FAR - Rocket flipping, every freaking time! Please help!

[bohemian]

Hi guys,

so I am having some FAR woes.

I am trying to send 3 communication satellites in KEO and I practically stole

. This is how my ship looks like, PF fairing and all. This is showing the payload, please disregard the high COL, it is because the fairing is open, the actual COM and COL points are shown in the first pic. Please also disregard the low delta-V, I know it is not going to reach orbit. The SRBs are adjusted to provide a TWR or 1.2 and the main engine (a Skipper) is also providing 1.2 at around 6k, which from what I 've gathered it the suggested TWR for FAR. This is showing the COM and COL, when the tanks are empty.

Those being said, the mother-flipping rocket flips every mother-flipping time!!! I 've tried everything! Fins, no fins, less fuel in the lower tank to bring the COM up, turning at 70m/s, turning at 100m/s, turning as soon as I leave the pad, turning at 1km. Everything. It used to flip at 6-7km high, but then I adjusted the thrust so that I wouldn't reach Mach 1 until I was 18-20k up. That way I made it past the 6km up to maybe 28km up, by veeeery carefully following the prograde (+- 10 degrees) but at that point if I tried to move outside the market it would flip like a dart being thrown upwards. Since I couldn't move upwards, the rocket started to experience heat effects from its speed. All these were of course possible only because SAS was on. Even the thought of turning the SAS off, would send my ship spinning wildly into every possible direction.

I 've tried pretty much everything posted in the forums, including checking the PF to see if all components were shielded. If needed I can upload a video of one of my failed attempts. Please help me, because at this point I am not sure if keeping FAR makes any sense at all....

[Tex]

Aha, but I think you may have partly answered your own question.

That way I made it past the 6km up to maybe 28km up, by veeeery carefully following the prograde (+- 10 degrees) but at that point if I tried to move outside the market it would flip like a dart being thrown upwards.

This is the main idea, and I think the problem is your rocket not being top-heavy enough. Fins and more weight help to a degree, but what you need is a heftier payload to make it truly as if you're throwing a dart. On a dart, the heavy part is at the front, and when it's thrown the weight is keeping it on course, the rest of the dart being used for stabilization. This is what you need to think of when building a rocket in FAR, and even in the new aerodynamics system. Once your stage is smaller it's not so much of an issue because you're out of the atmosphere, but try making your probes heavier and it should work.

[cybersol]

Disclaimer: I don't know anything about FAR, but I'm guessing these stock 1.0.2 principles may help you

Close to prograde generally means inside the green circle, or at the tip of the outer hashes in extreme cases.

A quick summary I posted previously:

Spinning out of control essentially comes down to a contest between drag and control. When the drag overcomes your control authority, you go for a spin. You can decrease drag by staying closer to prograde and building a more aerodynamic rocket. You can increase control using engine gimbals, stabilizer wheels, and control surfaces.

Also note that drag increases significantly right around mach 1 (340 m/s). So you should stay pointed as prograde as possible at speeds around 280-400 m/s.

On a rocket control fins should be as low and outside as possible. If you are using control fins on more than one stage, the CoL needs to stay below the CoM at all times, even as the tanks empty.

In your particular case in stock, for a quick fix I would add control fins on both the outer boosters and the middle stage. For example in stock this gets up with bad aero and high TWR:

Or use more engine gimbal or more reaction wheels.

[SAI Peregrinus]

+-10 degrees is WAAAY too steep. At 70m/s, turn very slightly. As you cross 100m/s you should be at about 5 degrees, at 150m/s at 10 degrees. Then let SAS lock the nose to prograde and touch only the throttle until you hit orbit. A TWR of about 1.6-1.8 tends to be a lot better than 1.2.

If you can, post your .craft file of the launcher (and list any mods used).

[Wanderfound]

Draggy nose, not a lot of control authority.

Swap the delta-deluxe winglets for some all-moving surfaces (AV-R8s, tailfins, etc) and get some gimbal. A few of the vernier engines (e.g. Thuds) are an easy way to add lots more gimbal authority.

As a demonstration of the stability-enhancing power of gimballed thrust:

Javascript is disabled. View full album

[Good_Cat]

Looking at your rocket once those booster go and your high enough that the fins don't have a great deal of influence as the air thins out the heaviest part of the rocket is the back end, so it'll flip.

Even looking at the rocket now the weight is all quite low, short and fat rockets aren't very stable.

As ^^ he said some fully movable canards (AV-R8s, tailfins) and Mach 1 at 18-20km?? how fast is that in FAR. As long as im not going transonic below 10-12km in the thick atmo or get reentry heat during assent the faster the better.

[cicatrix]

What's the 'dry' TWR of your boosters? I think either larger fins or stronger reaction wheels (or both) might help. Also, keep in mind, that SRB do not have gimbals and have no vectoring thrust capability so it's much harder to steer a rocket that flies on SRBs. Consider changing SRB for LF boosters

[bagofholding]

Having some sort of real second stage would help with the CoM issues as there would be heavier upper tanks and an engine above the first stage keeping the CoM high as the first stage empties.

The reason the drag is acting so bad and causing you to flip out is it has a longer "lever" to work with. Just glance at the distance from the tip of the rocket to the CoM (which will only increase in this design) vs the CoM and the CoL at the bottom. When drag applies to the top it can pull down with a greater force than can act with the lever that the CoL is using on the CoM. The thrust vectoring can help the CoL situation a bit but still likely not enough to save things in this case.

[seaces]

I also have issue with flipping rocket but in stock atmosphere. Two days ago I was trying to send a huge rocket that was carrying bunch of new modules for Mun Station, return pods for some crew members, fuel tanks for lander and some parts to repair lander etc. Basically to make it to orbit I had to keep it exactly in the middle of prograde which is very difficult. Little deviation and rocket flips over. I solved it by turning on RCS ports that I had for docking anyway I had ridiculous amount of propellant. The impact of RCS was huge I mean it was like flying very little stable rocket 100% control. So if you can try with some RCS.

[bohemian]

Thanks so much for the information guys!

I made some changes to my original rocket, namely the larger fins, which helped a lot and more fuel on the top, but more importantly the ascent profile. I tried starting my gravity turn at 70m/s and around 1000m, however it started dipping too fast which meant that if I followed the prograde to within the stability limits the ship would be horizontal at around 20km and at that point it would be folly to try to get into orbit with all the atmospheric friction.

I tried starting my turn around 6km and given my high mass I hadn't accumulated too much speed and it sort of worked out ok, leaving me with around 700 delta-V. Given that I need to get into KSO, I would say this was a failed attempt and I am not sure how I can reach it with this rocket. If I add fuel I won't have enough TWR.

I also tried a different design, 2 stages with the Swivel engine, based on the Falcon 9, with 4 SRBs but it is woefully underpowered and with even less delta-V. At this point I am not sure how a sleek 1.5 design is supposed to work. Its COM is obviously much better, it even flies relatively ok without the SAS, but with the Swivel it is never going to have enough thrust.

I also struggle to see how exactly a rocket like the Saturn V didn't have problems with COM position.

Unfortunately I can't add attachments yet, but here are the craft files for the first design and the second one. Let me if they are not working.

Here is my mod list:

1)6S Service Compartment Tubes

2)Active Texture Management

3)Astronomer's Pack

4)Atmospheric Sound Enhancement

5)BahamutoB Animation Modules

6)Distant Object Enhancement

7) EVE

8)FAR

9)Fusebox

10)Habitat Pack

11)Infernal Robotics

12)KAS

13)Kerbal Engineer

14)KIS

15)Kerbal Joint Reinforcement

16)KSP AVC

17)KW Rocketry

18)MechJeb2

19)MechJeb and Engineer for all

20)Module Manager

21)Procedural Fairings

22)Remote Tech

23)Stage Recovery

24)TweakScale

25)USI Exploration & Kolonization Systems

[MrDude]

Try adding more stages (Saturn 5 (or as XKCD.com calls in Up Goer Five) was not running scaled and tweaked physics. Kerbin is not anywhere near earth sized).

Your probably fighting Light Pointed Dart Syndrome - As Maximus described.

Solution options:

A) MORE STAGES - lock the fuel weights up front. (also try the AV brand fins)

Manually manage your fuel access with the right click fuel tank menu - lock out all tanks on the offending stages with the little play icon next to the resource and stop the flow. Then starting at the tail (probably after you've gone to single stack stage {balancing boosters in this fashion is hard}) right click and open each fuel type and let that tank drain to 95%. If using auto staging be more conservative since MechJeb see's blocked fuel as empty and will eject those stages.

C) Reconfigure the payload - Do the 3 probes weigh more than the capsule? Perhaps instead of Capsule - Tri splitter < 3x Probes. Try Tri-Splitter < 3 Probes > Tri Splitter Reveresed (with decouplers) and then capsule. Or if the fairings can be enlarged Hang the probes from the Capsule to concentrate the weight and then use structural parts to push that weight forward. Heck you could use fuel pipes to route fuel from a front tank for the last part of the accent and ditch it from infront of the payload when your at your desired location.

[Fengist]

I also struggle to see how exactly a rocket like the Saturn V didn't have problems with COM position.

Because real rockets work off the center of pressure, not the center of thrust. Page 9 may help explain it. Even they knew it was unstable. All I can say is, the Saturn V designers were rocket scientists. We're not.

Edited May 26, 2015 by Fengist