Unexplained structural failure

[SilentWindOfDoom]

Hello everyone, I've been plagued a rather annoying situation that causes linkages to fail for no apparent reason. The simple design as seen here: http://i.imgur.com/1LmXajx.jpg is just a testbed i've been using to see what happens. Even when i treat her gently (slow acceleration, a minimum of maneuvering) the link between the nuclear engine and the fuel tank above will inevitably fail at a time when nothing is really is happening. Steady acceleration through the last 10% atmosphere at 75% thrust? Nope! camera shifts, linkage fail. This happens regardless of the interstage adapter (rocketry) and seems to plague the nuclear engine especially.

The relevant mods i am using are Rocketry and Ferram Aerospace.

Quick edit: The structural failure takes place between the engine and the fuel tank above it, even if it is a stock rockomax tank.

Edited June 20, 2013 by SilentWindOfDoom

[Damazo]

I'm not sure if this is the same problem but this happened with me also. When about to make my gravity turn the camera would shift and the ship would more than likely blow to pieces, due to a linkage failure. This only happened when I was thrusting fast and on any kind of time warp, and when I stopped warping it stopped exploding. But if in doubt more supports works.

[SilentWindOfDoom]

Well, it isn't warping, i never do that during launch. It doesn't happen when the engine cuts out or kicks in either. It is just this one stage, blasting away until at some point during the trip the atomic motor wants to go off by itself.

[capi3101]

I've heard there are some bugs with Ferram Aerospace; this could easily be one of them. I don't actually use Ferram; best advice I can give is the standard "put struts between the two pieces"

[stupid_chris]

You normally also have the adapter decoupler right? If you set it right, you can add struts on the base of this adapter. What I usually do is that I add four struts from this adapter to the tank. I have a rocket that has this EXACT design and it works like this. Though I'm not using FAR.

[Jarin]

I've had a ton of trouble with nuclear engine stages, to the point where several of my rockets had only outer boosters so my inner fuselage could have a nuclear engine at the bottom without connecting anything. in the end, the secret is struts struts struts. If you want a really rock-solid connection, use the "cubic octagonal strut" in 6x symmetry, lined up above and below the engine, then connect them both vertically and diagonally (each strut should have 3 EAS-4 Strut Connectors running to or from it).

[ferram4]

I suspect the problem is that the nuclear engine you're using doesn't have its impact tolerance and strength increased enough for its increased size. Based on what I've seen, this is a problem caused by g-force spikes due to the nature of the numerical integration needed to simulate physics in KSP; you'll have to either reduce thrust, increase the strength of the part in its config file or use a lot of struts.

[SilentWindOfDoom]

You normally also have the adapter decoupler right? If you set it right, you can add struts on the base of this adapter. What I usually do is that I add four struts from this adapter to the tank. I have a rocket that has this EXACT design and it works like this. Though I'm not using FAR.

In this fashion?

http://i.imgur.com/vmHeA6w.jpg

Ive test flown it and it held together. I will require more testing, but you may have done it!

[Alinor]

As someone pointed out, you are missing struts.

Although things that snap together have some ability to stay together, you need to reinforce with struts to get stability.Otherwise they will slide side to side, relative each other. Weak structure usually shows on stage changes, as you stage, some, or all thrust is lost and next stage engines "kick" the vessel as they start the burn. That kick can be considerable, depending on the engine configuration and can be seen on the G meter and can break a rocket if it's structurally weak.

For smaller craft you can just slap some struts between stages, and it will stay together.

For large rockets/ships/stations struts need to be strategically placed to be most efficient and as few as possible, as you want to reduce part count on huge things.

Large number of struts will create lag especially during atmosphere part of the ascent, where you get allot of flex due stress on the vehicle.

Good luck.

[stupid_chris]

Well I meant with the sort of conic interstage one that goes from 3.75m to 2.5m, but this does I guess