The Liquid Bead Gas Pump (Bead Pump), or The Reverse Sprengel Pump

[mathmanican]

Using little droplets (beads) of liquid, we can create a percentage based gas pump that moves gases upwards. The principle is very similar to a real life Sprengel pump, just in reverse (thanks @Zarquan).  Maybe we should make the name "The reverse Sprengel pump" (comment below on your opinion).

Edit: It's very similar to a "Bubble Pump" which focuses on collecting air below liquid, but instead this pump focuses on forming little "beads" of liquid above a body of air. Both pumps are "air lift" pumps. 

This post attempts to explain (1) exactly how this principle works, (2) what it's uses are, (3) how to reproduce the effect reliably, (4) how to control the pump, and (5) some limitations. Of course, I'll aim for extreme case uses that massively exploit this mechanic - as always.

First, some history and credits:

Spoiler

• The recent discussion began with @tzionut's post "Pumping gases without a pump."
• @tzionut's application, the SSST, created more interest.  Great marketing, you got me hooked.
• Thanks to @blash365, I was able to find some videos by @socooo where he spent time optimizing this principle for the steam turbine. 
• Thanks to all of you who added to the recent discussion in these threads (@OxCD, @Angpaur, @Yunru, @nakomaru, @Craigjw, @0xFADE, @badgamer123, and those mentioned above). 

More thanks to @socooo for helping locate the rest of his earlier posts on the topic. 

 

What is a Liquid Bead Gas Pump?

Use liquid mechanics, one-element-per tile, and tile-swapping rules, to cause little beads of liquid to force gas upwards. That's its.  These are central mechanics to the game, so probably won't change.  I won't call it a bug, rather I'll call it an exploit.

There are probably ways to fix this, but I don't think the devs should focus on this.  There are much bigger things to fry. 

Why should I care, or How Can I Use This?

The maximum pumping speed has no upper bound (aside from game restrictions on maximum mass, if there are such things). The effective use requires a single gas (multiple gasses can cause problems). The pump can reach vacuum conditions quite quickly. So the uses require single gas settings where speed and/or high pressure are important.  Sounds like the steam turbine and gas storage centers are main uses. 

The fixed 10kg/s of the steam turbine is no match for the percentage based speed of the bead pump. If the bead pump is not fast enough to keep your turbine(s) running (yes - multiple), then add more steam. Eventually, the pump's speed will overpower any number of turbines (crazy!).  While you can provide the required pressure difference for essentially ANY number of turbines with ONE single pump, unfortunately this won't keep them all hot enough (but we have other fun exploits for that  - mwhahaha). Here's a few examples

Spoiler

This one is mine. You can add lots of turbines to the left, without adding any steam.  The pumps drain the top region of each turbine, dumping the contents into the bottom region of the turbine above. Adding heat is the only issue. 

This is @socooo's build.

Here is @tzionut's SSST. 

 The bead pump can empty a room to vacuum. High pressure storage is quite simple. Should the "liquid over vent" trick ever be fixed, this provides an alternative to unlimited gas storage.   

Spoiler

Personally, I think it looks pretty as well. The metal tiles are not needed, rather added for aesthetics. 

Other uses - I'll add them here as people find them.

Thanks to @Zarquan, we have a simple pump that can suck up 5kg/s CO2 from meteorite impacts, and collect it all in a tiny little room. Quite insane.

Spoiler

How do I form the beads reliably?

The beads automatically form when a liquid, located above another tile of liquid, is forced to move sideways, and then sits atop a layer of gas. This is most easily forced by having the lower tile of liquid being a different heavier liquid, however with a more tricky setup can be accomplished with a single liquid (see this for an example).

Back to the easier setup. Here's a simple setup that guarantees beads, and one that fails to form beads - the difference is the two walls left of the liquids. The second forms solid liquid trails (also fun, but that's another post for another day).

 

Now the explanation.

1. Place a heavy liquid (such as crude) in one spot on top of some solid tile (undug, air, regular, normal, door, etc.). Note that this heavy liquid will naturally drain all liquid away except for a small amount. We could call this the "natural minimum" or "overflow" mass.  I like to think of it as a measure of the liquid's viscosity - the liquid tends to bunch together in certain minimal size chunks. You can get chunks that are less than this, because of other external effects, but those aren't "natural." For both crude oil and petroleum, this "minimum" mass is somewhere between 350g and 400g. For water and polluted water, this value is between 35 and 40 (a factor of 10 less).
2. Put a tile on one side of this heavy liquid as well as a tile above that (these are right of the crude and petro above), so things will drain on only one side (this is important).  If things drain on both sides, eventually ONI liquid mechanics will displace the heavy liquid and your pump will fail. See the spoiler below.
3. Put tiles on the other side, leaving a gap for the beads to form. Both tiles are needed. Leave out the top tile, and the beads will disintegrate. Leave out the bottom tile and you might end up with a long liquid trail.  Could be fun to use for a new airlock pre-viscogel.  
4. Now add a constant flow of lighter liquid to the tile above this blob (petroleum is lighter than crude - water is lighter than polluted water). The top liquid accumulates mass and then every so often will push the excess liquid sideways. This creates the bead.
5. Increasing the flow rate causes the beads to form faster, up to a point.  This leads to the next section, "How do I control the beads?". 

Spoiler

Disaster struck because I wanted to have liquid fall down both sides. 

The opposing tiles are important. You don't want either of these (trail in first, regular water teleportation in second).

How do I control the beads (i.e. start/stop the gas pump)?

Note, the "overflow" mass discussed in the previous section is important for controlling the speed of the pump. You can quickly estimate this amount by dropping liquid over a single tile and letting the excess flow away.  The amount that remains on the tile is this "overflow" or "natural minimum" or "viscosity rating". This natural minimum value depends on the type of liquid, and appears to be independent of any other factor like temperature, surrounding liquids/gases, etc.

Spoiler

I'll add a table of minimal mass values for all the liquid types eventually.  For now, here's a rough idea.

• Crude/Petro - 350g to 400g
• Water/Pollluted Water - 35g to 40g
• Phosphorous - 6kg - 7kg
• Liquid metals - Above 50kg - not practical for pumping without an esher waterfall or a door pump. 

As a side note, these values above also appear to affect the speed of flow of liquids as they try to equalize pressure at the same depth, which affects the escher waterfall.

To control the speed at which the beads form, you just change the speed at which you drop liquid on the top tile. Valves are perfect for this, though an escher waterfall works too if you need more than 10kg/s to make things work (for example you want to work with liquid metals, magma, or whatever).  

• To stop the pump, set the valve to zero. 
• Anything above zero, provided it's large enough to not disappear when leaving the vent, will start to form beads. 
• You can watch patterns form based on valve settings.  Pause the game and look for a pattern such as "bead, gas, gas, gas, bead, gas, gas, bead, gas, gas, gas,"  If you ever see "gas, gas" in your pattern, then the pump speed can be increased.  
• Once you hit a little above twice the natural minimum mass (or overflow mass), you'll see the pattern becomes perfectly alternating as "bead, gas, bead, gas, ...". At this point, you've hit maximum speed. The room you are pumping from will reach vacuum, and stay vacuum (though don't 100% trust it will say vacuum, yet). For now, grab the number from the spoiler above, double or triple it, and set the valve to that value.  For petro, I used 800 for a while, but once saw gas slip.  At 900g, I haven't seen gas slip backwards at all. More testing needed.

To control the height of the pump, there are several options. 

• Have your pump stop on the floor where you collect the liquid. Simple and easy to setup. Just put the liquid pump (supplying the beads) in the pool on the floor, and let it run. Done. Or let an escher waterfall collect the liquid at high pressure, and use liquid pressure mechanics to send it back up. 

Spoiler

Whether you run the beads down one wall, both, or no wal, it does not matter.  Here I collect the beads at the bottom.  One picture has a room emptied to vacuum, so you can't even see the petroleum.

• Have a gap in solid tiles, on either side of the bead column. What?  When a bead forms, it constantly looks on both sides for reasons to unform. If there is a solid non mesh tile next to the bead when it forms, then you can continue building the wall as far down as you want, and watch the bead follow the wall. When you stop building the wall, the bead doesn't disappear. It only disappears when you build another solid tile, at which point the bead disappears and transports to the bottom of the fall in the usual, liquid teleportation, manner.

Spoiler

The beads appear on the walls, wall, and remain when the walls are removed.  However, the tile of water, above and to the left of the mech door, alternates between water and vacuum, every tick of the game.  I'm still trying to use this vacuum tile to make a matter converter, but haven't figured it out yet (as the above tile will always be liquid. 

If the pump is not vacuuming out a room fast enough, and you want the pump to increase speed in a different part of the room, just build an airflow tile (replace the mech door with airflow), and then the pump will focus on the area with highest concentration.  

• Let the liquid bead hit a floor, and then teleport for a while without being a bead, and then reform the bead. Do this several times along the downward descent, and you can create multiple pumps, all fed from a single bead setup.  This allows you to stack turbines and drain the top region of the turbine to the next turbine's bottom region, keeping the pressure difference quite large. It also allows for infinite gas in a fairly compact environment (the example below can be compacted even more).

Spoiler

The steam turbine build above keeps the pressure differences quite extreme.  I could probably add 3 or 4 more turbines to the left of each given turbine, without any issue. Supplying enough heat would be the only issue (delete the tricked neutronium block and supply heat with pipes, or add a row to each level and use aquatuners). 

What are the limitations?

This list is in the works. Here are a few.

• If you try to make the beads form on both sides simultaneously, it's possible for the bottom liquid to get displaced, at which point the whole process stops. Only use one side, and use a tile to prevent the bottom liquid from moving. 
• Avoid multiple gases. Consistently, you can only guarantee the pump works with a single gas. The pump stops working once you get a blob of gas trapped where the beads should form (and then you have an escher waterfall). Its fun to watch water fall at different speeds with some falling in bead form, and some falling in teleportation mode. The teleported water arrives sooner.  I'm sure we can find a way to exploit this too, though I'm not sure right now how (would having liquid arrive in 20kg/0kg/20kg/0kg patterns affect anything in game?). 
• If you use walls on one side, or both, remember that if you have any gap at all in the wall tiles along the downward decent, then the bead will stop being a bead once you start the wall up again. Gaps in wall tiles are fine, on either or both sides, but once the wall starts again, bye-bye bead. This is both a limitation and a feature.

I'm sure we'll find more limitations, more uses, and we'll find errors in what I wrote above.  Have fun playing with this. 

Here's my save file with experiments (very early cycles - so quick loading). Viscosity.sav

 

 

Viscosity.sav

[0xFADE]

Liquid loop gas pump maybe?

[Guest]

nice explanation, as always 

[0xFADE]

Waterfall pump sounds good. Trying to be descriptive with less words.

[beowulf2010]

I may not use some of these major exploits, but damn is it fun watching you all break the game. 

[mathmanican]

2 hours ago, 0xFADE said:

Waterfall pump sounds good. Trying to be descriptive with less words.

I would love to go this way, but if you use these words, then people confuse this pump with the escher waterfall, which is a very different mechanic. 

Does "Bead pump" work?  It's short and says exactly what's going on.  The full name would be "Liquid Bead Gas Pump" to signify that it is a liquid bead that pumps gas.  And then we just use "Bead Pump" for short.  Here are some other things I've thought of:

• Droplet pump (Maybe)
• "the drip effect that push gases up if you drip liquids on a wall using the conflict between 2 liquids" or "tdetpguiydloawutcb2l", from @tzionut
• Bead Pump (short, simple)
• Liquid Bead Pump (makes you think it pumps liquid.... wrong)
• Liquid Bead Gas Pump (long)
• Waterfall Pump (makes people think it pumps water using gas, not pumps gas using water, and gets the wrong principle in your head. 
• ....?  I'd love someone to come up with a snazzy great name. For now I like bead pump best of all. 

2 hours ago, beowulf2010 said:

I may not use some of these major exploits, but damn is it fun watching you all break the game. 

I want to put about 50 laugh icons on this one. I'm glad that this provides someone enjoyment.  I'm guessing there are also plenty out there that have put me on their "ignore" list, because they're sick and tired of all my exploit post.

I'd stop, but I can't stop.  This is way too much fun. 

[beowulf2010]

34 minutes ago, mathmanican said:

I'd stop, but I can't stop.  This is way too much fun. 

Please don't. I'm in the camp of "break it till they fix it". I just wish I had more time to contribute to breaking things. Keep up the good work. 

[goatt]

Hey professor, what do you mean by “bead gas gas gas”, “bead gas gas”, and “bead gas”, where to look for the pattern at game pause?

[goatt]

Another question: is this kind of rapid element swapping heavy to CPU?

[Yunru]

Honestly, I don't try and control it (and with an Escher Pump, can't really), I just have airlock doors cut off access for the gas.

[mathmanican]

1 hour ago, goatt said:

Hey professor, what do you mean by “bead gas gas gas”, “bead gas gas”, and “bead gas”, where to look for the pattern at game pause?

Let me know if this is enough (this is the perfect "bead, gas, bead, gas" alternating pattern)

39 minutes ago, goatt said:

Another question: is this kind of rapid element swapping heavy to CPU?

Tile swapping happens all the time every tick of the game, so I doubt this adds any extra computation time to the code. This is probably less computation intensive than regular pumps, because you don't have to compute heat interactions with pipes. 

[tzionut]

A very good post. Like @SamLogan series of how to. I like it. Maybe when you finish the post "how tho break ONI" using every exploits still not fixed devs will put more time in GQ (quality of the game) and fix them. It's every one choice if want's to use them or not. It's a single player game, and i think it will stay this way.

Best regards.

[mathmanican]

5 minutes ago, tzionut said:

A very good post. Like @SamLogan series of how to.

I was actually going to call it "The liquid bead pump: Everything you need to know", but decided against it near the end.  I'm not sure that @SamLogan wants these exploit posts to be attached to his series, so I backed off. This also means I'll need a new title for my post in progress on collecting the exploits. 

8 minutes ago, tzionut said:

Maybe when you finish the post "how tho break ONI" using every exploits still not fixed devs will put more time in GQ (quality of the game) and fix them.

There's currently so many bugs that could use attention. I figure that such a post (hopefully by the end of the week), with tons of links to help track things down, will help them rapidly isolate issues. Otherwise, they have to do all the work I'm doing. The difference is that I'm having a blast doing it for free, laughing all the way, whereas they'll probably be stressed out the whole time. Someone might as well enjoy tracking the bugs down for fun,  organizing them, categorizing them, and rating their severity. 

Time to go buy some PIE for the kids.  Happy PI day all. 

[mathmanican]

@Yunru, I thought you would like this find, as it relates to your comment over in this post.  The pump definitely is not completely symmetric. I was trying to make a new airlock and ran into this:

Both valves are set the same. I can't get the beads to form either on the left side, even if I add the extra tile, provided the room is in a vacuum. Resetting the valve does nothing while in vacuum.

I added gas, and deleted the extra tile to find that the right side forms bead, but the left does not. 

Adding the extra tile did nothing either. 

However, if I reset the valve to 0, then turn it back on, the beads form on both sides. Interesting mechanics.  They are clearly NOT symmetric. Thanks for your comment.

Oh, and once you delete the extra tile, guess what... the trail forms again, but only on the left. And the blob of crude oil you see in the bottom - it came from the left side, not the right.

[Yunru]

This is the smallest I could make it:

The valve is set to 1g/s, and the hydro sensor is set to turn on when it's below whatever causes it to flow:

Petroleum: 1kg

Water/Super Coolant: 0.2kg

 

Fun fact: Viscogel never spreads sideways, only up.

 

EDIT: New smallest:

 

[Yunru]

@mathmanican Something is going on here. This design:

Holds steady with 2 airflow tiles to the right, gains with 3, and loses with 1. The design before, however gains with 2.

O_o

 

EDIT: Actually it seems to be dependent on the pressure above somehow.

EDIT 2: Okay, I have no idea what's going on. This is somehow pumping the above to the side:

Edit 3: Think it's all to do with where the vacuum tiles form. 

[OxCD]

On 14/03/2019 at 10:01 PM, beowulf2010 said:

I may not use some of these major exploits, but damn is it fun watching you all break the game. 

I was about to say the same. I'm mostly playing "legit" ONI, but still I'm having guilty pleasure to see how you all are dissecting what was simple game mechanics into pure exploit ^^

[mathmanican]

1 hour ago, OxCD said:

I'm mostly playing "legit" ONI,

You might be playing legit ONI, but aren't these exploits legit?  

[Yunru]

It's no more an exploit than manipulating the forces of down-draft to make a lump of refined metals fly.

The game's physics are just a bit simpler than real life's is all.

[0xFADE]

This is a completely valid use of game mechanics.  Exploits are technically also completely valid uses of game mechanics.  The level of distinction falls completely on the user to decide.  I'd place the difference to something that can be accomplished without using the exploit. 

Can you turn one gas in to another?  No.  Exploit.

Can you pump gas from one place to another.  Yes, doing it differently is a clever use of game mechanics.

 

[mathmanican]

I guess intonation and  puns are hard to get across online. Definitions:

• legit - short for legitimate, as in "legit ONI"
• legit - modern slang synonym for "cool", "awesome", etc., as in "Aren't these exploits legit?"

I'd prefer not to enter another debate about whether something is an exploit, bug, feature, etc.    I'll be the first to confess that most of my work is legitimately an exploit, but man it's legit. 

For now, it's just sheer fun to take what we're given and break it as much as possible. When we know how to break something, we can aim to build the unbreakable. 

[beowulf2010]

1 hour ago, mathmanican said:

You might be playing legit ONI, but aren't these exploits legit?  

 

[0xFADE]

I havn't heard legit used for that meaning in awhile.

[Zarquan]

I would argue that this isn't an exploit any more than having a vacuum with a water lock access point.  I mean, you can do this in real life to a point (but the other way around). 

Sprengel pumps are real and they use a very similar mechanism, but in the other direction.  The Sprengel pump is even a percentage based pump IRL, but the limiting factor is vapor pressure and pressure differentials, which simply don't exist in ONI.  For those who don't know, they work by dripping mercury down a tube and catching atmosphere bubbles and pulling them down.  It can achieve 0.001 Pascals of pressure, where 1 atmosphere is 101,325 Pascals.

Also, everything is an exploit.  As a simple real world example, using gravity and the way solids work, I exploit a hollowed out cylinder with a covered bottom to hold liquid water without it spilling everywhere.  I can also manipulate the cylinder to have the water fall in an orderly fashion in to my mouth.  We don't think about it that way, but it is an exploit of real world physics and, if we were in a universe where surface tension caused water to climb up the glass (like liquid helium does IRL), then the "Earth Simulator" game forums would be full of people asking the devs to fix that whole "cup exploit" and say they play legit runs without exploits.  They would say "It makes liquid storage far too easy, you just put it in a cup and it will stay there for a very long time without making a mess and giving the humans the soggy feet debuff."

This is why I think the game should work for consistency rather than realism.  This pump is consistent with the rules of the ONI universe and, as a bonus, it has a real world equivalent, so it is completely fine.  It is fine because this is how the physics happen to work and it is the ONLY reasonable way is machine could operate given the fundamental laws of ONI physics (mainly the 1 tile rule and the liquid gas swapping rule).  The laws of physics would have to change at a fundamental level for this not to work.  Some features of the ONI universe are not consistent, like the factories that produce resources out of nothing, and it would be best if they were made to be consistent (i.e. removed).

Now, I am not opposed to removing infinite storage from the game.  The way I would do it is make it so if the pressure got too high, surrounding tiles would instantly break.  Like, anything next to a 10000 kg liquid or gas tile that blocks the gas's expansion breaks immediately and there is no way around it.  I would have doors and airflow tiles break too.  (The exact number could be on a per-element basis).  It would be a rule addition that is consistent with the rest of the game.

XKCD had a good comic about this recently:

(mouseover text: It's like someone briefly joined the team running the universe, introduced their idea for a cool mechanic, then left, and now everyone is stuck pretending that this wildly unbalanced dynamic makes sense.)

[0xFADE]

Holding liquids in a bowl is a clever use of game mechanics!