Atomic Science Fuel Life Config?

[Digdug83]

I don't know if liquid tesseracts have a limit, although they might. I can't think of a reliable way to test that as liquiducts are limited to 100 mB/t and buildcraft pipes are an unknown quantity. I'll see if I can get info from the authors on the subject.

Basically though you create a giant ring of liquiducts (outputs from the funnels won't input into tesseracts, only other liquiduct) and then space out your tesseracts so they can each draw from an output source and use most of their faces (probably only 4 or 5). This way the steam moves as rapidly as possible to your turbines/storage area with (hopefully) little loss. I have two of them eating steam from the inner ring of eight magnets and something like 20 eating steam from the outer ring.

Edit: According to some FTB threads, liquid tesseracts are limited output wise only by what you can input into them. That being said, the only information I can find on BC pipe capacity is 1.6 buckets per second. This translates (assuming they meant second and not tick) to 80 mB/t. That means you're still going to end up using liquiducts as they have a maximum flow rate of 100 mB/t. It should also be pointed out that liquiducts have been repeatedly stated as being less laggy than BC pipes as well.

[jakalth]

I have had the output of a fully powered reactor setup transfered to an array for 25 large turbines, buy a SINGLE line of liquaduct. it was enough to keep all the turbines maxed out for the test run. It seems the limit of a liquaduct is more the input/output rate then the flow rate through the tube it's self. That being said, I always try to have redundancy. Even with the reactor direct piped to the turbine array, I try for at least 4 feed lines just in case. The test reactor is not one of my bigger ones, but a smaller one I use for short duration testing of ideas. It's relatively low output.

Daemon_Eleuel: Your design looks right, but is a little bigger then what the plasma can flood out. Shrinking it's total radius by 1 block would be minimum for catching all the steam available. 2 blocks if you want to save on some resources. Your's currently has a flood radius of 8, 7 is the max plasma will spread out, 6 is a bit more efficient material wise. But yeah, that design should be a large output reactor.

[jakalth]

Ok got some pictures of what my server reactor looks like. There is a lot of stuff crammed into a small area so it is kind of hard to see everything. I'll start from the bottom. With an underside view of the reactor it's self.

Underbelly of the core.

Next we have the space between the core and the turbines. The area where all the steam is collected and buffered.

Steam chamber.

And lastly, the top of the reactor. Turbines a plenty.

45 Turbine array.

The whole reactor uses 45 large turbines, 244 electromagnets, over 125 steam funnels, 32 liquid managers, countless stacks of liquiducts and redstone energy conduits, 40 REC's for temp storage, 8 energy tesseracts, and a single fusion reactor fed deuterium cells from a computercraft computer hooked up to an autarchic gate. Total output varies a bit since plasma generation is being fiddley...

:EDIT: Ok, finally got some tests under the reactors belt. Power output is a lot more random then my 4 torus design, but also much higher. It all depends on how well the plasma generation decides to cooperate. After several tests and a little tweaking of liquaducts and conduits this is the results:

When you don't get the best plasma generation, the reactor can output steam for about 8 minutes. After the full 8 minutes have passed, it ends up generating roughly 9.5 million MJ of power. This is measured by starting with 40 empty rec's and checking how full they all were when the turbines stopped spinning. Every one had about 237500Mj.

When you Do get good plasma generation, the reactor fills the liquaducts and buffer, and runs for about 10 minutes. After the full 10 minutes have passed, it ends up generating roughly 14 million Mj of power. This is measured the same way as before. Every rec had about 351000Mj.

Each test was done by inserting 2 deuterium cells into the reactor after it had cooled off. And each time the measurements were done when the last turbine completely stopped spinning.

[Digdug83]

More tests, yay!

It appears that steam isn't lost at all and just sort of sits in the funnel waiting to be used/extracted. I wanted to know how much steam a diamond flood reactor with magnets on the top layer as well as the sides would produce. I constructed said reactor (taking jakalth's cue I attached an REC with a limited output of 50 to the core) and attached 4 tesseracts with 4 different frequencies (5 faces inputting each) to a series of 16 liquid managers and a gigantic pipe network. One cell filled this setup in almost no time flat (<10 seconds). So I added another 16 managers and slightly less ducting and then switched the tesseracts all to one frequency (same number of them and faces). Those managers were also filled up in almost no time flat. Seeing as how the ducts in the reactor were still crammed to capacity I simply started stacking managers in a tower with the 4 sides lined with ducts with a tesseract on each pipeline (again, all one frequency). I managed to fill up a few dozen of these and partially fill 14 more. In total, including the ducts themselves I calculate that the one cell generated somewhere in the range of 9075 buckets or 9,075,000 mB of steam.

I then decided to scale up my earlier test and the let the turbine run with more steam (ie a manager instead of a tank). I gave it one full manager (128,000 mB) and connected it to a turbine. Approximately 2'47" later (time varied +/- 2 seconds) I got around 145,000 MJ (+/- 1500). I then calculated that you get approximately 1.1 MJ per mB of steam when fed into an LT.

Edit: After running the total steam output test again I got a total of around 14,000 buckets or 14,000,000 mB of steam. I have no bloody idea how this thing generates steam lol.

[Flextt]

With everyones amazing help, especially jakalth via PM, I have finally reached a satisfying reactor output. however, the overall output dwarfs my buffer capacities, which I kinda need due to currently running in batch-mode.

Question at hand is: how can I have a liquid manager output into a liquid manager directly below it? Is it even possible?

[jakalth]

Only way to output from one manager to another is to use piping or liquaducts to transfer between them, I'm afraid... Since each one needs it's own input and output to move steam. But if you stack them, and run the liquaduct from your reactor up one side, and the output to your turbines up the other side, you can get the stack to act similar to a single larger tank.

[phazeonphoenix]

LMs are not designed to interact with each other. Due to the fact that they don't join like tanks and interconnect like portable tanks they can't do anything with out a pipe putting in or taking out liquid. So you're limited to at least two segments of piping between LMs. I think the LM comes from the Steve's Cart mod and it's designed to manipulate liquid like refueling and emptying an on board tank of a minecart. If you look back to about the 3rd or 4th page CharlieChop describes an arrangement of liquid managers that I use for large scale long term liquid storage. I've not yet attempted to hook that design to store steam.

[Digdug83]

I believe, but am having a hard time figuring out for sure, that the minimum steam flow rate needed to generate power is 25 mB/t. When I was running power output tests one tesseract being fed 400 mB/t was power 15 turbines pretty consistently. If anyone wants to fiddle with that it'd be helpful for turbine layout.

[phazeonphoenix]

So I decided to test storing steam in Liquid Managers and I found out some about how Atomic Science and the liquiducts interact. Hopefully I won't bore anyone and won't repeat anything I've said before.

First, It is very difficult if not impossible to determine how much steam can be produced by a reactor core design. The random aspect is too great. I started with 64 Liquid Managers in parallel and ended up with 128 and still ran out of storage a couple of times. This makes timer based fuel injection a little problematic since there could still be steam in the system after a large amount is created or a period of time where the reactor won't be running because of a small amount being created. This isn't as bad as it sounds though.

When using steam funnels and liquiducts no steam is lost. None at any point in the whole process. If your liquiducts are full, the remaining steam will remain inside the steam funnels. There's no apparent limit as to how much steam can remain inside the funnels. Turbines won't waste steam when connected to a liquiduct either.

As I partially revealed in another thread, I discovered Liquiducts are more like big custom shaped tanks than pipes. I think this was done to improve performance. Buildcraft Waterproof Pipes have the notion of where the liquid is inside the pipe. Liquid at the source end of the pipe has to travel to the outlet before it can be drawn. Liquiducts simply have a total capacity that is added to or drawn from by any connection at any point on the duct at any time. Each segment of Liquiduct adds 500mb to the total liquid capacity. Examine a liquiduct with a multi-meter to find it's capacity.

Each connector on a liquiduct is capable of moving 100 mb/t of liquid and multiple connections to a tank or duct will multiply the throughput. If you have more outputting connections verses inputting connections to your liquiducts they will appear empty and the actual output would be less than 100 mb/t per connection. The inverse is true as well. More inputting connections would result in a full liquiduct and the actual input to the duct would be less.

That fact is very important when attempting to use Liquid Managers to store steam since each LM has a connection filling it. My test design of a reactor core had 80 connections to the funnels pushing steam into the liquiducts and at first 64 then 128 connections filling my LMs. With 64 LMs the liquiduct running from the reactor to the LMs was full. With 128 LMs the duct was almost empty. In other words, my Liquid Manager storage array was capable of moving MORE liquid than the reactor core could produce. Interestingly enough, while all the LMs filled evenly, they drained unevenly. I eventually grouped my turbines and LMs and ran 4 parallel and separate liquiduct runs which helped even things out.

Since the liquiducts and the steam funnels hold unused steam you might be wondering why use Liquid Managers at all. There's one very important reason why. LMs work with gates. I was able to connect to about 6 different LMs in the storage array and use gates to emit a signal when the LM was empty. I connected the gates with RedNet cable and attached a glowstone illuminator. In every one of my tests the signal was triggered at least a minute before the steam drained from the system and the turbines stopped. This would solve the fuel injection timing problem. With 20 turbines attached the system ran for 15 minutes between refueling.

And finally (last thing I promise) I learned rather quickly that Liquid Tesserects won't work with moving steam around. I tried CharlieChop's tesserect driven LM storage system and discovered they simply do not move enough liquid in a single tick to be of any use. Steam is produced and used VERY quickly, faster than any other liquid.

[Moleculor]

Since the liquiducts and the steam funnels hold unused steam you might be wondering why use Liquid Managers at all. There's one very important reason why. LMs work with gates.

What's the advantage of Liquid Managers over standard Buildcraft Tanks, if I can necro this thread? Tanks work with gates, no steam is lost even without storage at all, so why not go with the most basic form of storage around just for gate detection?

[weirleader]

Having followed this discussion (though I've never used LM tanks, yet) it seems that the LM tanks hold much more in the same amount of space.

[phazeonphoenix]

Yeah that's the just of it. It's hard to compete with storing 128 buckets in a single block. The thing is that using LMs as just a liquid tank is like using a pocket knife as a letter opener. Yes it works and does the job right but it's capable of so much more. But it's the classic Tekkit mentality that anything worth doing is worth doing in the most complex manor possible, and even crafting the LMs is a complex ordeal.

[jakalth]

Agreed on both accounts. It's the sheer size of the LM's that make them appealing. They also do work nicely with gates, pipes, engines, and anything else that can detect and/or use liquids. But they are costly and complex to make.

[Moleculor]

Yes, but you don't need to store it, so why waste the materials on an LM when a buildcraft tank will do?

In fact, I just finished building my reactor/turbine system. I have a buildcraft tank, but it's just there to indicate the presence/absence of steam. it's all pumped straight to the turbines, no storage between the reactor and the turbines.

Reactor's there on the right, turbines on the left, liquiducts pipe straight through. A tank sits below the liquiducts filling faster than it drains, and the gate there tells me when the tank is empty. No steam is lost, because no steam is ever lost when using liquiducts and steam funnels.

[phazeonphoenix]

These fusion reactors are complex and there are multiple ways to achieve the end ultimate result which is large amounts of steady power. Do you run the reactor core constantly or in bursts? Do you store the steam or just let it clog the pipes? Do you use RECells to buffer an intermittent period of power generation or is it truly constant? It largely doesn't matter how you balance it long as you can achieve something effective. My experience with storing it has proven to me that it is largely pointless except to keep enough around to ensure the turbines don't spin down, ever. But then I didn't use RECells to buffer my power so... on and on

[jakalth]

I went with the excessive power storage so I can run mine for short bursts with longer delays between running it. But then again, I don't have the high power demand to require constant operation. And the steam buffers in mine are mostly for show, and so I can tap into some of the steam when I want to for other uses, like filling buckets with steam for giggles.