Fins are counter productive.

[KerbMav]

Now what happened here ...

The center of mass in the VAB was already pretty low - why ever that might be.

After launch the camera dropped two times, like falling down 5 meters.

Then it started to swing like a lasso around a center of mass far outside the rocket itself ...

[bakanando]

@Necrobones: Your design had little problems because the small fins have almost no lift compared to other fins I ran in the test, they indeed perform pretty well but still show the tendency to pull/push your rocket towards an arbitrary direction.

Indeed, fins+gimballing work very well to hide this problem. But that is not a fix, it is a workaround to an inherent problem in the game. I don't have SAS modules yet, so I can't really tell how much torque do you need to control your rocket and counteract the fins exagerated torque.

-Edit-

@KerbMav: That is hilarious, sorry. You might have moved the camera down with your mouse (middle button hold), and changed the camera to new chase mode. This would cause something similar to what you had. As for the tumbling, try holding closer to prograde at all times.

Edited May 11, 2015 by bakanando

[KerbMav]

.

Camera was untouched and it all started while my climb was still almost vertical, the low CoM made it rather unpleasant to steer, so opted to waste fuel.

After removing a couple of parachute and fairing mods (community bugfixes only, nothing fancy like RealChute or anything) the rocket attained orbit.

I forgot the best part: Afterwards I went into a test-frenzy just staged every stage some time after launch, the parachutes opened and although the engines were still pushing upwards they were oriented upwards like when landing and some time after that ... I cannot tell if the camera moved even further downwards or the ship zipped upwards, but the camera was tumbling through the air and nowhere nere any part of the rocket.

I need to get FRAPS running or something, just in case any bugs occur, maybe recording 10mins in a loop like a blackbox ...

Edited May 12, 2015 by KerbMav
forgot

[Randox]

I tested it with different wings. Canards and other control surfaces did not cause tilt. But if control on all three axis was disabled for the control surfaces, then the rocket did tilt towards the 270. Furthermore, the bigger the wings and/or the further back they were, the quicker it tilted. By using a launch clamp to start the rocket slightly tilted off to the 90, it exhibited the same behavior, but in the opposite direction. This has led me to conclude that the issue is simply that the center of lift is too far below the center of mass (too stable) for the available control forces to correct. Now, you may be asking why does it seem to have a tendency to tilt towards the 270 much of the time? I believe it is caused by the fact that if you point a ship straight upwards and try to fly with no input, the ship will appear to be rotating westward ever so slightly, when actually the planet is rotating eastward, and the ship is maintaining an absolute heading. (This effect is much more pronounced in low orbit). And as the ship tilts slightly westward (relative to Kerbin), its engine will start giving the ship a tiny westward velocity, which the extremely low center of lift will only exacerbate. On top of this, the launchpad is slightly tilted towards the 270. There are probably other factors at play though, as this does not happen 100% of the time...

I think this guy nailed it. Every rocket starts with a westward trajectory relative to Kerbin, so if you slap some unsteerable fins on a rocket that otherwise has very little control authority, it's just going to follow that initial westward heading. Someone else in this thread already pointed out that if you use launch clamps to instead give the rocket an initial eastward trajectory, it will behave exactly the same, except it will fly east instead of west.

It really is a PEBKEAC situation I think. If you could freeze Kerbin entirely in place (relative to Kerbol which I don't believe has any simulated movement relative to a Galaxy or anything) and tried this experiment again, you would almost certainly find that your rocket now ascends straight up. The moral of this experiment is that rockets should not be using fixed wings/fins that exceed their other control options, and as Ival has pointed out, you would probably have better luck moving the wings up closer to the CoM to reduce the fins own control authority down to a level that something like a gimbaled engine can overcome. What you have created in your experiment is a rocket powered arrow, with exactly as much control over it's vector as the real deal, leaving you entirely at the mercy of your initial vector and gravity.

Fins, gimbals, and torque wheels wouldn't seem to be the workaround to a problem in the game so much as the very real workaround to a very real problem. Your rockets are behaving exactly as one would expect, they're just operating in a context most people are familiar working in.

Edited May 12, 2015 by Randox

[gogozerg]

it's just going to follow that initial westward heading.

@SAI Peregrinus @NecroBones @others too

We may not have seen the same thing with similar bad rocket designs.

There may be good reasons for rockets to fly west (Coriolis pseudo forces, unperfect launch conditions, ...).

But a rocket that does a 90° tilt before reaching 700 m when and only when it has fins, it's telling something different, probably related to some unnatural/exagerated lift force.

Well, I'm done with it, going back to better designs.

Edited May 12, 2015 by gogozerg