from starting your pitchover right after clearing the launchpad. This allows you to fly a much more gradual turn, never deviating far from the prograde marker and never using aggressive steering input. That results in a very smooth and efficient ascent.

And if you're really good at what you're doing and your rocket is aerodynamically sound, you can just turn off SAS, sit back and watch. It's not easy to pull off, but the authentic, uncontrolled gravity turn does work under FAR. Scott Manley showed it off a couple of times already, the most recent example being in

. Notice how he steers a little at the start, then turns off SAS and just lets it fly. Observe the navball as the rocket turns, how close it stays on the prograde marker. Obviously Scott was aiming for a very high apoapsis for this particular launch, but it also works for classic low Kerbin orbits. If you know what you're doing. Don't ask me about it, because I've not succeeded in making it work yet Edited May 30, 2014 by Streetwind

[cantab]

A real gravity turn can work in stock too. I did it with a small probe launched on a long SRB. Tilted the SRB over 5 degrees on the pad and it flew by itself to a nice apoapsis where I circularised.

I don't know how to deliberately design a rocket to do this though.

[Drew Kerman]

However,I knew it was going to be too unstable once the fuel emptied (the forward tank empties before the rear one does) so on reentry I pushed all the fuel forward making the CoM far ahead of the CoL. This made reentry completely stable; the vehicle dove nose-first into the airstream as expected. But for landing, the vehicle does not have very much pitch authority given its nature, so I had to moved most of the fuel backward to get the CoM as close to default position as possible to give me more control in atmo. Positioning those two vectors is really critical to mission success. I know it's obvious, but it never really occurred to me at the time.

You'll want to check out TAC Fuel Balancer

[Hodo]

Thanks, your information was really helpful. I didn't think the CoL had to be behind the CoM with FAR, I thought that only applied to stock KSP aerodynamics. I see now that isn't the case.

And thank you G'th, this will be helpful in tweaking my rocket fleet. Actually, I should have been more clear, the rockets donb't always flip out. They only do that when I go too far past the prograde marker, which means I can only fly inefficiently. But I'll make sure to follow your advice. I'll give this another shot before giving up, if I decide to.

That actually applies to most aircraft in real life, the CoM is slightly in front of the CoL. The closer they are together the more unstable the aircraft is. If the CoL is to far back the nose will never pitch up and will actually be forced down at supersonic speeds.

FAR doesn't make the game easier, it just makes it more realistic. Which for most people that is easier.

Rockets are a bit more difficult in FAR, you need to start your gravity turn MUCH sooner than stock. I start mine at 1km altitude if you use MJ, or as soon as it clears the launch tower if I manually fly it. I never go more than 5deg past the prograde marker the whole time I am doing the turn. Also make sure your TWR isn't more than 1.2-1.4 at launch otherwise you will accelerate faster than you can control your craft, and it will push the back of your rocket around instead of pushing it where you want it to go.

Edited May 30, 2014 by Hodo

[LaytheAerospace]

Touching on CL in relation to CM again, I very strongly disagree with the notion that you should put the CL far behind the CM. It gives you a more stable plane, but stability isn't actually a good thing! Maneuverability is predicated on instability, so if you're too stable, you lose the ability to maneuver. You want only as much stability as is necessary to make the plane controllable.

I've been following a pretty strict rule for my small (<40 tons) spaceplanes, with great results.

First, your CM can't move more than 10% of the diameter of the CM ball as your tanks empty. Moving none at all is preferable, but a little movement can be tolerated. Second, put your CL ball just behind the CM ball, with the surfaces of the balls just touching. This will give you a good starting point to tune your ship.

Typically to tune, I turn SAS on at launch, then fly to 1km and turn it off. If the plane oscillates during this initial climb, then it's not well balanced or I have too much pitch authority (or both). You can fix balance problems with trim, but it's best to make it fly more or less straight on its own. You can use nose canards, which make it very easy to apply a precise upward pitching moment by moving the canards in relation to the CM for more or less torque. Rotated tail fins are another good option. Moving the wings to shift CL is another option, but the ultimate goal is to make the plane fly well without sacrificing the maneuverability granted by having them close together.

A well tuned plane should pitch down very gradually. The downward pitching moment should be low enough to be trimmed out for easy, level flight without SAS. A well designed plane will be just about as easy to control without SAS as it is with it. If you turn SAS off and your plane nosedives, then you need to take it back to the SPH and fix your pitching moment. Also, keep in mind that you'll tend to lose pitch authority (as well as develop a downward pitching moment) as you go supersonic and beyond. FAR's simulations are very, very handy for evaluating this before you spend a bunch of time getting your plane up to Mach 5 for hypersonic tests.

[Hodo]

LaytheAerospace is dead on about everything there.

If you set your CoL to far back the craft will never leave the ground untill the ground falls away from under it at the end of the runway. And even then it may not even pitch nose up even with full control lock for it to go up.

If you have it set to close to center, if your CoM shifts a great deal during flight then your craft will have a completely different flight profile when near dry then when full. Sometimes this is bad, other times it is good, it depends on how you built it.

Most of my craft the CoM actually shifts slightly forward when the craft gets near empty, making the craft a bit more nose heavy, which is good for me.

[BudgetHedgehog]

For rockets, the ideal ascent is a launch TWR of about 1.1-1.3 and keeping it around that through the lower ascent. If you go too fast, you're pushing hard against high drag atmosphere that will flip/break your rocket. If you look at it, it makes sense - say you're travelling straight up but are tilted over 10 degrees (through engine gimballing/SAS etc). You have a wall of air hitting the side of your craft at 200 m/s or whatever which creates instability/failure. Either it pushes your craft over or your craft can't take the stress so it breaks.

Anyway, a launch TWR of 1.2, start pitching over about 70 m/s. I say pitch over, but really, it's just a tap in the direction you want to go. As a rule of thumb, never pitch more than 5-10 degrees off prograde initially. Above 30km, you can start pitching/turning more aggressively, but still, be careful. As for the actual ascent path, for a low TWR craft, the Time To AP should be around a minute and for a higher TWR (1.7+), from 30-40 seconds. Try and follow the stock shape (be pitched over 45 degrees at 10km) but really, it depends entirely on your individual craft. I usually keep the gradual turn going until I'm pointing just above the horizon at around 35km. I keep this attitude until the AP is where I want it, cut throttle, point prograde and coast. Might need a few nudges to counter drag, but you'll likely be coasting for about 5 minutes. Once at AP, a circ burn of usually less than 400 m/s is needed, but again, it depends on your craft. Sometimes, I've only needed a burst of 30m/s at AP, other times, I've needed 500. A constant thrust ascent is doable, but very difficult.

For planes, as others have said, make sure the CoM is slightly ahead of the CoL. Don't overspeed low in atmosphere or you'll have a high dynamic pressure which is what tears you up if you turn. FAR has a few flight assistance toggles that can help you - DCA, or Dynamic Control Adjustment, will limit the input so you don't overturn your plane. However, in doing so, it does limit what the plane can do and doesn't let it get pushed to its boundaries (which is sometimes needed, at least for spaceplanes). FARs Flight Data contains all the info you'll ever need - it's far too much for me to explain here, hopefully the wiki has the info.

Talking of spaceplanes, I've found them incredibly easy with FAR. Make a simple deltawing with 1 RAPIER at the back (keep in mind CoM/CoL and everything) and while in the VAB, make it manual mode switching and put the following into action groups - AG1 = Toggle Mode + Toggle Intakes, AG2 = Toggle Intakes. This allows you complete control over your intakes and engine. When you start, press both AGs and get going. Go up high enough (around 10km or so, depending on how many intakes you have) and then fly about 5 degrees up. Build up speed and keep an eye on the Air Requirement Met % in FARs Flight Data. The faster you go, the higher it'll be, but the higher you go, the lower it'll be. Once it gets below 100%, you'll flameout unless you throttle down so keep going until about 120%. Your AP should be around 17km or so and you should be travelling at around 1.5km/s. Press AG1 to switch to closed cycle mode (and close your intakes), pitch up more and watch your AP. This is likely where you'll have to turn off DCA if it was on because it simply won't let you pitch up far enough. Don't pitch too much or you will break, just maybe 20 or so degrees up. Enough so your Time To AP rises at about 1s/s should do. If it's not going that fast, don't pitch up, just fly straight in closed cycle for a bit longer. Once your AP gets above 30km, you'll likely get to space today (as long as you have enough fuel).

I've honestly found making spaceplanes much easier now I know what to do. My current airframe is very capable and stable almost all the time. It's slightly more complicated due to having separate rockets and jets, but the principle is the same.