How to calculate the value for every other tick non-over pressure bypass pump

[guildcrest]

I've been creating a max power steam vent tamer that makes liberal use of the "Never Over Pressure" bypass pump as shown in this thread. In the post it is alluded that it moves gas every second or that one can slow down the tick rate depending on the single layered liquid flow rules. Now studying the ONiVerisity text and after playing with it for a while I understand how to slow the tick rate down but getting it to swap faster with the least liquid possible is evading me. I'm using petrol/naphtha and the swap seems to move at an uneven rate above 100g of petropl or if I turn the petrol valve up too high the swap "sputters" with the petrol sitting in the swap spot just draining until the mass is lower. So I'm wondering how to calculate the minimum amount that needs to flow through the valve to get the every other tick swapping?

[mathmanican]

Just noticed this.  I've been way too busy with other stuff.  The complete description of fluid flow on a single layer is here. 

Spoiler

 

If you just want to make sure you get the liquid to swap every other tick, then you need to guarantee that there will always be enough liquid over the vent to force liquid to flow right. For petro, this means you need over 400g on that vent for stuff to flow right. Liquid arrives every 5 ticks.  Since it happens every other tick, and you loose 1/4th of the liquid each time liquid flows, that should be enough to get your number. 

Note that petro and crude have 10 times the minimum flow as water and super coolant.  So you need 10 times the flow to get things going.  A valve set at 100g/s will work great for water and supercoolant, but not for crude/petro.  I generally just set my valve to 1000g/s for those liquids (it's higher than minimally needed).  You also need to realize that because of save/load issue, you have to plan for a possible 6th tick between ventings (if you care about the one possible missed issues on save/load).  

I go through a description of how to compute minimal flow in this waterfall post.

Spoiler

It's a linear difference equation. Search for those terms on the first page of the post and you'll see the computations. 

 

[guildcrest]

Thank you for the response. I ended up building an excel sheet model using the info in the OniVersity (and having an "AhHA" moment about the 4/9/3 rule means) to see what values would decay that would swap every other tick. However in testing I noticed that anything that wasn't every second (so 100g of petrol) was slower pulling gas to a vacuum. I'm guessing this has something to do with the rate of gas expansion into nearby tiles as you all were molding in the steam turbine thread.

[mathmanican]

The bypass pump is definitely something that could use more study. The fact that vacuum appears in the mix in certain spots makes for fun modelling problems. It also ivolves multilayer liquid flow (something I only have conjectures about how the exact mechanics work). Some day I may get around to putting out a post, but if you want to tackle that one and share instead, that would be awesome.  I'm currently stuck in a flaking rabbit hole that just keeps taking more turns.