Is a wobble-free vessel connected by docking ports possible?

[Kasuha]

You do know the Saturn V was joined with nothing more than essentially flat rings? The ISS too. Rigid connections don't need to be deep, they just need to be properly designed. (Or in the case of KSP, properly simulated.)

Saturn V connections are designed to withstand thrust of its main engines but are not really strong to forces perpendicular to it. If it didn't follow gravity turn well enough, it could break in half. ISS connections are not built to withstand much forces at all. Of course it gets orbit corrections time to time but these are applied using safe thrust which is very small. KSP joints are specific and are affected by game technology but overall they are in fact much stronger than their real world counterparts.

[Commissar]

personally, i use the 3.75m docking clamps from KW on my 150t tankers, for docking to space stations. nice thing about those ports, they can dock with clampotron sr's, and not just others of the same type. most everything else use standard ports, although i stick disposable sr. ports on my space station modules, so my tugs can tow them around without issue.

I think i did do the multi-port docking before, but I don't think i use that method for anything.

one time, i did fly 2 of my tankers to duna, docked together. they are double ended. the big problems I encountered had nothing to do with wobble, rather with the extremely sluggish turning, and the low TWR given by 16 lv-n's pushing ~300t of LFO and monoprop. i have another craft in orbit now to compensate in the future. pretty much just a docking port, probe core, and LFO tank witha 2.5m nerva.

[Jouni]

A ship flexing during reorientation is a different issue from wobbling under forward thrust. Wobbling under thrust is a result of the ship as a whole being dynamically unstable due to imbalances. Flexing during reorientation is the result of insufficient strength for the lateral forces and moments applied. A ship that flexes when turning may be wobble-free under thrust, although if it's going to wobble anyway, the flexibility of its structure may amplify the wobble in the same way that docking ports can. But both issues have the same root cause: design and/or construction deficiencies.

I've never seen a ship wobble under thrust in vacuum without physics warp or TWR over 2, so I could not realize it would be an issue to anyone. On the other hand, I have seen similar wobbling during violent maneuvers as physics warp causes under thrust.

My typical interplanetary ship design is like this:

The center of the mass is well within the engine block, while the main ship itself is relatively long and lightweight. The lander is attached to the nose of the ship due to balance issues, and as all reasonable landers designed for anything bigger than Minmus, it tends to be quite heavy.

During maneuvers, the lander usually trails a bit behind, especially if its fuel tanks are full. Usually this is not a problem, unless the maneuver is very violent, SAS is on, and the reaction wheels in the lander are active. In that case, things start going wrong, because SAS is doing the wrong thing. Instead of using the reaction wheels to keep the lander stable, it tries to turn the rest of the ship with them, which is pretty much opposite to keeping the lander stable. The ship starts to wobble, and SAS makes the wobbling grow worse and worse.

This was deliberately achieved with thrust during physics warp, but similar things happen, to a lesser extent, if you let SAS do its worst.

[DerekL1963]

Saturn V connections are designed to withstand thrust of its main engines but are not really strong to forces perpendicular to it. If it didn't follow gravity turn well enough, it could break in half.

Since we aren't talking about perpendicular forces... I'm not exactly sure what you're talking about. Turns lead to bending moment, not 'perpendicular forces'. That being said, there's still considerable stress on the joints and they are little more than flat disks. And the Saturn V doesn't wobble or bend.

KSP joints are specific and are affected by game technology but overall they are in fact much stronger than their real world counterparts.

They aren't anywhere near as strong as their real world counterparts - as evidence by the complete lack of real world rockets that bend and wobble on ascent.

[Kasuha]

as evidence by the complete lack of real world rockets that bend and wobble on ascent.

Main reason being, real rockets start exploding before any noticeable wobbling can be observed. And because real world engineers don't enjoy that happening they make sure the rocket is steered in a way that does not allow any wobbling to occur. Which is the point where things stop being a game and start being real rocket science.

[SSSPutnik]

KSP joints are specific and are affected by game technology but overall they are in fact much stronger than their real world counterparts.

Yes, yes they are: (Try doing that with a real ship)

[SSSPutnik]

Since we aren't talking about perpendicular forces... I'm not exactly sure what you're talking about. Turns lead to bending moment, not 'perpendicular forces'. That being said, there's still considerable stress on the joints and they are little more than flat disks. And the Saturn V doesn't wobble or bend.

The Saturn V was always balanced on thrust vs wind resistance etc. (IE net force was through the stack, not sideways). I guarantee you, if it had been subject to even small (in comparison) sideways forces, it would have snapped like a twig.

And the Saturn V *did* wobble. Not a great deal, but it did.

"Apollo 8 was the first manned flight of the Saturn V rocket. Bill and his crewmates Frank Borman and Jim Lovell were positioned at the very top of the mighty rocket. The biggest surprise to Bill was the violence of the lateral shaking as the F-1 engines gimballed back and forth to keep the giant rocket pointed in the right direction."

Also, it can clearly be seen in this video that it is not a rigid body. (Technically a perfectly rigid body is physically impossible).

Also:

"Preserving the mechanical integrity of the launch vehicle was one of its most important tasks. Active guidance was therefore suspended during the boost phase of the S-IC stage. The reason for that was, that during that phase the launch vehicle was travelling thought the dense layers of the atmosphere and was subjected to wind sheer and large aerodynamic forces. Additional lateral forces, which are applied as a result of the swivel motions of the large F-1 thrust engines, to make course corrections, might jeopardize the vehicle integrity. Therefore the launch vehicle went through a predetermined smooth flight path, controlled by a fixed program in the onboard computer memory. Deviatons from the desired flight path caused by wind sheer, were however sensed, measured and stored in the onboard computer for later retrieval. After ignition of the S-II stage in the thin upper atmosphere, the launch vehicle was actively guided and flight path deviations from the early boost phase could be compensated for."

http://home.kpn.nl/panhu001/Saturn%20V%5CSaturn%20V%20info%5CS-IU500%20info%5CS-IU500%20info%20home.html

Edited February 20, 2014 by SSSPutnik

[Jouni]

They aren't anywhere near as strong as their real world counterparts - as evidence by the complete lack of real world rockets that bend and wobble on ascent.

When it comes to joints, 'rigid' pretty much means 'weak and heavy', while 'flexible' means 'light and strong'. The more rigid you want a joint to be, the more mass you need to achieve the same strength. Of course, when you have many heavy rigid joints, you need to make them even more heavy and rigid to withstand the weight of the other joints. That's why tall buildings are made lightweight and flexible, especially if they are built in an area where strong winds or earthquakes can be expected.

[Geschosskopf]

I've never seen a ship wobble under thrust in vacuum without physics warp or TWR over 2, so I could not realize it would be an issue to anyone.

I'm totally with you on this. Wobbles under thrust are completely the result of improper design and/or construction methods. Those who build properly don't have them but a sizeable fraction of the KSP community seemingly doesn't know how to avoid this problem. You see wobbles under thrust all over YouTube and hear all the complaints about wobbly docking ports, and thus the huge "snake oil" industry promoting quantum struts, multi-port docking, using pulling instead of pushing engines, etc. All of which is just putting MOAR BandAids over the real problem, which is improper design and/or construction.

This was deliberately achieved with thrust during physics warp, but similar things happen, to a lesser extent, if you let SAS do its worst.

Well, you've got 2.5m parts on both sides of a 1.25m docking port. That's not a good thing to do because it creates a stress concentration in an otherwise 2.5m cylinder. Besides, a 1.25m port is scaled to stand torque and weight from 1.25m parts, so is overloaded with 2.5m parts anyway. It would be better structurally if you could turn the lander around and have it mate wit the main ship using a 2.5m port. Or if it's not possible to mount the bottom engines radially, to switch the 3-man capsule for a HitchHiker so you can put a 2.5m port on the nose. Not only would this eliminate the stress concentration and port overload you've got now, but it would save about 1.5 tons in lander weight and let you carry 4 instead of 3 Kerbals.

[Jouni]

Well, you've got 2.5m parts on both sides of a 1.25m docking port. That's not a good thing to do because it creates a stress concentration in an otherwise 2.5m cylinder. Besides, a 1.25m port is scaled to stand torque and weight from 1.25m parts, so is overloaded with 2.5m parts anyway. It would be better structurally if you could turn the lander around and have it mate wit the main ship using a 2.5m port. Or if it's not possible to mount the bottom engines radially, to switch the 3-man capsule for a HitchHiker so you can put a 2.5m port on the nose. Not only would this eliminate the stress concentration and port overload you've got now, but it would save about 1.5 tons in lander weight and let you carry 4 instead of 3 Kerbals.

Those ideas would conflict with my other design goals. I like to pretend that kerbals can access all parts of the ship without EVA, so the docking port cannot be attached to a fuel tank. Then I like to have a proper command pod at the nose of the ship, so I can't have a bigger docking port there without ugly structural things. Besides I like using the big command pod in my landers, as they usually don't need separate batteries, RCS fuel tanks, and reaction wheels with it. All those extra parts on the surface of the lander would make RCS thruster placement harder, as I like to have clean lines of thrust for them, even though that's not technically necessary.

It could be possible to replace the middle thruster module with a lander, as that's the only place where I can fit a 2.5 m docking port on the main axis. Then the ship would need more engines and more fuel somewhere else, and I still haven't found a good solution for that.

[SSSPutnik]

...and thus the huge "snake oil" industry promoting quantum struts, multi-port docking, using pulling instead of pushing engines, etc.

Sorry, I mildly object to this and agree with Jouni to a degree.

Aesthetic, practical or deltaV issues may stop you from having a perfectly balanced ship. (ie when dropping a probe, or one RCS tank instead of two etc).

And yes I can build wobble free ships, like this one for instance.. 770t, no wobble.

Edited February 20, 2014 by SSSPutnik

[Geschosskopf]

Those ideas would conflict with my other design goals....

Sorry, I mildly object to this and agree with Jouni to a degree.

Aesthetic, practical or deltaV issues may stop you from having a perfectly balanced ship. (ie when dropping a probe, or one RCS tank instead of two etc).

Well, then you have to accept the consequences of such decisions. The causes of wobble, either while thrusting or rotating, are pretty easy to understand. Once you understand them, you can build to avoid them and thus have no need of "snake oil". The docking ports themselves do nothing. They only react to what's going on with the rest of the ship, so should be viewed as diagnostic instruments of overall ship construction quality, not as malevolent wobble-causing robots. If you have wobble, it's because you built the wobble into the ship. A little wobble might be acceptable but if it reaches the point where the ship breaks up or is simply uncontrollable, then you did something wrong either in construction (symmetry accidents, part clipping, etc.) or design (poor structural design and/or weight/balance). You should go back and fix these problems, which might entail redesigning the whole ship. Using "snake oil" to make a bad design work anyway is just shoddy engineering. But in any case, don't blame it on the docking ports. It's your fault, not theirs.

This thread is called "Is a Wobble-Free Vessel Connected by Docking Ports Possible?" The OP and all the "snake oil peddlers" who responded seem to assume the answer to this question is "no". I was merely pointing out that the answer is actually "yes". The problem is, saying this and explaining that wobble is caused by the player, not the docking ports, is hard to do without a lot of folks taking it as an insult. But it is what it is.

[ekoo]

My typical interplanetary ship design is like this:

http://www.cs.helsinki.fi/u/jltsiren/stuff/laythe_launch.jpg

I really like this design. What's your procedure to get all that into orbit? I'm still trying to learn heavy lift vehicles, and I'm having more Rapid Unplanned Disassemblies than I am successes.