math-and-meta

[devout ingot]

i could share it with you in dms if you want to see how ridiculous it is

[fallow siren]

also, better just burn rocket than ionized

[devout ingot]

yea i know

[fallow siren]

with that amount, youre going to use more power making shards

than making net positive

[devout ingot]

would you like to see the model i have?

[fallow siren]

sure, just share it here

[devout ingot]

rules say i cant do it in server just dms

[fallow siren]

what?

youre sharing your production line, no?

[devout ingot]

yea just sent it in dms

[craggy birch]

the production lines in my creative world as of now
yes i got caterium before copper

[keen night]

should i be even thinking like this at phase 4?

[vapid gorge]

there's no names of anything on this. lacking any sort of labels. Modeler is not really something you can share info with

it's a visual mess to anyone who hasn't made the plan

[keen night]

Ai limiters, all parts for phase 2 elevator, advanced plates, steel beams, steel pipes, encased beams, rotors and iron frames

[vapid gorge]

if you need the things you need it?

but yeah, no info on numbers, or which recipes of what you're doing. Just don't expect people to pick apart a modeler plan and give advice on anything more than a 2 step process

[frosty owl]

What would be wrong with it? I don't really get what you're asking ^^

[still blade]

what are the benefits of the Bolted Iron Plates recipe compared to the basic Reinforced Iron Plates recipe? they both take iron plates and screws, and the alt uses 50 ppm

[wind spade]

more production per machine

[still blade]

interesting

still gonna automate Stitched though, i dont wanna deal with screws unless i have to

[unique cypress]

Compared to default plates? Bolted takes up less space

Compared to stitched plates though? Bolted has basically 0 benefits

[still blade]

stitched seems to be the go-to, especially when theres the Iron Wire alt

[wind spade]

uses screws instead of wire

[unique cypress]

Adhered is also decent

[still blade]

yeah but i dont have rubber yet

[hexed hornet]

what website for this

[keen night]

Satisfactory modeler on stean

[hexed hornet]

thx this game shits me & kills me lol

[delicate cradle]

Oh my god that’s a thing?

https://tenor.com/view/ash-ashh-baby-ia-gif-16729153814109504694

[delicate cradle]

@honest grove

Those white lines are mk5 belts, not 6

[honest grove]

thats so much fuel, might as well jump straight to turbo or rocket fuel

[grim citrus]

is it ok to send a pic of my alternatives to see which ones are worth claiming?

[mossy ibex]

Hey I read an analysis once that categorized assembly parts into those that pay forward into future products chains and those that don’t, eg smart plating is used in future steps so should be automated permanently. Can someone link me to that analysis? I can’t seem to find it now

I’m currently trying to plan the layout of an assembly tower and I want to know which floors will need to exist permanently and be periodically expanded vs which I can have basically temporary setups for

[unique cypress]

!wikisearch space+elevator

[brisk shore]

https://satisfactory.wiki.gg/index.php?search=space+elevator

[unique cypress]

In

Total number of parts needed

This assumes no sloops used though

[fervent shale]

@dusky dust as i posted before (using sloppy alumina and pure aluminium ingot)

the top path is the reused water

this is without that reused water

(i used an extractor cause moduler doesnt like when you bring stuff backwards

[dusky dust]

I'm.. sorry, but I really really loathe trying to decipher Modeller graphs

It is the absolute worst graph I have ever seen for conveying information to people

Write out how much Bauxite you're using, and what recipes you're using to process it. :)

[fervent shale]

300 bauxite,that will need 300 water for sloppy alumina

then i will get 360 alumina that will tak 180 coal to make aluminium scrap and give 180 water

[dusky dust]

What recipe are you using for the scrap?

[fervent shale]

base

i think other than the recipes being chosen, the graph makes it easy to understund?

[dusky dust]

So yeah, indeed, you need a total of 1.5 Sloppy Alumina refineries to process the 300/min bauxite. That'll give you 360/min Alumina Solution, which'll go into 1.5 Aluminum Scrap refineries, which produce 180/min water, so you're right on that.

[fervent shale]

(ngl that msg felt like something chatgpt would say XD)

[dusky dust]

The 1.5 Sloppy Alumina refineries will need a total of 300/min water, so yeah, you need 120/min fresh, and 180/min recycled will take up the rest

[fervent shale]

so 2 refineries total is correct?

[dusky dust]

So yeah, in that case you'd be able to get away with just two Sloppy Alumina refineries

[fervent shale]

just needing to underclock them maybe

[fervent shale]

YAY, i know how to math

[dusky dust]

One will be clocked at 60% (to take in the fresh water), and the one taking in recycled can be clocked to 90%

(.6 + .9 == 1.5)

[fervent shale]

OR, just let them both at 100 and let the bauxite handle the ratios

[dusky dust]

No, the point is wanting to keep the fresh + recycled water totally separate

So you want them configured to take in the exact amounts of fluids that you need

[fervent shale]

what i mean is

i prioritize bauxite to go on the recicled refinerie

[dusky dust]

(Technically, yes, you could run them at 100% instead, and have them go idle occasionally, but your system's gonna be a lot more stable if you clock them exactly as needed)

[fervent shale]

so that one always has bauxite in

and the second one takes fresh water gets the bauxite leftovers

[dusky dust]

There's honestly no reason to try and prioritize bauxite. The bauxite will stabilize very quickly

The 60% refinery will "back up" and the 90% one will start getting what it needs

[fervent shale]

to make sure bauxite doesnt go 150/150

[dusky dust]

It's the fluid that you need to be careful with, and you want those clocks to match exactly what they should be

[fervent shale]

you are restriting on the refinires

i am restriting on the intake

[dusky dust]

That doesn't matter at all; the one at 60% will be getting more than it needs, so it'll "fill up" and then the bauxite will self-stabilize

This is exactly why manifolds work just fine

The clocking is to make sure that your fluids are being handled properly. The solids will take care of themselves; no need to do anything fancy with 'em. Just a splitter in front of the two Sloppy refineries will work just fine after a tiny bit of initial stabilization

[fervent shale]

but the thing is

[dusky dust]

One other recommendation: clock the Aluminum Scrap refineries to be exactly what the Sloppy Alumina refineries are, and you can just run Alumina Solution pipes directly from one to the other

[fervent shale]

if for some reason there is more water in the system , the recycled refinerie being at 100 means it helps the system to clean itself

[dusky dust]

No need to complicate it by combining the Alumina Solution into a single pipe

[dusky dust]

Keeping the recycled + fresh water separate means there will never be more water in the system than there needs to be

The Aluminum Scrap refineries are producing 180/min water, and the Alumina Solution refinery using recycled takes in 180/min water

[fervent shale]

yeah, but if for some reason it stops, then some water acumulates

[dusky dust]

Yes, there may be a brief time before the manifold fills up where a machine might go idle, but it honestly won't matter

If you are that worried about it, just prefill the Bauxite input buffers on the Solution refineries

[fervent shale]

and it can break the system

[dusky dust]

Well, I mean, it sounds like you really want to do bauxite prioritization anyway, and I agree it's not going to hurt if you do. The system will stabilize regardless

But I'll just say once more that you honestly don't need to worry about it. I've been building aluminum factories since U3 and I've never bothered

[fervent shale]

that, and i want to avoid doing underclocking

[dusky dust]

Just manifold the bauxite in, it'll be fine.

[fervent shale]

i do know the ratios now

[dusky dust]

Heh, you're in #math-and-meta , you're gonna be advised to underclock. :)

[fervent shale]

but the thing is, will the system break if i dont?

[dusky dust]

Probably, actually, since your water usage is gonna get out of whack

But if you're dead set on it, go ahead and try and see if it works for you

[fervent shale]

well, even if i set the reused water to 100 and the fresh water to 60?

[dusky dust]

You asked for advice and got it. :D If you want a simple, stable aluminum loop which avoids basically all fluid problems, split the fresh+recycled, and underclock/overclock to suit.

It's up to you if you want to actually take the advice or not. :)

[fervent shale]

what i want to understund is: how important is the overclock/underclock

[dusky dust]

(Though I'll say again that trying to avoid underclocking/overclocking is, in the end, very self-defeating. Clocking machines is one of the most effective tools in your toolbelt, in this game)

[fervent shale]

because what i always did was let stuff stay at 100

and manifold just balances everything out

[dusky dust]

Well, as I say, Try It And See™. You've got the advice. Cheers!

I suppose I will also say that one huge advantage to clocking things appropriately is that you can then use the machines' status colors to gauge if you have any problems in your factory. All Green All The Time == everything's peachy. If anything ever goes to yellow then you know you have a problem to solve.

If you intentionally let machines go idle it's gonna be a lot harder to know if everything's working as it should or not

(And I personally would definitely not want that in a system where I have to process fluid byproducts)

[dusky dust]

xkcd.com/3126 :)

[fervent shale]

i didnt said you were XD

[dusky dust]

I just felt very Seen™ by that xkcd is all, and it came to mind. :)

[fervent shale]

lets hope it works

(ik, i still need to make the ingots

temporary sink while i make the next floor

[silent shoal]

i need help🤡
I produce 600 rocket fuel.
57 generators are connected to the pipe at 250%
And another one at 100*(6.249525/4.16667)=149.988480009216
(I put the OC into the generator as an equation.)
But it still turns out that the consumption is less than 600 per minute (something really small)
And there is a drop in the efficiency of the blenders that make the fuel (they fill up and go to idle)

4.16667= https://satisfactory.wiki.gg/wiki/Fuel-Powered_Generator#:~:text=14.4 sec-,4.16667,-m3/min

[unique cypress]

600 RF/min is exactly 144 gens at 100%

so 57.6 at 250%

or 57 at 250% + 1 at 150%

[silent shoal]

144*4.16667=600.00048
Am I missing something?

[unique cypress]

it's 4.1(6), not 4.16667

[silent shoal]

https://satisfactory.wiki.gg/wiki/Fuel-Powered_Generator#:~:text=14.4 sec-,4.16667,-m3/min

so, they rounded it up here and the consumption at 100% is 25/6?

[unique cypress]

yes, it's 25/6 exactly

you can calclulate it yourself from the energy value of RF and the fuel gen's power output

[silent shoal]

🤡 I thought they were nice and just set it up like that... and wow thanks

[light fox]

this jawn a good or bad idea

[silent shoal]

I would appreciate a brief explanation

[light fox]

oh its balancer man

hi balancer man

[haughty barn]

Bad

Besides the fuel

[unique cypress]

fuel gen produces 250 MW - so it consumes 250 MJ of fuel per second

rocket fuel contains 3600 MJ/m3

so a generator burns 250/3600 = 0.069(4) m3 of RF/s

times 60 for per minute consumption and you end up with 4.1(6) m3/min

[unique cypress]

hello

[tiny leaf]

is it better to have a full 300/min pipe or half full 600/min pipe

im guessing for 300/min mk1 would be better because full pipes = more pressure right?

[dusky dust]

I think the official "best practice" is to stick to mk1, though it's when you're right at a pipe's throughput limits that problems with sloshing and stuff can start to rear their heads

[devout ingot]

cant you use mk2 pipes anyway, you build less pipes and can just wait for the system to fill then turn it on

and of course have buffers

[wind spade]

Don't use buffers

[devout ingot]

forgot to say a storage at a higher elevation at the end of the line past the machines or is that not how its suppose to be done?

[crimson moat]

mk1 moves half as much fluid when it has the same amount of pressure, i.e. it needs twice as much pressure to do the same job. That's almost always a bad thing.

you should always use mk.2 unless you're trying to split fluid unevenly and you want one path to flow much faster than the other, in which case putting mk.1 on the side that you want to be slow and mk.2 on the side that you want to flow fast can reduce sloshing.

A 2-way junction for example will split 300/300 with mk.2 on both exits, but if you use mk.1 on one then it will do 200/400.

If you're trying to take 20 out of one side while sending 240 fluid out of the other junction exit (like a fuel manifold), the second case is preferable. A 33/66 split is closer to your ratio than a 50/50 split, and mk.1 pipe to a generator which consumes 20/300 fuel is fine.

Having your demand be as close as possible to the natural split numbers reduces oscillations (zero, if perfectly matched) as unmatched numbers cause the under-drained side to fill to a higher fill level (and thus pressure) than the other output, which causes its fluid to flow back out through the junction to equalise.

That equalisation actually works very well and helps us often, but it makes fluid flow count twice against the flow rate limit of the pipe so the pipe has to have enough headroom to handle that.

[tardy stream]

is there any way to limit an output of a belt to below the rated speed? Like if I only need 4/m or am I enslaved to forever using 60/m belts for all of my lower-rate items?

[unique cypress]

I use the fastest belts I have available for absolutely everything

[devout ingot]

im not spending 20 minutes to maximize the use of each lower tier of belt for each recipe for blueprints lmao

[vapid gorge]

info on planners? xD yes.
I cannot tell if this is sarcasm or not because people have murdered satire.

[delicate cradle]

;-;

[devout ingot]

or just spend a few weeks on satisfactory modeler like i did lmao

[vapid gorge]

clock the machines you want to put the right number of parts per min on the belts you want

[vapid gorge]

no just don't, just don't use buffers with pipe systems. They don't help and will often cause issues, is likely what greeny is talking about

[tardy stream]

strange hearing this in the channel with the description of "discussing material ratios and so forth" I thought all y'all nerds would maximise profits by not spending excess resources in your schematics/builds

[thorn trail]

time is a resource too, and by always using the maximum speed belt available you can avoid spending time debugging your factory/blueprint to find the one place you used the wrong belt speed. Some people prefer optimizing for lower build costs, some prefer optimizing for less debugging time.

[tardy stream]

but don't you want to reduce the costs of your schematics that you'll be copy-pasting?

[devout ingot]

lets say i want to use mk1 belts for 1 recipe based off a template i made earlier, i would have to change it for other types of belts for different recipes manually based off the template, sure the manifold would technically fill up faster but just using the fastest belts is just more convenient

[thorn trail]

belt build costs for blueprints are a one time expense, and in most cases are a fraction of the overall materials cost

[tardy stream]

If you load balance, each recipe doesn't change the rates it needs resources unless you add overclocking, I make a blueprint that takes in 240 steel, it'll only ever need 1 240 belt inputing steel, no?

like it's not hard to not mess up belt placement and tiers

[unique cypress]

The manifold would not fill up faster with slower belts

[devout ingot]

it would reach the last machine in the manifold faster i meant

[unique cypress]

Not the case either

Unless the belts are exactly matched to consumption

[devout ingot]

yeah you have a good point

[unique cypress]

I did the math on this once and the fastest filling manifolds are ones with smart splitters that fill the machines 1 by 1

[sharp lark]

Are we allowed to post links to reddit in here? I am having an issue with VIP junctions (because of course I am). I made a post on the reddit forum and some people have been trying to help but no luck in fixing it yet. I would love to get some more eyes on it.

[tardy stream]

you use high tier belts for an item only coming out at 15/m? you don't take the time to swap to lower tier belts? compacted coal is only consumed at 7.16ish /m, you use a 480/m belt for that?

@unique cypress

[unique cypress]

What does swapping to a slower belt gain me?

[amber umbra]

A lot lot of people just use their max current belt for everything.

[devout ingot]

you mean using overflows for each smart splitter?

[tardy stream]

less resource wasted on an unnecessary belt

[thorn trail]

by the time I would be using compacted coal for anything ( in my case, rocket fuel ), belt materials shortages really are a complete afterthought

[unique cypress]

Nodes don't run out

[tardy stream]

but your resources only come in so fast

[unique cypress]

I have infinite belt materials

[tardy stream]

and at the start of the game reinforced plates aren't infinite/m

[unique cypress]

I am making them faster than I can use them

[tardy stream]

oh you must handle power and route all your shit immediately, while also cutting down trees, while your buddy is also spending resources

[unique cypress]

You also don't use that many belts in the early game

[devout ingot]

dedicating like a few nodes for just conveyer materials for the highest belt you have unlocked lets you just not care about having to switch to lower belts

[unique cypress]

Switching to an "appropriate" belt wastes a lot of time not just on actually switching to the belt, but also on figuring out what the "appropriate" belt even is

[tardy stream]

we might play differently at the start of the game lol

[amber umbra]

It’s a sandbox game so caring about specific things is fine. I guess caring makes sense if you just play the game by speedrunning phase 1, 2. Otherwise nah.

[tardy stream]

its more mental, just wired to not waste resources on a belt that I can easily swap to

[amber umbra]

Wasting resource vs wasting time, mental effort

[tardy stream]

I believe in-game they even have a key to swap to different variants of the same item held in hand

[devout ingot]

holding e while building yea

same for blueprints under a category

[tardy stream]

so easy to swap to a lower tier belt, ain't even gotta open the menu

[unique cypress]

Still slower than just using the one belt I have in my hotbar...

[wind spade]

No, buffers are pointless for anything other than train platforms

[light fox]

45 foundries might be a lot

getting into that stage where build projects become headaches

[orchid brook]

ok fr this my first real time planning a factory how will did i do?

i wanted to use that 1:3 oil to plactic/rubber

can anyone help?

[oblique hollow]

This jsnt the 1:3

1:3 exclusively uses the heavy oil residue alt, diluted fuel alt and both recycling alts

The recycling part only works by feesing into each othee

You used the polymer resin recipe and turned resin into plastic - neither recipe are among the most efficient in regards to oil-to-plastic

Resin to rubber is prefered because it has more output

[vapid gorge]

https://www.satisfactorytools.com/1.0/production?share=cnssFRB8ITQbthiK9HT8

[orchid brook]

oh cool

thanks this gone help a lot

[vapid gorge]

modeler is.. certainly a program that people use.

[orchid brook]

that is way my math wasnt mathing i thought they both had the same output

[oblique hollow]

No they dont unfortunately

[vapid gorge]

I don't even know how you got to your outputs

you should b e able to squeeze 1800 from 600 oil with those recipoes

[oblique hollow]

And neither does the polymer resin alt make as much of an impact as the heavy oil residue alt

Poly resin is only a decent intermediary choice if you just want to do it for the sake of it

Turn all resin to rubber, then use the Heavy oil you get to recycle some rubber to plastic

I did that once.... needs a ton of refineries because the residual rubber or residual plastic recipes are so damn slow

And its only at like 1:2 (rather 1 to 1.75) at best when it comes to oil to plastic / rubber

[light fox]

oh no

[oblique hollow]

well

best get building some more generators

its only 4 fuel gens so you should be fine

[light fox]

dang only 4

[oblique hollow]

250 MW per gen

[light fox]

i was gonna do 200

[oblique hollow]

200 is lunatic levels of generators but if you wanna do it lol go ahead

[light fox]

im working on it right now

[oblique hollow]

but if you are gonna do it: keep your pipe networks as small and seperated as possible

[light fox]

believe it or not the two systems are gonna be connected

[oblique hollow]

even 5 x 600/min pipes is a lot to get working right

[light fox]

this is only my foundries at the start

[oblique hollow]

i suggest you make 25 generators per mk 2 pipe so its 500/min fuel

easier to work with than a full mk 2

[light fox]

im looking at these building numbers right now and realizing how absurd it is

but im already in too deep

like look at this that is WAY too many buildings

[fallow siren]

you could easily remove these machines if you plan to sink all the resin anyway

[light fox]

thats to get all the plastic containers i need for the packaged diluted fuel recipe

[fallow siren]

you only making 10 and those 10 goes into sink

[oblique hollow]

btw, for packaged diluted, the recommended strat is a bp per refinery with 2 packagers so its a single loop per refinery

[fallow siren]

literally whats the point

[light fox]

thats the extra

the plastic containers are recycled between the two packaging lines

[fallow siren]

youre looping the containers, just hand craft once and youre done

[light fox]

i am NOT hand crafting 4000 plastic containers

[oblique hollow]

then let them accumulate right now on the side while you build stuff

[fallow siren]

then use some of your plastics making the canisters in background?

no reason to add them in the production line

youre just overcomplicating stuffs

[light fox]

either way its a later problem

[oblique hollow]

make them now, not later

[light fox]

i dont even have all of the first refineries

never made a build this big i thought my equipment factory was large

or my motor setup

[fallow siren]

again, just grab some plastic and throw it into constructors making canisters in the background, they will fill up when you're building all those refineries

[oblique hollow]

btw you likely wont need 4000 canisters
its 4000/min
you can probably make do with 1200 or something

IF you make a loop per refinery
a big loop that collects canisters will likely need far more than 4000 canisters

[fallow siren]

as long as the belt isnt long, 20 per packager should suffice

my dpf bp

[limpid vapor]

graphs?

very large variation in trip time would be very bad for the system

they all leave from here

[civic kiln]

Why? What will a variation in trip timeee do? Is there a congested intersection somewhere?

[limpid vapor]

there are a number of potentially busy intersections

[civic kiln]

Satisfactory doesn't really give you a way to control train scheddules that perciceley, lots of things can cause them to slow down or speed up and eventually they won't be in sync.

Ddo you have them all timed to leave in a specific pattern.order?

[limpid vapor]

thats why i want this end of the system to run efficiently

nope, they should be at least somewhat spread out, the hope is that there is enough variation in distance that they are spread out well enough

but i wont know until its all running

i just dont want to have to solve problems on 2 fronts later

if i need to change my rail later, thats whatever, but at least i would know that on this end, its all good

[frosty owl]

The traffic should just balance itself out on its own.
I think the only way you can affect the system is by making sure that trains leave stations as few times as possible (ie: ideally a train should leave a loading station completely full and leave unloading completely empty). That will mean you'll have as few traveling trains as possible at any given time

[civic kiln]

I am 99% certain that any schedule and timing you set up will eventually fail due to rounding errors and non-deterministic imbalances building up. Youu can't get each train tto run at the exact same roundd trip time.

[limpid vapor]

thats not the goal

[civic kiln]

( Sorry for typos my keyboard is acting up )

[limpid vapor]

that IS the goal, as few trips as possible, in as short a time as possible

[frosty owl]

Ie: round-trip time should be as high as possible without hurting throughput

[limpid vapor]

some variation is inevitable, i know that, i am just trying to minimize the max amount of time a trip takes

full load and unload is the plan

[civic kiln]

Double containers with 2 belts will unload a station as fast as possible. If trains are set to wait for a full unload, they will still only make as few trips as possible

[limpid vapor]

thats what i am working towards here

[limpid vapor]

again, thats assuming there is enough room in all the involved containers to unload

which there will not be, if i dont do any balancing

[frosty owl]

I think you got that backward: the longer the round-trip, the more time trains spend in stations rather than traveling

[limpid vapor]

that doesnt make any sense

[frosty owl]

If you set trains to wait for 60 seconds instead of 30s at each station, you increase the round trip by 30x2 seconds

[limpid vapor]

what i am talking about is, for example, if there is only one train doing a trip, it has no obstacles, thats a minimum trip time, but if it has to deal with traffic at intersections, that increases its max trip time

[limpid vapor]

i do not want to use the timed waiting at all, its not efficient

only fully load/unload

thats the best setting

[frosty owl]

That's what I mentioned earlier: you can't affect traffic. That sorts itself out on its own.
You can only affect how often trains leave stations (thus how many are traveling at the same time)

[limpid vapor]

you misinterpret, i have no qualms with traffic, its inevitable, literally all i am trying to achieve is as short as possible trip times, and trains leaving stations only when needed

my goal is as smooth a flow as possible

[civic kiln]

Trains waiting at a station isn't a bad thing in of ittself. TRains don't have to be in constant motion, you just need to make sure the average is maintained. They can take longer for one trip and carry more in another, wwhich is what buffers help with.

[frosty owl]

Well, you can't affect traffic without modifying the railing, but I'm assuming that's not a possibility

[limpid vapor]

i cant know if there is a traffic problem, before starting up the system that involves the trains and this factory, but first i need to build the system

[limpid vapor]

its like you are not hearing me at all

[civic kiln]

Build in extra capacity for the train and it will all even out. If your machines consume 1,000 items/min and you trains can carry 10,000 items then you need to average a trip every 10 minutes. Adding buffers means one trip could take 12 minutes and you're fine because the next train will carry extra to cover it.

[limpid vapor]

the concept of using containers as buffers is not new to me, you just dont seem to understand my issue here

[civic kiln]

I am clearly not getting it, no. 🙂

[frosty owl]

I don't see the misunderstanding (on my part). You want as little traffic as possible - > you want as few trains moving at the same time as possible - > trains should stay inside stations (not moving) as long as possible (without hurting throughput) - > the time each train takes to make a round trip (travel time + waiting at stations) should be the maximum possible.
Is there an issue/misunderstanding in any of these steps?

[limpid vapor]

trains should not be sitting at stations for as long as possible

only as long as needed to fully load or unload

[frosty owl]

The moment they're completely full or empty coincides with the maximum time they can spend in a station without hurting your throughput (ie: not being able to transfer all items)

[civic kiln]

To make sure I am clear, you are talking about a potential problem in the future you are trying to avoid, NOT a problem with machines currently stalling or backing up?

[limpid vapor]

yes, future problem

[slender merlin]

cam someone explain me why this configuration is better than just 2 smelter at 100 % with merger and then spliter to the 3 constructor ?

[limpid vapor]

this is true, but thats the thing, i want that time to be as short as possible

[frosty owl]

Why? The shorter, the more often trains leave stations - > the more trains will be on the rails at the same time

[limpid vapor]

the time to unload depends on whether its possible to unload all containers at once, or if the train needs to wait for the last container in the line to empty

in the latter case, the train will sit for longer

i want this time to be shorter

as short as possible

[civic kiln]

How are you connecting your machines to the stations? Does each belt coming from a docking station go to a set of machines or are they all merging and then splitting again?

[limpid vapor]

it isnt, they will perform effectively identically

[frosty owl]

It uses less power.
I suggest refraining from using calculators to solve such simple problem as it's akin to using your phone calculator while someone (the game) is trying to teach you arithmetics (the basics of the game).
In this case, you have yet to understand how clocking exactly affects machines and thus why the factory planner is giving you this solutions rather than another one (the only difference is power used)

[civic kiln]

It will use slightly less power but will function the same as fare as speed

[limpid vapor]

well, so far the only merging i have done for the other stations is when i build balancers

[frosty owl]

This I do not understand. Can you expand on this detail?

[slender merlin]

Noted, thanks for the tips !

[frosty owl]

||Leaves for lunch 👋||

[civic kiln]

I'm trying to understand how a station not emptying fast enough will cause a problem. You have smelters hooked to all these belts? Are you worried some will starve of inputs?

[limpid vapor]

my train may have to wait a long time before it can leave, not just because it sits there waiting to unload, but also because it may have to wait for other trains to leave first, this means that the trip time can increase considerably, meaning eventually, yes, my machines hooked up to that station (effectively) might starve at points

[civic kiln]

If you make sure your trains can carry more than they need each trip, then occasional delays will be handled. Trains are inherently non-deterministic so you have to add in some extra capacity to handle fluctuations in round trip timee that wwill eventually happen no matter how well balanced you build it.

You can balance everything perfectly but eventually, no matter what, all your trains are going to wind up at the same intersection all at once. Which is why stations have larger storage than the trains, to handdle this over/underflow.

[limpid vapor]

they have some time for variation already

i just want the whole system to be as efficient as possible with regards to future rail expansion

like sure, i can already handle some variation, but i am trying to maximize the amount of variation the system can handle, if that makes sense

[civic kiln]

Longer trains wwith fewer trips is always going to bee the best for effeecency

[limpid vapor]

well, on paper, each of these 6 should have 43 minutes to perform a round trip

they will likely spend most of that waiting at interesections

[civic kiln]

How many trainss on the rail line are there, and how much planned?

[limpid vapor]

right now there are none, but there will be 19 going up and down this hill, incoming trains are a non issue, because each station only has one train, its the exiting thats a potential problem source

there will also be others (not to do with this factory), going around the right side

[civic kiln]

Congestion is fairly easy to solve wwth trains by adding bypass rails and shortcuts if it becomes an issue.

[scenic moon]

Hey team, I am having trouble here.
I have an input of wire 141.35 per minute
I am consuming 141. 375 per minute.
Yet the wire backing up and not consistent in flowing, is that normal? I have mk3 belts

[civic kiln]

Double check that your machines that are consuming are all running at 100% and seee if any have theirr outputs blockeed.

[unique cypress]

With such a tiny difference, any imbalance that you might've had before would take ages to even out

[scenic moon]

4 of the 5 machines are running 100%, one is 96%

Crreating wire

I did change a lot of things around - earlier, but that was like an hour ago. Maybe I'll take all the wire out of the machines that backed up, flick them on and off and see what happens

[unique cypress]

With a difference of 0.025/min, it'd take 40 minutes for one backed up item to disappear from the production side. So it'd take 20000 minutes, or 333 hours, for an entire stack to be consumed

[scenic moon]

Yeah okay, that's crazy. So it's probably something I shouldn't get caught up on or worry about

[limpid vapor]

thats the kind of thing that would drive me nuts, knowing that that kind of thing exists somewhere in my world heheh

[scenic moon]

Yeah it's really bugging me that my belts are stopping and starting even tho it's the same consumption and produce.

Been working on it for like 30 minutes now and I don't know what I am doing. Not the game for me if I am like this.

[limpid vapor]

im not sure i understand which way this is set up

"I have an input of wire 141.35 per minute
I am consuming 141. 375 per minute. "

are you overproducing or underproducing?

[scenic moon]

Sorry I am producing 141.35
Consuming 141.375

[limpid vapor]

if it bothers you, maybe you could set up an extra machine and direct overflow to a sink?

[limpid vapor]

on the contrary, i love obsessing over this kind of stuff, even if that obsession sometimes takes the form of annoyance

i think its the perfect game for this

[limpid vapor]

anyway, i got the impression you were trying to talk me out of designing a marginally more efficient system

but i think i have the answer, i do an even merge, and build a manifold off that

instead of doing something insane like managing 136 sub lines

[light fox]

are long manifolds good for refineries

[limpid vapor]

i just hadnt considered the fact that since i am under belt capacity with my importing, i can just do this

[limpid vapor]

good for anything

unless you are talking about long pipe manifolds, which i have had bad experience with

seem to work better if those are looped

i tend to build my machines, refineries in particular in a mirrored fashion on the input ends and then link the pipes at the end, i havent had a single flow issue since i started doing that

[limpid vapor]

or feed extra raw material into your line and overclock a machine

[oblique hollow]

Having too many machines on the same manifold tends to be troublesome.

Keep it small if possible

[civic kiln]

I haven't seen any issuees with huge manifoldds. Most of my factories are set up to have as many machines as can be fed by a 1200 belt at a time.

[oblique hollow]

Pipe manifold

Not belt

[civic kiln]

Yeah, those ARE a problem

[unique cypress]

So far, I have never once managed to create a 600/min HOR pipe
595 HOR? Sure. 600 water? Also yes. But 600 HOR? Nope

[civic kiln]

I try not to run pipes hardeer than 590/min.

[limpid vapor]

you know what, we need the ability to snap splitters and mergers to other splitters and mergers

just imagine, infinite line of splitters with no cap on rate of transfer

[scenic moon]

I just want some sort of Z height lock Or indicator of Above Sea level or somethin

[limpid vapor]

what for

[scenic moon]

I like straight lines, not going up and down every sand dune

with belts

Tho the new curves are pretty neat

[limpid vapor]

build foundations higher up

[grave escarp]

That may be too much, but a splitter at building outputs would be an awesome addition all game long

[unique cypress]

yeah, it'd be nice to know the height of foundations without using the console or save editor

[oblique hollow]

The thing is there is nothing that seperates liquids from one another. Not density, not viscosity

[unique cypress]

oh, but there is: the amount produced/consumed by 1 machine

I haven't had issues with water and oil, and I suspect it's because it takes 1-3 machines to fill a pipe
HOR takes 15

[oblique hollow]

Production is unique for many things. Purely in relation to properties though, liquids are identical

HOR probably suffers from the fact that large manifolds just have trouble staying full all the time. And of course the classic first junction water hammer

[frosty owl]

Sushi-load-balancing goes brrrr
And many others quirky setups one can create xD

[unique cypress]

I haven't had any issues with long consumption manifolds with an elevated pipe

It's the production side that's the issue with HOR

[oblique hollow]

Same principles apply to output manifolds and input manifolds

Use mk 1 while below or at 300/min, mk 2 above

[unique cypress]

care to explain then why I have no issues getting 600 oil/min out of an extractor but it's damn near impossible to get 600/min out of 15 HOR refineries?

[oblique hollow]

Output manifolds that feed down have the trouble that you cannot really get machine head lift - use pumps

[oblique hollow]

Would need to see the save or the pipe layout

Theres no network i cant really explain

If you want an inspection, put the HUB near the oil refineries and DM me the save

[frosty owl]

In combinations with fluids systems where you manage byproducts they can (in some rare cases) lead to designs that stall due to lack of buffering.
Eg: a recycled Plubber system can halt if the machines making rubber/plastic don't get enough solid inputs before the fluid inputs are full (this still depends on how one is connecting the machines though)
Again, this is very rare! I can recall only one time seeing a user having this issue (flushing the fluid once can be enough to never encounter it again as the system then stabilizes)

[oblique hollow]

A HOR to coke refinery manifold should be the simplest to troubleshoot

One / two issue with head lift that i can confirm atm, especially after receiving some more insight from the recent discussions on Pipecord:

Build order for pipes affects height for some reason. the same pipe built bottom to top vs top to bottom will result in different height according to pumps

And also Valves in series kinds just tend to kill head lift
One valve is fine, but 2 in series runs the risk of failure and getting no lift

Speaking of - Ven why are you not on pipecord

Slacker smh my head

[crimson moat]

because of this issue where valves behave inconsistently depending on the order of calculation, if a system involves multiple connected valves, we should probably just use pumps instead

[oblique hollow]

Pumpa between valves
If you need the valves to limit flow

Which, btw, is perfectly valid use for valves

[crimson moat]

for flow direction only, pump will do it without triggering the headlift propagation bug which can be triggered with multiple sequential valves between machines/pumps

[oblique hollow]

Pressure groups my beloathed

I still dont get why valves dont just save the last value from the last valid pressure group

Maybe they used to and thats why i discovered the bug that happens when rapidly changing the valve slider xd

(which is fixed nowadays)

[frosty owl]

Breaking news: it's okay to use valves as valves!

So, just use pumps to "reset" pipelines to a good state?
Is this a fair point to make?

[frosty owl]

Pls @ me if you're talking to me 😅
Sure, I'll give it a go if you think I can be of help.
I don't think I've ever seen a link to that server though (first time I hear of it)

[oblique hollow]

please DM me that, saves tend to ger nuked off here

And im not at home so cant download it rn

[crimson moat]

There's a bug that affects headlift propagation, so yeah. Pump sets it to a high value from itself forwards, without relying on propagation from further upstream which doesn't always work perfectly (especially with multiple sequential valves between the headlift source and the destination).

[frosty owl]

@unique cypress Do you mind if Sir Mc shares the saves with me?

[unique cypress]

Sure

I origianlly posted it here after all

[frosty owl]

But that's illegal!

[unique cypress]

only deleted it because McGalleon said it's gonna be deleted anyway

[silent shoal]

Does it make sense that the consumption is higher than the max consumption?
I will only be satisfied when the lines are aligned🫣

[frosty owl]

It's not uncommon, but it's still not supposed to happen. I don't recall what causes it, might be related to overclocking a machine after it has been connected to power (so the graph hasn't updated to the new max)

[fallow siren]

youre over 100MW, thats caused by hoverpack

or sometimes caused by trains

but likely to be hoverpack

[silent shoal]

you are right 🤡 😅

[silent shoal]

❤️

[frosty owl]

Now that's one smooth power draw

[light fox]

oh no

my new project is gonna take over half of what my power grid currently makes

good thing the project is an upgrade to my power grid

[silent shoal]

Do 2400 Rocket Fuel.
That lasts a long time...
But I don't think you're in the right tier so go for fuel...

[light fox]

yeah not rocket fuel i could only do turbo

this is what i have right now

[silent shoal]

I recommend you skip turbo fuel...
thats 50,000 MW no?

[light fox]

that should make 50k yes

5719.6 of which goes into making the 50k

[unique cypress]

That canister production and sink is totally pointless

You should do closed loops

One loop per refinery ideally

[light fox]

the canister production is how i get the canisters in the system

[unique cypress]

blueprints is how you should get canisters into the system

[fallow siren]

.

[silent shoal]

😞
There are 58 generators here.
57 at 250%
and one at 150%
they get one pipe of 600 RF.
The math is correct...

The problem is that the last generators don't get enough fuel and they turn off and on all the time...
And the blender that feeds the pipe also turns off every so often because its outlet fills up...
Right now the pipe comes and splits into 2
I'm trying to just connect it to a U at the end and create one really long pipe...
If you have any ideas I'm listening

P.S. Don't judge me... There's no room for that above...

[light fox]

good loop or not

[unique cypress]

If you've already got a loop, connect it to the other end of the blender manifold instead of the stat of the gens manifold

[light fox]

to clarify the bottom packager makes the packaged water and the top packager empties the packaged fuel

[silent shoal]

I don't really understand...
Right now everything is connected to one main line that comes out of a blender and is divided between all the generators

[unique cypress]

Instead of doing this

Do this

[unique cypress]

make sure to put a handful of canisters (full or empty) in any of the machines so they get pasted with the BP

and upload those to the DD

otherwise looks fine

could be compacted but w/e

[silent shoal]

Will this work?
h= The generators are under the blenders

[unique cypress]

it's still a 600/min pipe

those are janky

[unique cypress]

with this, it would be 2 300s instead

but who knows

sometimes 600/min pipes work

and RF is a gas so it should've worked from the start

[silent shoal]

300 in each direction?
Making a closed circuit would be really problematic because of how I built it...🥲

It seems to be slowly getting better...
The last ones (2-3) still can't fill up and that's causing drops...

[brittle talon]

Hi everyone. I am using the Small Refineries mod and trying to find out why I don't seem to be getting the expected output on my iron refineries. I have 12 of them stacked on top of each other in a group of 5, 5 and 2, each is running at 100% efficiency. Yet the output is not 780 on a tier 5 belt.... any ideas? https://i.imgur.com/AcJ7XzQ.jpeg

[silent shoal]

The miner produces 780 ore and is at 100% efficiency?

[brittle talon]

Yes.

All 12 of them, I checked twice.

Er the miner, plenty of ore coming in.

Good idea to check, but I have a huge amount of ore coming in lol, I used the balancer mod... and there is like... 15k ore coming in behind it lol (working on a new blueprint)

Hmmm

[silent shoal]

Are you using all 780 ingots? Are there any machines that stop because there is a shortage?

I ask because I don't know how accurate this gauge is..

[brittle talon]

All 780 are going into a sink.

The counter on the ore line, but I am certain there is enough ore going into the line-

Could it be the ingots backing up JUST enough to slow the total production a tiny bit?

741 on the ingot counter... 741/780 = .95, so, it is 5% less. Spread across 12 constructors, could that mean each one is backing up JUST enough to not register as 1% efficiency loss on the machine, while slowing the ingots?

....I think I found the error ><

While rebuilding part of it, I had one of the refiners at less than 100%.... so it was at 100% efficiency... but yeah >< lol

[tame harbor]

underclock some generators on that line for a while so the pipes can fill up

or Sloop some of the final refineries

tho that method might need some pipe-fu if you're maxing out the pipe by default

[light fox]

its so pretty looking

[unkempt hull]

how would i make a 4 to 10 balancer? i found a 2 to 5 balancer somewhere online (the one in the screenshot is my use for it), but i couldn't find a 4 to 10 one, and i'm not sure if it'd be more efficient to combine the 2 to 5 for the 4 to 10 or leave them separate

[unique cypress]

I mean I'd barely call that a balancer because it's bottlenecked to hell, making it useless for what I'd use a balancer for
But if you only want equal outputs, then I guess that works

[analog crag]

if the goal is exactly 120 per min, maybe you can just use a manifold and then have mk.2 belts into the containers

[unique cypress]

normally to build a 4:10 balancer, you'd take 2 copies of a 2:5 and balance them 2:2

but if you're fine with it being bottlenecked, then there might be an easier way

[wind spade]

Why have the balancer in the first place?

[delicate cradle]

ALUMINUMINUMS

[devout ingot]

over 900 sumersloops and 34 byproduct full water pipes...

[haughty barn]

“Why won’t the devs give us automated sloops?”
This is why folks

[versed violet]

and half of it could be avoided by using the other alu ingot recipe

[cloud sun]

Are the devs going to add a splitter where you can limit the throughput?

[marsh atlas]

That would be nice but they've been pretty opposed to it

[vapid gorge]

it would get rid of like 90% of the logistic design problem sovling, in a problem solving game

besides you can do that with clocking already

[devout ingot]

all their 1.004 recipes would just be easily done if they added that

[vapid gorge]

🕰️

[unique cypress]

They said that they aren't going to add them because that'd remove the logistics challenge

[devout ingot]

yea i was gonna say clocking is the intended solution

[unique cypress]

But frankly, there's very few logistics challenges that'd be made easier with ratio/rate splitters

After all, you can already build them and I don't see many people building then

I've modded them in once when I was a noob and thought they'd be useful
They weren't
Wasted more time than they saved tbh

[icy heron]

can anyone help me with this junction please there is one main lane where the trains gather bauxit and the should all end at the same trainstaion. yet i cant figure out a way on how to properly merge the trains again onto one lane

[vapid gorge]

hold a signal in your build options to show the 'block' colours

gonna guess there's some issues here though

and the stuff here looks like it's a bit of a mess

[icy heron]

jep indeed 🥲

[vapid gorge]

show the block colours though

[icy heron]

here you go

[vapid gorge]

could I see the first trouble spot I shared an image with in signal mode pls?

[icy heron]

sure

here

[vapid gorge]

I'd guess at least part of your issue is the tracks far too close to each other up ther e

and making it one block

[icy heron]

how did i made it one block?

[vapid gorge]

if rails are too close they'll be 'connected'

and turn into one block

to simplify it I'd probably just do something like this

[icy heron]

ok thanks for your help

[vapid gorge]

the closest you can have rails together w/o blocks merging is 2 foudnations next to each other, the rails in the middle of eaach

[tardy stream]

completely load-balanced 9 assemblers

[visual yarrow]

So I wanted to share my standard modular petrochemical setup and explain why I like it. Bear with me for a moment.

So the input is 270 crude oil/m. I use this number for several reasons, but the primary one is because I avoid using pipes at full throughput, and 270 happens to create a nice, clean ratio of machines. It's also the speed of a MK3 belt, which is nice for aesthetics when using packaged oil.

I adjust the clock speed on the HoR refineries to 112.5%, and the diluted fuel blenders to 90%, to provide 8 of each, so I can directly pipe one HoR pipe into one blender. I tend to also use one water extractor per blender when I have the space, since they have similar widths.

Finally, I use paired recycler refineries. Two refineries side by side, one rotated 180 degrees, leaves the output of each right next to the input of the other. This can be split 1:2 to easily keep producing plastic and rubber as long as fuel is piped in, but requires the machines to be "seeded" ahead of time.

There are pros and cons to this. The biggest disadvantage is obviously that you are making roughly equal amounts of plastic and rubber (it will be 5:4 rubber:plastic due to the polymer resin, which I did not include in the planner for simplicity's sake), but I personally find that due to how crude oil nodes spawn in groups, it's justifiable to make a lot of plastic and rubber at centralized locations, and then use it as needed, and there are enough uses for each that I can consume both in relatively even quantities.

But one massive advantage it has is that it's incredibly well suited to blueprinting because of how compact those paired refineries are.

And i'm saying this as a chronic non-blueprint user.

This isn't the best screenshot of this because it was moreso taken for showing off the aesthetics, but this is an example of paired refineries.

Another view.

[silent shoal]

By load balance you mean they start at the same time?

[visual yarrow]

Yeah. One belt split into three, which then split into nine.

Technically they'll start at very slightly different times due to the round robin nature of splitters, but the main idea is that, if you supply exactly as much input as is required, the belts will never stop flowing.

I'm assuming the important part of the example is that it was fit into a blueprint device.

[tardy stream]

finished!

[tardy stream]

yeah my goal was 10.8 assemblers for 270/m coal into compacted coal, but could only fit 9 while load-balancing :(

[visual yarrow]

You could also do 6 assemblers at 180% speed.

Okay, here's my secret weapon:

Take the total combined % of all buildings and divide by a target building count to get a clock speed.

[tardy stream]

you got the power slugs just lying around?

[fallow siren]

slugs are everywhere

[tardy stream]

I aint got a jetpack yet

[fallow siren]

5k power shards worth of slug in total (slooped)

[visual yarrow]

I haven't walked in that direction yet

[silent shoal]

Cool
I built something similar but in row (8 assemblers, main input into splitter 2 output into two assemblers input and third split output into next splitter and so on)
The last 2 assemblers had problems starting...😅
I just filled them mentally to help...
I think in the long run it would be fix but I just helped little 😅

[visual yarrow]

The nice thing about the quick search is that you can just swap out the number. So we can see that 4 machines is just over the overclock speed cap.

[tardy stream]

yeah that's manifold, mines balanced 😎

[visual yarrow]

You can also see what you would need to do if you hated yourself.

[silent shoal]

Sounds like load balancing is better 🫣 🤓

[tardy stream]

well depends 😂

I just prefer it

[visual yarrow]

I personally aim for always-flowing belts, which usually means balancing. But I have various tricks to simplify the process.

One of my favorites is throttling belt throughput with lower tier belts and smart splitters to subtract specific quantities from belts that you know will have a stable item flow.

[tardy stream]

with lighting

[visual yarrow]

Also a good use for the calculator! A full MK5 belt, with a MK3 belt "subtracted" from it, and then split in three.

Then I can merge the remainder belts as needed and work from what I have left.

I actually use these special buffer... things, at places like vehicle cargo dropoffs, to regulate the flow of items.

So like... I have a drone shipping in aluminum ingots to my rocket fuel setup (or, well, a nearby building on a tall outcropping) for the sake of making empty canisters for packaging.

(and also ficsite but putting that aside)

I'm sending 180/m over, so I have a storage container as a buffer outputting from a MK3 belt. This is split into three, and one of the belts is looped back around and priority merged back into the input.

End result: despite the bulk delivery of the drone, the container ends up outputting a constant rate of 180/m. Which is... not actually needed in this case. It just looks neat. I'll get a screenshot the next time i'm over there, if I remember.

...wait, there's a uranium node here?

Oh, right. I had forgotten, because the last time I used it was waaaay back before this place's graphical/gameplay overhaul.

[silent shoal]

Nice

[keen shard]

Hi, How does one calculate or figure out the burn rate of a 250% overcloked fuel general?

[fallow siren]

multiply it by 2.5x

overclock consumption are linear, the only thing that are exponential is the power draws

[keen shard]

it cant be that simple, but I dont have an easy time verifying it either.

[vapid gorge]

it is that simple

[fallow siren]

it is that easy

its simple math

normal fuel burn at 20/min at normal clock

250% means 20*2.5

so you get 50/min at 250%

[unique cypress]

The only thing that isn't linear with clock speed is power consumption of production machines

[keen shard]

when i looked at the wikipedia it went into burnrate and other stuff

but the math checks out with what displayed on the wiki

[unique cypress]

!wikisearch fuel+gen

[brisk shore]

https://satisfactory.wiki.gg/index.php?search=fuel+gen

[unique cypress]

^ Make sure you use that wiki

[keen shard]

Thank you, yes Im making sure of that. But I think the real problem was just my brain confusing itself with the answer being too simple to be obvious for it.

[fallow siren]

you can check the base consumption on the top right of input slot

[unique cypress]

Does that update with clock speed, or does it only show base consumption?

[fallow siren]

only base

[foggy pecan]

Just to be sure, there isn't enough s.a.m. ore on the map to do this right?

[fallow siren]

5k/min fics rod isnt even possible in first place

[unique cypress]

!wikisearch miner

[brisk shore]

https://satisfactory.wiki.gg/index.php?search=miner

[unique cypress]

Go to the max extraction section

[foggy pecan]

well if you have unlimited somersloop it is

[unique cypress]

There isn't 30k+ SAM on the map

[fallow siren]

theres only 10200/min sam in the world

and that alone isnt enough for max ficsonium build

[unique cypress]

With sloops, it barely is

[fallow siren]

yeah, its just so expensive

[unique cypress]

But max Ficsonium from 2100 uranium and without slooping nuclear stuff is ~153/min

Max without cheating is ~300

[foggy pecan]

Yes, I figured as much, I currently make the 250 plutonium fuel rods and have been sinking them for quite a while. I know ficsonium rods are not worth it but wondered if I could

Thanks for the info!

[unique cypress]

Hmm I should give optimizer like a million sloops and see how much Ficsonium is possible lol

Though it looks like 1530 assuming all SAM goes into Trigons

[mossy ibex]

I’m trying to think about how to sort deliveries and I guess I keep getting hung up on throughput. There’s really no way to maintain speed without basically having a fully separate line, full set of smart splitters, and a storage unit for each output line of all stations. Is there a better way?

[wind spade]

Not mixing things, dedicated lines

[deft lichen]

the general approach is that 1 freight car = 1 item type

and then, if you're transporting a lot of different stuff in small amounts, you can choose to have mixed cargo

[unique cypress]

I do 1 entire train = 1 item type

[deft lichen]

mixed cargo is kinda problematic because the train mustn't stop running, else one item type may back up and clog the others

as with everything else, "don't mix belts if you don't know how to separate them"

[vapid gorge]

delivery side:

platform -> buffer -> sorter -> individual part buffers + sink

[mossy ibex]

I can’t stomach the vast field of train stations it would take to make them one to one. I am using them, but I am actually sad—I’d been looking forward to trains forever, then I got there and realized how they actually worked.

Whatever, low throughput it is

[silent shoal]

At point A I produce 143 plastic per minute, what is the chance of moving it to point B and maintaining at least 120 per minute?
And how can this be calculated, if at all?

[unique cypress]

train or drone?

[silent shoal]

Using a drone or train
I was just about to write.

Or I think it would be better to build the plastic at point B...

[unique cypress]

I mean a train can always be expanded with either more wagons or more trains. but even a single train with a single wagon will be able to handle 120/min just fine. Hell, even 1000/min might be possible

with drones, the throughput is much lower, but because plastic stacks to 200, 120/min should be totally doable with just a single drone

[silent shoal]

Will this work with buffers (meaning 120 plastic per minute)? Distance doesn't matter here?
If in the future another train is added to the track, etc...

[unique cypress]

distance always matters. I'm just saying that at this distance it looks like you could maintain up to 1000/min/wagon with mk5+ belts and ~900/min/wagon with mk4s

adding extra trains won't affect it much, unless they cause significant slowdowns

[crimson moat]

i get around 4 stacks per min per drone (and 8 per dronePORT, if you have a drone leaving and a drone coming)

[unique cypress]

5 stacks/min is the limit at 0 distance

[crimson moat]

10.558 by my calculation, kilometers of distance taking it down to between 4 and 5

The takeoff+landing sequence is 51 seconds (TOTAL, not at each end), and they hold 9 stacks.

[unique cypress]

Efficiency
After the first round trip is finished, the stack throughput of the Drone will be shown in the Drone's home Port, alongside its maximum throughput and fuel required per round trip. The efficiency of Drones increases with travel distance, as the landing and take-off animation takes 51 seconds. If the route is too short, most of it will be the landing animations; it would take 102 seconds to move items from two Drone Ports immediately next to each other.

according to the wiki, it's 51 seconds per docking

so 5.29 stacks/min at 0 distance

[crimson moat]

I'm getting mixed messages from research, it's not 100% clear if it's 51s *2 or 23+28 seconds (23 at one end, 28 at the other)

I have measured throughput of over 800/min on one droneport with one drone leaving and one drone arriving though, using 100 stack items, which fits with the second case but would be obviously impossible in the first.

Needs testing ingame i think to prove which way it is

[unique cypress]

right after setting the route it says 1:58 round trip time and 4.55 stacks/min max

we'll see what it says after a few trips

[crimson moat]

That's 900 plastic per min, but you can double it (well, 1200 max..) by having one drone fly to the plastic and one drone fly from the plastic. So 2 droneports, 2 drones (instead of 2 droneports, 1 drone)

[unique cypress]

at 0 distance

[crimson moat]

Yeah, but 1820 plastic/min per destination port is pretty good for a baseline 😄

[unique cypress]

it can't decide what's the actual round trip time

it jumps between 2 and 2:20

and the transfer rate never showed anything above 415

also, fun fact, this is packaged turbofuel apparently 🙃

[crimson moat]

Yeah, plastic is double the stack size so that would be 830, but then 2x for 2 drones and it's 1660 i guess. Less if higher distance. 1200 belt cap

[unique cypress]

it went down to 367 and then back up to 394

don't make me pull out the fucking OCR to average the throughput monitor over an hour

[unique cypress]

that's from the drone port, not the monitor btw

[crimson moat]

That thing is a little wonky. What i would do is set up a bunch of ISC's on the output with a sink, then after the drone/s have been running for a few minutes, delete the sink and fire off a stopwatch.

and then after say 20 minutes, delete the output belt from droneport and count your items.

the longer the better since deliveries are lumpy, but 20m is plenty for a ballpark

[unique cypress]

or I could do what I did for testing pipes lmao

totally won't take like 5x longer

but it will be more accurate

[tiny leaf]

didnt see this till now but thanks

[icy smelt]

Since Im new and I think every1 here is like a mathematician genius, imma ask this: I’ve seen manifolds and I’ve seen load balancers, but is there a setup to combine consistent equal distribution of load balancers with the compact package of manifolds?

[crimson moat]

blueprints or ratio builds

(like machine A1 outputs to B1, while A2 outputs to B2, no combined belt)

[icy smelt]

Interesting thx

[steel mantle]

is there like a goldylocks zone for how much oil i should put toward making items and making power?

[unique cypress]

@crimson moat From a half an hour average from the throughput monitor, the transfer rate seems to be 381.46/min, and the drone port's reported transfer rate has not stabilised at all, even after half an hour, jumping between 310 and 410

Idk if drones are this inconsistent in general, or if it's because of the extremely short route, but it doesn't look good lol

not to mention the theoretical transfer rate is supposed to be just over 500

[frosty owl]

Maybe they should add a monitor that averages over longer times, for long-range transports?

[unique cypress]

maybe I finally won't have to OCR a recording of it frame by frame to get a long-time average 🙃

[crimson moat]

The throughput monitors are just inaccurate

[unique cypress]

how am I supposed to know whether the drone port or the monitor is accurate

[crimson moat]

they're both not. For the actual count, literally just count the amount of items and the amount of time

and then divide A by B

[crimson moat]

can roughly say 3 to 4 stacks per min per drone though 😛

(and up to 2 drones per port)

[wind spade]

not really, just "make what you need"

[unique cypress]

they are

counting items gave me 485.5/min

averaging the counter gave me 440.7/min

this is definitely not within margin of error

maybe I shouldn't've ran the game at 30 FPS to match the recording speed

but I don't have time to redo this today

maybe tomorrow

[crimson moat]

The counters are just all screwed

they are always wrong by like 10%

unless you have a very consistent and stable pattern on the belt (like 100% full, or every second slot has an item)

[unique cypress]

today I learned from McGalleon that a pipe showing 0/min can actually have flow

truly another moment

[crimson moat]

double for 200 stack size, double again for 2 drones and you have 1942/min at zero distance

(belt capped to 1200)

people underestimate drone throughput, especially if they don't know of the double drone trick

(drone from A goes to B, and B goes to A - 2x throughput per drone platform)

[unique cypress]

with a graph like this, I'd expect they'd be at least mostly accurate

[crimson moat]

yeah, they're sadly just not

[devout ingot]

damn they really dont want us to see accurate information

💀

[unique cypress]

Averaged it using a different method, not just a simple mean over the entire graph, but still got 450-460, nowhere near 480

Also, I'd like to point out that in one of my tests, the counter showed 380, in the other 440

Both tests - drone ports right next to each other, same arrangement

But for the second test, I moved somewhere else to have access to a miner

Different location, and already a 60/min difference

[light fox]

imagine this as someones power grid screen

genuinely thought thats what it was when i saw it

[unique cypress]

Actually, there's another thing I should check

When I used containers to measure items, I used a stopwatch on my phone to measure time

But the throughput monitors use in-game minutes

I probably shouldn't assume those are the same

[unique cypress]

a 600/min miner in 20 minutes produced 12002 items. 600.1 per irl minute

so IGT is accurate

[vapid gorge]

@young agate show me a screen shot of what you're doing

[young agate]

able to vc?

[vapid gorge]

not atm

but just send a screen shot

[vapid gorge]

oh if yo uwant to force it to use a specific recipe, uncheck the other base recipe

[young agate]

ohh

[vapid gorge]

if you have multiple recipes for the same item it'll judge by resource value.

as you've given the program 2 options

[young agate]

That fixed it. Thank u :D

[quasi grotto]

what's a more important resource going late game Sulfur or oil (wondering what rocket fuel recipe to use)

[regal falcon]

Do most satisfactory players keep their miners away from their constructed factory?

[vapid gorge]

it's often very convenient to make hbus near nodes

[vapid gorge]

sulfur is fairly niche, you can use almost none of it

and depends how big you go, honestly unless you're mad you won't run out of anything

[regal falcon]

Im scared of buses too, merging, splitting, ayaaahhh

[vapid gorge]

why would you make buses ? they don't work well wit hthis game

NOT building near nodes is more likely to have huge logistic lanes

[regal falcon]

So this is meant to say Hub, not Bus?

[vapid gorge]

ah yeah hub

avoid busses. Tons more work, no real benefit

supposed to be 'hubs'

[regal falcon]

Allgood to move to dms?

[vapid gorge]

if you want? but this channel is pretty quiet and you're right on topic xD

[quasi grotto]

Does anybody have the calculator set up for Uranium plutonium power set up so it gets the most power possible (with natural resource limit in mind)?

what ratio should uranium be distributed to each power (Im having a hard time calculating cuz the calculator took is janky)

[vapid gorge]

https://www.satisfactorytools.com/1.0/production?share=DnbQYumRt2vfOK3jyc48

[quasi grotto]

It does not show the optimal ratios

[vapid gorge]

define 'optimal ratios'

[quasi grotto]

uranium is the limiting resource of the world (at least from what I can tell)

its not optimal for power

[vapid gorge]

ah IIRC optimal power is max uranium for U rods.

make whatever P rods you can from the waste

[quasi grotto]

oh fr?

[vapid gorge]

so that would be optimal for power production

[quasi grotto]

I thought it was not

[vapid gorge]

If I recall correctly. Been a while

[quasi grotto]

atleast thats what i calculated

[vapid gorge]

I think they talk about it on the wiki

[quasi grotto]

do you have a link?

[vapid gorge]

or they used to

[quasi grotto]

let me calculate again ig

[brisk shore]

Official Satisfactory Wiki

This is the totally awesome community driven wiki for Satisfactory!! Come on in and check it out!
<3 @vapid gorge

[vapid gorge]

!wikisearch plutonium_rod

[brisk shore]

https://satisfactory.wiki.gg/index.php?search=plutonium_rod

[vapid gorge]

hmm they might have removed it.

But I'm still pretty sure your best bet is ALL the u rods you can make, then stretch out the plutonium with whatever alts work best

[quasi grotto]

Oh ok

I think your right I just re did calculations

I think some old yt video threw me off cuz it was doing it split

[vapid gorge]

plutonium rods were really only introduced so more people would interact with the nuclear process sicne folk didn't want permanent waste

and I strongly suspect ficsonium was onoly introduced because a number of people didn't have the self control to not burn P rods

[quasi grotto]

is it even possible to go no waste with ficsonium if your already maxing out on power with uranium and plutonium

or Im pretty sure the calculator is not maximizing the ficsonium Im using

[vapid gorge]

I think there might not be enough sam to process a full plutonium waste thing

though I'm nto sure if that's based on the 12~ rods pm, or the P rod 22 rods pm can be made

some ninny was bitching about how that completely 'negates the game mechanics, devs bad... wah wah wah'

[quasi grotto]

at that point your better off sinking plutonium no?

or spliting the plutonium so some of it gets sunk and the rest for ficsonium

[fallow siren]

ficsonium isnt really worth it if youre going for net positive

just sink plutonium rod, or use them as drone fuel

[random creek]

calculating setup for my ingot factory for copper and iron. 80 refineries for copper since leached isnt really worth it 💀

or i could cut down the space and use leached anyway

[quasi grotto]

So ficsonium is net negative when you add the plutonium production to it or?

[random creek]

with only 26 refineries but a bit less copper gained

[fallow siren]

no, fisconium is very expensive for a small net positive

you can get more power by converting those sams for more uranium

[quasi grotto]

Yea I know but Im trying to make 0 waste max power

[quasi grotto]

ohhh

[fallow siren]

you can do that by sinking plutonium rod

0 waste

[quasi grotto]

well if its gets the most power ig

yea your right lol

Just did calculation

[fallow siren]

max ficsonium build is barely possible even if you use all the sloop in the map, and that also require a fully empty world without nothing in it so you can use all the resources to make ficsonium

[quasi grotto]

yea no uranium makes the best ammount of power

[vapid gorge]

in general probalby just sink plutonium or use it in drones or trucks

[devout ingot]

i did some math a few weeks ago on the same model going for like above 100 fiscionum power plants lead to running out of sam for fiscite trigons lmao

[wind spade]

Ficsonium is net positive tho

[fallow siren]

never mention it being net negative in this convo

[fallow siren]

.

if youre going for no waste? go for it

if you want increase of power? not worth the effort

[wind spade]

"worth" depends on individual I guess

[lone igloo]

Like who wrote this algorithm for straight belts.... it aint that hard bro why does it struggle to understand this but default will do perfect placement...

[wind spade]

it may be because it actually cannot be straight by a few pixels like this

[lone igloo]

how do you even notice that O_O

[wind spade]

it's not that I noticed it, it's just my guess based on how the conveyor behaves

@broken minnow how is that choice "worst"?

[lone igloo]

low-key automated miners is goated

[broken minnow]

Ragebait

[unique cypress]

Max no-waste nuclear involves making exactly 0 Ficsonium

[lone igloo]

Pure ingot recipes are so fcking good 🙂

[unique cypress]

Eh, only if you're making a lot of ingots

I wouldn't bother with pure if I was making 200/min

But if you want 2000/min, then yeah, they're great

[lone igloo]

its more of I've had 1 pure node and smelters would not suffice the requirement without getting another node involved and I used pure recipe and made what I needed and still go so much of node left over.

[lone igloo]

low-key wish there was hotkey to swtich between jetpack and hoverpack on the fly

[haughty barn]

Make Ficsonium because it’s a skill solution

[knotty hornet]

Leached are better

[lone igloo]

yes but its tier 7, so when getting into oil that is so good.

I can do it but cba to have to make sulfuric acid for basic setup like quick HMF

[knotty hornet]

I just rebuilt my sulfuric acid sector of my factory in the Blue Crater cuz I want to be able to expand, and I wouldn't have been able to at ground level.

I leach everything, even use the instant scrap recipe for aluminum.

[lone igloo]

intresting as sulfur is most needed resource, surpised you using it for ingots when there is so much of just ore.

where pure ingot just needs water which basically infinite and sulfuring acid needs sulfur which is more rare resource and most needed late game.

Leached 100% doubles the ore but pure has 45% basically on top as 35 goes in 65 comes out.

[wind spade]

Sulfur is only needed if you do stuff like TF

[lone igloo]

Turbo fuel?

[knotty hornet]

Don't even need it for that, cuz the sulfur that doesn't get used for sulfuric goes to make rocket fuel via the nitro recipe, and then I get compacted coal out the back.

So i can make turbofuel and rocket fuel with the same sulfur and coal

And without crafting it into compacted.

[wind spade]

RF is same as TF, needs sulfur

[lone igloo]

This MF is immune to anything...

thought they made it so anything can be destoyed in this biome

[mossy ibex]

yeah.

[lone igloo]

well that single plant is invicible xD

[unique cypress]

Not really. Only leached cat is somewhat decent

[mossy ibex]

can someone talk to me about gas stations? I am seeing conflicting information and in fact getting conflicting results when I experiment. Simple question:

If I drive a vehcile within the loading hitbox of a truck station without stopping or explicitly loading (F), will that vehicle get fueled from the station or not?

[lone igloo]

Honestly once I get to place whre at now going into tier 8. I've mainly always use Pure ingots

[mossy ibex]

I think there may actually be a difference between manual drive and autopilot here too. Maybe manual vehciles get fuel, but auto vehciles have to stop? Soemthing like that?

[lone igloo]

you mean add a station in middle of a route to re-gas?

[mossy ibex]

sure for example

[lone igloo]

I mean I've never done that but it sounds simple. As long as vehicle needs fuel and it has the station alocated to stop (aprox 1 min) should get topped up

thats the theory never tried it. But it does have to stop at a station for the station to interact with the vehicle

[mossy ibex]

do you know that for absolutely sure because you tested it?

[unique cypress]

It should also be set to load and not have any cargo to actually load

[mossy ibex]

because my testing is showing inconsistent results

[lone igloo]

I've not tested it as said above "I've never done it" but should work... I dont see why not.

Station A (Load) -> Gas Station (Set to load / Wait time Aprox 30 seconds) -> Station B (Unload)

[mossy ibex]

but do I need to stop and press F to make it fuel like this, or can I just drive through?

I'm specifically confused about exactly only this issue: sometimes my vehicles get fuel by driving through, sometimes they do not. I'm trying to figure out why

[unique cypress]

The last time I used trucks was before the update, so maybe something changed, but back then, cargo operations happened at 2 stacks per second whenever a truck was within range

Pretty sure stopping wasn't required

But now we got those pause nodes that show up when you stop so maybe that's needed to move cargo

[knotty hornet]

Pretty sure they take fuel from any station they stop at that has it available, whether it's loading or unloading cargo (as long as it's the same as what they already have)

[lone igloo]

@unique cypress the vehicle MUST be parked at the station to get anything loaded or unloaded. I used normal tracktors for little while in 1.0.

[mossy ibex]

do you have a theory about why the following thing is happening sometimes:

I start recording a route at the gas station, noting that it successfully fueled. I do the route, stopping at my loading and unloading stations, and making my way back to the gas station. Now it's on autopilot, and doesn't get gas.

On a different route with the same set up, it works fine.

[mossy ibex]

do they have to "stop" like press F, or just be in the hitbox long enough to get fueled?

[knotty hornet]

That, i'm not certain of. I can count on one hand how many truck stations I've built, and I lost some fingers last fourth of July. (Second part is a joke)

[unique cypress]

If they have to stop, that means the pause node must appear

If you stopped but the pause node didn't appear, the autopilot won't stop

[lone igloo]

Honestly only time I use trucks is for ascetics to move finished produce from A to B or just have vehicles go round without any cargo. Makes certain builds feels more alive

[coral mauve]

i heard that pipeline junctions has priority inputs, does it also have priority outputs? and if the output or input is pointing down which is input or output is priority 1, 2 and 3?

[mossy ibex]

yeah I don't really use them either, I'm doing it for vibe reasons and it's giving me tuberculosis

[knotty hornet]

None of this is true

[coral mauve]

i heard it from kibitz and it fixes his problems

[mossy ibex]

it's sort of true

[lone igloo]

@coral mauve how can fluid have priority input / output?

[knotty hornet]

The only priority they have is to flow lower if possible.

[mossy ibex]

if you have a junction rotated up and down, the bottom pipe will be filled first, then the middle, then the top

[unique cypress]

Fluid flows down as much as it can. That's all the priority you get

[mossy ibex]

but I mean "first" loosely

it's not strict

[lone igloo]

that makes sense but who uses junctions vertical??? I've never used it like that and dont think I ever will.

[unique cypress]

Me, a lot of the time
Or at least at 45°

[mossy ibex]

I have used it a bunch?

[knotty hornet]

I do on every single refinery build

The main through-line is horizontal, and each machine pulls vertically off of it.

[lone igloo]

Intresting, Never had an issue with normal flat junctions ahah

[lone igloo]

so feeding from the top?

[knotty hornet]

Me neither, but vertical ones save space if you do it right

[mossy ibex]

the closest you'll get to priority outputs is facing the important pipe down and the lower priority one up, but whatever you're doing is probably too complex and you should keep it as simple as possible

[knotty hornet]

The pipe runs along the back of the machines elevated above the belts.

[mossy ibex]

yeah it helps prevent backflow leading up to the machines

I also use them a lot for water towers and other verticality

[lone igloo]

I usually try feed from the top on main setups but Never used verticle juctions are it takes so much effort to achieve a verticle one.

[mossy ibex]

I mean, rotating a few clicks?

[coral mauve]

it was like this, if you have water byproduct you put it in 1 and if you then need water from an extractor you put it in 2 that makes sure the byproduct water goes first

[mossy ibex]

that sounds incorrect

[knotty hornet]

That is incorrect

[lone igloo]

water from the top will fly down to the inn 1 at crazy speed lol

[mossy ibex]

seriously there are dragons in this exact spot, people try this and fail all the time. It's the classic thing to fail at for this exact part of the game

[coral mauve]

its weird but it worked for him

[lone igloo]

I honestly dont watch Kibitz but whatever he does is not always best or even close to true for alot of things

[coral mauve]

if it works it works, im just trying to figure out why it works

[unique cypress]

3 has priority over 2, not 1 over 2

[knotty hornet]

No

Stop

[lone igloo]

I love this channel

straight up carnage on certain topics

[knotty hornet]

Junctions don't work that way

[unique cypress]

That's how a VIP works lmao

Used it multiple times

Water from top pipe is used before the water from the level pipe

[mossy ibex]

yeah if anything 3 is priority, then 2, then 1, but it's not going to "prioritize" byproduct water the way you're wanting and you'll still end up with a deadlock probably

[coral mauve]

watch his explanation on it. https://youtu.be/t_sSjH8ezuo?si=QW7ScjB0RpQw0U8n&t=400

[heavy jay]

What is a VIP?

[lone igloo]

Isn't a junction just a free space for liquid to flow into any direction unless a fluid is already flowing through a direction and takes new fluid into the same direction???

[mossy ibex]

the "VIP junction" is a way of connecting fluid pipes that favors a particular pipe

[unique cypress]

Maybe it's not perfect, but it's been good enough to prevent every single one of my alu systems from deadlocking

[unique cypress]

Check the pins in this channel for the plumbing manual

[coral mauve]

the 3 year old one?

[mossy ibex]

surprising, honestly. You feed byproduct into the top, and extractor water from the side or bottom?

[heavy jay]

I should have known there was a manual

[mossy ibex]

the manual is still accurate, and the author posts on this discord and knows his shit

[unique cypress]

#math-and-meta message

I build it however the manual says to build it

[knotty hornet]

Why all the pumps?

[unique cypress]

Idk, I built it several years ago, it works perfectly, and I ain't touching it, lest it breaks

Actually, now that I'm looking at it, the byproduct water is connected to the bottom, not the top

[mossy ibex]

see, wtf?

[unique cypress]

Idek anymore

[knotty hornet]

See, it's just no

[coral mauve]

i think its weird that it works like that but for some reason it just works

[unique cypress]

It works perfectly tho. No leakage from the fresh side unless all byproduct is used.

I literally never once had this BP fail to work like a prio merger

Pipes are just cursed

Yesterday, I learned from McGalleon that a pipe showing 0 flow can actually have flow

[mossy ibex]

yeah it's true

I hate it

I thought they were going to land a pipe rework into 1.0. Like, that was the moment to unfuck everything. Alas

[unique cypress]

I sent them my 600/min HOR test and they did things I would've told you would definitely make things worse

And yet it works perfectly

I touched one thing and it fucking broke again

[coral mauve]

that's why I'm trying to fin a newer manual for fluids, because sins the previous was made i have seen like 10 new tips and tricks that works better then some of the things on the manual

[mossy ibex]

same here. Also with my HOR, haha

[mossy ibex]

well maybe try asking him, but I bet he's still right about everything

[coral mauve]

i would just like a manual that adds inn the new things

[mossy ibex]

my guess is that he would say the "new" things are fake

people say lots of incorrect things about pipes

[coral mauve]

after i found out about the priority input thing all my projects work better, 10x less problem with fluids, now im just trying to figure out if outputs are the same and what about multiple in and outputs on the same junction

[unique cypress]

Yeah, McGalleon's fixes to my test setup follow basically none of the most common advice

There's one loop, but it's between the last and second to last junction, not the first and last

And there are pumps on perfectly level pipes that most people say not to do

And a lot of weird junction rearrangement

[mossy ibex]

I'm glad we have him, but I also wish pipes just worked in a sensible way, lol

[knotty hornet]

Realistic Navier-Stokes equation pipe when

[heavy jay]

The pumps are working as both a check and to set pressure and prevent dead heading?

[coral mauve]

one more thing im wondering about is it true that you should not use full 600 pipes? the wiki says to not use more then 450pm to get a perfect setup, but then how are you supposed to use 1 full pure node? some times it works and sometimes it doesent

[mossy ibex]

horizontally they'd basically just prevent backflow, but will add headlift if they are close enough to a vertical section too

generally though, you don't use them horizontally

[mossy ibex]

well it's more like, don't pipe 600 and expect 600 to come out the other end. Because it's a bit stop and start, the actual throughput will be a bit less

[unique cypress]

From my experience 600/min pipes can work perfectly fine, but only if the pipe has no more than 5 fluid producers on it.

[mossy ibex]

if you want full throughput, oversize the pipes a bit. Not a lot because how filled they are matters, but like 50% is fine

[unique cypress]

Oil and water always worked fine for me, and they can fill a pipe with just a few machines.

[mossy ibex]

yeah so the idea here is that the cyclical chugging is covered by the 5 overlapping cycles

[unique cypress]

Same with aluminium water, which has, in ny standard setup, 3 refineries + a water supply from a VIP (1 or 2 extractors)

I've somehow managed 600 Fuel/min in the past, but I don't know how, and it took like 50 hours for it to actually be 600/min

Haven't had issues with 600 nitrogen either

And, like I said before, 600 HOR needed some janky shit I would've never came up with to work

And that has 15 producing refineries

That's just my theory, though. I just noticed that the length of the consumption manifold doesn't seem to matter much. 20 refineries being fed by a 600/min oil pipe work fine

[coral mauve]

i saw something like this, what do you think? you split the 600 pipe in 2 to have less pipe that is "full pm"

[mossy ibex]

what is the question?

I don't get it

[unique cypress]

Imma be honest, I consider this worse than just pulling 600/min to where it needs to go

Idk if I'm right, but another issue that could plague 600/min pipes is sloshing

If you have one dead end on a pipe, all fluid will be consumed before it gets there

[coral mauve]

im trying to use the 600pm from one extractor but it only works 4/5 times

[unique cypress]

If you have 2, you could accidentally get more on one side, then it reaches the dead end, reflects and you've got backward flow, which kills your 600/min pipe if it gets to it

Another thing I've found helps is putting the main pipe above the machines' inputs

And tilting the junctions towards the machines

[coral mauve]

i always do this

[unique cypress]

That way, they get filled one by one

[coral mauve]

and i have a buffer at the end

[unique cypress]

I never use buffers

[frosty owl]

I think that's a nice start for the 2 systems. I'd even throw in one pump on each side for good measure.
Making the pipes simmetrical as you did (at least between the junction and the first pipe supports/pumps) is also a good practice

[frosty owl]

That can help avoiding the aforementioned "backward flow", so long as the buffer is not full

[oblique hollow]

A quick summary on loops:
You dont ever really need a loop "from beginning to end"
all that matters is that the loop begins at the junction that will have 600 outflow or 600 inflow
where the second end of the loop goes is actually arbitrary, as long as it goes to any junction that has, mathematically, less flow than max

What i noticed specifically in that refinery setup you gave me and i adjusted, specifically version 2 with the mk 1 pipes everywhere and the mk 2 output near the middle, is the following:

1. Left and right side both usually have a max flow of 280/min. A pipe that is full wants to flow at max flow rate.
   this is a problem as the pipe will basically empty itself by flowing at 300/min when it only has 280/min input.

2. the one singular refinery output connected to the junction that has 600/min output WILL singlehandedly ruin flow if it doesnt stay full consistently.
   if it empties out, the junction notices the lack of pressure and redirects flow there - this makes the 600/min line drop below 600

[unique cypress]

Actually, since you're here, I'll ask:

Would it be better to merge 2 300/min pipes to get a 600/min one?

Because in a standard manifold, you're merging like 560 and 40 on the last junction. Would a more equal merge be better?

[oblique hollow]

yea 2 300/min mk 1 would be better since they will always be guaranteed to have max fullness and pressure - which you need to sustain 600/min

Equal merges or equal splits are always prefered

Asymmetric merges or splits become problematic because of - basically - pressure differences.
the side that produces less per minute does have less pressure.
Also, this is a case where water hammer can rear its ugly head.

if you dont have consistent flow from a pipe, you get the chance of the junction redirecting flow to that pipe and then water hammer happens, fluid flows back into the junction with greater speed and the junction is met with 2 huge pressure sources.

To put it more accurately in case anyone cares:

Flow rate creates pressure too. It is one of the most important pressure factors. This effectively means fluids have inertia. If it flows at a certain speed in a certain direction, it generally wants to keep doing that.

Note that when i say pressure i mean pressure.
Ive only had a hunch about it before @inner gulch came along with their pipe code analysis, but now its basically confirmed

Pipes HAVE pressure - it may not be measured in psi or bar, but it is pressure. Just converted to basically meters of head, which is a valid measurement for a given fluid density.

The reason raising pressure infinitely doesnt work is because the pipes have a flow rate cap - duh

further - pressure is only transmitted through full pipes. This was always the rule for head lift as well

[unique cypress]

@crimson moat Tested the throughput monitor at full framerate, got basically the same results. 490/min via item counting, 440-460 via the monitor

[mossy ibex]

monitor?

[unique cypress]

!wikisearch conveyor+throughput+monitor

[brisk shore]

https://satisfactory.wiki.gg/index.php?search=conveyor+throughput+monitor

[knotty hornet]

Very handy for checking that you did things correctly.

[unique cypress]

they're apparently incorrect if the flowrate is inconsistent

[mossy ibex]

wait sorry, what is "item counting" then? Just trying to understand what the different testing methods were

[unique cypress]

item counting means I'm dumping items into a container, measuring the time, and dividing the container contents by the time to get the average flowrate

as for the throughput monitor, I logged its display over 30 minutes and averaged it

I was testing drone throughput yesterday

And Aeryn brought to my attention that I shouldn't use the monitor for that because it doesn't give accurate values

and frankly, I'd trust item counting more

[crimson moat]

Yeah it's one less layer of bs

[unique cypress]

I mean I would prefer to use the monitor if it was accurate, because I can get throughput over time

but if it's gonna be 10% off, it's useless

[crimson moat]

yeah for 10% off i just use eyeballs. I need a counter (or external tools / writing) to be accurate to 1% or 0.1% range.

[mossy ibex]

I wonder if there's a thorny problem in there or if they should just fix it

[unique cypress]

I mean idk why it's even inaccurate in the first place

making an accurate one seems easy

it just might be more CPU/RAM intensive than they want it to be

[crimson moat]

At first i thought that the averaging interval was just not long enough or configurable, but your testing shows that there is an even greater error than that.

so there is some even bigger problem like rounding down

because it's not just the average fluctuating, it's actually failing to count 10% of the stuff going through.

[unique cypress]

I think imma test it with something more consistent, like a pasta accelerator's copper powder supply. it consumes 200 powder every 2 mins, so it should run at 1200/min for 10 seconds, then at 0 for 1 min 50 s

Maybe OC it so I can test for less time

[knotty hornet]

I think the issue you guys are describing is just cuz it does a rolling average.

[crimson moat]

Rolling average will give the correct answer after a long period of time. This doesn't. The mean on the display is much lower than the mean on the belt, even after 10x+ the rolling average time.

[wind spade]

you need to define a timeframe, and count amount of things in that timeframe. If the production is not consistent (e.g. recipe time is more than the timeframe), you get "issues"

[unique cypress]

Actually, do y'all think I should take the number shown in the GUI or on the outside display?

[unique cypress]

I mean my idea is to add a timestamp to a list every time an item passes by, then check the oldest timestamp in the list if it's older than the averaging period. if it is, remove it from the list

then average is the length of the list divided by averaging time

but at 1200/min and 1 min averaging time, that's a 1200 item long array that has to be stored for every counter

so that's probably not the best idea for something a user might place hundreds of

[wind spade]

what I was talking about is a problem no matter the implementation

if you make a batch of 10 every 64 seconds, and measuring time is 60 seconds, your measurements will jump up and down

[unique cypress]

I mean that's what already happens

[wind spade]

same goes for anything else - unless the output is fixed (e.g. from miner or full container), the "average" will change over time