Help with spacecraft turning over in atmosphere

[Wampa842]

This is Tapio II, my Mün/Minmus lander.

The LM has a range of scientific equipment attached to a Mk2 Lander Can, powered by 5 small engines; more than enough to get back from Minmus.

The TMI unit consists of one 2.5 m engine and four 1.25 m engines, a few winglets and RCS ports for stability in the atmosphere.

The main lifter has five identical tanks and engines in asparagus configuration, a 2.5 m reaction wheel on each, and lots of control surfaces.

I've got FAR installed, that's the reason for the number of winglets.

I've completed several successful missions to both moons using it. However, there's always a problem: around 15-17 kilometers, when I'd commence my gravity turn, it goes out of my hand and spins around uncontrollably. Despite the vast amount of winglets (there's still air at 17 km, right?) and reaction wheels, I can't regain control until I'm at 25-30 km, and even doing so means a great loss of velocity.

What is so fatally wrong with my design? Did overlook some details in aerodynamics?

P.s. I'm a noob when it comes to rocketry.

Edited October 29, 2013 by Wampa842

[TechnicalK3rbal]

There is still air, yes, but not much. Try doing your gravity turn at 10km or add RCS.

[JayCheetah]

There is one problem.

You are not using stock parts, so how would anyone not using your mods solve this?

My first guess would be:

Fuel imbalanced.

[capi3101]

Fuel imbalance is a strong possibility. Another is high center of mass - where is the center of mass on this ship, BTW?

That upper set of control surfaces is not helping you out any; I'd suggest losing those.

[jfx]

FAR makes heavy rockets flip-happy. make sure you have plenty control authority (not these puny stock control surfaces). Also your lander is pretty overengineered. my guesstimation is that you'll only need one of the five engines on your lander to land&return from minmus.

The FAeRodynamics of these fat landers will always be horrible, so the problem will most likely persist. I tend to launch my brick-shaped payloads at 5-10° and then pretty much leave it at that until 15k. Then I'll slowly test the waters while moving towards the prograde marker. This also requires a lot lower TWR than a stock launch, otherwise you'll end up with a very weird/inefficent ascent pattern and something like 1000m/s orbital velocity at apoapsis.

Edited October 28, 2013 by jfx

[Umlüx]

check the fuel pipings. maybe there is something wrong with the fuel drain and the whole thing gets imbalanced.

[Sirrobert]

I don't see any torque modules on there except for the landercan. You probably don't have enough to controll it once it starts tipping.

Try adding a few SAS modules, see if that fixes it

[Wampa842]

There is still air, yes, but not much. Try doing your gravity turn at 10km or add RCS.

RCS ports (modded, slightly stronger) were already present, they don't make such a great difference.

I tried starting gravity turn at 5° at 10km, ship flipped nevertheless.

Fuel imbalance is a strong possibility. Another is high center of mass - where is the center of mass on this ship, BTW?

That upper set of control surfaces is not helping you out any; I'd suggest losing those.

The center of mass is between the fuel pipes and the top of the first stage tanks. Removed upper winglets as you suggested.

FAR makes heavy rockets flip-happy. make sure you have plenty control authority (not these puny stock control surfaces). Also your lander is pretty overengineered. my guesstimation is that you'll only need one of the five engines on your lander to land&return from minmus.

The FAeRodynamics of these fat landers will always be horrible, so the problem will most likely persist. I tend to launch my brick-shaped payloads at 5-10° and then pretty much leave it at that until 15k. Then I'll slowly test the waters while moving towards the prograde marker. This also requires a lot lower TWR than a stock launch, otherwise you'll end up with a very weird/inefficent ascent pattern and something like 1000m/s orbital velocity at apoapsis.

So I should trim out the fat from my lander, and re-work my liftoff vectors. I'm afraid that although one tank of fuel should be enough to return from Minmus, based on my observations, it's not enough to reach the Mün's escape velocity.

I'm also installing the B9 pack, in hope to find sufficient control surfaces.

check the fuel pipings. maybe there is something wrong with the fuel drain and the whole thing gets imbalanced.

There are only four fuel pipes in the asparagus lifter, they're all set up correctly.

I don't see any torque modules on there except for the landercan. You probably don't have enough to controll it once it starts tipping.

Try adding a few SAS modules, see if that fixes it

I have one ASAS module, and one reaction wheel module on top of each radial fuel tank (four 1.25m, four 2.5m), but they're not enough to negate the torque generated by the drag.

[aven17]

When my rockets get wobbly it is either because I have fuel lines connected asymmetrically or I don't have enough control through engine gimbal or SAS. You can also hold off your turn until even higher altitudes when you have ditched more of craft's mass. It is also a possibility that mods are not cooperating with each other.

[Wampa842]

The problem is now fixed. It was caused by the lack of decent control surfaces.

I downloaded B9 Aerospace and Ferram's joint reinforcement plugin; even my largest rockets work like a charm now.

[KerbMav]

Put control surfaces as low as possible - stages for higher atmosphere cant use them as efficiently anyway.

Have only the center engines' gimbals active.

Have a low CoM - the higher your CoM the later you should start your gravity turn, throttle down when you do so.

Add some more reaction wheels.

RCS is only really doing something useful here in thin air/vacuum.

Procedural Fairings could help too, especially when using FAR.

Try building a less wide lander/smaller lander

Dont forget you can turn the lander upside down if you cant figure out how to put it onto your launcher - you may want to install a probe core upright and control from there though, as your pod's navball will be flipped 180 deg.

[Deathsoul097]

I am just going to say this for future reference. Nosecones.

FAR implements proper aerodynamics, meaning you need nosecones, which I see a distinct lack thereof.

[SRV Ron]

Experimenting last night with the tiny Rocomax engines set up as asparagus on the small NovaPunch tanks, I ran into the same issue. It is atmospheric effects that will cause the rocket to flip over the instant I start a gravity turn. With an extra SAS, I could make the turn but it required no vectoring to start it. It even happened on this design without an external SAS;

It may be an issue with NovaPunch as some fly well while others shake badly. You can stack five of their 1.25 x 6 meter tanks with no bracing, yet stack two 2.5 x 12 meter tanks and they will fall over on the launch pad.

[Hodo]

Ok I am going to ignore some of the ignorant answers you received and give you the simple answer.

FAR requires an aerodynamic shape to work properly. If you put rockets on the bottom of a brick, it will fly but it will fly badly. This is basically what you have done. Your lander is about as aerodynamic as a brick and your main stage at the bottom is just as bad. There is a GREAT deal of drag on the nose/top of your rocket, and when you get near transonic speeds mach 0.7-1.2 you start basically bulldozing your way through the air. This means the back of your craft will have less drag than the front at that point. So it will want to flip over and fly backwards.

The easiest fix is to use nose cones, or get the procedural fairings to make a aerodynamic capsule around your lander to help it cut through the air not fight it.

http://forum.kerbalspaceprogram.com/threads/39512-0-22-Procedural-Fairings-2-4-2-tech-tree-and-minor-improvements-%28Oct-19%29?highlight=procedural+dynamics

[SRV Ron]

I am just going to say this for future reference. Nosecones.

FAR implements proper aerodynamics, meaning you need nosecones, which I see a distinct lack thereof.

Doesn't work. Design is very unstable once one tries to turn. (Unless you add the extra 1/2 ton SAS ring.)

I even rearranged the solar panels, no go, note hard over control;

The results of that flight.

Had better luck with this one even though it was difficult to keep under control while in the lower atmosphere;

In orbit.

Yet, this one flies perfectly stable with no extra SAS ring;

All were using that mini Rocomax engine on NovaPunch fuel tanks.

Design has a lot to do with what works and what becomes unstable and wants to swap ends during the orbital turn or sooner. You only find out when they are built and test flown.

Edited October 29, 2013 by SRV Ron

[Lexif]

SRV Ron, as your tanks empty from top to bottom, your center of mass moves back. Your payload and upper stage are very light, so your center of mass will move behind your center of drag. When this happens, your rocket wants to fly backwards if drag is properly implemented (FAR). The solution is to add control surfaces at the tail end, move your center of mass up (maybe even ballast at the top), and decrease drag on the nose (fairing). Maybe also change your staging. Try looking at the lift and center of mass indicators in the VAB.

[Wampa842]

Ok I am going to ignore some of the ignorant answers you received and give you the simple answer.

FAR requires an aerodynamic shape to work properly. If you put rockets on the bottom of a brick, it will fly but it will fly badly. This is basically what you have done. Your lander is about as aerodynamic as a brick and your main stage at the bottom is just as bad. There is a GREAT deal of drag on the nose/top of your rocket, and when you get near transonic speeds mach 0.7-1.2 you start basically bulldozing your way through the air. This means the back of your craft will have less drag than the front at that point. So it will want to flip over and fly backwards.

The easiest fix is to use nose cones, or get the procedural fairings to make a aerodynamic capsule around your lander to help it cut through the air not fight it.

http://forum.kerbalspaceprogram.com/threads/39512-0-22-Procedural-Fairings-2-4-2-tech-tree-and-minor-improvements-%28Oct-19%29?highlight=procedural+dynamics

That's what I figured. Adding a simple fairing and a few nosecones made me able to launch even heavier landers to the Mün to rescue stranded kerbonauts.