why this simple rocket does not work?

[king of nowhere]

Premise: many of my rockets aren't all that great. because i am flying cumbersome payloads, or because i am trying to save money by using suboptimal engines, or because i didn't care enough to make it better when what i had was alreayd enough to get to orbit.

But in all those cases, i can see why my rockets are suboptimal. i can see where the problems are. and generally i can still get them to orbit, if i have to fight with the controls along the way.

which is why i was totally baffled yesterday when i made a very simple rocket, with a very simple payload, that should have a good aerodinamics and obey all the golden rules for making rockets, and it didn't work. at all.

here is the rocket

(i called it "save private rafnia" because rafnia was about to reenter atmosphere without a craft, i had to intercept him on the fly. though later i changed my mind and am now just trying to aerobrake an EVA stranded kerbonaut directly back from mun orbit. just for the sake of trying it)

the engine at the bottom is a reliant. the cargo bay contains a probe core (i need to leave room in the capsule for taking in rafnia, after all) and 2 batteries. starting TWR is aout 1.25, a bit low but enough. still, i only get the worst effects when i picked up enough speed, so a slow start should not be a huge problem.

i've tried many iterations of this simple design. the large first stage was originally split in 2, but the craft lost control when i separated the first stage. i treid removing the solar panels and add more batteries in the cargo bay. i tried fiddling with winglets. nothing to do.

always this rocket capsizes. it won't do it immediately, but as i start to tilt eastward, it will gain more and more tilt. despite me pumping hard on the reactions wheels to fight that tilt. it generally reaches 45° at 250 m/s, and by the time it's at 500 m/s it's flying almost horizontally, well before getting high enough for orbit.

i suppose i could have fixed it with a couple of reaction wheels, and i would likely do that on other flights. but i didn't want to do it in this case, because i wanted to figure out why this thing wasn't working.

in the end i only could make it work when i made a larger base and used a bobcat as first stage engine. and since i was there, i also split the large first stage in 2, with a reliant for second stage. the craft became very stable this time, getting to orbit with nothing to do but throttling the engines. incidentally, it also refused to steer, but at least it stabilized at 45 degrees, so it could enter orbit with a decently optimized trajectory. adding a couple reaction wheels would have been cheaper, anyway.

i would like to understand why such a simple, streamlined rocket had such huge aerodinamic issues. because all i thought i knew about rockets is telling me that this thing should fly like a dream; or, in the worst case, it should have problems steering but it should stay pointed in the right direction.

[ElWanderer]

The reliant engine has no gimbal, which means you only have the (admittedly overpowered) reaction wheels in the pod with which to steer.

Using the LVT-45 Swivel instead means you have an engine with gimbal, that will make your craft easier to steer. However, the Swivel has lower thrust which you'll have to make up to get off the pad, possibly with strap-on boosters.

You say your liftoff TWR is about 1.25, which is quite low and it'll only rise slowly as you have a lot of fuel to get through in that first stage. That means your flight profile needs to loft the rocket quite high - turning too much/too soon will result in you being horizontal way too low. The reaction wheels won't be enough to help lift the nose to maintain the trajectory you need. As such, it's probably very sensitive to how you manoeuvre at low altitude.

[Streetwind]

This style of rocket is one I frequently build and fly myself, and like you, I would be absolutely baffled to find that it has aerdynamic issues. Because there's honestly no way for it to do so. It should be perfectly flyable without fins, even. And using a Swivel isn't required either (and would in fact be counterproductive as it doesn't have the performance required).

So, I'm with @ElWanderer here - your TWR is probably too low for the trajectory you are choosing, meaning the gravity turn happens faster than you can accelerate. You're tilting over not because of aerodynamics issues, but because of a lack of thrust. The fact that you have fins simply pins you harder to this trajetory of falling over sideways, as you are in fact too aerodynamically stable to fight it!

You have two... no, three options now, actually. The first option is to replace the Reliant with a RK-7 Kodiak, if you have already unlocked it. It's basically a direct upgrade to the Reliant in every way except vacuum Isp, which is actually mitigated by the fact that sea-level Isp is higher, meaning its higher efficiency off the pad cancels out the lower efficiency as it rises. This engine should easily let you launch this rocket.

The second option is to just fly straight up for longer, and tip over more gently, so that you have a much steeper trajectory. It's kind of inefficient to fly this way though, costing extra dV to fight against gravity. And if you're losing dV either way, you might as well choose option three:

Remove a fuel tank. From my experience, the Reliant can push enough rocket into low Kerbin orbit that it easily has the dV necessary to perform multiple rendezvous maneuvers there. So there's really no reason to overload your rocket to this degree. My gut feeling when looking at the screenshot says: probably one of the five tanks in the first stage can go. When building rockets, it'sgenerally recommended that the mass fraction of all stages be relatively similar, and your first stage looks to have a higher mass fraction than the second. You may in fact even be able to eke out more dV if you add a FL-T100 to the second stage after removing a FL-T400 from the first.

Oh, and ditch the fins. You don't need them. That'll further improve sea level TWR and total dV.

[king of nowhere]

3 hours ago, ElWanderer said:

The reliant engine has no gimbal, which means you only have the (admittedly overpowered) reaction wheels in the pod with which to steer.

Using the LVT-45 Swivel instead means you have an engine with gimbal, that will make your craft easier to steer.

yes, i could have used a swivel, or i could have added reaction wheels. i didn't because i wanted to figure out what was wrong. a well-crafted rocket should not need either.

i generally accept that i'm flying an inferior rocket as long as it does the job, or that i'm flying a cumbersome payload and stability problems can't be avoided. but in this case i wanted to know.

1 hour ago, Streetwind said:

 

So, I'm with @ElWanderer here - your TWR is probably too low for the trajectory you are choosing, meaning the gravity turn happens faster than you can accelerate. You're tilting over not because of aerodynamics issues, but because of a lack of thrust. The fact that you have fins simply pins you harder to this trajetory of falling over sideways, as you are in fact too aerodynamically stable to fight it!

so, let me get this straight: i always considered the gravity turn a manuever to be performed, but perhaps i was wrong here.

you are telling me that all rockets will capsize, no matter how well balanced they are. and that a succesful gravity turn does not entail steering along the trajectory, but accelerating at the right speed so that your rocket will capsize at the right speed to make you turn successfully? seen in that light, it makes everything clearer. i don't marvel anymore that most rockets will capsize.

Quote

Oh, and ditch the fins. You don't need them. That'll further improve sea level TWR and total dV.

the fins were not there in the first design. i went through several iterations. in fact, the first attempt i made at this rocket was very much along the lines you suggest: higher TWR in the first stage, compensated by another intermediate stage higher on, and no fins.

it also didn't work. in fact, the first stage almost got there, but as soon as i detached the first stage, the rocket went wild.

anyway, i said i eventually did launch it with a bobcat engine.

[Spricigo]

10 minutes ago, king of nowhere said:

 a succesful gravity turn does not entail steering along the trajectory, but accelerating at the right speed so that your rocket will capsize at the right speed to make you turn successfully? 

Exactly.

[O5-7]

You should replace the basic fins with canards to make up for no engine gimbal. If you still have control issues on the other 3 stage variant, you should instead have a swivel engine on the middle stage. Of course you could always add a rediculous amount of reaction wheels, but that would be impractical since the craft has only 1.25 TWR. If you are short on money you could just do 'more boosters' and use a swivel for main engine with four canards. Believe me I used to have this problem all the time when I first started: once the rocket starts to tilt and can't recover it's all over. I hope that this helps with your issue! 

[ElWanderer]

Yes, ideally.

With SAS disengaged, your rocket should just follow surface prograde, which will lower towards the horizon as you go (the Wikipedia page for gravity turn tries to explain this, though I'm not sure it does a great job).

It's diverting from surface prograde that will produce aerodynamic effects that could flip the rocket, unless you can counter them.

With the right initial turn, you don't need to steer away from prograde at all, or at least until you're high enough that it doesn't matter.

Personally, I can't be faffed with getting the right initial turn for each design, so I use gimballed engines to retain control authority (and add small fins to prevent flipping). I am aware this leads to launches that are not as efficient as they could be.

Edited July 5, 2020 by ElWanderer

[Reactordrone]

Going up it's flyable, you just have to be gentle with your initial turn since it has very little control authority. Coming down it will want to flip head first so you might want to put an extra decoupler on to get rid of that service bay.

Edited July 6, 2020 by Reactordrone