Producing Extra Niobium By melting Germ Sensor

[goatt]

25kg Niobium + 50kg Plastic = Germ Sensor.

Melting Germ Sensor = 75kg Niobium.

Using rocket exhaust heat.

Some details.

Spoiler

Melting Niobium sensor using liquid steel

 

Melting Niobium sensor using liquid Niobium

 

It can be bigger.

 

[degr]

is it work for all buildings which contain composite material (steam turbine, 20kg gas vent)?

[gabberworld]

10 hours ago, goatt said:

25kg Niobium + 50kg Plastic = Germ Sensor.

Melting Germ Sensor = 75kg Niobium.

Using rocket exhaust heat.

Some details.

  Hide contents

Melting Niobium sensor using liquid steel

 

Melting Niobium sensor using liquid Niobium

 

It can be bigger.

 

nice discovery, now we need only the some automaton tool what makes self the sensors

39 minutes ago, degr said:

is it work for all buildings which contain composite material (steam turbine, 20kg gas vent)?

yes it works for steam turbine as well , but you need more Niobium for there, with gem sensor you need only 25 kg

[gabberworld]

there is one interesting thing about that, you not need build object for get the Niobium. atleast it seems like that but not sure

neh, this is not work, you need build first and then you add hot there, other-way plastic turns into sour gas

[goatt]

54 minutes ago, gabberworld said:

you need build first and then you add hot there, other-way plastic turns into sour gas

Exactly. That's why I used a door to push up the hot liquid to cover the sensor.

[gabberworld]

steam turbine you need 2 layers off doors , one for build and another is for transfer heat

you could use also one layer if you todo mopping but that's player decision

[Fradow]

Props for re-discovering this. It's been known for a while, and is explained on the wiki.

Several people worked on that, most notably @Tetrikitty who spent a lot of time on a germ melter more advanced than this (to minimize heat used and automate things more). I remember that they gave up attaining an "optimal" solution, though I don't remember the exact reason.

[gabberworld]

3 minutes ago, Fradow said:

I remember that they gave up attaining an "optimal" solution, though I don't remember the exact reason.

problem is that its manual build. you may give up this at some-point

[goatt]

2 hours ago, Fradow said:

Props for re-discovering this. It's been known for a while, and is explained on the wiki.

Several people worked on that, most notably @Tetrikitty who spent a lot of time on a germ melter more advanced than this (to minimize heat used and automate things more). I remember that they gave up attaining an "optimal" solution, though I don't remember the exact reason.

I was inspired by a reddit post to melt Niobium. This knowledge is not discovered first time by me for sure, but I'm not sure what do you mean by "more advanced". Did you mean it was more complex? Or did you mean it was melting more advanced material?

If it was more complex, I wouldn't be surprised. My posted build is just a concept which was intentionally simplified for demonstration. If you look closer, you'll discover this build will fail eventually due to the pump overheating. While this can be solved with very simple solution, the original post omitted it for easier reading. But there is a lot of room for modification and improvement.

If you meant it was melting more advanced material, I think Niobium is the metal that requires the highest temperature already, other than thermium which needs higher temperature. But melting thermium sensor is a bad idea. So I think Niobium is the most advanced material.

I posted the original post under the impression that if the build existed before me, it should inspired people to mass produce niobium to replace steel buildings since niobium has better overheating protection. But I haven't seen it during my limited research, so I thought I'd just share this idea. That is: more people should do so using a low effort rocket before finding a Helium Giant which takes more grinding.

[Fradow]

I meant more complex indeed, not with a more advanced material. If I remember right it was about mass-producing it in survival with minimal heat (there was issues with the transfer medium), without killing Dupes or breaking things. I don't remember all the details, but at the time there wasn't an obvious solution that scaled well.

Time will tell if this method catches on

[Tetrikitty]

There are two main considerations when building a germ melter:
1. heat efficiency, ie only heating up the sensor enough to melt it and no further;
2. player time efficiency, ie having a lot of sensors melt at once, in parallel, and automating the process so that the player can put down a lot of sensor blueprints and then ignore the build for a while until it all melts. This is probably the more important resource of the two.

When trying to fulfill the above, sometimes we also run into a third consideration, which is that controlled heating of the germ sensor needs mechanical airlocks which will not melt at high temperatures and having a lot of these to allow for parallel melting will make the build very expensive in steel or wolframite.

I have not tried to design a melter with niobium in particular, but one problem that I had with iron was that it was very viscous and so I needed to melt a lot of iron for it to start flowing over to where it could be extracted. Quite a few tons of iron ended up being "locked" inside the build for it to function, which acts as an additional cost. This is also something that may need to be designed around.

Currently we have this design by Peter Han, here:

The automation automatically drops in the liquid metal when the germ sensor is built, which saves player time. Once the germ sensors are melted, the door at the bottom opens to let it fall into the pool. Liquid metal is pumped from the pool and sent for heating from a heat source then to the vents at the top if needed, or leaves the system to be solidified into metal if not.

It is very expensive in steel and suffers from the issue where liquid metal is locked in the bottom due to viscosity. The steel cost might be cut in half by using horizontal doors with two sensors each, but more complex automation is needed.

 

And this design by me, here:

This is a simpler build with much less automation, where we directly inject heat from the top and cooling from the bottom. Due to the convection mechanic, heat spreads upwards more than downwards and helps to keep the heat differential. I've set the thermo sensor at the top to a bit above the melting point and the one at the bottom to a bit below the melting point.

I don't have an updated screenshot, but the sour gas can be replaced by chlorine for higher efficiency, and a row of tempshift plates along the top allows building two rows of germ sensors, which doubles the amount of sensors that can be melted in parallel. This design wastes some heat because the plastic absorbs a small amount of heat while the germ sensor is being built (less with chlorine), but all the liquid metal is solidified and available for reuse immediately. It also requires less steel doors per sensor.

 

Both of these designs are scaleable sideways to increase the player time efficiency by melting more sensors at once.

[DimaB77]

My version:

[Prince Mandor]

10 hours ago, DimaB77 said:

My version:

Tested this one. It doesn't work. If duplicants bring plastic first, it became sour gas before they start to build. Works only if, by chance, they bring metal first and starts building immediatly after bringing plastic

[DimaB77]

Hmm... Tested many times. It worked. Maybe there's some difference in the details, like there's no metal nearby or the priorities are a little different. I'll test longer, maybe I'll be able to understand the conditions of this behavior.

[Prince Mandor]

Just now, DimaB77 said:

Hmm... Tested many times. It worked. Maybe there's some difference in the details, like there's no metal nearby or the priorities are a little different. I'll test longer, maybe I'll be able to understand the conditions of this behavior.

Catastrofic situation:

• System already hot,
• plastic already brought and placed inside build,
• but some seconds passed before build actually starts.

50 kg plastic heats up to melting temperature really quick.

[DimaB77]

I understand exactly how and why this can happen. I also trust my eyes. And it worked flawlessly in the sandbox. And it took some time from the time I brought the material to the time I built the sensor. Built, of course, with sandbox mode off, duplicates.

[DimaB77]

Yes, you were right. I reworked the scheme a bit, minimizing the movement of dupes, and added a second option.

In any case, the scheme should be considered more of a concept.

[Tranoze]

Wait you guys dont melt tungsten? Why just stop at niobium?
Here a rough blueprint i usually use, but my new world i havent build it yet, so here is how you melt tungsten in large quantity

You can build it as tall as you want to put more germ sensors. The outer 2 layer can be any tile, as long as it meet 3 tile requirements.

[Prince Mandor]

1 hour ago, Tranoze said:

Wait you guys dont melt tungsten? Why just stop at niobium?
Here a rough blueprint i usually use, but my new world i havent build it yet, so here is how you melt tungsten in large quantity

You can build it as tall as you want to put more germ sensors. The outer 2 layer can be any tile, as long as it meet 3 tile requirements.

As I can see it is waterfall, so why it don't melt plastic if dupes bring plastic first and go to toilet second?

BTW -- we don't melt tungsten because abyssalite is abundant and has very little other uses

[goatt]

23 hours ago, DimaB77 said:

Yes, you were right. I reworked the scheme a bit, minimizing the movement of dupes, and added a second option.

In any case, the scheme should be considered more of a concept.

This design has 1 flaw allow schedule to mess it up. If the dupe is carrying plastic above niobium pool when down time just starts, the plastic will drop into the pool and mess up the entire room. This has happened multiple times in my run. That's why my setup has a small mesh time and only let them stand on the mesh time. When plastic drops on the mesh tile, everything is fine.

There is another thing: Hot niobium + plastic = 45C sensor. Cold niobium + plastic = also 45C sensor. This info is useful when you decide how you want to treat your solidified niobium.

 

3 hours ago, Tranoze said:

Wait you guys dont melt tungsten? Why just stop at niobium?

The reason I made this build is because it's very simple to setup. And it requires no advanced materials except for 25k niobium. I think everyone can do it early-mid game. They can set it up when they have rocket, and start to use it when rocket gets first batch of niobium.

3 hours ago, Tranoze said:

When the sensor's material has delivered and the sensor is waiting to be constructed, I think there is a chance the plastic delivered on site will melt into naptha, which then drop into the pool and turn into sour gas. Has it never happened to you?

[Tranoze]

15 hours ago, Prince Mandor said:

As I can see it is waterfall, so why it don't melt plastic if dupes bring plastic first and go to toilet second?

 

14 hours ago, goatt said:

there is a chance the plastic delivered on site will melt into naptha, which then drop into the pool and turn into sour gas. Has it never happened to you?

As i said, Construct the tile to stop the flow, which will stop the water fall, and you start building when water fall is stopped. Deconstruct it after you done building the sensors.

This set up can be done without rocket, and you just need to switch isolated tile with diamond tile and it still work the same.

You can also use constant water fall and build tile to move the water fall 1~2 tile away from it original position.

This setup should be used when you melt natural wolframite tile in it's biome, allow you to have bunch of liquid tungsten to start the water fall.

14 hours ago, goatt said:

This design has 1 flaw allow schedule to mess it up. If the dupe is carrying plastic above niobium pool when down time just starts

You should use auto sweeper to carry plastic and tungsten to construct location, and disable delivery + sweep for the dupe have access to the room to build them. auto sweeper is much more reliable and less time consuming.

[goatt]

6 hours ago, Tranoze said:

As i said, Construct the tile to stop the flow, which will stop the water fall, and you start building when water fall is stopped. Deconstruct it after you done building the sensors.

I see. Sorry I didn't see it. I'm not particularly a fan of pausing waterfall by constructing a tile.

1. The spot you suggested to construct a tile is too far inside to get the construction material out once deconstructed.

2. Depending what tile you construct, if it's insulated tile made of insulation, it's hard to get the insulation out after deconstruction. If it's glass tile made of diamond, the newly constructed glass tile has constant temperature of 45C which will suck tons of precious heat of out the system. I think they are the only 2 possible tiles to withstand heat that can melt tungsten.

Correct me if i'm wrong. I've never done it, so everything is my speculation. I think you can get the construction material out if you delete some insulated tiles to create a diagonal space for auto-sweepers to suck the material out thru that diagonal space.

 

6 hours ago, Tranoze said:

This set up can be done without rocket

Rocket can't melt Tungsten. Rocket temperature is capped at 3200k.

 

6 hours ago, Tranoze said:

You should use auto sweeper to carry plastic and tungsten to construct location, and disable delivery + sweep for the dupe have access to the room to build them. auto sweeper is much more reliable and less time consuming.

This is equally risky. The problem is not how fast the materials are delivered. It's how long delivered materials are sitting there in hot environment. The only way to make sure they don't sit in hot environment for too long is that make sure they do not sit in hot environment at all to avoid all accidents in such a fragile system.

[Tranoze]

29 minutes ago, goatt said:

This is equally risky. The problem is not how fast the materials are delivered. It's how long delivered materials are sitting there in hot environment. The only way to make sure they don't sit in hot environment for too long is that make sure they do not sit in hot environment at all to avoid all accidents in such a fragile system.

They are in vacuum, how can they get hot? Problem with dude delivery is when shift change, they drop what ever they deliver, and it wont happen with sweeper. Using sweeper is as good as building with dupe on top of mess tile.

 

 

31 minutes ago, goatt said:

1. The spot you suggested to construct a tile is too far inside to get the construction material out once deconstructed.

2. Depending what tile you construct, if it's insulated tile made of insulation, it's hard to get the insulation out after deconstruction. If it's glass tile made of diamond, the newly constructed glass tile has constant temperature of 45C which will suck tons of precious heat of out the system. I think they are the only 2 possible tiles to withstand heat that can melt tungsten.

It just a rough blueprint to show out you can use water fall to melt tungsten germ sensor, because no door can withstand tungsten melting point. You can add an auto sweeper on bottom left corner to move any built materials out.

You also should try your set up with thermium, as i think tungsten will drop down as solid.

 

35 minutes ago, goatt said:

Rocket can't melt Tungsten. Rocket temperature is capped at 3200k.

I mean it can be done without rocket's space materials.

[goatt]

16 hours ago, Tranoze said:

They are in vacuum, how can they get hot?

I didn’t know your tile construction /deconstruction to pause waterfall, so I assumed they were always in the hot pool.

But my point still stands. They are EQUALLY risky. Since auto-sweeper only speed things up, but doesn’t provide safety.

When plastic and metal are delivered on site but not yet constructed, the plastic and metal will exchange heat. 
 

if dupe deliver the material, they start to build immediately, and less heat will transfer.

if auto sweeper deliver the material, the construction usually happen not as fast. So in comparison, the auto sweeper only speed up the delivery process, but doesn’t provide extra safety. Maybe a little more risky if not handled properly.

 

17 hours ago, Tranoze said:

It just a rough blueprint to show out you can use water fall to melt tungsten germ sensor, because no door can withstand tungsten melting point.

Why not building pausing tile at the waterfall exit above germ sensor using insulation? It seems a lot more convenient.

17 hours ago, Tranoze said:

You also should try your set up with thermium, as i think tungsten will drop down as solid.

I’m not getting what you mean.

 

17 hours ago, Tranoze said:

I mean it can be done without rocket's space materials.

Yes it can. But it requires lots of work. One main reason I’m advocating my method is its convenience. It uses rocket as a shortcut heat source.

I believe melting tungsten require metal refinery right? That’s lots of set up and work. Correct me if I’m wrong. A little too much for me at least.

[Tranoze]

20 minutes ago, goatt said:

Why not building pausing tile at the waterfall exit above germ sensor using insulation? It seems a lot more convenient.

It have quite high pressure so it might make liquid tungsten go over pressure and move 1 tile away.

 

23 minutes ago, goatt said:

I’m not getting what you mean.

I mean use thermium instead of niobium for germ sensor, as thermium will melt into 95% tungsten and 5% niobium, and it melt at 2700, which can be done easily with rocket, and when it melt, it turn in to solid tungsten and liquid niobium. Since you are not building germ sensor base on any of it's material, and tungsten will go back to molecular forge, the new thermium will start at low temp and plastic will never melt.