[MOD 0.17+] Adamo Mods

[myricaulus]

I searched the forum for a bit and found that:

viewtopic.php?f=51&t=45001&p=279641&hilit=pipe#p279641

Of course i dont know if its still applicable.

Also Bob calls its Unstable Simulation:
search.php?keywords=unstable+simulation ... mit=Search

I found another case of unstable simulation: to fast liquids
viewtopic.php?f=25&t=63801&p=388975&hil ... on#p388975

[mmmPI]

I can't recreate the sloshing on my demo world. Tell me exactly the conditions you can create it, if you would. You mentioned it may be due to differently-sized pipes, according to something Bob had to fix. I'm not able to recreate the right conditions even with differently-sized pipes, so I'm going to play around with this more. I know I have a map where it's happening, so I'm going to pull that up next and recreate the conditions exactly. If it does come down to pipe sizes, I will definitely change them to all have the same size. They'll all have to be increased to 200 since 200 is needed for compatibility with vanilla megareactors under my physics mod.

I had experienced this behavior when fluids are updated every tick from 59.9 to 60 with a setup as the next picture, ( happened for steam and not for petroleum ) after thinking about it and reading the thread i am unsure it is due only to pipe size, since i was not using any different than vanilla. I had the hypothesis that it is due to the "loop" the fluid can do or a "back and forth", it is easy to make "8" or "0" shape with pipes with the reaction kettle without noticing. It seems to have happened to me more when i place reaction kettle directly adjacent to each other.

20190821231724_1.jpg (527.28 KiB) Viewed 7345 times

I ended up changing my setup to have the kettle not use their "liquid pass through" property and it seemed better ( and also to stop making sulfur out of petroleum gas )

20190830164717_1.jpg (742.34 KiB) Viewed 7345 times

(there are 2 of them at one end of the heat pipes )

[myricaulus]

I ended up changing my setup to have the kettle not use their "liquid pass through" property and it seemed better ( and also to stop making sulfur out of petroleum gas )

Oh the spaghetti is strong in you . Anywas i did the same for the same reason. I read somewhere that pass-through is like a pump to enable entity to entity flow. Maybe each "pump" pumps from one to each other, back and forth?

I experimented a bit around with the light oil to solid fuel recipe. This one is really bad sloshing as soon as it touches with any connector an non full pipe, only the check valve is able to "leave" it within if pointing towards it. This kettle seem to actively try to push its contents out, like a pump, while letting flow back from the same direction meanwhile.

[Adamo]

What I want to know is what setup you guys think is best for the kettle. We know what our options are, I think. What do you think is the most useful choice? Should it not share back onto a pipe? Should they not share between eachother? Is as it is now -- essentially, a very greedy pipe -- the best balance? I could also make them directional, so you could bring in fluids from either end but only one of the inputs shared back out. I thought about this, but I opted for the route we currently see because it provided the most design flexibility for the player.

[myricaulus]

Regarding the infos i found in viewtopic.php?f=25&t=63801&p=388975&hil ... on#p388975 about how pipe flow work, i would like to have it behave mostly like the vanilla boiler. That would be optimal. Currently because of this weird heat pass through mechanics the options for recipes utilizing the pass-through are very limited to 1 liquid in 1 liquid out recipes. Maybe a slow 2 liquid and one solid with long inserters could be possible to construct with vanilla equipment.

In order of my personal interest i would prioritize the functionality like this in descending order:
Top: Heat pass-through working seamlessly like the original vanilla heat pipe. That way i could omit the heat pipe increasing the possible recipes as well as drastically reducing the needed energy to heat up the system. My experiments showed, that the heat pipe "eat" most of the energy.
2: Liquid pass-through working like the vanilla boiler
3: Liquid pass-through working only between each kettle would be sufficient, unidirectional would also be ok i think
4: No liquid pass-through would be still an ok option for me, this would just not increase the possible recipes to the current amount and also increase the space and complexity needed to build it.

Currently i build them like they had no pass-through option for heat or fluids anyways and its okay. Actually its a fun logistical spaghetti minigame
How about a second type of kettle with no pass-through capabilities?

I dont understand why the kettle behaves differently than the gas furnaces which pass-through capabilities are basically like that of a boiler. Maybe it feels like they produce more flow resistance than the vanilla boiler between each other since they use up so little of fluid, but its okay. When i have a row of 24 gas furnaces and supply it with a full iron pipe of 200 gas out of a full tank, the last boiler takes several minutes to fill half up and never fills completely.

Also i am sure that the iron pipe are quite bad in regards of throughput: i connected 6 chemical plants for sulfuric acid, which would take up 600 Water/s after maybe 5 underground pipe section and all of the were connected together in a row by normal pipes and the last chemical plant didnt get any water until i installed two pumps. I can send you a screenshot if needed. AFAIK that could be caused by the increased fluid box size, i will test it later with copper pipes.

[mmmPI]

I don't know what would be the best way to use them.

Space efficency for the 1x1 , but requires quite some tinkering to find a way to reach the connections, heat pipes can't go undergound, outputing a solid without configurable inserter while needing 2 different liquid is a puzzle. ( there are tools to ease that in the modpack )

They can only host 1 module, and being so little you don't gain much by fully surrounding them , and it makes it even more of a puzzle.

Those i took as a hint that they were good early game when you don't have large scale building capabilities. It hints at using heat pipes early game too, which is even more true with the heat furnaces.

Now i realise they are convenient for dealing with fluid only, the greedy pipe option.

I think the balance is interesting as is, i was less confortable using the "fluid permutation" hotkey when making this spaghetti, i found it later in the controls settings, i think i haven't weighted all the options accordingly yet.

If there was some trade-off in flexibility to remove the sloshing , that could be balanced by adding another module slot from my point of view.

Or maybe make some receipe have better yields/ better receipe , depending on temperature so they are an option for specific case late game.

i would like to have it behave mostly like the vanilla boiler.

That's what i felt they were designed for.

The 1x1 size is appealing, i found it original and funny, and i learned the word kettle too .

[Adamo]

Regarding the infos i found in viewtopic.php?f=25&t=63801&p=388975&hil ... on#p388975 about how pipe flow work, i would like to have it behave mostly like the vanilla boiler.

Sorry, I guess I didn't explain well enough. It is impossible currently to have the kettle behave exactly like the boiler. That would definitely be ideal. I could get CLOSE to boiler behaviour by setting the kettle to be on the same level as the pipes, which would mean, essentially, that they fill up at the same rate as the pipes, rather than "grabbing all the available fluid" like chemical plants do.

The gas furnace's fuel should behave exactly like boiler water does since they both have the same type of connection: "input-output". This type of connection is not available on the assembling-machine prototype, on which chemical plants and the reaction kettle are based. The gas furnace's ingredient connection (which is not currently being used by any recipes I've introduced) should be expected to behave like the kettle's fluid boxes.

Yes, please let me know your test results from using the copper pipe. I will take a look as well. Thanks.

[Adamo]

outputing a solid without configurable inserter while needing 2 different liquid is a puzzle. ( there are tools to ease that in the modpack )

A puzzle for sure, but it is doable!

If there was some trade-off in flexibility to remove the sloshing , that could be balanced by adding another module slot from my point of view.

Or maybe make some receipe have better yields/ better receipe , depending on temperature so they are an option for specific case late game.

Hmm, OK. Adamo Chemical does introduce some better ratios on some recipes and gives them to the reaction kettle (and the process reactor). It might be subtle enough that people haven't noticed.

i would like to have it behave mostly like the vanilla boiler.

That's what i felt they were designed for.

The 1x1 size is appealing, i found it original and funny, and i learned the word kettle too .

This is what they were designed for, originally, until I figured out it was impossible. Of all of the imperfect options, what we have now is the best compromise I have been able to find. The only issue it really leaves us with, as far as I can tell, is the annoying sloshing.

[myricaulus]

That would definitely be ideal. I could get CLOSE to boiler behaviour by setting the kettle to be on the same level as the pipes, which would mean, essentially, that they fill up at the same rate as the pipes, rather than "grabbing all the available fluid" like chemical plants do.

That actually sounds ok for me, if it behave like a normal pipe. That way its not trying to get rid of its contents while at the same being totally greedy getting it from neighboring tiles. That would mean the "integrated-pipe"/tank would have a minimum supply level/pressure to start working, which is actually the same mechanic as with the heat. One could simply pump it into the kettle. And to be clear, i feel for its size its even adding to realism needing to do this. You cannot expect of such a small thing to have all that fancy "convinence" stuff like pumps and such. Its a simple chemical reactor, just a simple component, without auxilary equipment it couldnt work.

The pipe throughput issue:
I tested with copper pipes and everything worked like i expected, every plant got enough water even without pumps, i replaced it with iron pipes again just to confirm and the last one did not get enough water again. This seem to confirm the different post i read regarding the unintuitive fluid box size if pipes.

20191002152325_1.jpg (1.4 MiB) Viewed 7283 times

[Adamo]

That would definitely be ideal. I could get CLOSE to boiler behaviour by setting the kettle to be on the same level as the pipes, which would mean, essentially, that they fill up at the same rate as the pipes, rather than "grabbing all the available fluid" like chemical plants do.

That actually sounds ok for me, if it behave like a normal pipe. That way its not trying to get rid of its contents while at the same being totally greedy getting it from neighboring tiles. That would mean the "integrated-pipe"/tank would have a minimum supply level/pressure to start working, which is actually the same mechanic as with the heat. One could simply pump it into the kettle. And to be clear, i feel for its size its even adding to realism needing to do this. You cannot expect of such a small thing to have all that fancy "convinence" stuff like pumps and such. Its a simple chemical reactor, just a simple component, without auxilary equipment it couldnt work.

I was already considering doing this, and I think you've convinced me. Expect this in the next release.

The pipe throughput issue:
I tested with copper pipes and everything worked like i expected, every plant got enough water even without pumps, i replaced it with iron pipes again just to confirm and the last one did not get enough water again. This seem to confirm the different post i read regarding the unintuitive fluid box size if pipes.
20191002152325_1.jpg

Physics mod is changing the iron pipe size. The problem is that if I don't double the iron pipe size, vanilla megareactor blueprints will no longer function optimally because the system can't pass through enough steam. So I feel my hands are a bit tied on this. At first, I only provided the additional pipes at larger sizes (the stone pipe, for example). But I feel very strongly that I want my mods to be compatible with vanilla ratios and blueprints except in the places I clearly noted at the beginning of my original post, and I had to enlarge the iron pipe to maintain that.

To be honest, I'm sitting on my hands waiting to see what is going to happen with the game's fluid physics. There have been many allusions to a new fluid physics, which is now a "maybe" on the development map. I've heard there are some problems with the new simulation (I am a graduate physicist who's done work in computational hydrodynamics -- maybe I can help, guys! I know how to handle resonances.), and I don't know where they're going to go with it. It seems more and more likely that the fluid mechanics will stay as they are, which is not the worst decision. In that case, I think all I could do is, first off, check to see if doubling the iron pipe size was overkill to support megareactors, and then reduce that as much as possible, then consider any other tricks to maximize fuel sharing. Or I can decide to not support megareactors under physics mod except if the user is willing to use the larger pipes from my or someone else's pipes mod. This has the advantage of then leaving the iron pipe to function as it does in vanilla, although this would only be noticed by people (such as yourself) who look closely at the pipe mechanics. Then again, that argument holds for megareactors, too. There doesn't seem to be a perfect choice on this, and to be honest, I am torn.

[myricaulus]

and I had to enlarge the iron pipe to maintain that.

If your problem is throughput related, increasing the fluid box of the pipe could make the reactors with pipes suffer even more, since the throughput is lower with larger fluid-boxes as i showed you and other threads hinted me. At least on some distance, i dont know how it behave on very short distances. I also dont know exactly what you mean by megareactors, because i followed on reddit this trend with UPS friendly reactors reducing the amount of fluid/heat pipes as much as possible. I even saw one nearly without pipes, utilizing offline reactors by itself as a big heat pipe .

Anyways since some time u can easily change your blueprint with an upgrade planner, changing the pipes within some seconds easily on vanilla. I would even suggest to leave the iron pipe as it was in vanilla, because muscle memory dont get changed that easily

When i finally have some uranium cells i will test my old 24 Reactor 3,84GW squared-self defending-reactor design, then we will see Would you call this a mega-reactor? It has a lot of pipes and tanks, and is way inefficient in regards of UPS, but it worked for my last 3.1415 RPM mega-base. I actually could load up a sandbox and try it out. Could you tell me the code to change the fluid-box of the stone pipe to be set to 90 and plastic to 80 for data-final-fixed.lua? I would make my own pipes mod and try around a bit.

[myricaulus]

Ok i started a sandbox szenario and got this error :
Error while running event adamo-contraptions::on_player_joined_game (ID 43)
__adamo-contraptions__/control.lua:40: attempt to index local 'armorinv' (a nil value)
stack traceback:
__adamo-contraptions__/control.lua:40: in function 'reset_equipment_bonuses'
__adamo-contraptions__/control.lua:148: in function <__adamo-contraptions__/control.lua:144>

[Adamo]

You need to change the values, e.g.,

Code: Select all

data.raw.pipe["plastic-pipe"].fluid_box.base_area = 0.8
data.raw["pipe-to-ground"]["plastic-pipe-to-ground"].fluid_box.base_area =0.8

I have also seen a similar reactor blueprint to what you say. There's not much I can do for those people (edit: incorrect: after further consideration, I have increased the potential throughput of pipes even in these cases). If they're maintaining the pressure on the pipe, a double-sized pipe will have twice the throughput, so it just depends. This is what I would consider a minimum-sized megareactor: https://factorioprints.com/view/-LkDCyabpKplUMtkAAzI Since it maximizes the use of the water input pipes, this is the reactor I use for my testing. I can confirm I needed to increase the pipe sizes for the steam to flow through the entire network when using physics mod.

[Adamo]

Ok i started a sandbox szenario and got this error :
Error while running event adamo-contraptions::on_player_joined_game (ID 43)
__adamo-contraptions__/control.lua:40: attempt to index local 'armorinv' (a nil value)
stack traceback:
__adamo-contraptions__/control.lua:40: in function 'reset_equipment_bonuses'
__adamo-contraptions__/control.lua:148: in function <__adamo-contraptions__/control.lua:144>

OK, I released an update on the mod portal. That should fix it. It just needed to be aware that sometimes a player might join a game but not have a valid inventory.

[myricaulus]

Ok, i am sorry to inform you, that you have fallen into same unintuitivity trap about fluid dynamics in factorio as most of us were: Bigger fluid-box sized means less throughput. I just tested a Reactor of mine with 2.2 GW theoretical continuous power (in Vanilla), which was pretty good working back then. I started it up ran it for about 3 minutes with copper without load to fill the tanks. Then ran it for 15 minutes with max load. It floored just below 2.0 GW. I changed everything to iron pipes it immediately dropped down to 1.5 GW and floored at 1.3 GW. Next i will try the 90 Pipes. But i have the feeling, that heat changer is the bottleneck. On Full load the temperatures on the heat pipes at the end were around 505. Now they are 560°C.

20191002185109_1.jpg (1.15 MiB) Viewed 7249 times

[myricaulus]

Ok, i did test this reactor without your mods, so basically vanilla values, and it turned out, that the design isnt that efficient: it floored after 10 minutes around 1.9 GW. The main issue is, that the heat pipe are too cold when they reach the last heat exchanger, so it isnt working. So with your mod installed and using copper pipes makes it actually batter , i might even add 1 or 2 heat exchanger on the end and reach the theoretical power =).

Update: I added 2 heat exchanger on the end to see what power i can achieve and I tried the Pipes with 0.9 and 0.8 with strange results. I had no increase in efficiency by this change, but the power values where fluctuating way faster, between 2020 MW and 1990 MW within seconds, while with copper the values were are very stable only changing some kilowatts and approaching 2010 MW

[Adamo]

Hmm, OK, I'll take a look at it. Thanks. I made this blueprint on version 0.16, so it's possible things have changed slightly. I know it worked fine enough when I made it. We also use several of them on our server network, sometimes at max capacity, and they work fine. I see in your picture you had some extra heat pipes along your rows of heat exchangers. Under vanilla physics, I'm not sure that will work: I probably used the maximum length of heat pipes we can get away with at maximum output. So if it's off by a tiny bit of temperature at the end, that might be why. I think this should be less problematic under physics mod as the heat pipes can transfer heat at a higher rate. I'll set a new test world to see what I can come up with, and based on that I might start changing some values. Feel free to send me any more findings of yours, they can be useful.

Regarding the throughput, I don't think I made myself clear: larger pipes result in less throughput in many cases, I agree with you there. The case I mentioned is only if the player can maintain full pressure on the pipe. IF they can maintain full pressure on a larger pipe, THEN the larger pipe will have more throughput. I know this is a true statement because it is a critical requirement to use my reactor blueprint under physics mod, but it is a very difficult condition to maintain, so for most practical purposes for most people, it does not appear to work this way. For practical purposes, it works the way you have described, which is that the throughput goes down due to the difficulty of filling the larger pipe. Edit: I've just realized another thing I can try, so I'm going to test that.

Update:

I tested the reactor. Here is a picture of the reactor, running at full steam, under physics mod.

full_on_vanilla.png (3.79 MiB) Viewed 7219 times

So we can see that under the physics mods, at least, the reactor works as expected. The 4 steam engines that don't quite receive enough steam are expected: there is an overbalance of steam engines to work back down the steam storage when needed, and to make the design symmetric. If anyone is having trouble getting the reactor to work at full capacity under physics mod, then there's some sort of input issue. Happy to help if I can. Testing without physics mod, next.

Here is a picture of the reactor running with full steam under vanilla physics, also working as expected (note approximately 4 steam engines not at full steam, again expected).

full_capacity.png (3.87 MiB) Viewed 7219 times

So, if your reactor is having trouble keeping up, I think there's an input issue. Happy to help you troubleshoot it. Whether using physics mod or vanilla physics, getting enough water into this reactor is a challenge. It fully-utilizes every input line of water, so the water pipes must be kept at very high pressure, or the reactor will not run optimally. That's typically the first issue people have.


Update:

With my new pipe idea, I was able to improve the throughput of the reactor beyond the throughput under vanilla settings! I have it down to fewer than two steam turbines inoperable in each block. The tradeoff right now is that a direct pump connection is needed to feed water into the heat exchanger. I'm going to overhaul all of my pipes with this new method once I make a few decisions.


Update:

Yes, I've been able to further improve the throughput of mega-reactors by changing the way I'm altering plumbing. My new method should increase rather than decrease potential throughput in the other cases we've discussed, and I've already confirmed this is the case through the mega-reactor steam network. Let me know whether your results agree. I'll be updating the physics mod shortly, along with a few others.

The kettle is another story. I've tried almost every configuration I can, and if I want to allow pass-through between the inputs, it will cause sloshing. I've set the area, level, and height values of the kettle to default, which are the same values as the vanilla pipe. I think what I'll do is leave these settings as default settings forever, and then we're just dealing directly with the simulation the devs have setup as-is. Under these settings, the kettle will still work to pass fluids across, and they will force oscillation into any system they're connected to. This can be controlled by using pumps or one of the valves from the pipe control mods (I use Advanced Underground Piping) to isolate part of the system. I'll update this mod shortly.

[myricaulus]

So, if your reactor is having trouble keeping up, I think there's an input issue. Happy to help you troubleshoot it. Whether using physics mod or vanilla physics, getting enough water into this reactor is a challenge. It fully-utilizes every input line of water, so the water pipes must be kept at very high pressure, or the reactor will not run optimally. That's typically the first issue people have.

Now i am totally confused seeing your reactor. The normal offshore pump is producing according to tooltip 1200/s. You have 20 heat exchanger per pipe input.. makes 2400/s usage, this much is way more difficult to supply over one pipe. AFAIK the possible maximum fluid flow is from tank->pump->tank with 12000/s. I use this setup for my rail ways, yes it fills a tank waggon in 2 secs =).
In my reactor I connected one pump to 10 heat exchanger, then its easy to input the water even a bit longer distances. You can maintain this speed with one pump every 8 u-pipe sections. I have some pumps here and there. The input is not the limiting factor in my case, i can clearly see that even the last exchanger are over 2/3 full of water. In my reactor its most likely a heat pipe issue, since most the heat is generated in the middle of that reactor block with 400% neighboring bonus. Or its a steam issue, with all this maze from the turbines. But even with your better physics for heat pipes the distances are too great. After all i think its just impossible to construct this reactor with 100% efficiency with the current fluid dynamics. But still the neighboring bonus of its maximum configuration level would be 7200%. so even if its loosing a total of 10% due to fluid dynamics its still pretty damn fuel efficient =) Unfortunately it needs bobs inserter mod or similar to work.

Source about vanilla pipe throughput over distance:

If 1000 water/s is acceptable, you need a pump every 200 pipes. There is a very harsh penalty for going over 1000 fluid/s.

If you want to run your offshore pump at full capacity (1200 water/s) you need a pump every 16 pipes.

If you want to run two offshore pumps at full capacity (2400 water/s) you need a pump between every pipe.

viewtopic.php?t=57127#p338283
With bigger pipes it would be different though.

The kettle is another story. I've tried almost every configuration I can, and if I want to allow pass-through between the inputs, it will cause sloshing. I've set the area, level, and height values of the kettle to default, which are the same values as the vanilla pipe. I think what I'll do is leave these settings as default settings forever, and then we're just dealing directly with the simulation the devs have setup as-is. Under these settings, the kettle will still work to pass fluids across, and they will force oscillation into any system they're connected to. This can be controlled by using pumps or one of the valves from the pipe control mods (I use Advanced Underground Piping) to isolate part of the system. I'll update this mod shortly.

Thats unfortunate, i was hoping for a simple pipe like entity. But thats not how an assembler is supposed to work. Maybe if we cry loud enough, we will get a input-output option for assemblers?

[myricaulus]

Yes, I've been able to further improve the throughput of mega-reactors by changing the way I'm altering plumbing. My new method should increase rather than decrease potential throughput in the other cases we've discussed, and I've already confirmed this is the case through the mega-reactor steam network. Let me know whether your results agree. I'll be updating the physics mod shortly, along with a few others.

I am quite impressed by your updates. I was able to supply 55 exchanger in a single row with a single pump. Before the update i was able to supply max 21 Heaters. The main issue is still the internal resistance of the heater, but u seem to reduce it quite a bit. Theoretically i could provide 24000 water/s(the max capacity of a pump).

20191003214401_1.jpg (1.03 MiB) Viewed 7163 times

You see the blue line down on the map, thats all heat exchangers+reactors

placing one u-section of iron (200) in between reduced it to only 46.
placing two tanks in between reduced it to 44
placing one u-section of copper (100) in between reduced it to only 29.
placing one u-section of my plastic (80) in between reduced it to only 28.
So i am indeed convinced, that these bigger pipes do increase the throughput in special situations, for very high intensive and short connections.

Update:
I was quite shocked... as i saw my reactor working at peak efficiency 200 heat exchangers with iron pipes connected produced 2GW of power, like they actually should. I would like to know the magic you did there. Have you altert the water viscosity?

[BlueTemplar]

That's just the expected change from increasing pipe size ?