What is at 7000 m in Kerbin atmosphere ?

[ASnogarD]

Got my lifter flying straight up , canards flipping back and forth as usual with most craft... drop my first set of tank/engines, going up nicely... then I hit 7000 m my canards are all locked trying to stop a spin...unsuccessfully.

This is really frustrating me, I finally got a lifter with decent power, without 7 million struts holding the bloody thing together, it isnt even anything fancy ... a centre stack with 6 stacks in a ring set up in asparagus, the only deviation from the norm is the ring is spaced away from the centre with Modular Girder XL's and a Modular Girder with a Radial Decoupler (to make space in the ring for the payload and stabilising Modular Girders , otherwise my payload wobbles like a Saints Row Dildo bat).

The drop stacks are properly spaced apart, and at equal heights to each other, there is no deviation from each other to explain the spin... the fuel decreases evenly between the stacks, canards are placed in the centre stack at the CoM to try prevent spin, and added some more later on each stack... gimbals are turned off on the outer ring (the centre stack has a core of 6 areospikes and a T45).

Lifter goes up like a arrow, detaches first set without issues, then at about 7000 m .... my craft get delusional and thinks its a flipping cathering wheel going off.

[Flixxbeatz]

Got screenies?

[urablahblah]

Sounds to me like your craft is aerodynamically unstable. The front begins to have more drag than the rear, so it begins to flip around. Try putting some of the larger delta wings on the bottom of your drop tanks. Usually they can provide enough drag to keep a rocket from going topsy-turny.

[sporkafife]

Many of my larger asparagus staged vehicles suffer some sort of rotation on launch. It's probably because the ASAS we have at the moment isn't very well tuned to larger crafts. And the problem is always made worse when you drop any outer tanks, because it pulls the mass closer to the center of your rocket, increasing the rate of spin. Exactly the same as when a ballerina pulls their arms in to spin faster.

Hopefully the upcoming SAS improvements will mean the computers are better at handling the spin.

Edit: Ignore me, I completely misunderstood and thought you meant the craft was spinning on it's roll axis. One of the other responses about being aerodynamically unstable are probably true. Try moving all your canards to the back of the craft to increase the drag of the rear.

Edited July 7, 2013 by sporkafife

[sgt_flyer]

multiple possibilities : do you use an Asparagus design ? when you drop your first stack, asparagus has a tendency to start a slow spin. (one way to limit this spin is to disable and reenable ASAS just after dropping tanks)

second possilibity : something broke inside your rocket (check with F3 during ascent )

if something broke, it might be when your dropped stage unloaded from the game at 2.3km distance from your rocket. (that can cause a slight lag when the thing is unloaded)

in the end, where are your canard laced on your rockets ? (i usually prefer the small winglets for controlling rockets at the base than canards )

[ASnogarD]

I worked out what it was...

The usual answer 'Moar struts' *sigh*

The 6 outer stacks were attached to each other making a ring of struts keeping it stable, when I dropped my first pair of stacks, the remaining 2 pairs were joined in 2 separate stacks which caused the stacks to sway slightly causing a slight spin that just got worse and worse... had to add another set of struts to connect the 4 stacks after the first pair of stacks got released, this stopped the swaying which was barely noticeable and therefore stopped the spinning.

It wasnt specifically 7000 m, it just took that long for a slight sway to start a spin that couldnt be dealt with by the ASAS.

[Maddmatts]

The higher you go up, the less effect aerodynamic surfaces will have. 7km is just the breaking point of the balance between whatever's creating the twist and the diminishing force of the canards. I'm guessing the twist is from flexing of parts from the outer engines. Since you mention clustering I'm assuming the center stack uses a 3M tank. Try replacing the cluster with a single skipper or mainsail and see if the twisting goes away.

Edit: Ninja'd by the OP. Good work on the troubleshooting.

[ASnogarD]

Picture here to explain it better...

Initially the red struts wasnt on the craft, so when the stacks marked with blue arrows dropped the green struts broke away leaving the pair of purple circled stacks connected to each other, and they started to sway slightly which initiated the spin which was small enough at the beginning to be controlled but it grew until it was a uncontrollable spin.

Still annoying, its nothing really fancy and the craft wasnt rattling or shaking about like some I seen on the videos, just that spin which made gravity turns impossible. I had tested it with a Orange Tank and powerwise the lot got up to orbit with the centre stack having all its fuel left.

EDIT: The cluster is a single T45 , with 6 aerospikes ... the stacks on the ring have mainsails, I was / am planning to try phase out some of the mainsails for clusters so that when the initial lift is done more efficient clusters can take over.

Edited July 7, 2013 by ASnogarD

[Aragosnat]

Yes. More struts. As I would have struted that thing up completely to each other and the central tank. Wisdom of smaller rockets plays big parts.

[Vanamonde]

Your tanks are placed so far out to the sides that any rotation they may pick up is going to have an awful lot of angular momentum. I suspect that at lower altitudes, the air is thick enough that the canards can counter-act this, but once you get to thinner air, you're left with only pod torque and the ASAS trying to exert control. Replace those long struts with short ones, or even small hardpoints, and I suspect the problem will become much more manageable.

[justadude]

In my experience turning off the gimbals on all but the center engine eliminates any wobble that is not top heavy wobble (which is solved with more struts).

[Travisfv]

I too am finding that a small rocket can do just as much as a big rocket, perhaps more. All my rockets have been big ungainly things, that either spin out of control or tear themselves apart without the addition of a superstructure holding the whole thing together. Literally, all my big rockets have gains beams coming out of the them with a ton of struts to stabilize.

You know what I find really helpful with big rockets? Flight engineer of some type. Something that tells you TWR and D-V. Sometimes adding more rockets slows the whole thing to the point where you lose Delta-V and TWR. And you'll find engine you used to put in upper stages were no good, so you switch to something more compatible. And you may not need 6 Mainsails around the center rocket. You could get away with skippers, conserving fuel and not putting so much stress on a joint and twice their thrust will.

[ASnogarD]

I put the tanks far out so as to be able to make the 'cradle' with the smaller Modular Girders, to secure my future payloads... just my test Orange tank payload wobbled no matter how many struts I put on from the decoupler -> tank.

The radial decoupler with legs did not leave enough space to make the 'cradle' and relying on the stacks as places to hook struts to was useless in orbit.

I suspect the Modular Girders XL let the stacks twist a bit which creates spin, my solution I posted early works until there is only 1 pair left, then the spinning game begins... but its manageable but not desirable... I want a more firm lifter, so I can eliminate it from any future payload issues (basically want to be able to say this lifter can lift x tons with no issue...any issues will then be the payloads fault to fix).

[Vanamonde]

I'm not sure exactly what you're describing, but how about rotating the booster mounts so that they fit between the cradle parts, and so could be moved in without hitting the cradle? Or simply most the booster mounts down so that they're below the cradle arms when you shorten their mounting arms?

[Capt'n Skunky]

Canards are designed to go in front of your CoM/CoL. If you use them as tail fins, they will operate opposite of what you want, flipping your rocket. Use the tail fin parts made for this, AV-A8 I think they are called.

Cheers!

Capt'n Skunky

KSP Community Manager

[ASnogarD]

At 6 x symmetry there isnt enough room between tanks, and to place the tanks lower would a) remove any possibility of adding more fuel tanks on top if needed and mean I need to move the de coupler to the orange tank.

It has been my experience that if the de-coupler is high on the stack it is prone to simply breaking at random whereas lower on the stack it seems more durable... I have yet to have the craft disassemble itself due to structural failures whereas my other attempts have often had one of the stacks simply detach and go walkies on its own.

Why I need the cradle is because a standard tall rocket seems to wobble like crazy no matter how many struts I use between tanks and components... one experiment had over 90 parts with just the centre stack and a orange tank payload, adding the 6 other stacks would of made the craft very part heavy... and thats without a payload.

BUT as I said earlier it seems the girders cause the attached tanks to sway / twist so the engines introduce spinning... besides the thing is near uncontrollable in orbit without RCS, without pilot torque even adding a SAS module on top of my ASAS module didnt allow me to control the craft when the engines were offline.

Back to the drawing board for me.

EDIT: Thanks for the tip, I'll check out the part you advised, Skunky

Edited July 7, 2013 by ASnogarD
Replying to Skunky

[Whackjob]

Clearly you need MOAR STRUTS!

"Yo, dawg! I heard you like girders and struts, so I put girders and struts in your girders and struts, so now you can strut while you girder and girder while you strut."

[stupid_chris]

Aw crap! Skunky beat me to it :l

Yeah your problems are probably the canards. When you drop that pair of boosters, you loose stability, and after a while from burning fuel, the canards are powerful enough to act against what you'd expect them and throw everything off course. Switch for winglets and it should be good.

[HoY]

There's another problem with the control surfaces. Take them off the outer boosters and install them in 6-way symmetry at the base of the center stack. Like the others said, not canards.

With the outer boosters so far from the CoM, the location of those winglets gives them extra torque, and after your first stage is dropped you lose part of the control you had from those winglets. It also causes the remaining winglets to have non symmetrical control forces applied to the rest of the rocket.

With them in the center you maintain the same control for your entire ascent until your above the thick atmosphere.

[Whackjob]

And alternatively, you can get away with no winglets or control surfaces at all, if you use gimballed engines, SAS, and make the entire structure rigid with good strutting. My AEIA Mk4 hasn't got a single control surface, and it amasses somewhere around two or three thousand parts.

[NeoMorph]

In my experience turning off the gimbals on all but the center engine eliminates any wobble that is not top heavy wobble (which is solved with more struts).

This is essential for some of my rockets. Sometimes I leave them turned on and as they rip themselves to shreds I go "But that rocket is totally flyable... I've used it thousands of times... oh... gimbal lock. drat!"

[ASnogarD]

I gave up...

After failing with that lifter I tried a more normal one... first time 1 out of 6 identical stacks fell off at the launch pad, I added more struts and this time a stack , 1 of identical 6 stacks decided to break off and fly free right through the pay 1 min and 13 secs in the flight.

Why 1 ?

Why 1 min 13 secs into the flight will 1 , one , uno , ich , ein stack fail... wouldnt all the stacks fail because they are all the same?

Another try later, and the ASAS module collapses on itself despite having 8 struts connecting it to the fuel tank and have another 8 struts connecting to the probe component... need more bloody struts ?

I put struts to stop a fault, one flight it works next flight it doesnt...same flight profile (straight up)... I added more T45's to my cluster to see if I could get more control , even a SAS module and a 5 degree incline is understood by the game to mean flip over and fly straight down and refuse any further instructions.

... tested the very same rocket with a manned capsule and the craft controlled perfectly.

I give up for the time being, game is too finicky for me to deal with on a really hot muggy day, deleted the lot and will start fresh some other time.

[Whackjob]

Might want to restart with stock in case you've got mods. Otherwise, PM me with the .craft file for it and I'll work my strutting magic. You don't need to use what I give back, but you can use what I send you as something to examine.