How to build a heavy lifter that doesn't collapse?

[Troberg]

I have made a lifter that can lift around 100 tons, give or take. Now, I'm trying to do something that can take a heavier load, say around 200 tons, preferably with enough Delta V left for a planetary transfer burn.

However, whatever way I do it, tall or wide, the construction just tends to collapse. The radial decouplers are too weak, and struts does not help (not even lots of them, intelligently placed (I'm an engineer, so I know hot to place struts...)). if I use several radial decouplers to hold the stacks together, it behaves as if only one of them actually couples. If I build it high, the joints between tanks wobble until they fall apart, even with strut reenforcements. I've tried using beams to make it more solid, but it doesn't help.

Most of my tries has been with stacks of two orange tanks, usually with two "circles", an inner circle with 6 stacks and an outer with 12. Engines has been either mainsails or a combination of mainsails (for T/W ratio) and tricouplers with aerospikes (for Delta V, when T/W ratio is sufficient). Asparagus staging is used, but they seldom have time to stage before falling apart. Typical take-off weight is around 1500-2000 tons. My old lifter is similar, but with one orange tank stacks, and it works, if one is careful.

I've rebuilt various configurations over and over, from the start, so there is no standing oversight/messup in the construction. What am I missing? Some of the constructions, which are about as solid as I can make them with stock parts, fall apart even before the clamps are released, and most fall apart before even reaching 1000 m. This is starting to seriously frustrate me, I've had about 20 failed launches (but no kerbal lost).

Is it simply that stock parts are too weak to build vessels of that size?

So, hints? Tips? Tricks? Sample working vessels of that capacity that I can examine?

[nick1236]

Going thru this now myself, except I'm trying to lift 3 orange tanks bound together and get it to the moon. My payload with moon landing gear is 120tons. last night I managed to get to LKO but not enough to get me to the moon. What exactly are you trying to lift that weighs 200ton?

[Richy teh space man]

Are you using rockomax mainsails? If so try disabling the gimbal on them using action groups, or just have one in the centre gimbal. They can oversteer at times which shakes rockets to pieces.

You could also try flying manually, as the ASAS may overcompensate during flight.

[Penguinhero]

1. use lots of struts

2. don't go full throttle

3. don't put engines on orange fuel tanks

[Scrogdog]

Sometimes stitching tank seams together with struts helps a bit. Wobbles seem to grow harmonically or something. Something you don't see in real life methinks so their use in that way might not be intuitive.

In fact, I'll go so far as to attach radial hard points to get that done in some cases!

[Nao]

More is less!

My design that could bring 202t up to LKO with no struts whatsoever.

The more strut's there is the higher the maximum tensions will be. I always try to use as low amount of struts as possible.

Lately i started using a docking port-strut combination to attach payload to launching stage, to ease up the tensions. As long as there are no direct strut connections to the payload, there are much less explosions due to load concentration. Although it does induce wobble so that's another problem to worry about.

Nowdays i try to have 1 or 2 struts between each part in a "booster" stack to lower the wobble and allow higher loads. (especially since 0.20 brought up more explosions due to simple stack loads)

Then i use one radial decoupler and a pair of docking ports to hold up the booster, and then at the top a docking port and strut for connection to the cargo.

Whit this setup, and 25 mainsails i managed to bring around 300t of payload in one go into orbit. And the system is capable to increase to 49 mainsails ... (when i get better PC )

Good luck!

[Cranium]

1. use lots of struts

2. don't go full throttle

3. don't put engines on orange fuel tanks

Why don't you want engines on the orange tanks? What should be used instead? I have a similar ambition, to make "Fat Bertha", a rocket specifically designed to carry fuel up to a station. Currently, I have my design working, able to get into orbit with 20, yes 20 solid rockets, with four onion staged(yes, I know asparagus staging is better, but current design won't allow for asparagus without radical redesign of aerodynamics) orange tanks powered directly by one Skipper engine each. These are wrapped around a two stage stack of orange tanks, with the bottom one powered only with a single mainsail engine, switching to four white radial engines at the final orbit stage.

I can get Fat Bertha up to orbit with an almost completely full orange tank, but I would like to see how I can make it better.

[Cheaterman]

There is a strutting guide somewhere on the forums which helps a lot but basically you want to put struts even in between orange tanks in order to avoid them from collapsing - I did a Saturn V replica that was about 1'500T on the launchpad and worked:

The 1st stage was 2 orange tanks on top of each other with one big mainsail below it, the two orange tanks would have 4-symmetry struts in between them to prevent them from collapsing into each other, and I alt-copied the stack before putting radial decouplers in a 4-way symmetry on it and therefore duplicated the bottom stack.

Then strutted the bottom tanks of the "fuel boosters" to the parent tank (they would all detach at the same time in my case but for asparagus - or onion - you'd also put fuel lines there), cross strutted the tanks to one another in the middle, and even added a strut right above the extended radial separator since it seems weaker than the hydraulic detachment manifold - anyways it's not strong enough.

Finally I strutted the "boosters" to one another, once at the bottom, once at the top, could cross-strut the tanks in the middle as needed but it wasn't necessary in my case.

Before strutting this properly I had a 1/3 success ratio and needed to toggle the gimbal on my middle engine just like the real S-IC of Saturn V launch vehicles.

Now it's more like 90% - failures occur once in orbit

PS: In some rare cases you may even need to strut the engine to the tank. To achieve this, the strut must go from the tank to the engine as you can't radially attach anything to engines

Edited June 6, 2013 by Cheaterman

[Scrogdog]

Orange tanks tend to overheat when a high powered engine like mainsails are attached. You can get around that by adding one small tank of the same diameter to the bottom.

[Akalaël]

There is a point in the game you can't depass :

If you stack too many weight, the engine of the game considers that the structural forces of objets are exceed and will collapse. there is nothing to do about it, just stop stacking, and go sideway.

You got some addons that gives you struts 10th time more strong than the stock one (I think it's KW Rocketery, not sure, I don't use it myself)

Edited June 6, 2013 by Akalaël

[Troberg]

I have thought a bit more on the subject, and I feel that it would be nice with radial decouplers made for long stacks. Have a really sturdy (and proportionally heavy) decoupler that runs the length of an orange tank. That'll be on my wish list for some upcoming version! It's really hard to build large constructions when it feels like the entire construction is made out of rubber...

What exactly are you trying to lift that weighs 200ton?

Larger parts of a Duna base. In the future, multiple lander interplanetary missions. Also, I like to have some spare Delta V on my interplanetary missions...

3. don't put engines on orange fuel tanks

That could be part of the problem. On my earlier lifter, I used stacks of grey tanks, as I had problem getting the orange tanks to connect to the radial decouplers.

I'll make some experiments later. As far as I can see, performancewise, the grey tanks are equals to the grey tanks.

Then i use one radial decoupler and a pair of docking ports to hold up the booster, and then at the top a docking port and strut for connection to the cargo.

It looks like a very sturdy construction. It could probably be made even sturdier by using hexagonal symmetry, as there will be six sets of docking ports instead of four to hold it together.

However, how do you decouple the docking ports to drop stages? Right-click would be too slow.

Is it set up for asparagus staging?

By the way, try using three aerospikes instead of mainsails on some of the inner stacks. That gives better Delta-V at higher altitude when you don't have to accelerate as much any more. That has been successful for me. Just make sure you keep it symmetrical.

Anyway, it seems like Jeb likes it simple: http://rpglab.net/nobackup/fyf5.jpg

[Corbald]

Personally, I skip the use of Mainsails, altogether, and prefer engine clusters. These supply more efficiency at similar thrust levels, but require a better computer, as there's more parts. As to your question on decoupling docking ports, IIRC, you can assign them to action groups.

The comments about attaching Mainsails directly to orange tanks is no longer valid. This dates back to when orange tanks didn't handle heat properly and would cause rapid engine overheating. It's been fixed.

Generally speaking, I have had luck with a central single orange tank with an engine cluster, surrounded by clones in asparagus staging.

The Above lifted the Below (those tanks were emptied and considerably better trussed at launch) with little problem, if a bit slowly.

[Nao]

It looks like a very sturdy construction. It could probably be made even sturdier by using hexagonal symmetry, as there will be six sets of docking ports instead of four to hold it together.

However, how do you decouple the docking ports to drop stages? Right-click would be too slow.

Is it set up for asparagus staging?

By the way, try using three aerospikes instead of mainsails on some of the inner stacks. That gives better Delta-V at higher altitude when you don't have to accelerate as much any more. That has been successful for me. Just make sure you keep it symmetrical.

Dv is not a problem as even the 300t cargo ended up with a mainsail and an orange tank of fuel left in launch stage. and if i need more Dv it's really easy to add another 2-4-6-12 mainsail stacks.

As for the choice of engine 3x aerospike is only 1/3 of mainsail thrust. And it's mostly thrust not fuel that dictates how much we can get to space. Not only that but aerospikes require additional control surfaces to not flip the ship in atmosphere and also provide no gimbaling forcing use higher up so large amount's of RCS ports and tanks are required.

The design on the picture was using control groups and no radial decouplers. The one i have for 300t uses radial decouplers and docking ports decouple automatically when staged.

I tried doing hexagonal symmetry but it was harder to make, set fuel lines and figure out staging.

The stacks that are connected by more than 3 ports are not a problem, usually its the stacks at the rim or asparagus that cause problems. And with hexa setup that wouldn't really change much.

Ah an yeah the thing is asparagused in stacks of 4,4,4,4,2,2,2,2,1

@Corbald Yesterday i launched a ship with Mainsail below orange tank, and it exploded due to overheating, the "bug" (or fearure ) is still there.

[UmbralRaptor]

The comments about attaching Mainsails directly to orange tanks is no longer valid. This dates back to when orange tanks didn't handle heat properly and would cause rapid engine overheating. It's been fixed.

Actually it hasn't. They still explode when run at full throttle for ~20 s on an orange tank. (I just tested)

[Troberg]

As for the choice of engine 3x aerospike is only 1/3 of mainsail thrust. And it's mostly thrust not fuel that dictates how much we can get to space. Not only that but aerospikes require additional control surfaces to not flip the ship in atmosphere and also provide no gimbaling forcing use higher up so large amount's of RCS ports and tanks are required.

When I use the 3x aerospikes, I still use mainsails for the thrust, especially on the first stages. However, once I'm up to speed and at higher altitude, the thrust from the aerospikes is sufficient, especially as I need to carry less fuel.

As for control, I still have a few mainsails on the construction, so they provide the needed control.

The design on the picture was using control groups and no radial decouplers. The one i have for 300t uses radial decouplers and docking ports decouple automatically when staged.

Just for clarification, so I understand you right: If you put decouplers and docking ports to keep stuff together, when you decouple, the docking ports between the parts automatically release as well?

[Innuce]

Could always build a JetRocket! They're slow getting up to speed, but are incredibly fuel efficient!

In this video, it gets a full Jumbo Fuel Tank (complete with RCS and extra accomodations), and a large rover/lander to 100Km orbit around Kerbin...

Edited June 7, 2013 by Innuce

[Nao]

When I use the 3x aerospikes, I still use mainsails for the thrust, especially on the first stages. However, once I'm up to speed and at higher altitude, the thrust from the aerospikes is sufficient, especially as I need to carry less fuel.

As for control, I still have a few mainsails on the construction, so they provide the needed control.

Just for clarification, so I understand you right: If you put decouplers and docking ports to keep stuff together, when you decouple, the docking ports between the parts automatically release as well?

Yup and the radial decouplers are two times the medium's docking port height so its really easy to add them (It's actually possible to control "depth" while placing docking ports/radial decouplers, especially when you are in 30deg or more symmetry mode, it can help making things perfect and make things not want to place/connect so keep that in mind.

Also i tested the desingn in 0.20, and it looks like placing one strut between fuel tanks in a stack is now mandatory for the thing to not explode .

Looks like this.:

The test payload is connected by struts to docking ports. As a positive side of 0.20 it seems that having all 25 engines gimbal at the same time no longer make everything explode, for it's weight it made it quite nimble Although i still woudn't trust MJ/ASAS piloting this with tall load:P

[KevinW42]

Just so you know, you don't need fuel lines connecting your engines to your tanks to get clustering to work, so there's one way you can probably cut down on your parts count.

[Nao]

Hmm just found one small thing that *might* help any high part-count ship. CheatEngine's speedhack. Yes we have physics timewarp with physics-delta to slow down time if PC cannot keep up. But I've noticed that it works somewhat differently to speedhack. (for example you can punch 5x speed in CE on heavy and long objects under load and there will be no heavy wobble/bending that occurs with 4x timewarp KSP provides.)

I don't have much time on my hands right now to properly test the thing, but after setting speed to 0.5 in CE and launching a heavy rocket that always exploded during ascent, there were no explosions. (the clock was green mostly with flashes of yellow)

Since Cheat Engine is free, easy to use and quite reliable, It would be nice if somebody else could also take a look on it (especially since i use quite unique designs that may not give a good indication on viability of this method).

Cheers!

[Penguinhero]

hi i posted this in the wrong thread so i edited it because i dont know how to delete them

[Sirine]

Sorry I bring up this post....Just wonder...any news? About getting this 200 tons to standard Kerbin orbit?

[PrivateFlip]

There are lots of examples over the forums. Also check out asparagus staging.

I'd like to go for the simple brute force solution:

it can get around 200 tons in orbit.

The answer to not having them collapse is struts. There not visible in the picture above because they are all tucked away in the four gaps in the 'nine-pack'.

Add them wherever you think should be some sensible reinforcement. If the rocket collapse look at where it breaks or check the flight log an make more reinforcements.

[Ace_Of_Spades]

Moar Rockets

[rdfox]

My solution? Break it up into packages of a mass you can manage, then assemble it on orbit. You know, the way von Braun said to, and the way that almost all real massive cargoes are launched IRL...

[Monkeh]

15 mainsails and LOTS of fuel all held together with struts.

A fair few struts but just struts and decouplers, nothing else.

This may be a stupid question but you have added some big red clamps for the first stage?