Choosing train and track sizes
-
- Smart Inserter
- Posts: 1161
- Joined: Sat Mar 17, 2018 1:20 pm
- Contact:
Re: Choosing train and track sizes
I spent the day prepping it and just posted my mainline design / blueprint for the linear concept.
The forum main discussion page for it: viewtopic.php?f=194&t=63070
The factorioprints page: https://factorioprints.com/view/-LPOOJ-GsdY4h7KTnHgc
This might be useful for the work you are doing.
Also, even though it is not the plan to use it, it includes an interesting intersection that might (or might not) fit as a higher throughput replacement for your roundabouts (it is in the blueprint book).
This is a screenshot of the final version of the main line segment (with exit lanes):
And if you need my full rail blueprints let me know. I have not compiled them into a well organized book yet, but I can do that for you if after seeing the above stuff you still feel you might need them (probably not... I found myself using nothing but the above blueprint book now, well, plus my new stations... but those I don't consider "good enough to share" yet). But I will prep and publish if you need them.
The forum main discussion page for it: viewtopic.php?f=194&t=63070
The factorioprints page: https://factorioprints.com/view/-LPOOJ-GsdY4h7KTnHgc
This might be useful for the work you are doing.
Also, even though it is not the plan to use it, it includes an interesting intersection that might (or might not) fit as a higher throughput replacement for your roundabouts (it is in the blueprint book).
This is a screenshot of the final version of the main line segment (with exit lanes):
And if you need my full rail blueprints let me know. I have not compiled them into a well organized book yet, but I can do that for you if after seeing the above stuff you still feel you might need them (probably not... I found myself using nothing but the above blueprint book now, well, plus my new stations... but those I don't consider "good enough to share" yet). But I will prep and publish if you need them.
Re: Choosing train and track sizes
Nice! thankyou. I can see some benefits of the split to 6 rails as per your picture.
I've already started redesigning (below) with my new plan, and if that doesnt work out, i'll probably revisit that design you posted - thanks!
Heres my new intersection design. Only converted 5% of base so far, but immediately I can see massive potential.
Its perfect "exact 3 chunk" BP, and I also have a "exact 7 chunk wide" 4-track main line BP which doesnt have any u-turns, just merge between inner and outer tracks, and off-shot input/output tracks to each zone. (2 zones on one side of a track and 2 on the other side). Similar concept as earlier screenshots, but much bigger zones, and wider tracks (to match the intersection image below). Will post an image in a few days - things are bit of a mess right now lol.
Very happy with initial results, but we'll see after I finish converting everything (probably a week lol).
I've already started redesigning (below) with my new plan, and if that doesnt work out, i'll probably revisit that design you posted - thanks!
Heres my new intersection design. Only converted 5% of base so far, but immediately I can see massive potential.
Its perfect "exact 3 chunk" BP, and I also have a "exact 7 chunk wide" 4-track main line BP which doesnt have any u-turns, just merge between inner and outer tracks, and off-shot input/output tracks to each zone. (2 zones on one side of a track and 2 on the other side). Similar concept as earlier screenshots, but much bigger zones, and wider tracks (to match the intersection image below). Will post an image in a few days - things are bit of a mess right now lol.
Very happy with initial results, but we'll see after I finish converting everything (probably a week lol).
- Attachments
-
- Untitled.png (1.34 MiB) Viewed 5876 times
Re: Choosing train and track sizes
Concerning small ore patches. As said further out they get bigger. As you start with 1:2 trains filling them is no problem. Then as you expand and get bigger patches you can fill longer trains.
Another option is to mine multiple ore patches and combine the output into a single train. Either by belts or bots or by having a number of small trains deliver the ores to a trans shipment center that fills a large train.
Always have a locomotive at the front of the train because they have less wind resistance than train cars. Other than that it doesn't matter where you put locomotives. Acceleration and speed are the same no matter where the rest are. I recommend having a single locomotive at the start of the train. Keeps the train stop near the inserters and any circuit connection you might plan will be short.
Another option is to mine multiple ore patches and combine the output into a single train. Either by belts or bots or by having a number of small trains deliver the ores to a trans shipment center that fills a large train.
Always have a locomotive at the front of the train because they have less wind resistance than train cars. Other than that it doesn't matter where you put locomotives. Acceleration and speed are the same no matter where the rest are. I recommend having a single locomotive at the start of the train. Keeps the train stop near the inserters and any circuit connection you might plan will be short.
-
- Smart Inserter
- Posts: 1161
- Joined: Sat Mar 17, 2018 1:20 pm
- Contact:
Re: Choosing train and track sizes
I started using the multiple ore patches for a single train approach not long ago and it works great. I make a small station with up to 5 stops (pax, iron, copper, stone, coal or fewer if I produce only a couple ores at a time since I can recycle stops once the early ore is exhausted) and make it tap several deposits. This approach provides enough ore to support 4 wagon trains from the start... which keeps the smelters working better than with 1-2 trains.mrvn wrote: ↑Mon Oct 22, 2018 9:48 am Concerning small ore patches. As said further out they get bigger. As you start with 1:2 trains filling them is no problem. Then as you expand and get bigger patches you can fill longer trains.
Another option is to mine multiple ore patches and combine the output into a single train. Either by belts or bots or by having a number of small trains deliver the ores to a trans shipment center that fills a large train.
Always have a locomotive at the front of the train because they have less wind resistance than train cars. Other than that it doesn't matter where you put locomotives. Acceleration and speed are the same no matter where the rest are. I recommend having a single locomotive at the start of the train. Keeps the train stop near the inserters and any circuit connection you might plan will be short.
I like the 1-4-1SD (same direction locomotive) trains, for the reason you mention plus the ability to leave the rear locomotive on the entry curve rail... making the station a little more compact.
-
- Smart Inserter
- Posts: 1161
- Joined: Sat Mar 17, 2018 1:20 pm
- Contact:
Re: Choosing train and track sizes
@evopwr
This stuff gets big even without trying to make a megabase.
I run out of research even with my little starter base, so I built real beaconed science and rocket to at least keep researching artillery range and the like. Next thing I know I am spitting out satellites like candy (close to 400 now in about 2-4 days), consuming 3GW and it is very obvious that I need to expand my already pretty significant beaconed smelters for iron plate and copper plate. But before I can do that I need to extend my train line to the far distant good ore deposits (clearing massive numbers of nests in the process) and finalize my plans for processing and delivering the materials to my base.
On the other hand, good news... I am well past 100 1-4-1 trains and my rail network is handling it like if it is nothing.
I am mulling over the size of the long distance trains. I might just skip over the (1-4-1)x2 and go straight to (1-4-1)x4 trains for the distant trains... and design the system to support (1-4-1)xN, where N can grow to anything. I have a vague idea in mind. We'll see if it pans out. Sometimes "crazy" works
This stuff gets big even without trying to make a megabase.
I run out of research even with my little starter base, so I built real beaconed science and rocket to at least keep researching artillery range and the like. Next thing I know I am spitting out satellites like candy (close to 400 now in about 2-4 days), consuming 3GW and it is very obvious that I need to expand my already pretty significant beaconed smelters for iron plate and copper plate. But before I can do that I need to extend my train line to the far distant good ore deposits (clearing massive numbers of nests in the process) and finalize my plans for processing and delivering the materials to my base.
On the other hand, good news... I am well past 100 1-4-1 trains and my rail network is handling it like if it is nothing.
I am mulling over the size of the long distance trains. I might just skip over the (1-4-1)x2 and go straight to (1-4-1)x4 trains for the distant trains... and design the system to support (1-4-1)xN, where N can grow to anything. I have a vague idea in mind. We'll see if it pans out. Sometimes "crazy" works
Re: Choosing train and track sizes
@zOldBulldog,
heh, yeh, agreed.
I also ran out of research in my starter base, weeks ago - although I don't have biters on, so never bothered with combat research, and instead just did mining productivity and robot speed.
Having said that 2 days ago I doubled my labs (researching robot speed 13 at the moment: 256,000 science!!!), and surprisingly all science producing areas are still keeping up. So I now have 32 labs, each of them covered by 4 beacons (plus max science research speed), so I think they're +670% speed from memory.
I actually switched on research for combat stuff last night for a laugh (I'm using auto-research mod), and it just went "bing bong" every 2-3 seconds as it clicked through researching all the techs. They were all done (except infinite ones) in 2 minutes. lol. (wife got angry at all the bing bong noise interrupting her tv lol)
If you remember, my train design a week ago was a failure at the intersections, so I have spent most of the last week completely bulldozing and re-building the entire base (after giving you shit about your storage chests, I reckon I probably have more than you now lol). That's all done now, and happy to say its working fine (but whether it will stay that way who knows).
Currently using 3.5GW, but I just put down a lot more reactors last night in readiness for a big expansion. Also changed my perimeter ore track to be 3 wide, and expanded the south perimeter further out in readiness for expansion tonight. Also re-designed my Steel and Copper smelting to be more modular/expandable, and greater capacity, also in readiness. Still to do iron.
Silos are sustainably spitting out 1 rocket per minute. The launch time is quite frustrating. Making 100 rocket parts in ~40 seconds, and then the ~20 second launch animation kicks in.
Heres current. Still a bit of tidying up to do, but far better than last week:
heh, yeh, agreed.
I also ran out of research in my starter base, weeks ago - although I don't have biters on, so never bothered with combat research, and instead just did mining productivity and robot speed.
Having said that 2 days ago I doubled my labs (researching robot speed 13 at the moment: 256,000 science!!!), and surprisingly all science producing areas are still keeping up. So I now have 32 labs, each of them covered by 4 beacons (plus max science research speed), so I think they're +670% speed from memory.
I actually switched on research for combat stuff last night for a laugh (I'm using auto-research mod), and it just went "bing bong" every 2-3 seconds as it clicked through researching all the techs. They were all done (except infinite ones) in 2 minutes. lol. (wife got angry at all the bing bong noise interrupting her tv lol)
If you remember, my train design a week ago was a failure at the intersections, so I have spent most of the last week completely bulldozing and re-building the entire base (after giving you shit about your storage chests, I reckon I probably have more than you now lol). That's all done now, and happy to say its working fine (but whether it will stay that way who knows).
Currently using 3.5GW, but I just put down a lot more reactors last night in readiness for a big expansion. Also changed my perimeter ore track to be 3 wide, and expanded the south perimeter further out in readiness for expansion tonight. Also re-designed my Steel and Copper smelting to be more modular/expandable, and greater capacity, also in readiness. Still to do iron.
Silos are sustainably spitting out 1 rocket per minute. The launch time is quite frustrating. Making 100 rocket parts in ~40 seconds, and then the ~20 second launch animation kicks in.
Heres current. Still a bit of tidying up to do, but far better than last week:
- Attachments
-
- Untitled1.png (324.17 KiB) Viewed 5765 times
-
- Smart Inserter
- Posts: 1161
- Joined: Sat Mar 17, 2018 1:20 pm
- Contact:
Re: Choosing train and track sizes
Nice. I was having a bit of trouble keeping up with the plate demand, and rather than expanding my smeltery again I started on my next bring project... bringing stuff in from far away where deposits are bigger (and probably setup a monster smeltery there too). So I'm letting my research and rockets fail due to lack of resources while I do that.
I was debating how to bring materials "to the current base" without redesigning everything, so I settled on a transboard station. I think I'll bring in a 32-wagon (16 locos) train with all of the materials and let my trains that currently go to the original smeltery pick up from there (use the same name for the smelt out stops and the transboard out stops). I know it is not a good long-term design but it is a fun project... so I don't care Plus, I am designing it so that it is infinitely expandable by number of bays or length of trains... so it should be useful for any "transition" times between one size and the next. The basic concept is that each group of 4 wagons is one material... and some creative circuitry to make the process efficient.
As part of the project I am working on a way to isolate parts of my rail network, with circuits to ensure that any trains that try to sneak through the border will find a red light and get stopped by the border police Sure, I might eventually need to go clean up and send them to the right places, but it is a good way to discover if I have to beef up the rail penalties for cross-border traffic. Right now I am using I think 4 or 5 unnamed stops to make it look brutally expensive unless it is the only place to go on the route.
Crazy stuff, possibly sub-optimal, but very fun. If it is no good I'll destroy it.
As to my rail, I updated the blueprints in Factorioprints. The intersection now clocks about 60 trains per minute on random traffic, and about 109 on left turns (which is most of what I should be getting anyway). Plus I vastly reduced traffic to the intersection, almost everything is going straight.
Pretty soon I'll end up removing the intersection altogether.
I was debating how to bring materials "to the current base" without redesigning everything, so I settled on a transboard station. I think I'll bring in a 32-wagon (16 locos) train with all of the materials and let my trains that currently go to the original smeltery pick up from there (use the same name for the smelt out stops and the transboard out stops). I know it is not a good long-term design but it is a fun project... so I don't care Plus, I am designing it so that it is infinitely expandable by number of bays or length of trains... so it should be useful for any "transition" times between one size and the next. The basic concept is that each group of 4 wagons is one material... and some creative circuitry to make the process efficient.
As part of the project I am working on a way to isolate parts of my rail network, with circuits to ensure that any trains that try to sneak through the border will find a red light and get stopped by the border police Sure, I might eventually need to go clean up and send them to the right places, but it is a good way to discover if I have to beef up the rail penalties for cross-border traffic. Right now I am using I think 4 or 5 unnamed stops to make it look brutally expensive unless it is the only place to go on the route.
Crazy stuff, possibly sub-optimal, but very fun. If it is no good I'll destroy it.
As to my rail, I updated the blueprints in Factorioprints. The intersection now clocks about 60 trains per minute on random traffic, and about 109 on left turns (which is most of what I should be getting anyway). Plus I vastly reduced traffic to the intersection, almost everything is going straight.
Pretty soon I'll end up removing the intersection altogether.
Re: Choosing train and track sizes
@Bulldog, (apologies everyone else for this becoming a 2 person chit chat thread - feel free to join in tho!)
How are things progressing?
I've ramped mine up a bit, and seem to be sustaining my 32 +670% labs continuously, whilst simultenously producing approx 2.5 rockets per min.
Lots of tweaking on the way, but the main track system seems to be holding up, so far.
Currently using 6.7GW on Nuclear, and still never dropping under 60UPS.
180 trains at the moment.
Starting to see potential weakness on the perimiter ore track, which I dont think will allow me to double current Rockets PerMin.
With all the trains running out there, even with a 3 track (one way) perimiter, I can see congestion isnt far away. Plus the round trip is looong. (havent timed it, but estimate a minute or so).
Possible solution is a 2 way system, with efficient intersections, so trains dont have to go the whole way around the base (why go the whole way around when the ore its going to is right next to the smelting?). This would halve the number of trains needed, if the distance they need to go is halved and therefore halve congestion? great in theory... lol. I expect it all comes down to the intersection design, otherwise it will be a lot worse.
Not sure how much further i'll go - in the last few years Factorio has been my number 2 game, with Ark being my number 1. But I finished Ark, with all achievements, and defeated all bosses and end game stuff (3000+ hours). BUT... A new map/world is releasing in a couple of days, and the crew I played with are all getting back together to take it on. So i'll probably (?) be hooked on that for awhile (and as usual end up coming back to Factorio one day).
Hope your base is going well, send me a pic
How are things progressing?
I've ramped mine up a bit, and seem to be sustaining my 32 +670% labs continuously, whilst simultenously producing approx 2.5 rockets per min.
Lots of tweaking on the way, but the main track system seems to be holding up, so far.
Currently using 6.7GW on Nuclear, and still never dropping under 60UPS.
180 trains at the moment.
Starting to see potential weakness on the perimiter ore track, which I dont think will allow me to double current Rockets PerMin.
With all the trains running out there, even with a 3 track (one way) perimiter, I can see congestion isnt far away. Plus the round trip is looong. (havent timed it, but estimate a minute or so).
Possible solution is a 2 way system, with efficient intersections, so trains dont have to go the whole way around the base (why go the whole way around when the ore its going to is right next to the smelting?). This would halve the number of trains needed, if the distance they need to go is halved and therefore halve congestion? great in theory... lol. I expect it all comes down to the intersection design, otherwise it will be a lot worse.
Not sure how much further i'll go - in the last few years Factorio has been my number 2 game, with Ark being my number 1. But I finished Ark, with all achievements, and defeated all bosses and end game stuff (3000+ hours). BUT... A new map/world is releasing in a couple of days, and the crew I played with are all getting back together to take it on. So i'll probably (?) be hooked on that for awhile (and as usual end up coming back to Factorio one day).
Hope your base is going well, send me a pic
- Attachments
-
- Untitled.png (330.36 KiB) Viewed 5663 times
-
- Smart Inserter
- Posts: 1161
- Joined: Sat Mar 17, 2018 1:20 pm
- Contact:
Re: Choosing train and track sizes
@moderator:
- Maybe this should be moved to the Train/Rail Show Your Creations forum, as it shows thinking and evolution of the rail strategy for large bases.
@evopower:
I find it very cool how while we were both on the same trail to figure out how to do trains "well" we evolved in different directions.
Here is my current state and plans:
- I am still "not trying" to do a megabase, nor trying to launch a bunch of rockets. But from my choices of going the "rail and outpost" approach, building "beaconized" and mostly producing and carrying "full 4-wagon trains" of product I seem to have ended up with some kind of a megabase.
- Close to 120 trains (no congestion at all). Passed 900 satellites while typing this. Approx 400 Science per minute (when I have enough smelted materials). Biggest ore deposit found so far is 164M iron. Biggest oil deposit found is 32811%. Currently using two 1-4-1 artillery trains and it take a while to clear an area (I am starting to understand why people use huge artillery trains), but what is most surprising is how long it now takes the biters to reach me after I blow up their nests.
- I am happily progressing through my trial and error process of figuring out the best designs. You will probably notice very different approaches to the stackers as I try one and another, although I seem to be drifting towards a couple of designs that work well in most situations.
- My biggest challenge was to clear space for building. I just didn't want to build on top of all those small ore deposits. Now the area south of the main line is mostly clear, with much fewer ores that need to go. Sooner or later all of the stuff right of the transboard station and left of the Green Circuits/Oil factories will get removed/replaced with a cleaner layout. But all those ores have to go first.
- This is my transboard station under construction. 32-wagon trains will come in with coal, stone, uranium, iron plate, copper plate, steel and give the materials to the 4-wagon trains used in the base. The base and west rail networks are isolated, with crossovers only permitted for Pax/Supply trains or trains I am permanently moving from one network to the other.
- And here is the full map, showing the west branch I cleared to reach the good mining deposits. Around the far end of the west branch will be the mining loops with their ore load stations, a bigger and infinitely expandable smelter (I am still not producing enough iron and copper plate to fully support my needs), and if necessary a n-wagon to 32-wagon transboard station (I don't know yet how big are the trains that can be supported by the mines in this area... guessing 8-wagon, but maybe bigger).
I started using some quality of life mods that I expect to always use:
- Bottleneck, Progressive Running, Squeak Through. Simple stuff, but essential.
- Miniloader, because I am tired of building complex train load/unload beasts to compensate for the game's limitations. With this one in 1 block I get a full blue belt load/unload.
- I have not started using it yet, but I plan to use Luzivras Factorio Power. It gives 8 tiers of solar panels and accumulators with the top one being over 100GW. My hope is that it will let me concentrate all power production into one chunk, without the know UPS hit of Nuclear. With any luck it is a matter of building a production line that constructs them while I am doing something else. Then once I have a solar panel and accumulator that can replace my nuclear plants... plop it down and "good bye power concerns".
- I am also intrigued by Helmod and Stop That Silly Robot. They are in my "to test" list.
- Maybe this should be moved to the Train/Rail Show Your Creations forum, as it shows thinking and evolution of the rail strategy for large bases.
@evopower:
I find it very cool how while we were both on the same trail to figure out how to do trains "well" we evolved in different directions.
Here is my current state and plans:
- I am still "not trying" to do a megabase, nor trying to launch a bunch of rockets. But from my choices of going the "rail and outpost" approach, building "beaconized" and mostly producing and carrying "full 4-wagon trains" of product I seem to have ended up with some kind of a megabase.
- Close to 120 trains (no congestion at all). Passed 900 satellites while typing this. Approx 400 Science per minute (when I have enough smelted materials). Biggest ore deposit found so far is 164M iron. Biggest oil deposit found is 32811%. Currently using two 1-4-1 artillery trains and it take a while to clear an area (I am starting to understand why people use huge artillery trains), but what is most surprising is how long it now takes the biters to reach me after I blow up their nests.
- I am happily progressing through my trial and error process of figuring out the best designs. You will probably notice very different approaches to the stackers as I try one and another, although I seem to be drifting towards a couple of designs that work well in most situations.
- My biggest challenge was to clear space for building. I just didn't want to build on top of all those small ore deposits. Now the area south of the main line is mostly clear, with much fewer ores that need to go. Sooner or later all of the stuff right of the transboard station and left of the Green Circuits/Oil factories will get removed/replaced with a cleaner layout. But all those ores have to go first.
- This is my transboard station under construction. 32-wagon trains will come in with coal, stone, uranium, iron plate, copper plate, steel and give the materials to the 4-wagon trains used in the base. The base and west rail networks are isolated, with crossovers only permitted for Pax/Supply trains or trains I am permanently moving from one network to the other.
- And here is the full map, showing the west branch I cleared to reach the good mining deposits. Around the far end of the west branch will be the mining loops with their ore load stations, a bigger and infinitely expandable smelter (I am still not producing enough iron and copper plate to fully support my needs), and if necessary a n-wagon to 32-wagon transboard station (I don't know yet how big are the trains that can be supported by the mines in this area... guessing 8-wagon, but maybe bigger).
I started using some quality of life mods that I expect to always use:
- Bottleneck, Progressive Running, Squeak Through. Simple stuff, but essential.
- Miniloader, because I am tired of building complex train load/unload beasts to compensate for the game's limitations. With this one in 1 block I get a full blue belt load/unload.
- I have not started using it yet, but I plan to use Luzivras Factorio Power. It gives 8 tiers of solar panels and accumulators with the top one being over 100GW. My hope is that it will let me concentrate all power production into one chunk, without the know UPS hit of Nuclear. With any luck it is a matter of building a production line that constructs them while I am doing something else. Then once I have a solar panel and accumulator that can replace my nuclear plants... plop it down and "good bye power concerns".
- I am also intrigued by Helmod and Stop That Silly Robot. They are in my "to test" list.
-
- Fast Inserter
- Posts: 123
- Joined: Sun Feb 26, 2017 10:48 pm
- Contact:
Re: Choosing train and track sizes
@evopwr
Xterminator did a series on sending supporters to space. The latest video shows one rocket every 10 seconds, so a decent sized base.
One thing he did was take all iron from south east of his base and all copper from south west. So if he found a great copper patch in the south east he ignored it. This kept the iron and copper ore trains from crossing paths and reducing congestion.
What you could do is use copper North West of your base and Iron North East of the base and straight into your smelters. If you then move the smelters outside the perimeter ring you should reduce the traffic on the perimeter.
I do not know how this would work with the constraints you are imposing on yourself, and it would make part of the base outside your pretty square.
Xterminator did a series on sending supporters to space. The latest video shows one rocket every 10 seconds, so a decent sized base.
One thing he did was take all iron from south east of his base and all copper from south west. So if he found a great copper patch in the south east he ignored it. This kept the iron and copper ore trains from crossing paths and reducing congestion.
What you could do is use copper North West of your base and Iron North East of the base and straight into your smelters. If you then move the smelters outside the perimeter ring you should reduce the traffic on the perimeter.
I do not know how this would work with the constraints you are imposing on yourself, and it would make part of the base outside your pretty square.
Re: Choosing train and track sizes
I'll let the topic there, it's somewhere between a discussion, a gameplay help, and a showcase of what you have done, so I feel General discussions still is the right place for it to be .zOldBulldog wrote: ↑Sat Nov 03, 2018 1:44 pm @moderator:
- Maybe this should be moved to the Train/Rail Show Your Creations forum, as it shows thinking and evolution of the rail strategy for large bases.
Koub - Please consider English is not my native language.
-
- Smart Inserter
- Posts: 1161
- Joined: Sat Mar 17, 2018 1:20 pm
- Contact:
Re: Choosing train and track sizes
Okie dokie.Koub wrote: ↑Sat Nov 03, 2018 10:59 pmI'll let the topic there, it's somewhere between a discussion, a gameplay help, and a showcase of what you have done, so I feel General discussions still is the right place for it to be .zOldBulldog wrote: ↑Sat Nov 03, 2018 1:44 pm @moderator:
- Maybe this should be moved to the Train/Rail Show Your Creations forum, as it shows thinking and evolution of the rail strategy for large bases.
Whatever this really is, I hope it is useful to others. Bouncing ideas with evopwr and trying to find "a better way" to do rail has certainly been useful at least to me.
There is so much info about the micro level designs but too little discussion and best practice information for the "overall strategy" topic.
Maybe this will inspire others to focus on and discuss the macro level design strategies
Re: Choosing train and track sizes
Thanks Greybeard, thats actually quite clever, yet nicely simple. Will look in to that I think!Greybeard_LXI wrote: ↑Sat Nov 03, 2018 4:51 pm One thing he did was take all iron from south east of his base and all copper from south west. So if he found a great copper patch in the south east he ignored it. This kept the iron and copper ore trains from crossing paths and reducing congestion.
@Bulldog,
My trains are all still <CCCC< style. Just lots of them
164M ore deposit! Holy... I havent gone too far from main area, so the most i've come across is about 50M from memory - but i've got 7-8 deposits of each iron/copper, and I typically put 2 outposts at each ore deposit (less congestion on belts and chests etc, so faster mining).
It still stuns me how much water you have on your map. Filling all that in would drive me nuts.
Agreed on the mods. I'm using Bottleneck and Squeak as well. Also "Auto Research", which is quite handy. YARM is another I find very helpfull, and Long Reach (dont have to be standing next to something to build or deconstruct it - as long as its on screen, you can still build/deconstruct).
Also using Water Fill, which allows you to build water! I got really tired of running pipes everywhere to the nearest lake for my Nuclear Plants, so now I just create water next to boilers, and put the tap there right next to them.
Been spending a bit of time playing with circuits so load balancing in to buffer chests, and from buffer chests, remains balanced. This is so the trains are loaded all balanced as well, and cargo wagons empty at the same time as well. Its been fun figuring that all out.
My train schedules would have to tbe simplest possible ever - and it mostly works. I used to get all complicated with circuits, reading from trains, sending to trains, staging stations on inputs and outputs, etc. None of that now.
Each area (lets use my Red Circuit area 2 as an example), has 3 input tracks. Copper, Plastic, Green Circuits. They are named as "Red2, Copper", "Red2, Plastic", etc.
The trains are all "asssociated" to the destination station. Its schedules would be (eg) "Red2, Green", "Empty cargo" - no time passed conditions or anything else.
Then 2nd station would be simply "Green" - "Full cargo". (there are multiple Green stations).
So, the train (or multiple trains) are dedicated to the input/unload stop. It stays there until its empty, and then goes and gets more from the closest available station. I disable the output stations if they dont have sufficient volume, causing the trains to go to a different option.
Its incredibly simple, yet works fairlywell. There are sometimes scenarios where a train goes to get stuff, and the last station goes inactive, causing the train to return still empty, but its pretty rare.
-
- Smart Inserter
- Posts: 1161
- Joined: Sat Mar 17, 2018 1:20 pm
- Contact:
Re: Choosing train and track sizes
Indeed, keeping it simple seems to work best.
- My in-base trains are all 1-4-1SD (same direction locos).
- Typically doing one blue belt per wagon, via Miniloaders, simple 4x4 balancer, and a couple chests to buffer each belt. Then use 3 lamps per train to give me an instant sense of which materials are running low at a factory. This approach usually leaves a train sitting at the unload station and one at the stacker, with plenty of time for new full trains to arrive, and reduces congestion on the tracks.
- With few exceptions, my intermediate materials factory lines are sized to produce 4 blue belts, so that they load onto trains easily. Again, I normally use a simple 4x4 balancer, buffer chests and 1 miniloader/belt per wagon.
- There are some locations where I use more belts/Miniloaders per wagon to speed up load/unload onto the buffer chests. But I have not needed it much.
- A critical design goal for each factory is to be infinitely expandable. That led me through a weird evolution of my stacker designs and factory layouts. If you look at the science layout, I think I almost got it. I could easily make the stacker 10x longer and tile the stops and lines in the science area 10x. I don't plan to expand science at all, but I wanted that expansion ability for everything.
- Most stations use no circuits at all.
- OUT stations for the same product use a simple circuit to close when occupied (except for the furthest one that never closes).
- The "customs security" between isolated rail networks does use circuits to open the path for authorized trains.
- The transboard station (32 to 4 wagon trains) and new main smelter (N to 32 wagon trains) under construction will use circuitry to manage the flow of trains, to ensure that trains do their job from front to back of the long stopping area (so that they don't get on each other's way). I am probably not making much sense and we might have to wait until I finish it so that I can show pics. I think it is a type of station that was never attempted before, a single large scale station design that is adjustable to trains of different sizes without touching the input/output belts. So, the same smelter could be used with one 32 wagon train, two 16 wagon, four 8 wagon or eight 4 wagon trains. The concept is powerful and I think I can make it work, but it will take some creative circuitry to coordinate the entry/exit dance so that load/unload happens in parallel with the smaller trains.
- My in-base trains are all 1-4-1SD (same direction locos).
- Typically doing one blue belt per wagon, via Miniloaders, simple 4x4 balancer, and a couple chests to buffer each belt. Then use 3 lamps per train to give me an instant sense of which materials are running low at a factory. This approach usually leaves a train sitting at the unload station and one at the stacker, with plenty of time for new full trains to arrive, and reduces congestion on the tracks.
- With few exceptions, my intermediate materials factory lines are sized to produce 4 blue belts, so that they load onto trains easily. Again, I normally use a simple 4x4 balancer, buffer chests and 1 miniloader/belt per wagon.
- There are some locations where I use more belts/Miniloaders per wagon to speed up load/unload onto the buffer chests. But I have not needed it much.
- A critical design goal for each factory is to be infinitely expandable. That led me through a weird evolution of my stacker designs and factory layouts. If you look at the science layout, I think I almost got it. I could easily make the stacker 10x longer and tile the stops and lines in the science area 10x. I don't plan to expand science at all, but I wanted that expansion ability for everything.
- Most stations use no circuits at all.
- OUT stations for the same product use a simple circuit to close when occupied (except for the furthest one that never closes).
- The "customs security" between isolated rail networks does use circuits to open the path for authorized trains.
- The transboard station (32 to 4 wagon trains) and new main smelter (N to 32 wagon trains) under construction will use circuitry to manage the flow of trains, to ensure that trains do their job from front to back of the long stopping area (so that they don't get on each other's way). I am probably not making much sense and we might have to wait until I finish it so that I can show pics. I think it is a type of station that was never attempted before, a single large scale station design that is adjustable to trains of different sizes without touching the input/output belts. So, the same smelter could be used with one 32 wagon train, two 16 wagon, four 8 wagon or eight 4 wagon trains. The concept is powerful and I think I can make it work, but it will take some creative circuitry to coordinate the entry/exit dance so that load/unload happens in parallel with the smaller trains.
Re: Choosing train and track sizes
First of all, thank you all but specially zOldBulldog and evopwr for sharing your ideas and blueprints!
I had been following another thread from zOldBulldog on vanilla train strategies and started playing with the pyramid idea from Malandark (now on its fourth try), and was thinking of sharing some ideas and questions when I found this thread.
A couple of questions: first, how do you fuel your trains? I tried a dedicated sector for fueling, but so much wasted time with trains going for fuel when not really needed (no way to say repeat pick up green, drop off at red, until low on fuel), and the fuel sector became first a bottleneck and then a deadlocked traffic jam. So now I have a fuel train that delivers fuel to every sector. There are what seem to be a crazy number of fuel stops, and I have to manually add new sectors to the fuel train schedule. Is there a better way?
Second, I find that trains are lined up to pick up resources at one sector even though there are lots of resources available at another sector (the shortest path algorithm for trains I guess). Yes I am disabling the station if there aren't enough resources for a full load, but if station A has 10K (two and a half loads) and station B has 125K, the trains still go to station A. So my factories at station A are working full bore, and the factories (or furnaces) at station B are at a standstill. I am wracking my brains over how to do a general better load balancing that will still work when I add station C and D and... without being crazy complicated. Any ideas much appreciated!
Thank you all again, it is so wonderful to have all a place to exchange all these mind-stretching ideas and share the Factorio love!
I had been following another thread from zOldBulldog on vanilla train strategies and started playing with the pyramid idea from Malandark (now on its fourth try), and was thinking of sharing some ideas and questions when I found this thread.
This made me laugh because after many tries and much frustration with bottlenecks, and traffic jams, and "no path", this is exactly the solution I came to for trains inside the pyramid - one train, one purpose, one source, one destination, simple names that show exactly what the train is doing! So simple, and with 120 trains and counting, it just works. And if (when) I need more Green circuits, I just plonk down another Green circuit sector.evopwr wrote: ↑Sun Nov 04, 2018 6:44 am My train schedules would have to tbe simplest possible ever - and it mostly works. I used to get all complicated with circuits, reading from trains, sending to trains, staging stations on inputs and outputs, etc. None of that now.
Each area (lets use my Red Circuit area 2 as an example), has 3 input tracks. Copper, Plastic, Green Circuits. They are named as "Red2, Copper", "Red2, Plastic", etc.
The trains are all "asssociated" to the destination station. Its schedules would be (eg) "Red2, Green", "Empty cargo" - no time passed conditions or anything else.
Then 2nd station would be simply "Green" - "Full cargo". (there are multiple Green stations).
So, the train (or multiple trains) are dedicated to the input/unload stop. It stays there until its empty, and then goes and gets more from the closest available station. I disable the output stations if they dont have sufficient volume, causing the trains to go to a different option.
Its incredibly simple, yet works fairlywell. There are sometimes scenarios where a train goes to get stuff, and the last station goes inactive, causing the train to return still empty, but its pretty rare.
A couple of questions: first, how do you fuel your trains? I tried a dedicated sector for fueling, but so much wasted time with trains going for fuel when not really needed (no way to say repeat pick up green, drop off at red, until low on fuel), and the fuel sector became first a bottleneck and then a deadlocked traffic jam. So now I have a fuel train that delivers fuel to every sector. There are what seem to be a crazy number of fuel stops, and I have to manually add new sectors to the fuel train schedule. Is there a better way?
Second, I find that trains are lined up to pick up resources at one sector even though there are lots of resources available at another sector (the shortest path algorithm for trains I guess). Yes I am disabling the station if there aren't enough resources for a full load, but if station A has 10K (two and a half loads) and station B has 125K, the trains still go to station A. So my factories at station A are working full bore, and the factories (or furnaces) at station B are at a standstill. I am wracking my brains over how to do a general better load balancing that will still work when I add station C and D and... without being crazy complicated. Any ideas much appreciated!
Thank you all again, it is so wonderful to have all a place to exchange all these mind-stretching ideas and share the Factorio love!
My own personal Factorio super-power - running out of power.
Re: Choosing train and track sizes
If there is oil around you can produce fuel locally. Crack the heavy oil to light and then make solid fuel from the light oil and petroluem
-
- Fast Inserter
- Posts: 123
- Joined: Sun Feb 26, 2017 10:48 pm
- Contact:
Re: Choosing train and track sizes
I only refuel at the delivery station. And I plan it for every delivery station.Amarula wrote: ↑Fri Nov 30, 2018 4:01 pm A couple of questions: first, how do you fuel your trains? I tried a dedicated sector for fueling, but so much wasted time with trains going for fuel when not really needed (no way to say repeat pick up green, drop off at red, until low on fuel), and the fuel sector became first a bottleneck and then a deadlocked traffic jam. So now I have a fuel train that delivers fuel to every sector. There are what seem to be a crazy number of fuel stops, and I have to manually add new sectors to the fuel train schedule. Is there a better way?
So if I have a smelter line I would put the input ore (and coal if using coal) station at one end with an output plate station at the other. The input station would have a coal siding and all coal for the ore trains would be loaded into the engines there. You could use belts from the coal train unload to the engines at the ore unload or do what I prefer and cover the stations with a bot network. The bot network does not need to overlap any bots in the production sector unless that is needed for output from an unload station to a production factory.
You could have the coal delivery at the output pickup but then you will have to make sure the ore fields have coal delivery and that would be a big hassle when adding new or removing depleted ore fields.
But do pick one or the other. If I tried to minimize the number of coal stops by picking and choosing which would get fuel at unload and which would get fuel at load I would have a problem. For example:
* Train 1: Pick up at A, Deliver to B, fuel at B
* Train 2: Pickup at C, Deliver to D, fuel at C
Then because A has enough capacity add:
* Train 3: Pickup at A, Deliver to C OOPS.
Maybe part of the solution to too much manual maintenance on sectors is to make bigger sectors?
Re: Choosing train and track sizes
The simplest design I found for this is to disable stations that have a train present. Then instead of trains all waiting at station A they re-path to station B. Problem is with the simple design you got no waiting bay and no train waiting right before the station to be filled when the current train leaves. A long while back I posted an example for disabling stations with a train present and have a waiting bay. In my example the station was designed for 2 trains, one filling up, the other waiting. Any further train gets re-path somewhere else. But you could extend this to 3, 4, 5, 6, ... trains as needed.Amarula wrote: ↑Fri Nov 30, 2018 4:01 pm Second, I find that trains are lined up to pick up resources at one sector even though there are lots of resources available at another sector (the shortest path algorithm for trains I guess). Yes I am disabling the station if there aren't enough resources for a full load, but if station A has 10K (two and a half loads) and station B has 125K, the trains still go to station A. So my factories at station A are working full bore, and the factories (or furnaces) at station B are at a standstill. I am wracking my brains over how to do a general better load balancing that will still work when I add station C and D and... without being crazy complicated. Any ideas much appreciated!
Thank you all again, it is so wonderful to have all a place to exchange all these mind-stretching ideas and share the Factorio love!
The idea then is to enable only as many waiting bays as you can fill. So station A would have 1 waiting bay open so 2 trains can be at the station. After that the waiting bays and stations are turned off and all further trains re-path. Station B on the other hand would have 10+ waiting bays open so trains would re-path to there. Hopefully they get far enough towards station B that they don't turn around back to station A every time a trains leaves station A.
And then there is the LTN mod which solves this problem with lua scripts.
Re: Choosing train and track sizes
Thanks for taking the time to share your experience. I agree, you have to choose pickup or delivery, mix and match is a recipe for trains running out of fuel.Greybeard_LXI wrote: ↑Sun Dec 02, 2018 8:26 pm I only refuel at the delivery station. And I plan it for every delivery station.
My own personal Factorio super-power - running out of power.
Re: Choosing train and track sizes
Thanks mrvn I will give this a try.
My own personal Factorio super-power - running out of power.