A very unwieldy rocket--either pitches over or spins out [KWRocketry & FAR]

[birrhan]

Since this is a hybrid mod post, I'll post it here and let the mods sort out the mess. The short of it is: I have a somewhat simple rocket using KW Rocketry, and using FAR as the aerodynamics model. Four stage lifter (12 x Globe V SRB, 4 x Titan XX liquid fuel, and 2 stages each of the Titan T1). this is all pushing a payload of 90-110 tons. At 90 tons, I can reliably get payloads into orbit. at 110 tons, one of two things happen:

1) The craft will pitch over (engines get ahead of themselves) at around 20-25km, speed 500-800 m/s. I have adjusted every combination of throttle and gravity turn angle, fin number and placement, as well as RCS thrusters (those high powered ones). Craft here:

The inevitable orientation here:

It appears that it starts to lose responsiveness as the topmost tank empties. The last image shows just one of my attempts--placing fins further down get ripped off by the launch tower or SRBs, and raising the SRBs off the liquid tanks appears to lead to rapid spinning out. What's killing me is that the behavior is indicative of high drag, but my ship ought to be compatible with FAR (designed for low drag).... so what gives??

2) Spin out. Begins immediately after launch and quickly accelerates out of control.

I'm fairly experienced with the game, what's throwing me for a loop is fixing the problem "legitimately"--using the FAR mod for aerodynamics and maintaining correct orientation without, say, a nose-mounted RCS block. Any number of smaller ships work though, just this one is being buggy.

[Audiopulse]

well, i guess 110 tons is just too heavy. Up to the height in the second picture the atmosphere was dense enough for the rudders to keep your ascent under control - but at around this point it becomes too thin for the controlflaps to do anything so the heavy nose topples over. Thats my guess.

Also - Ive never used FAR, but I hear its a little buggy sometimes. Drag at the tip of your rocket (Fairings not being counted by FAR + high-drag payload) might still be the answer, too.

That reminds me - I think you locked the Gimbal on the outer engines, no?

[stupid_chris]

You're just doing your gravity turn too aggressively with FAR. Start early (nearly at the launch pad) but be extremely gentle with your rockets if you don't want this to happen.

[Autochton]

I noticed you have some fins on the upper part of your inner ring of boosters. When you shed the outer ring, those will shift your center of lift radically forward, which will make your rocket go ninja: Flip out and kill people. Basically, check your CoL vs. CoM against every stage that will operate in atmosphere. If at any point the CoL is in front of the CoM, you will not be going to space today.

As well, your rocket will actually have relatively high drag, given the very wide configuration you have going on. Every rocket stack will add to it, this is why you see real-world rockets often looking quite thin and tall. Even a monster like the prospective Falcon Heavy only has two side-stacks. So it seems to me like there's a fundamental problem in your rocket design here.

Edited July 22, 2013 by Autochton

[birrhan]

I noticed you have some fins on the upper part of your inner ring of boosters. When you shed the outer ring, those will shift your center of lift radically forward, which will make your rocket go ninja: Flip out and kill people. Basically, check your CoL vs. CoM against every stage that will operate in atmosphere. If at any point the CoL is in front of the CoM, you will not be going to space today.

As well, your rocket will actually have relatively high drag, given the very wide configuration you have going on. Every rocket stack will add to it, this is why you see real-world rockets often looking quite thin and tall. Even a monster like the prospective Falcon Heavy only has two side-stacks. So it seems to me like there's a fundamental problem in your rocket design here.

Oh I'm sure there is, but I'm modeling it after this post, and he got it to work just fine: http://www.reddit.com/r/KerbalSpaceProgram/comments/1ie1fz/3kerbal_395ton_aerobraking_lander_with_remote/

110 tons even.

I have tried removing the fins, putting many fins down low, all to no avail. The magic (tip) happens after the SRBs drop and after I'm halfway through my liquid fuel. it's precisely when the upper tank empties. Which makes me think there's a coding error involved with the empty tank.

[Donziboy2]

I have managed to get this into space on several occasions, the top portion has 6 large atomic engines and im using the kethane pack other then that its pretty stock. The top half is 187T with no kethane added.

At 831T its a chore to get it into space, the trick is to pump most of the fuel from the top tanks first, as your lower engines suck down fuel your weight distribution changes to your upper stages. Once you cut engines while you wait for your AP burn you empty the lower stages back into the tanks and dump them. Fuel balancing mod pays for itself a hundred times over.

Here it is mated to my 321T tanker currently orbiting the Mun. I need to get a shuttle up there with some Quantum Struts since the silly SAS makes the engines jump around constantly and causes lots of wobble unless I disable most of them during burns.

[birrhan]

I have managed to get this into space on several occasions, the top portion has 6 large atomic engines and im using the kethane pack other then that its pretty stock. The top half is 187T with no kethane added.

At 831T its a chore to get it into space, the trick is to pump most of the fuel from the top tanks first, as your lower engines suck down fuel your weight distribution changes to your upper stages. Once you cut engines while you wait for your AP burn you empty the lower stages back into the tanks and dump them. Fuel balancing mod pays for itself a hundred times over.

Here it is mated to my 321T tanker currently orbiting the Mun. I need to get a shuttle up there with some Quantum Struts since the silly SAS makes the engines jump around constantly and causes lots of wobble unless I disable most of them during burns.

I see you have MechJeb but not FAR. I strongly suspect this is a problem with FAR coupled with the massively overpowered engines in KW Rocketry.

[birrhan]

double post

[BananaLee]

Pretty much what Autochton said. FAR is quite unforgiving if your Centre of Lift ever gets ahead of the Centre of Mass.

I haven't built massive designs like yours (so far, still on Earth Orbit practice) but I've learnt to try to fix it in three ways: -

1) Shift winglets as far down as possible to push CoL down

2) Shift side stages upwards to bring CoM up

3) Actively manage fuel distribution, i.e. shut off the top fuel tanks until you get out of the atmosphere

[birrhan]

It appears that the problem was that my payload extended a wee but past the fairings, and is those pieces lacked nosecones they introduced drag up top. I put nose cones inside the fairings, and I'm about 3 orbits for every 5 launches.

It was drag, all along. I think. I also notice that when it does pitch up and over that FAR generate a severe stall warning. On a rocket. But whatever, it's a party.

[Lexif]

It was drag, all along. I think. I also notice that when it does pitch up and over that FAR generate a severe stall warning. On a rocket. But whatever, it's a party.

It's the winglets that stall. I got that too and it's because they get deflected beyond their stalling angle in the thin upper atmosphere, when ASAS was flapping them about like mad. (ASAS is not FAR's friend.) It might also have something to do with the transsonic effects as it reliably happend around mach 1. I got it under control by constraining their deflection angle to 6° in the VAB FAR menu.

[birrhan]

You're just doing your gravity turn too aggressively with FAR. Start early (nearly at the launch pad) but be extremely gentle with your rockets if you don't want this to happen.

I do. I'm at 65 degrees around 19000 m, when this all happens. How shallow do you suggest? Degrees and speed at what altitude?

[rpayne88]

Four stages to orbit? I personally think that is a bit much, but that's my opinion. Looking at your first screenshot, I believe your outer liquid engines and your SRBs ignite at launch. The SRBs burn out and you ditch them. That leaves a good portion of your mass above the geometric center of your rocket. I'd download the Treverios (sp?) Pizza and aerospace mod. This allows you to place rescaled stock control surfaces that are about three times larger. I've done this before with FAR and had astounding success. By the way, what is that thing on the other side of the runway? Just curious.

Edit: I'm also assuming those are KW fairings by the title of this thread. I know that Novapunch fairings are not FAR complaint and will still cause parts to create drag.

Edited July 24, 2013 by rpayne88
fixed a typo

[birrhan]

Four stages to orbit? I personally think that is a bit much, but that's my opinion. Looking at your first screenshot, I believe your outer liquid engines and your SRBs ignite at launch. The SRBs burn out and you ditch them. That leaves a good portion of your mass above the geometric center of your rocket. I'd download the Treverios (sp?) Pizza and aerospace mod. This allows you to place rescaled stock control surfaces that are about three times larger. I've done this before with FAR and had astounding success. By the way, what is that thing on the other side of the runway? Just curious.

Edit: I'm also assuming those are KW fairings by the title of this thread. I know that Novapunch fairings are not FAR complaint and will still cause parts to create drag.

Correct. SRBs flame our around 13km, and the rest of the first stage gets me most of the way there. Ideally. What just worked was keeping the pitch to 75 degrees (15 degree off vertical) until 32 km, then pitch over to 20 degrees pitch at full throttle. of course, this is not efficient--lost about 250 dv from what I expected, so there's room for improvement. keep plugging away....

[birrhan]

By the way, what is that thing on the other side of the runway? Just curious.

Nothing there. Nothing to see there. move along.... that's the, um, pizza delivery rocket.

[zanzibar196]

Lose the winglets, and your rocket will fly like a champ. As of right now, too much lift is being generated by them, and you will tumble...

[birrhan]

Lose the winglets, and your rocket will fly like a champ. As of right now, too much lift is being generated by them, and you will tumble...

Sorry, thought I'd clarified--that was one of many versions, most of which lack winglets there (or near there). All of them pitch over, actually, regardless of winglets (position or number). It is simply the size of the rocket using FAR. I actually replicated the behavior on a much smaller rocket--if you do a gravity turn too quickly, you're gonna have a bad time--and "too quickly" is much slower than you'd think at first. I do it in sub-5 degree increments now, and tab the opposite direction after each adjustment to make sure I'm still in control.