Posts by gorzak

  • Mark 1-1-suc that puts out 120 Eu-FullCycle

    I see a lot about what to do with power from a CASUC, but not alot addressing the reactor design.


    I am no expert on CASUC, but a couple things jump out at me. First, 6 chambers leaves no room to pipe in ice, 1 ice /s input means manually adding ice stack once a minute? You cant even add ice that fast, the other ice is not yet consumed. Am I missing something, like pipes aren't needed?
    Second, Air cooling instead of water? What is the reason?
    Whats with the abundance of heat dispensers?


    Here's my take for the same U config/ power output:
    http://www.talonfiremage.pwp.b…vwciiatwm0fs4qp1f65j8708x
    2 less chambers... 14 less heat dispensers. Same cooling cells, same internal cooling. Better external cooling. Unused space is for pipe. Feed of ice timed to consumption to not waste power producing unneeded ice.
    Compare my design to yours. My goal isn't to radically change anything, I have only changed air cooling to water and then removed the unnecessary, they are nearly functionally identical.


    http://www.talonfiremage.pwp.b…48lqjnlqo1b6lodb6bcgzzgg1
    Thats with slightly better cooling, consuming ice 1x per 8.5s


    Over all though, I'd prefer Duezmanns mark II
    http://www.talonfiremage.pwp.b…4k87s7mw1b5s3rmkfdlebhmv5.
    Using SUC when the fuel has expired is an option, there are 5 empty slots with the fuel expended. This would be the most efficient use of the SUC if you are not continuously applying them. A run of his reactor generates 280,000 excess heat, so pack the ice in. It will take a bit of time for the dispensers to suck up the cold, maybe ~10 -15 min?


    If you make Ice via compressors, don't forget that it costs eu. Factor that cost in and you get lower eu/t than the 120 you claim.


    Referencing the guide in this forum, ice removes 200 heat from the hull if it is over a threshold. A stack of 64 ice will remove a total of 12800 heat over 64 seconds. If it is not replaced/resupplied during the run, that changes to 12800 heat over 10000 seconds. A coolant cell removes 10000 heat over 10000 seconds at 1/s. It also can absorb up to 10k heat, making it responsible for a heat buffer of 20k in designs that won't melt parts. So a coolant cell is capable of handling 7k more heat than a single stack of ice, and any design that would melt parts with that cell in would melt with the ice in it's place. Even @ 300 heat per ice, that is not enough to make it competitive with a cooling cell. This is why you see Continuously Applied Single Use Coolants... or none at all. Real "single use" of coolants would be to pop them in a hot empty reactor to cool the parts down for another run of a mark II design.

  • Reactor planner version 2

    Quote from Talonius


    - A finer heat bar would be dobable


    Thanks!


    The heating and cooling happens in granular 1 second intervals. For heat generated as an exact factor of cooling, that works, but when they aren't the shorter the timing the more that granularity is apparent. I'll give an example of what I mean.


    http://www.talonfiremage.pwp.b…q=1k101010015015111r11r10 176 heat, 66 cooling 200 active eu/t 3.33 energy efficiency.
    Minimum timer 1 sec generating / 2 sec cooling. 66.667 eff eu/t (because 22 cooling is wasted every 3 seconds using this minimum timer)
    Maximum timer 33min 15 sec generating / 55min 41 sec cooling 74.8 eff eu/t This should actually be shortened to 55min 40s, waiting that last second for it to finish cooling means the first second it is generating wastes 66 cooling, effective eu/t is 75
    8.133 eu/t is gained by going from the minimum complexity timer to the maximum. 80% of that is 6.506. What is the minimum timer that gets 6.506 better then the simplest possible timer, and no more than 1.627 worse than the most complex timer possible?
    2sec on / 4 sec off is 66.667, same as minimum.
    3sec on / 5sec off gets 75, same as maximum. This is the minimum complex timer that gets at least 80% of the difference. It happens to get 100% of the difference, being perfectly equivalent to the best timer.


    With a linearly increasing timer complexity, for people like me who don't use redpower/ computer controller mods, this simplifies the construction of the timer by a factor of 665. Going with the simplest one is a wise choice if not a lot of cooling gets wasted, but in this case a timer less than 3x as complex gets >10% output boost. Being able to contrast simplest, most complex, and most of the way toward best output at a glance lets you make decisions about what kind of timer to construct. That was my thought at least. I don't know how many mad scientists make mark IVs without redpower or other smart controller mods. I may be the only person that uses that feature.




    Quote from Talonius


    - I'm unsure what you mean with the last idea, the planner already displays the amount of 'top-up' lava needed for pre-heated designs that cooldown over time even when generating. (In the Breeder info panel)


    Instead of the user using a slider to create a target starting heat, that slider becomes a target for target heat during operation. With a perfect breeder, starting heat is operational heat, so this is more for positive breeding. So instead of 6k starting heat and seeing how long it can run before meltdown, I put the slider at 9k heat. This would tie in to the timer suggestions to come up with the required starting heat for each suggested timing interval.
    The minimum timing interval would likely be heat loss over time, so would likely need to have a starting heat over the target. It may even have heat requirements that would melt parts and thus not be doable, knowing that tells you you either need to aim lower or make the timer more complex/better.
    The 80% good timer would probably be closest starting heat to the target 9k heat.
    The long timer would probably have a starting heat around 8k, since it would spend half the time under and half the time over.


    It lets the user aim for what they really care about, and tells them how to get that result. I like what's there now too, you can plug in a starting heat and see what happens. What I'm suggesting excludes that, as average operational heat dictates starting heat, and vice versa. I suggest a toggle feature, where you can change the slider from target heat to starting heat and back.

  • Reactor planner version 2

    Ok, feature requests.


    Finer tuning on the initial heat setting. Instead of a 4 interval slider, ideally it would be a smooth 10k ( interval 1) slider, though I'd be happy with 10 (1k intervals)


    The next 2 feature requests kind of tie in to each other.


    The one that would benefit more people would be an output for timing suggestions for mark III+. Minimum, maximum and 80%*, effective EU/t for each. Minimum has heat generation set to 1s, then cooling the minmum # of seconds to hit 0. Maximum is heat generation the way it is now, until critical or melted parts, then as long as it takes to cool off completely. They have different EU/T, Minimum is lower but much easier to set up. 80% is .8 * (Maximum timing's eff eu/t - Minimum timing's effective eu/t) + minimum timing's effective eu/t as your target effective eu/t and find the minimum timing interval that gives at least that. I wager 80% will usually be fairly close to minimum and show you how much returns you get on increasing complexity of your timer. This will promote designs where cooling is a factor of heat generated, or just over.


    The next feature suggestion requires the previous. It would be a toggle for starting heat, changing it to average target heat. Setting this to say 9k for breeding would then tell give you the same above options, but include starting heat for each to maintain that average heat during operation. Cooldown time would be excluded, as breeding does not occur.

  • How to preheat your breeder

    Continued.


    With precision heating, you can set up a positive breeder as a mark II. This allows you to design energy producing configurations that breed isotopes, hybrids that essentially run on isotopes. Breeding requires heat, so to breed efficiently, you want to end the cycle as close to max heat as possible (risking meltdown). Positive breeder designs that start cold are much less efficient, as they spend less time in the breeding sweet zone and take more time to cool down. The hotter you can start a hybrid, the more efficiently it breeds, and you can drive that efficiency gained into energy production. If you don't know what heat you are at with some precision, knowing whether you are starting a mark II positive breeder hybrid or mark IV positive breeder hybrid is something of a crap shoot. If you know your system is starting at a dispersable 45% heat and generates 45% heat during operation, you know its a mark II. When you come back to decent energy efficiency and 1 refined isotope / uranium consumed, you start to understand why this benefits you.


    When we are heating up parts, we don't want to fight external cooling. 5 lava is enough to cancel external cooling entirely. Isotopes will generate heat even while the reactor is off, so with designs that include isotopes, I include external cooling to offset the heating, leaving the hull a constant temp when it is not in operation


    Heat dispensers -
    Impossible to heat up seperately, as they are tied to the hull. Most of them will be sucking 25 heat /s from the hull, ones adjacent to hot uranium are a bit harder to work out. Nearby adjacent could be dispensing/absorbing heat, but in practice only do so to correct a heat imbalance, like if they started at different temps or one is next to hot uranium. At 400 seconds to melt if enough heat is being generated, I tend to pop in all the ones I need and heat them all up at once with no cooling cells. With a hull constant config and no cooling cells, they don't cool when you shut the reactor off. You should shut the reactor off when you store these, or it will explode, as removing them means the remaining ones will be overwhelmed and not remove heat from the hull as fast as it is being generated. Figure out how much dispensers your config needs, multiply by 25 and that's the heat the dispensers can keep up with. Make a config that comes in at UNDER that heat, the closer the better for example 14 dispensers would be http://www.talonfiremage.pwp.b…g=10101015114010101001010. Remove external cooling with lava, hit the on switch, and wait for the last red dot to disappear. THEN HIT THE OFF SWITCH OR BOOM. This also preheats the hull to this amount. I do this first for a hot hull for the next step.


    Cooling cells - Heat dispensers are bad, but cooling cells are worse. They are actively working against our goals of heating. They do not seek out or accept much heat. A cooling cell next to 1 dispenser heats up by ... 5/s that is 2000 seconds if it is next to just one dispenser to heat up. This is where the 40 minute heat up time comes from. We can do better than that. If we surround it with dispensers, we get 23/s (only while the cell is much cooler than the dispensers), which is still ridiculous. That's longer to heat up than the dispensers, at a ratio of > 1 dispenser per cooling cell. So to efficiently heat up coolant cells we have to feed them heat from uranium. Which is fine, since we don't want that heat going to the hull, no dispensers are removing it. You can pop a coolant cell in 60 seconds. I came up with a faster design that kills a coolant cell in 25 seconds... but we can use neither of these designs, because for 1 we don't want to flip the on off switch each time we switch out a cell, and 2.) 390 heat into 1 part is too much! I came up with a design I call Crazy Ivan, just for this job. With the hull already at ~9k from the last step breeding effectiveness should be good, and as long as the plating lasts, no heat goes to the hull. The idea is, when the last spot of red disappears from the bar of the coolant cell, pop a cooler cell diagonally from it, and pick up the hot one and put it in your inventory, not the reactor. Repeat until you have the number of hot coolant cells you need. At 160 heat/s into cooling cells, this is more efficient at heating cooling cells than a checkerboard of 19 dispensers and 10 cooling cells. This is a manned configuration, you do need to have your finger on the button for as long as it takes to heat up each cell, up to 1 minute per cell, less for reheating warmed ones.


    From a cold start, this method has reduced warm up from 45 minutes to 30, and gives a much greater level of precision for reheating warm parts. With hot parts stored, you can do your normal breeding run. You could also use a mix of hot and cold parts for a target heat. For example, I want to target 5k starting heat for a positive hybrid breeder. 47 parts + 1 hull, If I start with 24 hot and 24 cold, I get 5k (really 4.5k, as the parts are more like 9k hot than 10k) heat average that can be dispensed during operation. With that hybrid design, I would heat the 13 dispensers and the hull, then crazy ivan 9 coolant cells for a warm up time from cold of around 25 minutes.


    If you have a method I haven't mentioned, or any refinements, please don't hesitate to chime in.

  • How to preheat your breeder

    I'd like to present this as something of a comprehensive guide on both available and preferred methods of heating up a reactor.


    Methods:
    1. Lava buckets
    2. Heating configuration
    3. The art and science of precise heating.



    1. Lava Buckets


    I'll start with BE CAREFUL! because people often have to make thier own mistakes, but I'll also go into DON'T DO THIS!
    As this is a typical breeder, I'll be using it as an example: http://www.talonfiremage.pwp.b…i=1m101010114010101001019


    A lot of people explode their reactors when they try to use this method because they don't have a good understanding of how it works! How do people make this mistake? Well, you set up a configuration you see, say for a perfect breeder with 4 chamber 23 cooling cells and 7 heat dispensers. All this work into all those parts, and now you will finally get the increase in uranium yield you've been working towards. Now it's time to add lava buckets. Maybe you know how many you are supposed to add, maybe you are just winging it. Maybe you know that lava adds 2000 heat to the hull, maybe you don't. Maybe you know it's supposed to be somewhere around 9k for breeding. So, you add buckets. And nothing visible happens. So you add more. And you see the particle effects of smoke. You're looking at the parts, and they aren't getting hot. Maybe the smoke goes away, maybe the fire particles turn back to smoke, but it seems like you aren't making progress. Maybe this is when you find out you need > 240 buckets of lava to heat this config. So you add more lava. Still not really making progress it seems. So you start putting in a few at once. The first 2 you put in are fine. You put in 2 again, its fine. You put in 2 again its fine. You put in 3... Still the parts aren't getting hot. this is taking forever. If you were this patient, when you put in 3 again you will start being irradiated and have to flee the reactor room. If you were slightly less patient and bumped up to intervals of 4 or 5, you probably blew everything up. I irradiated myself this way, plenty of people have this exact experience. This is pretty natural behavior... don't do it though.


    Lava buckets add 2000 heat directly to the hull, the part that is responsible for the meltdown. Hulls have 10k to 16k capacity for heat, depending on the chambers. 5 buckets of lava AT ONCE will vaporize a zero chamber reactor, no matter the cooling config, make a 2 chamber reactor go critical (chance of meltdown, or melting nearby blocks)or start radiation burn in up to a 4 chamber reactor. 8 buckets at once vaporizes all reactors. So how are you going to add the 240+ lava buckets that it needs to heat up to the right temp? Instead of adding them at once... add them over time. Heat dispenser pull heat from the hull into themselves, and transfer that heat to the cooling cells next to them. They try to equalize their temperature to the hull temperature, but each can only suck 25 heat out of the hull per second. When you start by adding a bucket of lava, waiting ~11 seconds and adding a second one, nothing visible happens, but each dispenser sucked in about 275 heat, dispersing that among ~4 to 5 available parts and reducing the hull temp from 2000 to around... 50 or so. so each part & the hull is now around 50 heat, which is not enough to move any bars or make anything visible happen. What you are supposed to do is continue adding 1 bucket every ~11 seconds to keep this increase flat and linear, 240+ times. That is 44 minutes of adding lava. That same 240 buckets of lava in a geothermal would net 4.8 million eu. People do not do this. Those that know, do not use this method, those that don't know... They overload the heat dispensers. As nothing visible happens, you think it can handle more. When you go to adding 1 every 5 seconds, the hull is gaining ~1k heat for each you add. When you take a second to ponder, it might catch up a bit, but when it catches up, still likely you have not made enough of a dent in what is required to move bars. So more lava faster gets the particle effects, because the hull can't shunt heat fast enough, and when yet more is added, the nasty happens.


    If you haven't tried this method, save yourself the trouble.
    If you want to know why it's hard to make this work it's because time is needed for each part to receive the heat you are adding to the hull, and the total heat needed includes the parts 30 parts + hull to 9k heat. 279000 total heat with the ability to dispense slightly less than 175/s. 44 minutes of pouring 4.8million geothermal energy down the drain.


    2. Heating configurations



    Most people graduate to this method when they realize that buckets of lava are less than reasonable. It's premise is simple. Efficient consumption of fuel generates lots of heat. Set up a energy producing config to produce heat, then switch to breeder config when it gets hot enough.
    This method is much simpler, and somewhat safer. It is still not entirely safe.


    There is a trade off between safety and time. The safest way would be to set up a mark II config that runs for 1.05 cycles, and show up right when it finishes, almost 3 hours later. 3 hours to "heat up your breeder" no chance of explosion. Just pop in the breeding config using the hot parts and you are golden. Most of us don't want to wait 3 hours, but if you are a planner you can make that work. Most people set up a mark IV config and let that run for a bit, then when it gets hot paying closer attention to it. Heating configs should have the same number of cooling cells as thier breeding configs have, and enough dispensers to soak all of the heat generated. Heating up the parts is kind of the point of heating up your breeder, do not set up configs that overwhelm heat dispensers, the reactor will explode before the parts are hot. Divide the heat generated by 25, and add 1 to be safe. Cooling cells can't be next to extra hot uranium, and you get bonus points if most of the parts can stay in the same place between configs.


    For example:
    http://www.talonfiremage.pwp.b…c=1m101010114110101001019 Breeding config
    http://www.talonfiremage.pwp.b…g=1m101010114010101001010 Heating config


    So you run the heating config for ~40 minutes, then swoop in and change it up to breeding config. Note that this is around .25 cycles with 4 uranium. This means heating it up this way takes about 1 uranium, but that uranium is consumed with efficiency of 3 with 120 eu/t output with around 6million energy produced. The istopes in the heating config do not breed very well during that time, they are mainly there to pump up the heat produced.


    You can substitute other energy efficient designs, as the heat that is their waste is the real goal, but keep an eye on how many hot parts you need in your breeding config. Heating up 16 dispensers and 15 cooling cells would mean you have to swap in 8 cells from storage. If those aren't simlarly hot, you'll have to heat them up by running the heating config more.


    Some may be surprised that heating via mark IV config and heating by pouring lava take a similar amount of time... but it all comes down to how efficiently heat is dispersed among the parts. With the same number of heat dispensers, heat is distributed at the same rate. Adding more heat will not heat the parts faster, it will just heat the hull to explosion


    Remember when I said this wasn't exactly safe? I believe it to be what most people do, but... You are setting up a mark IV to run for >30 minutes unattended. This takes time management skills, or some other kind of safety mechanism (like my chicken kill switch) because @ 30 minutes you have pretty good breeding, but should probably let it sit for a while. At 40 minutes you have very good breeding, time to reconfigure. At 50 minutes, you have a crater. So yeah, heating up a reactor with uranium takes a heat generating config. Heating it up with a config as quickly as is possible means generating heat as quickly as it can be dispensed. Leaving a reactor that is generating heat quickly alone for too long means boom.



    3. The art and science of precise heating


    In this section, I am going to go into when you might want to precisely heat a config, how to heat individual parts, and a tense game I like to play with a mark IV breeder on a short fuse.


    You may have realized by now that you could save a lot of time/energy if you stored one set of hot parts for breeding and cool set for energy generation. Invariably, the energy generation ones get hot and the breeder ones cool and the stored ones end up mediocre heat. At some point, you have to reheat parts that are at unknown starting heat, and you either give up on storing hot parts, sit on the reactor for 15-30 minutes watching it like a hawk, or figure out how to precisely heat different types of parts. I went with that last one, and I am going to share with you how you can do so. This offers a benefit for 2 reasons. First, as mentioned, reheating those parts for that breeder config. 96% of the work is in those stored parts, if all that was cold was the hull, that is just a few buckets of lava and problem solved. Secondly, with a little math, you can use those hot stored parts for precision heating.


    ... to be continued.

  • Mark II, Industrialcraft Only (4 core)

    BBQRoast already covered it, but I wanted to say the same thing a different way.


    Active EU is what the reactor produces when it is on. You don't get a 95.60 eu output out of 100, you get a 100 eu output while it is on and consuming uranium.


    Then it shuts off, and you get a 0 eu output while it doesn't consume uranium.


    If it was NEVER off, it would explode, and you'd get 0 eu output anyway, along with loss of all initial resources. This is true of any heat positive design (mark II +)


    So if it is required to be off for some amount of time, it's overall output will be lower than the 100eu/t output that you would expect if it was on all the time. That is effective eu/t. Over a long span of time, the eu/t you would get if you ran it as much as the design would let you.


    Note that this does not affect the efficiency of uranium consumption. When we are talking about efficiency, that is what we mean. The more the reactor is off, the longer it takes to consume the uranium. A design like this: http://www.talonfiremage.pwp.b…1k101010037ps011111101110 really does get much more energy out of the urianum... however it takes ~5x as long to do so. It is off every 4 seconds out of 5. Which is why even though it spits out 330 eu/t, you should expect an average of 67.8 eu/t. If you planned to consume 330eu/t, the active output of the reactor, your machines would be off 80% of the time like the reactor is. If you plan to consume 64eu/t ( 2 batboxes in parallel), your machines could remain constantly powered with a small surplus, for 5x as long as a normal cycle, as long as you have redstone shutting down this reactor at the appropriate intervals. Energy efficiency, fuel effieciency, consumption efficiency, all have to do with how much total energy you get from each uranium. Output is energy delivered over time. Efficiency and output are separate and distinct.


    There is a trade off between efficient consumption of fuel and power output. Generally you can get more effective eu/t by using less efficient designs, as they require less cooling downtime. More fuel efficient designs have higher active EU/t but must be shut down for cooling more frequently and thus have lower power outputs when averaged over time. Check out this thread http://forum.industrial-craft.…page=Thread&threadID=3660 for crazily inefficient very high effective eu/t designs.


    Pick a point where you like the trade off between output and consumption efficiency. If you want 4u, I'd suggest duezmans for fuel efficiency (it is the maximum efficient mark II) or ragan's for output. If you are willing to bump it up to 5 or 6 u, Ragan has a couple good mark II's with higher output with 2 - 2.33 fuel efficiency in the first post that rick maintains .


    Don't forget you can always double both output and consumption by making a second reactor. That should slide you more towards efficient consumption, as you gather more resources. If you are willing to produce more reactors, Ricks mark 1 with 40eu/t , and my mark II with 78eff eu/t scale energy production up more cheaply than anything fancier. 2 of my 4u designs cost almost the same as ricks slightly more efficient 5u design. Those would burn 8u for slightly more than 150 eu/t... but I don't recommend just scaling up with multiple reactors that way. I'm a fan of shielding the reactor, and that really changes the cost equation. If you wanted to roll unshielded, multiple reactors is the way to scale up uranium consumption for more EU output. That is still true including shielding, but the cheap designs make a lot less sense when you are sinking tons of alloy into reinforced stone as overhead.

  • Transformers - How do they work?

    Do I have to worry about energy being lost if the pull from my consuming machines/empty storage is less then the push from the pushing producers/full storage?


    Examples:
    1. 1 full MFE -> Lv transformer -> empty batbox
    If I understand correctly, the batbox requests energy, the transformer with an empty buffer requests energy, the full MFE transmits its full 128 eu packet, which is it's maximum per tick. The Transformer makes 4x 32eu packets, which it stores in a buffer. As only 1 32 eu packet is requested, it transmits that and has 3 remaining in the buffer. The next 3 ticks, the empty batbox continues to request, and is fed from the transformer buffer without the transformer making a request from the MFE. As a result, the MFE sends 1x 128 eu packet every 4 ticks, and the batbox recieves 32 eu/t every tick


    2. 1 full mfe -> lv transformer -> split among 4 empty batboxes
    4 batboxs request energy, the transformer with an empty buffer requests energy, the full MFE transmits its full 128 eu packet, which is it's maximum per tick. The Transformer makes 4x 32eu packets, which it stores in a buffer. As 4 32 eu packets are requested, it transmits that and has an empty buffer. This happens every tick, with the MFE sending 1x 128 eu packet and the each of the 4 batboxs recieving 32 eu/t every tick


    Is this correct?

  • Project - making an MK2 with 3,33 eff

    There is a question that the OP has about whether this is possible or not.


    I will help by providing a way to answer that question.


    Consider this fictional reactor focused on cooling
    http://www.talonfiremage.pwp.b…=1k10101011111521s1r11r19


    -73 cooling, no heat generated, 54 parts - 40 cooling cells, 14 dispensers. Not very useful, as there is no uranium, unless you wanted to cool down parts quickly.
    Why are we considering this? If we figure out how much heat it would take to melt this, we have a line in the sand. If this design can't cool that amount of heat, no design can.
    First, even though the hull has 16k heat tolerance, for purposes of cooling it should be considered 10k, as dispensers seek to equalize and will melt if the hull goes above.
    So that makes 55 parts to heat up to 10k, or 550k heat to melt parts. In the 10k cycle required to be a mark 2, 55 excess heat would melt parts. 55 excess + 73 to overcome cooling is 128 total heat.



    128 total heat/s total would melt this optimal cooling reactor, that is our line in the sand. While slightly below that may not be possible, we know above that is definitely not.
    http://www.talonfiremage.pwp.b…0=1k101010110010101001010
    3.33 efficiency. 176 heat generated. Definitely not possible to be a mark 2.


    You are talking about popping a stack of ice in as an alternative. To be clear, we are talking about SUC, not CASUC, there is no continuous supply. Ice does not compare favorably. If the reactor is a mark II, a cooling cell had the opportunity to cool 10k heat in a 10k cycle and soak 10k heat. Ice cools by 300, 64 ice cools by 19200. That is slightly less than the 20k the cooling cell takes to melt in a mark II design. A single stack of ice is less effective than a cooling cell in a mark II.


    There are 3 ways to make ice better than a cooling cell.

    • Continuous supply. This should need no explaination. 67 ice in a slot is better than a cooling cell, but takes supplying during operation
    • Discard mark II requirements, supply additional ice between operation. Ice front loads it's cooling, so it lets mark 3+ run for longer. Run a preset partial cycle with auto shutoff, then restock ice before you restart it.
    • Replace enough components with ice that you can safely replace dispensers. 19.2k > 10k, dispensers are only a heat soak and don't provide cooling. They do link the cooling cells to the heat. If the optimum layout has a spot with 2 spaces, coolant cell + dispenser is 30k heat, 2 ice stacks is 38.4k heat, slightly better.


    You excluded all 3 in your design specifications. Based on your design specs, using a small amount of ice would be counterproductive. That's fine with me, I don't like using ice in my designs.


    This is definitively not possible.
    In fact, because both energy efficiency and heat generation are dependent on uranium cell configuration, and it looks like your suggested config is the heat minimum for > 3 efficiency, we can say definitively 3.0 is the highest energy efficiency a mk II reactor can get with current mechanics.
    I encourage exploring the edge of possibility. Part of doing that is knowing where the edge is. With mark II, aim for slightly less than 128 heat generated. Each cooling cell replaced lowers the threshold by 2, each dispenser by 1.
    These should be outside the range of possibility:
    http://www.talonfiremage.pwp.b…0=1k101010110110101001010
    http://www.talonfiremage.pwp.b…4=1k101010110110101001010
    This should be just barely possible:
    http://www.talonfiremage.pwp.b…w=1k101010110110101001010
    It would have lower effective eu/t than duezmans because of a longer cooldown, but the ability to breed just slightly less isotopes than it produces for a slight fuel price cut & less time running a breeder.

  • A list of good reactor designs (IC² v1.103 and earlier)

    Quote from Rick


    Mk1:
    2. http://www.talonfiremage.pwp.b…=1o10101001501521s1r11r10 The upgrade to the starting reactor it outputs 80 eu/tick at 2 efficiency. Cost wise its better to upgrade your starting reactor to this than to make a 2th starting reactor. It can run 25 cycles in a row if you somehow instantly refuel it. So its pretty much only possible in theory that this reactor blows up. First gen (24,75 cycles) made by me then later on slightly improved by raGan (24,94 cycles) and then improved again by me (25 cycles).




    Near infinite cycle reactors (mk2 E):


    Reactor #:
    http://www.talonfiremage.pwp.b…=1o10101001501521s1r11r10

    • Eu/tick: 80
    • Average Eu/tick: 78.19
    • Efficiency: 2
    • Cost: Iron 180, Copper 347 and Tin 208.5
    • Cooldown per cycle 3 min 52s


    I consider this a slight upgrade to the quoted reactor. Slightly higher average eu/t, slightly faster cooldown time, slightly cheaper. It can do 23.87 cycles vs the originals 25, which is still effectively infinite.

  • Shielding the reactor

    The melting to lava thing happens at 85% hull temp in 5x5x5, so I don't build shielding there. Realistlically, it doesn't stay there long, as 100% of heat generated is going to hull by that point, all heat dispensers will have melted. Chances are low that everything would melt, but since I don't know what will and what won't I don't build anything in the 5x5 that I don't expect to melt.

    If you are talking attenuation of the explosion in general, that depends on what it goes through before hitting the reinforced stone. Using http://www.minecraftwiki.net/wiki/Explosion as a reference, this is what I get.


    If it's going through air, you'd have to solve this:


    Miniumum Block resistance = ((1.3 × power - attenuation steps × step length × 0.75) / step length - 0.3) × 5.



    for attenuation steps
    whre the block resistance 150 must be > or = minimum block resistance
    Where step length is probably 0.3
    and the power of the explosion is not something I know
    so the maximum attenuation steps for


    150 > ((1.3 x power - attenuation steps x 0.3 x .75) / 0.3 - 0.3) x 5.


    which simplifies to


    attenuation steps >= 5.778 * Power - 40.4


    but still need to know the power of the explosion, I don't.
    edit: Nukes don't use minecrafts explosion code, apparently. Trial and error!

  • A list of good reactor designs (IC² v1.103 and earlier)

    Quote from Rick

    you need 9.625 uranium per cycle for your reactor (9 uran cells and you need to craft 0.625 ore into 5 depleted cells you get 3 depleted cells back from those 9 uran cells each cycle). You get 8 uran cells back with breeding thus you consume 9.625-8 = 1.625 uran per cycle instead of the 9 you normaly need. So thats a 5.5x increase in efficiency.


    Further each cycle makes 9*3.67*2 = 66.02m eu. This means 1 uran cell makes 40.6m eu which is a efficiency of 20.3 which is nice but nowhere near a max efficiency breeder combined with a 3.67 eff reactor. It would be somewhere over 30 efficiency in that case.

    Again, I have to agree with part of what you are saying, but point out something that changes the conclusion. The hybrid design is an energy producing design, it is not meant to be a net gain in uranium, but an energy efficient means of consuming uranium


    Basically you are comparing the net loss hybrid with losses shored up by raw uranium with the lossy energy only config with losses fed by a perfect breeder.
    What if you compared my net loss hybrid with losses fed by a perfect breeder with the lossy energy only config fed by perfect breeder?


    I crunched numbers for an hour over this. I came up with this result. The method you suggest is better yeilds 92.127 million eu per raw uranium. The design I suggested is, as you state worse at 86.226m per raw uranium, which came as a surprise to me, I expected better. I thought the better energy efficiency on average while breeding would outweigh the more time spent in the lower efficiency breeding config. Then I realized I wasn't measuring efficient production, only efficiency of consumption. So I spent like THREE HOURS comparing methods, and came to this conclusion. Existing configs require:


    12857.143s (1428.5s per ucell) @ 10 eu/t for 2.571m energy


    48571.4285s consuming 9 bred ucells at 69 eu/t for 66m energy



    68.571m in 61428s @ 55.814 effective eu/t



    my suggestion



    1428.5s @ 10 eu/t for .286m energy
    55556s consuming 1 bred ucell @ 59 eff eu/t for 66m (the sim disagrees on the eff eu/t, i think because the dispensers are overwhelmed by more and it assumes continuous runtime, this is for both configs)
    56984s to generate 66.286m @ 58.162 eu/t



    The differences in consumption efficiency are slight ~-5%, effective production slight ~+5%, so its roughly equivalent overall. In practice, the method I brought up has one breeder run with a checkin halfway at the Hour and 50 min mark to replace finished isotopes and 8 long production cycles, with a checkin halfway around the 7 hour 45 min mark to replace isotopes. The tried and true method I'm comparing it to has 9 breeding runs with checkins every hour and half, and 8 long runs with no checkin for isotopes. Because they are roughly equivalent, you could use either option, whichever suits your playstyle... but one is not significantly better than the other when compared on equalish footing.

  • A list of good reactor designs (IC² v1.103 and earlier)

    http://www.talonfiremage.pwp.b…1l101010037ps011111101116
    Mark I - O EB
    40 eu/t active & Effective. Energy Eff 2.0, Breeding Eff 3/20, Enrichment Efficiency 3/20
    This reactor produces it's own fuel, enriching as much as it consumes, potentially more. Note that there is a preheat, if you fail to preheat, you do not come close to breaking even on fuel.

    Quote from albijoe

    I have two more in-depth questions on breeder reactors:
    1. Do Multiple Uranium cells surrounding depleted isotopes charge them faster?


    2. Is the production of uranium from depleted isotopes a constant process, or a random process?


    Side: Could there be a reactor type created for reactors that "create their own fuel"? I thought would be a good idea and worth a new reactor type.
    Maybe a "cycle" reactor because it creates its own fuel during a cycle, as opposed to a "breeder" reactor that creates more fuel that it uses? It could also be labelled a "UD" reactor if it has greater than 1.0 efficiency, for having Uranium and Depleted. The relevant stats would not just be efficiency / output / time before meltdown, but also ratio of fuel used to fuel produced (assuming infinite depleted cells available). Therefore you could have a true "breeder" that has low power efficiency and high fuel ratio, all the way to a full production reactor that has high efficiency and 0 fuel ratio. Along the way would be things like a 3U-1D reactor that might produce .5 fuel because the D is next to 2U.

    1. According to the sim, it does. My tests in game confirm
    2. It is a random process, that I don't fully understand yet. Isotopes do not finish at the same time, or predictably, but because of the large number of random events that go into their enrichment, you should not get a lot of outliers (especially early or late), and they should be predictable within a range with some good degree of accuracy.

  • Shielding the reactor

    My reactor uses almost exclusively IC2 & vanilla parts, though I do have an EE alchemical chest instead of a vanilla chest for more space. I do not have a pipe into it, it is six chamber.


    I have not tested this by exploding the reactor, as I am playing it SSP hardcare and don't want to die, but it should in theory provide total containment of all but extreme urainium filled combos that I can't use anyways because I dont have logic pipes. Every line from the center must go through 3 reinforced stone minimum (top and bottom too). Add another layer of stone and it should work even then in extreme cases.

  • Shielding the reactor

    It is possible, and my design incorporates it.


    I don't know if you want to figure it out yourself or just have the answer, so I'll give you a clue, and then the answer in the spoiler. Also, do not build shielding in the 5x5 around the reactor, it will melt and stop providing the same absorbtion.


    Hint: In minecraft, explosions happen as lines drawn outward from the central point, and are absorbed by things in that straight line.