Wallows like a drunken cow.

[Vanamonde]

it should not start moving or rotating unless a force is applied

Thanks, that's a more concise way to phrase one of my major points. Similarly, any forces the controls are exerting should be enough to eventually stop a rotation, unless something is sustaining the rotation. If a rotation is getting worse in direct contradiction to the control forces, then something is wrong with the simulation rather than the ship design. I mean, I'm sure there's room for improvement in my ship design, but it's terribly frustrating to be struggling to overcome a problem that shouldn't be arising in the first place.

That being said, I'been playing the game for 3 months and have never had problems with the kraken before, I suppose because I like to build the smallest ship that will get the job done. One suggestion was to orbit higher. Any other tips for avoding the kraken?

[Ziff]

When I've got RCS firing, an ASAS, the capsule's inherent attitude control, and 3 SAS modules all trying to hold the ship still and the random rotations are not even slowing down, that's a problem. And the rotations are not constant anyway, as if forces are acting on the ship while it's in freefall vacuum, which is not a piloting challenge but a physics flaw. Furthermore, the SAS and ASAS modules need to be reworked anyway because they are too narrow to form stable structural components between the wider tanks and capsules. In short, it might be a challenge to try to fly a 787 with the steering controls of a Cesna, but it's not much fun.

Regarding your original post, get rid of the SAS. Only use SAS on ascent stages if they ROLL uncontrollably. (Not yaw or pitch.)

Regarding the .craft file you uploaded..

Your ship doesn't have a roll problem that I noticed when I flew it. It did have a yawing and pitching problem. You have a lot of weight at the top, and middle, and all the thrust at the bottom. This makes it easy to lose control especially with those FL-T16 tanks in the middle that are full. You want to try to keep the weight on the bottom, this is where designing stages becomes important.

I did manage to get into orbit, and as soon as I detached the orbit insertion stage and FL-T16 tanks I started flipping like a drunk Kerbal. First, the SAS and ASAS aren't playing well together. I don't know why, but when I removed the SAS and put another RCS tank on there to keep the spacing, the flipping problem almost completely went away. The second issue is you don't have a gimbaled engine on your injection stage which makes it really hard to control if the ship is out of balance. Basically, your lander and injection stage have too much fuel for just going to the Mun or Minmus, and the added extra weight is hurting you just about everywhere.

I calculated the delta-v of your injection stage and your lander. It's pretty nutty. You are just shy of 3k of delta-v in the injection stage, and the lander is well over 2k. You could redesign the lander to 3 external tanks/legs instead of 4, and then yank off the FL-T200 tanks off the injection stage. You will still have 1900+ delta-v in the landing stage, enough to land and return to kerbin, and the injection stage will be down to 2400ish, still more than enough for what you need to do. The overall weight reduction is a huge benefit for the control of the craft, also, you still have plenty of fuel in the TMI stage to de-orbit and get you almost to the ground, leaving your lander almost full of fuel.

I too lost the engines on this thing when it came out of time warp. No idea why, the 4 engines just waved goodbye and floated off into the sun. I made a few small changes and the issue went away. Here is the modified .craft file. https://dl.dropbox.com/u/93355756/Mk26%20Alite-RD.craft

As far as the original ship you posted, lose the SAS. Here's a pro-tip I learned from someone else whose name I forgot. (SORRY!) If you look down at the top of the capsule you can place the parachute just a hair off-center. This allows you to place an inverted small decoupler on over the parachtute and an ASAS on top of the decoupler. This is a temporary fix for the stability issue of trying to squeeze that between the larger 2m parts. Detach the decoupler on rentry to allow the parachute to deploy.

Again, start with your lander first. It has 6 radial engines for the landing/ascent stage. That gives it a Thrust:Weight ratio of 2.0 on Kerbin alone! 3 radial engines is more than enough to land on the Mun, with only 3 of the new landing gear as well. (They are BEEFY!) and all the weight you just removed increases your delta-v too! The new 2m engine only has a specific impulse of 330 in vacuum, so you can probably find a way to use a more fuel efficient engine for your injection stage that will also save some weight. Any chance you can upload that other .craft file? I apologize if the above is in any way confusing or sounds stupid, I'm a little burnt out from studying for summer finals.

[Vanamonde]

Thanks for the feedback, Ziff. I'll try those suggestions and let you know how they work out, though it may be a while because I'll be at work most of the day tomorrow.

I realize my ships are wonkily overbuilt, but my practice is to cobble together something that works and then trim it down where it proved to be excessive. Since the big one can't reliably get anywhere, I haven't been able to see what's left over when it gets back. Doing the math takes the fun out of it! Okay, to be honest, I'm too impatient and not sophisticated enough in such matters to do the math. I'll trust your figures. And really, I don't suck at rockets as much as that .craft might make it seem. That one was only a rough draft, which I only posted in its unfinished state to illustrate the deciduous engine problem.

I too lost the engines on this thing when it came out of time warp.

Weird, huh? Actually, that's part of the reason the mass distribution is suboptimal on the other ship; I've beeen afraid to use side-mounted engines in .16, though so far it doesn't seem to have cropped up again. Edited August 8, 2012 by Vanamonde

[Ziff]

I realize my ships are wonkily overbuilt, but my practice is to cobble together something that works and then trim it down where it proved to be excessive. Since the big one can't reliably get anywhere, I haven't been able to see what's left over when it gets back. Doing the math takes the fun out of it! Okay, to be honest, I'm too impatient and not sophisticated enough in such matters to do the math. I'll trust your figures. And really, I don't suck at rockets as much as that .craft might make it seem. That one was only a rough draft, which I only posted in its unfinished state to illustrate the deciduous engine problem.

I actually like the design because of it's uniqueness. That's what makes KSP awesome in my eyes. I don't think they're wonky, and I see exactly what you are doing. It's the same thing I did when I built a lot of my rockets before I started calculation things out, I refer to it as the boyscout method. Me: Hrm, never landed on the Mun before. How much fuel do I need? 'BRING ALL THE FUEL!' screams a little voice. YEAH! That's good. I'll stack all this fuel on here so I can hit multiple landing sites and just in case something bad happens... And then my ships just get huge from there. Once you are able to practice Mun landings by hopefully starting in an orbit around the Mun, you will be able to refine your landers with ease.

Then I started calculating delta-v and realized, wow, things get out of hand quick once you have an oversized lander for what you need to do. Check this post out, it is probably the most informative one I have read on how to build rockets with science and math. [Thread=15796]Kerbal Rocket Science[/thread]

Also, I use a handy little delta-v calculator found here http://www.strout.net/info/science/delta-v/ I just input the total mass, and empty mass after fuel is burned off, and select specific impulse and put the Isp in the field and whammo, instant delta-v calculations for the stage. The KICOIT Plugin was a fantastic program for doing the calculations in game, I hope it gets updated for .17. In fact, just an in-game calculator that I didn't have to alt-tab to use would be handy. Man, wish I had stuck with those programming classes I took years ago..

Edited August 8, 2012 by Ziff

[Zephram Kerman]

Sometimes if parts overlap, they can cause strange rotation forces. It's probably there all the time, but ASAS is able to compensate until you shut off the gimballed engines. Try shutting off ASAS during the ascent to verify this. The only way I know to fix this is to find and relocate the offending part. If all else fails, rebuild the same design from scratch.

[BillWiskins]

I see your point point BillWiskins, but it's not that just it's more difficult. When I've got RCS firing, an ASAS, the capsule's inherent attitude control, and 3 SAS modules all trying to hold the ship still and the random rotations are not even slowing down, that's a problem. And the rotations are not constant anyway, as if forces are acting on the ship while it's in freefall vacuum, which is not a piloting challenge but a physics flaw. Furthermore, the SAS and ASAS modules need to be reworked anyway because they are too narrow to form stable structural components between the wider tanks and capsules. In short, it might be a challenge to try to fly a 787 with the steering controls of a Cesna, but it's not much fun.

Aye, quite right sir. As I said - there are clearly control issues which need sorting. I'd just hate to get a new version one day and find that massive vehicles were able to turn and handle as easily as a tiny one. I realise that that is an oversimplification and not what you were getting at, really. Still - I hope that any revisions made still retain the feeling of 'heaviness' for larger craft. It should still be the case that if you design a massive vehicle, you should be prepared for it to feel like steering a glacier (and be required to adjust relevant design aspects accordingly).

[Fendleton]

I've had problems with ships spinning wildly with ever-increasing speed, with no ASAS ability to stop them (or even slow them. If anything, the ASAS tends to make things worse)...this was also most noticeable with the larger parts. The trick I used was to go to X10 time acceleration for a few seconds....going on-rails stops the rotation, and then you can (try to) keep it more steady when you come back into X1..without using stability aids. Doesn't always work though, sometimes things start up again...at which point apart from decoupling stuff to throw off mass or using landing legs to correct some of the excess roll (if you're lucky enough to be just rolling), there's really nothing to do except abort the mission.

[Vanamonde]

Lookie what just made 2 successful test trips to Mun and back:

Dispensed with SAS, made the upper stages lighter by drastically reducing fuel, rearranged things to discard the first 2 stages before reaching orbit so I'm not trying to steer with a county strapped to my behind, orbiting Kerbin at higher altitude, and other stuff you helpful guys suggested. (Fendleton, I couldn't use that trick because it kept claiming it was "under acceleration" and wouldn't go in to warp. ) Thanks a bunch for the help, gentlemen! (And ladies, if any of you were.)

Edited August 11, 2012 by Vanamonde

[Ziff]

Now that, good sir, is a nice looking rocket! Congrats on your success! Now off to Minmus you go!

[Vanamonde]

By the way, the problem isn't solved; it's just been made manageable. When I took one of the guys out for EVA during the transit just now, he floated around watching the ship do a pollen dance.

[Vanamonde]

Ugh. This again. Due to a download problem and then a heavy work week,

is my first 3-man rocket in .17. Despite 4 SAS and a large ASAS, it requires frequent manual corrections just to keep it flying straight. It wallows around in atmo on takeoff, then spins end over end in orbit, then slowly wanders off of attitude in intermoonal space. It's not the same as my previous problems (from the beginning of this thread) because I can always get it back under control with manual corrections. But as soon as I stop personally steering and turn it over to SAS/ASAS, it wanders off attitude again. Clearly, if I can get it back under control with manual commands, the ship has the attitude control authority to fly straight. But it's as if the SAS just isn't trying hard enough. What's the deal, and are other people experiencing this?

[EndlessWaves]

Are you flying with mechjeb? That doesn't seem to like using lots of SAS.

circular struts between the engines (well, fuel tanks) around the body of the machine may also help a little but it looks like those fuel tanks are directly attached.

[Vanamonde]

I have pinpointed the source of the problem. It's pretty technical, so follow me closely here. If you get confused and think the new RCS tank is the new ASAS, so that your rocket has no ASAS module at all, this will cause the rocket to behave as if it has no ASAS module at all. The problem can be remedied by installing the new ASAS module that you thought you had installed in the first place. In short, nevermind.

Edited September 27, 2012 by Vanamonde

[draeath]

I was gonna say...

A single Advanced SAS and 8 RCS quads (4 top, 4 bottom, centered around stage's center of mass) around my final stage has no problem steering and stabilizing my 100-ton rocket! I don't even need to use the RCS - the main stage's gimbals keep it straight (the Advanced SAS drives them) - and the final stage can still be controlled superbly decently (~ 4 seconds to execute a 180-degree pitch or yaw from standstill to standstill). The RCS is there "just because" - after writing this I'm actually thinking of removing it as the fuel tanks I gave them are a good quarter of the final stage's mass.

[Ziff]

I was gonna say...

A single Advanced SAS and 8 RCS quads (4 top, 4 bottom, centered around stage's center of mass) around my final stage has no problem steering and stabilizing my 100-ton rocket! I don't even need to use the RCS - the main stage's gimbals keep it straight (the Advanced SAS drives them) - and the final stage can still be controlled superbly decently (~ 4 seconds to execute a 180-degree pitch or yaw from standstill to standstill). The RCS is there "just because" - after writing this I'm actually thinking of removing it as the fuel tanks I gave them are a good quarter of the final stage's mass.

I am a firm believer in that RCS should only be used on craft that are going to perform orbital rendezvous. Otherwise you are just wasting precious fuel moving mass you won't really need. Particularly in the case of interplanetary travel. All the delta-v required to move that extra mass really adds up when you are trying to go from Kerbin to another planet and back.

[Vanamonde]

The ship I pictured, by the way, is the minivan that will pick up the kids from soccer practice and the Kerbalnauts from other planets. I'm planning on keeping my lander weights down by making them mostly-one way: get to, land on, then meet the bus in orbit for the ride back. That's why this ship has RCS (for rendezvous), no legs, and delta-V out the wazoo so that it can achieve orbit just about anywhere in the solar system and bring the boys home.

[draeath]

I agree. I put them on as a matter of habit... as when I last played we didn't have those launchpad stabilizer arms and I needed to use that to fight "rubber rocket" on the first few seconds of launch.

[Grizzlol]

If you can manage without the SAS modules then they are only dead weight and adds to the total height.

[RoboRay]

I am a firm believer in that RCS should only be used on craft that are going to perform orbital rendezvous. Otherwise you are just wasting precious fuel moving mass you won't really need. Particularly in the case of interplanetary travel.

I could not disagree more. RCS is ideal for trajectory correction on long flights, where tiny changes make big differences. A single puff from the main engines may be too much for the fine adjustments needed to line up an aerocapture at the proper periapsis and inclination. With RCS, however, it's easy.

Edited September 28, 2012 by RoboRay

[Ziff]

The ship I pictured, by the way, is the minivan that will pick up the kids from soccer practice and the Kerbalnauts from other planets. I'm planning on keeping my lander weights down by making them mostly-one way: get to, land on, then meet the bus in orbit for the ride back. That's why this ship has RCS (for rendezvous), no legs, and delta-V out the wazoo so that it can achieve orbit just about anywhere in the solar system and bring the boys home.

That is exactly what I did as well. It work's great as long as you are decent with orbital rendezvous, which I am not. I got it though, just took awhile.

I could not disagree more. RCS is ideal for trajectory correction on long flights, where tiny changes make big differences. A single puff from the main engines may be too much for the fine adjustments needed to line up an aerocapture at the proper periapsis and inclination. With RCS, however, it's easy.

I have never had that kind of issue, especially with the new NERVA engine. Maybe you have an unnecessarily high T:W ratio on your transfer stage? It could just be my minimalist approach , I don't bring anything I don't need, including ASAS.

Edited September 28, 2012 by Ziff