ANSWERED -Space Tug issues - need serious help

[ValleyTwo]

Have had some problems last few days.  I put up a few station cores and now need them attached.  I tried my old Orbital Tug design but it went crazy - over-correcting every time right at the launch and up into orbit - it almost slammed into one of the cores.  I do use MechJeb2 so let me say I always click on the top Senior Docking Port to make sure it knows which way is which.

I have tried a couple of designs (replacing the base fuel tanks with jet fuel tanks) and even used HyperEdit to put one into orbit, turning off MechJeb2, and telling SAS to do the work.  It also over-corrected and then counter-over-corrected.  It looked like that last scene from the Last Star Fighter.  And nothing I have tried, no design, has made anything any better.  Some get into orbit looking like they may work but the second in stable orbit they act just as nutty.  They don't even stand still when they have nothing to do - they tumble.

This is my latest design:

I am was starting to think it was the mass of the side mounted nuclear engines were making it hard for MechJeb2/SAS to control the craft but the designs before this one used lighter non-nuclear engines, on the sides, and it had the same problems in orbit as all the rest.  Sluggish or quick to spin they all over-react like drunk bumble bees.  

I have decided it must be that something has changed in the game itself to effect how I design my tugs.  In other words, there is something I am not grasping.

What I would like is:

1. An example of a working Orbital Tug that works in KSP 1.1.2.

 2. Could somebody explain to me what has changed and why I can't seem to get any of my Tug designs to work.

My Basic Booster I made to place station cores into space worked perfectly and I did use testing rockets before this to see how such things, as re-entry, may have changed.  

I am not bothering to download any of the designs because at this point I have made over 15 different Orbital Tugs and boosters.  I feel the problem is my knowledge of the game is somehow outdated or flawed now.

 

Edited May 27, 2016 by ValleyTwo
Answered.

[bewing]

I only use SAS, so I can't speak to MechJeb. But SAS overuses the RCS system. If SAS can only use one reaction wheel to steer, then things are much more stable. You can temporarily disable rotation control for all of your RCS ports with an action group. That way, when the tug is not attached to anything and has low mass, you disable the RCS. Then when you attach to something big, you can turn it back on again to wrestle your component around.

[ValleyTwo]

6 minutes ago, bewing said:

I only use SAS, so I can't speak to MechJeb. But SAS overuses the RCS system. If SAS can only use one reaction wheel to steer, then things are much more stable. You can temporarily disable rotation control for all of your RCS ports with an action group. That way, when the tug is not attached to anything and has low mass, you disable the RCS. Then when you attach to something big, you can turn it back on again to wrestle your component around.

It never touched the RCS - they were not used during any of the flights.

[Mjarf]

Why are the engines placed so far up? Do you attach the station module on top of the "tug"? I use a very different design to assemble and re-arrange station modules in orbit. 

[ValleyTwo]

8 minutes ago, Mjarf said:

Why are the engines placed so far up? Do you attach the station module on top of the "tug"? I use a very different design to assemble and re-arrange station modules in orbit. 

I was trying to balance the engines so the thrust and center of mass lined up together.  And most of the time I use the under-carriage docking port (in the past anyway). 

[Mjarf]

Center of thrust should always be behind center of mass to get optimal stability this is because you're constantly fighting gravity. If you need to use the lower docking port occasionally, just place the engines further out from the hull so they don't damage whatever is docked underneath.  

My base design for a space tug for LKO is an orange Rockomax tank with a Poodle engine, gives more than enough delta-v if launched with a strong lifter, a probe core, I like the Okto 2, nice little core, plenty battery cap, 4 or 6 small ox stat solar panels, one of the largest reaction wheels, and strategically placed Vernor thrusters for translation, they are stronger than traditional RCS thrusters and use the same fuel as the main engine to keep things simple, the reaction wheel takes care of rotation.

Edited May 26, 2016 by Mjarf

[Tarheel1999]

I can't remember if the nukes are gimballed but if so turn the gimbals off. I have a somewhat similar design and had to turn off the gimbals on the engines to get control while thrusting. 

[ValleyTwo]

Just now, Tarheel1999 said:

I can't remember if the nukes are gimballed but if so turn the gimbals off. I have a somewhat similar design and had to turn off the gimbals on the engines to get control while thrusting. 

Well, that is good to know, but engines are not the problem.  It tumbles and over-reacts.  Engines are off at the time.  There is no problem when the engines are on.  

 

15 minutes ago, Mjarf said:

Center of thrust should always be behind center of mass to get optimal stability this is because you're constantly fighting gravity. If you need to use the lower docking port occasionally, just place the engines further out from the hull so they don't damage whatever is docked underneath.  

I think you are thinking about aircraft when you say I need to put thrust behind center of mass.  And I have always used the lower docking port and the engines have never damaged anything before this design.  And they have not done so with this design because it has not been used yet.  I am not worried about cargo right now.  I am trying to figure out why this design, which should work, is not working.

 

Does anybody HAVE a working Space Tug they can SHOW me?  That would be nice.

 

[Mjarf]

Placing center of thrust behind center of mass is rocket design chapter one actually. 

Here's my Heimdal class orbital utility craft, the probe core is clipped into the tank FYI:

Edited May 26, 2016 by Mjarf
added some info

[ValleyTwo]

That is a big tug but as I am going to be moving large station cores and not half-empty fuel tanks I guess I should go big and see how ti works.  Thanks!

[Mjarf]

This one should handle modules up to 50 tons or so, you can always scale it up to move bigger stuff.

Here's a launch I assembled just now. I had to re-construct my Heimdal craft as it was on my old science save that got deleted when I updated to 1.1, so I should test it to see if it still works as well as in 1.05.

[Tarheel1999]

31 minutes ago, ValleyTwo said:

Well, that is good to know, but engines are not the problem.  It tumbles and over-reacts.  Engines are off at the time.  There is no problem when the engines are on.  

 

If you are not using engines, not using RCS, and not running out of power, then the answer is simple. You need more reaction wheels. That is the only thing that will provide more control authority under these circumstances. Or you could reduce the weight by getting rid of at least half of the mono in that tank. 

Edited May 26, 2016 by Tarheel1999

[ValleyTwo]

16 minutes ago, Tarheel1999 said:

If you are not using engines, not using RCS, and not running out of power, then the answer is simple. You need more reaction wheels. That is the only thing that will provide more control authority under these circumstances. Or you could reduce the weight by getting rid of at least half of the mono in that tank. 

I can try it.  Thing is I tried it before.  As of now I am on design number 17 (or 18)  and still it tumbles.  

[Mjarf]

This is why I like to add several sets of Vernor thrusters on my lifters, keeps the rocket stable with brute force. It uses a bit more fuel, but allows me to launch some real hodgepodges into space.

 

[Tarheel1999]

It tumbles on launch or in orbit? As those are different problems with different solutions. 

[ValleyTwo]

It tumbles in orbit, as explained in the first post.  It isn't a problem with the lifters or center of mass or anything like that.  I know how to make rockets and planes and understand all that - at least I thought I did - but this is in vacuum in orbit, no air and micro gravity.  This is not about getting it off the planet (as a few of the launches had no problems at all)  It has a probe and TWO SAS and a MechJeb2 and when I tell it to do something, when it goes to align itself for a burn it tumbles.  Now with two SAS is spins VERY fast and does recover but it takes time and if it did that when approaching a station it would slam into it before it corrected itself.  I assumed it was not MechJeb2 because that was working earlier (thought MJ does have a habit of breaking on its own) and I still think it may be the mass of the engines causing centripetal acceleration.  But maybe I will wait till MechJeb2 is updated or the game is.  Sad, I was kind of hoping to play the game again.  

[Kerbart]

1 hour ago, Tarheel1999 said:

If you are not using engines, not using RCS, and not running out of power, then the answer is simple. You need more reaction wheels. That is the only thing that will provide more control authority under these circumstances. Or you could reduce the weight by getting rid of at least half of the mono in that tank. 

Not necessarily. There is such a thing as too much torque. Especially when you have that wobbly construct in between. Remember, structure wise the quad-coupler only couples one set of LF tanks. The rest attaches at the top, or the bottom, but not on both ends.

So you have this structure:

[RC-L01] [Wobble Wobble] [Large Reaction Wheel]

Whatever the large reaction wheel is doing, it might not be what the SAS unit thinks it's doing because there's a bit of spaghetti noodleness in between. You could try to attach the reaction wheel directly to the RC control unit, and see if that improves things.

EDIT: also, even if that helps, consider connecting those two quadcouplers with a set of four struts (preferably corner to corner, if the editor lets you, otherwise from the middle of the straight edge down to the corresponding center of the straight edge of the other coupler) to firm that part of the construction up as much as possible.

Edited May 26, 2016 by Kerbart

[Tarheel1999]

If the problem is that the tug is slow to turn or slow to recover from a turn, you need to reduce your mass or increase your control authority. Your tug is 38 tons which seems excessive for Kerbin orbit. Try reducing it to a 32 16 tank, two radial mono tanks, and two nukes, and the other bits and bobs. That should be enough for LKO.  A big heavy ship is going to turn slowly. Also I prefer the 4-ways for RCS.

@Kerbart

Too much torque can be a problem but that is not what OP is describing.

Edited May 26, 2016 by Tarheel1999

[ValleyTwo]

32 or 16 Tanks - hmmm..those have liquid oxygen in them, don't they?  I am using nuclear engines.  Unless you are suggesting I empty the oxidizer out?

[Kerbart]

@Tarheel1999

I bring 25T vehicles to orbit on a regular basis and have no problems turning them in the right direction with a 1.25m stabilizer, which offers half the torque of the 2.5m one. And that's usually with a jumbo tank attached, with at least 1/4 fuel in it, and a skipper or mainsail attached to it.

It doesn't exactly have the handling of a F1 racing car, but "tumbling out of control" isn't occurring either. You might be right of course, but I'm very skeptical that a second set of reaction wheels will fix the issue.

[bewing]

2 hours ago, ValleyTwo said:

Does anybody HAVE a working Space Tug they can SHOW me?  That would be nice.

My space tugs are all 1.25m klaw-based things. They work fine, but a picture of one wouldn't help you at all.

[ValleyTwo]

Kerbart has suggested I share the latest design to let people look it over and maybe see what is wrong.  This is the latest design - in fact untested - and I think version 22?  There were more, but this is the nuclear engine version.  It is MechJeb2 free, so if it does work then I have to assume I have wasted two days trying to fly with a faulty MechJeb2.  Right now, I don't care if it is underpowered - I just want to know if it works in orbit.  Sorry, it is so new I don't even have a lifter for it.

It will look tiny compared to the design in the first post as I was trying to strip it of mass.

Dropbox:

https://www.dropbox.com/s/y93977bxrewlmnw/Orbital%20Tug%20Ver_%2022.craft?dl=0

 

(You should be able to download a copy - in the right hand corner is a 'MORE' button which gives you access to 'Download')

 

Also, I would like to apologize if I came across as rude.  I know you are all trying to help me and I have become a tad frustrated with the tug.  I have designed three of the most compact and simplest station cores I think I have ever designed and they even look kind of nice.  And to NOT be able to put them together is getting me down.  Also I was always proud of my old tug designs in my old saves and to find I can't make one now to save my life is also frustrating.  

Edited May 26, 2016 by ValleyTwo

[Tarheel1999]

16s with the oxidizer removed. Should be enough but you might need the 32s depending on what you are doing. After thinking about this a bit, it occurred to me that you were using the nuclear engines to control fore/aft for docking. If this is the case you should be using RCS and then you won't have to turn around to control your speed. You would need to use the 4-way RCS or pit linears fore and aft to do this though.  I disable the yaw/pitch/roll on the RCS and docking unbalanced loads are must easier than it was in 1.05. 

Edited May 27, 2016 by Tarheel1999

[ValleyTwo]

I only use the RCS for docking.  Or MechJeb2 does.  It only uses the Nuclear Engines for travel.

[bewing]

@ValleyTwo: Here, I took your ver 22 apart and put it back together my way. It makes it to orbit OK (moar boosters would save a little fuel). You don't really need 4 RTGs, do you? You aren't planning on taking this out to Eeloo, are you?

After I put it back together, I am not having any trouble maneuvering it. Most of your parts were hand placed, and some of the symmetry was not good, I think. And there was a bit of clipping with the RTGs -- not sure if that caused any problems.

More notes: sometime before you decouple the top, you should manually transfer the fuel in it to the little tank above the docking port -- I think it's nice if tugs have a little oxidizer to share. I think there is a bug in the hardpoints -- I have them set to crossfeed, but no matter how I install them they do not crossfeed. So you will have to manually transfer your fuel with them, until the bug is fixed -- or put a couple of your fuel ducts back on. (But the fuel ducts have a bug with very high drag right now, and you should not need them with this design, darn it!) Of course there are the bigger hardpoints that you could try, too.

I'm not sure if having the nukes hang down as far as I do will cause you problems -- but if you want to use this design at all, I expect you will modify it to make it work more like what you want. Some of the nosecones and things that I put on are just for pretty anyway. So it should be easy to move things around.

All SAS modes besides "stability" currently have a lot of high frequency noise. So if you have RCS and SAS active, and SAS set to "target" or something, and the RCS thrusters have rotation turned on, then they will constantly make thousands of tiny thrusts. So I set the RCS thrusters "rotation" to off.

http://www.virtualrealitytoursllc.com/pix/Orbital_tug.craft

 

Edited May 27, 2016 by bewing