Perfect Division Belt-Fed Storage Facility

[PyroFire]

So, i'm trying to figure out how to perfectly split a max-compression express belt into a perfect load balanced storage facility, which can then unload perfectly onto a max-compression express belt.

This is my original design:


But as you can see there's a few small problems with it.
As material enters inserter range at the bottom, the inner inserters will always receive priority over the outter ones resulting in a poorly balanced offload.

Based on tests and information from the wiki, it is confirmed that 9 blue inserters for both sides of the express belt will fill it to maximum capacity, with the 9th inserter on both sides operating at around 60% capacity.
In other words 18 total inserters.
Tested and confirmed that removing 1 then load balancing the 17th will not fill an express belt to max, you really need 18.
Visually, this is the optimal layout for a logistics-fed express belt loader/offloader:


Now onto the math problem.
How do we turn 2 into 3?

9/2 = 4.5

We can visually see this as the 4 whole numbers (1 per belt on each side) then the 0.5 belt in the middle in the following picture;

In other words, the middle belt only needs half (0.5) of what the 4 outter belts (1 per) need.


After some testing, it appears there is no difference with throughput when switching to smart inserters.


How do?

This is the best i've come up with so far, with the help of the inserter stack size bonus.


But even this has a problem of resources building up at the very last chest meaning it will not always output a max capacity express belt.


** SOLVED **

With some help from my friends, it was suggested to me to use a train as the offload point which solves any possible resource distribution problems from the previous image.
The following is the best load balanced express belt onload/offload storage facility that can possibly be made that doesn't have any distribution issues.

[siggboy]

Use your original design (top picture), and use this circuit (not by me, Neotix made it): 23125 (this is a very simple and elegant circuit, you only need two combinators for the entire setup)

This will distribute the input evenly among the bottom row of chests (they need to be smart chests in 0.12 and of course the inserters need to be smart as well).

The rest of the setup looks fine, at the top (exit) a simpler belt setup is sufficient to make a compressed belt:

Screenshot-2.png (299.28 KiB) Viewed 7363 times

[PyroFire]

Use your original design (top picture), and use this circuit (not by me, Neotix made it): 23125 (this is a very simple and elegant circuit, you only need two combinators for the entire setup)

This will distribute the input evenly among the bottom row of chests (they need to be smart chests in 0.12 and of course the inserters need to be smart as well).

The rest of the setup looks fine, at the top (exit) a simpler belt setup is sufficient to make a compressed belt:

Screenshot-2.png

This has the same problem as my original design;
The chests at the edges aren't evenly filled, which means if the input belt drops below max capacity, so does the output.

Original post has been updated with the most efficient and effective design for a max-compression express belt storage facility.
Finally figured it out

[siggboy]

This has the same problem as my original design;
The chests at the edges aren't evenly filled, which means if the input belt drops below max capacity, so does the output.

Have you tested it? Are you aware that the circuit guarantees a balanced load in the chests?

[PyroFire]

This has the same problem as my original design;
The chests at the edges aren't evenly filled, which means if the input belt drops below max capacity, so does the output.

Have you tested it? Are you aware that the circuit guarantees a balanced load in the chests?

I don't see an input belt used in your design.
And having a look at the linked thread, it seems to be rather messy compared to what i've come up with.
Though i must say the output belts on mine need improving :p

Upon closer inspection, the thread linked shows belts that don't exist in vanilla;


And there are 2 major load-balancing problems with his prototype design, also seen in his final design;


I thought of using a self-feeding splitter to solve the 2 into 3 problem, but then i realized doing that would only back-up the input belt.
If you have any resources finding itself back to the input lane, then you're not fully emptying the express belt.

And finally, if the input belts are not flowing at max capacity, then the inserters at the bottom will not feed the chests correctly resulting in said load balancing issue.

Seemed like a cool design but after breaking it down it's more complicated and not as effective

[siggboy]

I don't see an input belt used in your design.
And having a look at the linked thread, it seems to be rather messy compared to what i've come up with.
Though i must say the output belts need improving :p

So you haven't really read what I have written, neither did you try to understand the essential part of my solution (chest balancing circuit), and yet you dismissed the proposed solution out of hand... ugh

[dee-]

Hi, I just saw your thread.

This is the layout I've been using for some time now:



It's balanced and unloads trains (or chests for that matter) evenly up to blue belts speed -- not fully compressed but quite full. Adding perhaps two more train trailers should bring it up to compressed speed (?).

There's also a corresponding loading layout which distributes incoming blue belts evenly into chests.

This is one module from my station blueprints.

[siggboy]

What's this "nixie tubes" style display? Is it the original "Nixie Tubes" mod with modified graphics? I love those graphics, where can I download it?

[dee-]

What's this "nixie tubes" style display? Is it the original "Nixie Tubes" mod with modified graphics? I love those graphics, where can I download it?

Yes, they're beautiful and I prefer them myself. Link

[siggboy]

Thanks man, it was on page 7 of the thread, I would not have found it alone .

BTW about your unloading station: you have 48 fast inserters, which is not enough to fully compress 4 blue belts. If the belts are compressed it's because they are backed up.

[dee-]

Thanks man, it was on page 7 of the thread, I would not have found it alone .

BTW about your unloading station: you have 48 fast inserters, which is not enough to fully compress 4 blue belts. If the belts are compressed it's because they are backed up.

Indeed not fully compressed blue belts.
I edited above post

[PyroFire]

I'm surprised how Everyone in this thread managed to COMPLETELY miss the point i was making.

As i said, and allow me to repeat myself and make it clear,
Belt-Fed storage facility that can Receive any amount of material, regardless of compression, and still Output with 100% Compressed Blue Belts until it is empty.
As in, i was talking about the mathematical precision and thus exploitable blueprint that can be created to achieve this.
Not a "it's close enough to work".
Maths, boys.
Maths.

I wasn't talking about trains, i wasn't talking about making belts max capacity for the sake of it, nor ensuring any sort of fill methods.

I was talking about one thing very specifically.

How do you get a fully compressed blue belt to offload into chests for storage/buffering, and then back onto the blue belts without any loss of compression?
AND accounting for the, if the input belt drops below max compression, due to the perfect load balancing, the output belt from the chests regardless do not drop below max belt capacity until the entire storage facility is completely empty.
EVERYONE who posted in this thread completely missed this point.

But anyway...

This problem is now obsolete with the new stack inserters making it extremely easy to build belt-fed storage facilities.
However, i would like to illustrate with this thread how it used to be extremely difficult, near impossible prior to the stack inserter update.
My OCD can rest easy now.

Thanks factorio devs!