We need Moooore gimbal

[royying]

After 1.0 update, it is hard to launch a "long rod" shape rocket, rocket are flip over frequently.

Until I use TCA mod's steering gain function, this function can make more "vector thrust" by control each engine throttle, it make every "long rod" rocket launch success.

I think the stock engine should have larger gimbal range, make it suitable to use in 1.0 "more real" aero system, as I know the real rocket engine have much more gimbal range than KSP's engine.

[Alshain]

Gimbal is fine for all purposes except a space shuttle. Not everything will fly, if you're having trouble with it then you should design a better rocket. I've built some incredibly tall rockets that I've had absolutely no trouble with at all. However, as in all games, there are limits.

Edited November 7, 2015 by Alshain

[royying]

Gimbal is fine for all purposes except a space shuttle. Not everything will fly, if you're having trouble with it then you should design a better rocket. I've built some incredibly tall rockets that I've had absolutely no trouble with at all. However, as in all games, there are limits.

Why it can't fly?

COM COT aligned

5 vector engine

have tail fin

perfectly symmetric

[silent_prtoagonist]

A center of thrust misalignment isn't your problem. The problem is that your center of mass is so low on your rocket, and therefore behind the center of drag, so your rocket wants to fly backwards (COM in front of COD). Note that what the blue marker in the VAB shows is only center of lift, which only accounts for lifting surfaces on your vessel and not general drag. Currently there's no way to display the center of drag on a vessel, which makes solving this kind of problem very frustrating. But the general solution is just "add bigger and bigger fins until it works."

EDIT: You can also redesign your rocket to move the center of mass farther forward (try to avoid big long fairings with low density parts inside them like the one you have), but this can be difficult for certain pyloads. Or you can often fly even an unstable rocket if you never let the nose get more than a few degrees from the velocity vector, but that can be tough to do without some sort of autopilot.

Edited November 7, 2015 by silent_prtoagonist

[Alshain]

Your fins are control surfaces, best used for planes. Those are going to react to all control inputs and cause an almost over correction effect. Replace them with the Basic Fin or the AV-T1 (or any of the static plane wings even) and rely on the engine gimbal by itself. Alternatively you can right click those ones and turn everything off.

The radial booster tanks are not aerodynamic, they will create drag which will have a tendency to pull the front end around. This raises your center of drag and as silent_prtoagonist said, that puts the center of drag in front of the center of mass. More fins (but not control surfaces) may also help negate this problem, but start by eliminating sources of excess drag first by placing nose cones on top of the tanks.

Edited November 7, 2015 by Alshain

[Snark]

Aero control fins work great for rockets, they're not just for airplanes. They provide lots of control authority, equivalent to having a strong engine gimbal. Depending on ship design, they may be too strong (just as engine gimbal can be too strong, for some designs), but they certainly have their place.

That said, they won't help your current design much. Look at how far forward your fairing is sticking out above your CoM (a long long way), and then look at how far behind your CoM the fins are (not very far). It's a much shorter lever arm than that huge long bulky fairing sticking out, and with a placement like that, I kinda doubt that any amount of fins would help you.

The fundamental problem here is that your CoM is far, far too low; this is a problem that many folks have, but I gotta say that the craft you've shown is the most extreme example I've seen in a long time.

So, what can you do?

Well, I don't know what you're hiding in that fairing, so I don't know how much flexibility you have there. If you can make it less tall, or move its CoM upwards, that would be a step in the right direction. But probably not enough.

The other thing you can do is move those radial boosters way, way down, e.g. sticking way down below the bottom of the Skipper. Then put steerable fins down at the bottom of the boosters, as low as possible. AV-R8 winglets would be good. By putting the fins very far back, you give them a longer lever arm to work with. Moving the boosters down will also make their gimbal much more effective, again by giving them a longer lever arm to work with.

Make sure that the fuel lines draining the boosters into the center are connected to the TOP booster tank. This will make your Skipper drain the bottom booster tanks first, which will help stability by moving CoM upwards. Unfortunately your booster engines will drain from the top down, which is bad, but there's not a lot you can do about that. At least you can configure it so that your Skipper is making the problem better rather than worse.

Another possibility, if you have a really awkward payload and making it more aerodynamically stable is impractical, is to settle for a less efficient ascent profile and just eat the extra expense in mass. Launch straight up, avoid going too fast, and wait until you're a lot higher than usual to start your gravity turn. Doing it this way lets you postpone the risky flip-prone turn until you're up where the air is thin and aero stability will be less of an issue. The downside is cost. For example, on your current design, you could stick a bunch of Thumpers under it (one under each radial booster, and four under a quad-coupler under the Skipper. Set their thrust appropriately so you don't go too fast too low. Then just ride the SRBs straight up.

Edited November 7, 2015 by Snark

[fireblade274]

...center of mass is so low on your rocket, and therefore behind the center of drag, so your rocket wants to fly backwards

ive actually never thought about that, this could help with some designs (it would be great if we had a COD display in VAB)

[Snark]

ive actually never thought about that, this could help with some designs (it would be great if we had a COD display in VAB)

It would... the problem is that CoD is a hard thing. The amount of drag depends on how your craft is oriented and also on how fast it's going.

For example, consider fins, which are vitally important for aero stability. When you're flying perfectly prograde, they have almost zero drag (because they're edge-on); whereas their drag goes up by a lot when you're pointed off prograde. That's precisely why fins are so great for stability. But how would you display that with a CoD indicator in the VAB? You can't calculate a CoD at all unless you assume a certain orientation and speed; the only orientation that makes sense to assume (i.e. pointed straight up) would cause the fins to have just about zero effect on the positional display marker.

That's not to say it would be completely impossible to have some sort of "you're making an unstable contraption" indicator, just that it's not at all a trivial design problem to figure out how it should work.

[More Boosters]

A rule of thumb is that unless your CoM is periliously close to your CoL you'll be fine. Put that CoL way below.

[BudgetHedgehog]

Ok, back to the topic at hand:

Yes, more gimbal would be nice. Buuuuuut only if stock aero models aerodynamics as well as FAR does (which includes a patch to increase engine gimbal). Real life engines have a large gimbal, but KSP air isn't real life air so it doesn't need as much.. it'd helpful, sure, but then, so would removing drag or making reaction wheels even more unrealistic.. So yeah, I'm all in favour of moar gimbal, but it'd need the atmosphere to also slide towards realism because otherwise, it's just making things easier.

[Snark]

Ok, back to the topic at hand:

Yes, more gimbal would be nice. Buuuuuut only if stock aero models aerodynamics as well as FAR does (which includes a patch to increase engine gimbal). Real life engines have a large gimbal, but KSP air isn't real life air so it doesn't need as much.. it'd helpful, sure, but then, so would removing drag or making reaction wheels even more unrealistic.. So yeah, I'm all in favour of moar gimbal, but it'd need the atmosphere to also slide towards realism because otherwise, it's just making things easier.

I'd tend to come down on the side of "no, we do not need more gimbal." The OP wants it because his spaceship is grossly unstable aerodynamically. However, any design that needs more aero stability can get better results from steerable fins (already available). Also, how much gimbal is "right" depends a lot on the ship design. I would contend that the existing engines are actually over-gimbaled for a lot of designs-- for example, when I'm running a Mainsail under a big orange tank with a payload on top of that, I almost always need to damp down the Mainsail's gimbal to about 30% or so if I don't want the rocket to start flexing in the middle.

I think Squad did a pretty good job of balancing the gimbals on the various engines. You really don't need much engine gimbal for a "reasonable" rocket design, and that's part of the charm of KSP: you need to think about what you're doing and there's some challenge to building effectively.

The one scenario where having a lot more gimbal would be helpful is for building space shuttles, but that's why they're adding the Vector engine in the 1.0.5 patch, which I get the impression is going to have a pretty big gimbal range precisely because it's intended as a shuttle engine.

(And if you really can't stand it and want the gimbals for everything increased, it's trivially simple to mod it yourself with a snippet of ModuleManager config.)

Edited November 8, 2015 by Snark

[ferram4]

I'd tend to come down on the side of "no, we do not need more gimbal." The OP wants it because his spaceship is grossly unstable aerodynamically. However, any design that needs more aero stability can get better results from steerable fins (already available). Also, how much gimbal is "right" depends a lot on the ship design. I would contend that the existing engines are actually over-gimbaled for a lot of designs-- for example, when I'm running a Mainsail under a big orange tank with a payload on top of that, I almost always need to damp down the Mainsail's gimbal to about 30% or so if I don't want the rocket to start flexing in the middle.

Except this isn't a problem with gimbaling engines itself. This is really the result of 4 problems interacting in very nasty ways:

1. Gimbals don't generally have response speeds, though the coding exists for them. This leads to the thrust vectoring snapping instantly to what's been commanded, creating sudden torques on the vehicle under manual player control or aggressive SAS control.
   
2. Joints still flex too much, particularly between parts with very large mass ratios. The extreme flex between the fuel tank-decoupler-engine-fuel tank group between stages is still there, and while it isn't as bad as it was before the first batch of joint passes, it is still bad.
   
3. SAS has no idea how to handle gimbal response speeds (uncommon), nor does it know how to handle flexing (very common). This means that if it starts oscillating because of one of those it will cause the vehicle to lose control.
   
4. SAS starts off controlling way too aggressively, ensuring that it runs into problems with the above. Less aggressive initial controls would make things function better.
   

Fixing those issues would remove all the issues with "too much" gimbaling. The problems are never about having too much control authority, but instead the controller being unable to make good use of it.

[Snark]

Except this isn't a problem with gimbaling engines itself. This is really the result of 4 problems interacting in very nasty ways:

1. Gimbals don't generally have response speeds, though the coding exists for them. This leads to the thrust vectoring snapping instantly to what's been commanded, creating sudden torques on the vehicle under manual player control or aggressive SAS control.
   
2. Joints still flex too much, particularly between parts with very large mass ratios. The extreme flex between the fuel tank-decoupler-engine-fuel tank group between stages is still there, and while it isn't as bad as it was before the first batch of joint passes, it is still bad.
   
3. SAS has no idea how to handle gimbal response speeds (uncommon), nor does it know how to handle flexing (very common). This means that if it starts oscillating because of one of those it will cause the vehicle to lose control.
   
4. SAS starts off controlling way too aggressively, ensuring that it runs into problems with the above. Less aggressive initial controls would make things function better.
   

Fixing those issues would remove all the issues with "too much" gimbaling. The problems are never about having too much control authority, but instead the controller being unable to make good use of it.

Definitely, and I don't think anyone's arguing that the PID tuning of SAS in KSP is anything to turn handsprings over. I wouldn't mind seeing engine gimbals being given a reasonable response speed, or better tuning for SAS. My point was not so much a wish that they'd "fix the problems by reducing gimbal" ... just that we don't need more gimbal than is there already.

[royying]

The payload stage is very light relative to the booster stage, and the lower part of rocket often attached some radial booster, make the CoM very low

How to make the CoM higher?

[nikokespprfan]

I think the stock engine should have larger gimbal range, make it suitable to use in 1.0 "more real" aero system, as I know the real rocket engine have much more gimbal range than KSP's engine.

Doesn't 1.0.5 add a new engine with a huge vectoring range? It's the S3 KS-25 "Vector" Liquid Fuel Engine.

[Alshain]

The payload stage is very light relative to the booster stage, and the lower part of rocket often attached some radial booster, make the CoM very low

How to make the CoM higher?

You have two options, lower your center of drag, or split your payload into multiple parts and dock them insitu. Artificially adding mass to the tip of your payload would technically work but isn't the greatest way to do it.

- - - Updated - - -

Doesn't 1.0.5 add a new engine with a huge vectoring range? It's the S3 KS-25 "Vector" Liquid Fuel Engine.

That's a bad solution for the reasons ferram mentioned above. Watching the streams I've noticed it's hard to control a conventional rocket with that thing at full gimbal.

[MAFman]

One word for you, my friend. ModuleManager. That is all