Fluid buffer

[Gurgel]

We already have a regulator that lets a maximum per packet through. It would be good to have one that requires a minimum per packet (useful to make sure liquids do not freeze or evaporate) and stores input until that minimum is reached. Could also be implemented as "let exactly this amount through per packet" or have an upper and lower limit and store until enough if too low and store excess if too high.

[BlueLance]

I would personally love this is well

[Oozinator]

From Saturnus (i think)

[BlueLance]

yeah I know about this, but this wouldnt work for the OP because they want to let only a select amount through when there is enough of the material

[Kasuha]

1 hour ago, Gurgel said:

We already have a regulator that lets a maximum per packet through. It would be good to have one that requires a minimum per packet (useful to make sure liquids do not freeze or evaporate) and stores input until that minimum is reached. Could also be implemented as "let exactly this amount through per packet" or have an upper and lower limit and store until enough if too low and store excess if too high.

Assuming it would be a valve-like pass-through building, it couldn't guarantee that size of the packet because if packet of different element comes in, it has to release whatever packet it has collected so far even if it's not big enough. Plus the partial packet it has to hold could mean a problem too, it can freeze or evaporate while waiting.

If you want to prevent packets freezing or evaporating, just build the pipe out of abyssalite.

It's possible to build something using "pipe logic" that does the trick, as long as you have only one element guaranteed in the pipe and can use another element as blocker. I could try to construct it later today when I get home.

[Oozinator]

20 minutes ago, BlueLance said:

yeah I know about this, but this wouldnt work for the OP because they want to let only a select amount through when there is enough of the material

Mhh but a "room/liquid storage" after the overflow(pic), with pressure sensor and pump could be not the same?

[blash365]

Good suggestion. A workaround exists (as posted by @Oozinator), but i would also like a device-ish solution.

[Zarquan]

I've been wanting this for a while to.  Something that will block the packet until it is at least a certain mass.  You would have to be careful though, because if your gas isn't pure, you could have a problem if there is a small contaminant blocking the pipe, it could clog the whole system. 

[Oozinator]

7 minutes ago, blash365 said:

Good suggestion. A workaround exists (as posted by @Oozinator), but i would also like a device-ish solution.

Fluid mechanics works nice for me. I use this sort of overflow cooler/buffer. Not optimal, but works very well. Germy water is entering from the top left and self "balancing" level, after coolness kills the germs ^^

[blash365]

@Oozinator: I was talking about a device that takes an input of y/s with y <= x and creates an output of x/t with t <= s. E.g.

100mg/s -> 1kg/10s

Of course this would not be trivial for mixed contents. A simple solution would be that the device puts out the first element of which it aggregates the amount x.

[Kasuha]

Okay, so here's the design. It uses "mechanical filter" technology so if you're familiar with them it will be easier to understand.

Without icons:

And with some legend and description:

Let's assume you're "quantizing" packets of water and use polluted water as stopper.

Left green arrow: input

Right green arrow: output

Blue rectangle: buffer

Red arrow: overflow

Orange rectangle: stopper

Blue arrow: backfeed

Stopper liquid is only moving within the orange rectangle. The valve inserts small packets into the pipe, and the mechanical filter grabs them back.

While the stopper liquid is cycling, the bridge to the middle of it from the buffer is blocked and the liquid collects in the buffer.

When there's too much liquid in the buffer, it overflows and extra packets of water go along the red arrow, enter the pipe before the stopper and block the valve through which the stopper feeds polluted water. These packets then continue under the bridge from the buffer, and this bridge makes sure they're full packets. Then they continue further out of the stopper, and there another mechanical filter steals part of the contents and sends it back, setting the size of the packet on the output to the desired amount. Setting on the valve of this filter sets up the size of produced packets.

Liquid grabbed by feeder goes back towards the buffer and overflow, and has precedence over liquid coming through the input pipe.

Note that it only works if only one kind of liquid comes through the input. It can be any liquid except the stopper liquid (assumed polluted water) but it must be only one liquid.

Also note that the buffer needs to be bigger than the length of the overflow. It's not big enough in my image, I just wanted to keep things clearer.

But honestly, this is just an exercise. It does not have much of practical sense designing or using it. The main obstacle is that it relies on the output pipe being free. If it blocks and the packets start merging in it. the size of last packet is undefined again. Worse yet, if the consumer draws irregular amounts from it. A lot of effort going into producing uniform packets is wasted on such situation.

Also in my experience, it's best if supply is regulated by demand. Meaning, the pipe should be full all the time. Whenever something draws some liquid from it, it should be replenished again. If the pipe is empty, your demand exceeds supply and you have insufficient supply for your needs.

[Gurgel]

You mistake the purpose. This is for heat-exchangers running through WW or Nullifier-coolded Hydrogen at temperatures far below 0C. If a small water packet goes in there, they will break. Unfortunately, when loading a game, a small packet as the first packet is a regular occurrence. Now, when doing a regular switching of a pump, that can also happen, but that you know and you can simply redirect the start of the output of the pump back into the tank it came from for the first few seconds, no problem at all and same for switching the pump off. But loading you cannot detect, as far as I know. 

As to the valid problem what to do when mixed fluids go in, there are a number of possible solutions

1. Just pump out anything of the old fluid if a new one comes in, regardless of whether it meets the criterion. Advantage: Simple. Disadvantage: Can still produce small packets. But put a liquid-filter before and it is not a concern.

2. Block and require deconstruction to unblock. Pretty bad, but would still be viable and solve the problem.

3. Combine with a liquid filter. In fact this could be done as an "advanced liquid filter" that accumulates its target liquid until a set amount is reached and only then outputs it. Or it could be a setting on the regular filter.

4. Combine it with the pump: Pitcher pumps already do this: You just get one liquid even when there are several available. The pump then would have a "packet size" and only outputs liquids that have enough available of that size. 

I think there are enough good options here. Personally, I think 4. would be most elegant, as you need the pump anyways. 1. would still work for my application, and I frankly see not many more than that heat-exchanger use where mixed liquids will usually not be a concern in the first place.