Flying straight.... in space...!

[tek_604]

I can successfully perform enough rendezvous/docking in Kerbin orbit, to make a "3 stage" probe to go to Moho (the only inner solar system target I have not yet reached).

The first and second stages are basically the same unit, and a probe docked on the very end.

My first try at this, was using the stock clamp-o-trons, and that ship was not flyable... As soon as the thrust was applied, the whole thing would basically go "head over heels". Today, I've tried using the "Common Berthing Mechanism" parts, thinking they would keep the ship together better. But, again, as soon as thrust was applied, the whole thing pulled to the right (going right, or head over heels, are all relative to your POV, but anyway...! ).

I understand it is to do with the centre of thrust. If the thrust vector is off centre, the worse it gets with longer ships? What I don't get is... how do I ensure that ships created in orbit through docking have a balanced centre of thrust?

I would assume that when building the parts in the VAB, with symmetry turned on, would build your capsules with the balanced centre of thrust. But this doesn't seem to be.

Maybe its as simple as not enough SAS? Advice is most welcome!

[llamatoes]

Pictures would be really froody, then we could really help.

[Akalaël]

It may be a problem with you center of mass off center... Or an other problem due to the wobling of your ship when it's on thrust...

I use Quantum struts to provide the spagetti syndrome on my ships.

Can you provide a pic ? Hard to help you without see your creation

And do you use SAS or ASAS ?

[SecondGuessing]

It's almost certainly a CoM issue, but it could also just be the SAS unit getting twisted out of linearity by the way SAS units fights and then the whole ship correcting to that direction, rinse and repeat.

[Alistone]

I always contemplated moving fuel from tanks on one side or the other of the CoM in order to balance it. But balancing thrust AND making strong connections to withstand the thrust requires a fair bit of planning. Sort of has to be planned as a mothership instead of just a space station.

[austin4050]

Might sound dumb, but do you have RCS on both ships?

[Supernovy]

Are you using ASAS? Is the centre of thrust ahead of the centre of mass? This configuration causes engine gimbals to turn the ship opposite to what they should.

[Alistone]

Are you using ASAS? Is the centre of thrust ahead of the centre of mass? This configuration causes engine gimbals to turn the ship opposite to what they should.

hmm..... so that's why my "tug" turns so slowly; the atomic engine vectored thrust is counteracting the command modules' turning force. I'll try locking their gimbals to see if it works better.

[tek_604]

Here is my ship, assembled in orbit... Be gentle

RCS was used during the assemble in orbit stage, but, was ejected, don't need to take the extra mass. The only SAS module I have is on the probe.

I am thinking of a large SAS module on the injection stages, that might sort it out?

Also, I think I know what made this worse. During my injection burn, I forgot to turn off cross-feed on the docking module between the injection stages, so the first stage drew fuel from the 2nd stage. If I remember to turn this off, it should also help?

[Evil Acrylic]

As far as i can see your rocket looks fairly inline, I know those engines have alot of Vectoring power 3Dec if memory serves.

Only 2 things i can suggest is 1 Which Probe is in control as in (Control from here) if the docked probe is the master the ship will want to right its self as you are flying backwards make sure the probe that has the currently thrusting engines is in control.

Secondly using MechJebs Kill Rot negates this but i have always treated a SAS Module as a gyro to keep things stable but has little effect on your Engine vectors and Control Planes where as a ASAS does this and with the engines you are currently using it should have no issues even if it was off balance.

[Morzak]

I think the Kethane Mapper and MapSat aren't the same mass, so that would give you some imbalance, but with that overall mass I highly doubt that this would be a big issue with the transfer stage. Like Evil Acrylic mentioned, make sure that you select "control from here" on the probe of the transfer vehicle, if you use Mechjebs Smart A.S.S. it will turn you're ship 180 degrees otherwise.

[Akalaël]

I gess you center of mass is align whith the center of thrust.

The problem may come because you probe is not attatch enought : when you thrust, your probe goes right, and MechJeb (who is a terrible psychotic tool) compensate to the left, and your bodie's probe then goes left, and MJ compensates, and so on...

Just add 3 or 4 struts between your probe and the propulsion core, that may solve the problem.

And dont mix SAS and ASAS, as far as I can see you have an ASAS module whitch is a good thing for a vessel of this size.

EDIT : I dont have the MapSat nor the Kethane mod... Are the antennas heavy ? If yes, then I remove my first sentence : your COM is not aligned with you COT.

Edited May 21, 2013 by Akalaël

[tek_604]

I selected the first injection stage for "control from here". I was wondering if the antennas were causing the imbalance. Checking the parts:

KE-S210 Compact Survey Unit: mass = 0.075
ISA MapSat: mass = 0.01

This mass difference seems quite small when we're talking the whole thing being close to 50t.

I think I'm liking Akalaël's suggestion, in that the probe is slightly off centre compared to thrust, and MechJeb is doing a bad job of compensating for this, and made worse because of wobble. I say this because at full thrust (480kn), the problem was a lot worse than at 25% throttle. Also, if I have the same probe, attached to an injection stage through a normal decoupler (I use the same probe to go to the other planets, using the boosters 3rd stage as injection stage), then this does not veer off course. Only when using docking ports.

Struts was the suggestion, to remove the off centre wobble. Problem is, the injection stages are already pushing my lifting capability, and I'm not sure I can lift the probe and 2nd (middle) injection stage in one go. I seem to have picked up that Quantum Struts might be the solution, to add struts in orbit. Am I right?

Other than that, I guess I've got some experiments to do! Thanks for the help guys, I'll post my result when I've done them.

[J.Random]

Am I seeing a bunch of stacked toroidal fuel tanks there (gold-ish rings)?

Edited May 21, 2013 by J.Random

[Akalaël]

Yep, Quantum Strut let you add some struts in space via EVA, but never tried it myself...

[kodack]

It's a simple thing to check but make sure all of your thrusters are turned on. If even a single one is off it will cause you to flip out of control. You can also make sure that the thrusters are straight and haven't been tilted when you built it. As someone else suggested double check your COG in the building area and make sure it's dead center. One more thing, if you are using both sets of thrusters....don't. A rocket is most stable with thrusters at the back, pushing.

[tek_604]

Am I seeing a bunch of stacked toroidal fuel tanks there (gold-ish rings)?

Yes, I like the look of them

[J.Random]

I suggest to either strut the hell out of them or change them to normal fuel tank. You have around five tons of fuel and engines holding on that tower of toroids and single docking port.

Another example of similar design decision and what it can lead to:

[tek_604]

Thanks Random! I had a feeling you were going to say something like this (didn't realise they would bend, but thought the stacking in the VAB might be off centre), so I redesigned my probe to have a single 90 ltr. tank. I think I'll try that as my first solution, and if that doesn't fix it, then I'll go for Quantum Struts, as I said before.

Another thought I had was... if you use those gold tanks, you can stack them around a beam, I wonder if that would stop the bending?

[Chruschtschow]

Another example of similar design decision and what it can lead to:

You built a wibbly wobbly space adult toy?!? With KSP? You're my idol!

And to say something productive: You want to go to Moho and yet I don't see any nuclear rocket engines. If you want to make your life easier, use them. They are highly efficient in vacuum.

And about the topic: I had some problems with a rather heavy station and slightly unsymmetrical solar panels and sensors. It can have some massive effect, even when the masses are small compared to the whole ship. Especially if you add that oscillating gold tank thingy. A beam might help. Not using them at all might help more, since they have a very bad fuel / mass ratio.

Edited May 22, 2013 by Chruschtschow

[tek_604]

You built a wibbly wobbly space adult toy?!? With KSP? You're my idol!

I didn't want to say anything, but, now its been said, I was thinking the same thing

And to say something productive: You want to go to Moho and yet I don't see any nuclear rocket engines. If you want to make your life easier, use them. They are highly efficient in vacuum.

And about the topic: I had some problems with a rather heavy station and slightly unsymmetrical solar panels and sensors. It can have some massive effect, even when the masses are small compared to the whole ship. Especially if you add that oscillating gold tank thingy. A beam might help. Not using them at all might help more, since they have a very bad fuel / mass ratio.

Right now, I don't want to use the nuclear engines. Call it role play if you will, but, I think they're a little bit cheaty. The oscillating tanks I have now solved with the new Explorer Na series of probes; not tested yet, but designed. And when I looked at the fuel/mass ratio, I thought they were lighter than the 90 ltr. tank for the same fuel load? God I'm bad at maths if my thinking was wrong

[Chruschtschow]

And when I looked at the fuel/mass ratio, I thought they were lighter than the 90 ltr. tank for the same fuel load?

The ratios for (mass of fuel / mass of tank) are:

Round-8 Toroidal 4.44

Oscar-B 4.24

FL-T200 8

FL-T400 8

FL-T800 8

Rockomax X200-8 8

Rockomax X200-16 8

Rockomax X200-32 8

Rockomax Jumbo 64 8

Actually that's the first time that I made the math myself. I only read about it earlier. It's worse than I thought for the two small tanks.

[Francesco]

The ratios for (mass of fuel / mass of tank) are:

Round-8 Toroidal 4.44

Oscar-B 4.24

FL-T200 8

FL-T400 8

FL-T800 8

Rockomax X200-8 8

Rockomax X200-16 8

Rockomax X200-32 8

Rockomax Jumbo 64 8

Actually that's the first time that I made the math myself. I only read about it earlier. It's worse than I thought for the two small tanks.

how did you get those results? they're totally wrong.

[stupid_chris]

Right now, I don't want to use the nuclear engines. Call it role play if you will, but, I think they're a little bit cheaty.

Why do you call nuclear engines cheaty? They exist in real life and the efficiency is proportional. You can search on Wikipedia for the NERVA engines, some plans to go to Mars were made using NERVA engines and the Saturn V rocket. But still, they're not boosted or hacked over their real life equivalent as far as I know!

EDIT: Just reread my thing and I sound a bit rude, sorry. But I really want to know why you think they are cheaty

[HoY]

Might I suggest a change in your upper stage as well? Instead of having two engines and two tanks separated by a decoupler, use one LV909 engine on an FLT400 tank, and have two FLT200 tanks on radial decouplers outside of it with the fuel piped into it. This way you save the 0.5 tonnes from the extra engine, still get the same overall fuel and stage time but a more stable platform and more deltaV out of them