Is engine heat blowing up my fuel tanks?

[celem]

This is a lander i've been fiddling with. The radial tanks are on the decouplers that hold things away from the central piece.

The engine in the centre is an LV-N. It's staged to fire with the decoupler sheathing it. On the radials are 48-7S's, they are not igniting with the nuke.

Problem:

The stage that triggers the decoupler and lights the nuke causes the lander to detonate, in most cases losing 2 or all 3 of the radial tanks, though not always destroying the nuke.

Manually decoupling releases the dead-weight launcher (not shown) fine, but lighting the nuke will blow the lander anyway.

Is it heat from the nuke blowing the 800's? I thought they were held out far enough, they are quite low since I needed to get the legs low enough to land with the nuke.

Edited November 19, 2013 by celem

[SRV Ron]

This is a lander i've been fiddling with. The radial tanks are on the decouplers that hold things away from the central piece.

The engine in the centre is an LV-N. It's staged to fire with the decoupler sheathing it. On the radials are 48-7S's, they are not igniting with the nuke.

Problem:

The stage that triggers the decoupler and lights the nuke causes the lander to detonate, in most cases losing 2 or all 3 of the radial tanks, though not always destroying the nuke.

Manually decoupling releases the dead-weight launcher (not shown) fine, but lighting the nuke will blow the lander anyway.

Is it heat from the nuke blowing the 800's? I thought they were held out far enough, they are quite low since I needed to get the legs low enough to land with the nuke.

The shrouds from the LV-N are causing the explosion when you stage. They are flying straight into the radical fuel tanks. Rotate the decoupler 90* and test again.

[Geschosskopf]

Next time it happens, hit F3 and it will tell you exactly what happened.

But my guess is that the shock of starting the nuke causes the side tanks to lurch, so that their bottom ends slam together and they explode on impact. Or maybe the tank in the lower stage explodes from the nuke's exhaust and takes the side tanks with it. So I suggest 2 things. First, add MOAR struts between the side tanks so they can't flex. Next, put the nuke and the stack decoupler on different stages so you can let the lower stage get clear before the nuke lights. Even if it doesn't explode from fire, the fire can still toss it sideways into a side tank and blow it up that way.

I'm 100% certain, however, that it's not heat build-up from the nuke. That takes a long time to develop, it doesn't happen immediately.

[sal_vager]

SRV Ron is right here, it's the two engine fairing shells that are causing this, they are being fired off from the LV-N, impacting the drop tanks and rebounding, destroying the engine as well.

The fix is to arrange things so the fairing halves do not impact anything, you will see a slight seam between them which will help you when rotating them on your craft with A and D (and shift)

[SRV Ron]

Now that I have a better look at the picture, the fuel tanks are not hooked on to radical decouplers. They have clipped inside them. At the least, staging will fail there as they do with the close in design decoupler. Most likely, that is causing the explosion since those decouplers are now clipped inside the fuel tanks.

If that is a tri arrangement, rotation is not going to solve that issue. You will have to do something different or lose the LV-N for a lander.

Edited November 17, 2013 by SRV Ron

[Designer225]

This is a lander i've been fiddling with. The radial tanks are on the decouplers that hold things away from the central piece.

The engine in the centre is an LV-N. It's staged to fire with the decoupler sheathing it. On the radials are 48-7S's, they are not igniting with the nuke.

Problem:

The stage that triggers the decoupler and lights the nuke causes the lander to detonate, in most cases losing 2 or all 3 of the radial tanks, though not always destroying the nuke.

Manually decoupling releases the dead-weight launcher (not shown) fine, but lighting the nuke will blow the lander anyway.

Is it heat from the nuke blowing the 800's? I thought they were held out far enough, they are quite low since I needed to get the legs low enough to land with the nuke.

Uh oh. Someone forgot the LV-N fairing logic. The LV-N fairing, once separated, acts as debris themselves (unlike most in-game fairings), so if you place something that blocks the direction the fairings separate, in most cases the fairings went off to destroy the things blocking their direction. If this is the case, you have to place the fuel tanks in a place so that when fairings separate, it does not cause damage to ships. The nuke isn't the problem, the placement of the tanks is.

[TechnicalK3rbal]

Nope. It's the shrouds from the Nerva. Try rotating it 90 degrees so they go in different directions.

[SRV Ron]

Consider that he has a triangular setup of saddle tanks with lander legs on each, He may not be able to find a rotation that doesn't affect at least one of the external tanks.

[celem]

Yup, ran another launch to double-check with f3 and it is indeed the nuke shroud. Thanks guys, gonna mean a redesign (it is tri-sym and theres no clear shroud ejection setup that still can put legs below the nuke) I'll probably split the lander/interplanetary stages. Half the issue is that I wanted to land with the damned nuke, guess i'll brush up on rendezvous/docking

edit: also, yes. those radials did turn out to be clipping and wouldnt stage

Edited November 17, 2013 by celem

[TechnicalK3rbal]

Consider that he has a triangular setup of saddle tanks with lander legs on each, He may not be able to find a rotation that doesn't affect at least one of the external tanks.

Oops. Low resolution screen + quick glance = weird optical illusion.

[Smidge204]

Instead of three long tanks, use six shorter tanks for the same amount of fuel but half the height. You can squeeze six of them on with no clipping if using those decouplers. That, plus sliding them up just a little more, should clear the LV-N's shroud completely.

You can still keep the three monopropellant tanks, either don't use symmetry when placing them (easiest) or use two sets of three-way symmetry placing the other tanks and decouplers. If you plan to ditch those side tanks when empty to save mass, then going the second route + asparagus staging is a better option!

Aside: It's not clear why you're using parachutes on the tanks that will be decoupled. Anyplace you land that an LV-N will be useful will not have an atmosphere or much gravity, making the parachutes useless... and anyplace with an atmosphere that the chutes will help, the LV-N will be useless. If the goal is to "salvage" the side tanks by letting them down gently after ditching them, use the smaller chutes to save mass. If the goal is to save the *entire thing* then I wouldn't use decouplers at all.

=Smidge=

[celem]

Thanks for the comment on the chutes Smidge, it's given me something to think about.

As for why it's a weird design mixing nukes with chutes:

It was originally intended as a prototype interplanetary-lander-return for Duna. Designed to go lko-duna surface-kerbin surface as seen in the OP screenshot. Idea was to use the chutes on the radials to help the landing there, then ditch those sub-assemblies completely on ascent from Duna and use the main pod chute for Kerbin return. (It's a career ship. The monoprop tanks are actually goo, and the central body below the pod is 2x materials bays and a 400 tank, it's a sci-ship with no transmitters)

Im aware the nuke alone wont do twr>1 on duna, but was hoping the 48-7S's on the radials would give me enough grunt to get off duna (then drop em empty and nuke back to kerbin). That said I havent actually crunched the numbers, I tend to fire it and see, then plan a glorious rescue if/when it all goes wahoonie-shaped. If I can just get out of Duna's atmosphere and up towards the 40km mark then im hoping the nuke can take over having developed a better TWR

Ultimately I did switch out the 800 drop-tanks for a ring of 400's and moved them up out of the way, using girder segments to then extend the legs past the nuke. I also after several tries managed to get the nuke positioned so that it's shroud doesnt hit the legs and the design now 'works' in that it doesnt suffer crit failures. Got a couple of them circling Kerbin now, going to test them at the next Duna window.

Edit: gonna switch the flag since OP has been answered

Edited November 19, 2013 by celem

[Smidge204]

As for why it's a weird design mixing nukes with chutes:

Ah, that makes a bit more sense. only concern I'd have, then, is the parachutes ripping off. Duna should be okay but if your ship is really heavy...

What I've taken to doing lately is building a two-part vessel with a lander and interplanetary thruster, a-la Apollo, held together with a docking port. Following the less is more principle, the idea is to leave the heavy LV-N and its fuel in orbit so I don't have to get all that mass back up off the planet. The interplanetary stage has a remote control unit, RTGs and RCS system as well. The lander gets back into orbit and the interplanetary stage does the rendezvous... again so I minimize the mass of the lander.

=Smidge=