Maybe the solution for 0.21 flight stability problems

[Baenki]

I have no access to my pc right now so i can not try it.

My idea is that the engines have a little force pulling you sideways, resulting in unstable flight.

What about inverting your engines in the VAB/SPH. Put one engine on and rotate a second one 180º clockwise.

The idea behind is this: http://imgur.com/1zOSbQg.png

Circle = engine

Red = force of the engines

Blue = resulting force

Sorry this was drawn on my phone.

Maybe somebody could try that an tell if it works.

Edited July 25, 2013 by Baenki

[BeefTenderloin]

I'll try this, but I doubt it is the case.

[whaaw]

its not that.

the stupid people dont realize that they have to put rcs winglets or reaction wheels to make the ASAS usefull.

the new asas is just perfect...

[BeefTenderloin]

TEST 1 - Both engines aligned same way. No gimball or control inputs.

IN THE VAB

AT TWENTY SECONDS

Craft had deviated up and left of the centrepoint.

Edited July 25, 2013 by BeefTenderloin

[BeefTenderloin]

TEST 2 - One engine rotated 180 degrees. No gimball or control inputs.

IN THE VAB

AFTER TWENTY SECONDS

Deviated slightly down and left. Most stable of the three configurations.

Edited July 25, 2013 by BeefTenderloin

[Chrisd857]

To BeefTenderloin: Do those engines gimbal? Will you try adding reaction wheels or winglets?

[skyace65]

its not that.

the stupid people dont realize that they have to put rcs winglets or reaction wheels to make the ASAS usefull.

the new asas is just perfect...

I've done all of that, and it still completely sucks

[BeefTenderloin]

Sorry for the break in testing, my mouse ran out of battery.

To answer Chris' question,

To BeefTenderloin: Do those engines gimbal? Will you try adding reaction wheels or winglets?

These are the non-gimballing LV-T30 engines. As this is a test of whether engines produce a mystery force, I will not add more winglets or reaction wheels. I am going to repeat these tests, however, with the reaction wheels disabled, and see if that makes a difference.

[BeefTenderloin]

TEST 3 - Non-Gimballing engines, No SAS used, no control input other than throttle.

IN THE VAB

TWENTY SECONDS INTO FLIGHT

At twenty seconds, the craft had deviated wildly up and left, and had also rolled a little.

[RatBeer]

the new asas is just perfect...

I dont know why people say this. How can it be perfect when it cant even keep you locked onto your heading

[Neomatt]

These are the non-gimballing LV-T30 engines. As this is a test of whether engines produce a mystery force, I will not add more winglets or reaction wheels. I am going to repeat these tests, however, with the reaction wheels disabled, and see if that makes a difference.

Just to be sure it starts straight up: can you add 4 launch clamps to the sides?

[KerbMav]

Hm, two engines doesnt "feel" very stable - could you please try three or four?

[BeefTenderloin]

delete this post please

Edited July 25, 2013 by BeefTenderloin

[BeefTenderloin]

Just to be sure it starts straight up: can you add 4 launch clamps to the sides?

Good idea. Will redo tests, but I feel difference will be negligible.

Hm, two engines doesnt "feel" very stable - could you please try three or four?

If you feel like drawing up all the different arrangements for four engines, I'll do it.

[swiftgates24]

The solution is add one SMALL tank of RCS and 8 thrusters around the ship. It uses next to none RCS, and it keeps it straight. Before my rocket was also leaning to the right, and that fixed it. And as for power for the SAS - One battery bank (Large or Small), and 2 RTGs.

Try this Beef.

Edited July 25, 2013 by swiftgates24

[Tingle]

I have the weird issue that the new SAS doesn't even use RCS.

[swiftgates24]

Haha! That is the good thing - it looks like it doesn't, but it does! Right here my rocket is draining RCS at a rate of 0.01, but you can't see any white spurts:

This is that rocket, and as you can see, like I said, one small RCS tank, 8 thrusters around the ship, and that's all you need. Flies like a beauty.

Edited July 25, 2013 by swiftgates24

[magnemoe]

Just to be sure it starts straight up: can you add 4 launch clamps to the sides?

note that with no gimbals and no reaction wheels or control surfaces your rocket is dumb like an firework rocket, an small offset will grow large over time, this is common in real world.

An longer rocket would be more stable, rotating it would also help, this is done on unguided artillery rockets like the ones launched from Gaza towards Israel, note they miss towns from 20 km away.

Anything more complex like the V2, scud missile uses some control system to keep on track, they are not smart missiles but uses something like asas to keep direction straight.

[BeefTenderloin]

The solution is add one SMALL tank of RCS and 8 thrusters around the ship. It uses next to none RCS, and it keeps it straight. Before my rocket was also leaning to the right, and that fixed it. And as for power for the SAS - One battery bank (Large or Small), and 2 RTGs.

Try this Beef.

I was only testing if the engines produced a mystery force like the OP suggested.

[urablahblah]

I hope the people complaining realize that the 2m SAS module doesn't have reaction wheels in it. They are easy to add though.

That being said, I do miss the aggressive stay-on-target behavior of the old ADV SAS. If your rocket payload is not centered perfectly, it does take some work to keep it steady. For planes though, it is a wonderful new guidance system IMO.

[swiftgates24]

I was only testing if the engines produced a mystery force like the OP suggested.

I only found that SRB's do. So SRB's and engines? Strange...

[Tingle]

I also just did some rocket testing. Wiglets create a sluggish SAS behavior, RCS is barley used to stop angular momentum - or actually anything is barley used to do so. And when I micro manage my heading into a certain bearing, lets say 45°, if I then roll the rocket, it looses that bearing and starts tilting away/down, so that's proof that its borked and not holding single angels like its suppose to.

I wish Squad would say something about this, the new SAS pretty much ruined the experience - the new system makes the game unplayable in the sense of 0.20 behavior.

Edit: It also seems that the higher the velocity is, the less effective the SAS is - it keeps dragging toward the pro-grade vector.

Edited July 25, 2013 by Tingle

[Baenki]

Well it doesn't seem mystery force is the problem. To be honest i didn't even think of mystery force, but the exhaust nozzles on the engines now doing their job. Then they would provide a litte thrust sideways.

Anyways thanks for testing my idea BeefTenderloin

[KerbMav]

If you feel like drawing up all the different arrangements for four engines, I'll do it.

Ahm - just four engine - in a circle/square.

* *

* *

[Arran]

TEST 2 - One engine rotated 180 degrees. No gimball or control inputs.

IN THE VAB

AFTER TWENTY SECONDS

Deviated slightly down and left. Most stable of the three configurations.

You only have a asas, and you rocket diviates due drag..

I just added a sas unit too it, launched it, and it went perfectly straight up, all up too 23km without even going a half degree off..

It fell then again straight down, on the launchpad exactly where it launched from..

Its not the sas thats flawed, but your design.

asas and the sas are 2 different things, and they work now in the new logic in combination.. I think alot here that just dont understood this right..

Also added more sas doesnt help now by adding torque, but adding asas units help in making a rocket nimbler..

So is your rocket sluggish, with just a CM, add a asas module.. this will make them alot stable and better turning..

Here is an image of your rocked, how i tested it, how they should work

https://www.dropbox.com/s/turenz9t20dhv26/redesign.png

Grr, site is pretty slow, so added the link directly, image doesnt seem too load

NM.. did realize there was a update allready, guess this fixes most people that had problems their problems..

Edited July 25, 2013 by Arran