Stability/control issues in very low tech craft

[UmbralRaptor]

Or, why does this flip wildly after a bit if I tell it to go straight up? (And/or try to pitchover) Especially as nothing is sticking out near the front, and there are fins in back to move the CoD.

 

For additional context: while the examples are in sandbox, it's for a Munar flyby in a career with greatly limited resources (so an LV-T45 is iffy, and an LV-909 or steerable winglets are right out). What I'd like is to get something so seemingly simple to be stable without the fins (freeing up part count for additional FL-T100 tanks)

Alternate configurations (eg: preventing propellant flow from the uppermost 2-5 tanks, placing tanks around the sides instead of fins) can help with stability for purely vertical flight, but not if I want to pitchover. Yes, I'm being careful. But at some point, the aerodynamic forces just rip the nose out from inside the prograde marker...

 

[RizzoTheRat]

My initial thought was it's using fuel from the top first so the CoM moves down as you fly, but your test of stopping flow from higher tanks should help that.  How easy would it be to lock all but the bottom couple of tanks off, and then manually activate each tank up the stack in flight to keep the CoM as far forward as possible?

Presumably with that many parts in the stack it's pretty flexible so as soon as the nose goes off prograde it's going bend until it snaps.  What happens if you pitch over straight off the launch pad and then hold prograde? 

[Boris-Barboris]

Use staging, that sausage is too long, CoM shifts way too much towards the engine, behind true CoL and the rocket becomes unstable, not to mention dV qualities.

Don't use drag as tool to enforce of stability, use lift. Lift torques are dominant over drag for long crafts and don't waste energy. If you rotate those three tanks at the bottom the way that they are oriented like fins, you will make it more stable, and can drop the fins.

Edited March 11, 2016 by Boris-Barboris

[Wemb]

Yeah, what Boris said - that thing is just not suited for anything, possibly except going straight up - but even then you'll go so high you'll probably have problems with re-entry.  Try and turn at all and It'll be flexing, with almost no way to correct itself, it's CoM will be moving dramatically during it's flight and you have barely any control authority on it so you can't steer it and even if you could it'll probably snap, and even if it didn't and you could steer it at some point, by the time the fuel has drained, you'll probably no longer be able to steer it...

You have stack decouplers and SRBs - use them. Get the science, and get some AV-R8's and bigger fuel tanks - the AVR8s will provide you will a huge improvement in control, and with bigger tanks you'll be able to build stronger less bendy rockets.  

But, basically. if you want to get high in the early game, use stages, and make your first stage an SRB. Till you get thrusters, reaction wheels, and steerable winglets, you're reliant on the reaction wheels in your pod for steering - they're only going to be useful on modestly sized ships..

Wemb

 

Ah - have just seen you have 1392 posts - I suspect you know all this already...:-)

Edited March 11, 2016 by Wemb

[GoSlash27]

I think the problem is it's too structurally noodly. You have the attitude reference and reaction wheel in the nose, and thrust vectoring in the tail. Since it's flimsy in the middle, this sets up a control loop oscillation, especially with SAS engaged.

 I make a similar vehicle for my first orbit, but mine uses the LV-T30. Since there's no thrust vectoring, it doesn't oscillate.

Best,
-Slashy

[Nich]

Is this a joke or troll?

[GoSlash27]

8 hours ago, Nich said:

Is this a joke or troll?

Not at all. @UmbralRaptor is a widely recognized and respected space pirate.

The design looks ridiculous because it's super- early career (single node unlocked) and this is all that's available to work with. I consider myself doing well hitting orbit at this stage of the game and he's going for a Munar flyby. Pretty badass if he manages to pull it off. I don't go for a Munar flyby until a good deal later in the game.

Best,
-Slashy

Edited March 11, 2016 by GoSlash27

[Nich]

So I made this with 30 parts 

And got it to orbit with a little more then 897 dv which should be enough for a free return trajectory.  I messed up the assent and ended up in a 95x95 orbit.  I would never be able to hit that without patched conics.  I am playing with life support as well and I would never be able to make it there with enough electricity to survive.

 

[GoSlash27]

1 hour ago, Nich said:

So I made this with 30 parts 

And got it to orbit with a little more then 897 dv which should be enough for a free return trajectory.  I messed up the assent and ended up in a 95x95 orbit.  I would never be able to hit that without patched conics.  I am playing with life support as well and I would never be able to make it there with enough electricity to survive.

 

Good stuff!

 Is it under 18t? The lack of electricity and patched conics is workable in stock. How did you get it stable?

Best,
-Slashy

[Nich]

I used the reliant (x2) it was super stable.  I started a new career and only got 10 science for the first 2 nodes.  It was VERY hard to get to orbit.  Tip too soon and you burn up, tip too late and you cant get a good gravity turn going.  Some where between 40 and 45 m/s tip to 5 degrees worked best for me.  I suspect less joints for tanks helped and the fact that it drained from the middle down keeping some weight up front for the second stage. MJ assent guidance would have been nice

Edited March 12, 2016 by Nich

[UmbralRaptor]

@Wemb, @Boris-Barboris:Yes, and no. Using boosters means kicking the mass out above the 18 tonne limit, or dropping a fair amount of high-Isp tankage. It might help with a partially fueled RT-10? (Though of course, a shorter stack also means less wobble). Tanks as fins is promising for stability, but ultimately needs fuel pumping. (not available at that point in career mode, and if it were, one can apparently gain enormous stability from pumping fuel up!)

@GoSlash27 Yeah, bending might be a bigger issue than I thought -- shorter designs (albeit with more fins) have made it to LKO. Er, and it's just using an LV-T30. Control authority is sharply limited.

15 hours ago, Nich said:

Is this a joke or troll?

A completely serious attempt to explore part of the design space.

 

For a more general update, it turns out that a partially filled booster, 3-4 upper tanks locked down for the ascent, and very careful attention to keeping the craft inside the prograde marker can get to LKO. And then run frustratingly short of enough ΔV for a munar flyby. (I estimate 840-860 m/s, with my best attempt getting 820 m/s. For most of my ascents it seemed like I could get ~740 m/s with 25 tanks and an RT-10 with 30% fuel, but only ~670 m/s with 23 tanks and 50% fuel in the RT-10.

edit:

It is deeply bothersome when an approach suggested by someone who starts off by insulting you works best. But using 2 stages of LV-T30s resulted in bringing almost 990 m/s to LKO. I ran 11 tanks in the upper stage and 13 in the lower from a best guess on split efficiency in my spreadsheets. The ascent was comparatively unremarkable, with pitchover starting at 50 m/s, and then just staying inside the prograde marker until I got a 70ish km apoapsis.

Edited March 12, 2016 by UmbralRaptor
moar testing

[GoSlash27]

UR,

 Congrats on making it. That's quite a feat! And I'm sure Nich didn't intend to insult you. I could picture myself thinking the same thing if I didn't understand why you chose those parts.

Best,
-Slashy

[Nich]

To be honest I did not even think Munar flyby was possible with tech that low I am normally happy just to get to orbit at that point.  And yes I was originally confused by your first post because you did not address the known noodling issue that is a design challenge for tech that low and you didn't use the FLT-200 which is available at the same time you unlock the swivel.