"Stable" reactor design melting?? (ic2 experimental)

  • Hi everyone.


    The forum banned my main account accidentally, so I had to post on an alt one. Basically I copied a design off the reactor design thread (http://www.talonfiremage.pwp.b…6ucaceovy6uixc92gxt2grpxc) which is stable according to the planner.


    Ingame is a different story. Instead of running at 0 heat, it runs at 32 for some reason. It does output the correct EU/t (300) so I didn't think much of it, but I came back a few hours later to find it had shut off via the thermal control. The Heat Exchanger in the bottom right, and the heat vent to the left of it had both melted. I tried to recreate in a creative world, and sure enough you can see the exchanger durability dropping as the reactor runs.


    In short, why doesn't this design work?? I am not using MOX fuel in this, everything is exactly as the design above shows (yes, I triple checked the components).


    Also as a side note, I tried the design in FTB Mindcrack pack (1.5.2) using the GT computercube, and it runs completely fine, as expected. The GT computercube is broken in my 1.6.2 version and I can't use it to test builds. I've had multiple people tell me the "vanilla" reactor mechanics had not changed in ic2 experimental as far as heat/eu, so does anyone have the knowledge to explain to me what the problem is here? :) Thanks a bunch.

  • Ok I opened your reactor design and set the heat to....8,100 heat. I can take a guess at what is happening. the Heat Exchanger(HE) Is pulling 4 Heat from the core, and spreading it to the three adjacent Over Clocked Heat Vents(OCHV) the OCHV pulls 36 and vents 20. The excess heat from OCHVx(3)+HE is 52. Now a Component Heat Vent(CHV) cooling is 16 (irrc) meaning all nearby CHVs above and left are probably being used be the other OCHVs leaving one CHV dealing with 52 Heat. so the components are pulling the heat from the core. Dipping 32 in and being pulled out at the bottom right of the reactor. This causes the HE to retain the heat and burn out. While Displaying an Internal Temp of 32


    My guess is that 52-16 =36, 4 heat is being sucked out one of the other CHV and 32 is bouncing into the bottom right HE.




    Data Source(s)
    Alright, so: MOX reactor designs.

  • Hmmm. I did notice as you increase the heat in the planner, that it's that one Heat Exchanger that melts. I don't understand though, why it works in 1.5.2, when the mechanics for heat apparently did not change...

  • If you change that one Heat Exchanger to a Gold one, it should allow the Reactor to run a full cycle.


    I admit, that the Reactor in the current EXP. Build does seem a bit buggy atm.

  • It still loses durability and melts eventually (by "gold", I assume you mean the component heat exchanger). I'm assuming this is a bug and not intended?


    I can remove the bottom uranium cell for now and just run 6 dual cells I guess.


    Edit: Even with one of the uranium cells removed, it still isn't working properly. Heat is fluctuating around the 20's constantly. This is weird!

  • From what I understand, 4 Heat Exchangers Trade 4 heat each with the hull, and the two Advanced trade 8 each. or (4*4)+(8*2)= 32 heat. Dropping the one Heat Exchanger for a Component Heat Exchanger Gives it 28 Heat, now the 1:1 ratio with Adv Heat Exchangers and the Heat Exchangers is changed to 4:3. which isn't as clean as the 1:1 which would produce fluctuating numbers of heat in the hull, around 16-28.

  • The real issue is how the heat interacts with the hull.


    This 0 Core Heat Design
    http://www.talonfiremage.pwp.b…ng7i1fswewvg7413wsr9qes7o
    Works perfectly, for any fuel. It prevents any heat from entering the hull and works VERY predictably.


    -edit-
    The coolant cells are just "plugs" for a heating mechanism (two dual uranium), as it is a MOX reactor

    Interestingly, I tried another of the reactor designs in the stickied thread (http://www.talonfiremage.pwp.b…adh05nlzbpykw84kwczan05q8) and it also no longer works. I guess you're right, and the heat interaction with the hull has broken a lot of the more complex designs.


    I am actually using a very similar design for my MOX reactor(s), I think made by Omicron: http://www.talonfiremage.pwp.b…vyd4gbkpqtiu8lwrf87m6uh34
    I may just use 2 MOX-ready reactors and dump uranium into them. Pity I just wasted so many resources on the overclocked vents, hah. But looks like they're not practical to use right now.

  • Interesting. Which build do you guys use?


    I used this one in a build 230-something world a few months back, and it worked flawlessly full cycle. I'm now going to test in on build 298, which is the last one that had any significant reactor changes, I believe.


    EDIT: test says it is fully stable. Must be something introduced in a post-300 build.

  • Well, if it happened in a rnage of ca. 10 builds, that's a fairly good narrowing down already. I'll have confirm it with one of the newer builds when I have time later, then I can make a proper bug report.

  • Well, if it happened in a rnage of ca. 10 builds, that's a fairly good narrowing down already. I'll have confirm it with one of the newer builds when I have time later, then I can make a proper bug report.

    Not sure if you made a typo, build 298 does NOT work properly. I tried build 215 and the reactor works as intended, so I will have a specific build number for you shortly. *rubs hands* Time to get to work!


    EDIT: BUILD 288 IS THE CULPRIT! 287 works fine, but 288 melts components. Sure enough, on the page for 288 there are a lot of reactor changes: http://ic2.player.to:8080/job/IC2_experimental/288/changes


    I'm not sure how to view the code changes of each file, SVN-style, but that gives you somewhere specific to look.


    If you want to test it yourself, it can take a long time for components to melt, or to even notice that there is a problem at all.


    Instead, set up a reactor with this design, connect it to an industrial display panel using a sensor card, and turn it on. It is supposed to run at 0 heat, but in build 288 or newer, it will show 32 heat... for some reason more heat is being generated.

  • +10 Internets to you!


    Specifically, we could probably narrow it down to the exact files... My guess being
    api/reactor/IReactorComponent
    core/block/generator/container/ContainerNuclearReactor
    core/item/reactor/ItemReactorHeatStorage
    core/item/reactor/ItemReactorHeatSwitch


    I might see if I can dig around in them later. My guess is an error in an expression.


    Now, the other idea is that maybe the new updates rendered the old reactor plannerv3 "bad"

  • Not sure if you made a typo, build 298 does NOT work properly. I tried build 215 and the reactor works as intended, so I will have a specific build number for you shortly. *rubs hands* Time to get to work!


    Erm, no. I did not make a typo. I am on build 298, and the reactor I have in my test world works flawlessly. As you can see from the remaining lifetime, it has been on for more than half its duration - i.e., for more than an entire cycle in pre-buff terms. None of the components show any signs of failing anytime soon. Mouseover damage values for all components are constant and/or flicker around a constant median. This reactor will not blow up no matter how long I run it.


    Instead, set up a reactor with this design, connect it to an industrial display panel using a sensor card, and turn it on. It is supposed to run at 0 heat, but in build 288 or newer, it will show 32 heat... for some reason more heat is being generated.


    This is perfectly normal. No reactor that employs any kind of hull transfer will stay at perfect zero at all times - after all, if there is nothing in the hull, then a component like an overclocked heat vent cannot draw anything from the hull, and therefore it cannot do any cooling and will be skipped. And a skipped cooling component will result in the next cycle being slightly more heat to handle, since the fuel rods still dumped their full rating. But since there is now excess heat in the hull, all components can draw when it is their turn and the reactor is stable.


    This is especially easy to see when it comes to heat exchangers of any kind. They balance their own heat percentage with whatever they exchange with, so if whatever they exchange with has zero heat, then the exchanger keeps itself at zero heat as well and doesn't do anything. The hull cannot ever be at 0% heat by definition if a heat exchanger is supposed to exchange heat with the hull. In fact, the heat exchnager will transfer any heat it can find into the hull in order to attempt to lift the hull heat up to the same percentage as the other components in the reactor. Refer again to the screenshot I linked above - that reactor makes heavy use of exchangers, and as you can see, many components have some amounts of heat in them. But nevertheless, they are all stable on that heat level. That is simply because that is the amount of heat they need to make the entire cascade of heat exchangers from the hottest to the coldest place in the reactor work properly.



    I will try and test this design next, maybe that will fail on me. But the one I am currently testing is rock-stable in my build.

  • going to try this(reactorplanner) with IC2 build-2.0.316


    It broke,


    So did http://www.talonfiremage.pwp.b…rmfbaz49yjahwrmav2hc98jy8
    So my guess is, the Over Clocked Heat Exchangers values have changed.


    Edit
    the second reactor did work, the components stopped taking damage after ~:200-340


    Edit
    After Omicron's information from the below post,
    http://www.talonfiremage.pwp.b…siqdffvmzt5i4sg3mi2s9kr9c
    Is a stable design, even though RPv3 determines it will burn out after 4 minutes.

  • Okay, testing that design, it indeed melts down in short order. Replacing it with any other kind of vent also fails eventually. Interestingly enough, replacing the reactor plating with a reactor heat exchanger does not fail. That one runs perfectly fine, even though the reactor planner says it shouldn't, and hand-math also says so. The fuel rod is dumping 120 heat into it, and it can only transfer 72 - even with the 12 extra from 3 surrounding component vents, it shouldn't survive that. The only way this can possibly work is if the reactor exchanger got buffed.


    Hmm... buffed?


    Observe this: http://www.talonfiremage.pwp.b…wf7zgoihr5w97ifu0820jwef4
    All I did there was replace the basic heat vent in the fourth slot of the bottom row with an advanced one. This advanced one has twice as much transfer as the basic one, and because of this, it starves another cooling component, with the result that that component isn't working at full capcity, causing the reactor to cool less heat than it outputs overall, and eventually the weakest link (the bottom right corner) fails.


    So that means, if all heat exchangers were somehow buffed, then the basic vent now acts as if it was an advanced, and voila - you have your reason why this reactor is no longer stable!


    And how could it be that heat exchangers were buffed? Well, quoting the changelog of build 288: "Reaktor run 2 time per cycle a heat and eu calculation run"
    If the heat calculation runs twice, then heat exchangers tick twice, bringing their transfer ratings to bear twice. In order to maintain the previous status quo, the transfer values of exchangers would have to be cut in half, but they were not cut in half. Hence they now work as if they were twice as good as they were before build 288.


    There is your bug, ladies and gentlemen. ;)


    (But is it a bug, or is it a feature? Maybe it's intended. I'll try and get Thunder to look at this thread, maybe he'll share what his intention was. In the meantime though, you can get your reactor stable by employing a reactor vent in place of the plating, as demonstrated above.)


    EDIT: Bug report submitted.

  • Thanks so much Omicron & TwitchUriah, for both the short-term fix, and for taking the time to work on this problem and writing up a proper report.


    I offer you a slimeball each! MinecraftSlimeMinecraftSlime