Efficient Duplicant Checkpoint Airlock Designs?

[lunazone]

I dislike how easy water locks make gas containment, so I was curious if anyone has made any efficient designs that keep gasses out.

[Zarquan]

Do you mean with a vacuum on one side or just separating gases?  Because a CO2 lock could solve your problem.

[Gus Smedstad]

I’m averse to liquid locks as well (I won’t say water because you often can’t use water going into a steam environment, like a steam geyser).

What I usually do is a small room with a gas pump and two airlock doors. I generally filter the output so one gas goes back into one side. I.e. all hydrogen goes back into the wheezewort room, and anything else gets dumped out into my base. If the other side is vacuum, I just dump everything into my base.

It eats some energy, but it feels correct even if it doesn’t have a proper door interlock system, and it’s very unusual for any gas to escape. Gas migration takes time and almost always the pump empties the airlock chamber before the gas makes it across 4 tiles.

The main place I’ve seen it not work was a volcano chamber with water and steam. Quite frequently enough steam escaped to condense into hot water once it made it into my base. I later shifted to a vacuum for the volcano chamber, which was much less of a problem.

You can also use transport tubes as airlocks, though it’s a bulky approach that eats a lot of energy.

It’s worth noting that visco gel seems to have no purpose beyond creating liquid locks, so I guess the developers see liquid locks as the normal solution.

[peachkillu]

The checkpoints sound like a great idea the problem is dupes lose their current task and drop everything they are hold when the checkpoint or door infront of them locks 

[OutlinedVermin]

2 hours ago, peachkillu said:

The checkpoints sound like a great idea the problem is dupes lose their current task and drop everything they are hold when the checkpoint or door infront of them locks 

actually checkpoints locking do not interupt tasks they only force the dupe to wait, but doors locking will.

[peachkillu]

53 minutes ago, OutlinedVermin said:

actually checkpoints locking do not interupt tasks they only force the dupe to wait, but doors locking will.

O wow I didn’t know this so you could actually make an airlock it would waste a ton of dupe time but is actually possible I did not know this  

[WanderingKid]

53 minutes ago, peachkillu said:

O wow I didn’t know this so you could actually make an airlock it would waste a ton of dupe time but is actually possible I did not know this  

The problem you have is the dupe comes to the checkpoint, and waits.  You cycle the airlock (lock the doors) and the dupe no longer has pathing, so they're now sitting around idle until they find a new task.  The strongest method you have, other than waterlocks, is tube access.

[0xFADE]

You would need to use tubes to guarantee no gas exchange 

Liquid locks are likely the simplest example of a physical design replacing a specific building. 

[Gus Smedstad]

I’d assume that the point of the checkpoint is that you don’t have to lock the doors.

I’ve never played around with the checkpoint, but I speculate that if you wanted to make a foolproof airlock with one, you’d leave the doors unlocked, but prohibit movement using the checkpoint until all the air (or other gas) was pumped out.

The problem is waiting for a proper vacuum is very slow, even in a 2x2 airlock chamber. You’d have to have good reason to want to do that over letting the Dupe just walk through while the pump was going.

[Gus Smedstad]

So I threw a quick example together around one of my regular airlocks.

Here I'm separating my main chlorine storage from the rest of my base.

I learned a couple of things the hard way. You can't put the checkpoint adjacent to a door, because the Duplicants will open the door and keep it open while waiting at the checkpoint. You have to be careful about linking the logic at the last moment, because it's easy to have the checkpoint stop 3-4 Dupes who will never get through because the logic isn't set up correctly.

The principle here is that the airlock is always in vacuum or near vacuum before a Dupe can open the door. Any gas let into the airlock chamber is pumped out before it can reach the other side, and a filter moves it to the correct side of the barrier. The atmosphere sensor checks whether the pump has achieved a vacuum.

Set the sensor to activate at 1g or below, and you can be reasonably sure no gas is going to cross in either direction, or any value you deem "safe." If you want to be absolutely sure of a vacuum, set it  to "above 0g" and add a NOT gate.

An alternate setup would be to make the airlock chamber 4x2 instead of 2x2 with a notch, and put in two sensors, adjacent to each door. Trigger the checkpoints only if both sensors register a vacuum - that way, you know there is no gas adjacent to either door.

Such a setup is not only larger, but slower because there's more volume the empty. It depends on how foolproof you want it to be.

[Lilalaunekuh]

On 21.3.2019 at 10:22 PM, Gus Smedstad said:

An alternate setup would be to make the airlock chamber 4x2 instead of 2x2 with a notch, and put in two sensors, adjacent to each door. Trigger the checkpoints only if both sensors register a vacuum - that way, you know there is no gas adjacent to either door.

If you don´t have unsual high pressure (>20kg/tile) on one of the sides or the temperature(>150°C) is a concern:

2 small gas pumps should be more power and more time efficient.

 

But if it fits my design, I prefer 3 small gas pumps in a 3x2 room wortking till a complete vacuum.

(At least under my test condions it was a perfect seal for a chlorine storage.)

Spoiler

If you want a real 2x2, 4x2 or maybe 3x2 room like we call it, without the additional tile with the atmo sensor:

Use a smart battery transformer setup to register the power consumption of your gas pump

 

[SharraShimada]

On 21.3.2019 at 10:22 PM, Gus Smedstad said:

So I threw a quick example together around one of my regular airlocks.

 

And now, adapt this, build it three times in a vertical line, and lock the doors, if there is no vaccum in the airlock. 1 of the 3 should be always ready to let dupes pass. If you want to be really sure about that, add automation that way, 1 airlock is always accessable as an emergency passthrough. 

For example a "super airlock" with 4 pumps and one of the normal ones on each side, that only becomes usable, when every other airlock is sealed. 

Keep in mind, airflow is faster, if the pressure-difference to the surround is high. What it want to say is, a low pressure airlock will be flooded, but will noch leak itself on the other side at the same time. You do not need to create a 100% vacuum to prevent gas flow in a critical area.

[Gus Smedstad]

I’m not sure I can envision a case where I’d want to build 3 airlocks for the same transition. Usually the spaces I’m protected aren’t big enough to support that many vertically. Oh, sure, the surface is pretty big, but I don’t remotely care about losing a few grams of oxygen if the pump misses them.

Locking the doors is bad because it screws with pathfinding and causes dupes to forget what they’re doing. The point of the checkpoints is that you can beef up an airlock’s effectiveness without affecting those. In practice, though, I’m fine with a slimple 2x2 room with a pump in it and nothing else.

[Grimgaw]

Last time I tested, this:

guaranteed no gas exchange for 0 athletics dupes.

[Oozinator]

On 21.3.2019 at 10:22 PM, Gus Smedstad said:

So I threw a quick example together around one of my regular airlocks.

 

I learned a couple of things the hard way.o be.

I did it similar and now i use waterlocks, because the airlock mechanic is not working fine / uses energy.
Not sure @mathmanican what about the gas / deletion / duplication bug. Here are two gases in different pressures and vacuum is intended, could it be triggered here too?

[mathmanican]

5 minutes ago, Oozinator said:

what about the gas / deletion / duplication bug

The two gases will race inwards at the same rate, regardless of pressure differences (10-12% each tick).  If your airlock were 3 tiles wide, instead of 2, then you would have matter conversion in the middle tiles (a very small amount). The extra tile at the top with the pressure sensor I'm not sure about. I think with this exact design you won't find any matter conversion, but I'd have to play with it.  In any case, the matter conversion here will be very small, because you are looking at 10% of 10% (so 1%) of each side of gas colliding. 

[Gus Smedstad]

3 hours ago, Oozinator said:

I did it similar and now i use waterlocks, because the airlock mechanic is not working fine / uses energy.

I’m pretty sure the airlock doors are working as intended.

You could build an airlock that worked the way real airlocks do, using two airlock doors and checkpoints, if you put the checkpoints on the inside instead of the outside as I did in my example. It’d be pretty big and slow, though. Minimum of 4x3 if you used mini pumps, since the checkpoints can’t be adjacent to the doors.

The problem with such a scheme is that the dupes would spend a lot of time standing in the airlock, waiting for the air to get pumped out. Real airlocks are pretty slow for the same reason, waiting for the pumps.

As I said earlier, I’m not really feeling the need. The 2x2 version with no checkpoints is actually pretty effective.

I know waterlocks are popular, but I personally don’t like them. If nothing else, they give the “sopping wet” de buff, which makes the Dupe stop and complain and causes stress. Not a huge deal, but I’d rather spend a little energy instead. Particularly since large scale energy production isn’t difficult, and airlock costs are not significant compared to everything else.

 

[Yoma_Nosme]

On 21.3.2019 at 2:38 AM, Squeegee said:

I dislike how easy water locks make gas containment, so I was curious if anyone has made any efficient designs that keep gasses out.

Do you dislike as well waterlocks that push gas up by water with doors underneath and achieve vacuum this way?

Hope you can imagine what I mean...

[wachunga]

On 3/21/2019 at 2:43 PM, Gus Smedstad said:

The problem is waiting for a proper vacuum is very slow, even in a 2x2 airlock chamber.

You might find this interesting.

The 2x2 pump doesn't pull in gas from all tiles of the pump's footprint. The gas in the top right corner first has to diffuse to one of the other tiles before the pump pulls it in. This is what makes pumping so slow. You start at grams and slowly drop into milligrams and micrograms until it's so low that the micrograms diffuse into a rounding error.

Mini pumps follow the same rule but their footprint is 2x1 and they can be rotated. With a bit of clever placement, you can make a chamber (including an atmo sensor) where all tiles get pulled into the pumps. It still takes a little bit to go from 200g to 20g for example, but the milligram and microgram stage is eliminated and pumping to vacuum is much faster as a result.

 The bridges are just to provide room for wrong gases to wait if they can't join the main line, so that the secondary pump doesn't stop.

 

And in honor of the poster formerly know as Rsomethingorother:

Klei, fix the bugs in your game. It's shameful if you release it in it's current (or near to) state.

[Gus Smedstad]

The milligram and microgram stages drive me a bit nuts when I'm creating a vacuum inside a turbine chamber. And I've learned the hard way that you absolutely must have an absolute vacuum for a turbine with only 1 port open, because otherwise inevitably a few grams of oxygen stuck in front of that single port will cause it to stall.