Solution to the Heavy Rocket Problem

[swamismurf]

I\'ve been thinking about the problems caused by heavy rockets and the much more 'realistic' fuel tanks by the modders. The problem, as I see it, is that the parts are too heavy for Unity to deal with.

The solution is simple. Divide all masses and thrusts/moments by 10. You could then have 10 times more fuel/engines/boosters without the parts pogoing and melting.

@Harvester: Mod the parts folder so everything is 1/10th its current mass. If you modify the original parts, then everyone else will 'balance' theirs to match and the problem of super-heavy collisions will be ameliorated.

I\'m going to try this on all parts packs I download from now on.

[jdogtulsa]

Are you serious?

[GroundHOG2010]

Too heavy for unity????

There is no such thing (as far as I know, though I know heaps of people who use the free version). The mass (not weight) is a Integer (or possibly a float or a double as things may be) and has no problems with crashing the game. The thing that would possibly crash the game is amount of polygons in the model * a number of these models + a not strong enough CPU/GPU/RAM.

[Kryten]

The problem is that the parts are far heavier than those the connection model was meant to deal with-not some fundamental Unity issue-and it simply needs updating to reflect the large masses of many common parts.

[Tiberion]

Its not Unity in general, just the way its implemented in the KSP engine, specifically the attachment node system; When the cumulative weight of the stack reaches a threshold, the connections give up - sometimes it may snap the final connection between the tank and engine on the pad, sometimes the upwards force of the launched engine breaks a higher point and the stack pancakes, or it gets wobbly, or snaps in half.

Its not user error, and there isn\'t much to do about it other than strapping on struts over joints.

Btw Swamismurf: You can\'t just lower weight, or the engines will be super-powerful. You have to adjust fuel capacity of tanks, fuel burn rates, and engine thrust along with it.

[swamismurf]

By Unity, I meant KSP\'s connection implementation.

I stated: 'Divide all masses and thrusts/moments by 10.' this means all force/resistance type numbers should also be divided by 10. Until Harvester implements some type of connection fix which would allow massive tanks (perhaps the connection size variable would control this behavior) this might reduce the 'part dancing' enough to allow players to keep building bigger rockets with higher delta-v budgets.

[Tim_Barrett]

The only people who have this problem are those who don\'t know what surface pressure is...

I\'ve built big-ass rockets before. Heavy too.

Basically, six engines on the ground instead of one will have one sixth of the weight bearing on it.

[what-the]

If you really need to, build a stand out of parts for the rocket to sit on. I have done it. The middle engine was too low and all the weight was on it. So I attached decouplers and fuel tanks to the bottom of the others. And used boosters to get a safe distance away to fire my liquid engines. And when I mean safe, I mean the moment I can blow up the detached fuel tanks without blowing up my rocket.

[Mictaste]

Blowing stuff up just to get off the pad, totally jeb approved.

Doing a complete re-balance would fix nothing because the attachment points are set that way on purpose if you lower/edit everything you would also lower what the attachments could take, solving nothing.

I use detachable stands on alot of my recent rockets, your just talking about making easy mode where nothing breaks.

[ludsoe]

This is why we need custom launch pads.

[Sordid]

The only people who have this problem are those who don\'t know what surface pressure is...

I\'ve built big-ass rockets before. Heavy too.

Basically, six engines on the ground instead of one will have one sixth of the weight bearing on it.

Alternatively, use lander legs on radial decouplers. They\'ll hold your engine off the ground so it doesn\'t get crushed and you can jettison them as soon as you lift off.

[CarolRawley]

By Unity, I meant KSP\'s connection implementation.

I stated: 'Divide all masses and thrusts/moments by 10.' this means all force/resistance type numbers should also be divided by 10. Until Harvester implements some type of connection fix which would allow massive tanks (perhaps the connection size variable would control this behavior) this might reduce the 'part dancing' enough to allow players to keep building bigger rockets with higher delta-v budgets.

You do realize that if you divide all the integers by 10, the problem will be unchanged, just making it harder to balance mod parts. Duh.

[crekit]

Someone needs to link that show about the girl who was afraid of weight. She wouldn\'t put stuff in her fridge cuz she was afraid the floor would collapse, among other crazy stuff.

[Tiberion]

I think you guys are a little confused - its not JUST a launchpad problems - the connections are prone to giving up during flight too, as the ship accelerates once leaving the atmosphere the forces they experience can still overload the already-wobbling connections between the heavy 2 or 3m tanks and ANY part. Its possible to make this happen with the long tanks from Nova and Sunday Punch too, but it takes several stacked together - its just that the big K&W parts are so heavy that one tank/engine combo overloads the internal part-connection system.

Lowing the overall weight while maintaining the ratios WOULD HELP, because connection strength isn\'t something the guys can mod up to compensate, the impact tolerance does affect it tangentially, but its not a direct modification of the 'glue-factor' between 2 parts.

[Xobotun]

As for me, I use long H-shaped beams every 7.5-10 meters and LOTS of strut connectors between these beams and each part (excluding radial-mountable batteries and fuel tanks:)) This helps to remove wobbling, and these beams are very useful for placing additional RCS fuel (in small white spheres) and RCS thrusters, but the rocket become a bit uglier. You can also place beams on decouplers that don't leave visible joints.

Also, reducing size of parts is not a panacea - you'll have to rescale the whole Kerbol system and gravitational constant xD