Processing u235 automatically

[Nailfoot]

This is driving me crazy! I have come up with an automatic processing build using robots, but its the best I can do.

http://steamcommunity.com/sharedfiles/filedetails/?id=913872164

Each blue chest requests 40 u235 for the catalyst. The inserter I have highlighted works if there is less then 40 u235 in the logic system. It then loads 40 into the yellow chest so there is a bit of a backup for the catalyst.

This would work ok if we could shrink the roboport's logic network down so that each centrifuge system was its own network. It will also work if you just plan to space them out. However, once they start intermingling with other centrifuge arrays (or, put another way, if you have more than a few centrifuges in one array) you run the risk of the bots spreading the u235 "too thin" so that none of the centrifuges will work at all.

This solution requires that you babysit it and check back on it far too often for my tastes. Of course, once you have hundreds or thousands of u235 in the system, it will work flawlessly as long as you are careful about how much you syphon off for the reactors.

Someone out there HAS to have a better idea!

[Varenvel The Festive Dinosaur]

when i get to that stage , ill propobly have something , right now im using this for refining , it may give you some idea's

http://steamcommunity.com/sharedfiles/filedetails/?id=913885142

[Varenvel The Festive Dinosaur]

Have you not thout that instead of "it being instantly loaded into the logic port it first goes into "buffor chest" who feeds the centrifuge , and if buffor(whos requester chest for uranium 238 and does not request uranium 235) already has enough , exes is removed using smart inserter to remove exes 235(you may need to use number overide of stack inserter to not remove too much)

That way 235 is in closed loop system , and exes is removed into logistic network

you have to hand feed initial amount of 235 in che chest , and then it should stay there forever


and excuse my terrible english , its late and not slept too well today

[Nailfoot]

Originally posted by Varenvel:

Have you not thout that instead of "it being instantly loaded into the logic port it first goes into "buffor chest" who feeds the centrifuge , and if buffor(whos requester chest for uranium 238 and does not request uranium 235) already has enough , exes is removed using smart inserter to remove exes 235(you may need to use number overide of stack inserter to not remove too much)

That way 235 is in closed loop system , and exes is removed into logistic network

you have to hand feed initial amount of 235 in che chest , and then it should stay there forever


and excuse my terrible english , its late and not slept too well today

That's what my system does, except all of the centrifuges share a common loop and the extra is syphoned off.

I am producing 430 U-235 a minute with this setup. It is consuming way more U-238 that I am making, luckily I had backstocked over 10K of the stuff. I am working on increasing its production and finishing the research to reclaim the used uranium fuel cell.

http://steamcommunity.com/sharedfiles/filedetails/?id=913899106

[impetus_maximus]

i'm working on a combinator setup that will feed 40 U-235 into the centrifuge, feed 40 of the 41 produced back in, then push the excess into the next centrifuge and so on.

i know about memory cells (arithmetic combinator [*each +0 each] self wired), and counting +/- of items via pulse. even S-R latches for steam power, and oil management. but GD, all i managed to do so far with this is give myself a stress head ache. no matter what i've tried i get 41 back into the centrifuge.

[Nailfoot]

Originally posted by impetus maximus:

i'm working on a combinator setup that will feed 40 U-235 into the centrifuge, feed 40 of the 41 produced back in, then push the excess into the next centrifuge and so on.

i know about memory cells (arithmetic combinator [*each +0 each] self wired), and counting +/- of items via pulse. even S-R latches for steam power, and oil management. but GD, all i managed to do so far with this is give myself a stress head ache. no matter what i've tried i get 41 back into the centrifuge.

I can't do it either. If we could only count how many items an inserter has touched.

However, once you have robots and a few hundred U-235 to feed the centrifuges, the process to automate them is trivial. I would like to do it without robots but I just cannot.

[impetus_maximus]

Originally posted by Nailfoot:

I can't do it either. If we could only count how many items an inserter has touched.

However, once you have robots and a few hundred U-235 to feed the centrifuges, the process to automate them is trivial. I would like to do it without robots but I just cannot.

wire an inserter set to read/pulse to a arithmetic combinator. wire the combinator output back to the input of the combinator. set the cominator to *each +0 *each. the combinator will keep track of how many items the inserter has passed.

i really want to do this without bots.

[Nailfoot]

Without bots would be nice. Otherwise you will need a bot network every single place you want to refine uranium.

Honestly, though, this stuff refines so quickly and is so slow to "burn", I cannot see us needing huge arrays of them like we did with solar. Even once you are into the megabase arena, I think these will be easy to keep running.

Here's another attempt, just trying to stop the bots from looping forever.

http://steamcommunity.com/sharedfiles/filedetails/?id=913916637

[GMC]

Originally posted by impetus maximus:

i'm working on a combinator setup that will feed 40 U-235 into the centrifuge, feed 40 of the 41 produced back in, then push the excess into the next centrifuge and so on.

This is roughly what I'm doing (except that the excess just gets pushed out to use in fuel cells; multiple centrifuges are in parallel rather than series). I posted a blueprint string in this thread.

The basic principle is a standard feedback-based counter: an arithmetic combinator which adds zero to (the sum of) its inputs, with its output being one of the inputs. The other inputs are the contents of a belt cell (pulse mode), and the hand contents of an inserter (again, pulse mode) multipled by -41.

So each item on the belt adds one, each item removed by the inserter subtracts 41. The inserter is enabled when the running total exceeds 20.

The thing that tripped me up initially is that the inserter is both a producer (hand contents are written to the network) and a consumer (running total is used to enable/disable the inserter). The connections need to be made with different wire colours to avoid an extra feedback loop.

[GMC]

Originally posted by Nailfoot:

Honestly, though, this stuff refines so quickly and is so slow to "burn", I cannot see us needing huge arrays of them like we did with solar.

One centrifuge with a couple of speed-3 modules will produce a U-235 every 25 seconds (plus a few seconds to feed the catalyst back in). That's more than enough to retire my 200-engine coal power plant and use the coal for liquefaction.

It's just the time taken to get the initial 40x U-235 to use as catalyst, at 0.5% (or worse) . That's 8000 crafts and 80k ore. Which is why I went for a circuit network solution, as you're up and running the moment you get the 40th U-235.

From that perspective, it's worth putting down miners on every available patch of uranium, carrying barrels of sulphuric acid out to the patches and emptying the chests into the car. Use all of your centrifuges (including the ones earkmarked for enrichment and reprocessing) to refine ore. Once you have 40x U-235, you can rip everything up and build a stable, automated production line served from a single patch (each fuel cell is basically 22x U-238 = 220 ore).

[impetus_maximus]

GMC did you end up getting it working? to subract with an inserter connect it to a arithmetic combinator set to [*each x -1] between the inserter and the memory cell.

here[forums.factorio.com] is a solution that uses 3 chests with just some inserter setting (stack size) and a couple of wires.

[edit] i tweaked it again. my U-238 was jamming. new blueprint string
http://steamcommunity.com/sharedfiles/filedetails/?id=913981078

i'm going to work on getting a few of them in a row automated.

blueprint string [UPDATED] removed super speed module (creative mode mod)

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

[MisterSpock]

You need less than 10 ore to get a fuel cell with productivity and enrichment. Only the starting is harsh.

[GMC]

Originally posted by GMC:

each fuel cell is basically 22x U-238 = 220 ore).

Wait, that recipe is for 10 fuel cells, so it's only 2.2x U-238 (22 ore) per cell. Or 1.6x (16 ore) if you account for reprocessing (3x U-238 for 5 used cells = 0.6x per cell). You could get it even lower with productivity modules, but it's ridiculously cheap as it is.

Originally posted by impetus maximus:

GMC did you end up getting it working?

Yes. The blueprint in the thread I linked has been running all day.

[MisterSpock]

Originally posted by GMC:

Originally posted by GMC:

each fuel cell is basically 22x U-238 = 220 ore).

Wait, that recipe is for 10 fuel cells, so it's only 2.2x U-238 (22 ore) per cell. Or 1.6x (16 ore) if you account for reprocessing (3x U-238 for 5 used cells = 0.6x per cell). You could get it even lower with productivity modules, but it's ridiculously cheap as it is.

Originally posted by impetus maximus:

GMC did you end up getting it working?

Yes. The blueprint in the thread I linked has been running all day.

a fuel cell costs 1,9 U238 and 0,1 U235. You get something lake 30% back. Its actually 40% because there are productivity modules in the assembler.

Because you have a infinite stack of that effect (its a geometric series) this multiply your effective number of fuel by 1,6

[Stormshado]

My head exploded while trying to read this thread. Sticking with solar (for now) :P

[Nailfoot]

Also, you can refine the used fuel cells BACK into U-238, so its really even cheaper than that by a small amount.