Factorio Forums

www.factorio.com
https://forums.factorio.com/

4-way intersections: Throughput and deadlocks [image heavy]

https://forums.factorio.com/viewtopic.php?t=46855

Page 4 of 36

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Tue May 16, 2017 8:23 pm
by torne
aaargha wrote:Hopefully that is understandable and of interest for someone looking to get those last 5-10% extra throughput out of their intersection :)
Oh wow, that's a really good explanation; thanks a lot for posting it. I can see how these rules make sense :)

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Tue May 16, 2017 8:29 pm
by Shokubai
Good write up. One of these days Ill find time to pull some of my actual junctions for testing from my last railworld.

I disagree on principle on the placing of the normal signal you have circled on the lower part of intersection #7. I would rather he wait on the East-West line for the junction to clear. Otherwise he will block West-East (and probably E-W) traffic while waiting for W-E Right turning trains to clear. Normal signals on the outside and in non-crossing positions make more sense.

Would you be willing to share your save file for personal testing?

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Tue May 16, 2017 8:57 pm
by aaargha
Shokubai wrote:I disagree on principle on the placing of the normal signal you have circled on the lower part of intersection #7. I would rather he wait on the East-West line for the junction to clear. Otherwise he will block West-East (and probably E-W) traffic while waiting for W-E Right turning trains to clear.
The chain signals will make sure that no train is able to stop there as long as the output block is properly sized. The train taking that path will reserve all blocks on it from the first chain signal up-to-and-including the output block so barring extreme situations like disabling all stations on the schedule (and signals can't fix that anyway) it should be fine.

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 2:49 am
by AndrewIRL
Admin should change aaargha's username to "Mr Railway Science".

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 6:18 am
by Aeternus
Distelzombie wrote:
Aeternus wrote:Avoid 4 way crosses in favor of T junctions with at least 4 or so train lengths between any intersections that involve the parallel tracks simultaneously - much better throughput that way.

Now I remember hearing that somewhere. You're right. But it doesnt really matter unless you have a megabase.
If you don't have a megabase with a high traffic volume on the rail grid, why worry about throughput on rails at all? If you're running low volume, deadlocks are your only concern - use the simple "one block intersection" designs to ensure none occur in that situation, those are the most simple solution to that requirement (single block roundabout or direct cross).

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 1:49 pm
by Shokubai
Aeternus wrote:
Distelzombie wrote:
Aeternus wrote:Avoid 4 way crosses in favor of T junctions with at least 4 or so train lengths between any intersections that involve the parallel tracks simultaneously - much better throughput that way.

Now I remember hearing that somewhere. You're right. But it doesn't really matter unless you have a megabase.
If you don't have a megabase with a high traffic volume on the rail grid, why worry about throughput on rails at all? If you're running low volume, deadlocks are your only concern - use the simple "one block intersection" designs to ensure none occur in that situation, those are the most simple solution to that requirement (single block roundabout or direct cross).
It's really a matter of taste. For me, I tend to decentralize production around blue science or slightly after. I want a separate smelting facility and i will train in Plate/Steel rather than ore. It creates more trains but fits my own habbits. I do the same with Oil. I want it away from my base and i train in Petro/Lube. I train out acid for uranium...etc...etc...etc... So even if I build all my STUFF at home I still have a good number of trains everywhere. On a true railworld map where i literally build all base and intermediate components in their own camps (I even make a refueling station) Its easy to have hundreds of trains and throughput on the perimeter of my maps is as important as it is in 4x4 highways junctions. At that point, the speed of a junction means less trains total since it takes less time to get A-B.

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 4:51 pm
by leoch
All LHD, and all with huge spaces between the lines!

Would you care to test these? All RHD with no gap between the lines.
crosses.png
crosses.png (3.47 MiB) Viewed 19513 times
Blueprints

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 6:48 pm
by aaargha
leoch wrote:All LHD, and all with huge spaces between the lines!

Would you care to test these? All RHD with no gap between the lines.
Almost all of the ones in the OP are actually available as both LHD and RHD blueprints. I've moved the info a above the pictures to make that clearer.

Anyway, I've added Compact cross and Double diamond, the last one is already tested as Squareabout. These were really interesting in that the unusual design made the pretty hard to reason about with regards to deadlock rating,especially before I fixed all the broken signals :). I initially had double diamond rated as C due to a four train right-to-straight deadlock, it was only after I had added it to the OP that I remembered that only two trains can enter the intersection at a time, so it's actually safe.

If you plan on using any of these design you may want to take a look at the repaired versions in the OP. Anyway, thanks for sharing.

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 7:54 pm
by leoch
Thanks for testing. Yes, I'm not great at signalling. Hopefully this one's better.

It's inspired by the Celtic cross, but unfortunately doesn't have the same aesthetic appeal.
Blueprint
I couldn't think of a good name; if you think of one go ahead and name it. I'm trying to find a design which works well with juxta-positioned tracks.

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 8:12 pm
by leoch
The "wide" layouts are new?

Here's another. It's actually the same as Wide except with the lines closer together.
Blueprint

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 9:40 pm
by aaargha
Added those two as "Crushed Celtic knot" and "Crushed wide" with only minor signal edits. They behave pretty much like their 2-spaced counterparts so I'm guessing they're about as good as it's going to get unless you want to look into reducing input/output dependencies like in Shuriken or Cross.
leoch wrote:The "wide" layouts are new?
Nah, they've been around for quite some time it's just that I hadn't found anything that looked like an original source yet. If you happen to know one for the ones missing it please let me know as I'd like to give credits where it's due.

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Wed May 17, 2017 10:50 pm
by impetus maximus
leoch wrote:Image
the one on the right looks like the signals are on both sides of the lanes. :|

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Thu May 18, 2017 1:20 am
by FrodoOf9Fingers
My solution is rather wide in comparison, mostly to allow inner train turn-abouts, but I think it turns out pretty good (see attached txt document for pattern).
doubleWideBubbleCross.txt
(3.91 KiB) Downloaded 266 times

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Thu May 18, 2017 1:36 am
by iceman_1212
aaargha wrote:The way I and Shokubai have signalled the intersections do not give any lane priority like this. The signalling you suggest, iceman_1212, does give priority to the outer lanes (turning right, or going straight from the east) to avoid this, if you want to, I'd remove the last rail signal on each output and replace the chain signals coming from the crossings with rail signals.
Agree - this is a holdover habit of mine from 0.14 when we didn't have train acceleration on steroids (aka rocket fuel) and I would want to deprioritize left turns. As I mentioned, your signaling is the canonical one. What I wanted to emphasize was that all the back-to-back chain signals with no crossings in between are not really doing much in the way of either throughput or deadlock prevention.
Shokubai wrote: The entire point of using chains exclusively is to prevent a train stopping and blocking the intersection. It's exactly like preventing gridlock in the city. Stop at the red light not in the middle of the intersection. There is no need to add priority UNLESS i wanted , for instance, to make the East to West Left turn have priority at the expense of my West to East traffic. This is generally no desirable reason for me to ever intentionally gridlock a junction.

By making them standard you intentionally allow a train to stop AT that signal...in the middle of your junction. Again, this may be desirable if priority is really what you are after but I've never designed a system where I needed to block one direction to let another train move.

Standard signals should be used if your track segment is long enough to hold an entire train. This junction will allow traffic through but not stopping for my standard 2-4 or 1-4-1 trains.
Yes, this is the point of using chains...but it is possible to make a deadlock free intersection without the chains that I've circled (as shown in aargha's signaling).

It sounds like you're under the impression that the canonical (i.e., "standard") T intersection design (i.e., aargha's) and/or my priority T intersection allow trains to stop in the middle of junctions and gridlock trains. Neither the standard design nor mine will ever gridlock/deadlock because neither allows a train to stop in the middle of an intersection thanks to the chains in the middle (i.e., the ones I haven't circled). I cannot think of a single reason to use back-to-back chain signals on a stretch of track which has no crossings.
aaargha wrote:
Aeternus wrote:but I agree with sentiments posted earlier: Avoid 4 way crosses in favor of T junctions with at least 4 or so train lengths between any intersections that involve the parallel tracks simultaneously - much better throughput that way.
Distelzombie wrote:Now I remember hearing that somewhere. You're right. But it doesnt really matter unless you have a megabase.
Do any of you have a source on this with some actual data on this? It's starting to feel like this is some kind of urban legend, I see it here and there but so far no one has had anything to back it up with. I'd be most interested to see if it actually has some merit.
Afaik, this thread is the first time I'm seeing this level of rigor when it comes to testing train intersections. My observation is that people who are focused on building big bases have it in their best interests to apply the 80/20 rule when it comes to train systems that work, simply because train throughput is so high to begin with. Also, from personal experience, train networks / intersections are probably one of the most annoying things to refactor in an established base, so it's not surprising that there is a tendency for folks to stick to what they know works.

That said, testing that is more "academic" in nature (and less focused on building a base) like what aargha is doing in this thread is exactly what leads to better understanding of factorio systems as a whole. For similar bodies of work in the game, see Davemcw's posts on pipe length and productivity modules as well as XKnight's post on fluid dynamics.

Until this thread, I also believed that 3-ways gave better throughput than 4-ways with no exceptions. But the first post of this thread literally has results showing otherwise...

And on that note, mods, I feel that this thread is worthy of a sticky, especially in context of the madness that we got a sneak peek of from aargha this morning on reddit. I realize it's not a democracy on that front, but that's my vote :lol:

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Thu May 18, 2017 11:12 am
by Distelzombie
in their best interests to apply the 80/20 rule when it comes to train systems that work
I dont know what that means or what this rule is. Could you please explain me? :)

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Thu May 18, 2017 2:52 pm
by Rapier31
I'd like to see some 8 lane designs please. I'm trying to go about it in my game, but I'm having some trouble with the 4 way intersection...

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Thu May 18, 2017 3:35 pm
by torne
Distelzombie wrote:
in their best interests to apply the 80/20 rule when it comes to train systems that work
I dont know what that means or what this rule is. Could you please explain me? :)
It's https://en.wikipedia.org/wiki/Pareto_principle - in this context it's a guess that you can probably get ~80% of the theoretical throughput of your train system for only ~20% as much effort as painstakingly designing every intersection to be perfect, and so people opt to not worry about it too much :)

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Thu May 18, 2017 4:51 pm
by aaargha
FrodoOf9Fingers wrote:My solution is rather wide in comparison, mostly to allow inner train turn-abouts, but I think it turns out pretty good (see attached txt document for pattern).
Rapier31 wrote:I'd like to see some 8 lane designs please. I'm trying to go about it in my game, but I'm having some trouble with the 4 way intersection...
Support for multi-lane intersections is on the way. I'm done with the big stuff, now I just need to get it to adapt to anything from 2 to 8 lane intersections. It's still probably at least a few days out.
iceman_1212 wrote:Afaik, this thread is the first time I'm seeing this level of rigor when it comes to testing train intersections. My observation is that people who are focused on building big bases have it in their best interests to apply the 80/20 rule when it comes to train systems that work, simply because train throughput is so high to begin with. Also, from personal experience, train networks / intersections are probably one of the most annoying things to refactor in an established base, so it's not surprising that there is a tendency for folks to stick to what they know works.
That makes sense. Also, as the 4-way intersections that usually pop up mostly seem roundabout based it's not bad advice most of the time :)
iceman_1212 wrote:That said, testing that is more "academic" in nature (and less focused on building a base) like what aargha is doing in this thread is exactly what leads to better understanding of factorio systems as a whole. For similar bodies of work in the game, see Davemcw's posts on pipe length and productivity modules as well as XKnight's post on fluid dynamics.

Until this thread, I also believed that 3-ways gave better throughput than 4-ways with no exceptions. But the first post of this thread literally has results showing otherwise...

And on that note, mods, I feel that this thread is worthy of a sticky, especially in context of the madness that we got a sneak peek of from aargha this morning on reddit. I realize it's not a democracy on that front, but that's my vote :lol:
Thank you, I'm glad it's been so well received.

It's always a good day when you can welcome someone from pagan faith of "T-intersecionism" into the one true faith of "4-way crossianity" :lol: Or was it the faith of crossinsanity?

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Fri May 19, 2017 7:03 pm
by VFaalcatnodriiro
Hey aaargha, great work you did here...
I'd like to submit my "little" behemoth of an intersection too....

Code: Select all

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
It's made for a 4-lane 2-4-2 (or 2-6) train-network. There is space for a whole train inside each of the branches. However, the signaling at the exits may need a little tweaking to be deadlock-proof.

It'd be awesome if you'd test this one too

Thanks & greetz

Vala

Re: 4-way intersection testing: Throughput and deadlocks

Posted: Fri May 19, 2017 10:45 pm
by impetus maximus
VFaalcatnodriiro's little behemoth.
950px pic
All times are UTC
Page 4 of 36

Powered by phpBB® Forum Software © phpBB Limited