ReactoCraft Interesting dual fusion setup

[PartyRockVamp]

Hello guys and girls, today I started experimenting with ReactorCraft and I though about something. Then I realised it and it worked. Its just 2 fusion reactors side to side and one feeds the plasma to the other and vice versa.
Here are the screenshots:


Also I was looking at a thread named
Fusion reactor maxed out? Try this.
and so I tried it. The result is amazing

There is so much fusion events occurring that I hear crackling sound
It should melt neutron absorbers instantly but because of the lag it reaches its melting temperature and then melts after a few sec. The bedingot blocks are there to minimize the lag I got from it (by no means reactorcraft is laggy but this setup is)

and somebody in that thread asked if you can make a "torus-less reactor" and the answer is no. It just doesn't fuse even with 6+ rows injectors blasting at each other from 2 sides to join the plasma in the middle.

 

[zemerick]

The fusion reactor is indeed capable of way more than most people think. It's modular, allowing quite a lot of variability.

With some work, you can actually remove the second solenoid magnet, and make the reactor even bigger. You have to put some thought into it to prevent melting absorbers, as you have noticed. Understanding how each piece really works is critical to this.

You also only need 1 preheater btw. Each 1 can feed something like 50 injectors.1

The injectors can be squished in tighter, actually overlapping 1 block.

You can see a few of my various designs: https://imgur.com/a/sBy0j

 

[PartyRockVamp]

The fusion reactor is indeed capable of way more than most people think. It's modular, allowing quite a lot of variability.

With some work, you can actually remove the second solenoid magnet, and make the reactor even bigger. You have to put some thought into it to prevent melting absorbers, as you have noticed. Understanding how each piece really works is critical to this.

You also only need 1 preheater btw. Each 1 can feed something like 50 injectors.1

The injectors can be squished in tighter, actually overlapping 1 block.

You can see a few of my various designs: https://imgur.com/a/sBy0j

Good to know this "You also only need 1 preheater btw. Each 1 can feed something like 50 injectors.1" and this "The injectors can be squished in tighter, actually overlapping 1 block."

I also tried to do it with the other solenoid removed and it didn't bend the plasma and removing 1 magnet makes it not work at all

 

[Erumaaro]

WOW!!
And i thought i was beeing creative by "supercharging" my reactor from the side.

Ahem... would you consider sharing the method/tricks needed to make those modular reactors?

 

[zemerick]

Good to know this "You also only need 1 preheater btw. Each 1 can feed something like 50 injectors.1" and this "The injectors can be squished in tighter, actually overlapping 1 block."

I also tried to do it with the other solenoid removed and it didn't bend the plasma and removing 1 magnet makes it not work at all

WOW!!
And i thought i was beeing creative by "supercharging" my reactor from the side.

Ahem... would you consider sharing the method/tricks needed to make those modular reactors?

Sorry for late responses, I don't come here much lately. It's also been awhile since I've done anything in Reikas mods, so I might have forgotten some specifics/etc.

Each of the internal pieces ( the Injector block, and Toroid blocks ), have a direction built in, that points "forward" in the reactor.
In the case of the injector, if you are on the injection side, it points 2 blocks to the right.
In the case of the toroids, it depends on the rotation. You can find the table used on the github somewhere, but it's essentially 3 possibilities: 2 right; up 1, 2 right; Up 2. Then there are the rotations of this to fill it out.
The injector and Toroid are interchangeable for most all purposes. The catch being that you need a toroid every now and then, else the plasma will become "escaped" and can not fuse. ( This is to prevent reactors without any toroids working just as well as reactors with them. )

The above is very important for the next 2 concepts:
First is the checking for completeness. Basically, there is a check to make sure that the reactor makes a complete loop of some fashion.
This is done with a check that starts at a toroid, and simply goes to the "next" spot forward, following the above rules. This continues up to a certain number ( something like 100, but I don't remember exactly. ) If it ever reaches the first toroid, it succeeds and sets the reactor to be complete. ( Note, there are some specifics that can cause issues with this. If it SHOULD be complete, but isn't, start spinning toroids all the way around. Eventually it will update and work. ) If it fails to ever reach the first one, then it is not complete and the plasma will not be bent.

Second is "hasSolenoid", which is setting whether or not a solenoid is present and powered. IIRC it starts by going 14 blocks East, and 2 blocks North of the solenoid. Then, it follows the previous rules for going around the reactor and setting "hasSolenoid" on each of the toroids. ( If you have the setting in the config turned on to show NBT data when holding TAB, this is a great way to see where any issues are. You'll see hasSolenoid=true for awhile, then change to false at the problem. )

You can see in 1 of my bigger toroids where I used Injector blocks to take the solenoid check over to where the toroids actually are.

For a Toroid to bend plasma, there are 5 factors:
hasSolenoid must be True.
hasNext ( the completeness check ) must be True.
Coolant must be > 0.
Charge must be > some number I don't exactly recall, though it's relatively small.
Finally, the angle can not be too great. ( This prevents making a super small reactor, though I DID manage to make a smaller one, thanks to several Toroid rotations having the same location for the Next piece, meaning I could skip certain rotations. Plus, not having injectors inside the torus, so it was pretty much all toroids turning. )

The final limitation is that plasma will only live for something like 100 blocks ( taxi cab ) from the injector. ( Distance from that is, so half way around the reactor must be within the 100 blocks taxi cab, not total distance traveled around. )

There was also a bit of a limitation before from a bug though. I have no idea if it was ever fixed. It seemed to be something to do with packet limits on sound easily being exceeded by the reactor. Once this happens, it causes increasing server lag that very quickly gets to be extreme. ( Like a minute plus just to flip a lever. )

Now, if you're going to run quite a few injectors, you need to put some thought into your design and cooling. For example, the reason 1 of my reactors looks kind of like a diamond, is you don't want straight lines where lots of fusion can occur. ( Neutrons from fusion, just like from fission, travel straight N/S/E/W ) This would cause extreme hot spots that easily melt.

Also remember that neutrons ignore boilers ( so you can place the boilers inside of the absorbers as well. ) Heat then travels up to I think 3 blocks from an absorber, including up and down. ( Technically, there's no limit as heat is just shared from 1 to another...but the specific temperatures and functions means more than 3 doesn't do anything. ) On top and below an absorber you need to use more absorbers, not boilers, since you have to put water in the bottom, and steam out of the top. Watch out for steam pipe loops, as they can cause performance issues.

That should cover the bases, but if you have more specific questions, just let me know. I'll get to them...eventually

 

[Erumaaro]

Thanks!
Gotta try that out (once i finish my OpenComputers powered ME-Teleporter system...)

I do have a few extra questions/additions:

First is the checking for completeness. Basically, there is a check to make sure that the reactor makes a complete loop of some fashion.
This is done with a check that starts at a toroid, and simply goes to the "next" spot forward, following the above rules. This continues up to a certain number ( something like 100, but I don't remember exactly. )

Do you know how this first toroid is selected? it would make sense if it was the first non-connected toroid if sorted by coordinates.

I saw one of your designs has toroids in a straight line,is there a limit to how many toroids you can have in a chain without injectors?
The 100ish number could be 128 (2^7).

You can see in 1 of my bigger toroids where I used Injector blocks to take the solenoid check over to where the toroids actually are.

hmm..., do you mean the diamond shaped one?
those 2 lines of injector-blocks joining the toroid-circle?
So it stops when it reaches a block of the loop that already has a solenoid, and forgets those it passed,even if they werent in a valid loop.

The final limitation is that plasma will only live for something like 100 blocks ( taxi cab ) from the injector.

Is this updated when the plasma passes through another injector, or is it always reletive to the injector it was created by?
taxi-cab == manhatten-distance?

For example, the reason 1 of my reactors looks kind of like a diamond, is you don't want straight lines where lots of fusion can occur.

If i recall what i read in the ReactorCraft GitHub, plasma entities are counting each other in a cuboid area rather than in a spherical one. This means that plasma travelling straight and diagonally to the x or z axis will have a higher "density" => higher fusion chance than if travelling parralel to the x or z axis. Curved trajectories are better than a straight line, and a curve that goes dioganilly tops them all in terms of "density".(which one is the longest path through a cube: arc vs straight line, arc/line at 0° vs. arc/line at 45°).
This would explain the # pattern of the "regular" design.

Making diagonal "wiggly" lines would probably be best in terms of density vs neutron absorbtion area.

Now, if you're going to run quite a few injectors, you need to put some thought into your design and cooling

I think thats the most important aspect concerning power output.
Ideally all boilers should be below NH3 detonation temperature(650°C) but above 100°C and 50°C above ambient.(according to GitHub)
Heat-transfer is highest when the absorbers are as hot as they can safely get without melting or causing detonations.
Any boiler thats runs too cold will waste heat without working.
My rule for removing a boiler is, if it doesnt boil once every 2 or 3 seconds, it can be removed, unless doing so results in NH3 detontion or a molten absorber...
(when it comes to effisciency, i get temporary ocd...)

I was experimenting with a design, where the absorbers were arranged in a zig-zag pattern(front,back,front,back,etc.) each-one coverd with boilers from all sides. this way there is no heat transfer between absorbers, and heat can be absorbed from all 4 sides.
Since boilers need to be connected top+bottom to pipes, extracting heat vertically canonly be done with empty sodium-heaters as an intermediat(they activly transfer heat and wont need fluid connections if empty...), so there are 3 block high boiler towers directly next to the absorbers..
In the locations where neuton-flux was too high, i used irradiation-chambers to absorb neutrons and make the reactor self suffiscient (a perpetuum-mobile...oh well).



Also, what are those block below the overlapping injectors?

 

[zemerick]

Do you know how this first toroid is selected? it would make sense if it was the first non-connected toroid if sorted by coordinates.

IIRC there are 2 methods checking it, and 1 of them is started from each individual toroid. I can't recall it ever mattering which was the first. ( Especially since it must be a loop. There is no "first" point in a circle after all. )

I saw one of your designs has toroids in a straight line,is there a limit to how many toroids you can have in a chain without injectors?
The 100ish number could be 128 (2^7).

No specific limit, just the life of the plasma/total part limit. The part limit is just a hard coded number Reika chose, not related to binary/etc. I just double checked and it's actually 60.

hmm..., do you mean the diamond shaped one?
those 2 lines of injector-blocks joining the toroid-circle?
So it stops when it reaches a block of the loop that already has a solenoid, and forgets those it passed,even if they werent in a valid loop.

Yes, the diamond with the 2 lines of injector blocks ( only 1 actually works though. I just didn't like the asymmetry and added the second line anyways, lol. )
The solenoid check has nothing to do with loops. It just goes from 1 object to the next, and any time it finds a toroid, it sets "hasSolenoid" to true.

I just double checked this, and there is indeed a little bit there I forgot to mention. It only does this up to 38 parts. However, the toroids can copy between each other to bypass this. Any time you turn a toroid, it will check "behind" it for a toroid, and copy the solenoid state of it. ( If it's an injector, it'll check back from the injector until it finds a toroid. )

Is this updated when the plasma passes through another injector, or is it always reletive to the injector it was created by?
taxi-cab == manhatten-distance?

Always from the origin point of the plasma itself. ( So, effectively the injector that created it. )
Yes, Taxi-cab is where you simply add the distance up and over together, instead of using pythagoras' formula.

If i recall what i read in the ReactorCraft GitHub, plasma entities are counting each other in a cuboid area rather than in a spherical one. This means that plasma travelling straight and diagonally to the x or z axis will have a higher "density" => higher fusion chance than if travelling parralel to the x or z axis. Curved trajectories are better than a straight line, and a curve that goes dioganilly tops them all in terms of "density".(which one is the longest path through a cube: arc vs straight line, arc/line at 0° vs. arc/line at 45°).
This would explain the # pattern of the "regular" design.

Making diagonal "wiggly" lines would probably be best in terms of density vs neutron absorbtion area.

Pretty much yes, but the simpler version is it's the fact that neutrons only travel on 1 axis that matters. So, if you have a line going straight west ( like the normal design does ), you will wind up with a bunch of neutrons going east and west. ( The standard design has straight lines on each side, hence the # pattern of hotspots. )

Not sure what you mean by "wiggly". The best is diagonal lines like I did. You can't turn the toroids back and forth to "wiggle", they can only curve inwards. ( As far as I could get to work anyways. )

I think thats the most important aspect concerning power output.
Ideally all boilers should be below NH3 detonation temperature(650°C) but above 100°C and 50°C above ambient.(according to GitHub)
Heat-transfer is highest when the absorbers are as hot as they can safely get without melting or causing detonations.
Any boiler thats runs too cold will waste heat without working.
My rule for removing a boiler is, if it doesnt boil once every 2 or 3 seconds, it can be removed, unless doing so results in NH3 detontion or a molten absorber...
(when it comes to effisciency, i get temporary ocd...)

I was experimenting with a design, where the absorbers were arranged in a zig-zag pattern(front,back,front,back,etc.) each-one coverd with boilers from all sides. this way there is no heat transfer between absorbers, and heat can be absorbed from all 4 sides.
Since boilers need to be connected top+bottom to pipes, extracting heat vertically canonly be done with empty sodium-heaters as an intermediat(they activly transfer heat and wont need fluid connections if empty...), so there are 3 block high boiler towers directly next to the absorbers..
In the locations where neuton-flux was too high, i used irradiation-chambers to absorb neutrons and make the reactor self suffiscient (a perpetuum-mobile...oh well).

Don't bother trying to do ammonia. The random and high energy nature of fusion neutrons can exceed the detonation temperature before the heat transfer can even occur. So, you just need to avoid melting absorbers. ( Ammonia can work fine for like 10 minutes...then POP. Don't trust it. I tried it a few times, but found I had to turn the power down so much/limit neutrons that water would have been better. )
If a boiler does not touch the air, it will lose virtually 0 temperature, so no need to worry about efficiency. Pile on the boilers
Boilers can pass water and steam through each other, so you can stack them as high was you want.
When using only 1 heater, a single NIC in a very "cool" spot of your reactor will make it self sufficient btw. The reactor really uses very very little fuel. It seems like a lot when you first run it just because it has to fill all of the different tanks.
Zig-Zag/checkerboarded absorbers is definitely useful for hot areas. It makes a world of difference on temperatures.

Also, what are those block below the overlapping injectors?

Just wireless receivers to make turning sections of injectors on and off for testing easy.

 

[Erumaaro]

Pretty much yes, but the simpler version is it's the fact that neutrons only travel on 1 axis that matters. So, if you have a line going straight west ( like the normal design does ), you will wind up with a bunch of neutrons going east and west.

actually, neutrons are always emitted in all 4 horizontal directions.
each fusion creates 3 neutrons travelling in a random direction.

The standard design has straight lines on each side, hence the # pattern of hotspots.

the hot spots are not inline with the injectors but rather with the diagonal sections, though?

If you inject just enought plasma to have fusion at all, the neutrons will only appear at the diagonals (and travel parrallel to the x or z axis afterwards, of-course).

a fusion occurs if the amount of plasma entities in a cube (centered on them with edges 2,5 blocks long) is >= 15-20(random).
(I double checked the code!)

Don't bother trying to do ammonia.

I ran a "regular" fusion reactor with 12 injectors (4x2 side injected) with 29 absorbers on each edge in a line(no neutron radiation free blocks in between) covered with 1-2 ammonia filled boilers on either side, all running below 300°C. There was a very hot spot inline with each injector, so i used NICs there to reduce the neutron intensity before the stream hit the absorber.
It would run 32 HighPressureTurbines @12,78GW each, and still building up steam.
luckily i put an ifinite reservoir, otherwise the building up steam would deplete the reservoirs and cause a detonation.

I just piled on boilers(1block high) and then reduced the amout to keep all of them at +96°C.
Runs steadilly for hours.

I used pipes from immersive-Engineering to feed in the ammonia, since they "teleport" the liquid to any reservoir connected (unlike rotary pipes, which rely on pressure deifferential).

but it was laggy as hell! I had to use tickratechanger to slow things down to 5tps, or plasma would constantly get bend in weirs directions and create geometric patterns outside the toroids...

Boilers can pass water and steam through each other, so you can stack them as high was you want.

well, its the only way to make stacked fission reactors.
But since fluid transfer is proportional to the difference of the amounts each boiler contains, very high stacks will cause a bottle-neck (learned that the hard way with a breeder-reactor...).