Exact valve calculation

[stlnegril9]

This question is really related to the valve information on the GG/wiki, and specifically the info in the section under tips.

Understood that valves have discrete values (call it 2.3622) and results with one decimal precision.
So actual limits and resulting precision are, e.g.:

1 2.3622 2.4
2 4.7244 4.7
3 7.0866 7.1
etc.
Fair enough.
And the examples of a valve set to 240 rounding up (2.3622 * 102) to 240.9, or set to 36 rounding down (2.3622 * 15) to 35.4, follow suit.

However, the tip describing a 3:4 Sloppy Alumina setup having a valve set to 120, resulting in a value of ~118.1 is not making sense.
(2.3622 * 51) = 120.5, so why wouldn't it round up to 120.5, similar to the rounding up of 240 set to 240.9?
If input is 120, result of 120.5 is closer to 120 (if we're defining rounding that way) than 118.1 is to 120, so how does the calculation actually work?

A valve set to 119, should result in actual value of 118.1 (2.3622 * 50) (as should 117 and 118 as well), but not seeing why 120 should.

Note this is not a question on how to use valves, or even whether to use them at all; the question is on how the calculation works exactly.

===edit, 1st post had, incorrectly=====
"rounding down (2.3622 * 36) to 35.4, follow suit."

[Senrezu]

Just don't pipe your water residues into your fresh water supply and you'll never need a calculation or a valve ever.

[Dirtyshadow]

Abandon the idea that flow is a constant. Everything that flow into and out of a pipe is a tick rate of production.

Different machines operate at different rates so you get this back and forth. So what you end up with is if you graphed it is a Wave graph... its oscilating.

So for it to average 120... it will fluctuate between a value slightly above and below it. 119 120 121 120 119 120 121. (example using integers, but the osciallation is a floating point problem)

So if you hard cap 120... it will go 119 120 119 118 119 120... and that will average lower than LESS than 120.

What this creates is a cascading error, cause machines on the other end will be slightly short... they will idle take time to restart, then that creates a sloshing.

The pipeline manual McGalleon suggest flow compensators designs to smooth these oscillations. Creating closed loop systems is still the best as pipellines as manifolds and using valves as splitters just creates weird floating point math problems.

[stlnegril9]

Thought I posted this earlier... trying again.
(and thanks for input though I'm, perhaps pedantically, after how the valve flow calculation works, not after how to make factories avoid issues with valves)

Tried in-game verification.
Using 1 Water Extractor at 100% (120 m^3 min) and piped toward a (not-yet-connected) fluid buffer.
Once pipe is full, connect to Fluid Buffer and add valve.

Setting the valve limit to 120 settles on a flow rate of 120.5.

Dropping the valve limit to 119 (or 118 or 117) settles on a flow rate of 118.1.

Maybe the gg/wiki tip example info was just off slightly (though the logic is sound) ?
==========
Note - am thinking through the oscillating, tick-by-tick explanation, TY, but have yet to calculate how that would result in a flow of less than 120 m^3/min, with the valve set to 120 valve limit. (and assuming Fluid Buffers don't have the fluctuations caused by consuming machines)

[Kage Goomba]

Don't need buffers or valves for water recycling.

Got a aluminium plant with sloppy and it works fine.

Waste water is piped back in under the building then back up into a junction back into the same building - fresh water comes in from overhead.

Water extractors idle out when waste water overtakes.

Works fine without issue or complications.


Using same system for my Uranium fuel plant - with sulfuric acid.

Got screenshots on my profile if your curious.

[stlnegril9]

Note this is not a question on how to use valves, or even whether to use them at all; the question is on how the calculation works exactly.

My answer. The wiki entry is slightly wrong.
It should read something like:
"a valve set to 120 has the actual value of 120.5"
(though that doesn't fit the context of that tip very well).

[curtyuiop]

Don't forget to factor in evaporation into your equation.