Some experiments with nuclear explosives

I have my base underground at 16 height now, but want it to be on the surface. And I'm too lazy to move all of my stuff 50 meters up, so I decided that surface should come to me:) So, it's vital for me to know how I can do that with least amount of uranium and walls. Also, I need to know how many walls,distance should be between me and reactors.

So, I conducted some experiments. I found following:

1. In some small radius (around 5-6 cells) nuke kills, regardless of coverage.
2. Scenery is overall more sturdy than people. Chests, machines, torches, even redstone can survive in blast from 5 cells behind two layers of reinforced stone, but you will die.
3. Blast is very random. If one layer of RS (reinforced stone) can hold one blast, it can be breached by another exactly same one. EDIT: this seems to be result of melting. Reactor can melt virtually anything, including RS. So, if RS wall is too close to reactor, it can melt off some holes on one layer of it and then breach that weak points with explosion.
4. One layer of RS can protect things at 5meters and people at 10 meters.
5. Two RS layers can sometimes be breached at one or two places, but blast will be significantly weakened
6. Reactor blast seems to be a little weaker, but much more uneven than bomb blasts. For example bomb removes exactly one layer of RS, and is very weakened after it, destroying one layer of stone. While reactor leaves most of RI layer intact, but punch some holes and many layers of cobblestone/dirt behind it.
7. Haven't noticed much differences between 18 and 54 uranium cells reactor explosions. They seem exactly same. haven't tested below 18 cells, though.
8. Blast destroys RS at up to about 10-12 meters of air.
9. Uncovered bomb blast is big, but not THAT big. Only about 36 meters radius in air. 9 meters in cobblestone. If screened by one layer of RI, it destroys only one meter of cobblestone behind it
10. Bomb "jumps" when is primed, so if you plastered it to wall, it will fall off.
11. Priming bomb with lever gives much more time before explosion than if it is done by dynamite stick. In first case, you can easily outrun, or even out-walk explosion.
12. "Damage rays" seems to be emitted not from exact center of bomb, but from a sphere of about meter radius. So, it can "go around a corner" a little.
13. After many experiments, 5-chamber reactor full of uranium could not breach two layers of RI. Each time all chambers and all uranium cells and even lever was salvaged. only core itself was lost.
One layer of RI, though was breached even by chamberless reactor, along with many layers of stone behind it.

Many things are still fuzzy for me, though. If you have experiments/information/links about it, I will be happy to hear.

14. When there are few cells in reactor (like, breeder), then explosion is, indeed, much much smaller. Breeder reactor with one uranium explode only doubly as strong as a common TNT (7 meters radius)
15. Plating reduce blast radius A LOT. 23 plates reduced explode of 3-uranium reactor to a couple of meters - less than on dynamite stick. With 12 plates and one uranium it didn't even leave a dent in ground.
16. ENERGY Armor DOES protect from nuclear blast. On normal, quantum armor protected from point blank explode of 27 uranum completely. Nano helps too, especially agains MK 1/2 ones. Metal armor protects much less, died in diamond set.