Rockets falling over

[ishootdeadppl]

Version 1.0.2

Windows 7 64 bit

Once I get to a certain point (or building a rocket possibly too high, not sure which) my rockets just start to fall over. I even tested this with a capsule, goo pod, antenna, and parachute on a single Flea. The rocket fell over! I NEVER have problems at first, then things seem to get progressively worse. Possibly also related (might be me though) is that liquid rockets seem to go bonkers the more I play as well.

[sal_vager]

Hi ishootdeadppl, aerodynamics has been updated and is different to .90, you will need stabilising parts such as fins to start, and you will need to be careful with your turn until you get some gimballing engines.

As rockets burn fuel their tops get lighter so will tip over and want to fly backwards, use staging so there is a full fuel tank and engine at the top to help with stability.

Start your turn much sooner and tilt over gradually, large movements are more likely to cause the craft to lose balance.

[micha]

Is this on the launchpad, or while flying?

My rockets have been falling over (on the launchpad) a lot more in 1.0.2 as well (never got far enough in 1.0 and 1.0.1 to have noticed) - to the point that I use the poor-mans launcher until I get the launch clamp. (Stack-separator with a couple of trusses attached)

I can't recall this ever being an issue with 0.90.

[gogozerg]

As rockets burn fuel their tops get lighter so will tip over and want to fly backwards, use staging so there is a full fuel tank and engine at the top to help with stability.

Exactly.

ishootdeadppl, as a good example, just build a simple 1 stage rocket with a long 1.25m tank, a LV-T45 (heavy!), and some fins. When you reach 10-15 km altitude (depending on speed), the rocket becomes completely unflyable, because the tank is almost empty.

In the VAB, empty the fuel tank and take a look at the CoM... You'll immediately understand why it can't fly as expected.

[ishootdeadppl]

This problem is actually both on the launch pad as well as in flight. They're both similar in some ways, but could also have roots in different things.

I understand the concepts of shifting center of mass, but this still seems just crazy (though I'm totally new at this game). I've played around a bit with it. I did a three stage rocket. The first two stages were two 400 fuel tanks and one Reliant engine. The third stage was two 400 tanks and one swivel rocket. Capsule with stuff on top. I tried one launch and lost it about halfway through the second stage. The second time I lost it when the third stage fired. Could be bugs, could be operator error, and I honestly have no clue.

[sal_vager]

Can you please take some screenshots of the craft you use and pot them on imgur, it'd help a lot to see it.

You could also put the craft files on dropbox.

[Northlight]

I find that you must have fins on the first stage. For your small rocket, the small rocket fins should do fine. This will keep it stable in atmosphere.

If the 1st stage does not get you high enough, you might need fins on the 2nd stage too.

One thing is that you have 2 T400 fuel tanks on all three stages, this is propably not a good idea. Stages tend to get smaller the further up the rocket you go.

For eyeballing it, I make every stage about 2/3rds for the rocket. As in; 1st stage is 2/3 of everything. 2nd stage is 2/3 of what is left after 1st stage is dumped. And so on...

[gogozerg]

The first two stages were two 400 fuel tanks and one Reliant engine.

Don't forget the Reliant engine has no gimbal for thrust vectoring.

[ishootdeadppl]

Hopefully the screenshot works. I'm working on the .craft file now.

Edit: Grr, must be getting tired if I can't post a screenshot!

- - - Updated - - -

OK, for now I'm going to say thanks. Might play with it a bit more using the comments here, and see what happens. I'll update either late tonight or early tomorrow.

[Northlight]

For a quick suggestion about the stages, move 1 T400 from the 3rd stage to the 1st stage, that will give you ~620 more ÃŽâ€V.

Your Current rocket:

3rd stage: 2659 ÃŽâ€V

2nd stage: 466 ÃŽâ€V

1st stage: 312 ÃŽâ€V

Total: 3437 ÃŽâ€V

After move 1 T400 from 3rd to 1st:

3rd stage: 1691 ÃŽâ€V

2nd stage: 1300 ÃŽâ€V

1st stage: 1065 ÃŽâ€V

Total: 4056 ÃŽâ€V

[vonKerbin]

For a quick suggestion about the stages, move 1 T400 from the 3rd stage to the 1st stage, that will give you ~620 more ÃŽâ€V.

Your Current rocket:

3rd stage: 2659 ÃŽâ€V

2nd stage: 466 ÃŽâ€V

1st stage: 312 ÃŽâ€V

Total: 3437 ÃŽâ€V

After move 1 T400 from 3rd to 1st:

3rd stage: 1691 ÃŽâ€V

2nd stage: 1300 ÃŽâ€V

1st stage: 1065 ÃŽâ€V

Total: 4056 ÃŽâ€V

Weird question, but how do you know that? I mean, what did you use to find that out? I'm still pretty new, just had my first Mun landing last night. So just trying to read everything I can and this looks like good info to be able to use!

[Northlight]

There is a basic set of rocket equations that come in handy when playing KSP.

(But you get a lot of the answers for free if you use Kerbal Engineer Redux add-on.)

To find the ÃŽâ€V of a stage: ÃŽâ€V = ISP · g · ln(m0/m1)

ISP is the ISP of the engine in that stage. (You find that info in KSP under the [more info] for the engine)

Note: There is a differens ISP for sealevel and vacuum. This is because rocket engines become more efficient in vacuum.

g is the standard gravity. For Kerbin (and Earth) this is 9.80665 (or 9.81)

ln (LN) is the natural logaritm. You find it on more advanced calculators. On Windows calculator you find it if you switch view to scientific.

m0 is the weight of the stage before firing.

m1 is the weight of the stage after firing. (Basically the same as m0 minus the fuel spent)

So for your 2 T400s pr. stage in the rocket:

(I see now I made an error in my previous math)

S3-m0 = Pod(0.84) + parashute(0.1) + decoupler(0.05) + 2xT400(4.5) + T45(1.5) = 6.99 ≈ 7

Each T400 has 2t of fuel+ox, so S3-m1 = S3-m0 - 4 = 3

S3 ÃŽâ€V = 320 · 9.80665 · ln(7/3) ≈ 2659 ÃŽâ€V

S2-m0 = S3-m0 + decoupler(0.05) + 2xT400(4.5) + T30(1.25) = 12.8

Each T400 has 2t of fuel+ox, so S2-m1 = S2-m0 - 4 = 8.8

S2 ÃŽâ€V = 280 · 9.80665 · ln(12.8/8.8) ≈ 1028 ÃŽâ€V

S1-m0 = S2-m0 + decoupler(0.05) + 2xT400(4.5) + T30(1.25) = 18.6

Each T400 has 2t of fuel+ox, so S1-m1 = S1-m0 - 4 = 14.6

S1 ÃŽâ€V = 280 · 9.80665 · ln(18.6/14.6) ≈ 665 ÃŽâ€V

Total: 2659 + 1028 + 665 = 4352 ÃŽâ€V

If we move 1 T400 from the 3rd stage to the 1st stage:

S3-m0 = Pod(0.84) + parashute(0.1) + decoupler(0.05) + T400(2.25) + T45(1.5) = 4.74

Each T400 has 2t of fuel+ox, so S3-m1 = S3-m0 - 2 = 2.74

S3 ÃŽâ€V = 320 · 9.80665 · ln(4.74/2.74) ≈ 1720 ÃŽâ€V

S2-m0 = S3-m0 + decoupler(0.05) + 2xT400(4.5) + T30(1.25) = 10.54

Each T400 has 2t of fuel+ox, so S2-m1 = S2-m0 - 4 = 6.54

S2 ÃŽâ€V = 280 · 9.80665 · ln(10.54/6.54) ≈ 1310 ÃŽâ€V

S1-m0 = S2-m0 + decoupler(0.05) + 3xT400(6.75) + T30(1.25) = 18.59

Each T400 has 2t of fuel+ox, so S1-m1 = S1-m0 - 6 = 12.59

S1 ÃŽâ€V = 280 · 9.80665 · ln(18.59/12.59) ≈ 1070 ÃŽâ€V

Total: 1720 + 1310 + 1070 = 4100 ÃŽâ€V

So... Basically I screwed up my math, and your way was better.

I think I was thrown off by the 3 stages to orbit. 2 is normally enough.

[KerbMav]

Even with fins at the first stage my rockets tend to tilt during flight if they have no gimble and only marginal reaction wheels - e.g. a probe core with a tourist in a one-man-pod on a medium solid booster with fins.

Larger rockets - about equal to four to five orange tanks in height - not only wobble (I expect nothing else really) but also get "bent" by launch clamps, the marker on the navball is visibly not on 0.0 - some parts sink into the launchpad even (cannot remember right now if I already have the highest upgrade for it in my current save).

[vonKerbin]

Thank you for the math breakdown.

Science is fun! You ask a question, you get unexpected results, and you learn something!

My expected answer to the question was "Here is where (in KSP) you go to bring up that info / or / here is the mod you use" (which I don't use any mods so far).

However what I got was that you have to do it yourself ... but here's how!

And that is pretty darned cool.

[DocMoriarty]

My first rockets tilted too in career in 1.0. But after i started adding tailfins (not the small ones that got added in 1.0.1) it never happened again except yesterday when i wanted to launch new designed 45t tug with nukes consisting of lift-body LF fuel tanks. Added some more tailfins and it worked too (the lifting body parts are nice for airplanes but on top of a rocket can be really deadly in lower atmosphere). Wasted like 450k funds until i finally got one into orbit. So the new basic rule is: moar tailfins!

[ishootdeadppl]

Well, took a little longer to get back that I said, but... I think that there is still a launch pad bug, but there's both strategy and luck needed in design and launch as well.

- - - Updated - - -

And wow you guys were busy with this thread! I'm REALLY starting to see just how much there is to learn about this!

[PnDB]

Thank you for the math breakdown.

Science is fun! You ask a question, you get unexpected results, and you learn something!

My expected answer to the question was "Here is where (in KSP) you go to bring up that info / or / here is the mod you use" (which I don't use any mods so far).

However what I got was that you have to do it yourself ... but here's how!

And that is pretty darned cool.

Northlight provided a very nice treatment of how to use the rocket equation to compute deltaV. As you use it, you may find that repeatedly computing deltaV by hand gets a bit tedious once you've got the hang of doing it. If so, there are at least a few mods to the game that automatically compute deltaV while you are constructing your rocket in the VAB, and also show you how much deltaV you have remaining during flight. MechJeb and Kerbal Engineer Redux are the first that come to mind.

Also worth noting is that the rocket equation can get a bit complicated if you have a rocket which uses two or more different engines/boosters at the same time, especially if they burn out at different times during flight. It's not so much that it's harder to compute, it's just that you need to keep track of more things in your equations, and thus it becomes prone to error.

[Stephen Locksley]

If the rocket has a high amount of mass, and put up into the sky using VesselMover, then it will fall, probably killing everyone onboard the rocket.

[Vanamonde]

2 hours ago, Stephen Locksley said:

If the rocket has a high amount of mass, and put up into the sky using VesselMover, then it will fall, probably killing everyone onboard the rocket.

Thank you for attempting to help, but by now the initiator of this thread has either solved the problem or moved on.