Mizuno reports increased excess heat

  • No, this is the wrong calculation. The heat from the cylinder is much lower where it is surrounded by insulated foil that reflect back a good proportion of the heat. You might perhaps get a 20X amplification factor (taking extreme values) due to emissivity difference and insulation if that was very high. But that is still ball park only 5W.


    Okay, so do the right calculation. Show us! You have 5 W left after the air removes most of the 50 W. Show us how that could all go out the hole, and fool two RTDs, a thermometer and two thermocouples.


    Put a number on it. Do a quantitative analysis.


    How the hell could it be 5 W reach a 5 cm hole 65 cm above the cylinder when 45 W are removed by the moving air? Does all of the heat bounce around, get reflected back, and then it all shoots right out that hole? Is the foil in the box in the shape of a parabola aimed at the hole? None of the heat radiates out from the wall? How would that work, anyway? Give us a model, instead of more hand waving, and baseless estimates of 5 W.

  • Okay, so do the right calculation

    THHnew writes nine hours or so posts with words only,,, no calculations

    Yes. the adjacent wall is insulated. The blower (entrance) is not (although if Jed says this is baffled by similar insulation, this anomaly can't happen, I agree). Remember radiation from the reactor makes the inner insulation foil go up to +80 or something. Expect something similar for the blower: its emissivity is lower (which means it will gain more heat from radiation) however it is obviously less well insulated on the outside.


    So THHNew ,maintains that the blower could be at room temperature plus 80C?= 105C


    105C!!!!

    With 35C air at 4m/s blowing through it?



    This is the problem with words.. and no calculation

  • Remember radiation from the reactor makes the inner insulation foil go up to +80 or something. Expect something similar for the blower:


    No I don't remember,


    Expect something similar for the blower:

    Show calculation expecting this pls THH???

    Convective losses from the blower body to air flow at delta T = 80C........use HTC= 120? fully turbulent

    Convective losses from the blower outside to ambient at delta T = 80C ...... use HTC =4? passive convection


    The operating range of the blower is 20 to 60C.. how can it get to 25+80= 105C?

    Replicators.. pls check THHnew word calculations


    If the blower is too hot to touch... THHnew is correct.

  • Let's try a little calculation.


    sigma = 5.67e-8 W / m^2/K^4


    According to Stefan Boltzman. Att T = T0 + 300 degrees surplus at T0=300K


    black body radiation wattage/m^2 from the inner surface:

    j1 = sigma T^4 - sigma T0^4= 6889 W / m^2


    No let's say the outer box radius is 5 times the inner box radious assuming everything is a spheres and all black bodies, then on the outer sphere


    j2 = j1 / (R/r)^2 = j1/ 5^2 = 276 W/m^2


    Now assume the blower facing the reactor is a square of area 0.1*0.1 m^2 = 0.01m^2 then


    w = j2 * 0.01 = 2.8 W


    Does not look important to me

  • The measured energy out is the energy from the blower wall plus the energy from the air to the blower which both comes from the reactor so the effect this has on the system should be really really small e.g. leakage out of the system besides air flow. I don't see the importance of this issue when we are talking about 250W.

  • Here is an example of how closely the blower power correlates with the air speed measured by the anemometer:

    9025-pasted-from-clipboard-png

    Air temperature and other factors have no visible impact on the air speed. No doubt they do have an effect, but the fluctuations in electric power are so large they swamp these effects. You would need much more sensitive instruments to detect them.


    Wow! This surprising diagram speaks by itself and says: I'm impossible!


    There is no way in which 2 measured quantities are correlated in that way for every single value of 2 numerical series of about 50 data each. Especially when one of these quantities is the velocity of a turbulent flow!


    That diagram reveals that the air speed is clearly derived from the blower power throughout a mathematical formula.


    Who presented this diagram as a correlation of 2 measured quantities has little idea of the real behavior of measured signals. If it came directly from the experimenter, it raises a great concern - another one (*) - about his ability to correctly collect and interpret the experimental results.


    (*)
    Mizuno reports increased excess heat

  • Jed, my problem is, under what conditions was this done?


    To avoid THH's last escape questions a control reactor that uses the same input power as the active one is not the best solution.


    I would always use a feedback (driven by delta outlet of active/control) power supply that heats the control reactor with as match power needed to give the same outlet T as the active. Given the same heat loss this is way more accurate.


    This, in the future, avoids silly questions. (Of course triple reversed clamps with coupled oscillator and .. more creative nonsense are still possible questions...)


  • It's not fixed errors, looks like the errors is in the order of 0.025% for wind speeds, sounds small to me, but one has to verify this against the relative errors of the actual anemometer (precition and not systematic error)

    which should be the crucial measurement here. Then also one need to consider if these measurements are taken momentary or is a mean values for say 1min duration then perhaps these numbers are ok.

  • THH,


    Tell me what I am missing; the R13 (reported on at ICCF21), R19 (run for 111days this Feb-May 2019), and the brand new R20, are essentially the same reactors, using the same calorimetry set-up, and operated similarly (except power input). Only thing different, was that with the R20, Mizuno made one major change to the reactor, and adapted a different mesh cleaning procedure, and adjusted the operating pressure. This is how he described it:


    "The R20 reactor is the most effective version we have developed (Figs. 13-17). The major difference between this and previous versions is that it is heated internally with a sheath heater.

    This change, along with changes in the methods and pressures, apparently enhanced the reaction"


    So all things being basically equal between the experiments; if there were an error with the R20 as you suspect, would not that same error have manifested itself with the R13/R19? Instead the R13 produced 12% excess, the R19 45%, and took off to an astonishing 600% with the R20. And also, could the R13/R19 tests be considered a control for the R20 experiment?


    I may be totally off base, and sound stupid asking all this, but this has been driving me crazy so figured I would risk it and ask.

  • the R20 experiment?


    The 600% is due to mainly an excessive reading of the outlet air temperature.

    The blower body heats up to as much as 105C

    and mysteriously radiates heat to the outlet RTD causing it to give a temperature reading in excess by 30C.


    This is despite the fact that the operating range of the blower is - 20 - 60C

    and that the blower has 40 C air flowing through it at >> 4m/s which can remove much more than 100 W of heat by convection,

    which would make it impossible for the blower body to get to 105C.


    Of course Mizuno never even noticed the fact that the blower stopped operating

    and was too hot to touch.

    If only Mizuno had done those THHnew new physics word calculations,:)


    Replicators beware... do the touch test.. and install those THHnew baffles.(patent pending)

  • Why are we needing to focus on this?


    ask THHnew .,,focussed on it for 9 hrs


    105C temperature of the the blower body..

    64 % error velocity traverse...

    3mg deuterium combustion supplies megajoule of heat.

    laminar flow = flat velocity profile

    R20 instability requires outlet temperature feedback..

    Industrial heat needs to replicate..


    it amounts to an endless distraction for replicators


    here is the 2017 blower spec.

    I doubt whether Mizuno envisaged running it above 60C.

    https://www.sanyodenki.com/arc…_pdf/San_Ace_97BM33_E.pdf



  • That diagram reveals that the air speed is clearly derived from the blower power throughout a mathematical formula.


    Nope. It is measured with the anemometer. It does seem impossibly close, so I zoomed in some more to the individual data points. You see varying separation. Then I subtracted watts from air speed for each 5 s interval. That is, of course, a meaningless number. But if one number was derived from the other, the difference would be the same in all cases, I think. The differences were as large as 1 part in 10, and randomly distributed. Then I looked at what percent watts were of air speed. It varied, randomly. Watts were not multiplied by any single number to equal air speed.


    I do not know what other derivation there might be, or how you would detect it.

  • THH, whilst I think your theory about radiation reaching the RTD would have an effect in some theoretical geometries, it would seem not to be the case with this one, mainly due to the length of the exit pipe.


    And this statement still holds true:


    The outlet air is definitely 10 deg C hotter than it is in the 50 W calibration. This is confirmed with thermometers. So something is adding 250 W to the air. Even if the heat comes from the reactor and magically goes into the blower, it is still anomalous, and far more than input.


    Although, thinking about it, if you assume the perspex box in fig.5 is clad with several centimeters of rigid insulation, this wouldn't fit between the exit pipe and the wall of the main box. This would lead to radiative heat from the reactor leaking into the exit pipe - there only appears to be a fairly thin wrap of insulation preventing this.

    But even though, as Jed essentially says above, excess heat can't appear magically from nowhere, it would sem that this is a mechanism that could artifically boost the reported COP, although not by enough to negate the peaks of fig.4 to a significant degree.



    There is no way in which 2 measured quantities are correlated in that way for every single value of 2 numerical series of about 50 data each. Especially when one of these quantities is the velocity of a turbulent flow


    Turbulent flow has an easily measured average velocity. Yes there may be small fluctuations in local velocity, when measured at the scale of the eddy dimensions, but in practice, these are averaged by the inertia (and diameter) of the anemometer's fan blade.

  • The anemometer info is here, by the way:


    https://www.kk-custom.co.jp/emp/CW-60.html


    You can Google translate it.


    but in practice, these are averaged by the inertia (and diameter) of the anemometer's fan blade.


    It is a hot wire type. But the same principle applies.


    The HP data collection gadget does some averaging. I am not sure I recall how it is set up, but I think it takes 20,000 readings over a second, averages, writes one value, and then repeats after 5 seconds. That will smooth things out. I suppose it will diminish the lag between the blower motor speeding up and the air speed increasing.

  • Nope. It is measured with the anemometer. It does seem impossibly close, so I zoomed in some more to the individual data points. You see varying separation. Then I subtracted watts from air speed for each 5 s interval. That is, of course, a meaningless number. But if one number was derived from the other, the difference would be the same in all cases, I think. The differences were as large as 1 part in 10, and randomly distributed. Then I looked at what percent watts were of air speed. It varied, randomly. Watts were not multiplied by any single number to equal air speed.


    It is possible I am confused about this. I intend to ask Mizuno when I get a chance. But as I said, I searched for a function that converts watts into air speed, and I cannot find one. There are small, random differences between them.


    As I said, I suspect the HP gadget is homogenizing the data.


    It is possible he massaged the air flow with data from the motor plus the anemometer, but I don't see anything about that in the notes. I have extensive notes on every field, which will be published in an upcoming paper. (Who knows when.) Some of the computed fields include complicated functions with multiple variables and physical constants. Unfortunately, in the transition from the old Japanese software to U.S. Windows 10, the functions are lost and only the computed results appear in the spreadsheet. So the spreadsheet is filled with copies of equations and polyglot notes saying, for example, "3 ohm resister here, so multiply by 0.3333." It is messy. Eventually I plan to convert all the fields back into computed values and upload some sample spreadsheets.

  • THH: Expect something similar for the blower:


    RB: Show calculation expecting this pls THH???

    Convective losses from the blower body to air flow at delta T = 80C........use HTC= 120? fully turbulent

    Convective losses from the blower outside to ambient at delta T = 80C ...... use HTC =4? passive convection

    The operating range of the blower is 20 to 60C.. how can it get to 25+80= 105C?

    Replicators.. pls check THHnew word calculations

    If the blower is too hot to touch... THHnew is correct.


    RB (& Jed): you are asking me to guess here. I was just pointing out that the inner foil must go up to 80C or so. I don't, obviously, expect the blower to be so high because it is not insulated on the outside, however OTOH it has higher emissivity. I'm not comfortable guessing emissivities for foil and blower. But an upper limit on radiative forcing could be got assuming worst case hemispherical visibility of reactor, reactor emissivity = 1 (too high), blower emissivity = 0 (EDIT from !) (possible), and comparing this with convective cooling of blower to get max temp rise above ambient. RB why not do this?


    My point is that in this setup - because of radiative heating which is v small at low case temps but becomes relatively 10X more significant at case temp = 380C, the blower could be significantly hotter than the air stream for a reactor it can see, and this effect depends on where the reactor is relative to the outlet.


    Having now looked at the direct radiation issue in some detail I'd conclude (joining this as a summary to a few other things):

    • Artifacts due to blowers, mounting brackets or RTDs, anything exposed to case radiation or in thermal contact with such are an issue in this system and must be carefully checked: they could lead to false positives. However the blower heating up (not heating the RTD via conduction) can't introduce a false positive, it is just another way to heat output air (see mea culpa below).
    • If the control and active reactors are exactly symmetrical in the case, with symmetrical exhaust and inlet, the excess measurement here would mean underestimating calorimeter heat loss through sides, making results dependent on exact match between control and active setups, and making the "first principle" calculation not relevant
    • If symmetry does not exist this could also skew control vs active - however the other mechanisms to do this are any asymmetry of reactor shape, size, color or position in case and what that does to the relative cooling of each reactor by the air stream
    • There is a first principles positive error of up to about 20% based on the varying velocity profile of air in a tube at Re ~ 5000 - 20,000. More clarity about the tube and airflow measurements would allow this to be determined more clearly, and perhaps reduced.
    • There is (GSVIT v clever idea) an issue with false readings due to room temperature change interacting with the time constant of the reactor body, which is long enough for thermal inertia to "remember" older input temperatures and thus skew results up or down by maybe 10%. Needs checking.
    • Temperature and power are not equivalent. High case temperature could be either more power out, or less efficient combination of forced air convective and radiative cooling.
    • The blower heating up and indirectly imparting heat to the exhaust air is an almost non-issue (and to my shame I did not see this for a little while above, the way one often does not) because that is measured just like heat from the reactor: if the air heats up that represents the same amt of power. There is a small caveat - the heating via the blower might be less even than from heat via the input air. However, that is very unlikely because you'd expect the blower fan blades to do most of the heating and mix well at same time. And there remain RTD mount conduction artifacts to check here.
    • Overall I still think these R19 results look pretty good, but there are enough question marks that I'd badly want a proper write-up of the exact conditions in the system used to generate the stated measurements, with control and active done at same time, and with each of the possible error mechanisms carefully considered and eliminated.
    • Overall these details don't matter with the single larger sample result as claimed for R20.


    Everyone else.


    I have said (early in this thread) and have been repeating it. Clearly from posts above I need to put it in my signature, from the number of people arguing about R20.


    • R19 results are high enough to be solid with this method of calorimetry - if all holes that can easily be plugged are plugged. I highlight things which because of lack of info about R19 methodology and/or lack of precise analysis (may be beyond me - or just impossible due to lack of info) could result in significant artifacts that explain R19 results without extraordinary nuclear reactions.
    • R20 sample result is so very high that this is not a matter of calorimetry. I've not been addressing this. And I've said so many times.
    • R20 result seems weird to me if real and reaction rate is determined by temperature - the system would become hysteretic (unstable) very easily. Mizuno would probably notice that and indeed triumphantly point out that it generated heat without any input - the generated power is 6-10X the input so adding 10% - 20% to the thermal insulation would make this self-sustaining. This information - mostly not commented on here or rubbished - is highly significant if you think R20 results real. It means this system is controlled by something other than temperature.
    • I'd like to highlight in some comments here the tone. I can understand those who, like anonymous, reckon replication of R20 is the thing and therefore analysis of R19 irrelevant. I'd point out (to those people) the issue about triggering and R20, which is obviously relevant to any replication because it strongly hints at possible mechanism. What however is the point of simultaneously arguing strongly that the R19 results are sound in all respects, and noting that the R20 results are so significant R19 is irrelevant? If you consider R20 as sound, then R19 is indeed irrelevant, why argue. The detailed analysis of calorimetry in R19 is only needed if you replicate and get similar results. If you replicate and get R20 level results it is still (almost completely) irrelevant.


    It is a lot of effort (for me, at least) to understand these (R19) results at the level of: what are all the possible errors and what magnitude could they have. Jed here has done that already, spending several years with much closer knowledge of the setup: but Jed is one person, and fallible, so he will welcome - and needs - others. I do this because it is fun: it stretches my abilities (I never learnt all this heat stuff at uni, so I need to look it up) and it is like a puzzle. I see few posting here on it with the same interest, instead they treat it like a political debate where you want one side to win. I should also point out that there are others (e.g. GSVIT) who do these things more professionally than me. GSVIT have a tone more hostile than this site would accept, but all they are doing is subjecting extraordinary claims to the type of critique they should get: exploring every possible loophole through which artifacts might slip. Replicators not expecting COP = 6-10 might want to check all the GSVIT caveats in addition to what has been said here. I've deliberately not been remembering the GSVIT analysis when thinking about this, so will have different issues.


    Where I disagree profoundly with some here is on the utility of such strong criticism. I think it is hugely beneficial to those who reckon LENR is likely (a judgement call) and therefore see any one of these positive results as potentially indicating useful info about how to make LENR happen. Obviously, in that case, determining which of these results are real, and which artifact, is essential.


    Where I don't disagree is with those who have decided they want to replicate something (say this - which lends itself to replication). They should do this without worrying about how strong are the published results. HOWEVER - they should be even more interested in a careful analysis of artifacts so they can design those out. If they dismissed artifacts - as many do here - they will have much less strong results themselves than could be. These artifacts, from this relatively straightforward experiment, are all easily controllable and/or design outable: and this can be done while also faithfully replicating Mizuno's setup.


    Anyone who wants to replicate LENR should see that as top priority if they want their results to be useful - whether positive or negative.


    THH