#math-and-meta
1 messages ยท Page 213 of 1
But yeah, what we do know for certain is they're storing the liquid contents, as m3, as float.
short of disassembly or stepping through in a debugger
which isn't bad
float is adequate for the range 0-whatever that represents the volume of fluid in a pipe segment with max length of 50? meters
I.. mean.
Double precision floats are adequate for anything through e308. ๐
I'm actually not sure what the max volume is on a pipe. The minimum is 5.
it would be better and faster if they stored the representation as integral milliliters or something
Which they don't.
with pipe welding idk if there's a max
Best we can hope for is maybe they do math on Liters instead of m3.
well, i mean, that's pretty much x/1000
Casts to int, too.
i have no doubt that there's precision problems in it all
we are getting there :3
1440 aluminium scrap pm complete, now time to save scum hard drives till i get pure ingot alt
but as with all things involving fp math, where the problems actually lie is always very non-obvious
Yeep. The scale is also interesting.
I could see FP error accruing causing pipe networks to very slowly fill or empty.
Very slowly.
i've observed that
Are you sure that's what you observed, though?
Pipes also have hidden volume.
So they're damned liars about what's actually in them.
wasted a lot of time in the u5-u6 era looking at that stuff
Fair.
the code has changed a bit since, but those things are still around
there's a reason i build pipes a certain way and follow some unwritten rules
i have a set of solutions that work and don't rely on voodoo
ofc, power stutter from the hoverpack just blows it all to crap
That's mostly just "don't recycle" and "don't rely on 600 throughput".
You do those two things, it's very easy to have reliable pipes.
i do recycle, but pipes at 600/min need special care
If you do recycle, I'd bet money you've baked the voodoo into your unwritten rules.
It's sneaky voodoo.
top feed only and only use mk1 pipes for it ๐
Those aren't unwritten, you just wrote them.
Bear in mind, things as simple as "how you connect pipes together, and in what order, and with what structures" can invoke the voodoo.
My working theory is that people that never have issues, build their pipe networks intuitively in a way that invokes the voodoo.
Also things like "top-feeding" are simple robustness measures. Shouldn't make a substantial difference for this conversation.
i think what screws a lot of people is that there's some form of 'memory' that pipes have when you start reconfiguring them - same thing you build may work differently if you had other things on a pipe network and changed it
what those hidden variables are is beyond me
Not as much as you'd think, actually.
but if you have experience, you're building it once in a manner that works and not doing any monkeying around with it after you initially flood the pipes
i suspect that something of that happens with many problems
it needs to come from right side to the station
Your station is backwards I think.
the locomotive needs to dock into the station
Take a top down picture.
@somber venture see #math-and-meta message for alt recipe choices
Anyone that can tell me if this is a bug or something im not aware of. im working Rocket Fuel, and as you can see my buffer is full with 2400 in it, but it only takes out 468 / 600 ? and i try to put a valve on it becuase it is a gas so it dosent go back up, is it becuase a gas as rocket fuel can move with 600 pr min ? or is there something im not aware of? im asking becuase i have set up around 1700 fuel generator for it, but the pipes only go with 468, and becuase of that, not all of them are getting the amount i have calculeted :/ Hope it makes sens what i mean ๐
Buffers don't work well with gasses.
don't use buffers and valves
hmm, so my problem is a im using trains to move all of this rocket fuel..
If you want to do that, package it.
ROger
Buffer it with containers, then unpackage it into your generators.
Shees that alot of packing to do xD
The long and short is that pipes can only move at full flow when they're full. Because they don't care about gravity, buffers can't produce full output until they're full.
It works fine with liquids because they can cover the output pipe easily.
i see, i will try to remove the buffers and let it work from the train station only, and if that doesent work. i think i might need to look into packing all the gas :p
thanks alot for the help
That's going to be problematic too: train stations lock outputs while loading/unloading, that'll cause interruptions to your flow.
If you want to move gasses, your best bet's always going to be to package it.
Also why, generally, it's recommended to not try to haul gasses anywhere.
gas package loop isnt that bad
I mean, I'm anti trying to haul fuel for my power generation anyway.
Are the heavy oil residue + diluted fuel alt recipes more power effective than just making fuel?
Very much so, yes.
Thank you
if you just count the machine power draw, no
if you count the power generated from fuel as well, then yes
I'm just interested in the total difference
then yes, it makes much more power per oil
but if your goal was "make some fuel as cheaply (power-wise) as possible", then base recipe is the way to go
this is what we're aiming for ๐
Dont really know guys, the fluid dynamics is what might kill the game for me..... 4 OC refineries, directly linked to 5 OC blenders, the last 2 blenders in the pipe do not get enough HOR , this is just poor mechanics...
@vapid gorge threw a single sloop in each rocket fuel blender to really overproduce and I think it's working? Starting to stabilize now that I did that and found (hopefully) the last production bottleneck
yay 100% ugly 20 hmf p min, was quite the journey
yeah pipe manifolds perform not nearly as well as belt manifolds especially when you get close to the max limit of the pipe
the fun of the game for me is to get numbers in the lines of 4000 plastic and 4000 rubber in a single spot
and the system constantly crashes because of those pipe mechanics...
There are definitely limits to what is possible in the game ๐
I am sorry but the limit being fuild code mechanics is not something to enjoy,
No, that is not an artificial limit put in to increase player choice, it's a technical limit that the devs can't really do anything about
Ehhhhhhhhhhhhh.
what's it with factory games and incomprehensible, janky, fluid mechanics.
Clips for the May 7th, 2024 Livestream originally streamed on https://www.twitch.tv/coffeestainstudiosdevs
Reminder: This is an unofficial channel, support requests should be directed to https://questions.satisfactorygame.com/
00:00 Intro
00:26 People have noticed there are issues with the Mk.2 Pipes
01:12 However...
01:38 Stop! (Water) Hammer...
@robust vessel ๐ This was from May. It's definitely a choice they've made, and not really a technical limitation.
@vapid gorge looking really good now, about 20 min in
Anyone know how far a path signal reserves a path? Cause I cannot figure out if it's to the next block signal, to the next path signal or to the destination.
Assuming constant elevation can a pipe be extended perpetually and continue to fill?
until the next signal
It reserves a path through to a block signal, including any intervening path signals.
block or path or EOL
And if that block signal is red, it does not move, right?
Correct.
yes
It can't reserve it until its exit block is available.
First.
can they not wait at paths then?
Cool, thx
I think, sort of?
If two trains are going to the same exit, then only one can reserve a path to that exit at a time.
I might have to go experiment
If two trains are going to two different blocks, I think they can both reserve paths, and whoever's second will go until it hits the first train's path, and once it clears that path, it'll continue to its exit.
I was under the assumption that paths and blocks worked almost identically just that paths would allow a second train through if it would not collide with another train in the process
No, paths behave substantially differently than blocks.
When you have a path-block, as in a section of rail that has path signals entering it and block signals exiting it, any train that wants to cross it has to reserve a path successfully to do so.
And it can only reserve a path if it can exit.
I'm not sure whether it can reserve intersection paths or not. I think wiki says it can't, and my memory says it can. ๐
might have to experiment a bit more about where they stop and when they reserve the path but I think I get what you're saying
So it's worth experimenting with that.
They definitely reserve the full path before they enter the first path signal.
make the block before the path longer because it is considered a red signal and trains will brake 250m in adfance
if you have a 4 way intersection I know they can reserve multiple paths through it provided the paths have enough space
Actually as soon as they enter the block before the path signal.
so opposite corners or opposite straights
Yeah, it's all about exit blocks.
Yeah that's kind of what I meant. It's not like the devs are intentionally adding some weird thing that makes pipes not go to their full capacity, it's that the way fluid dynamics are implemented means that some ways people use pipes they end up with unexpected behavior that the devs can't really just "fix" by itself
This nuance trips people up, too: if you put a block signal too close to a path signal, trains will nearly stop before reserving a path.
oh ok
You can strategically prioritize paths by varying the lengths of the entry blocks that way.
I was thinking "would that make it best to do block then immediate path then junction then exit block" but between the two of you that is basically the worst thing you can do
Yep.
For example: I'll have a train coming down hill that I don't want to slow down, I'll make sure the whole hill is in a block signal, so it can reserve at the very top of the hill.
how much are you over producing though? try putting it back to normal prod and see if it still balances out or if it starts starving
This means trains can't cross until it clears the intersection, but it also means it won't slow down.
is there some kind of trick to getting an aluminum factory that recycles the water back into alumina machines to run at 100%? i got it going at around 90% but there is always a slight fluid backup for some reason
are you recycling into input or into dedicated wastewater system
I wasnt saying that this is a bug, i am just saying if this stays this way the enjoyment of the game for me will be gone and every time i want to load the game i think of the fluid stuff and makes me not want to load it
yeah you have some machines only use recycled and some machines only use fresh
the only recycled machines can use their own output in the case of it being uranium and sulphuric acid
burn the wastewater in a seperate system
the devs have said they will consider changing fluid dynamics to remove sloshing but that's not just an overnight change
or just yeet it into a nuclear reactor lol
wow we still havent evolved past this?
instead of it being "Satisfactory" with perfect numbers and design, it is more of a chose trying to fix a feature that is not givign you correct output
recycling into input is finnicky,
you need some pump bs, not worth the effort
the best choices are to sink it, use it in a generator or use it in a different system yep
Well with the sloops it doesn't matter, but by quite a bit lol (+25% over target)
there are methods to keep it all in one system but it's finnicky and breaks easily as you've discovered
i never got it working in the first place lol
yeah i have that
that is a really good guide to fluids in general, one of the pages descibes the VIP system which is what you want to keep it in one system
i am using a VIP junction with the recycled on the lower pipe
Can't you just underfeed the system by however much water it produces?
was thinking maybe adding flow equalizers to the fluid loops would help?
it's super unstable still
Like if you get 30/min wastewater just cut the required water by that much ๐ค I never check my aluminum system for efficiency tho. 100% efficiency when fluid is involved is a fool's errand
like you would think that's it but nope fluids
I would just go with solution 2 on page 14
oh sure, but for the sake of science, it woudl be good to know if properly flooding a system would keep it stable w/o having to always over produce
wdym mean 'evolve past'? It's the most robust system.
by far the most reliable, and basically unbreakable
how can you just "burn the wastewater"? like, packaging and sinking it, or putting into a reactor?
best is if you can get a limestone node nearby and make a wet concrete recipie and sink that
that or making a seperate production line only running on it
or burn it with some petcoke in a coal gen
it's effectively teh same thing as using waste water in a seperate bauxite process, except needing extra materials taht you usually sink instead to use
that's a choice as well, but I usually don't consider people running coal gens after they get fuel and nuclear
people usually limit their options through weird preconceptions, yes
and people for some reason don't avoid the issue entirely by simply reusing it xD
Speaking of sinking, when you use a sushi-belt, do you always need sinks to kill overflow so stuff doesnt get stuck?
at least at the very start you generally do
once it stabalises you don't. Though if you're particularly paranoid you might want to leave a single over flow for yoru whole factory in case accidental item dupes on the belt eventually causes an issue
though it's unclear if that happens anymore
if you build sushi correctly, overflow isn't required
I mean, otherwise it will clog up
yea for sushi you may need it
For sure, will do a bit more testing later
perfect ratios are easy though with clocking,
not buffers
machines are literally flow valves
you have perfect flow control of what goes on belts
I know, but we got the valve, which is also unnecessary, right through that arguement.
valve is literally unnecessary tho
Just some consistency here would be nice
the fluid valve is also pointless because the consuming machines control the flow
a bette argument would be to get rid of the fluid valve
That's what I said. "which is also unnecessary"
If the fluid problems persist I kind of hope they add something like a pressurization tank. Something sized about the size of a regular fluid buffer or maybe a bit smaller. Give it 10mw of power and it will act like a packager/unpackager and give you a "pure" source of fluid
Makes sense that just stringing a bunch of pipes in spaghetti wouldn't work, but there should be options to make it work if players are smart about it
I dont really know.... it is just something that 600 in does not equal 600 out for me is just not Satisfactory
or you built it wrong ๐
Not that the game bothers even remotely suggesting what you did wrong or anything.
I am building it 7.5 Refineries into 6 blenders, single pipe horizontaly, if that is wrong....
looped the pipe? fed from above?
one level not a santimeeter going up or down
4 meters from eachother, not a finle pump, valve, floor hole nothing
looped the pipe?
always was
See, the trick when trying to use 600 in a pipe, is to not use 600 in a pipe by instead splitting it in two and using 300 in two pipes.
fluid bounces back when it reaches dead end, which reduces flow
looping the pipe makes sure that there's no "wrong way" for the fluid to go
ok let see if it solves it, I am willing to try anything now
a way to build pipes so that it's almost guaranteed to work:
- don't use buffers or valves
- loop the pipe manifold
- feed from above, not from below
- keep system as simple as possible
- prefill pipes before running the system
- don't use floor holes
- only use pumps if going up
They fixed the floor hole bug though. ๐ฆ
yep, I think they did fix it
is kinda old message that I quoted... tho I've seen some people still saying floor holes are bugged, idk if that's true or not
I haven't seen anyone saying that because they actually tested it and saw a bug.
I've only seen people repeating that because we like sacrificing goats around here.
I use floor holes all the time, never an issue but I do prefill and balance the input on the network. Manifold with liquids is finicky
manifold is the only way to do liquids. You can't balance them
Sure you can, think on the 360 water issue with Mark I pipes
The looping is a sort of balance too
by "balancing" I mean "building junctions with assumptions that fluids split equally between two exits (and that they only ever flow in one direction)"
how to loop the Oil extractor?
looping applies when doing pipe manifolds
I do need Valves thou, are those still a problem?
what do you need them for?
to fix usage
not sure what that means?
i will try to find a way around it
Hey guys, pretty new to the game. How would i get these 6 smelters to split 100 ingots to their respective constructors and the remaining 80 to the others? Thanks!
make two groups of smelters, one making 100, one making 80
Well, liquids in Satisfactory are not a real simulations, sure we have flow and throughtput but here liquids flow in all directions at all time so you can certainly balance a liquid manifold.
If you feed 8 coal generator with Mark I pipes and you connect the all the water on 1 long pipe, 2 things will happen, 1 generator will stop working and another will work half the time and 1 water extractor will not be online full time because of pipe limit while adding and additional pipe looping the system will allow everything to work at 100%
that's not balancing... that's just normal manifold
merge 3 on one belt and 2 on another, then split the last one 3 ways, 2 of them going and merging to the 2 smelters and one way going to the 3 smelters
anyone know if there is any reason to invest in trains instead just rushing to drones? i used to go full train and make a interconnected network since batterys are complex to make but now that drones can use regular fuel should i go for drones?
I always have to remind myself that it is not a simulation when I see "flow rate max but pipe is half empty" ๐
drones only have 9 slots
while one freight car has 32
if you merge the outputs of the smelters to 1 belt you can make this to split it up correctly
the thing i drew on the left is supposed to be an 80, fyi
lesgo were making space parts with box factories xd
wait i started thinking during my stall face im having so one slipliter one in and one out shuld be 1 to 1 and if im thinking right a slipliter with two outs shuld be half the in going? on both belts? right
Easier to ride a train than a drone.
Also trains can move liquids without packaging it. Drones can't.
Is there a way to see the current head lift on a pipe? Ideally, net of gravity.
Look at the pump providing the head lift.
If there isn't one: it's 10m from whatever is providing the head lift.
anyone making spaceparts like me?
my brother ๐ซ
so this pipe that this pump is on is experiencing 3.7m of head lift?
It's providing 3.7m of head lift right now. Which is to say that the liquid reaches 3.7 meter high. It can provide up to 50m. So you can run your pipes 46.3m higher before you would need another pump.
Assuming your pipes are full right now.
That number is the actual head lift being provided right now.
That gives me anxiety
im loving it
this is the furthest ive got into the game, the first phase 4 parts are rolling out!
Yeah. I need to suck it up and put together my factory for them. Last night got my drone running for bringing Copper to the aluminum factory. Just startin to feel overwhelmed with it lately.
Builds are gettin so big.
more BPs required
Ok. I downgraded the pump to Mk2 and set the pipe height to 22m, obviously it stopped working.
I then applied another pump and it was able to deliver to 22m. I know I can't stack them right next to eachother to get 40m of head lift, but how is the headlift calculated per pump in a series? Is the lower pump just calculating head lift to the next pump?
Pumps just straight up reset head lift.
No calculating happens.
in series: from one pump to the next
Should I make even amounts of plastic and rubber???
They just ignore all previous pumps and set their own head lift.
pump head lift does not add, it just ENDS at the next pump
Do you need even amounts of plastic and rubber?
I dunno was just wondering
Is the lower pump just calculating head lift to the next pump?
So yes, is the answer to this.
Greeny I was about to give up on the game after 1200 hours, thanks for the looping tip, it saved everything
Pumps dont really calculate head lift as much as they just measure it
"im here, how high does the pipe/fluid go before there is another pump or exit?"
They arguably just produce it, if you ask me. ๐
That makes sense
They produce and then measure
Tomato potato.
Yes I make extra and leave it on the floor to make a bounce house.
they always apply a fixed amount
namely the thing they are set to: 20 m or 50 m
the applied amount never changes. if a pipe is flat on the ground the pump still applies 20m
its just that it only gets to measure 1.3 m
if pumps constantly had to self regulate and increase the headlift they apply that would be a nightmare
so CSS went with the simple answer: just always apply the max and then just measure how high the fluid actually goes
the fact that head lift stacks with gravity is not some pump magic, its just the pipe code taking care of it
and some clever height measuring
Is it really clever, though? ๐
The manufacturer said 20m, but if I paid for it I'm going to get 110% out of this pump!
Now how will my brain make this work 1080 oil to fuel
Fair warning: if it goes above its designed limit it reduces flow.
The fact this game can take gravity into account when calculating total head lift is something I find very interesting.
Wait until you learn it takes viscosity into account too.
I don't plan to actually push it that far. I'm just testing out fluid dynamics to make sure I have a better understanding before I start my massive build out.
i put signs just for you
managed to finish the Fourth slice and realized I actually needs less than 6.4k rubber per minute from recycling
that's way too little
I only need 5124.3 from recycling :(
Would y'all build a factory floor to the near exact height of the machine on the lower floor or would you go for some even floor count like 32m in the case of the refinery?
i usually go for the next wall height (4 walls for Sm, Fo, Co, As; 5 walls for Bl, Ma; 8 walls for Re)
I just make all my floors 10 high cuz i like jumping on things :]
seems like the right play, but I kinda like the added benefit of "capping" the smoke stack so I release less polution into the enviroment.
I coulda swore it used to do that. ๐
At least on foundries or smelters or something.
hey all. is there a way to make a train stay on the station until there's at least something in the platform to load into the car?
I tried "One load/unload has been completed" and "wait for 0 seconds", but this just makes the train to leave as soon as it stops on the station
i need it kinda like factorio's "Until X amount of Y item"
time it manually /foreheadtap/
or set it to wait until full/empty
but manual timing is superior ino
imo
what do you mean time it manually? like, precisely denote how much it should wait?
yes.
this wouldn't work because it's a packaged N2 train, and if I do that - it would have to wait until all gas has been (un)packaged
ok I'll try to think something about it, thanks
then why not do that?
because after the entire gas is (un)packaged, the packager would have to wait until train leaves, gets a batch of empty tanks and comes back, and until then it would do nothing and won't supply/deliver any gas
ISC?
what is ISC?
Whats wrong with just having a full buffer of canisters/packaged nitro, to (un)package on either side?
Industrial storage container
basically nothing wrong, I just firstly decided i wanna go with exactly 32 stacks of tanks to fill the entire car
but it appears I'll have to make more so I wouldn't have to deal with the current problem
Well.. not having saturated buffers is a big issue for getting throughput through a train
๐ค
because freight platforms lock up, and belts in/out will stop. If you dont have a buffer to collect your continuous feed, that backs up and pauses production. (or starves production on delivery side)
(assuming you dont already have an entire full 2 belts of production that won't fit into a platform anyways)
i know about that, I just wanted to try to make a continuous loop similar to factorio
but oh well
i dont know what you mean by continuous loop
If you leave trains on "complete a single load", they'll just pick up whatever's there and then drop off whatever they have, running continuously.
how far on the map can I build?
- the train takes as many empty tanks as available
- goes to N2 station
- unloads the empty tanks and takes as many filled tanks as available
- goes back
- unloads the full tanks and stays on the station if there's no empty tanks, otherwise repeats the cycle
in my case the train doesn't stay on the station and continues to go here and there even if the cars are empty
ok I found out lol
theoretically - as far as you can. practically - until the border starts to kill you
Ya I found that, border be aggresive.
I kinda like building out here though, no geography to navigate or monsters to kill
you're trying to make one train do 2 different things? "take only what's available" vs "wait for a set amount" and satisf. trains are dumb and cant be told to use logic like that
:(
@ashen girder i just figured out how to solve fluids
its very simple really
just put a ||packager infront and behind every machine||
Don't max out Mk2 pipes?
not maxing out mk2 pipes doesnt help
i got the same issues i got pushing 400 through a mk2 pipe that i also get pushing 600 through a mk2 pipe
literally no difference ime.
return the game to the glorious update 3 days
im sure u could wrap it all up in a neat blueprint
Anytime I do a big build that involves fluids, I wanna quit the game after
small builds are even worse because the problems dont acumulate and its harder to spot them imo
Always run into issues, even when sticking to best pipe practices
Yah, but big builds require a bigger time commitment in total
"best pipe practices" is just mythology, sacrificing goats to the pipe gods. the real way to fix pipes is pumps.
I'm starting to setup a massive refinery factory and man am i dreading all the pipe work.
And it's just demotivating when stuff doesn't work after spending hundreds of hours on it
find me one problem thats not VIP or math based and ill fix it with pumps. id bet u 5$ on that
literally every pipe problem ive encountered so far (thats not VXP or math based) could be fixed with pumps
or well, pipes not connecting at all but thats a different issue too
I have builds that worked without pumps in update 7, now requiring pumps to properly work in 1.0
I'm setting up 18 fuel refineries. This requires 1080 crude oil a minute. Without doing the crazy math to figure out when I can merge in the second pipe, can I just connect all the refineries together and feed one 600/min tube at the start and one 600/min tube at the end?
Yes
is there a different recommend method that I'm not foreseeing?
I intend to do that, but may not be perfect. This only becomes an issue if I'm near max head lift yeah?
Sometimes I miss the times when fluids were items tbh
i wouldnt mind the wonky rules of pipes if they just told us what those rules are!
Fluids prefer lower elevation, so if you have parts that are lower, they will fill first and prefer to be filled
Or you can, you know, fix or simplify those rules so we won't need master's degree to understand them
but i WANT, complexity, i crave it. i want to have to write programs to simulate stuff to optimally build my factories.
i just dont want to have to play detective to figure out what those rules are.
I build on foundation so I'm not forseeing this be a huge issue. Regardless, if the pipes are saturated and there's enough throughput when does lower pipes become an issue?
I guess if I had one machine way lower that took all the flow rate?
okay, how does that explain the fact that u can make a VIP junction on a flat surface just with junctions?
I see both points. They want people to discover and experiment. If you give them all the rules there's less to figure out.
How do I pre-fill a machine without it producing?
but literally noone knows the rules. theres people here that wrote simulations for the games belt system to figure out how manifold filling works, calculators to calculate the optimal recipe chains to use and maps that map literally the entire game
but noone knows why hte fluid doesnt go up.
after years. people are resorting to prying open the game and trying to look at the code to figure it out and still nooe truly understands fluids
its not just that its hard to figure out by yourself, noone knows it.
thats what im assuming
i mean, they obviously know the base rules of the system, but that doesnt mean they fully understand the deeper implications of it. which is understandable and totally okay. jsut tell us the base rules so we can figure it out ourselves
Pretty sure there's one guy that understands it, at CSS. ๐
theres variables for viscosity in the game. what is viscosity???
maybe that one janitor
probably !
like, just because u write a system doesnt know u can accurately predict how it will act. its not that simple. thats my guess at least
After all, janitors work with liquids in one way or another 
equal to 1 for all fluids
Then what's the point of it wtf
it was a thing in earlier updates
instead of removing it from all code and places, setting it to 1 effectively removes it with much less work
Gustav, apparently. ๐ https://youtu.be/COJV4Ubn1Ac?t=106
Clips for the May 7th, 2024 Livestream originally streamed on https://www.twitch.tv/coffeestainstudiosdevs
Reminder: This is an unofficial channel, support requests should be directed to https://questions.satisfactorygame.com/
00:00 Intro
00:26 People have noticed there are issues with the Mk.2 Pipes
01:12 However...
01:38 Stop! (Water) Hammer...
How do you know this?
Docs.json have included that variable for several updates before the variable was removed from Docs (as it's pointless now)
Yeah, I was wondering if it was specified in there. Good to know, thanks.
so technically it's possible they may have changed it again, but I don't see why would they do it (and there's not really any difference in fluid behavior between different fluids anyway)
not counting liquid vs gas
i watched that video before and...
im not a fan of it
I know you aren't, but he still mentioned G2. ๐
whos g2?
Gustav.
i see. we need to talk to gustav
(the urge to ping a random person named gustav in this server is very strong)
๐ If he's orange, I'd say it's both the correct person and something you should absolutely not ever, in any universe, do.
If he's white, well, he's an innocent bystander. Best not throw goats at him.
You can and should throw goats at Snutt and Mikael though.
yes, none of them are orange thats why itd be funny.
Snutt promised us another video in that video.
i see...
I don't think that other video ever came into existence.
whens the dev stream again?
Tuesday.
i see. im going to try to bully them into having a conversation with mcgalleon or something like that
I'd settle with bringing Gustav and ELI5ing the fluid system on a video or stream. ๐
idk, im worried they'd just go "ah yes, theres the waterhammers. also, dont forget headlift! fluids cant go infinitely high. you silly goobers. thats it. ive explained the fluid system. good bye"
Yeah, that's fair.
trying to find a pic of a hammer made out of water to post and be like "who are these waterhammers and why are they in my pipes" but i cant find one and i dont have my ai image generation stuff set up, so just imagine iposted that and it was really funny
honestly i love how they depict waterhammer in these images. the liquid just goes crazy
I love this video about it: https://www.youtube.com/watch?v=xoLmVFAFjn4
He straight up broke his pressure gauge by accident.
Hydraulic transients (also known as water hammer) can seem innocuous in a residential setting, but these spikes in pressure can cause major damage to large pipelines and industrial pipe networks. In this video, we briefly discuss how water hammer occurs and how engineers mitigate the effect.
Watch this video and the entire Practical Engineering...
yea i thought mabye i missed something about waterhammers but after this video im even more confident that they shouldnt result in what we see in satis
You think 150 crystal oscs is enough for a casual endgame experience?
yes.
Yeah, the water hammer thing is specifically only talking about what happens when liquid tries to backflow into a junction and disrupts full volume flow.
It has nothing to do with the other failure modes, as far as I'm aware.
its only for flow rate, yes
(in SF)
Yeah, IRL it likes to break shit and makes lots of scary noises in old buildings.
yea but with this setup here, wheres the water hammering?
and why is it fixed if i add a valve
Beats me. I don't think water hammer explains all sub-optimal flows. Just the most common ones.
pumps shouldnt fix waterhammer, but for some reason every piping issue can be resolved with a pump (besides pipes that arent connected, math issues or VXP situations)
find me one situation like that that cant be fixed by a pump and ill give you 0.01$ (i wont)
That's low confidence 
not low confidence, laziness.
A machine outputs 600 fluid and feeds a long straight manifold of machines each taking let's say 30 of the fluid.
The last few machines get starved eventually. And the output machine is not emptying well.
Where would you put the pump to fix this?
between the machine that outputs 600 fluid and the 20 machines that take 30 fluid each
i would guess 100 meters of headlift would be enough.
(assuming bottomfeeding, if ur sidefeeding then way less should suffice, if u even need any)
We don't know if it's top or bottom feeding, yeah. ๐
I've tried this and it doesn't fix it. You need to loop the end of the manifold to fix it.
i doubt that
See, this is where the water hammering explanation involves itself.
ur really making me prove u wrong... brb
No mods tho 
(i hope im proven wrong)
no mods !
can i use two machines to create 600? water extractors only go to 300
loop on the start is usually the best bet
aka just a bypass from the first junction down to the fourth or fifth. or to the middle of the manifold
im loopphobic, both in the homophobia sense of the word phobia and the arachnophobia sense of the word phobia. as in, i am deeply afraid of them but i also hate them
then put a valve on the bypass, then its not a loop but a split that merges again
like with belts
Take a rocket fuel blender OC to 200% and sloop it, it will output 600 rocket fuel. Now try feeding all of that rocket fuel to a long straight manifold of fuel gens.
A pump at the start will not prevent the last fuel gens from starving.
because its a gas
Gasses don't have water hammer.
you cannot use pumps on gas
in satisfacotry they should have too as they dont compress either i think
it ignores them
Ok, nvm, not rocket fuel, let's say regular fuel
why not water?
Ok, water
i have a split-less manifold for water i could show
Does it reliably support 600 input?
my test setup has 16 machines and bottomfeeds them successfully at 600/min, but just for you i added 4 more machines and did sidefeeding instead (apparently thats harder now)
Oh, I have a very simple example to try.
6 blenders doing diluted fuel, single line manifold, and then feeding into single line manifold of 30 fuel gens. Everything side fed at the same foundation level as the machine.
Using only pipes and junctions.
I can guarantee you the last fuel gens in the chain will get starved.
building an entire diluted fuel plant is not simple
here is my new test setup, does that fulfill all ur demands?
Oh did you see the screenshots I pinged ya with last night?#design-and-architecture message
if u send me something like that ill make it work with pipes. i dont mind getting saves in dms
Aite I'll cook something up tonight
yea i did. im too lazy to do that so ill just place them manually i think
is my test setup correct?
Yeah its iffy, but like the one workaround for BP madness
Ehhh... I see pipe elevation changes. Also water extractors sit at a different level than the machines
At least with that design I coppied, could delete the template parts for placing the extractor.
well, that should make it harder, right?
I'm talking about a baseline test that exposes flaws with Mk2 pipe at 600
I use two extractors to place a vertical pipe for my coal plant, the point between the machines.
i said i can fix everything just by adding pumps and u said i couldnt fix that by adding pumps.
or did i misunderstand what u meant?
here.
this setup here seems to be running flawlessly with 600 water/min. i didnt even add a pump.
oh nvm, the last one isnt being fed enough
Panic!
strange that this happens even when sidefeeding
nono, i said i can fix everything with pumps. i believe that. time to do my magic trick
Okay, fine. ๐
oh nvm, i remembered i didnt flood hte new ones
Rude.
it probably just didnt warm up yet
Here....
Pretend I have 6 blenders doing diluted fuel, and 30 fuel gens all down that line. You cannot fix this with a pump.
Them's fightin' words around here.
Or 31.
You can prefill the system, but eventually, the last few gens will starve.
I tried adding a single pump on the pipe between the last blender and the first fuel gen.
The only thing that helped me in such a scenario was looping the last fuel gen to the first one.
I bet looping the first fuel gen to the second fuel gen would work. ๐
Also, the first blender tends to have trouble emptying as well.....
So looping the first blender to the last blender also helps.
send me the save and ill fix it with pumps.
My old save had that problem 
or just look at my setup, i have 36 refineries and 8 blenders without a single loop.
hmmmm
Out of curiosity, is your framerate stable @ember fractal?
Yes, 90 fps
Just checking.
oof turning 20 machines on by hand is annoying
Mine ends up being much lower, ya know, running a NVIDIA GeForce GTX 660 Ti from 2012 
very curious, sidefeeding actually doesnt make it work. im getting horrible flow
My theory is that fluids require some leeway to flow at either a higher or lower rate than needed, and the pipes provide the buffer.
So let's say you have a mk2 pipe feeding a factory consuming 400/m. You'll see that pipe's flow vary between like 200 and 600 or something.
However, if your factory requires 600 all the time, that's the max already for the pipe, so there's no margin for error.
It can flow slower, but not faster, therefore unable to make up for any inconsistencies, whatever those may be....
๐
yeai thought sidefeeding was ez mode and made everything jsut work without pumps
Top feeding is better.
no, bottomfeeding is harder dw.
maybe topfeeding is the true easymode
It is, you're basically using the pipe to extend the input buffer of the machines.
last one dropped down to 16
the fourth to last is at 50 tho
now dan what would u say if i added pumps and they made it work flawlessly?
Then I will go back and try it again in my designs
I like that you said it wouldn't work and you didn't actually try it. ๐
But from all the pumps attempts I tried, it didn't fully fix it
first two machines idling. ill add pumps, not even flood it again. just pumps. and see what happens
I did try it many times lol
Ah. With how many pumps though?
Single pump between the fluid production, and the fluid consumption factories
So you just weren't ambitious enough.
Maybe...
๐
But logically, if adding one pump doesn't work, why would adding more work?
It's Complicated โข๏ธ.
Thanks, helps a lot/
yes, thats my point
I'm not a philosopher. I don't know the why.
the setup i posted has 8 meters of height differenece between the water extractors and the machines
the headlift of the water extractors should be enough, right?
You would think, yah
so any issues in flow shouldnt come from headlift
You're essentially bottom feeding?
Don't forget: junctions eat head lift.
no it was real bottomfeeding
anyway, with just the 10m of headlift from the 2 water extractors the water extractors were at 42 and 41% efficiency
then i added a mk1 pump. i changed nothing else, just a mk1 pump. after that they were at 71 and 74%
i added the pump at the height of the water extractors, so there were 20 meters of headlift instead of 10
then i replaced the mk1 pump with a mk2 pump
the efficiencies went to 97 and 89%
50 meters of headlift instead of 20, to lift fluid up 8 meters
while it is untrustworthy, this does seem to run fine
yknow what i did then? i added 370 meters of headlift. efficiencies went to 100%, everywhere
the large buffer is there to monior flow rate interruptions
mk 2 pumps on input and output
Buffer reeeeee
i always run my pipes 2m above the inputs and have very few problems
That's top feeding.
Yes, I do this a lot, especially on machines that have solid inputs or outputs.
@ember fractal after adding 20 pumps, one for each machine, this baby runs at 100% efficiency
Yah, I'm not adding 20 pumps to make a factory work lol
just by adding pumps i made this thing run at 600 water per minute constantly, with 20 machines. nothing else. no looping, no buffers, nothing
i mostly do it just to keep the pipes off the ground for walkability
i went overkill, 2 would porbably be enough
HMm ok
one at the beginning and one in the middle of the eating machines should work i think
depends on the length. more machines == more pumps. why? i dont know.
you either have to raise them up or drop them down to get them out of the way, and i'd rather not have a floorhole for every machine
This has also been my theory. I wonder if pumps are able to "smooth" out the variable flow rate somehow and prevent those drops below 600 from occurring/being as drastic. Would also align with 0laura's reports of one pump increasing the efficiency an amount, but not up to 100%, and another pump increasing it again, but still not up to 100%. We only get 100% with 370m of HL
i did some research on the effects of headlift on bottomfeeding, i didnt think they'd apply to sidefeeding too but maybe they do:
The longer the manifold, the more trouble it has keeping flow all the way to the end.
but the 370 meters ofh eadlift was by going up, there were no pumps between machines.
also, breaking stuff into smaller modules and using mk1's is something i favor
i used this abomination to get 370 meters of headlift:
Right, I wonder if it is the headlift provided by the pumps that smooths it out. That same headlift is supplied by the 370m tower
This is all just my speculation though
Perphaps the pressure in the pipe drops as the manifold gets longer and longer and machines have already consumed some of the water.
i know ur gonna say it made it work because the big pipe acts as a buffer, but if i placqe a pump at the end of it on the bototm (thus resetting the headlift) it stops working again.
so, adding pumps can also make it worse
So maybe it's the headlift that fixes things, not pumps per se
yep. if i have issues with flow now i just slap on needless pumps, even on horizontal sections.
Still think it's not Head Lift making it work. ๐
100% in theory we speculate that perhaps most probably with extreme confidence we can conclude that head lift fixes all pipe problems
100% in theory is a bit oxymoronic (deliberately so) ๐
very easy to keep things sized right
the fact that you can get it with non-pump-based headlift seems to indicate that headlift itself helps
still sticking to my "headlift is weighted as a variable when the game calculates which direction to send fluids from junctions" theory
tbh, to me it seems like headlift affects the order in which pipes are processed sequentially
Was just about to say something similar. It does make me wonder if HL is used for more than how high a fluid can go, but also "pressure" (I know some people believe there may even be some other hidden variable related to pressure we don't know about). My thinking is HL is still just the one variable, but is used for more than we think it is
But if everything is in a straight level line, wouldn't the headlift remain constant throughout, or at least remain at the value the output machine gives the fluid?
btw, one thing i wanted to note about pumps:
they are technically a machine like refineries and such
thus, pumps always try to output at the maximum flow rate.
Unlike valves where it depends on how full the input is
so, in a way, they are pumps, they do try to boost flow
but they cant boost it if they dont get it
it should
you don't possibly have 2 levels of stuff consuming on the pipe, do you?
I wonder if maybe that is where the "smoothing" comes in from. Because in the 370m headlift example, pumps were still needed to get the water up, and that may have smoothed the flow rate in doing so
something else i do for reasons of space is OC refineries and blenders to keep manifolds as short as possible
so the back end of a pump is basically a vacuum
it may not be actively programmed to create pressure, but since it greedily consumes any fluid that reaches it without passing it back, that ends up being an effect
i wonder, do pumps have internal fluid buffers?
also, i sometimes see pumps active and reporting flow of around 200 m3/min even when no pipe is connected into them...
Then having multiple pumps in a row stacks whatever smoothing effect they have to eventually fully smooth out the flow
maybe there's a data structure in one of the game header files?
Do we have access to the .h files? I wouldn't figure we did
We do.
Oh neat
In the community resources.
i think there's some out there, yeah
mod writers couldn't do stuff simply w/o that
Oh yeah, I guess that makes sense
the one thing that smoothed out flow for me in an output manifold was:
- use mk 1 pipe until flow rate was greater than mk 1 speed
- use mk 2 pump at the halfway point and at the final mk 2 output
consistent 600/min from what i can tell
but mind you this is packagers, where every sideline is mk 1
I wonder if putting a pump at every fluid producing output, before it goes into the manifold...
unironically considering putting a packager infront and behind every one of my machines...
just go back to the days before pipes were a thing. return to monkey update 3
i'm finding that the only place i don't have issue with mixed pipes is when i do this to split a 600 line in 2
or.... you know.
just go with the simplest way and dont max out your pipes
that one always works well. stay well below the limit
my pipes ARRENT maxed out, thats the thing.
If a machine is producing at 100% and is able to fully empty, then theoretically it should be able to feed a manifold of any length, right? Where else is the fluid gonna go hehe
yeah, well miner on a pure node...
im pushing 400 HOR through my pipes. and 400 fuel.
It can backup the producers so the fluid just isn't created.
why does it sitll not work without stupid pumps
Try the same but with 600
im easily pushing 600 oil through my pipes, with pumps.
i know you probably dont wanna hear it but.... mk 1 hidden somewhere?
pumps made it work tho
unless ur implying pumps can push 400 through a mk1 :p
each one of those sinks eats 800 plastic per minute. 6400 plastic from 2400 oil ๐คค
and it works only due to pumps.
im basically rebuilding my old oil factory, just bigger and with more pumps. and this time it actually works (because of hte stupid pumps)
youre a bottomfeeding lover tho so idk the game must just hate you
it works with bottomfeeding tho !
thats the neat thing. dubious amounts of pumps just make it work.
thats why it hates you
ive also noticed that sometimes valves make things work that they shouldnt make work, for some reason
brown is sub 600 flow, green is 600 flow, pink is fluid sink and yellow is fluid spawner.
its because they randomly remove the 'is bottom feeding for laura' bit that is set on the pipes ๐
this system is not meant to be analyzed on paper with a pen the way you can with belts and imo thats not good for it
placing a valve there shouldnt make it work, but it does
valves could, in update 4 even, enforce priority on lines without pumps on them
just by existing on that line
If pipe stands can, I don't see why valves can't.
pipe stands cant, i tried it :/
Oh, I thought it was.
at least i couldnt get them to change anything.
Fine, if junctions can then. ๐
only junctions still hold that power
something else that i'm finding is that really short pipe segments caused by splicing pumps or valves on a line can do funny things
the game has a minimum buildable pipe length for a reason, i guess
It also has minimum volume, so those short pipe segments all hold 5 m3.
question.... whats the shortest pipe in that setup
The real answer is just never ever move fluids upwards, if you have to, nuh uh you dont
uhh no clue, why?
im not even moving anything up the pumps are just there for emotional support
just go and find out, i must know
in my test setup or in my oil factory?
I have built all my plastic, oil fuel gens water and with 0 pumps after the first time they fucked up and literally haven't had a single pipe issue since
Fuck pumps
Problems solved!
in ANY i wanna know the shortest pipes in a system that works and the shortest ones in a system that doesnt
Sacrifice MOAR GOATS!
if you want: dont tell me their length but just their volume if thats easier to find out
Are you thinking like a machine consuming more than what the short segment feeding it can hold? So it ends up trying to suck extra from the main manifold?
5m3 would be the shortest one i think.
Idk I use literal .2 m long pipes or whatever the min distance Is always still 0 issue
its totally straight so u can probalby calculate the length from it
i have lots of 5m3 big pipes in my setup.
5 is the min, so you can't calculate the length from it.
Whatever sticking a junction basically into a output/input is
mk 1s and mk 2s?
its in the setup htats feeding my oil refineries, all the pipes that can be mk1 are mk1 for those refs
i didnt do that anywhere else, i think its probably unnecessary but why chnage it now
try to find any factory that doesnt function and try to remove mk 2s that are 5 m3 only
No. That's worse than junction magic.
if its due to the manifold being too close to the machine: welp cant fix that without tearing it up but it would be a good move
well this one does work and has 5m3 pipes
i think i had the brain blast i needed
this one here works too
you know, fluid flow is usually just a number and it moves fluid through the pipe
all fine and well
but.
what is 600 and 300 mยณ/min in mยณ/sec?
because the game does all the math in that range
10 and 5?
Here is my unified theory about what's going on with pipes.
- All pipe behaviour boils down to the function that determines which pipe at a connection/junction will give fluid, and which will receive it.
- Differential: fuller pipes pass fluid to emptier ones. This is what results in sloshing, as the system spends a bunch of time evening itself out, rather than actually passing fluid onward to the destination.
- Gravity: Higher pipes pass fluid to lower ones. This is why bottom-feeding works poorly; fluid does not want to go up the branch pipe until it's absolutely finished sloshing around to fill the main pipe.
- Headlift Limit: Fluid cannot be sent above the cap from a pipe that is below it.
- Headlift vs Gravity: The further the pipes are below the headlift cap, the less weighting Gravity has (ie, go far enough below the cap, and high/low pipes eventually get equal priority).
- Vacuums: Machines and pipes consume fluid without passing it back; they are automatically treated as the "emptier" pipe as long as they are not completely full.
Thus, pumps help where they "shouldn't", because they:
- Counteract gravity by adding headlift
- Create vacuums that prioritize forward movement
- Impede sloshing
yep
now, a little trivia about pipes
every pipe lies and is actually 40% bigger than what it says it is
in terms of volume
makes sense, otherwise full pipes wouldnt flow
so a 5 mยณ pipe is actually 7 mยณ
thats the overfill thing
overfill is for sustaining flow at max value
fluid decides where to move based on base fullness, but it can also fill overfullness, right?
at least htats how i plan to write it
overfill is for max flow stability afaik
if you dip from overfill down to normal fill, you enter the zone of sub-600 or sub-300 flow
the overfill on a 5 mยณ pipe is 2 mยณ
that is 200 milliseconds to empty
so what you're saying is that shorter pipes make the system more volatile and less stable?
on a pipe thats at least 7.5 mยณ, you get a total max volume of 10mยณ if you include overfill
and the overfill portion is 3 mยณ
im just saying.... you cant really fit 1 second of flow into a pipe thats smaller than 10 mยณ in total
I assume the update rate is at least 30 ups, which is more like 32ms/update.
and you cant buy yourself enough time / stability on an overfill portion that is smaller than the "best safety" imo of 10 m3
There's no way they calculate flow at the second interval, it's way too jumpy.
so a 25mยณ pipe would have the best overfill portion for mk 2 pipes
as thats at least 10 m3
but that shouldnt apply unless overflow is a thing, where flow can exceed 600
oh, almost forgot to mention: the overfill portion isnt FULLY used for pressure.
it has a percentage only thats used for pressure
the rest is safety
Is the issue worse for machines that are overclocked and consume a large amount of fluid in one go? Maybe the segment feeding the pipe doesn't have enough volume to provide a full cycle of fluid.
otherwise the only way to lose overfill buffer would be for the previous pipe to be below 600 flow.
and that could only happen when theres already an issue
all im saying is
Its good to have a bucket that can hold at least 10 mยณ when you get 10 mยณ every second handed to you
But.. I don't get 10 m3 every second.. I get some small fraction of that every few dozen milliseconds. ๐ฆ
ye
That actually make sense. Imagine 2 machine very close together, and they both take in 50 fluid, at the same time (synced cycle). The manifold might not have enough buffer capacity available at that point in time....
That's more or less why bottomfeeding is so hard.
i disagree, its just as hard with constant consumption
buffers make sense. without a buffer, I figure the game would have to calculate all pipes sequentially from emptiest to fullest each tick. and performing that sort seems not efficient
but again, 5 isnt a magic number. and all my systems have a 5 in them and they all work.
if that was true then none of my systems would work. but they do.
5 is only magic in that its the minimum pipe volume
and its the default junction volume
imo junctions and pipes should be 10 mยณ minimum
i dont see how that would fix anything.
some of my systems that dont work had no pipes that small, all my systems that do work have pipes that small
remember that fluid volume also affects how fast you accelerate towards max flow
It's the local buffer amount. It needs to be bigger to smooth out flow.
a bigger volume for everything would also dampen jumpiness in flow
Also take longer to fill up. ๐
again, i have seen 0 correlation between systemfunctionality and lowestpipesize
Same as how belt splitters and mergers have internal buffer that smooth out belt flow
yes thats just pipe life for you
not good for long distance
btw heres junction manifold nightmare test
Pop quizz: whats the flow on the very pipe at the top if 600/min are supposed to enter from the bottom
Okay, so after a bit of poking around the header files to see how the pipe subsystem may work, here are my thoughts:
There is only content, flow and pressure. It seems that pressure is what we call head lift. Those are the "high level" values often used in debugging and simulation.
Fluid is simulated using fluid boxes that have volume and pressure. Those seem to be the only things that determine how fluid flows. Pressure is derived from three sources it seems:
- It can be added by a machine such as a pump, or from gravity
- It can be added by overfilling the fluid box. Fluid boxes can have their volume overfilled by a certain percentage (the header has 40%). This varies based on the fluid box. Some of this overfill can add to the pressure based on a global percentage. This is how we achieve normal flow by adding fluid to the system.
- There is a global dampening factor that causes a loss of pressure applied to the fluid boxes.
Interestingly, when we have talked before about junctions removing HL from the system, I am starting to think it was more due to how fluid boxes dampen pressure. But we could really only observe this by using junctions, hence we believed it was due to junctions. This may also be why the dampening factor seemed to be in a range. In reality, it is not in a range, but since we could only view that factor at relatively wide steps, our measurements are inconsistent.
what is that. thats satanic.
also like the pespective is really trippy
uhhhh, 600? the pipe indicator seems to show full flow
oh nvm i misinterpreted the pipe indicator. yea idk, id expect horrible flow because similar way simpler setup have resulted in sub 600 for me.
here for example, the slushotron9000
answer iss.... 0
yea, doesnt surprise me
Y'all should just do like me. I put all my fluids into 300 Industrial fluid buffers. I call it the water tower.
a lot more
ive noticed that before. pumps get wonky like that in those situations. ive seen them display 2 meters or more on a flat plane
i just brushed it off but i think u just uncovered that its a bigger thing
definitely above 55, right?
no
?
wdym its less than it should?
i thought it should be 20
well on that manifold maybe
but not on that giant tower behind it
thats 53 tall
nope
???
im missing 6 m head lift
yes to this, no in the sense "indeed, it does not see a continuous connection"
can anyone show me a 6 to 6 load balancer?
well... why do you need it? ๐ค
i want my 7200 alu scrap to flow properly
just hook each belt to machines that need the amount that is on the belt
no reason to balance
hm
A B C D E F -> (AB) (CD) (EF) -> AB AB CD CD EF EF
then,
(AB CD EF) (AB CD EF) -> ABCDEF *6
so 3x 2:2 balancers followed by 2x 3:3 balancers lol
but yea no need to balance if the clock speeds are the same and they are fed the same amount
3x 2:2 balancers followed by 2x 3:3 balancers lol
i can work with that
What is the best beginner setup for a turbofuel power plant?
there's a 3D example on the wiki
those pics look quite convoluted tbh
I want to say a normal oil node can run 12 fuel generators with 2 shards apiece, but I forget the amount of compacted coal needed
That's a normal node with 3 shards on the extractor
what Alts do you have?
Heavy oil residue
no packaged diluted fuel?
IMO do diluted fuel and skip straight to rocket fuel. If you want to do turbo now though, diluted fuel will make it better anyway. Diluted fuel is one of the more powerful alts in the game.
The thing you're asking for is quite convoluted.
reggt
Aluminum scrap is a weird place to be fussing over balance, as a manifold that should saturate very quickly if you just put an ISC at the end of each refinery while you're putting up the smelters/foundries.
By the time you figure out balancing it, if you had just done a manifold it would be saturated and producing ingots.
i don't suppose the online map has a good way to figure out what machines i left all my sommersloops in?
Need help with math.
I'm unable to properly figure out how to sloop a constructor that makes 30 reanimated sam to produce an extra 25 reanimated SAM. Like how much do I overclock it?
I'm curious, how do people determine how many factories to use in a manifold?
Under clock to 27.5 per min and sloop to make it 55 per min which would be the 30 + 25 needed
uhhh
depending on interpretation, do you want to use up all the inputs (impossible w/ 1 constructor) or do you want to underclock the miner? (what Ryan said)
whatever fits the best in my chosen footprint in my case
that and belt speed
Make it possible with one constructor please. I'm already using 2.5 constructors, and I'd like 1-2 perfectly math clocked to work at the ideal speed. The fewer machines I have to alter the math on the better.
exactly as many as I need for next step
then what ryan said, 27.5 x2 = 30+25
Hm. So for iron plates with a MK2 miner and T3 belts... actually, should I have one pure node be dedicated to all those iron plates?
The math on this is fuckin' weird, but I trust yall.
So I still have 12.5 constructors, but two of them are just underclocked+slooped?
think so
you don't really need more than like 15-30 per minute plates
Yeah if you want it to be one machine at 27.5 slooped for the 55 but if you want 2 perfect math machines then it would be different
not sure if I parsed this sentence correctly
one constructor at (27.5/30) = 91.667% will make 55 reanimated sam per minute if slooped
That does help clear things up, thank you.
how many drone ports would i need to transport 200ppm about 6km? thank you for your help I cannot find much info on the math for drones to belts
the thing is drone ports will dynamically calculate their round trip time and it depends on distance and fuel type
Unrelated um it had a stroke
if the distance is certain, try just building one and put in the fuel you'll use. I think a single port will handle 200/m easily but if not, you'll just have to add a second anyways?
heard, thank you for the assistance. i appreciate it
known bug due to float math imprecision, can safely ignore
definitely enough
I have a lil problem in my pipe setup. left is extractors, pulling 530 per min. there are 3 sets of these extractors, hence 3 pipes. the go through a mk2 pump, up a water "tower" and back down. then, that water gets split into two, where one side uses 320 water, and the other is combined with the other "overflow" from the other two pipes to make one last pipe of 320 for the fourth partition of buildings. each partition of buildings has its own loop, where the red box is how the pipes are set up, with refineries on the left and right side of the boxes, pulling from there, and the top and bottom connecting both sides. the yellow lines indicate the "overflow" from the three pipes going to the fourth segment of refineries. there are some elevation changes within this setup besides the water tower, but nothing that goes above the water tower, and some at the same level as the bottom of the water tower. the buildings are a few metres above the bottom of the water tower. I am running into issues with pipes showing 600 flow, some showing 0-10, and im unsure where to begin troubleshooting
greeny your calculator is going a little haywire
Best control rod recipe?
Cross posted in #satisfactory
i have the rocket fuel recipe unlocked in the MAM but dont have nitric acid in particle enrichment can i still get the rocket fuel alt from hard drives?
it's not indicated here
looks like it only checks for blenders, possibly, otherwise you could unlock this as soon as you have rocket fuel unlocked in general
YES!
I did this and had the same question, but then I was making RF before I could make nitric acid
I save-scummed like a madman
yeah i used like 7 hard drives didnt get it
went for another run now
well atleast my depot storage is maxed now too
Interesting, I'm still toying with how I want to do nuclear, I might end up at 36 uranium rods using 1500 uranium, and 21 plutonium rods using 500 uranium to mix in fertile uranium. No instant cell, but yes to the plutonium fuel unit alt.
how many nuke plants is that
180 uranium, 210 plutonium. But some amount will be allocated to drones.
good lord
i finally got supercomputers running yesterday so i can finally do nuclear but i absolutely don't know what sort of scale i want lmao
pick a size between 300, 600, 900, 1200, 1500, 1800, and 2100 and use that much uranium, everything else rattles out from that
I'm in a weird place of trying to figure out how to generate around 900GW of power, as someone who doesn't care about just storing waste. Maybe I should just store uranium waste?
you're probably not going to get much better than 600gw from uranium alone unless you convert bauxite
btw I just noticed that the nitro rocket fuel can be unlocked in earlier milestone than the original recipe ๐
I guess I need some sort of plutonium power either way, yeah. Maybe I'll burn 18 of the 21 plutonium rods so I have an equal number of uranium and plutonium reactors. That's 900GW of power, which should be plenty with strategic underclocking of particle accelerators and so forth, and still gobs of drones.
because you need just gas instead of acid which is unlocked in particle acceleration
yeah, you can't really make it earlier though since you need the well pressurizers
oh, oops, i gotcha
yea but the well pressurizers can be unlocked earlier, together with cooling systems and fused frames
The game has many similar scenarios IIRC.
i'm still working off of turbofuel, lol
I really should've started building my rocket fuel factory around nitro instead of default ๐
done with phase 5 and like never bothered even with rf, lol
I think lots of refinery and foundry recipes are unlocked before you can build those?
tbh, default is better when coupled with blended tf
yea, but pure/alloy recipes are unlocked before the respective building, while nitro can be unlocked after unlocking blender but before unlocking nitric acid
I think I remember other stuff being gated awkwardly by that same milestone. I could just be remembering the same thing you're observing now though.
hi, what makes blended turbofuel more often preferred over diluted + the basic recipe?
so there's like shorter gap between unlocking the recipe and being able to utilize it
All forms of turbofuel are good with diluted fuel.
side-by-side, it isn't a simple comparison
I know, I am asking about the perceived preference
Turbo blend is oil+sulfur, default is oil+coal+sulfur. Mostly it's personal preference based on that.
nitro rf uses a lot more sulfur and coal whereas blended tf+default rf uses more oil and no coal for a fraction of the sulfur
blue crater is probably the place you wanna build it one way or another. you have 2100 sulfur right there, 1500 nitrogen, 2250 oil and 2400 coal all within belting distance (along with enough iron for the nitric acid plates), so you're build needs to fit within those parameters
If I have production numbers, with a parenthesis, does that mean that's the true number?
you'll find that if you stay within those limits for short range logistics, you make more power off of default nitro because it is kinder on nitrogen
is it dumb to train in a bunch of resources as a mixed bag and just use a smart splitter to break it up? logic is to have it all easily coming in at one place?
you can but you run into some horrible latency on delivery and need to set up a buffering system on your sorter output to stabilize that
on top of that, belt speed caps what you can import pretty quickly for things like quickwire, concrete, etc
It's the buffering that you need to be sure is set right. Otherwise a backup of one item backs up the whole line.
i've done it on small scale with trucks and tractors, and really you run into belt speeds limiting things pretty quickly
another thing you can do is set trains to only drop off specific items at train stops and do some pre-filtering
i've never put that strategy to the test, but keep thinking about it as an option
If the cycle time on a building is an exact number of seconds, does the repeating decimal in a clockspeed not actually matter?
I've seen that be wonky on water extractors, so I might play with it on aluminum.
sounds good didnt even think of the buffering or belts, makes total sence, could dual belt it top and bottom but that sounds almost more difficult than just having each freight container / train freight item specific
yeah, the belting of it all gets pretty tricky
When I've done it, I set the smart splitter to [product], any undefined, and overflow. You can get by without the overflow setting if you're confident the stream will never back up because there's a sink somewhere down that line.
It's generally a lot less work to just make another truck/train stop though.
good point sounds way more of a hassle than just separating it via train or belting
Sorry had a few calls come in. Appreciate the wisdom gents
don't overlook train's filtering ability on pickup/dropoff
i was building in creative mode, but i got bored and automated powershards lol
a loop like that is how i made the dmc's for mission complete
my inputs are just coal for diamonds/tc's and then quartz for the shards
uses like 40k mw
im having an issue with liquids and was wondring if anyone had a fix. I have 2 groups of 77 fuel generators running. both are set up exactly the same. one runs fine and the other is struggling to run 4 of the gens. for the life of me i cant figure out whats going on. its not a headlift issue, as theyre everything is on the same level. anyone having similar issues?
with some sloops the dark matter kicked off makes a ton of crystal
rule #2 of pipes is keep pipe manifolds small
ah damn... little too late for that lol
if you gotta long ass pipeline you might be able to use valves to prevent slosh
time for another 2 rows
try disconnecting the generators from power and let the pipes fill until the flow rate in them drops to zero before reconnecting the generators
or you can use buffers and let them fill befor you start the generators
alright, ill give both of those a try. thanks!
how do I stop satisfactory tool from making 22 computer byproduct
gotta use more of the computers
or make less
Define "small," please.
relax, I had this ๐ฌ #math-and-meta message
what the fuck, how, what is satis tools on
on mushrooms bacon agarics, I'd assume
stais tools looks useful but i would just feel like someone else is playing the game for me
if you consider raw math a game, then yea, kinda
yeah i know, im crazy
fair enough, I mostly use it for large factories that would burn out the last brain cell I have left and figuring out which recipe combos work best
blender recipes are based
radiation?
which ones are worth taking
Not using maximize should be enough
All of them.
Yep, I'm just gonna buy a second golden nut. Forget the rest of the statues. Nuts are way easier now, but I guess because it's an achievement it needs to be easy or else there will be an avalanche of complainers.
I wish I could take them all ๐ฆ
It's hard to give you advice on alts without having any sense of what you're planning to do.
Pure aluminum is the only one of those I'd pick without a second thought.
Outside of pure aluminum, depends on what, where and how much are you going to produce
WIthout knowing what you don't have, there's no way to estimate the value of a rescan, either, but I'd be hunting for diluted fuel if I didn't have it, for instance.
I usually don't question "is it worth it".
I question "Why it's not what I need"
the fuel recipes are nice
You can, just get more drives. Every single recipe can be gotten.
For some people that uranium alt is going to be essential. But if your goal is just to get the last delivery done and all achievements, nuclear power is just a distraction.
the ones that skip production steps are nice
instant scrap:
yes
These are the ones I use:
Recycled Plastic
Radio Control System
Caterium Wire
Pure Copper Ingot
Pure Aluminum Ingot (this one first)
Instant scrap is not simpler IMO because you're routing another resource, and the sulfuric acid step basically makes it the same as alumina -> scrap.
nitro rocket fuel
Is there any round perfect ratio for rocket fuel production? This is a nightmare
diluted fuel
no blenders arent simpler
They really are :/
You're skipping the alumina step and adding the sulfuric acid step and dealing with 3 raw resources instead of 2.
It's a perfectly fine recipe, but it's not simpler.
@vapid gorge unslooped and looking good. we are now slightly overproducing, but only a tiny amount (up to the nearest round number), 702 vs 720, so overproducing by 2.5%
still a few machines waiting to saturate, but I think they'll fill up eventually
this is essentially the plan of the tower
There's no way you did not do that on purpose
Don't look up the hex to decimal conversion of my pioneer designation ๐ ๐คฃ
Pics of gen tower-
#design-and-architecture message
80 fuel generators takes up so much space. How tf are y'all organizing these
By going to places with space?
I have MK3 blueprints and use this, but I'd just break it into 2 pieces if I were limited to MK2 blueprints still. #design-and-architecture message
stack up up up
Maybe I should go unlock MK2 before I keep building
8 gens a floor, that's 10 floors, easily below build limit
For sure get MK2 so you can do 2 generators side by side.
Then also you can preoverclock them.
are they already overclocked?
nooo
just do it
Should I be overclocking these guys
That shrinks down to 32 generators if you clock them to 250%.
should sloop your powerslugs too. i have like 400 shards rn
That's more shards than I have tho
Sloop 10 purple slugs and you're there. Raid the red forest.
theres a bunch on the northwester part of the map
Or raid the spire coast, they stick out easily at night
I'm not sure I have slooping tech yet
shards are infinite resource later on, just spend the limited ones you have on fuel gen
its worth it
isnt pure aluminum a terrible choice? wouldnt you want to use silica to get more alu ingots per bauxite?
and how many nodes on the map of quartz do you think there are
it hasn't flooded yet?
Useful for saving a lot on complexity, footprint etc