Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
The author of this topic has marked a post as the answer to their question.
Discussions Rules and Guidelines
Report this post
Furthermore, you are dramatically limiting your throughput, as only 1 train can use a straight section of track at a time.
The layout & signal setup you show in your pics can only support 3 trains total safely, if you try to use more than that & they attempt to all cross the same junction / section of track at once, it will very likely result in a deadlock situation.
Generally speaking: yes, if you want decent throughput (though a 2-way rail line is an excellent budget option for early game, when steel is scarce & throughput requirments are very low);
& yes ~ they are smart enough to avoid collisions with other automatic trains ~ they actually "reserve" track ahead of themselves up to their braking point (notice the signals turning yellow when an automatic train approaches, but they do not turn yellow if a manually-driven train approaches). They will try to avoid manually-driven player trains too, but sometimes there is just not enough braking room / alternate route for them to avoid a collision.
Though, it should be easy for you to convert that double-track system to double one-direction rails. Like an IRL highway system / motorway. One side is outbound lane, the other is the return lane. ~ This allows a much denser level of traffic on the rails (trains can be nose-to-tail), thus far higher throughput can be achieved for similar construction costs.
edit ~ formatting errors
If you use personal construction bots, you can program a deconstruction planner to remove only the unwanted signals (R-click the red planner to set filters), & use a blueprint to apply the new 1-way signalling for your junctions &etc afterwards. I found them to be a great time-saver, while designing my own rail blueprints book.
There are a few ways to do this:
1> Belt splitter system: Take the mine output belt & split into 2 belts, split each of those to 2 more. Giving 4 total belts, each with 1/4 of the mines output. Route each 1/4 belt to a different cargo wagon for even loading. Scale-up as needed based on mine output Vs number of wagons. (You are lucky you chose a power-of-two number of carriages, as that is super simple to split-ter evenly. Odd numbers take up more room, due to the belt balancers requiring feedback loops.)
2> A simple control circuit: 1 combinator linked to the loading-into-chest inserters, that keeps them roughly evenly balanced based on the chest contents, thus ensuring even wagon loading. ~ Has same effect as #1, but smaller footprint & can support odd numbers of wagons much more easily. (Can supply a sample BP or pic, if you require.)
3> Logi Bots. (Though this works much better imo, when combined with #2 ~ using a "set requests" control circuit).
You can use similar setups for unloading too, with a few tweaks.
Personally, I prefer belt based with smart circuit (#2) for loading at mines & other low-throughput items, and prefer to use bot based (#3) for unloading (especially bulk items), as the station footprint is smaller (though the power requirements are much higher when using bots).
I have a very-high-speed station design featuring active providers & circuit-based inserter throttling. Kinda requires a good understanding of the logistics network & (virtual) circuit mechanics to setup / work well, but always keeps the wagons perfectly balanced & gives excellent throughput (above 4 blue belts per wagon for stack50 items such as ores, 5 per wagon for stack 100, 6-7 per wagon for stack 200).
Personally, I went and looked up a book of belt balancer blueprints.
I usually run LCCCC trains, but I'm trying LLCCCC in my current game (with one LTT for sulfuric acid to uranium mines; and a couple LCC for a few other purposes). So however many lined of belts I've got coming off of my ore patch, I run it through a balancer designed with that many belts input, and 4 belts output. If there are 6 lines of belts from the mine, I run a 6 to 4 balancer. If the patch is tiny enough I only get 2 or 3 lines from it, then I'll run that to 4 balancer. This ensures that the cargo wagons get an even distribution of the raw material.
Unloading I tend to run them through balancers as well, before they go into the smelting column. This allows for relatively smooth unloading of a train, as long as at least one smelting column is looking for the ore.
You can try to synchronize the arm movements of your stack inserters if you wish, via the circuit network. I don't really see any point in doing this. Of course, once you start using bots for your loading/unloading, balancing becomes a different proposition entirely, one more about placement of supply chests vs demand chests, since bots tend to take from the closer box if multiple boxes containing teh desired item are present.
https://steamcommunity.com/sharedfiles/filedetails/?id=1590470222
A lot of people rag on roundabouts but I've never had any issues with them.
You actually stated the #1 reason for synchronising inserters in the same paragraph:
This is why I use throttled inserters into active providers ~ they only remove as much as is needed from the train to meet the current demand, meaning the chests never get over-filled & so the wagons always unload evenly (unlike with a basic passive provider setup). Chest position does not matter at all with this approach.
It is actually easier to control the inserters via the logistic network, no wires or combinators are needed to make it work.
No. AI pathfinding will advance as far as possible, and since you allow trains from either side to stop in either lane they can block both lanes.
Wait, are you saying that you want nearly empty trains to hang around instead of finishing the job and doing something useful? Are we trying to minimize or maximize throughput here?
Trains are very useful, once you get the hang of signalling & settle on a consistent design to use in that world.
I suggest you take Alex's advice & re-design your 4-way & T-junctions though, so you can fit chain signals inside to break up the blocks & allow better train throughput. It really makes a difference when you have dozens of trains zipping about.
I think you misunderstood the purpose behind it.
I use a 10-seconds-worth item buffer amount programmed into the station (to cover train out-in cycle downtimes). As items are placed onto the output belts, the same amount is removed from the train wagons evenly, maintaining that 10sec buffer but avoiding buffer overflow.
The setup I briefly described is designed to cope with any level of throughput requirement from zero to maximised, automagically.
Trains are set to "wait until empty". (Why would I send a partly-unloaded train anywhere else? It is already doing something useful by being at the unloading stop with some cargo ~ exactly where it is needed.)
edit - formatting. again.
That rails guide is good, but there is a better / more-comprehensive one for rail networks available on reddit / imgur ~ https://www.reddit.com/r/factorio/comments/4f38sk/factorio_train_automation_complete_parts_23_and/
Here are a couple of pics for you. I use a 6-tile-gap LHD network, but the signalling / blocks principle is the same for a 4-gap &/or RHD versions (though the track positions might need adjusting to fit in the chain signals, especially for a RHD 4-way).
4-way ~ https://i.imgur.com/5qJoC3L.png?1
Ultra-compact, has the same footprint as a roundabout (making it easy to use the cheaper roundabouts in early game, then upgrade to a full junction later when there are more trains active).
T-junction ~ https://i.imgur.com/Ayrvuhw.png?1
Note: the signals on that T-junction are in slightly odd positions, as it is designed so that the 4-way can simply be placed over the top to upgrade it, & without having to remove anything. (They could be placed closer, for a slight throughput gain.)
I also have some straight-to-diagonal V- & Y- junctions ~ https://i.imgur.com/qA22tIa.png?1
But, securing more mines than you currently need is a good idea regardless ~ the more you link up, the better (within reason, heh). Miners have zero electricity drain when idle, thank the devs, so there is not really any downside to building extras before you actually need them.
You can get a rough eyeball estimate by counting the number of full belts of ore output from each mine, & comparing that you your smelter inputs (assuming the belts are flowing freely & mostly fully compacted). The (P)roduction overview screen might also help to give a better idea.
For a rough ratio (if similar ore patch amounts): 4x iron ore : 2-3x copper ore : 1x coal : 1x stone. Scale-up & adjust the ratio as needed for your mapgen / tech progress / &etc.
At least you can drive the car (or tank) ~ that is one part of the game I still suck hard at. >_<