Anyone else have "unrealiable" rocket designs

[OUScooby]

I've been experiencing some problems launching some of my rockets lately. One time they will launch fine and get my payload into orbit and other times they will seemingly randomly explode or break apart or have parts fall off of them. This is with no redesign to the rocket. One time it will launch, the next nine times something will go wrong, and then it will launch fine on the tenth try. I've tried launching with MechJeb which should fly the rocket the same each time to rule out user error, and still I get the same results, sometimes and orbit, sometimes a Boom!

Has anyone else made unreliable rockets? How did you finally find the problem, and how did you fix it?

[J0hnsmine]

I fixed my giant ships with 45-90 more struts.

works 80% of the time for me.

[Sephta]

Normally ships don't fall apart randomly. There has to be a reason somewhere...

Take a look at the log and try to figure out, if there's one part failing most of your crashes. If a rocket crashes regularly, there is a weak spot somewhere, you can fix.

I had a problem with one rocket once, that would only start, if I manually helped Mechjeb to start it - otherwise the engines would overheat and explode. Got around that with minor tweaks...

I still have a rocket i can only start if mechjeb steers and I do all the fuel adjustments... (Yes - I have to pump fuel between stages to get this thing flying: Jeb does not approve most of the time...)

[NeoMorph]

I've had ships that have flown many times without problems just go haywire. I know when this happens the only way to fix them is the close KSP and restart. Something just goes bonkers you see. I think it's probably a corruption with the stuff in memory.

That said, I haven't had it happen once in 20.2 as of yet.

[FlamedSteak]

Weak spots dont always fail. Check your end mission log after a failure, the answer is usually in there.

[OUScooby]

The problem I'm having is that the rocket CAN work. I've launched successfully with it. But when I attempt other launches with it they end in failure. An engine will get knocked off when a stage decouples, or a radial decoupler will fail and whole engines and tanks will fall from one side of the ship. Other time solid boosters will come lose and fly up and into the rest of the ship, a single fuel tank will come disconnected but still stay in place causing one engine on one side of the side to run out of fuel before the other sending the rocket spinning uncontrollably, and then sometime it will just all fall apart in one spectacular fail. But then other times it will launch with no problems. All this without changing a single part of the ship. It's hard to diagnose the problem when something different goes wrong each time, and then just as randomly it will all of a sudden work.

I'm wondering if anyone else has made designs with these sort of results.

[GreenDragon]

Reinforce those problems one at a time, this just means that there are several problems, start with at least one strut from the top of SRBS to the tank they are atached to to keep them from random decoupling. Decouplers not being strong enough elsewhere? They never are, struts are needed for reinforcement. Engines getting knocked off during decouple? Then add some decoupling srbs, you know, the tiny ones to push the part away before yor ships spins add the engine gets ripped off. Do all of those and then see if there are many more problems. This seems a pretty basic set of problems, one fixed by reinforcing the spots undergoing the most stress. Struts are your friend.

[Sephta]

Ok, some things you can do:

* check your staging (perhaps a decoupler is triggered too soon?)

* fit radial stages with sepratrons and pull them away (exspecially if they get separated in your gravity turn!)

* MOAR Struts strut all parts with high stress on connection points - if possible divide the force to different parts of the rocket.

Check one thing after the other - perhaps you solve more than one problem with only one fix you do...

[jwenting]

Currently using one that consistently has the boosters exploding when I stage, but just flies through the debris with no damage done. Hairy experience, hearing a large bang and seeing a cloud of flame surrounding my rocket for a few seconds at first stage separation.

Have another one somewhere that used to be able to reach Minmus in 0.19 but now in 0.20 can't even get >20km on ascent from Kerbin.

[Flixxbeatz]

Unreliable, like this?

Then struts would be the answer.

[Johnno]

All I can add is that I stick to the KISS principle - Keep It Simple Stupid. I muck about with complex designs when I feel for a change or when my crude launchers don't cut it but otherwise I stick to the same old proven design(s).

For example I was telling a friend last night how I build my main lifter (40 ton payload to 85x85km or so), I counted the steps in building it and it turned out to be 17 with struts and all (my favourite number to boot, coincidence?). Simple, fast to build, crude, but well balanced, only two stages, no radial decouplers or solid boosters, just simple raw red tank + mainsail power with pancake staging, which I've personally found to be safer than aspargus staging.

As someone already mentioned, weak spots don't always fail. You can see it as that there's degrees of reliability, if it fails once in a blue moon then it's just odd chance of several factors aligning, if it fails as often as it works you have weak spots, if you can only manage to fly it sometimes then you're just lucky that the weak spots didn't fail that time.

It's all down to your design process. Like GreenDragon here mentioned, struts are your friend. Generally just slapping something together and going on full thrust nonstop will end up in unreliable rockets. You have to learn how to fly specific rockets and what their limitations are.

[OUScooby]

Well after 30+ attempts late into late night all ending in all kind of failures, each one unique like a snowflake, I decided to head to bed. This afternoon I tried again, with the same rocket, same design, no new struts or anything and it flew right the first time, no problems. Got my payload into orbit and docked with my ship I'm constructing in orbit. Then I went to launch the next section of the construction. This uses the same heavy lifter for the bottom stages but this time with a heavier payload. That made it into orbit no problem as well. I went back to the rocket that was giving me so much trouble last night and decided to try it again just to see what happens, got it up and in orbit again with no problems.

Thirty-some unsuccessful launches last night followed up by the very first three launches today going perfectly, with no changes to the design. What is the variable I'm missing here?

[7499275]

As for the craft breaking apart as it goes up ( or on the launch pad sometimes ) I would never put struts before .20 because 4 struts and my pc would lag like crazy but now that performance is better adding struts usually works for me... If not, I just go back and redesign the entire rocket! Best of luck to you though!