Recovering Giant Rocket Cores

[GigaG]

Due to the massive payloads that I am sending skyward on columns of explosives, I need to be able to recover my core rocket stages. These usually make it to orbit with a bit of fuel left. The basic core uses the giant decoupler, 2 S3-14400 tanks, and 1 S3-3600 tank, along with a KS-25x4 cluster. So I simply attach 16 radial parachutes, along with a probe core, and some control systems. Splashdown is vertical at just above 9m/s... then all the fuel tanks cave in due to their low impact tolerance. How do I recover giant rocket cores? I'm thinking of 2 ways to engineer this-

-MOAR CHUTES!

-MOAR CHUTES organized to have the stage land sideways.

I will also make sure to attach batteries and an RTG to ensure a controlled landing.

So which of these options makes more sense? I think the sideways landing does, because when the stage tips over in the water in a vertical landing, the tanks will probably get destroyed. Can anybody link me to somebody who has successfully de-orbited and recovered a giant rocket core?

[WWEdeadman]

With enough chutes, and landing sideways, you should be good. I just have no actual idea how many chutes you'd need.

[quasarrgames]

Or you could simply give it a little more fuel so that it can have a chute-assisted powered landing. i always do that.

[GigaG]

The issue with a chute-assisted powered landing is what happens when the booster tips over.

[NecroBones]

More chutes, landing legs, and possibly even a powered landing (in the last few meters). These can all help, for sure.

[metaphor]

The issue with a chute-assisted powered landing is what happens when the booster tips over.

Maybe add some legs, and make sure to hit land instead of water? I think having enough parachutes for a non-assisted landing will be hard since the parachutes are pretty heavy.

[quasarrgames]

you can also try putting a LOT of reaction wheels on it so that it doesn't tip over, or so you at least have enough time to recover it before it does. I never have less than 100 or 200 units of torque on my lifters for such an occasion.

[Pecan]

+1 more for chute-assisted powered landing (I only use two drogues on a 200t SSTO with 100m/s deltaV left after de-orbit).

NOTE: Splashdown is a really bad way to go. In KSP the impact tolerance is tested for every part when you land in the water. Land on, er, land and it's only the ground-touching parts that are checked. Girders, if you need them, have a very high impact tolerance so you can safely dustdown at a higher speed - although I still aim for 5-10m/s.

[Russoft]

More chutes, landing legs, and possibly even a powered landing (in the last few meters). These can all help, for sure.

http://www.necrobones.net/screenshots/KSP/KSP%202014-07-31%2022-05-32-73.jpg

What NecroBones said. Just stick legs on the bottom. I usually widen the landing base with girders. Then I aim for land rather than water.