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

Smart setups of railway stations, intelligent routing, solutions to complex train-routing problems.
Please provide - only if it makes sense of course - a blueprint of your creation.
linkined
Manual Inserter
Manual Inserter
Posts: 2
Joined: Mon Mar 09, 2020 12:22 pm
Contact:

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

Post by linkined »

Thanks for your answer.
The intersections in the OP are solid. However there are intersections with higher throughput. I have posted multiple in this forum.
Depending on how long your trains are and how much ressources you plan to put on trains you might not need high throughput intersections.
Most intersections in your base should be basic ones, but areas with high loads need high throughput intersections.
If you want a type of highway you might want to make highway rules. Like all intersections on a highway are high throughput ones and you try to merge as many lanes as reasonable possible before they get to enter the highway.
Regarding 4 lanes.
2 lanes with high throughput intersections has a really high throughput, about 33 per lane with 2-6, while 4 lane with no buffer intersections with many crossing gets about 7 trains per min per lane.
Unless you have really long trains I would recommend 2 lanes with buffered intersections over 4 lanes with unbuffered intersections.
I will focus on buffered two lanes according to what you recommend. My trains are: <C, <C<, <CC<, <CCCC< (with C: car and <: locomotive with direction).

I like to test things myself that's why I tried to use your intersection test bench but I failed to make it works. I found the command panel and try to launch train but they are all stuck "no path".
Are some mods needed and I missed it (in this case sorry) ? Thanks

Edit: Installing Creative mod and Automatic Train Development fix the problem
Edit2: Should I choose EQUAL or PLANNED Right Hand Drive ?

User avatar
hansjoachim
Filter Inserter
Filter Inserter
Posts: 252
Joined: Wed Apr 26, 2017 7:03 pm
Contact:

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

Post by hansjoachim »

I prefer testing with equal and left only (right only for LHD). But if you want to compare your intersections to the OP list you can use planned aswell.

prdfndr
Inserter
Inserter
Posts: 45
Joined: Sat Mar 21, 2020 6:23 pm
Contact:

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

Post by prdfndr »

I am wondering if someone could test my mixed LHD/RHD compact intersection:
mixed intersection.jpg
mixed intersection.jpg (583.02 KiB) Viewed 11155 times
Blueprint:

prdfndr
Inserter
Inserter
Posts: 45
Joined: Sat Mar 21, 2020 6:23 pm
Contact:

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

Post by prdfndr »

And another one LHD/RHD mix:
intersection2.jpg
intersection2.jpg (762.74 KiB) Viewed 11146 times
By calculating possible collisions, this intersection should be a bit slower than "compact"

User avatar
hansjoachim
Filter Inserter
Filter Inserter
Posts: 252
Joined: Wed Apr 26, 2017 7:03 pm
Contact:

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

Post by hansjoachim »

Hei!
I made a Major update to the test bench. To make it easier to use. I would love some feedback on it.

New auto tester
- Runs 3 tests when started.
- Checks automatic for RHD/LHD
- Clears intersection after use
- You can switch between auto and manual mode. Manual mode also got an upgrade.
- Everything is now nuclear
- Now everyone can test with the same results.
- Adaptable to run quicker tests.

See signature to download

prdfndr
Inserter
Inserter
Posts: 45
Joined: Sat Mar 21, 2020 6:23 pm
Contact:

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

Post by prdfndr »

hansjoachim wrote:
Mon Mar 30, 2020 1:51 pm
Hei!
I made a Major update to the test bench. To make it easier to use. I would love some feedback on it.

New auto tester
- Runs 3 tests when started.
- Checks automatic for RHD/LHD
- Clears intersection after use
- You can switch between auto and manual mode. Manual mode also got an upgrade.
- Everything is now nuclear
- Now everyone can test with the same results.
- Adaptable to run quicker tests.

See signature to download
Hey, thank you! I checked my intersections together with the "compact", and got following results (equal train direction distribution):
1) Compact. 54 trains/m
2) Super-compact LHD/RHD mix 49 trains/m
3) Compact LHD/RHD mix 51 trains/m

User avatar
hansjoachim
Filter Inserter
Filter Inserter
Posts: 252
Joined: Wed Apr 26, 2017 7:03 pm
Contact:

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

Post by hansjoachim »

Nice, how did you deal with the mix of LHD and RHD on the test bench?

prdfndr
Inserter
Inserter
Posts: 45
Joined: Sat Mar 21, 2020 6:23 pm
Contact:

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

Post by prdfndr »

Just directed a rail at the start to the right track and renamed the station on the left to "!XL-XTPT North 1". (instead of 4)
And did similar thing on the bottom of the base
tracks.jpg
tracks.jpg (2.68 MiB) Viewed 10931 times

User avatar
hansjoachim
Filter Inserter
Filter Inserter
Posts: 252
Joined: Wed Apr 26, 2017 7:03 pm
Contact:

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

Post by hansjoachim »

prdfndr wrote:
Tue Mar 31, 2020 7:51 pm
Just directed a rail at the start to the right track and renamed the station on the left to "!XL-XTPT North 1". (instead of 4)
And did similar thing on the bottom of the base
tracks.jpg
Interesting.
That gives me an idea of how I can implement support for that without breaking it.

Sopel
Long Handed Inserter
Long Handed Inserter
Posts: 65
Joined: Mon Sep 24, 2018 8:30 pm
Contact:

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

Post by Sopel »

A 2 lane RHD 6-car intersection of size 4x4 chunks.
I was unable to find any smaller with the matching throughput. I think the smallest one that matches this one in throughput is about 5x5 chunks. Last time I tested it was ~87/min. It would be nice to get it verified by someone else and included in the list.
Image
https://imgur.com/a/FLUoJyZ


User avatar
hansjoachim
Filter Inserter
Filter Inserter
Posts: 252
Joined: Wed Apr 26, 2017 7:03 pm
Contact:

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

Post by hansjoachim »

Sopel wrote:
Wed Apr 01, 2020 12:36 pm
A 2 lane RHD 6-car intersection of size 4x4 chunks.
I was unable to find any smaller with the matching throughput. I think the smallest one that matches this one in throughput is about 5x5 chunks. Last time I tested it was ~87/min. It would be nice to get it verified by someone else and included in the list.
Image
https://imgur.com/a/FLUoJyZ
Thanks for the submission. That is a nice intersection. It looks like the multicross just compressed.
I tested it briefly and gets about the same as the multicross. We are working on making a new topic since aaargha isn't updating and I'll add yours. Could you make a blueprint that is easier to stample down? Like about 1/4 of the intersection.
Also a LHD and 3 rails between the the incoming and outgoing lanes would be nice but not necessary.

Sopel
Long Handed Inserter
Long Handed Inserter
Posts: 65
Joined: Mon Sep 24, 2018 8:30 pm
Contact:

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

Post by Sopel »

hansjoachim wrote:
Wed Apr 01, 2020 9:00 pm
It looks like the multicross just compressed. I tested it briefly and gets about the same as the multicross.
Yes! I was playing with multicross and was bugged by how much space is wasted in the corners. This should have the same performance if I didn't screw up. Should have probably mentioned that. Could even name it "compressed multicross" or something.
hansjoachim wrote:
Wed Apr 01, 2020 9:00 pm
We are working on making a new topic since aaargha isn't updating and I'll add yours. Could you make a blueprint that is easier to stample down? Like about 1/4 of the intersection.
Also a LHD and 3 rails between the the incoming and outgoing lanes would be nice but not necessary.
I agree, it would be nice. I'll get to it and have for the new thread.

Factoid247
Manual Inserter
Manual Inserter
Posts: 2
Joined: Fri Apr 03, 2020 8:08 pm
Contact:

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

Post by Factoid247 »

Spent the last couple of days trying some new intersection designs that have a square space in the middle for a military outpost, is minimal in its approach, and has a throughput of more than 75/min.

After testing I was able to get the following throughput numbers.

1 - 80 (15 min run)
2 - 85 (15 min run)
10 - 95 (5 min run)

Would love to hear comments on the design.
Screen Shot 2020-04-03 at 12.24.46 PM.png
Screen Shot 2020-04-03 at 12.24.46 PM.png (1.67 MiB) Viewed 10691 times

User avatar
hansjoachim
Filter Inserter
Filter Inserter
Posts: 252
Joined: Wed Apr 26, 2017 7:03 pm
Contact:

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

Post by hansjoachim »

Factoid247 wrote:
Fri Apr 03, 2020 8:17 pm
Spent the last couple of days trying some new intersection designs that have a square space in the middle for a military outpost, is minimal in its approach, and has a throughput of more than 75/min.

After testing I was able to get the following throughput numbers.

1 - 80 (15 min run)
2 - 85 (15 min run)
10 - 95 (5 min run)

Would love to hear comments on the design.
Looks good!
What version of the testbench did you use 2.12?
Also you are kinda breaking an intersection rule, to allow 1 train length clearance after chainsignals. But sure if you dont mix trains of different length it shouldn't be much of a problem unless you mainline clogs.
You can remove many of those extra signals and still have the same throughput

Factoid247
Manual Inserter
Manual Inserter
Posts: 2
Joined: Fri Apr 03, 2020 8:08 pm
Contact:

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

Post by Factoid247 »

Thanks for the thoughts! Pretty sure it was one of the 2.1x variants of the test bench.

When you say you need at least one train length, do you mean one full 2-4 train or a single car? I checked all my intersections and they have a single train car of clearance.

After watching the intersection for some time I noticed I was close enough to get one more train to buffer for left hand turns and three cars for going straight. here is a BP with that design plus the additional signal spacing (3 cars as I may run 1-2 trains).



hansjoachim wrote:
Sat Apr 04, 2020 10:57 am
Factoid247 wrote:
Fri Apr 03, 2020 8:17 pm
Spent the last couple of days trying some new intersection designs that have a square space in the middle for a military outpost, is minimal in its approach, and has a throughput of more than 75/min.

After testing I was able to get the following throughput numbers.

1 - 80 (15 min run)
2 - 85 (15 min run)
10 - 95 (5 min run)

Would love to hear comments on the design.
Looks good!
What version of the testbench did you use 2.12?
Also you are kinda breaking an intersection rule, to allow 1 train length clearance after chainsignals. But sure if you dont mix trains of different length it shouldn't be much of a problem unless you mainline clogs.
You can remove many of those extra signals and still have the same throughput

User avatar
hansjoachim
Filter Inserter
Filter Inserter
Posts: 252
Joined: Wed Apr 26, 2017 7:03 pm
Contact:

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

Post by hansjoachim »

If the longest trains you are using are 2-4 you need a 6 car clearance.

Bocian
Burner Inserter
Burner Inserter
Posts: 10
Joined: Tue May 12, 2020 6:50 pm
Contact:

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

Post by Bocian »

When working on some rails for a book, this intersection came to life. It's pretty similar to "TRADITIONAL" intersection. It has less space between lines, but losses some throughput on set 2.

After some tests, these are the numbers I got:
Set 1: 62 (15min)
Set 2: 67 (15min)
Intesection.png
Intesection.png (4.39 MiB) Viewed 10033 times

User avatar
hansjoachim
Filter Inserter
Filter Inserter
Posts: 252
Joined: Wed Apr 26, 2017 7:03 pm
Contact:

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

Post by hansjoachim »

Bocian wrote:
Tue May 12, 2020 7:34 pm
When working on some rails for a book, this intersection came to life. It's pretty similar to "TRADITIONAL" intersection. It has less space between lines, but losses some throughput on set 2.

After some tests, these are the numbers I got:
Set 1: 62 (15min)
Set 2: 67 (15min)
Which version of the testbench did you use?
How did you test it without the outer lane being able to turn left? Did you use a lane changer?

Bocian
Burner Inserter
Burner Inserter
Posts: 10
Joined: Tue May 12, 2020 6:50 pm
Contact:

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

Post by Bocian »

hansjoachim wrote:
Fri May 15, 2020 9:48 am
Bocian wrote:
Tue May 12, 2020 7:34 pm
When working on some rails for a book, this intersection came to life. It's pretty similar to "TRADITIONAL" intersection. It has less space between lines, but losses some throughput on set 2.

After some tests, these are the numbers I got:
Set 1: 62 (15min)
Set 2: 67 (15min)
Which version of the testbench did you use?
How did you test it without the outer lane being able to turn left? Did you use a lane changer?
I've used 2.12 version, and I've indeed used a lane changer. The one provided with the save file.

dacendoran
Manual Inserter
Manual Inserter
Posts: 4
Joined: Fri May 15, 2020 3:57 pm
Contact:

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

Post by dacendoran »

You guys should totally start the new topic!

Locked

Return to “Railway Setups”