Visco-gel, why? (How does liquid flow work in mixed puddles?)

[myxal]

It's common knowledge that visco gel must be protected from other liquids, otherwise it will happily spread sideways and ruin the intended liquid lock. But why does this happen? What are the mechanics of mixed liquid pools?

 

I want to use zarquan's recent steam extractor on a pair of hot steam vents which are used as a power source (ie the steam is stored indefinitely, and temps at the extractor can exceed visco-gel's phase-up point). Going by my trivial understanding (MinHorizontalFlow ≈ viscosity) I need a trio of liquids stable across 200-500C, in which, if ordered by density, the middle one can flow sideways on every other tick to power the swap-pump, while being replenished by a pump, at max. rate of 10 kg/s. AFAICS there's no such combination of liquids - other than petroleum, everything else that can stand 500C is too viscious, and petroleum is lighter than all of them.

But maybe if I could somehow make naphtha more runny (like the known visco-gel failure), I could use that as the working liquid, with nuclear waste as edge, and petroleum as the vent seal?

Edited March 31 by myxal

[Tigin]

Visco gel stacks vertically the way it does because its speed (roughly translates to maxHorizontalFlow) is actually less than its minHorizontalFlow. I don’t really have a good understanding of what you’re trying to do but if you make your visco gel less than 40kg it might do what you want 

[myxal]

14 hours ago, Tigin said:

Visco gel stacks vertically the way it does because its speed (roughly translates to maxHorizontalFlow) is actually less than its minHorizontalFlow. I don’t really have a good understanding of what you’re trying to do but if you make your visco gel less than 40kg it might do what you want 

I've added a short clip of the phenomenon in the OP, demonstrating with nuclear waste and mercury. (sweepy included to convey game being paused vs running).

What I wanted was some some arrangement of 500C-stable liquids that extracts steam from a vent (the geyser, not the building) like this:

Truth be told, I can do without the liquid seal (visco-gel in the above pic boils at ~480C) and then naptha/LNW + petrol will do just fine, but I still want to understand the mechanics of liquid movement in mixed pools, I might utilise the understanding later elsewhere.

EDIT: I got some weirdness, though not sure if it's usable as a gas extractor:

Spoiler

And 3-liquid version:
 

Spoiler

 

 

 

EDIT2: So far as I've observed:

• A thick liquid will lower its minimum horizontal flow to match adjacent runny liquid - only to the side, not above/below.
• The thick liquid will flow, displacing runny liquid until the masses in neighboring cells are roughly equal.
• The reduced MHF does not affect the opposite side of the thick liquid
• Thick liquid will displace the runny liquid only if the amount of runny liquid is enough to satisfy MHF. A 10g bead of mercury will not allow 75 kg bead of LNW to spill, but a 300g will.
• Heavy thick liquid will never push the last bead of runny liquid off an edge in vacuum, but will do so in presence of gas which can be compressed/displaced:
  Spoiler

   

Edited April 1 by myxal

[Zarquan]

14 hours ago, myxal said:

It's common knowledge that visco gel must be protected from other liquids, otherwise it will happily spread sideways and ruin the intended liquid lock. But why does this happen? What are the mechanics of mixed liquid pools?

 

I want to use zarquan's recent steam extractor on a pair of hot steam vents which are used as a power source (ie the steam is stored indefinitely, and temps at the extractor can exceed visco-gel's phase-up point). Going by my trivial understanding (MinHorizontalFlow ≈ viscosity) I need a trio of liquids stable across 200-500C, in which, if ordered by density, the middle one can flow sideways on every other tick to power the swap-pump, while being replenished by a pump, at max. rate of 10 kg/s. AFAICS there's no such combination of liquids - other than petroleum, everything else that can stand 500C is too viscious, and petroleum is lighter than all of them.

But maybe if I could somehow make naphtha more runny (like the known visco-gel failure), I could use that as the working liquid, with nuclear waste as edge, and petroleum as the vent seal?

That is a weird behavior I've seen before, but never actually experimented with it.  I first saw it being annoying in one of my pipeless counterflow heat exchangers, where I used a liquid naphtha as a blocker and a starting point for beading, which was fine in debug, but annoying to actually build, as the setup process I developed for survival caused any large mass of naphtha to spread out.  I actually tend to use pretty small masses of liquids I want to stay put for that reason.

Spoiler

EDIT: I got some weirdness, though not sure if it's usable as a gas extractor:

  Hide contents

 

And 3-liquid version:
 

  Hide contents

 

 

I would say that if they extract gas reliably without losing gas or losing stability, they are good gas extractors in my eyes.  Is there a reason you doubt them?

12 hours ago, myxal said:

  Hide contents

 

 

Funnily enough, I had that problem on my latest petroleum refinery's flaking boiler.  I needed the compression of an escher waterfall, so I built a zebra-structure by layering oxygen and polluted oxygen so that the crude oil couldn't push the petroleum off the ledge like your nuclear waste pushed your mercury.

Edited April 1 by Zarquan

[myxal]

49 minutes ago, Zarquan said:

I would say that if they extract gas reliably without losing gas or losing stability, they are good gas extractors in my eyes.  Is there a reason you doubt them?

They don't, unfortunately. The liquid-shuffling show keeps happening on its own schedule, and steam extraction is mostly incidental when the eruption event lands on the tick with the vacuum in the right place.

[Zarquan]

18 minutes ago, myxal said:

They don't, unfortunately. The liquid-shuffling show keeps happening on its own schedule, and steam extraction is mostly incidental when the eruption event lands on the tick with the vacuum in the right place.

I see.  That would do it.

13 hours ago, myxal said:

EDIT2: So far as I've observed:

• A thick liquid will lower its minimum horizontal flow to match adjacent runny liquid - only to the side, not above/below.
• The thick liquid will flow, displacing runny liquid until the masses in neighboring cells are roughly equal.
• The reduced MHF does not affect the opposite side of the thick liquid
• Thick liquid will displace the runny liquid only if the amount of runny liquid is enough to satisfy MHF. A 10g bead of mercury will not allow 75 kg bead of LNW to spill, but a 300g will.
• Heavy thick liquid will never push the last bead of runny liquid off an edge in vacuum, but will do so in presence of gas which can be compressed/displaced:

I think that the runny nature of the liquid helps make this happen naturally, but it isn't necessary.  This is because the runny liquid naturally flows away at low masses, leaving two consecutive tiles of the same liquid that the more massive liquid is able to push out of the way.  In the video, the viscogel only stops because it isn't massive enough to push the liquid anymore. Naphtha is by no means runny, but it is easily displaced by the viscogel.  And it works in reverse too, with naphtha displacing viscogel.

Actually, an interesting observation I just made is that this process can naturally create liquid tiles of less than 10 grams from normal liquid flows (which was instantly deleted), which is interesting.  I didn't know that could happen.  It happened when I built my setup from the video and put 30 kg of naphtha on the heavy side and 10 g of viscogel in the other tiles.

 

Edited April 1 by Zarquan

[Zarquan]

In your 2 liquid video, I notice that the critical tile is exposed once every 3 ticks.  A steam vent emits once every 5 ticks.  If steam ever appears, then it should appear once every 15 ticks, or at 1/3rd the rate.  If it doesn't make steam, then there is probably some check happening before the liquid falls and you may need two ticks of non-liquid.

In the 3 liquid video, it appears that you have an 8 tick cycle.  If you could adjust it to get that to a 5 tick cycle, then you could get a 100% uptime structure in theory.  

[myxal]

33 minutes ago, Zarquan said:

In the video, the viscogel only stops because it isn't massive enough to push the liquid anymore. Naphtha is by no means runny, but it is easily displaced by the viscogel.

Huh. So compression into fewer cells is not restricted by the MHF. Good to know. (What I meant by my note is that a single 10g bead, with vacuum on the far side was not pushed away by a bead of thicker liquid which was more massive, but still below the threshold of its own MHF)

[Zarquan]

1 hour ago, myxal said:

Huh. So compression into fewer cells is not restricted by the MHF. Good to know. (What I meant by my note is that a single 10g bead, with vacuum on the far side was not pushed away by a bead of thicker liquid which was more massive, but still below the threshold of its own MHF)

If I had to guess, there is some kind of mid-tick thing happening, where the liquid spreads, then is immediately pushed to the side before the tick ends.

[Tigin]

Got a chance to look at this. This is liquid pressure displacement. To be precise, for each pair of liquids which elements are not the same, if the liquid's mass is greater than the sum of the liquid's mass to be displaced, then 12.5% of the larger mass moves into cell, and the two liquids that were displaced are merged into one. Displacement is required here (so this doesn't occur if one liquid can't be displaced and merged into another same liquid)

Concretely, these naptha cells are 2kg each for a total of 4kg. Any liquid exceeding 4kg will displace and merge the naptha into a single cell of 4kg. (4.01kg of viscogel will split off into 501.3g in the middle)

:

 

Edited April 5 by Tigin

[KittenIsAGeek]

On 3/31/2026 at 10:18 AM, myxal said:

But maybe if I could somehow make naphtha more runny (like the known visco-gel failure), I could use that as the working liquid, with nuclear waste as edge, and petroleum as the vent seal?

My experience with nuclear waste is that it is Always a Problem.  It doesn't matter how I try to use it, it always causes problems.  It does weird things with other liquids, and with pipes, and reservoirs, and pumps.. In fact, at this point, I make it a priority to keep it frozen so I can dump it in a hole and forget about it.

 

But hey, if you find a way to make it useful, I'll at least be interested!

[myxal]

On 4/7/2026 at 5:39 AM, KittenIsAGeek said:

My experience with nuclear waste is that it is Always a Problem. ...  I make it a priority to keep it frozen so I can dump it in a hole and forget about it

Haha, tell me about it...

That said -

1. I'm pretty sure there's no problem with using nuclear waste in "simple" plumbing systems which only contain pipes and bridges (and valves, shutoffs, vents... basically anything that doesn't store the liquid) - hence my fascination with yielding loops in the other thread. LNW is practically the best heat transfer medium for environments between 30°C and 530°C, I'm just about to plumb up a meteor harvesting build with it. (Technically - a rail with alum. atmo suits can shift ~2.5x more heat in a packet for a given temperature delta, but that's horrendously expensive, especially for large loops)
2. As a high-SHC liquid, in atmospheric form (1-deep pool of waste) it's excellent at equalising temperature in wide steam rooms, and doesn't leak heat into imperfect insulated tiles like steam would. I'm making a habit of spilling some near airlock heat/steam injectors, submerging aquatuners, and turbine water return vents in a pool of waste as a way to eliminate temperature swings around thermo sensors controlling said components of a build.
   

As for the thread's topic though, that seems impossible - the thick liquid will only spread at low masses into the thin liquid, and I don't see how that can arranged into a setup where the thick liquid can perform corner swaps without the thin liquid also falling down - at which point I can just use the thin liquid by itself.

I might look more into the discrepancy of single beads being pushed over the edge into atmosphere vs not when pushed "against" vacuum, but all that will be able to accomplish is a once-per-second(-per-vent) displacement, which misses the goal of high-rate displacement with spreading liquid pool.
(todo clip)

Edited April 8 by myxal

[Zarquan]

On 4/6/2026 at 9:39 PM, KittenIsAGeek said:

My experience with nuclear waste is that it is Always a Problem.  It doesn't matter how I try to use it, it always causes problems.  It does weird things with other liquids, and with pipes, and reservoirs, and pumps.. In fact, at this point, I make it a priority to keep it frozen so I can dump it in a hole and forget about it.

 

But hey, if you find a way to make it useful, I'll at least be interested!

 

2 hours ago, myxal said:

Haha, tell me about it...

Why not make that nuclear waste in to a solid tile for radbolts?

What you can do is make a solid nuclear waste tile like you would with aerogel, then put the nozzle of a dispenser or two in to the tile.  Then, load it up with all the debris, then save and load.  Boom, super dense solid nuclear waste tile to put your radbolt generators next to.  I would also use such tiles for GMO plants if I ever cared enough to make GMO plants.

You can also use nuclear waste for a supercharged version of the ethanol chiller that is orders of magnitude more efficient if you have a thermium or niobium aquatuner, but that doesn't require a continual supply of nuclear waste.  This is because nuclear fallout has a much lower condensation point and thermal capacity than liquid nuclear waste has boiling point and thermal capacity.  I used it in my self-sustaining rocket interior.

 

Quote

(Technically - a rail with alum. atmo suits can shift ~2.5x more heat in a packet for a given temperature delta, but that's horrendously expensive, especially for large loops)

You just opened my eyes to a whole new world of possibilities.  Thank you.  I am totally using this for heat transfer now! 

I firmly believe that looping pipes are the best way to move heat as long as you can get the required throughput.  Better than doors because doors have heat bugs.  Though I'll probably use steel jet suits or iron atmosuits instead of aluminum atmosuits because my heat sources are way hotter than 660 C.

Also, you can layer these piped heat transfer mechanisms, so why not both liquid nuclear waste AND atmosuits?  And steam or gaseous super coolant gas loop just for fun.

Edited April 8 by Zarquan

[Tigin]

41 minutes ago, Zarquan said:

 

Why not make that nuclear waste in to a solid tile for radbolts?

What you can do is make a solid nuclear waste tile like you would with aerogel, then put the nozzle of a dispenser or two in to the tile.  Then, load it up with all the debris, then save and load.  Boom, super dense solid nuclear waste tile to put your radbolt generators next to.  I would also use such tiles for GMO plants if I ever cared enough to make GMO plants.

You can also use nuclear waste for a supercharged version of the ethanol chiller that is orders of magnitude more efficient if you have a thermium or niobium aquatuner, but that doesn't require a continual supply of nuclear waste.  This is because nuclear fallout has a much lower condensation point and thermal capacity than liquid nuclear waste has boiling point and thermal capacity.  I used it in my self-sustaining rocket interior.

 

You just opened my eyes to a whole new world of possibilities.  Thank you.  I am totally using this for heat transfer now! 

I firmly believe that looping pipes are the best way to move heat as long as you can get the required throughput.  Better than doors because doors have heat bugs.  Though I'll probably use steel jet suits or iron atmosuits instead of aluminum atmosuits because my heat sources are way hotter than 660 C.

Also, you can layer these piped heat transfer mechanisms, so why not both liquid nuclear waste AND atmosuits?  And steam or gaseous super coolant gas loop just for fun.

Klei took away our precious oakshell molts, which was 500kg of genetic ooze and was by far the best. We still have 100kg seakomb leaves which is better than aluminum atmo suits, but not nearly as conductive.

Edited April 8 by Tigin

[Zarquan]

1 minute ago, Tigin said:

Klei took away our precious oakshell molts, which was 500kg of genetic ooze and was by far the best. We still have 100kg seakomb leaves which is better than aluminum atmo suits, but not nearly as conductive.

I find that genetic ooze is so non-conductive that it isn't particularly effective.  I think you need decent thermal conductivity to make a good heat transfer substance without using a massive amount of space for the ooze to soak up and dump heat.

[Tigin]

1 minute ago, Zarquan said:

I find that genetic ooze is so non-conductive that it isn't particularly effective.  I think you need decent thermal conductivity to make a good heat transfer substance without using a massive amount of space for the ooze to soak up and dump heat.

We don't even have that anymore - seakomb leaves are also (now?) 20kg . Curse you unit mass fixes

[myxal]

2 hours ago, Zarquan said:

Why not make that nuclear waste in to a solid tile for radbolts?

Radbolt production is the exception where I'm mostly OK with super-compressed liquids. I see it done so often I just assume that use of waste needs no mention

A super-compressed solid block... eh, the level of dupe safety I strive for would have me wall it off to prevent dangerous exposure, at which point it's not meaningfully advantageous over super-compressed liquid, methinks.

Now GMO crops is something I might go for one day, with industrial plants. For food I've got more than my colony and Experiment 52B know what to do with.

1 hour ago, Tigin said:

We don't even have that anymore - seakomb leaves are also (now?) 20kg . Curse you unit mass fixes

The dev solid "pump" can put liquid bottles onto rails. Maybe someone can find a glitch to do that without dev mode, and put 20kg supercoolant bottles onto rail?  EDIT: On 2nd thought, that would be hardly better than the alum. suits...

 

2 hours ago, Zarquan said:

Also, you can layer these piped heat transfer mechanisms, so why not both liquid nuclear waste AND atmosuits?  And steam or gaseous super coolant gas loop just for fun.

The reason I avoid conduits is lag anxiety, I guess.

No conduit beats a level-44-strength dupe carrying ~3 tons of magma to the destination per trip.

Edited April 8 by myxal

[Zarquan]

Dupe safety should be assured by denying access.  And remember your duplicant radiation safety handbook.  "If the radiation doesn't hit your feet, it doesn't count!"  Perfectly safe! 

You can ignore those messages about how that nuclear reactor is on the verge of meltdown, they are lies spread by the safety margins conspiracy.

I think we might have different definitions of dupe safety...  

8 hours ago, myxal said:

The reason I avoid conduits is lag anxiety, I guess.

No conduit beats a level-44-strength dupe carrying ~3 tons of magma to the destination per trip.

True, but I'm doing relatively small loops.  Or long one-way runs through atmospheres.  But I have started doing long hydrogen loops for some things, like my tungsten volcanoes and my 5x geotuned iron volcanoes to keep them below temp.

Edited April 9 by Zarquan