How do you get radbolts out of a research reactor?

[oosyrag]

Probably the only place I've actually needed radbolt joint plates were to get radbolts out of a sealed research reactor chamber... but the chamber is easily 300c+ and the plates are made out of plastic, which melt at 160c?

I guess I could built it in space and waste all the steam...

[Nogard78]

There are probably other ways that people know about, but plastium will defonietly work here. Yes it requires thermium and brackwax but its easy to implement.

You can insted build radbolt generators outside of the chamber, but then you loose quite a lot of efficiency. 

Another way you can do this is to just send them through a liquid lock that is somwhere down and its safer, or create double liquid lock somwhere near the top so you can get them out there. 

[blakemw]

I consider there are three basic techniques, which I've mocked up in sandbox:

• Single tile liquid lock (top-left). These tend to be very safe when "sandwiched" between two solid tiles because nothing can get in there to interfere. Naphtha is the best choice by far. You can use a double liquid lock if you want to prevent heat leakage, though truth be told if you have a Nuclear Reactor you can easily brute force the cooling, the heat leakage through a single Naphtha tile is not very great, it's like 10 kDTU/s, pretty much nothing. These are easy to add after-the-fact even when the chamber is full of hot steam by pre-placing the liquid then deconstructing tiles, like in the position I have mine you can literally just place it on top then deconstruct the insulated tiles so it falls down into position.
• Full liquid lock (bottom-right). You can simply "zig-zag" with Radbolt Reflectors. You can see if you can get away with diagonals, the collision rules for diagonal are understood but a bit arcane. Either way, Radbolt Reflectors are efficient the penalty you pay for bouncing the radbolts around is very small. As with the single tile liquid lock you may choose to double-lock, or just ignore the roughly 10 kDTU/s which leaks through (for Naphtha, it'll be significantly more for Crude or something).
• Actively cooled Radbolt Joint Plate. Plastic is a low conductivity solid, so the steam doesn't exchange heat with it that quickly. If you embed like an Aluminium Radiant Pipe in the plastic tile it'll pull plenty of heat out of it and thus keep it well below the melting point. You likely already have active cooling in the area. You can expect to pay about 100 kDTU/s for a basic Radbolt Joint Plate, or roughly 1/6th of an Aquatuner or roughly 200 W, which is completely fine for brute-forcing. However if you immerse the "dongle" in Naphtha instead of Steam heat exchange will be slashed to about 1/20th of that with Steam, here the lower mass of Naphtha the better. The main reason this works so well is that Gas:Solid heat exchange has a 25x multiplier, while Liquid:Solid has no multiplier, and both Steam and Naphtha have about the same Thermal Conductivity, so you're getting rid of a 25x multiplier and bringing the heat leakage in line with the liquid lock solutions. However the Radbolt Joint Plate is clearly inferior to a single tile liquid lock sandwiched in the wall and the effort to make it as good is basically utilizing a single tile liquid lock anyway. So honestly if you want the best solution just use a single tile liquid lock of Naphtha.

Edited February 14 by blakemw

[oosyrag]

Thanks for the detailed solutions!

[Farsight]

Radbolts will pass through a diagonal corner if the reflector shoots diagonally upward from a tile away.

I don't have an image handy, but you can probably find one with a search, or just play around with 'em to see.

I always shot the bolts out of my reactor room without issues.

[pnambic]

1 hour ago, Farsight said:

Radbolts will pass through a diagonal corner if the reflector shoots diagonally upward from a tile away.

I don't have an image handy, but you can probably find one with a search, or just play around with 'em to see.

The authoritative guide to all things corner is afaik Corners: Do they exist? by sakura_sk; It has a section on getting radbolts through corners, too.