Belt Designs for ColonelWill

[MadZuri]

Worked on this a bit lat night. I decided to post results. Enjoy:

[MadZuri]

addendum

and also a closer look. Those fast belts limit edge throughput and forces items into center.

[MadZuri]

edit: accidentally deleted this picture...

[Piranha]

Your Design with Percentage

..

My Simplified Design with Percentage

..

But would this not work also with far less material

..
Yellow lines could have another splitter that wouldn't do anything except warp the material through a bit faster then the belts that are there

Also, I feel like the last one could be improved with underground belts to make it more compact but MS Paint is hard enough already.

And lastly, I would have to guess a design that doesn't have a single percentage over 100 would be the best result because technically anything over 100% is actually just causing a small back log so although you have 100% compression it is a result of backing up the throughput at that specific point to force compression to be full again.. but I guess the corners are losing some compression.. so you do need a bit over 100.. if someone could figure out the exact number you are losing aka the compression percentage after a belt turn then you could tweak the design again to fully optimize the throughput.



If I am missing something here please let me know
I do not work on any of this in-game but enjoy reading these discussions on the forums.

[MadZuri]

Tested your designs:

Your compact design

..

Your fractions are off on your math. You're putting 2 different numbers for the 2 outputs of some splitters. Just add the 2 inputs and split them evenly. Also, compression loss is ~33% on corners, so anything above ~67% will hang up. The max output/input of any one side of the splitter is 100%, but back-calculating the uneven split is difficult.

Your first design

..

Your first one worked surprisingly well, only minor delays occasionally. The compression loss on the corners forces the entry splitters to force more out to the wings, decreasing the number of entry splitters required. I over-engineered that, and got the numbers to almost exactly 67% on all 6 lines that turn. I suppose that wasn't necessary, and you used basically the same exit that I did. Good job.

[MadZuri]

If you want a relatively cheap and effective corner, this will work:

Really

..

It relies heavily on that back-pressure mechanic, and works at about 95% throughput.

[Piranha]

Interesting results, thanks for testing

And your last one is the same as my last one just with the last 2 splitters reversed.. also interesting.. and that 33% is now a number I can think about.. Thanks

[ryannayr140]

I believe madzuri's design from last night would work best. The left is where the ore will attempt to go and the right is where I believe ore will go with corner bottlenecks in the center.

http://i.imgur.com/9EWQS15.png?1

[Xecutor]

The best way to handle so much iron is to use boxing mod

[ssilk]

Definitely. It adds also a new type of game-play, cause you have to handle the empty boxes.

[MadZuri]

Some more designs for Will:

so many things

At first I didn't think that the combiner would work if I just literally put the splitter in reverse. After designing an entirely too complicated merger system, I tried it, and it works magnificently. There is a slight change, but I was surprised it worked. And yes, that 4 to 12 splitter system equally divides the stream. Ignore the thing to the right, don't know why I included it in the pic, just messing around with strange lossless belt designs.

[MadZuri]

16 Lane count-perfect rebalancer freaking finally

ant sized

[ssilk]

That pic is barely readable. Turn on daylight/place lamps, before make screenshoots! :)https://forums.factorio.com/wiki/inde ... and_videos

[ColonelWill]

I can see it perfectly clearly and from that pic recreated it for use in my mega base. cheers Madzuri !

[Talguy]

LMAO Madzuri, once again you have outdone yourself. I mean, wtf.

[MadZuri]

anything I can do, I can do better....

compact

well lit

[Koub]

anything I can do, I can do better....

well lit

Hahaha love that one

By the way, you're a true psycopath Beware : one day, you'll build a balancer so big it won't fit on screen, and you won't be able to screenshot it :p

[MadZuri]

Hahaha love that one

By the way, you're a true psycopath Beware : one day, you'll build a balancer so big it won't fit on screen, and you won't be able to screenshot it :p

Well, nobody has asked for a 100 lane rebalancer, and I doubt anyone has ever had the need for one. I did complete the T junction I said I was going to build. It may look familiar:

2x4to8

[quinor]

This is mine 16-rebalancer, 24x35 tiles with 4 tiles of horizontal slide. I couldn't keep it such "small" without it

16-16 evener

[quinor]

Also, I just came up with some small (9x13 tiles) 8-rebalancer. Smaller 16-one feels quite hard...

8-8 evener