Connection between heavy loads keeps to break

[Tokay Gris]

So I build a huge lifting stage. Basically, to get a Laythe-lander in Orbit. 130 tons.

As I usually do, it is overpowered. The lander got into orbit with the center asparagus stage still full and two remaining side-boosters almost half full.

So I decided to try something heavier. Duck-taped five Jumbo-tanks together, put them on top with a decoupler in between, strutted them many ways, came up with a TWR of 1,7, enough delta-V to get to orbit...

And the connection with the decoupler breaks (decoupler to payload, to be exact)

More struts, try again. Same thing.

So I put boosters on the payload. Works. Until the boosters run out and the decoupler breaks again.

Next, I put sideengines on the payload. Kind of defeats the exercise, because I want to get the fuel in orbit, not use it to get into orbit.

But doesn't work either (before anyone asks: Yes, gimbal was disabled on those engines). This time the payload broke apart.

(Then I had to get to work, which is far far away from a computer that can do this.)

Any ideas? How can I strengthen the connection between the lifter and the payload?

Boosters in sequence? I can't decouple them, because they would fall on the (VERY broad) liftign stage. But this might be an idea.

(Funny that. Using boosters not for the lift, but to reduce the strain on the payload...)

I did use cubestruts, strutted them, strutted directly, struts between the asparagus tanks and the payload, struts EVERYWHERE.

Didn't do zilch. With the boosters on the payload, I managed to get of the pad, otherwise it broke to piece without any thrust.

To show what I mean:

[capi3101]

Struts, struts everywhere....but I can't help notice the number of struts between the payload and the booster (you know, where the problem is occurring) looks a tad paltry. Try adding some running from the tops of those X-32s on the booster to the bottom of the Jumbos. And I mean from all of them (the ones outboard as well) to all of them...

Some more struts between the orange tanks would help if it starts shaking itself apart on a regular basis. Those orange tanks have notorious problems structurally.

[Xaiier]

You could try replacing the decoupler with two large docking ports. Same tonnage and a bit more strength, you'll have to watch the crossfeed though.

Your problem could be too much thrust, whats the TWR as it leaves the pad?

[Tokay Gris]

Yeah, I know. That was the only picture I had.

It was the first try and obviously it didn't work out.

Believe me, later versions look more like a pack of spaghetti. Struts from the payload to the asparagi, from the payload to the center stage, in between the payload, and more struts, whereever I could think of.

[DerekL1963]

It's not just about MOAR STRUTS!1!!11!... it's also about eliminating the sources of the vibration that are causing the breakage (I.E. relieving the cause rather than putting a bandage on the symptoms) - in this case, all the links between those X200-32 tanks on the booster. Try replacing them with Jumbo-64's.

[Tokay Gris]

TWR isn't the problem. The ONLY design that actually got off the pad was one where I had boosters on the payload.

This pictures was taken before physics were calculated.

It breaks apart without ANY thrust.

Even with loads more struts, the connection between the payload and the seperator breaks. Without thrust.

Too heavy? Sure! But Others get heavier stuff in to orbit with success.

[Tokay Gris]

In this case, all the links between those X200-32 tanks on the booster. Try replacing them with Jumbo-64's.

You think so? Like I said, the decoupler doesn't seem to stand up to the weight. And the problem seems to be how to reduce that weight with struts.

In RL I would build additional structual elements, but since I can't do that in KSP (the root "problem")...

[Tokay Gris]

You could try replacing the decoupler with two large docking ports. Same tonnage and a bit more strength, you'll have to watch the crossfeed though.

Ah, that could work!

One docking port would be enough, anyway... the lifting stage doesn't need to connect to anything, after its job is done.

Also, crossfeed is not an issue. The idea is to get the fuel into orbit, not to use it.

[Sᴄɪɴᴛɪʟʟᴀᴛᴏʀ]

If you look at the flight data it will tell you which parts are failing, find them and strut them.

[Tokay Gris]

If you look at the flight data it will tell you which parts are failing, find them and strut them.

I KNOW what part failed. I strutted it to excess. Later designs look like somebody glued a pack of spaghetti to it.

It is the connection between the payload (the five Jumbos) and the decoupler. It can't withstand the weight on the launchpad, let alone under acceleration.

[birrhan]

I KNOW what part failed. I strutted it to excess. Later designs look like somebody glued a pack of spaghetti to it.

It is the connection between the payload (the five Jumbos) and the decoupler. It can't withstand the weight on the launchpad, let alone under acceleration.

That is what I would have guessed. Replace the decoupler with a mated pair of Sr Docking ports. Bonus--you'll have to redraw all the struts between the lifter and payload, and now you should need much MUCH less. Only disadvantage is you'll have to manually "decouple".

[capi3101]

Only disadvantage is you'll have to manually "decouple".

Which you'd probably be able to overcome by setting up an action group. Worth a shot in any case.

[Tokay Gris]

The docking ports can take more stress?

In previous crafts, it seemed they "wobbled" a lot more. But that doesn't mean they break earlier... Hm. Will try that.

[Sᴄɪɴᴛɪʟʟᴀᴛᴏʀ]

The docking ports can take more stress?

In previous crafts, it seemed they "wobbled" a lot more. But that doesn't mean they break earlier... Hm. Will try that.

If that fails double or triple up on the connection. In the current stock version of KSP with the bad aerodynamics model wide rockets have a big advantage over tall rockets since they can spread the load more effectively with no negative drag effects.

[capi3101]

My understanding is that the way the Seniors have been built, they make a more rigid connection than earlier docking port modules. Now, caveat here: I haven't actually played with them myself. If that doesn't work, then you can try multiple ports. Three works, but you need to make sure and set up an action group so you can get them all decoupled quickly. Maybe toss a couple of seperatrons in for good measure.

[HoY]

If your willing to install quantum struts I've got the perfect solution for you

[Klajan]

Actually the Decouplers don't need that many struts. about 8 should be enough, but you have to strut the payload somewhere else. Try strutting it to the outer stages and to the fuel tank below the decoupler.

Maybe analysing this might help you (if you Computer can handle it that is) I designed it so you can strut the Payload to the nosecones to distribute the weight more.

[John FX]

My understanding is that the way the Seniors have been built, they make a more rigid connection than earlier docking port modules. Now, caveat here: I haven't actually played with them myself. If that doesn't work, then you can try multiple ports. Three works, but you need to make sure and set up an action group so you can get them all decoupled quickly. Maybe toss a couple of seperatrons in for good measure.

They are much more secure. Here are some images of a 281T payload on a lifter to use for ideas.

Struts

Closeup. Notice 281T is balanced on a probe micro core.

Takeoff

The lifter was one I got from a heavy lifter thread so i can`t take any credit.

[iEvermore]

I have a similar design and had a similar problem. (That's a lot of payload fuel mind you...) What I did was set up a scaffold system coming from the top of the top lifter tanks and then strut the scaffold girders to the payload and lifter at multiple points. Once you get high enough you lose the scaffolding due to stage asparagus separation you should be ok because the lifting thrust will no longer be as powerful.

If by chance that doesn't work. Add two midsize engines with small tanks to the payload with a fuel line to the lifting stage to save your fuel in the payload.

If all else fails, you can share your craft file and I'l fiddle with it for you or you bring that puppy up in 2 parts and assemble in orbit.

Docking ports are much stronger yes, However I prefer the clean look of separators, especially on ships I have no desire to dock. It's just personal preference.

Edited June 13, 2013 by iEvermore

[nhnifong]

Try controlling from a part near the center. All other things being equal, it tends to reduce wobbling by reducing the average delay between error measurements and corrections.

[Tokay Gris]

Try controlling from a part near the center. All other things being equal, it tends to reduce wobbling by reducing the average delay between error measurements and corrections.

Like I said, it breaks down BEFORE I control anything.

I will try with docking ports and different struts.

Bringing so much fuel in orbit is useless anyway. I don't see how I can possibly dock it to anything.

Basically, it was just a test, how heavy a payload I could get into orbit.

But such a weight is a step towards a lander for Eve.

[Tokay Gris]

If all else fails, you can share your craft file and I'l fiddle with it for you or you bring that puppy up in 2 parts and assemble in orbit.

Thanks, but that is not the problem. The fuel-payload is basically just a test.

The issue is not to put together a modular craft (I can get about 100 to 130 tons in orbit and assembled 600+ tons ships in orbit (after refueling), but to see how heavy a load I can get into orbit in one piece... With this one, it is fuel, but I could have stacked hitchhikers for testing instead.

[DerekL1963]

You think so? Like I said, the decoupler doesn't seem to stand up to the weight. And the problem seems to be how to reduce that weight with struts.

Yes, because vehicles on the pad wobble too... and that wobble can tear the ship apart in unexpected places. You're assuming the problem is weight.

[HoY]

The Supernova can get 160 tonnes into LKO easily if you don't mind using someone else's ships. It's 100% stock, and can be easily modified with srb's to get nearly 200t up.

http://forum.kerbalspaceprogram.com/showthread.php/33381-0-20-2-Zenith-rocket-family-%28modernised-for-0-20-x-with-perfect-subassembly%29

Edited June 14, 2013 by HoY

[Rickenbacker]

Someone might have mentioned this already... The big decoupler is actually only connected at a single point in the middle, it's much better to use the smaller decoupler and surround it with eight struts to connect even the biggest parts.