Experimental Evidence on Rossi Devices

    Re python.


    On windows download the standard distribution of 2.7.


    comment out the references to numpy and matplotlib (if you want to plot nice graphs, good for visualisation, matplotlib is the only decent tool but it has annoyingly large dependencies with numpy:

    Code
    #import scipy.interpolate as sp
#import numpy
#import matplotlib.pyplot as plt


    comment out the section that does the plotting:


    (sorry for the BB introduced heading - somone does not like programmers!)


    Finally comment out the plot function from main - it still does some decent text printout:


    Code
    #interp1(alumina_band_e_pts)


    All will then be good. If you look at the details you will see that for second order terms - like the contribution of convection - I have some hacky adjustments. the point being I don't have the raw data, but I can take the quoted report powers and work out a best guess adjustment ratio based on the temperature change and and the theoretical power dependence on T (e.g. T^1.6 or whatever).


    For the active tests radiant power at higher temperatures is so dominant second order terms don't matter (I still do my best to estimate, but much higher errors can be tolerated). For the dummy test this is less true - but in any case this calculation for dummy is not very relevant - we know it was fudged by the testers.


    Interestingly - consider the quite large error they notice after fudging. They had to adjust emissivity from book value, and so would use the adjusted (total) emissivity in the power calculation. The temperature must be correct because independently measured in this case. But the total emissivity would be adjusted affecting power out. Maybe you'd get a better fit if you used the correct (non-adjusted) emissivity? Anyway that is really not of much interest, and especially at low temperatures where there are questions about the convective loss which is much more significant, their calculations are not reliable.


    Best wishes, Tom

    Encore, http://www.elforsk.se/Global/O…er/LuganoReportSubmit.pdf


    Executive Summary

    This is an alternative evaluation of the Lugano E-Cat test allowing for an explanation based on classic physics.
    The argument below rests on the assumption that the dummy test was performed correctly.


    Using only data from the report we calculate that for the main test the true electric input was around 2960W, implying that the COP was 0.96.


    Marcia Longa

    The electric power that was fed to the three heating elements internal to the so called reactor was delivered from a three-phase SCR power controller from Compact Fusion that was connected to a 400V AC outlet. Note: No transformer was used.


    The following entities are defined in the Lugano report.


    R1: Resistance of each of the 3 wires collectively called C1 (Figure 4, page 14)
    R2: Resistance of each of the 6 wires denoted C2
    I1: AC current through each wire in the set C1
    I2: Half of I1


    The heat loss in the wiring during the dummy test is calculated in the report:
    Wtot.dummy = 3*R1*I12 + 6*R2*I22 = I12*(3*R1 + 1.5*R2)


    Data from the main test is given in table 7.


    The current I1 for the main test is not to be found in the table, but we can calculate it inverting the formula for W entering an average “Joule heating” from data in table 7 of 41.7 W.


    The unreported current I1 for the main test is thus evaluated to be 49.0A.
    This agrees well with “the current through the resistor coils, normally 40-50 Amps” on page 3.


    For unregulated three phase AC we now use the standard formula: Power = Voltage * Current * 1.732 (square root of 3)


    For the dummy test the current I1 is reported to be 19.7A. If this current had been supplied at full throttle from a 400 V outlet the power would have been 13.6kW instead of 0.5kW.


    From this we can deduce that the SCR was working with a very small conduction angle supplying just short current pulses to the heating elements. This means that at any time only one of the three heating elements was powered and that makes it easy to understand what is going on.


    In Figure 4, look at the middle C1 wire. During one AC period the current goes up and down once to the left resistor and the same for the right (but not at the same time).


    The effective current I1 is defined as the square root of [(time integral over one AC period of current^2) / (time interval)].


    The same formula is valid for I2 as well, you cannot just divide I1 by 2 to get it.


    Since all the time i1 = i2left + i2right (lower case i -> momentary values) we realize that the I2 integrals each equal half the integral giving I1. So I2 will be I1 divided by the square root of 2 instead of just 2 as is done in the report.


    This error is of no consequence for our calculation of the current in the second test. All we did was to use the given output value to retrieve the unknown input. But we can use the corrected currents I2 to get a better value for the ohmic losses, and they will be 8.4W and 51.8W respectively, not needed below though.


    The simplification of non-overlapping currents is valid also for the main test. The conduction angle needed to get around 3 kW is only 16 degrees and overlap starts first at 60 degrees and a power of 30 kW.


    So now we can look at a simple resistor circuit connected to a power supply:


    SCR o ---- R1 ---- R2 --------- R0 ---------- R2 ---- R1 ---- o SCR
    - - - - - - - - - - - - - - -| - - - - - - - - -|
    - - - - - - - - - - - - - - - ---R0 ---- R0 ---


    The corrected I2 current for the dummy test is 13.9A. Let R3 be the resistance of the three heating elements connected as shown above, each with the resistance R0.


    R3 = 1/(1/R0 + 1/(2*R0)) = 2/3*R0


    The total resistance in the chain is Rtot = 2*(R1 + R2) + R3


    Supplied electric power to any one of the three circuits is Pin/3 = Rtot * I22


    For the dummy test, Pin = 486W giving


    R3 = Rtot - 2*(R1 + R2) = Pin/3/I22 - 2*(R1 + R2) = 0.821Ohm


    R0 = 1.23Ohm


    For the main test with I1 = 49.0A we get Pin= 3*R3*49.02/2 = 2959W


    For an average measured heat output of 2840W this gives a COP of 0.96 whereas we got 0.93 for the dummy test.

    H-G - re these two explanations. I would not put it past Rossi to try for both, but since both account for the X3 we must choose one.


    In both cases there is an apparent inconsistency. However the joule power inconsistency can be explained easily and completely by the setup changing connectivity (Wye versus Delta) between the dummy and active tests, or maybe a few other setup changes. The emissivity error has no other explanation and you are left with a COP of around 1/3!


    it is quite plausible the experiment setup changes - and it should make no difference to power measurements given that clamps are not reversed or anything completely wrong like that!


    I actually note another error which was not significant here but would maybe have become significant had they not made the emissivity error and turned up power much higher.


    In that case, it is possible the current clamps would have saturated (but see below) so giving a spurious higher COP given the higher currents required for the lower phase to phase resistance Delta configuration used in the active tests.


    Now - one reason I'm not confident about the electrical analysis (maybe you have more info and so can be) is this. We don't know enough about how the black box is configured to control power:
    Is it phase conduction duty cycle control, or voltage control (by altering phase angle) or both? How it goes depends very much on which part of the waveform is used.
    Is it PID control (with fixed waveform) or one of the above?
    Does the control switch on all phases simultaneously, or just two phases at any one time, or (applies only to Wye) just one phase at a time?


    For me there are too many unknowns: but there are plausible ways both to get X3 and to get correct power (with possible errors due to clamp saturation at high crest factors).


    Best wishes, Tom

    Thomas, I can’t see how the Delta – Wye theory could help explain anything. I do not think that the PCE 830 instrument cares if the configuration is one or the other. A delta can be replaced with a wye with other resistor values and if done in a black box nobody will know the difference.


    I propose that my explanation model has a considerable advantage compared to pure speculations that lack hard support. This is because it rests on data from the report and gives a COP indistinguishable from what the null hypothesis predicts. This could of course be a strike of luck given the uncertainty of the … let us call it calorimetry, for lack of a more fitting word.


    Speculations regarding e. g. the emissivity or transmittance of alumina are not possible to quantify as I think you admitted a few posts ago and I do not think it is possible to give an accurate estimate of what size the errors could be.


    To be honest, I was not the first to observe this gaping hole in the data, but I have done my best trying to clarify its significance and spread the word.

    H-G,


    I like your analysis, and it is most certainly good for me as a person to work through the details. I need a little bit more time to grasp the details. However, I will already now ask one question, and that is about the Table 7. I would think that the "Consumption" column is a measured quantity by the PCE-830 and can thus be trusted (with the risk of being called "wildly gullible"). Would it then not be more natural to assume that Joule heating has been wrongly calculated insad of the other way around?


    I went through Thomas' analysis and if you boil it down there are simple and good reasons to believe that the reactor was colder than stated in the report. The short version is like this (from memory):


    1. The optris camera is sensitive between 7 and 14 micron.
    2. The emissivity of alumina in that wavelength range is close to 1 instead of 0.4 used in the report
    3. Adjusting for the emissivity gives a much cooler reactor.

    Hej Urban,


    Whatever the root cause for the self contradictory data is, we can say one thing for certain: the Lugano report cannot be taken as a proof that Rossi’s technology works. If anything, it is yet another indication that it still does not and the eternal silence from the test group does not help Rossi’s case either.


    But I find little reason to suspect that the joule heating should be miscalculated. The PCE 830 logs the currents. The expected way to arrive at the wire loss data would be to enter these data into a spread sheet using formulae that the test group would have developed for the dummy test. What could possibly go wrong, as they say just before the crash.

    One item which may require some consideration for the electrical version of a COP of roughly 1 is, is it possible to make ceramic tube/ heater assembly of the Lugano device dimensions survive nearly 3 kW of electrical input through normally available (not special magic wire) resistance heater wire?
    The area of the wire could be worked out, over a range of diameters, and 3kW divided by it to arrive at a range of W/mm^2, which could be compared to existing wire specifications.
    I have not yet seen any replication attempt get even close to 1500W input, let alone double that. By the time around 1000 W is being fed in, the tubes are well over 1000 C in every experiment I have seen, or performed.

    Quote from Paradigmnoia

    I have not yet seen any replication attempt get even close to 1500W input, let alone double that. By the time around 1000 W is being fed in, the tubes are well over 1000 C in every experiment I have seen, or performed.


    If memory serves me, the detail about Inconel wire forces a choice: either the inside of the E-Cat did not go beyond ~ 1100 C, or something other than Inconel wire was used. Perhaps someone who knows more about Inconel can comment on the allowable range.


    I agree with H-G Branzell when he says that the Lugano report cannot be taken as a proof that Rossi's technology works.

    Thomas wrote:


    Quote

    Does the control switch on all phases simultaneously, or just two phases at any one time, or (applies only to Wye) just one phase at a time?


    Indeed, the control box (which is a black box) outputs three voltages V1, V2 and V3. For example, it may choose to only use two out of the three outputs, leaving the third one floating, causing only one out of the three coils to be used, and reducing the ratio of the wiring resistance to the heating resistance. It can do that and cycle between the three coils, for some thermal reason, or some weird electrical reason (i.e. if Rossi believes that this causes some specific electromagnetic field.) Or one of the coils may have failed.


    Thus the Joule heating anomaly, while interesting, does not allow us to reach any conclusions.

    Quote

    Thomas, I can’t see how the Delta – Wye theory could help explain anything. I do not think that the PCE 830 instrument cares if the configuration is one or the other. A delta can be replaced with a wye with other resistor values and if done in a black box nobody will know the difference.


    The observed anomaly to be explained was apparent resistance change of the heating element, as determined by comparison of Joule heating and element power values. Either one reversed clamp, or a delta/wye config change, will explain this anomaly.


    Quote

    I propose that my explanation model has a considerable advantage compared to pure speculations that lack hard support. This is because it rests on data from the report and gives a COP indistinguishable from what the null hypothesis predicts.


    My explanation does exactly that, with no need for anything unusual. Your explanation does not do that unless you make the same calculation error that the testers did.

    Quote


    This could of course be a strike of luck given the uncertainty of the … let us call it calorimetry, for lack of a more fitting word.


    There is not much uncertainty in the thermography. At IR wavelength alumina always has a high emissivity with little variation. Therefore the results obtained, correctly processed, can determine the temperature of the alumina to within a relatively small tolerance. There is more uncetrainty working out the power out from the temperature I agree.


    There is no great merit in reconciling the active and dummy results because the testers admit that they fudged (sorry, "altered" - to use their terminology ) the dummy calculations.


    Quote


    Speculations regarding e. g. the emissivity or transmittance of alumina are not possible to quantify as I think you admitted a few posts ago and I do not think it is possible to give an accurate estimate of what size the errors could be.To be honest, I was not the first to observe this gaping hole in the data, but I have done my best trying to clarify its significance and spread the word.


    It is always difficult to get accurate error estimates but also always possible to give error bounds. Since when the calculations are done correctly there is no anomaly to within even very agressive error bounds there is no room for some other hypothetical mechanism that changes the results. The only motive for such a mechanism is to fit results.


    Tom

    Quote

    I agree with H-G Branzell when he says that the Lugano report cannot be taken as a proof that Rossi's technology works.


    It cannot be taken as evidence that Rossi's technology works.


    It is strong suggestive evidence, though not proof (how could one ever get that?) that it does not work.

    Quote

    One item which may require some consideration for the electrical version of a COP of roughly 1 is, is it possible to make ceramic tube/ heater assembly of the Lugano device dimensions survive nearly 3 kW of electrical input through normally available (not special magic wire) resistance heater wire?The area of the wire could be worked out, over a range of diameters, and 3kW divided by it to arrive at a range of W/mm^2, which could be compared to existing wire specifications.I have not yet seen any replication attempt get even close to 1500W input, let alone double that. By the time around 1000 W is being fed in, the tubes are well over 1000 C in every experiment I have seen, or performed.


    Are you comparing like with like? You need ridges, and the correct dimensions.


    To first order power dissipated at given temperature scales as reactor visible surface area (ridges don't increase this but do increase the effective emissivity so allow maybe 20% more power for given temperature). So double the diameter you double the power.


    Lugano was about 900C for input of about 860W.

    Branzell wrote:


    Quote


    No, Antoine the box is not black, it is red:


    I meant black box in the sense of an unknown device (see https://en.wikipedia.org/wiki/Black_box).


    You are saying it's a Fusion Power device, that may be true but it's not stated in the report. And even if that is determined to be true, assuming the device was unmodified, we don't know how flexible that device can be in the way it drives its output: to what extent can it be programmed?


    Quote


    About nearly 3 kW from the dogbone: my gut feeling is that it would glow much more than in the pictures that we have seen from Lugano.


    Do we know if the three pictures showing the reactor glowing were taken during the dummy run, or the active runs?

    What I mean is I would be impressed if someone could build a device the size of the Lugano device and power it with 3kW, with 1.23 ohms resistance, and get any sort of life out of it. The resistance limits the wire choices. The temperature limits the alloys. The size, shape, and composition limits heat rejection of the tube body.
    Makng such a thing run for 28 days would be quite the challenge. Making one last for 24 hours would be quite the accomplishment, in my opinion.

    • Official Post

    At the :56 minute mark of Lewans Webinar the other day he spoke of Lugano. He said: "measurements on methods not calibrated properly" or something like that. Lewan knows Essen (one of the testers) well... having introduced him to Rossi, so Lugano is officially dead it seems.


    However! there are always "howevers" in LENR LOLs, Lewan did say the fuel ash composition would: "be extremely difficult to copy". And we all know that if the fuel/fuel ash transformation stands, it is nuclear.


    So hear ye all believers; we have our new Lugano talking points...it's all about the ash! Oh, and don't forget that Ecat X too! :)

    Quote from Antoine10FF

    Do we know if the three pictures showing the reactor glowing were taken during the dummy run, or the active runs?


    Indeed. Do we know when the pictures were taken? Were they taken when the E-Cat was at full power? Were they taken when it was still heating up? Did the camera that was used capture the color in a way that a human would see it, or did it compensate for the brightness? Many unknowns. Without answers to these and similar questions, I draw zero conclusions from the color temperature shown in the pictures in the Lugano report.

    The electrical anomaly does not prove anything, although it's interesting. The reason it does not prove anything is that we do not have recordings of the waveforms coming out of the control box and we do not have statements as to how they changed between the dummy run and the active run. And thus it is possible to assume that the waveforms would have changed between these two. For example, the controller may be using only one heater at a time for some reason. It's not unreasonable to have that kind of behaviour, even if it's not optimal or even if it does not make sense, as it's the kind of thing an engineer would try if they believed it's helpful. Another thing is that maybe one of the coils failed. We don't know, and there are no claims regarding the stability or the nature of the waveforms. In the absence of data, It does not imply that the device is producing excess energy or not.

    Quote from H-G Branzell

    About nearly 3 kW from the dogbone: my gut feeling is that it would glow much more than in the pictures that we have seen from Lugano. But I think we have better stop racking our brains over the Lugano report.


    I thought that too so I did some number crunching.


    Below are luminous and visible radiant flux bounds for a reactor-shaped and sized blackbody at different temperatures. These were obtained by integrating the blackbody spectral exitances multiplied by the cylinder area for l=200 mm and Ø=20 mm over the 400-700 nm range and against the CIE 1924 photopic spectral luminous efficiency function.


    These assume unit emissivity in the visible and neglect any absorption by the alumina or shadowing by coils, etc. and are therefore upper bounds.





    It seems that, unless there is a mistake in these plots (which should be checked) that the reactor shouldn't be glowing any stronger than a 60 W ordinary incandescent bulb at the maximum temperature, and would be quite a bit dimmer at 1260°C.


    Quote from Eric Walker

    Indeed. Do we know when the pictures were taken? Were they taken when the E-Cat was at full power?


    The plots seem to indicate that no light would be discernible at the dummy temperature of 450°C, especially not against the room illumination. We can therefore assume that they are in the active phase at 1260 to 1412 °C in all the lit glowing images.

    Quote

    Do we know if the three pictures showing the reactor glowing were taken during the dummy run, or the active runs?


    Antoine - are you arguing that reactor was outputting 3kW? It is not possible, given the low temperature.

Subscribe to our newsletter

It's sent once a month, you can unsubscribe at anytime!

View archive of previous newsletters

* indicates required

Your email address will be used to send you email newsletters only. See our Privacy Policy for more information.

Our Partners

Supporting researchers for over 20 years
Want to Advertise or Sponsor LENR Forum?
CLICK HERE to contact us.