I keep getting structural failure during launch

[Kelderek]

I need help. I have a large rocket (3370t, 473 parts) that keeps having structural linkage failures during my launch ascent. It happens at the same point every time I launch: shortly after all 8 of the big boosters have all been separated (at about 32-40km altitude). The reason why this boggles me is it happens more than 10-15 seconds into that stage and there is no wobble - I have added about 40-50 extra struts and a bunch of girders all over the place and my center stack still fails at this point in the launch. The top-most Kerbodyne S3-14400 fuel tank fails first followed by both of the Mk3 adapters above it.

As you can see from my first image I should have more than enough struts to hold it together. In the second image, all of the girders and struts have broken off and those 3 sections of the center stack are now all debris while the lower portion of the rocket is still intact with engines running at full power.

Can anyone help me figure out why this always has structural failure?

Javascript is disabled. View full album

Edited March 29, 2015 by Kelderek

[Slam_Jones]

You might want to consider moving your later-stage boosters higher up the ship. Pushing from the back works great most of the time, but in space it can tend to wobble. Imagine pushing a piece of string from the back with your finger. Now imagine pulling that same string from the front.

[Deutherius]

By the looks of it you got too much thrust on the bottom and too much mass on the top. A slight bend in any direction will induce huge forces on the middle part, which breaks. I reckon if you don't do a gravity turn at all and just go straight up, this won't happen (well, unless the top wobbles).

I'd suggest throttling down a bit or making a slight design change - do you really need such a long middle part? Shortening the mid part and strutting from boosters to payload should increase the structural integrity enough. Probably.

[Kelderek]

There is no wobble at all, the whole thing is steady and solid when all of a sudden the girders fall off and those top tanks become debris.

Here's some more pictures. I added even more struts and as you can see everything is fine in the first image, no major pitch, yaw, or roll movement. Then 10 seconds later the girders have broken off and the top big tank and the two Mk3 adapters have broken off.

(I used MJ for the autopilot so I could focus my attention on why this is falling apart. I'm fairly certain the problem has nothing to do with piloting)

Javascript is disabled. View full album

- - - Updated - - -

By the looks of it you got too much thrust on the bottom and too much mass on the top. A slight bend in any direction will induce huge forces on the middle part, which breaks. I reckon if you don't do a gravity turn at all and just go straight up, this won't happen (well, unless the top wobbles).

I'd suggest throttling down a bit or making a slight design change - do you really need such a long middle part? Shortening the mid part and strutting from boosters to payload should increase the structural integrity enough. Probably.

I'm not sure how I could have too much thrust at the bottom, my TWR numbers are mostly on the low side. I tried launching straight up with no gravity turn and that is no better. I guess I need to turn this thing into more of a pancake - I was hoping to avoid that, but I don't have much choice.

[Deutherius]

I'm not sure how I could have too much thrust at the bottom, my TWR numbers are mostly on the low side. I tried launching straight up with no gravity turn and that is no better. I guess I need to turn this thing into more of a pancake - I was hoping to avoid that, but I don't have much choice.

I'm not saying too much thrust to counter gravity, I'm saying too much thrust for the structural integrity to handle. Top of the rocket is simply too heavy, and is putting too much stress on the weak middle part (hundreds of struts aren't going to do you much good if the part below them snaps and disconnects them). I'm also not suggesting to make it a pancake lifter, I'm suggesting removing the structural integrity weakness that is the middle part. If you need the extra fuel it holds, you can just add some half tanks on top of the boosters - that won't make the rocket fatter, its going to make it look and behave more sane.

[bakanando]

In all honesty, out of the pictures alone I can't understand what is happening. Please post more pictures of the ascent stages and comment on each of them for more clarity.

Taking out the big tank and adding the fuel to the side boosters will reduce the length and help with balance. It looks that you decouple some stages before the pics so I can't really say how can it be done without breaking the whole ship.

The rockomax decoupler has some reputation as being extremely wobbly, try changing it for a separator (the blue one) instead.

Disable any SAS modules or control surfaces in the upper stages, the SAS could be countering itself and creating even worse bending.

Use some RCS for control, it can help keep the ship stable without too much applied force in any one section.

Some control surfaces at the bottom stage can help stabilizing and turning too.

[Kelderek]

Thanks for the help all, I ended up doing a significant redesign which ended up being a bit of a pancake, but it was very stable and accomplished what I wanted.