Rossi Lugano/early demo's revisited. (technical)

  • The datasheet give a thermal conductivity of 15 BTU-in/hour.F.Ft^2

    Converting that value to SI units gives 2.16 W/mK (Using the engineering toolbox converter)

    However different sources are giving a value of about 35 W/mk at room temperature for Al2O3

    That's a large difference !

    Any clue ?

    The Cotronics data sheet shows 1.2 W/mK for pure alumina

    35 W/mK sounds rather high. That would be equivalent to the thermal conductivity of Pb.

  • The Cotronics data sheet shows 1.2 W/mK for pure alumina

    35 W/mK sounds rather high. That would be equivalent to the thermal conductivity of Pb.



    https://en.wikipedia.org/wiki/Aluminium_oxide gives 30 at room temperature.

    Some references say it is between 28 and 40

    Tabulated values give about 35 at room temperature (see for example http://www-ferp.ucsd.edu/LIB/PROPS/PANOS/al2o3.html )


    Indeed about the seame as lead

  • Lugano dummy run FEM simulation - First result


    Based on a configuration with three heater coils of 10 windings under the ribbed area, I post here the first result of a thermal FEM simulation of the Lugano ECAT.

    In that simulation I assume that halve of the convective and radiated heat of the side of the end caps flows into the rods.

    The result of the simulation can be seen in the following figure.



    As can be seen the center temperatures of the simulation are somewhat higher then the center temperatures of the Lugano dummy run.

    Also the temperature profile is somewhat more parabolic (less flat) then is the case for the temperatures reported in the Lugano report.


    In my opinion the only way to get the profile of the simulation more flat is to increase the thermal conductivity inside the ECAT.

    Increasing the thermal conductivity increases the lateral thermal flow which in turn makes the thermal profile flatter and provides more heat to the end caps.

    Experimenting it was found out that standard alumina, even with a somewhat increased thermal conductivity, will not make the profile more flat.

    However if possibly another material with a much higher thermal conductivity is included in the ECAT, then the profile will be probably even more in agreement with the Lugano profile.

    The following comment in the IH patent indicates that a higher thermally conductive material could have been used inside the ECAT :


    -------Those skilled in the art will understand based on upon the

    -------present disclosure that the attributes of a thermally conductive

    -------material included in a reaction device may vary .....


    One of the highly thermal conductive ceramic materials which could have been used is AlN (Aluminium Nitride) ceramic.

    This material has compared with alumina a density of 3260 instead of 3900 Kg/m^3 and a thermal conductivity of 140 - 180 instead of 35 W/(m.K).

    But since the outside of the ECAT was, as shown by analysis, made up of alumina, in that case an outside alumina layer must have been added to the ECAT.

    Such a layer would also protect the AlN ceramic at high temperatures from degrading.


    Nevertheless the current FEM simulation indicates that the temperatures are already close the Lugano temperatures.


    Next thing to do is to spend some more time on investigating what the influence of included materials with a higher thermal conductivity has on the thermal profile.

  • Lugano active run period 16 recalculation - The ECAT worked - COP was 4.98


    With the FEM model developed we can analyse, using finite element analyses, the thermal behavior of the dogbone shaped ECAT for an active run.

    We are doing this analysis for the last active period 16.

    The electrical power data for this period given by the Lugano report is :


    Total power consumption was 906.31 Watt

    Joule heating was 41.25 Watt


    This leaves 906.31 - 41.25 = 865.06 Watt for heating coil wire.

    Since about 4 cm of heating wires continue in the rods, the total power in the ECAT itself is somewhat less and is calculated as being 735.08 Watt


    The average measured temperatures reported for active period 16 are


    ----------Temperature (C)-------Temperature (K)

    Cap 1----------611.09-------------------884.24

    Cap 2----------595.15-------------------868.30

    Body--------- 1412.31-----------------1685.46


    The report only gives for the active periods only the accumulated powers of the end caps , body and rods, not the seperate ones.

    However using the average temperatures given above we can calculate the approximate powers.

    We must indeed redo these calculations since for the body the Lugano team did not take into account the correct total area of the ribs, the view factor, the emissivity change due to the view factor and the correction needed for the convection of the ribbed surface. Also the temperature measurement of the body by the Optris camera was influenced by the change in emissivity due to the infinite reflection method.

    So the first thing to do is recalculate the average body temperature to the correct one.


    For the mentioned temperature the used emissivity on the Optris is .950

    View factor between the ribs is 0.428

    This changes the alumina in band emissivity to be used on the Optris from .950 to 0.971

    Using both the original and the corrected emissivity we can calulate the correct body temperature.

    The found body temperature is 1389.91 C ( n value 1.518).


    Having found the correct average body temperature we can now determine the emissivity and the convective heat transfer coefficient for the ECAT body


    Emssivity--------------------------------------------0.392

    Convective heat transfer coefficient-----14.542


    And with both values we can calculate the convective and radiated heat power of the body :


    Radiated body energy-----------3453.17 Watt

    Convective body energy----------523.55 Watt (uncorrected)

    Convective body energy----------269.10 Watt (Corrected with factor .514 extrapolated from earlier simulations)

    Total radiated and convected power of the body area is


    3453.17 + 269.10 = 3722.27 Watt


    For both caps we find :


    Cap 1

    ----------Radiated cap energy------101.79 Watt----( e = 0.592 )

    ----------Convective cap energy-----34.05 Watt----( h = 11.481 )


    Cap 2

    -----------Radiated cap energy---------95.36 Watt----( e = 0.597 )

    -----------Convective cap energy------33.06 Watt----( h = 11.457 )


    The total thermal power comming from the ECAT after recalculation becomes then


    3722.27 + 101.79 + 34.05 + 95.36 + 33.06 = 3986.53 Watt


    This total power is much higher then the 2886.18 Watt reported and this increase is due to the recalculation of the power of the body area.

    For the rods the Lugano report states a value of 88.47 watt due to radiation and 87.94 watt due to convection. The convection is overestimated since the testers used a correction of .667 instead of .561. Thus the convection shoud have been 87.94 x (.561/.667) = 74.00 Watt

    Total rod power for one set of rods then becomes 88.47 +74.00 = 162.47 Watt

    For two sets of rods the total power becomes 2 x 162.47 = 324.94 watt


    Total power for both the rods and the ECAT then becomes 3986.53 + 324.94 = 4311.47 Watt

    The new calculated power leads to a COP of 4311.47/865.06 = 4.98 , higher then the COP of 3.74 mentioned in the report.


    By now assigning the calculated ECAT power of 3986.53 Watt to the heating element of our FEM model, we can with the model calculate the approximate internal and surface temperatures and compare them with the data in the Lugano report. The results are :


    ---------------------------------------------Lugano---------FEM simulation----------Difference (%)

    Setpoint temperature---------------1400----------------1436----------------------- 2.6

    Average body temperature--------1412----------------1367----------------------- -3.3


    Note that the actual average body surface temperature found by the FEM simulation was 1339 C

    However for comparision with the Lugano report we must correct this temperature due to the misreading of the Optris as defined by the infinite reflection method.

    The corrected temperature is 1367 degree C (n factor 1.570) and this is the temperature reported in the table above.


    The setpoint temperature is the temperature in the inner core of the ECAT.

    In the Lugano tests this setpoint temperature is measured by a thermocouple and by adjusting the power the testers where able to arrive at the required temperature setting.

    Note that the setpoint temperature measured from the FEM data is close to the reported setpoint value of 1400 degree C the Lugano testers used for active period 16. Also the reported body temperature is close to the simulated temperature.


    Note that the above calculations are approximations since they where based on average temperatures reported.

    Despite using the average values in the above calculations my conclusion from the close agreement between reported data and the FEM simulation data is that the ECAT indeed produced excess energy with a COP of about 4.98 during period 16 of the Lugano report.

  • LDM ,

    The over-reported temperature using the wrong IR camera emissivity (0.392, or whatever) must be corrected to the correct IR camera emissivity temperature before using the correct total emissivity to calculate radiant power with the correct temperature rather than the exaggerated temperature.

  • Paradigmnoia


    I disagree about wrong emssivities for the following reasons :


    1. Dummy run


    The calculations and simulations of the dummy run showed that it is very unlikely that wrong emssivity settings where used


    2. Real surface temperature if wrong emissivities where used


    Using the .971 Optris value and a broad band emssivity of 0.393 for the 1412 degree C reported, gives a non inflated temperature of abouut 800 degree C (n = 1.895)

    Doing a FEM simulation with an ECAT power of 735.08 Watt gives an average temperature of the ribbed area of 628 degree C

    These temperatures are not even close.

    Thus for the active run period 16 the power does not match the non inflated temperature.


    3. Core temperature


    The average core temperature from the simulation is 609 degree

    (That this value is lower then the average surface area is due to the fact that the core continues under the end caps with their lower temperature)

    Since this temperature of 609 is much lower then the setpoint temperature of 1400 degree, the testers would have seen on the thermocouple display that they could not reach this temperature.



    About that a set point of 1400 C internally, if achieved, would melt the thermocouple .

    It depends on the type of thermocouple used.

    For such high temperatures I would use a type R or S thermocouple or even a B type.

    The B type has a gray color coding and in the thermocoupls wire in Lugano seems to have about that color.


    That the heater coil wirewould would melt at 1400 degree is not true.

    Kantal has a melting point of 1500 degree C and one of the companies i worked for has been running Kantal elements just over 1350 degree without problems


  • 1. At the Dummy temperatures the emissivity changes have a minor effect. This was clear on my graph. There is enough wiggle room that the temperature-power error can be fudged away. Regardless, the Professors wrote numerous comments about how they adjusted the emissivity settings and re-iterated away. They explicitly explain the flawed method, and include examples.


    2. Perhaps your simulation is still flawed, as good as it might be. I cannot evaluate that.


    3 Core temperature where? Anyways, Rossi always uses Type K. They are cheap. Note that the mini connector plug is coded yellow for type K, as is the extension wire. (see below). The thermocouple shown may have a stainless braided sleeve making it look grey (note the reflection in the 2nd thermocouple image below, and the Professors' comment below Figure 2.

    If the outside temperature is anywhere near 1400 C, then the heater wires are hotter than that. They might survive a while at 1400 C, but not 20 days.

  • I have probably enough Durapot to make a Cap.

    Maybe cast a nub of rib section on one side...

    How deep shall I install the thermocouple? Maybe make it moveable?

    I can easily wind up an appropriate heat input arrangement, and a bit extra.

    I won’t be able to use 15 ga Kanthal wire, braided, though. To big. I can match the watts range however.

  • How deep shall I install the thermocouple?


    If you are using one of the low-cost stainless-jacketed K-types you will need to put it into a 'thermowell' like the one shown below to protect it from oxidation if regularly run at more than 6-700C. If you can cast a tight-fitting hole for it into your test-rig that would also help, since an open tube furnace is a highly oxidising environment. Worth also mentioning that it is the junction at the tip of the thermocouple that produces the temperature signal, so that decides how deep it needs to be - too shallow and the outer jacket will conduct heat away from the junction and give you a false (low) reading.



  • Paradigmnoia


    1. At the Dummy temperatures the emissivity changes have a minor effect. This was clear on my graph. There is enough wiggle room that the temperature-power error can be fudged away. Regardless, the Professors wrote numerous comments about how they adjusted the emissivity settings and re-iterated away. They explicitly explain the flawed method, and include examples.


    The problem is that indeed the changes in emissivity are small at the dummy run temperatures, but also convection is involved.

    And at lower temperatures convection plays also a major part.

    From the dummy recalc we see a difference of about 10% between the inflated and non inflated calculation. While not very large not small either.


    Concerning the method the testers used, I already proposed an other method which also closely followed what was written in the Lugano report.

    Look also at the references 4 and 5 the testers referred to which are about emssivity and temperature seperation used for example in earth survey. These algorithms are also based on the total amount of radiation received which was also the basis for the alternative method the testers might have used and which I proposed.


    2. Perhaps your simulation is still flawed, as good as it might be. I cannot evaluate that.


    That could be the case. And it is as good as the guy who makes the model and does the simulations.

    So what can I say ?

    However my simulation of your round rod gave a temperature within a few degrees of what you measured. So it can't be that bad.

    And what would be the chance that with an incorrect model all the simulations are giving about the same results as what is reported in the Lugano report?


    3 Core temperature where? Anyways, Rossi always uses Type K. They are cheap. Note that the mini connector plug is coded yellow for type K, as is the extension wire. (see below). The thermocouple shown may have a stainless braided sleeve making it look grey (note the reflection in the 2nd thermocouple image below, and the Professors' comment below Figure 2.


    I agree from the info you provided that they indeed probably where K type


    If the outside temperature is anywhere near 1400 C, then the heater wires are hotter than that. They might survive a while at 1400 C, but not 20 days


    In my opinion Kantal wire will survive much longer if they, as is the case in the Lugano ECAT, are embedded in ceramic which prevents further oxidation by outside oxygen at high temperatures.


    I have probably enough Durapot to make a Cap.

    Maybe cast a nub of rib section on one side...


    From all simulations I did with different configurations the average temperature of the ribbed area seems to be the most constant factor.

    That is the reason why in my last analysis I used the temperature of the ribbed area

    The heat distribution at the caps is influenced by the rods and more difficult to forecast.

    Also the radiation of the caps is interacting with the radiation of the ribs, and this interaction is greater where the ribs end in the end-caps.

    So in my opinion a cap with a short stub of ribs will not be representative.

    I think the only way to get representative results is to have a full sized ECAT model.


    If you are short of Durapot for a full sized ECAT, you might consider using standard ceramic tubes and only cast the outher layer and ribs


    An other approach would be not to us the ribbed area, but based on the expected temperature to calculate based on view factor, effective area and expected convection a representative tube diameter. (if that is possible) Otherwise stated, use a tube, wind the heater coil and cast the outher layer.


    How deep shall I install the thermocouple? Maybe make it moveable?


    I would certainly try to make it moveable. What I see from the simulations is that under the end caps the core temperature is significant lower. (676 C in the middle of the core, 422 degree 2 cm in the end cap, only heater coil power applied)

    Starting with a standard rod with a smaller plug in tube at the end with an inner diameter a little bit larger then the thermocouple sheet diameter will allow you to move the thermocouple.


    I can easily wind up an appropriate heat input arrangement, and a bit extra.

    I won’t be able to use 15 ga Kanthal wire, braided, though. To big. I can match the watts range however.


    From the simulations it seems that the mass of the kantal wire has no measureable effect.

    So you can wind it from any Kantal size you like. However as you already explained, the heat distribution is important, so you have to make sure the heat distribution in the body and in the end caps are correct.

  • They explicitly explain the flawed method, and include examples.


    If they used wrong emssivities, in my opinion they would afterwards have considered this to be the most importantant shortcomming in their measurements.

    However I was told by somebody with good contacts with the Lugano team what they considered to have been their major shortcomming.

    And this major shortcomming was not that they had used wrong emissivities on the Optris, but something else which I considered to be less important.

    It made me conclude that their method of temperature measurement must have been correct.


    Sorry that I can't give more details, but I don't want to compromise my source of information

  • LDM,


    Several months after the Lugano report was released, Lewan rumored the team were sending out the alumina for a full E analysis. That was in response to the controversy surrounding the issue, and the move as intended to resolve that. Never got a follow up on whether or not they carried through with it. and if so, the results.


    Good to hear you have a source that may help shed a little light on it.



  • LDM


    evidence against this:


    "wrong emissivity methods" accurately explains claimed Lugano COP and mysterious acceleration in COP noted by experimenters and used by them as confirmation of an exothermic reaction.


    Lugano-style reactors, with Rossi fuel, were used by IH and extensively tested. Using the Lugano power measurement methodology a dummy reactor showed high COP, identical to the active reactors.


    That is pretty damning.

  • Good to hear you have a source that may help shed a little light on it.


    The information above was told to me without me asking for it

    I am not asking my source for further information, but can confirm from what was told to me that the E analysis of alumina was not the number one on the list either.

  • LDM,


    Several months after the Lugano report was released, Lewan rumored the team were sending out the alumina for a full E analysis. That was in response to the controversy surrounding the issue, and the move as intended to resolve that. Never got a follow up on whether or not they carried through with it. and if so, the results.


    Good to hear you have a source that may help shed a little light on it.

    I know that ”They” did emissivity testing with alumina, because someone directly involved with Lugano told me that my results were consistent with theirs.