Avoid roll and ASAS flexing during ascent with vertically staged rockets?

[zarakon]

When I make somewhat large/tall rockets, I tend to get two annoying issues during ascent.

1. The rocket starts to roll as it picks up speed, and I can't overcome it unless I add fins. Even with the fins, it's annoying that I need to keep fighting against it. What could be causing this roll, and is there a good way to just design it out?

2. Using ASAS tends to make my rocket wobble and flex a lot. Even when it doesn't break it apart entirely, it's probably pretty inefficient to have the engines vectoring back and forth constantly. Adding a bunch of struts helps, but is there any way to just get the ASAS to behave better and stop overcompensating? Does it help to have more of them? Does it matter where they're placed? Would it help to add regular SAS modules too?

[Saberus]

Those are some big rockets.

I can see how a little roll can be irritating, but in reality a bit of roll can help the rocket maintain a stable flight path. If you need to counter it and are having a hard time, try using ALT and the roll key that counters your roll. ALT sets the controls to trim mode, so after you set enough counter-roll, you can let go of the keys and the ship will hold that trim.

The ASAS has trouble with the smaller adjustments, unless you hit caps lock to activate the fine controls, then it can usually settle out the rocket's wobble. Conventional SAS can help, but they add a good deal of weight as multiple are needed. SAS modules have the greatest impact when placed at the top and/or bottom of the stack you're trying to stabilize.

[zarakon]

So activating fine controls affects ASAS behavior?

[Temstar]

Rolling is generally caused by radial stages flexing obliquely relative to each other, so instead of pointing straight up they become a helix and induce roll forces.

To fix this you need to add diagonal X shaped struts between the radial stages like so:

ASAS causing wobble via oversteering caused by too much control authority. For example in your case you already have lots of mainsail which gives you already too much steering, adding steerable fins will make it even worse.

Edited October 16, 2012 by Temstar

[EndlessWaves]

I've never had problems with ASAS's amount of control, for me it tends to be bang on. The engines will make fine adjustments all the time, that's just how it works and shouldn't affect anything.

What can cause problems is having the command capsule not in line with the centre of mass. ASAS seems to compensate based on the readings from the command module so if your thrust vector isn't aligned with your centre of mass to command capsule vector you will see it giving the wrong corrections.

[zarakon]

I'll give the X-shaped radial struts a try, thanks

ASAS causing wobble via overstair is caused by too much control authority. For example in your case you already have lots of mainsail which gives you already too much steering, adding steerable fins will make it even worse.

Too much control authority coming from the thrust vectoring of the main engines?

[Temstar]

Too much control authority coming from the thrust vectoring of the main engines?

Yes. Basically ASAS doesn't know how maneuverable your rocket is, developers have set some hidden value on how hard it steers based on some assumptions on how they think the "average rocket" would handle. So if you rocket is way more maneuverable than how Squad envisioned a rocket should be then ASAS will constantly oversteer when it tries to adjust for small errors, overshoot the correction then oversteer the opposite direction, hence the wobble and the constant back and forth gimbling of the engines.

That back and forth gimbling is very wasteful of delta-V by the way, because any time your engine is not thrusting you directly up it's using some of its thrust to steer the craft and you lose that bit of thrust. In this case it's doing a whole lot of steering to try to get the rocket stable, yet the very fact that it's steering so hard is what's causing the rocket to wobble in the first place and you lose a non-negligible amount of delta-V to steering loss. One of the big hidden disadvantage of mainsail is that it has too much control authority for it's own good. First stage engines (of which mailsail is one, since it has high thrust and low Isp) need no to very little vectored thrust, since:

1. In an asparagus rocket, high Isp gimbling engines that fire together with first stage engine like LN-V or LV-T45 provide enough control authority. During first stage burn you are not turning the rocket, so you only need enough control authority to keep the rocket upright.

2. If for some reason you do need more control authority, you can alway add steerable fins since first stage burn is through the atmosphere

So the fact that mainsail has very powerful vectored thrust (as it has both big gimble range and very high thrust) actually works against it and destabilises the rocket.

Now the steerable fins on your rocket makes things even worst because:

1 - you already have too much control authority from the mainsails, adding steerable fins makes things even worse. If you have to control spin with fin then use the non-steerable fin

2 - The fact that fins are at the very bottom means that when they move to steer your rocket they give the bottom of those radial boosters a sideway force, thus twisting them even harder into a helictical shape. You will see this when staging. You know how when you stage radial stages without fin then more or less fall straight down? Yet if they have fins on them when you stage they instead twist away from your rocket? That twist is also happening when the stages are burning and unless the radial boosters are very stiff and resist the twisting your whole rocket will start to spin.

Edited October 15, 2012 by Temstar

[AmpsterMan]

I'm starting to think it also has something to do with placement of struts. I am doing some experiments on that now to see how it is.

[foamyesque]

ASAS will always oversteer on large rockets; it's been a problem since it was introduced; I believe because it takes angular momentum as an input instead of angular rates, resulting in it treating a large rocket tilting a fraction of a degree like a small rocket doing backflips (and, therefore, applying maximum reverse power). The resulting humming-bird back and forth will tear apart otherwise perfectly fine rockets. It's not an issue of having too much control authority at all; rockets that manouver like flying bricks on manual control will still have ASAS-induced destructions.

For roll, strut anything on a radial mount.

[Temstar]

ASAS will always oversteer on large rockets; it's been a problem since it was introduced; I believe because it takes angular momentum as an input instead of angular rates, resulting in it treating a large rocket tilting a fraction of a degree like a small rocket doing backflips (and, therefore, applying maximum reverse power). The resulting humming-bird back and forth will tear apart otherwise perfectly fine rockets. It's not an issue of having too much control authority at all; rockets that manouver like flying bricks on manual control will still have ASAS-induced destructions.

For roll, strut anything on a radial mount.

I can show you a 100ton to LKO rocket that doesn't wobble at all if you like. It's only steered by 7 LV-N, 4 SAS and command pod.

[foamyesque]

"4 SAS".

Was that intentional or did you mean ASAS?

[Temstar]

It's intentional, it has 4 SAS that the spacecraft itself carries all the way to orbit. It's a 9600L capacity interplanetary craft and I figure I need a bit more steering than just the command pod for when the engines are not firing. Aside from the LV-N (which are lit at lift off and fire all the way to orbit), the 4 SAS and the command pod the rocket's lift vehicle are all asparagus staging radial boosters and none of them have any steering at all. If i remember correctly the left vehicle around the spacecraft is:

Aerospike x 10

Big SRB x 8

400L tank x 64

200L tank x 4

No vector thrust engine or fins at all. During the burn to the orbit the rocket is only steered by the spacecraft itself and it flys and gravity turns fine.

I'll put the craft file up once I'm home.

Edited October 15, 2012 by Temstar

[foamyesque]

SAS are magic torque machines and not what is under discussion here.

[Temstar]

SAS are magic torque machines and not what is under discussion here.

The point is too much control authority makes your rocket unstable, so I wanted to show a big stable rocket which ASAS can control very well with very little control authority.

[foamyesque]

The point is too much control authority makes your rocket unstable, so I wanted to show a big stable rocket which ASAS can control very well with very little control authority.

Okay, two points:

1. 9600L isn't that big a rocket.

2. You don't actually have an ASAS unit mounted.

[Temstar]

2. You don't actually have an ASAS unit mounted.

Oh I do, I have ASAS on all my ships.

I have seen lot bigger ships of course, but you can extrapolate the same conclusion upwards.

In my view, mainsail is only good for very outer stage of huge asparagus staging rockets, sort of like a rich man's SRB. It's Isp is too poor for upper stages and it's vectored thrust is too powerful for small rockets.

[foamyesque]

Oh I do, I have ASAS on all my ships.

I have seen lot bigger ships of course, but you can extrapolate the same conclusion upwards.

In my view, mainsail is only good for very outer stage of huge asparagus staging rockets, sort of like a rich man's SRB. It's Isp is too poor for upper stages and it's vectored thrust is too powerful for small rockets.

Okay, so you do have ASAS on that rocket? Go tell me what its doing to your command inputs, and I'll tell you what it does to large rockets: Constant maxing oscillations in either direction.

It will do this even at launch even with no control inputs on a large rocket. Even if that rocket's stable. Especially if it's stable, in fact; if it's tilting, ASAS will actually compensate consistently, which is actually a better scenario for ship stability.

You can make a small ship with all the CA in the world-- hook a mainsail to a minitank!-- and ASAS can and will fly it fine, making only fine, small adjustments. Build a big rocket with less CA, and ASAS will rip it to shreds.

[Temstar]

Okay, so you do have ASAS on that rocket? Go tell me what its doing to your command inputs, and I'll tell you what it does to large rockets: Constant maxing oscillations in either direction.

Well no, not quite. It does oscillate quite a bit at take off but not quite max in all direction.

You can make a small ship with all the CA in the world-- hook a mainsail to a minitank!-- and ASAS can and will fly it fine, making only fine, small adjustments. Build a big rocket with less CA, and ASAS will rip it to shreds.

Fine, fly it and see for yourself if ASAS shakes this ship up.

The one I attached is the stock version without the docking and fuel transfer parts. To get it into orbit:

1. Turn on ASAS and throttle up to max then fire the engines

2. Stage SRBs when they run out, stage first set of four liquids when you see 28 empty tanks

3. Once you've stage the first four liquid boosters, nose down to 45 degree until the 2nd set of 4 liquids run out (another 28 tanks), stage those

4. Nose down to about 15 degree on your last two boosters and then watch your apoapsis until it gets near your target orbit then nose down to 0 degree. Circularized orbit as normal.

Edited October 16, 2012 by Temstar

[kellven]

Is the root of the problem, I believe. Unfortunately, my fix to the over-wobbly part connections just doesn't work as well as it did in 0.14. Inflight, it's mostly fine, but the ground contact code we have now makes stuff go crazy if the part connections aren't allowed to wobble and slide all over.

Only safe way to deal with stack flexing is to extend something radial, and strut the parts to the distal ends of that, then strut the radial parts so they stay mostly in place.

[tuguley]

This is how I prevent FLEXING in my ship, due to ASAS.

It adds a little bit of weight, but it distributed the thrust vectoring on the mainsail engines throughout the entire ship.

I have never have a ship flex itself apart using this design. Even tested it with asparagus staging with 6 main sails (That's A LOT of thrust vectoring) and an even longer ship.

http://imgur.com/a/F7etE

[Rickenbacker]

Fine, fly it and see for yourself if ASAS shakes this ship up.

ASAS Can't shake your ship up, because there are no vectored thrust engines. The problem is with big rockets and powerful, vectored engines. ASAS just bangs the gimbals against the stops over and over, eventually bending and breaking a tall rocket, even if there's nothing wrong with it otherwise.

[Temstar]

ASAS Can't shake your ship up, because there are no vectored thrust engines. The problem is with big rockets and powerful, vectored engines. ASAS just bangs the gimbals against the stops over and over, eventually bending and breaking a tall rocket, even if there's nothing wrong with it otherwise.

I agree it's the big vectored engines that sake up a ship, but my argument is if you have the right amount of vectored thrust (ie, a lot less control authority than a ship powered only by mainsails) then a big ship can fly fine too with ASAS.

In my case, it's the 7 LV-N cluster at the center that's gimbled on my ship.