Alumina Emissivity and the Lugano E-Cat Test (Bob Higgins)

The following post was submitted by Bob Higgins, member of the Martin Fleischmann Memorial Project team.

I updated my paper analysis of alumina emissivity – applicable to the Lugano report.  I recently had a technical reviewer that found an issue and have recalculated the graphs to correct the problem.  I am enclosing the updated paper (comments happily received).

The revised experiment is making more sense of the HotCat measurements – the new data reduces the inconsistencies in the Lugano experiment.  The paper shows the temperature was 1100C, not 1410C.  The COP was approximately 2.36, not 3.75.  This temperature is compatible with an inconel heater wire.  The results are also commensurate with Parkhomov.
One of the things that didn’t add up was the heater wire.  You see resistance wire coming out of the reactor (glowing red in operation) and are told that it is inconel (OK, inconel is a resistor wire).  If the heater coil were made of something else, capable of very high temperature, why would Rossi have put resistor wire for the leads?  He wouldn’t.  You could use low resistivity Ni wires and not have any heat in the lead wires.  The only conclusion is that the lead wires are the same as the coil; presumably inconel.  Inconel heater wires will not withstand 1410C for any length of time – failing on the order of minutes.  Also the wire temperature is higher than the surface temperature, probably near 1500C.  The same is true for pretty much any ductile metal wire you want to use for the heater – by 1410C wire temperature, you have almost no life.  That’s why MFMP looked into other heating core technologies, to try to get to 1410C.  Only after my analysis, it never has to get that hot.  It appears that the maximum sustained Lugano HotCat operating high regime temperature was close to 1100C; well within the medium term temperature range of an inconel wire heater.
The paper is also available on Google drive at:
  • Obvious

    This is bound to be an interesting discussion in the following week.

    • Gerard McEk

      Good work Bob! I just wonder why the IR meter was not compared with a K-type thermocouple during the whole test. Does anyone know that? (you can buy them up to 1430C). These professors should also know that Inconel does not allow such a high temperatures. I wonder what went wrong.
      This all means also that the total net energy was also considerably lower, but I think that it is still much more than can be explained by any chemical reaction.

    • Ged

      Awesome details, Bob! Thank you for sharing this. It really makes the pieces fit better, and refines the temperature target to something more realistic.

      • Warthog

        We have far more indications that there will be than not. Why would the Lugano testers still be working on things related to the test, if not to ultimately publish. Why would the testers obtain a sample of the alumina, which they have already specifically said was to be used to directly MEASURE the emissivity, if they didn’t intend to do (and according to Mats Lewan, they are indeed doing) so?? And yet EVERY “analysis” I have thus far read which comments on the “erroneous calibration” even mentions that tiny little FACT.

        All this baloney is sheer speculation, based in assumptions made by the various skeptics. Wait for the facts!

        • right, not calibrating at high temperature was very serious error.
          McKubre was clear on that, as many experts in calorimetry.

          anyway we have available dat on calibration à 450C, and calibration at 800W.

          GSVIT claims under-exploit, like others, those values, even if it is not easy because one have to integrate convection factor estimated from calibration, and an unknown emissivity which is probably stable between 800W and 900W

          I tried to exploit those data, but not integrating convection , and concluded that temperature was much lower but COP>1.5 …

          anyway this kind of reanalysis will not be accepted by bad will opponents.

          anyway a muchbetter analysis is the following :
          Lugano test was screwed up by an error of the physicist team, not by magic of Rossi+Darden. The electric installation was setup by the physicists and not screwedup.
          So one can conclude that Rossi was not expecting to trick.
          This reasonably apply to Ferrara test, with additional same argument.
          If one eliminate the Pomp conspiracy as a tinfoil theory, then the Ferrara test is much better calibrated (black dots, thermocouple, full calibration), and the meltdown add to the credibility.

          the absence of expected fraud (the mistake was unpredictable and independent as the test) for the Lugano test, gives also good credibility for the isotopic measurement.

          this is complex, and as I said before, above the capacity of a bad will opponent.
          anyway a businessman, an engineer can understand.

          1- you don’t give a prototype to test if it does not work in your lab, or if you are sure you control the protocol to apply a well controlled trick. thus if the physicist screwup that is not your fault, nor your plan.
          2- if you fraud isotopic results, at least make it look realist and don’t be surprised. By the way again, design a protocol where you tricks is not based on luck.

          whatever did the scientist, the device abandoned for independent test was supposed to work from IH and Rossi point of view.

          That the physicist forgot key calibration, or used dummy instruments made in wood, does not change that point.
          Game theory decide that the reactor was supposed to work, in donator’s mind, when abandoned to the testers.

  • Leonard Weinstein

    Why anyone would directly use IR emission detection to measure power out is beyond me. Parkhomov did far better by coupling the radiation (and convection and all other effects) output to a container and measuring the water boiled off, which is a far more direct and believable process. Also the use of mixed radiation and convection and the large feed wire losses made the Lugano approach poor science. You must use a good calorimetry approach to get a reliable result. If radiation is desired to couple the output to the surrounding, a much higher emissivity is desired. Coating the outer surface of the device with SiC powder would double the emissivity. Also, why would you use Nichrome as a heater? Tungsten or Molybdenum wire would heat as high as you want. When coated with ceramic cement, they are protected from oxidation.

    • Obvious

      These and others are all good questions.
      But perhaps a better one is, without designing a device for ease of testing, how or what is the best approach to determine the heat energy coming from a source that cannot be tampered with for one or several reasons?

      • Andreas Moraitis

        It would have been possible to get watertight results with a calibration run over the whole temperature range. Apparently, the testers had planned that, but they were told that this could damage the heater coils. With a complete calibration, all the thorny math would have been superfluous.

        • Obvious

          Indeed proper calibration is critical.

    • Warthog

      LOL. Because “detecting IR emission” is a well-developed non-contact method for measuring temperature……that’s why. And with respect to the comments above…..there is no such thing as “resistor wire”. Or more correctly, any and all metal wires are “resistor wires”. Particular alloys are used because they are more or less inert to oxidation. I will remind you of two statements….. one from the Lugano report, and one from Mats Lewan. From the Lugano report (paraphrased because I’m too lazy to go to the trouble to look it up)..the testers OBTAINED A SAMPLE OF THE ALUMINA IN ORDER TO DETERMINE ITS EMISSIVITY. And from Mats Lewan…….the Lugano testers are currently working on doing precisely that (probably among other things).

      Why not simply wait for the reported results instead of indulging in all this totally unfounded speculation???

  • result is coherent with many critics of both sides.

    For example if you add SS mode which probably multiply by 3, you get a COP of 6-7, which is compatible with Industrial Heat confident behavior, both in investment and in allowing this tests in Ferrara and Lugano.
    Lugano and to a lesser degree Ferrara open test protocol rule out all the conspiracy theory of fraud, but let open emissivity errors.

    the temperature is more realist and answer some critics about metal melting, and question on resistors.
    we are fully inside SiC domain, and maybe in inconel domain, especially as said in doped inconel domain.

    anyway the testers have to rewrite the report

    • I’ve read the report and this seems to be a very good analysis, to be confirmed by experts.
      It first confirm my understanding of what is an Optis, an array of bolometer sensible to 7.5-13um.
      as expected there is some unknown in the shape of the sensibility of the optris in that range.

      to add complexity the export explain that emissivity is changing drastically over the bandwidth.
      I have made a work similar to them, but assuming a flat emissivity. they make a better work by adding this step in emissivity plus a probable change in sensibility over the bandwidth.

      finally they remind us that the emissivity used by lugano testers is the global “averaged” emissivity , independent from the one over optris range.

      their approach match my own naive vision, with more precision.

      with my assumption (flat emissivity, square sensibility window), the optris signal was proportional to the temperature (minus 217C). with them, it seems not linear at all…

      another key point of the report is how to manage the transparency of alumina at high temperature.

      normally reflectance+transparency+emissivity=1
      some skeptic have reminded that the e-cat was transparent for visible light, and we answered that it was opaque in the opriris range.
      both was true.
      in fact transparency is nearly null for optris, and emissivity is near 1 as MFMP and GSVIT explains.
      but over the whole range and especially above 1000C, transparency is not negligible.
      the key to the answer is given by Bob Higgins, as he remind us that transparency reduce the heat radiated by the alumina, but not by the E-cat. the light that goes through the alumina, is radiating energy out of the e-cat. we don’t care if the radiation came from alumina or from the nickel or coil, or from the alumina inside.

      if the optis was sensible to the wavelength where the alumina is transparent, this could have an impact on the estimated temperature, as we should remove the transmitted radiation.

      on the opposite the effective emissivity of alumina should be adjusted to integrate the transmitted radiation from the inside of the reactor

      without good calibration, the method I would use is the following :

      – get the emissivity curve in the 7.5-13um
      – prepare few possible shape of windows (square, 3dB/octave 6dB/octave, 3then6dB/octave …)
      -> for each shape of window compute for each temperature from 450 to 1450C the integral of plank law adjusted by wavelength emissivity and sensibility window
      -> see if the curve is changing much according to the kind of window, if yes, we have a problem… best is to ask Optris for the window shape.
      -> for each response curve, find an interpolation curve (linear, square, exponential, polynomial) and see if it fits well, and if the parameters don’t change too much with the window
      – if possible, take an average fair interpolation formula, and estimate the error, from window choice and interpolation R^2

      estimate the “response” of the optris from the measured temperature and the Lugano assumed emissivity.

      from that estimated response, recompute the implied temperature from the 7.5-13um band emissivity.

      from that newly estimated temperature, deduce the wideband stefan-boltzman emissivity, and thus the radiated energy.
      the adjust the convection .

      the way to treat the transparency question is not clear.
      since the e-cat produce energy from inside, I would consider that transparency add to emissivity, but this may be abusive. shoul ask an expert.

      tu summarize we need few answers :
      – what is the sensibility curve of optris over the 7.5-13um
      – how to manage alumina transparency in the computation of total radiation

      • Obvious

        I wonder what the effect of the fins is. The bottom of the valley between the fins approaches that of a black body test hole, but the tips of the fins themselves are quite different, with a range between the two being the majority of the visible surface.

        • the fins are much bigger than the wavelength where most energy is radiated.
          however it have important effect on convection.

          what is more important is the surface state at um scale…
          this is why lack of calibration at high temperature is really a problem.

          • Obvious

            I have been looking at quite different results, although with metallic aluminum. Regular triangular groves with a 60 degree V were shown to increase emissivity by 3.3 times. This is with a highly reflective surface, obviously. Essentially, the extra surface area radiates direct to e observer, but also reflects fron either side of the V grooves, then to the observer in both single and double reflections.

          • yes I forgot that fins can cause successive reflection and absorption…

            it should reduce apparent reflectivity, probably the apparent transmissivity is reduced (prism effect, multiple bounces), this increase apparent emissivity logically.

            for sounds and radio the effect is well known.

            once again need expert and calibration.

            from all we speculate it seems
            – that emissivity is nearly 1 for the Optris
            – that total apparent emissivity is much higher than what we imagine, probably near 1, because to alumina emissivity one have to add alumina tranmissivity from hotter black core, plus fins effect…

            so it seems that optris emissivity is same as total emissivity,
            which means finally that
            -> temperature is much lower (1100C , as Abd-ul eyeball thermograpy report)

            -> but power radiated is much more (>twice) than expected at 1100C because effective emmissivity is much higher… and report is nearly exact, by accident

            T^4 at 1100C isn 2.5 less than estimated by Lugano
            but since emissivity is now aroud 0.9 and not 0.4, the total radiation is only 90% of Lugano testers computation…

            if you add to that the possibility that the core is much hotter than the alumina, radiating more by transmissivity than what alumina would do at 1100C, then maybe we are at 100% of Lugano report result…

            #LOL #FACEPALM

            problem is that there is many question, first of which check that my assumption that the dogbone (not the alumina) have an effective emissivity around 90%+

            that is a hell. 8(

          • Obvious

            The Lugano fins seems quite sloppy, so perhaps they aren’t so very effective at increasing the apparent emissivity as they could be, as in a commercial or lab test device. But they may be required if the device isn’t perfectly smooth and black. If the device was black and finned, it may overheat from its own radiated heat, nullifying the potential convective increases. If a device of this sort is loaded into a metal cylinder, convection cooling simply isn’t going to happen. Again something deceptively simple looking is deeper than the simple appearance first suggests.

          • Obvious

            Just a guess, but black v groove valley bottoms and peaks, with reflective sides, might maximize the potential emissivity.

          • yes, in fact if you accept that transmissivity for the whole dogbone is negligible, then transmissivity is 1-emissivity.
            the fins reduce the reflectivity, because multiple reflections absorb more radiations…
            probably diffusion amplify the effect…

            question is how much?

          • From reading various advices about emissivity from camera manufacturer, they always say you have to calibrate, because even when you have table with “clean copper” “oxidized copper”, in fact the results can be very different

          • Obvious

            I am working out an idea for on-the-fly calibration. See my reply to Andreas, above.

      • The transparency issue is important. It invalidates the method of calibrating temperature values read by the camera with a thermocouple attached on the surface, and it makes the calculations with Stefan Boltzmann Law complicated.
        However, you could also see it this way: A thermal camera basically measures radiated power and translates it into temperature through an algorithm and an input emissivity value. Calculating radiated power from that temperature value and the same emissivity value should in principle make the calculated radiated power independent of emissivity, and also of possible transparency (with marginal errors, depending on the algorithm used in the camera).

        • yes, or at least it introduce a negative bias, as more energy is radiated than what the temperature says…
          basically the method , if it is confirmed, is simply intractable.

          the good point is at least that transparency is not a problem for the IR cam since it is opaque in the bandwidth.

          what is possible is that the whole dogbone have a very high emissivity near 0.9, far from the 0.7-0.4, while alumina have a low emissivity 0.7-0.4, because it is transparent.

          I remember a law: transparency+emissivity+reflectivity=1 (is it right?)

          in fact for the whole dogbone, it is not transparent. reflectivity is mostly the alumina, so , emissivity is simply 1 minus reflectivity of alumina…

          the inside of the dogbone is hotter than the outside, so estimating it is radiating like the outside alumina is minoring the radiated power.

          anyway i’m not competent, just interpreting what I read.

        • Anon2012_2014


          I agree with you. What matters is not the temperature, but the energy radiated.

          If we imagine instead of a single sensor say .5 cm in diameter, that the device is enclosed in a sphere that is able to absorb radiated energy, we need to estimate how much energy is passing through the sphere.

          By having a .5 cm disk at a known distance away and knowing the energy in the pass spectrum of the sensor, we can extrapolate out the remaining parts of the spectrum that we are missing essentially from the shape of the energy spectrum (assuming grey body), and use that for the integrated energy produced. All independent of emissivity — all that matters is the energy received by the sensor, and then the factor to multiply that by to get the sphere, and then the spectrum correction factor (based on the peak fitted to a standard grey body energy vs wavelength function).

          • right, and what you describe looks like a bolometer ..
            is it that ?

  • Dr. Mike

    Thanks for the excellent explanation of the alumina emissivity measurements in the Lugano report and their effects on the calculated power outputs and temperatures. I believe that this error would have been discovered by the Lugano authors if they had calibrated the reactor with a control run up to a temperature that was close to the expected active run temperature. Also, it appears the reactor was still operating at the end of the experiment. The reactor could have been vented to release the hydrogen, then a full calibration curve could have been run at the end of the active run in an unsealed reactor.
    It certainly didn’t seem reasonable that the Lugano reactor was operating for 20 days at an external temperature that was so close to the melting temperature of Ni.
    Dr. Mike

  • Don’t take these kinds of articles seriously. “Experts” will disagree on any test procedure and claim they know things that others do not know. If you test with a camera, they say test with water instead. If you test with water, they will say test with a camera. The real test is the economic test at the factory. Does the E-Cat save them money? Has the E-Cat become too expensive in the current configuration with so many computers? That is the test to watch.

    • Omega Z

      The boilers & such will be part of the cost of any system.
      However, the E-cats & there controls will obviously cost more then gas manifolds & control valves. The question is how much the fuel savings will offset the cost difference & how quickly.

      If it takes 3-5 years to cover the cost differential & is good for 20 years plus, It’s an easy call. You go with E-cats.

  • builditnow

    Thoughts on an air heated and cooled reactor test setup.
    See diagram attached.
    The basic idea is to use hot air to heat the reactor and then cooler air to extract the heat one the reactor become exothermic. Burning Gas could provide the heat. Magnetic stimulation could be provided by insulated windings designed to optimally provide magnetic pulses efficiently. If the reactor becomes exothermic, the only added heat would be from the magnetic pulses and the circulating stirrer fans. This could be very small and the system could run with no added heat while producing plenty of heat.

    Once exothermic, the COP would be very high and obvious.

    • Builditnow

      Frank, my diagram keeps disappearing, any reason for this?

      • ecatworld

        I’m sorry, but I don’t know the reason. I don’t see any image with your posts — did you do an upload, or use a link?

  • Robert Ellefson

    I think I’m developing an allergy to calorimetry.

    • Bob Greenyer

      Much better to get it running a long time and then see isotopic shifts, much more cut and dried – but that can’t be done at home.

      • Pekka Janhunen

        Heat after death (like Parkhomov had, although inadvertently) is also good because it’s fast to do and not dependent on calorimetric input-output balance calculations.

        • Bob Greenyer

          That too, but that has been challenged. Isotopic shifts can only mean one thing – and if the net energy balance of the in and out supports a positive energy output during operation, then we are in a strong position to argue anomalous heat.

          • Andreas Moraitis

            That was my first thought when I saw the report. But changes in the isotopic composition can also result from fractionation. Remember the discussion about fluorescent light bulbs we had some months ago. In order to minimize the risk of misinterpretation it would at least be necessary to analyze both the powder and the condensed material from the inner reactor walls.

          • Bob Greenyer

            Refining isotopes like this, if true, would be valuable in its own right.

          • Fyodor

            Speaking of which, the other Bob said that you would be doing the next set of Parkhomov replication attempts. Can you give us any sense of when that is planned for?

            Thanks for all your hard work.

          • Bob Greenyer

            Bob Higgins will be running experiments that should add really good detail on the pressure – time – temperature – heating relationships, this will give much needed information on what is going on inside these kind of systems now that the “Bang!” reactors SEM and EDX revealed that we are in a new kind of reaction matrix (An ionic Li-Al-Ni-H on solid Ni in an H2 environment)

    • Eyedoc

      Yes, I agree, constant overexposure can do that ,<)

  • I should comment that the actual COP has not been adequately re-calculated yet. It may be in the range of 1.9-2.4 when all of the contributions to the COP have been revised for the lower measured temperature. For example, the convection calculations must also be revised, not just the radiant power (although at those temperatures the radiant power dominates).

    The intent is to extract the maximum useful value from the Lugano experiment – not to point fingers or criticize. By helping to extract the true values, we all add to the experiment’s value. We should consider it no other way.

    • Sanjeev

      What is your analysis of the dummy run with corrected emissivity?
      Why did they get a COP=1 for the dummy, when the incorrect temperature shown by the camera should have lead to a COP=2 roughly ?
      I’m puzzled by this, does the error cancel out in case of the dummy?

    • Tom H

      What surface area would you use to calculate the radiant energy from the wavelengths that are transparent to the alumina reactor shell (ie; the heated wire or the area of the outer shell)?

    • Obvious

      The Idea that the spectral emissivity epsilon is the correct one to use is incorrect. The camera has a built in grey body algorithm that is used to calculate temperature. The normal total emissivity is to be used. Calibration is needed to adjust the internal grey body calculations to match the actual material tested. The spectral peak is used to match the camera optics to the most emissive portion of the spectra, so that the calculations use the most effective part of the spectra for the internal calculations (IE: avoid a dead spot in the emission spectra). Calibration dots and black body holes ensure that the Epsilon of one or the designed emissivity of the dot is being read properly by the camera. When the emissivity is a selective grey body, a calibration should be done at each temperature step to verify the camera response, however the total normal emissivity of alumina has been tested many times. The transparency of the alumina will however affect errors in the grey body approximation if only the Alumina epsilon is accounted for, and underlying materials can radiate at frequencies that are transparent to the alumina, and in an area that the alumina is a poor emitter, increasing the actual output in an area of spectrum If it is not visible to the camera. The total normal emissivity would be higher in this case, most likely.

  • BTW, I posted a new revision of the paper this morning to correct a typo. I am leaving the old versions in the folder so that the links don’t get stale. Here is the link to the folder:

    In that folder, select the latest version.

  • Andreas Moraitis

    The following method to determine the released energy should be practicable:

    1. Run the reactor in active mode and monitor the radiative output with the Optris.
    2. Remove the fuel and run the reactor again, switching the power stepwise up until the camera shows the same reading as in the active run, or until the heater wires fail. (A failure of the wires could perhaps be prevented by using an auxiliary, internal heating element.)

    Now use the raw data of the camera as a reference. If the maximum was reached, the energy level could be inferred from the electric input in the dummy run. In case of a wire failure one would get at least a reliable minimum. This should work without all the complex calculations regarding emissivity, convection, and so on.

    Did I overlook any pitfalls?

    • Obvious

      Whether the reactants can actually effectively removed from the reactor seems to be an open question. Rossi seems to have been able to do this, but the “replication” attempts seem to show otherwise (coatings on tube walls, sintered fuel, etc.). Perhaps this is one area where the replication attempts are not as well developed.
      The method I have been looking at involves a IR mirror that is alternately facing towards and away from the device. The true emissivity is calculated by comparing the difference between the values obtained when the forced re-absorption of reflected heat is compared to the non-reflected heat emissivity of the heat source. (Greatly summarized version).

  • Obvious

    Obviously these are conceptual replications. The info for exact replication is not currently available. However, if successful conceptual replications occur, then the generality of the effect is tested. If all conceptual replications fail, then at least the experimentalists are familiarized with many of the technical considerations and skills required by the time the exact details that are required in order to exactly replicate the effectdevice become available. The chances that an exact replication will be successful the first time it is attempted will be greatly improved if the experimenter is knowledgeable in the preparation aspects and potential problems that might be encountered. Failed experiments are a great teacher.
    Also exact replications are rare in most sciences. One cannot repeat an ageing study on humans exactly, or repeat geological conditions, or build a solar system, or make a desired type of star, or crash a car quite the same way… Or for that matter, use the same Lugano testers with the same ideas they had at the beginning of the test all over again. Even they wouldn’t test it exactly the same way again, knowing what they know now, if given the chance.

  • Warthog

    He doesn’t even know the heating coil IS inconel. That is totally an assumption on his part based on flawed logic (“…….why would he use resistance wire (as the lead-in wire). He wouldn’t…..”).

    All he or anyone else knows with certainty is that the bare wire leading into the E-cat is inconel, probably chosen due to its stability against oxidation exposed to air, and to make the transition from copper to whatever the heating wire is.

    Kanthal would be a better choice for the actual heating wire, but that is totally speculation on my part. It could be anything……..

  • Warthog

    Balderdash……these are professional scientists, not politicians or journalists.

  • Obvious
    Re: transparency of alumina with inconel

  • Bob Greenyer

    He is waiting on a few key parts and must then assemble and do preliminary tests

  • Dexter

    Plot 5 and 9 in your report have no physical sense. You can’t overlay two normalized plots in that way and there can’t be a body with emissivity higher then the black body that is by definition the perfect emitter and absorber. I presume that there some errors in your reasoning.

  • Dexter

    The gsvit blog page ( not a real report ) has been found full of errors.
    I attach here the original answer from Bert Abbing ( probably a pseudonym of a group working in the field ) just for clarity:
    Dear Dr. Rossi,maybe you are interested about the short email note that we have sent today to Mats Lewan, Frank Akland and daniele Passerini. We can’t stand that a work like yours is denigrated in a blog I have discovered today in the most possible unscientific way. Here is the thread.

    Dear Dr. Lewan I have written this letter to Daniele Passerini and in CC to Frank Akland regarding the horrible blog page by “GSVIT”…/tpr2-calorimetry-of-hot…/
    appeard today. I think that this short considerations could interest also you.
    Dear Daniele, Dear Frank
    I (we) write here some observations about the disgusting blog page you (and Levi) have have indicated to me.
    Those pages (and all the site) are purposely designed to appear as scientific to a layman so to divulge disinformation and ill formed concepts.
    They pretend to be written by an “official institution” that should appear seroius and scientific for the outsider and the common man.
    In fact the “institution” ins NOT a research institute and does non have any contact with internationally recognized research institutions or any University.
    No surprise if they received NO answer from any of the research group.
    The most disgusting thing is that their page appear purposely written in a way, mixing up real information from literature, omitted information and absolutely FALSE statements so to “demonstrate” a (false) conclusion. This demonstrate that the group that have written this pages is far to be scientific but has an agenda and a precise goal.
    Let us review just some of the points:
    The main FALSE information they try to transmit is that when measuring a temperature with a non contact thermometer one should use the “Spectral emissivity” and NOT the total normal emissivity.
    This statement is absolutely WRONG. Due to the fact that detector sensitivity is far to be a flat function and usually differs from pixel to pixel in an IR camera all that information is handled by the internal software of the instrument and to the user is requested ONLY to input the value of the TOTAL NORMAL EMISSIVITY which is “the ratio of the energy radiated by the material at a temperature T and the energy radiated by a black body at the same temperature” over ALL wavelengths. (you can find that in ANY textbook! eg: G. Gaussorgues Infrared Thermography)
    So all the argumentations in the blog page about integrating only in the “measurement window” are ILL-FORMED, wrong and misleading!
    Note also that is just by chance that Allumina has a constant spectral emissivity in the window of sensitivity of the detector. Many other materials have not! And the “spectral emissivity” is NOT available for many materials. This would limit tha use of non contact thermometers just to few special cases, and this is not true! Total Normal Emissivity tables for materials are available from many vendors showing similar values independently from the detector!
    The authors of the also MISS to explain WHY if the AA of the TPR2 would have done such a tremendous error (they have NOT) all the measures done with the DUMMY (uncharged) reactor match the input power! THIS was in fact a calibration and a confirmation that the method was good.
    Another point they MISS to cite is that when they have measured the emissivity of the with the reference dots of the external allumina pipes the have found a values (0.69…. 064) that are in PERFECT agreement with the literature. Note that is even possible for the TPR reader note that the reference dots have a higher emissivity then the pipe because the are much britgher that that.
    Is NOT surprising on the other hand that MFMP have obtained different values. Cement materials, even if they have a high percentage of alumina can have a very different emissivity because the presence of metals (Mg).
    The TPR authors have analyzed the material by X-ray spectroscopy and found that was pure alumina, so they applied correctly the data for that material.
    Is quite WIERD that the blog page authors have found an emissivity near to 1 (in contrast to any emissivity table!) at that low temperature. Or the material was not pure or they have done an error! (bad thermal contact of the PT100 or K probe could eventually lead to that!)
    In conclusion.
    We have found that blog page far to be scientific and probably part of a “disinformation plan”. They make evident theoretical and experimental errors probably on pourpose.
    We will ignore it and go on working.

    • Henry

      “The gsvit blog page ( not a real report ) has been found full of errors”

      Really funny, in this way mr. Bert Abbing attached also Bob Higgins of MFMP because he got quite similar findings of GSVIT.

      Did you think about this aspect or simply copy Bert’s ranting here just to show to be another Rossi’s repeater?

  • Obvious

    Bob, I am almost 95% satisfied that you have this right, after understanding all this better.
    However, after looking at the numbers, I have a couple of questions.
    Have you tried the math using the spectral emissivity range if the IR camera was using the 2-5 range, instead of 7 to 13? Are we certain that the latter range is what the testers used? It seems like the values the testers were using is more compatible with the higher frequency. This is compatible with the authors suggestion that they looked up the range in the Optris reference, where I did not see any LW suggestions for emissivity epsilon for alumina of any kind.

  • Obvious

    Where I got hung up was on the spectral T^nlambda, which was hard to read in the Optris document, and is rare to find in most treatments of black/grey/bodies.

  • Sanjeev

    Both Bob and GSVIT are silent about the dummy run. I guess they cannot explain why the testers got a perfect unity result in dummy run in spite of using wrong emissivity values. This is a big hole in their otherwise excellent analysis.

    So far both have not replied to my questions. Hopefully they will address this issue.

    • good point.

      anyway you cannot ask an attorney generat to consider evidence for the defence. they are not honest they are attorney general in the worst sense. they will hide data if it oppose their goal. however they spot real problems alos.

      the emissivity story revealed that Logano approach is broke on the paper however the calibration at low temperature , plus complex knowledge we gathere may reveal that in fact it work…

      the error we made is that we considered emissivity as one numbern greybody emissivity.

      we considered opacity as perfect, which is true for the IR cam, but false for full range.

      we considered emissivity was as told by the tables, which may even be true on global spectrum .

      in fact what seems possible is

      IR cam see a nearly perfect opaque blackbody (0.9) in the 7.5-13um.

      note that IR cam don’t care of the full emissivity, so no risk of error.

      as Lugano testers assumed 0.7 the 0.4 emissivity temperature is erroneous…

      question is how much?
      this is where the calibration period cause problem… at 450C, emissivity is assumed 0.7, and should be about 0.9 in my hypothesis… there should be 25% error …
      my model for the optris is linear with zero at 220C, so the real temperature would be 400C… it is not what is measured…
      so emissivity was 70%, … that is a big problem… or the optris is not behaving as I imagine.

      now at 900W the emissivity was assumed 40%… was it 70%, or 90%

      good question for the temperature. the eyeball thermography seems to say temperature was more about 1000C than 1450C, but troubled by translucidity

      if emissivity was stable it would be compatible with eyeball thermography.
      however opposite to what skeptic were saying, this is where translucidity increase the hope of real COP.

      IR cam measures the temperature of the alumina, but if the alumina is transparent to the most energetic part of the spectrum, maybe is the radiated energy mostly from the inner core of the reactor, at high temperature, much above 1000C.

      finally the test seems flawed, it may even, and it seems to be very positive, but how much ?

      finally the only reason I trust the E-cat is because Industrial Heat gave freely a prototype and let scientists test it the way they liked…
      This mean that they are convinced it works.

      • Sanjeev

        GSVIT guys censored my comment about the dummy analysis and sent me an anonymous email asking for my ID (real full name) ……….irony !

        Probably they don’t want anyone to spoil their party. I think I will put them into “useful but untrusted” category for now.

  • Thomas Clarke


    The GSVIT team have replied to your comment below about their behavior on their blog page:

    As far as the dummy run goes, there is a simple reason why the radiant power error noted by Bob Higgins and also the GSVIT team would be much less in the control case.

    For the control temperature is much lower (450C) and therefore from the Lugano Report the emissivity value used would be much higher (0.69). So the error (0.69 vs 0.9) is smaller. Also at this lower temperature the dependence of band radiance on temperature is closer to T^4 – which if true would make the error cancel. Finally at this lower temperature the contribution of radiance to total power dissipation is lower compared with convection.

    The Lugano control result was not perfect, as you say, but 446 vs 486, about 10% error. In fact, in addition to the radiant power error highlighted here, there are many other assumptions (such as that the shape of the fins does not matter, and that the reactor is an infinitely long perfect cylinder) in the convective heat calculation. The systematic errors from these assumptions have not been included in the error analysis given in the Lugano paper which is in any case ad hoc and not quantified, because only a few of the (smaller) errors have been quantified.

    Therefore it is I suggest unsafe to take the control results as any indication of radiant power at this lower temperature, and the data both experimental and theoretical from Bob and GSVIT is not contradicted.

    • Sanjeev

      Thank you for your analysis and explanation.

      I hope the more light will be thrown on this issue during the next round of testing by the Levi group, that will (probably) start next month.

    • Obvious

      For any Canadians out there, Canadian Tire has their IR Temperature reader on sale for 20 bucks for the next few days (80% off). It only goes to 480C, but does have an emissivity adjustment. Measures from 8 to 14 um. Not the best quality, but fairly good, and possibly OK for cheap and rough testing some theories for the dummy emissivity if compared to a thermocouple.