Test

The principle of the scheme is based on bug, in which the melting of the germ sensor, made of 2 materials: 25 kg of niobium + 50 kg of plastic, we get 75 kg of niobium. In this way we can swap the less rare plastic, for the rare (space) niobium.

The principle is not new and was previously used to produce, for example, steel. For niobium it was proposed by goatt. Below you can see my slightly simpler version.

We make a scheme as on the first screen. The main tiles are diamond, in between are copper/gold tiles, etc. The rest of the tiles are obsidian. The bottom of the rocket has 3 Tempshift plate of diamond. The middle tile under the rocket can overheat and melt, so it will have to be rebuilt periodically (every 2-3 rocket launches). Make sure that its temperature does not exceed 1700ºC before launching.

After the copper tiles have melted, you can build germ sensors (Fig.2). This can be a liquid, gas, or atmospheric germ sensor. After all the sensors have melted, you must wipe off the copper (Fig.3) otherwise the dripping niobium will solidify the block, which will break the circuit.

Fig.4 shows the finished circuit. The switch is needed so that the pump would not work idle (you can not use hydrosensor - it just melted in the niobium). In 2 Storage bin stored plastic and niobium, so that the dupes do not have to run far. Also, you need to prioritize "Construction" a little higher.

Don't build new sensors if the temperature of the diamond tiles has dropped below 2500ºC, otherwise you risk getting a solid block instead of liquid niobium, which can break the circuit.

The output of the circuit is niobium with a temperature >2000ºC, which can be cooled by this way or simply by running a conveyor with it under the turbines.

Blueprint
Producing niobium