Problems with a payload.

[Draconiator]

First off, this:

Biggest payload I've ever launched (love that lifting vehicle, best one I've ever made), however, as soon as I'm down to the final pair of boosters, and I have added lots of struts to correct this, no dice. But what happens is the rocket starts swaying. And trying to correct it midflight only makes it worse. Then usually what happens is after a while of swaying, all the xenon tanks fall off, and that is usually the beginning of the end for the whole thing. I've tried putting reaction wheels everywhere, nothing. How do I fix that?

As for what it is...a pretty ambitious interplanetary ion ship. 32 engines, and enough xenon tanks to fly virtually forever. I've also taken all the fuel out of fuel tank parts so it's as light as possible.

[Liowen]

Wel if you do not need the engines to gimble you could turn that off, or add more SAS modules. I noticed you do not have one on the center tank below the decoupler, you might want to consider putting one there as well. Also the tanks your ion engines are attached to should have some strutting to each other to maybe steady it some. Lastly the batteries worry me some, I tend to use the smaller square ones now and hide them inside the big SAS module. Without seeing where the swaying is happening does make this a bit harder to answer though, but maybe some better rocket builder can see something I am missing.

[Xavven]

Give us a screenshot of your rocket at the time swaying starts. Usually it means your rocket is just too long (if anyone says, "that's what she said" I will hurt you). One way to fight this is to turn SAS off. Under many circumstances it can increase swaying and instability. Another way to fix it is by changing your design such that the payload is lifted by radial boosters that are attached at or above its center of mass (it's "pulled" to space instead of pushed).

[SkyRender]

Good grief... I think it would be simpler just to assemble your ship in orbit in two pieces. Or even better, use something more powerful than ion engines.

[RobotsAndSpaceships]

There are so many things wrong with this picture.

[Draconiator]

lol, yeah, I finally got it up there but it took some really intense rocket wrangling....and then when I got to orbit, the thing had so many pieces the game crashed. Back to the drawing board I guess...

[Specialist290]

Looks like this one belongs in Gameplay Questions. Moving

[Anquietas314]

Wow... at that scale, I think you would have a far better time (both launching and flying, not to mention refuelling) if you used a (sizeable) LV-N cluster and LF+O. You should use the 2.5m tanks instead of the 3.75 ones though - they have a better wet/dry ratio (= moar deltaV!). If you insist on using ion engines, the stack tanks are marginally more efficient in terms of wet/dry than the radial ones.

Interesting launcher design too, but probably overengineered . Is it at least SSTO? It's hard to tell from the screenshot.

[Draconiator]

I DID record the whole flight if anyone's interested, from launch to where KSP crashed on me. you can see my launcher in action too. it's probably not safe for anyone on the ground though...heh, when I jettison the boosters, they fly clear of the rocket but smash into each other a bit later. I'll get that uploaded tomorrow.

[Anquietas314]

I DID record the whole flight if anyone's interested

I'd love to see that.

when I jettison the boosters, they fly clear of the rocket but smash into each other a bit later.

Boosters? What boosters? Oh wait... do you mean the big engines at the bottom? Or heaven forbid the orange tanks?

[Draconiator]

The orange tanks, I told you, it's BIG. might be a bit over engineered really. works wonderfully though. I might have to edit the video down if it doesn't upload to Youtube

[Anquietas314]

The orange tanks, I told you, it's BIG.

Well, I guess at least Jeb would be proud; moar boosters!

As long as the video's 15 mins or under you should be fine - if it takes you that long to get to orbit (after accounting for lagsbane) you're doing something wrong, even with spaceplanes

[Draconiator]

Bleh, the movie recorder obliterated the file, gah

[Draconiator]

Going to launch something else with the same lifter, really want you guys to see that thing

[Anquietas314]

Going to launch something else with the same lifter, really want you guys to see that thing

If you're having difficulty getting a youtube video to work, you could always just upload the .craft file (this is stock, right?) and we could try it ourselves.

Except me... gaming rig's broken

[Draconiator]

That okay if I name it that by the way? "The safety of our new launch vehicle is impressive. The safety of people on the ground after it launches...not so much." -Jebediah Kerman

[Anquietas314]

That okay if I name it that by the way?

LOL! I had an impact on the name of someone else's youtube video . Yes it's okay lol.

*watching video* You used actual SRBs as sepratrons? Nice.

Your TWR is way too low though; you want it to be about 1.6ish at launch in stock aero and maintain at least that most of the way up (MOAR BOOSTERS!). Honestly it'd probably fly just as far if you just ditched one orange tank from the core and each booster. Gravity losses bite .

You want to be a lot more aggressive with the gravity turn too if you're going for efficiency. More TWR will allow you to do that. Ideally you want to be nearly horizontal by ~40km with your apoapsis at or above 70km in my experience. If you install the Kerbal Engineer Redux mod that'll help you optimize your flight path ; you want to keep "atmospheric efficiency" at 100% during the vertical ascent, and then reasonably close during the rest of the flight (it becomes impractical/unnecessary at higher altitudes to keep it there since terminal velocity skyrockets).

You're also better to launch into as low an orbit as you can without grazing the atmosphere (due to oberth effect). Obviously with that rocket it's not really practical though because it's so heavy / TWR is too low . Add struts between the core and the payload though - that'll help with the flexing late in the flight

[Red Iron Crown]

That payload is ambitious. I spend most of my play trying to get one or two kerbals somewhere, OP is taking over thirty. All aboard the Ion Bus!

As for the wobbly control, try backing off the throttle a bit when it starts to wobble. If that doesn't work, I'd put a 2.5m probe core near the decoupler to the payload and controlling from there during ascent.

[Xavven]

I laughed out loud at the RT-10's as sepatrons! Very Kerbal, but on the other hand, quite effective. Bravo

As suspected, the wobble as you're circularizing is due to overall rocket length vs width. Since you asparagus'ed away all the other boosters, there was less thickness to stiffen the craft. Of important note is that you're very top heavy at that part of the flight which aggravates the issue dramatically.

I'm not saying you should design a pancake rocket because those aren't realistic and actually don't work well with FAR if you're running it, but the shape you end up at the 10 minute mark is the problem.

[Draconiator]

What about the smallest? Again, video title says it all. Watch for the perfect dismount at 13:48. Micheal Phelps eat your heart out!!!! I had actually accidentally hit Space a bit too early, and oops.

[Th3F3aR]

As many other already have stated, it's the tail lenght that causes the swinging. Therefore, try to cope with that by strutting the tail to your payload. That might help a bit.

Other suggetions are:

-Raise your TWR by adding moar thrust about 1.6 should be sufficient -> You will end up faster in space, reducing deltaV requirements by gravitational drag and atmospheric drag

-Reduce your weight by removing one or two fuel tanks at your middle engine (just to reduce length and weight for your last stage) -> result is higher TWR and less overfuel

^- I assume you have a good part of fuel to spare as your g-Turn/circularization were not that efficient and you still had plenty o fuel left

-Try to pack your lift-off stage more, separate it into flatter (no pancake sh**) and not that long rocket

--> I think if you'd split one long fuel tank line into 2 you might reduce swing a lot, i think of something like this :

../\/\

.| | |

.| | |

.\.../

..| |

../..\

.| | |

You might get the idea

But nevertheless, pretty good idea with the srbs and nicely done putting that beast into orbit

[Jouni]

Contrary to what the others have said, the length of the rocket is not a problem. Tall rockets can behave well, if you build them right.

The real problem is with the mass distribution. You have a heavy payload at one end, and a heavy lower stage at the other end, and the two are connected by a narrow middle section. Designs like that are inherently wobbly, as the middle section flexes too much. Use only 3.75 m fuel tanks in the central stack, and connect the payload with a 3.75 m decoupler, and the rocket should work much better.

Turning faster is a good idea, as it can make the launches much more efficient. My usual stock ascent profile is to turn to 45 degrees at 10 km, and then lower the nose another 5 degrees every 5 km. The initial TWR could be a bit higher, but it's not as much a matter of efficiency than personal taste. Some people like rockets that launch quickly, while others like to see them ascend slowly from the launchpad.

[MarvinKitFox]

*watching video* You used actual SRBs as sepratrons? Nice.

++points for style, using SRBs as seperatrons.

But why make them so heavy? Drain the fuel until there is only about 2 seconds of burn left.

Your TWR is way too low though; you want it to be about 1.6ish at launch in stock aero and maintain at least that most of the way up (MOAR BOOSTERS!). Honestly it'd probably fly just as far if you just ditched one orange tank from the core and each booster. Gravity losses bite .

His TWR is quite respectable, after 10 seconds his velocity is 63m/s.

That is **at least** 1.7 TWR at launch.

I think what is fooling you is the deceptively slow temporal flow in his universe. He is running at 3-4 seconds per second, or even worse.

[Anquietas314]

The initial TWR could be a bit higher, but it's not as much a matter of efficiency than personal taste. Some people like rockets that launch quickly, while others like to see them ascend slowly from the launchpad.

No, this actually has a large effect on efficiency since if you have low TWR, you're suffering from quite large gravity losses all the way up. Of course the reverse situation is also very inefficient - crazy high TWR (and no throttling to compensate) will make you suffer drag losses. Install mechjeb and use its gravity/drag losses readout for a low thrust rocket and compare it to one at around 1.6 launch TWR to see for yourself. For an airless body the effect can be as much as 70% of your fuel wasted for an unrealistically optimal ascent profile: source. With an atmosphere it's even worse because you have the atmosphere limiting your acceleration, even ignoring drag losses.

His TWR is quite respectable, after 10 seconds his velocity is 63m/s.

That is **at least** 1.7 TWR at launch.

I think what is fooling you is the deceptively slow temporal flow in his universe. He is running at 3-4 seconds per second, or even worse.

No, it's not. Watch the mission timer in the video. Having flown many vessels with a TWR of 1.6 in the stock souposphere, he should be ascending much faster than that.

EDIT: Just checked the video. With mission timer at 10 seconds, velocity is 35.6m/s, so the TWR is approximately 1.3-1.4 which is too low for stock atmosphere (absolutely fine for FAR though).

Edited January 13, 2015 by armagheddonsgw

[Red Iron Crown]

No, this actually has a large effect on efficiency since if you have low TWR, you're suffering from quite large gravity losses all the way up. Of course the reverse situation is also very inefficient - crazy high TWR (and no throttling to compensate) will make you suffer drag losses. Install mechjeb and use its gravity/drag losses readout for a low thrust rocket and compare it to one at around 1.6 launch TWR to see for yourself. For an airless body the effect can be as much as 70% of your fuel wasted for an unrealistically optimal ascent profile: source. With an atmosphere it's even worse because you have the atmosphere limiting your acceleration, even ignoring drag losses.

You're measuring efficiency in delta-V to orbit. If you measure by payload fraction, lower TWR rockets come out ahead. More dV expended, but less fuel consumed (this results in lower cost, too).