HOW NOT TO FLIP: Some Crazy Things I've Gotten to Orbit, And How

[Mister Dilsby]

Rather than reply to every "Help! My rocket won't stop flipping!" or "KSP is impossible now!", I thought I'd put up a few Crazy Things I've Gotten to Orbit with some comments on how I did it. I bet the rest of the community has ships bigger, uglier, prettier and crazier than mine.

1) "Ore Piggy" Delivery vehicle

This one carried a pair of ore transporters to my Minmus mining operation.

Problems: Large mass, uneven internal payload

Solution: Thrust vectoring of the LFBs helped a lot. Aero shape and minimal fins were enough for the core stage.

2) Payload in a big fairing

I've used a design like this to deliver several space stations and surface labs.

Problems: Large mass. Wide front means lots of front-end drag even with fairings

Solution: No fin was big enough. I put four delta wings with elevons on the end of the core stage. You can see them sticking out between the strap-on boosters.

3) Big-S Space Station

This is my main fueling station around Minmus. The 'ore piggies' from craft #1 cycle to it from the surface.

Problems: Huge mass, not streamlined. Docking ports and fuel pods everywhere.

Solution: Forget about efficient ascent. Fins and SAS to hold course straight up to 30km, then turn east. Lots of struts connecting stages--typically running between two sets of four radially placed cubic octagonal anchor points (one set on each of the two mating stages).

4) Not-SSTO Interplanetary Explorer

This is my Duna ship: VTOL capable with full surface mining/processing facilities. I did not want to carry the mass of jets to Duna, so this is a rocket plane only.

Problems: At first I tried carrying jet pods on the wings and ejecting them at altitude, but the highly variable thrust of V1.0.2 turboramjets made that impractical--center of thrust moved faster than I could react, flipping me out.

Solution: Drag can lift as well as balance--horizontal launch with "White Knight" style expendable first stage rocket plane. There is a Skipper engine mounted dorsally to the LFB (and behind the orbital vehicle, not connected to it) to bring the center of thrust closer to center of mass. Constant adjustment of LFB thrust required to keep thrust offset angle near zero.

Everything can still be done, you might just have to do it a different way. Possibly by taking a solution from real life, possibly by doing something #lolsokerbal, but there is ALWAYS a way.

Edited May 19, 2015 by Kuzzter

[Chewy]

That's why I love this game. It constantly gives you new challenges, and encourages you to engineer solutions to your problems created by exploring new ideas and trying new things.

I've typically leaned towards designing my payload to fit the rocket rather than the other way around, but this post might motivate me to try some more unwieldy payloads in the near future.

[peachoftree]

I once got this thing into orbit. thats a 10m faring with 3.75m boosters!

[KSK]

Great post! Nice selection of examples too, ranging from 'yeah that looks OK under any aerodynamics' to 'good to see that kerbal contraptions can still fly'.

[Mister Dilsby]

I once got this thing into orbit. thats a 10m faring with 3.75m boosters!

oO wow! I actually have not unlocked the big tanks yet in career, but that looks like a great way to use them!

[peachoftree]

oO wow! I actually have not unlocked the big tanks yet in career, but that looks like a great way to use them!

If you plan on doing this, it only works if you have 2-way symmetry, as it is the only symmetry that is not really glitchy on a procedural faring.

[DuoDex]

Punted to Tutorials

Also, thanks. I've been having trouble with this for a bit.

Edited May 20, 2015 by DuoDex

[Captain Sierra]

I'll contribute.

I needed to get this fuel depot to the Mun.

The problem:

All that stuff made the drag highly irregular, very unstable, and generally a pain.

The solution:

While the fairing does absolutely nothing to reduce drag, it homogenizes the forces into something more uniform & easily counterable.

The ascent pattern was fly up till booster sep, keeping a ~1.7 TWR, then fireball it, set SAS prograde, and switch orbit mode.

The payload got to the Mun fine.

[TheSaint]

Here, this is one of the craziest things I've sent up:

Under the fairing was a 106-ton station with a spin habitat. (I was feeling lazy, didn't want to send it up in pieces.) The key for me was to start the gravity turn early, it started only a couple hundred meters up. Very gradual, the nose was never really out of the prograde marker until it was well over 30k up. After that I heeled it over to the horizon and burned to orbit like normal. The first orbit was sloppy, 131k x 75k, 1.5 degrees, but it was in orbit. I sent up a little service module later to circularize it at a 100k x 100k.

And people say the new aero is unforgiving. LOL.

[Mister Dilsby]

While the fairing does absolutely nothing to reduce drag, it homogenizes the forces into something more uniform & easily counterable.The ascent pattern was fly up till booster sep, keeping a ~1.7 TWR, then fireball it, set SAS prograde, and switch orbit mode.

Very good lesson here: Kerbin is small enough and rockets are powerful enough to push almost anything through to orbit, provided the forces are balanced. The challenge for new players is to understand all those forces and make use of them.

- - - Updated - - -

The key for me was to start the gravity turn early, it started only a couple hundred meters up. Very gradual, the nose was never really out of the prograde marker until it was well over 30k up.

Quite different profile from some other big fairing designs. Maybe the length of fairing relative to rocket spreads out your drag forces in the vertical direction, whereas ones like Capt. Sierra's and mine are all at the front? (requiring an ascent with late gravity turn to avoid flip)

[Captain Sierra]

Very good lesson here: Kerbin is small enough and rockets are powerful enough to push almost anything through to orbit, provided the forces are balanced. The challenge for new players is to understand all those forces and make use of them.

Those are not stock engines though, so I have a MASSIVE advantage over stock. They're cryo engines. My sustainer engine outputs 2650kN of thrust (150 more than a Rhino) and peaks out at a vac Isp of 405. Heck, my upper stage maxes out at 455. Also, these burn LH2/Ox (InterstellarFuelSwitch). My fuel tanks thus have a worse dry mass/wet mass ratio, but weigh less overall, shifting the CoM higher up & making my rocket more stable.

Quite different profile from some other big fairing designs. Maybe the length of fairing relative to rocket spreads out your drag forces in the vertical direction, whereas ones like Capt. Sierra's and mine are all at the front? (requiring an ascent with late gravity turn to avoid flip)

The discussion came up of what fairing size is too large. From some documents from Boeing (at least I think it was Boeing; it pertained to the Delta IV Heavy at least), we found that realistically, a fairing can be up to 167% the diameter of its stack and behave just fine. For the largest stock fairing, that will enclose a 6.2m payload. We also found that in the real world, you can go as large as 190% before aerodynamic forces become difficult to counter. Thats up to a 7.1 meter payload.

ALso, stock aero is pretty easy to fly under if you keep the forces balanced. I managed to get this up through a proper gravity turn:

[TheSaint]

Quite different profile from some other big fairing designs. Maybe the length of fairing relative to rocket spreads out your drag forces in the vertical direction, whereas ones like Capt. Sierra's and mine are all at the front? (requiring an ascent with late gravity turn to avoid flip)

Dunno. I've always launched that way, even under 0.90 with FAR. It was not stable or easy, it took a couple of tries and a lot of concentration. And you'll notice the liberal use of large delta wings as stabilizing fins. But it got there.

[spaceBrezel]

I tamed it

launcher with 28 boosters for more safety

Its a small space station. With lots of space instead of fuel.

Tried the classic "stick an engine at the bottom and some boosters to the side" variant first but it flipped and exploded like crazy (1000 - 5000 m).

Soon i realized it needs more stabilization to the side and i built a "cage" so more fuel and weight is in the front section. Spontaneous explosions at ~15 km so i changed my ascent profile...

Do not touch anything till 20 km -> turn and burn. Lots of fuel wasted but im in orbit. Repeatedly now with several variants ...

Its now in space, pretty useless but its there.

[nfpinto]

biggets thing i flyed. no need for fairings tho. guess the mass would critically increase

[Warzouz]

I was working on SSTO rocket familly (from 12 to 236 tons payload). I noticed that bigger rocket are easier to fly thant smaller ones. When you understant to fly small rockets, it's easier with bigger ones. I usually don't fail orbiting big stuff, but I sometimes mess things with smaller one. Hopefully, I love big stuff (don't hask why ) so I don't use small rockets on kerbin anymore.

I'm working on a 180T LKO space station (regular one with 2 huge fuel tanks already filled.) Using my biggest SSTO rocket (236T) it goes in one shot to LKO. But I'm still fine tuning the station (especially on docking ports)

My maine problem when ascending is not the new aero : it's wobbling. I hate adding struts everywhere. Fairing should prevent wobbeling for parts inside it. But I've solved the problème with Joint reinforcements mod.

New aero ot more fun than old one. In the 0.9, everything goes in orbit with proper dV and TWR. IT was so dull, that I use MJ for all ascent. Now I fly all rockets myself, it's very pleasent.

[basic.syntax]

@ Warzouz: Wobble. Ugh, yes, I don't like it, either. But I have no hesitation: struts everywhere, lol.

To all: I thought I would share this example of a "thrown together" design, that finally got me to the Mun and back, fairly early in career.

I had fun trying different stuff... it's kind of messy. I wouldn't try to win any contests with it. But I think it fits the thread, another example that tall craft can work, & you don't need fancy stability parts like reaction wheels. (But they sure are nice.)

Many notes in the album.

Javascript is disabled. View full album

[Geschosskopf]

My craziest thing to orbit yet in 1.0.2:

NOTE: The explosion in the orbital pic is intentional--it's just the spent SSTO booster becoming not-debris thanks to TAC Self-Destruct (best mod ever) initiated when staging it away . Anyway...

PROBLEM: Get 2 large-winged Basic Jet probes to Laythe. Large wings means problems for aero if put on the front of the rocket. The probes lack RCS so also had to go up on their tug, which consists of a full orange tank with 4x LV-Ns, and I'm too lazy to bother sending the tug up empty and refueling in orbit before departure, so that had to go up full and not be used during ascent.

SOLUTION: Use the wings of the probes as the tail feathers of a rocket. Build a large rocket above the tug/probe payload, and make it SSTO so it retains more mass up high to counter the mass of the tug as the bottom. Add delta wings and BIG elevons to the nose for control and disable the control surfaces of the probes.

RESULT: Flew just fine using normal gravity turn trajectory like a conventional rocket.

[Mister Dilsby]

SOLUTION: Use the wings of the probes as the tail feathers of a rocket. Build a large rocket above the tug/probe payload, and make it SSTO so it retains more mass up high to counter the mass of the tug as the bottom. Add delta wings and BIG elevons to the nose for control and disable the control surfaces of the probes.

Thanks, this is an important point that hasn't been raised before on the thread--sometimes turning a control surface off is better than keeping it on, because they like to fight each other! For my planes I always make sure the rudder does just yaw, the elevators only pitch, and nothing responds to roll except the ailerons. Disabling one or more (or all) control directions on a rocket fin may help with control problems as well.

[nfpinto]

Heard you guys like huge fairings (yet to me deployed)

idk how to properly share the album here

[basic.syntax]

... idk how to properly share the album here

It's described in the forum FAQ, here.

[Mister Dilsby]

Another technique, just used to orbit my new asteroid miner/hauler:

PROBLEM: Heavy load without a center attachment point. (Klaw is in the way) No place to put boosters.

SOLUTION: Radially attach the lower stage(s) as needed, leaving the center blank, and tie it all together with struts. Keep symmetry, and use all other techniques for stability. This one could not gravity turn in low atmosphere. Went to 25km, turned east and burned. 4x nuclear motors in the main ship took over at 60km to finish orbit.

ETA: This is my 500th post!

Edited May 27, 2015 by Kuzzter

[StrandedonEarth]

I've recently tried to launch some ungainly rockets that would have been just fine pre-1.0. After many test flights fiery crashes, I think I've figured out how to stabilize these beasts all the way to orbit.

First and most important, Throttle Controlled Avionics is required. This is a very useful mod, especially for VTOLs that may not be perfectly balanced. TCA adds a button to the right-click tweakable menu that cycles between 'Main Engine,' 'Maneuver Engine,' and 'Manual Control,' and will throttle engines as necessary to maintain balanced thrust.

The first thing I tried to launch was this LFO-moving rover, to move LFO around my Minmus operations.

Fairings didn't help, as I guess it was too far off balance. I tried different things, like fins on the 2nd stage, but still no go. Even going nuts with vernors wasn't enough. I already had TCA for landing the rover, and with a bit of tweaking, I was able to make orbit.

All of the main lifter engines were set to 'Manual Control' . The 'Twitch' engines on the rover were left as 'Main Engine' and assigned to a hotkey. Some big control fins (delta wings and elevons) on the bottom got it through the first stage burn. When I staged, I also activated TCA (press 'y'), RCS, and the Twitch engines. This was enough to keep the flame side rearward the rest of the way to orbit.

My next difficult launch was sending a science package of probes to the Eve system. Each long cargo bay holds a lander and a scanning satellite, and I wanted to launch a set for both Eve and Gilly in one launch. The result was this tall skinny thing, which was totally flip-happy. The lifting-body cargo bays probably didn't help.

So I did the same thing. I put some Twitch engines at the top of each cargo bay, and set up the engines like I did previously. I lit the Twitches and activated TCA at the first sign of an impending flip. It was stable all the way to orbit and is currently on the way to Eve.

I hope this helps somebody.

Edited June 2, 2015 by StrandedonEarth

[cybersol]

These are all career examples, which force some interesting compromises.

The craziest things for me tend to be stations. Below is the first example, where I launched the lab module, habitat module, and tug all at once. This one is notable just because its a very long and draggy fairing with relatively light weight components inside, which is why I wanted to launch all three components together. The key here is mostly 20 AV-R8 control fins placed as low and outside as possible (along with the 5 mainsail gimbles in the rocket SSTO launcher). Also note that all struts are as aligned with the airflow as possible.

Here is another station, and the hardest thing I've had to launch. First of all, I didn't have low tech like no fairing that could hold it and only modular girders for structure. Secondly, most of the connections use docking port juniors, which are all that I had but are VERY wobbly. Here aligning the struts with the airflow was key. When the struts were in an X configuration they wobbled until it all exploded, but switching to the vertical airflow aligned configuration below eliminated the wobble. I also using 20 low and outside control fins here. Notably I tried wings here, but they did not help nearly as much as just adding more control fins.

Finally, a different class of problems comes when hauling things with their own lift to orbit. Here I am lifting my standard nuclear tug. I like this configuration with 2 wide nukes, because they are stackable (need more TWR, just add more tugs), *limited by wobble. However, the wide arrangement doesn't fit in a fairing, so I had to overcome the lift generated by the radiators fins. Here I used 4 delta wings generated enough lift below the CoM to offset the lift in the upper stage. The launcher is my medium 3 mainsail rcoket SSTO.

Cheers,

CyberSoul

[Cirrus]

Thanks, this is an important point that hasn't been raised before on the thread--sometimes turning a control surface off is better than keeping it on, because they like to fight each other! For my planes I always make sure the rudder does just yaw, the elevators only pitch, and nothing responds to roll except the ailerons. Disabling one or more (or all) control directions on a rocket fin may help with control problems as well.

Thank you for this tip. You just fixed all my planes.

I didn't realize that all the control surfaces were trying to work at the same time when I was changing only one axis.

[Ehco Corrallo]

I'm not excellent at designing rockets, but I have found that the FAT-455 tail fin is great for bigger launches. They provide plenty of control and don't bump up the part count too drastically. 

This is not a bigger launch, but it is easily the least stressful rocket I've ever flown. 

Edited January 14, 2016 by Ehco Corrallo