Uranium enrichment is weird

[Ubertwink]

Currently, you have to grab the output items grom the centrifuge and then put them back in.
I think the products should be put back into ingredients slot, unless there is enough there to start a new enrichment cycle. In that case, the remainder is left in the output slot.

[Nemoricus]

I'm of two minds on this one. On the one hand, the need to feed back the products of the enrichment process to keep it running is an excellent reason to make use of circuit network logic, and introduces an interesting level of complexity.

On the other hand, this makes nuclear power much more difficult to use, since you have to balance enrichment with mining already and this adds a layer of complexity that people may find annoying to deal with. Since the net recipe is 3 U238 to 1 U235, it would be much easier to simply require 40 U235 and 2 U35 as part of the recipe to make an enrichment centrifuge to begin with. However, that would involve adding another entity.

Perhaps there could be a new type of recipe that consumes a fixed amount of resources when the first cycle runs, and then only requires a small amount of resources for each additional cycle?

[Ranakastrasz]

Could add a "catalyst" list to recipes, which must be present, but do not get consumed. Anything from molds to actual catalysts would use it.

[DaveMcW]

This accurately simulates the inefficiency of centrifuges. You need to move a large amount of low-grade Uranium-235 to obtain a small amount of pure Uranium-235.

In real life there are hundreds of enrichment steps as your sample increases from 0.7% to 100% Uranium-235. You should be happy we don't have to track all those intermediate products.

[Neemys]

Currently, you have to grab the output items grom the centrifuge and then put them back in.
I think the products should be put back into ingredients slot, unless there is enough there to start a new enrichment cycle. In that case, the remainder is left in the output slot.

What if people just wanted 1 craft and take all result elsewhere ? By making the step with inserter, people are free to put back in everything or part of it as they like.

[Nemoricus]

More seriously, the current version of enrichment interacts oddly with productivity modules, as can be seen here: viewtopic.php?f=7&t=44635

Since this is likely unintended behavior, I think that a different recipe model is needed. The catalyst model is perhaps the simplest way to do this, at least conceptually.

Currently, you have to grab the output items grom the centrifuge and then put them back in.
I think the products should be put back into ingredients slot, unless there is enough there to start a new enrichment cycle. In that case, the remainder is left in the output slot.

What if people just wanted 1 craft and take all result elsewhere ? By making the step with inserter, people are free to put back in everything or part of it as they like.

This is Factorio. Why would you ever only want one craft?

[DaveMcW]

The empty oil barrel recipe works oddly with productivity modules too. The solution is to disable productivity modules.

[Ranakastrasz]

The empty oil barrel recipe works oddly with productivity modules too. The solution is to disable productivity modules.

There is a reason why productivity modules are disabled for emptying oil barrels.

Unless, I suppose they forgot to add that part in. Thing is, its a white list allowing Productivity modules, so by default it should have been disabled.


A catalyst using recipe would ignore the catalyst when it comes to productivity modules, so that issue would certainly go away.

[Ubertwink]

What if people just wanted 1 craft and take all result elsewhere ? By making the step with inserter, people are free to put back in everything or part of it as they like.

If you want to do one enrichment cycle, you can as well track it by yourself.
Even if you miss it, all you lose is a few lumps of U-238, to make some extra U-235. That's not even a loss tbh.

[Neemys]

What if people just wanted 1 craft and take all result elsewhere ? By making the step with inserter, people are free to put back in everything or part of it as they like.

If you want to do one enrichment cycle, you can as well track it by yourself.
Even if you miss it, all you lose is a few lumps of U-238, to make some extra U-235. That's not even a loss tbh.

That one was an example, my point being not everyone need all result being put in the ingredient slot everytime. So it's not a good idea to force that. Remember that you can't take from ingredient slot with inserter.

[Ranakastrasz]

Catalyst ingredients would stay in the input side, while the rest would go to the output.

[AileTheAlien]

Personally, I feel like how the enrichment acts fits into the same mental model as burner miners on a coal patch. That is to say, that they produce a resource which goes onto a conveyor belt, which can then be fed back into them by an inserter. So, my vote is to keep the inserter-ness of enrichment as it is right now. :)

As for the productivity modules, I think that's a pretty easy decision too - they should be disabled for the enrichment recipe (and maybe some others too), just like oil barrels. Basically, any recipe or set of recipes, which would allow for infinite resources. :)

[impetus maximus]

productivity modules are now disabled for uranium enrichment in 0.15.2

i'm working on using combinators to limit the 235 to 40 being fed back into the centrifuge.
it's not going well.

[Marqee]

Took a while to reach the 41 needed to start the process, but then nearly all the U-238 was enriched.
i had to cut a piece of belt to make it stack again.

[Tekky]

i'm working on using combinators to limit the 235 to 40 being fed back into the centrifuge.
it's not going well.

It didn't go well with me, either. By the time I had finally managed to create a working combinator setup which does this, I didn't need it anymore, because most of my uranium 238 had already been upgraded to 235.

[impetus maximus]

person3triple0 AKA Adorin (youtube) came up with a clever 0 combinator circuit.
i tweaked the layout a bit, and controlled the 238U going back in, but all credit for the clever bits of the closed loop of 235U goes to person3triple0.

feed this 40 235U directly into the centrifuge (and 238U of course), it will process it, dump 41 235U into a crate, take 1 out, load 40 back in.
it's a thing of beauty. it will keep spitting out 235U as long as you feed it 238U.

Adorin.improved.IV.png (202.3 KiB) Viewed 6359 times

blueprint string

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

i was trying to chain them, but after realizing it can supply enough to stress test my 463MW of reactor power @100% why bother?

[undarl]

person3triple0 AKA Adorin (youtube) came up with a clever 0 combinator circuit.
i tweaked the layout a bit, and controlled the 238U going back in, but all credit for the clever bits of the closed loop of 235U goes to person3triple0.

This is a thing of beauty! Thanks to Adorin for creating it and to you for sharing it. I've been trying to wrap my head around that very problem for several days now.

[impetus maximus]

This is a thing of beauty! Thanks to Adorin for creating it and to you for sharing it. I've been trying to wrap my head around that very problem for several days now.

my pleasure. i got such a stress headache trying all day to do that. really wanted to come up with something on my own, but oh well.

[Kelderek]

person3triple0 AKA Adorin (youtube) came up with a clever 0 combinator circuit.
i tweaked the layout a bit, and controlled the 238U going back in, but all credit for the clever bits of the closed loop of 235U goes to person3triple0.

feed this 40 235U directly into the centrifuge (and 238U of course), it will process it, dump 41 235U into a crate, take 1 out, load 40 back in.
it's a thing of beauty. it will keep spitting out 235U as long as you feed it 238U.

Adorin.improved.IV.png

blueprint string

0eNrNV9uOmzAQ/Rc/gwQOuSxS+yPVCjkwSawFG9km3TTi3zuGhBDAu9Ct1L4EYXvOzDlzwbmSfV5BqbgwJL4SnkqhSfzjSjQ/CpbbNXMpgcSEGyiIRwQr7FsKwih+qI5Aao9wkcE7icPa+9TQKCZ0KZXx95CbnjGtXz2CqNxwaENoXi6JqIo9KER3YXiklBrNpLBeEcrHoxd8rBA94wrSdm/jEaRnlMyTPZzYmUtlDVKu0oqbBATb55BkXNsniQ8s1+B12wpYlpyYyBILgrFhkEZVvRP39fZoITNoJLFORRuDtv5C+6Mg63Pk+LapX+vaKjjgTTveB6aNz4UGZXDDSZtOoaweKDxH8w9wgjvMk3qRd7Nsk3NPeQdbYVJ4Vfh0tSPIZBxBNDt/jvRNYK5nYwb/UUlMENnMJkL/hAjuZR3IgSttklGfnrkyFa50kbQnfDiDupgTF0fSFrM2zE6LwL4UJVPMWGfkW7PtUO5ZlinhBtpOKRfMb6bQ0Uzbjt1PiZjCT0+gp0bIqpV5Mlm7h0SGpW8ftWQLE0yhvHQoBWS2cyBHYoqnfilRsjEWdWOFwdRkdjSBlRELhzdZf5NnpuDdB4GeTwUa+6WSKWhNpvyEA+qfDpNb1H74XK909jRZT0+TkC7LpGsQHxWAGFXPquexWXAN5/AxVz+tAlp/vUmfP6ZPKvV7Mxo253dSLxKAWr4ekdj7imeQNNnGcH7B7TM/1CGa3xO0S8jfHWAztQnHc2uRNGtXKayXVeRqkdfI5XUzpwBp5/GfSB58UXLqIr9deiccSECnQHcLLxVzMF8WXn5mYNJg4Z1hhIkX7iaBce9PgEew5XVbH1FEo22wCaJdXf8Gbm8pDg==

i was trying to chain them, but after realizing it can supply enough to stress test my 463MW of reactor power @100% why bother?

Wow, this is weird, I came up with almost that exact same design for my base yesterday. Same idea with 3 boxes, take 1 out and move the other 40, I just ran my belts differently. I did chain a bunch together, about 7 centrifuges I think and I am now sitting on about 2200 units of U235 now That is in addition to 8 centrifuges processing the raw ore. It is worth chaining them to get this kind of stockpile, not for the reactors, but for atomic bombs instead. I'm researching that right now and they will take 30 units each. I'm a strong believer in overkill, lol.

[impetus maximus]

oh, nukes right. lol yeah i guess if you're trigger happy more than one going would work.
does your chain limit to 40 going in each machine, or do the internal buffers fill up?