Gimballed engines cannot perform roll maneuvers?

[flux.faraday]

I have a ship with several active gimballed engines (say 40 mainsails for example). I can pitch and yaw just fine, but I cannot roll. I don't *think* I'm doing anything wrong. Is this a missing feature? Or maybe there are real life safety concerns being modeled here? Or something else I've missed?

I'm not stuck, I can always add fins. I'd just rather not if there's a solution. The most recent posts that look something like this are a year old, so I'd guess it's not something that comes up a lot.

Thanks!

[Temstar]

Yes vectored thrust does not support roll yet. The devs did say once before the forum purge that eventually they plan to put it in.

[CosmicJoker42]

With only 2 axis of rotation, a vectored nozzle is not physically capable of rotation on its own. Several engines could counter-gimbal, producing rotation, but it's far simpler and easier to use RCS or SAS to rotate. Also, using thrust vectoring would mean giving up the ability to rotate while not burning.

[flux.faraday]

Awesome. Thanks for the reply!

[RoboRay]

WTF needs 40 Mainsails?

But no, all engines gimbal to the same orientation, so you have pitch and yaw control but not roll.

[Kromey]

WTF needs 40 Mainsails?

"The correct question to ask is, 'WTF doesn't need 40 Mainsails?'"

-- Jeb

[Fuzzy Dunlop]

I guess the devs didn't realise that people would be using so many gimballed engines, one on it's own wouldn't be able to roll anyway (cos its in the middle - can't generate any tourque)

[purpletarget]

I guess the devs didn't realise that people would be using so many gimballed engines, one on it's own wouldn't be able to roll anyway (cos its in the middle - can't generate any tourque)

Kinda odd....my inclination as soon as locking gimbals was introduced was to lock the inside engine and allow the gimbals on the outer boosters to handle the roll control issues once the s/c was too high for fins to be effective....oops.

Yes, Engine Gimbal controls still have a couple bugs with them....lacking roll authority was listed in bug 292, and a similar issue preventing them from working properly for skycranes and puller tugs was identified in 386.

I'll be interested to see when they get fixed...it'll be a subtle change that won't be easily shown off or sexy as shiny new parts. But just like RCS balancing, it will immensely improve gameplay when they do get around to fixing it.

[Temstar]

it will immensely improve gameplay when they do get around to fixing it.

This. Currently for large probe core'd rockets the only way to control roll during ascent is by either control surfaces or RCS. Roll control via gimballed engine will make an enormous difference in large rocket design.

[Jason Patterson]

it's far simpler and easier to use ... SAS to rotate

I'm not sure if it was your intent or not, but just to be clear, SAS cannot induce a roll, it can only stop one. The source of induced rolls is command module torque, which is defined by the variable rotPower in the part.cfg. You can add additional command modules to a vehicle to increase your handling as concerns rolling.

[numerobis]

Another fun thing to keep in mind is that gimbals above the center of mass do not compensate for their position, so the controls are reversed for them alone. Which usually leads to !!fun!! if you forget to lock those gimbals.

It is a simple matter of programming to fix both issues. However, the roll issue can be solved by accepting a rolling rocket: it actually helps you go straight, since any imbalance averages out as you roll. And the other issue can be fixed by putting your engines below the CoM.

[Edit: I misconstrued your question -- you want to roll, I assumed you wanted not to roll.]

Edited May 15, 2013 by numerobis

[ZingidyZongxxx]

With only 2 axis of rotation, a vectored nozzle is not physically capable of rotation on its own. Several engines could counter-gimbal, producing rotation, but it's far simpler and easier to use RCS or SAS to rotate. Also, using thrust vectoring would mean giving up the ability to rotate while not burning.

Well, several mainsails would give considerably more roll power than a few RCS thrusters, especially on a large rocket. Also, you wouldn't waste your RCS fuel. And you don't give up the ability to roll when not burning in any way, since RCS and command pods still work perfectly fine with thrusters off.

[flux.faraday]

Lots of fun replies.

RoboRay

WTF needs 40 Mainsails?

Kromey

"The correct question to ask is, 'WTF doesn't need 40 Mainsails?'"

It looks like I need around 120. My current goal is to single-launch a manned lander-and-return to Eve. So I've been working on a 30 stage by two asparagus lifter that starts with 60 mainsails. Sadly, this has only been able to lift about fourteen mainsails and orange cans of fuel into orbit around Kerbin which probably isn't enough to land on and return from Eve. Still working on this I have a 45 stage by two, but that runs into KSP/Unity/PhysX limits and dies on the pad.

Why do I need to roll? I've been using MechJeb for ascent. It handles throttle much better than I do.

numerobis

However, the roll issue can be solved by accepting a rolling rocket: it actually helps you go straight...

That's what I think as well, but MechJeb appears to get rather cranky if it can't roll the craft so that 90 is to starboard. Ah, I just realized I can probably adjust MechJeb to boost straight up a little higher and then take over the gravity turn myself. Well more to do!

[Temstar]

It looks like I need around 120. My current goal is to single-launch a manned lander-and-return to Eve

As with any mission, the best way to get more delta-V is to reduce weight from the top down. To start with if you're not leaving an Eve-Kerbin return vehicle in Low Eve Orbit you better go make one, it makes a huge difference in delta-V having to carry that return vehicle down to Eve's surface and then back up.

Then once you have a return vehicle waiting for you in orbit, there's always the option of not using a command pod at all and just get Kerbals to hang onto ladders back to Eve orbit. If done this way it's actually possible to land and return two Kerbal to Eve (using the 6.5km mountain top) with a vehicle under 30 tons.

Here for example is my Eve lander, capable of returning two man back to LEO if landed at 4.7km altitude or higher. The crew lands separately on a rover and is currently at the beach collection ocean "water" sample. When they're done they will drive to the rocket to get back up.

[capi3101]

Uh......question. There's no command pod on that lander, right? So how do you control it?

[zarakon]

There's probably a small probe core in there somewhere

[Temstar]

Uh......question. There's no command pod on that lander, right? So how do you control it?

Probe core

[numerobis]

I wonder why MechJeb is cranky -- I see issues when it fights the roll so hard with aero surfaces there's no control authority left to pitch and yaw with (in which case I usually can fly fine), but I haven't seen issues when not rolling.

MechJeb 2 is nicely set up for a semi-manual ascent. You can let it stage and throttle and tell you which way to go, while you drive the pitch/yaw/roll.