Experimental Evidence on Rossi Devices

There are uncertainties in the calculations due to alumina transparency. The problem being that we know neither the effective optical surface area nor the effective emissivity of the underlying structures that would then emit radiation. Probably the heater would be this structure.

Still, this affects only part of the emitted power, and the alumina is only partly transparent. But it is one of the many reasons (documented here) why this experiment has significant unknowable errors.

You have to ask how could the testers possibly not at least have tried to get some direct temperature measurements with thermocouples! It is as though they were deliberately trying to avoid anything that would reduce uncertainty in their results. Of course, hindsight is a wonderful thing, but the lack of control which even they noted is a serious matter in such an experiment.

Alain,

I started this thread because you said that the Ferrara and Lugano tests effectively proved that Rossi's device works.

You had arguments for why the Lugano test must be positive. I have I hope now addressed them in a "simple" manner. (And I apologise for not seeing things clearly enough to do this sooner).

Would you acknowledge now:

(1) That the Lugano test results show no independent evidence of excess heat.

(2) That they have certainly given that impression. So whether Rossi is deluded or deliberately deceptive, allowing these tests has been in his interests.

I would like to see the experiment done properly with thermocouple measurement of temperature, and the control device measured at the same temperature as the real device and with the same electrical setup. in that case we would get a tighter bound on errors. Like many, I've thought that the Rossi effect would quite quickly be resolved if we had accurate measurement of Rossi reactors.

It is not possible to put a tight bound on how much Rossi's devices resemble electric heaters without an experiment that accurately measures input and output power. Still, you get more information from the Lugano experiment than from no experiment, because this was Rossi's choice of device and also his choice of calorimetry (the previous device was painted black, for example). For Rossi to have a null test under this circumstance means that either his device does not work (which he may or may not know, since he could make the same mistake as the profs) or he wants the test to be inconclusive for some difficult to comprehend reason.

Either way, it means the sum total of Rossi's tests gives us no information that his devices work

@'Alain

I'm still wondering whether you now accept that there is no independent experimental evidence that Rossi's devices work, contrary to your previous statements on other threads that the Ferrara and Lugano tests proved they worked beyond doubt?

To summarise why the various arguments for the Lugano test being positive, as I note Rossi and all the financial types who reference him claim,don't work:

(1) The wrong emissivity used by the profs (0.4) inflates apparent temperatures

(2) The work of Bob Higgins, while valuable, does not include the nonlinear effect of the Planck curve and so underestimates the real COP inflation. Once this is included we have COP ~ 1

(3) The change in power of T from 1.6 (assumed by Optris due to wrong emissivity) to 2.1 (real) accounts exactly for the apparent acceleration in COP that is so superficially impressive.

(4) The "adjustment" of book emissivity values to match independent and accurate measurement of temperature at low temp values means that this error does not affect the dummy test and explains why it is fairly accurate. it is also true that the effect of wrong emissivity is smaller at these temperatures anyway because the unadjusted book value is much higher. Still the fact that the profs needed to do this shows that even so the effect was substantial. This adjustment was not done at high temps because the testers had no independent source of temp measurement for these temperatures.

(5) Although the transparency of alumina makes emissivity at higher temperatures variable, and is an additional source of error, this cannot much increase COP. That is because the error has nothing to do with real emissivity at high temperatures, but only depends on the book emissivity used by the profs that inflated apparent temperature.

(6) The isotopic results are extraordinary if real, but they are the one part of the experiment where Rossi was heavily involved:

• He inserted the fuel
• He removed the ash
• He had the ability to swap the reactor (and there are unexplained electrical changes that hint at this though it is in no way proven and my guess would some simpler sleight of hand on fuel removal or insertion)
• The Lugano testers are on record as not checking for such a switch - because they did not consider deliberate falsification from Rossi possible
• The previous isotopic test from a Rossi reactor was later dismissed by Rossi as wrong due to contamination. But it stood as an unexplained anomaly for many months with Rossi saying nothing against it. So even considering no deliberate falsification the possibility of contamination again with Ni-62 that we know Rossi previously used as a fuel must remain.

PS - for those not following this thread from start:

Numerical solution in my paper at LENR CANR here gives COP ~ 1.07 +/- 20% (depending on exactly how you choose various parameters which are not clear, the code displays bracketed results for different parameter values). The solution includes convection and radiation but there are various approximations made necessarily hence the large error bounds.

The fact that acceleration goes away, and the rough fcator that radiative power reduces by, can be found above on this thread shown by a web Planck curve band integration calculator.

Bob Higgins work is referenced in my paper - I had the benefit of Bob's work on emissivity but then did a more thorough job of working out what that meant for COP.

@Urban - thanks for a considered and thoughtful reply - I'll address this in my next post, and answer Alain's points here.

@Alain.

See below

Quote

the emissivity is a key factor, and the physicist screwed up, in that context, not to check the tables with a good calibrations.about the temperature assuming emissivity is 0.9, we all mostly agree, that it is around 750C.

Good.

Quote

You story about T^2.1 seems very strange for me... you seems to say that the bolometer sensitivity is strongly wavelength dependence... I imagine the documentation is referenced in your report...

This is the key misunderstanding. The factors of 1.6 and 2.1 DO NOT COME FROM THE BOLOMETER. They come from the Planck curve when evaluated at the bolometer pass-band.

Try it yourself: https://www.sensiac.org/extern…d_radiance_calculator.jsf
And use radiance from 7um to 13um (or thereabouts, exact limits don't matter).
See how Band Radiance R(T) changes with Temperature T
For a given T0 and T close to T0 we have R(T) ~ T^N where N is the exponent which depends on T0.
To get the exponent (roughly) try this at two temperatures T and T+50. the exponent is then roughly: R(T+50)*T / [ R(T) *(T+50) ]
Remember that T here must be in Kelvin, not Centigrade.

Quote

My computation is based on a flat wavelength response, and is quite robust under that assumption.

That is what I suspected, and it is why your results were incorrect. It turns out that assumption is way off.

Quote

Note that some state that you use the low temperature version of the Optris documentation, and I never considered that argument under my model of flat response. if response heavily depend on the bolometer array characteristics, it is a point to restudy.

There is some dependence on the bolometer, but it is relatively small. You can for example try the above experiment using different radiance bands`:

7-10
10-13

You will see that the exponent ratio changes very little. The absolute change in radiance does not matter, because that is accounted for by the Optris calibration. Also the absolute change in exponent. But the ratio in exponent between the two temperatures does matter.

Quote

About calorimetry, Ferrara test is perfectly calibrated, and fairly independent (reactor could be inspected and was inspected without opposition), so it worked.

I have not said much about the Ferrara test. It was criticised (rightly) at the time because there was no indication of how input power was measured. There are many ways that could have gone wrong. The testers, you will note, took this criticism seriously and greatly improved the input power measurement for the next test.

Quote

Lugano was fully independent, abandoned alone to the testers, and unfilled by the testers (despite the misunderstanding).I don't exclude the test was badly done, but it is very strange that Rossi did not notice his reactor was so cold.

I agree the excess heat results were independent. However the calorimetry method used (and that has this bad flaw if the calculations are done naively) was presumably also used by Rossi, and determined by the apparatus which unlike previous apparatus was not painted black.

Sure, maybe Rossi himself understood the issue and did the calculations correctly. But in that case surely he would have alerted the testers to this possibility of calculation error? They were in communication with him about the test. Remember that the thermography method was introduced by Rossi for the previous test and the testers followed Rossi. While they would of course check that his methodology seemed correct they were not the experts on thermography, as their mistakes here show.

My guess is it is most likely that Rossi made an identical error himself. You can see how easy it is to make this error! Whether he ever realised it was an error, or not, would be speculation.

Quote

and if he was fooling the testers, this protocol was suicidal.If a scientist used a good multiband bolometer, a thermocouple, or known emissivity dots he would have seen the problem - they even tested known emissivity dot but they fall of it seems - so Rossi did not oppose that.Theory of fraud is impossible to support honestly.

Alain, any analysis of psychology here is unsafe. There are many different permutations of error or deception possible here - you are making bad assumptions if you think this is impossible. I'm more comfortable not speculating about fraud (Urban was right above that my comments about the isotopic analysis, where i indulged myself, are not safe). If you wish to debate the issue of psychology and fraud relating to Lugano perhaps a separate `(speculative) thread could be used for that purpose?

Quote

Even the hypothesis of emissivity at 0.9 or 0.4 seems to neglect the effect of short wavelength transparency on radiation, and internal metallic reflectivity. The Dogbone are really complex for optical calorimetry, and calibration was required.

Agreed. However it turns out that the errors here are smaller than the main error, which DOES NOT DEPEND ON THE ALUMINA and wavelengths where it has variable characteristics. The main error depends only on what the Lugano Profs THOUGHT was the alumina's characteristic.

At the relatively low temperature used (750C) the alumina has a high emisivity and therefore is not much transparent. Any error here due to transparency is therefore low - essentially the frequencies at which this happens account for relatively little of the radiant power.

Quote

I would propose that since the dogbone is fat and opaque inside, then the emissivity should not assume any effect of transparency... Fins and surface state may also change emissivity (some say it reduce reflectivity, thus increase emissivity).It does not impact normally the temperature since alumina is opaque in Optris range.

I agree 100%. There are two affects, the change in radiance in the bolometer range, and the change in total radiance.

The bolometer is at a frequency where the alumina is well characterised and errors are small. However any error here has an amplified affect on the results.

The second has an affect, but it is not amplified and just means that the COP is inversely proportional to the real total emissivity.

The big error is known and relatively constant, due to the difference between the book value used by the profs (0.4) and the bolometer effective emissivity (close to 1 - I think 0.9 is a pretty safe value and 0.97 - 0.85 are safe limits).

Quote

It looks like a screw-up caused by abuse of theory, but it is not even sure.

I agree it is a theoretical misunderstanding. But the affect of it, given the recorded results, can definitely be calculated as here. The uncertainties remain, but they relate to the experiment itself, not to the miscalculation.

Quote

About Isotopic measurement, Bo Hoistad say clearly that Rossi was present but did not extract the samples himself. He just controlled the quantity was minor. You show your bias by refusing that fact again.

Urban was right that my comments, in this thread, about the isotopic results were inappropriate. I could properly summarise this by saying that while the excess heat results are definitely independent, the isotopic results are equally definitely not independent. I realise you will maybe disagree - but arguing that case involves a lot of speculation and counter-speculation. Let us do it an another thread. I'll start it.

@Mary Yugo
I think your speculation here is harsher than is necessary, but I'd like that matter to be on a different thread. it is better to keep this one for the known science. If you feel strongly you could start such a new speculative thread now and I'll contribute, also I'll answer Alain's comments about psychological plausibility.

@'Urban

Re investigation the initial issue on emissivity was raised by GSVIT and independently Bob Higgins - both referenced by me. I just put it together following through the quantitative implications which are surprisingly difficult to work out when you first consider the matter (well, they were for me - and others just did not do this).

Quote

Point (5) I cannot completely agree with you. I don't think you can assume that the reactor has the same emissivity curve as pure alumina. What if the alumina is largely transparent at certain wavelengths and there is Lithium vapour within the tube? How would that affect the emissivity over wavelength and the COP calculation? Can we really be certain without having measurements on the acutal device? Maybe we can, but I would appreciate if you cold perhaps elaborate a little more on that so that I can understand.

The issue is that at the Optris bolometer pass-band (deep infra-red) the alumina characteristiscs are very clearly almost a black body. With high emissivity anyway surface microstructure does not much matter and the bolometer real emissivity is pretty safe.

Therefore the temperature calculation, and the relative temperature calculation, is pretty safe. There is some error of course, and there are errors from other causes. We could try for a complete quantitative error analysis but I'd rather just say +/- 30% from this and various other (see below) errors.

The key thing (rather surprising) is that the big error depends on the ratio of the band-pass emissivity (near 1 and at a wavelength where alumina behaves fairly nicely though I agree there will still be some small differences between different samples) and the emissivity the Profs keyed into the Optris. This from their description of methodology must come from their curve and be very close to 0.4.

There is then the fact that at a given real temperature power out and hence COP depends linearly on total emissivity. At 750C or so some fraction of this (maybe 20%) is at optical frequencies where the alumina becomes transparent and everything is unclear. But the emissivity here is bounded by 1 - 0 and since only 20% or so we get that much error - say +/- 10%. We should maybe look at this more carefully for a proper error analysis. My numbers get a value of around 0.8 for this (alumina from book curve at 750C) and it is difficult to see it going a lot higher than that since so close to 1.

Quote

Point (6) feels a little bit out of place, can we really get anywhere discussing this?

Agreed, see my comments above

Error analysis.

The COP headline value depends on the (real) banned emissivity as follows:

At the (used in Lugano) temperature R scales as T^1.6

A fractional change of epsilon in band emissivity:
e' = (1+ epsilon)e

Causes a Radiance change of 1+epsilon and therefore a displayed temperature change of (1+epsilon)^(1/1.6) or a
total power change of (1+epsilon)^(4/1.6) assuming constant total emissivity.

At the actual temperature, total emissivity scales as as T^(-0.7) (looking at the slope of the "book" graph from the Lugano report - this is a decent first guess and the change here will not alter conclusions much).

So we get total power change of (1+epsilon)^(4/1.6)*(1+epsilon)^(-0.7/1.6) = (1+epsilon)^2.

So the dependence of COP on epsilon is roughly as 2*epsilon. Any uncertainty in band emissivity gets multiplied by 2 to become uncertainty in COP.

A band emissivity value of 0.97 - 0.83 (+/- 7%) turns into a COP uncertainty of +/- 14%. That is relative to a band emissivity of 0.90 which I seem to remember is what numerical integration of Bob's data gave, and is close to that accepted by most people.

The total emissivity at the actual temperature is also variable, and partly independent of the band emissivity - though there must be some correlation between the two. If we suppose +/- 10% on total emissivity (reasonable given the ballpark figure of 0.8) we get a total error bound of:

+/- 30%.

Tom

Quote

There is also the incontrovertible fact that Rossi's early experiments with Focardi had COP's of 200 or more, the experiment with Levi in February 2011 gave a consistent COP of 30 for 18 hours, the experiments with Lewan gave COP's of 6 or so, and the current ones even at best, are claimed to have COP's of 3. This is a classical pattern for a free energy scam. It fits perfectly. The closer you look, the less it works. Or perhaps Alain thinks COP's going from 200 to 30 to 3 is what one expects of development and improvement over almost ten years?

I agree with this. It is not speculation, nor psychology - there is a pattern of the apparent COP going down as specific experimental errors are noted and (usually) not used again.

The fact that new errors occur in the later demos shows real inventiveness.

However that fits both a typical self-deluded inventor and a scammer.

I'm then not going to argue about all the other evidence. Just saying that distinguishing the two cases above is less easy than noting that one or other applies.

Quote

All this [lexicon]conversation[/lexicon] and disagreement about thermal camera measurements would be entirely moot had any of the illustrious professors properly employed a thermocouple, much less a calibration run over the full operating temperature range of the device.

Again, it is not speculation that the profs made a bad optical measurement error only possible because they had no control or calibration at temperature, and that in a serious test of this sort that lack is highly unusual.

It is not speculation that they adjusted their book emissivity data so it would give the right result at low temperatures when used with the wrong method. The fact that they had to do that, and still viewed the book data at higher temperatures as reliable, shows bad judgement. I think that is why they emphasise the acceleration in COP. They'd tell themselves (this is speculation, but fairly safe) that even if emissivity were wrong it could not give this. That is the really big mistake, and a lot of people have made it.

It is finally not speculation that they have refused to respond to cogent and serious criticism. That shows a lack of concern with the truth and is an unusual attitude in a scientist.

I have long ago stopped trying to make sense of Rossi's statements. They defy analysis.

Do we take from this that he is saying he thinks the Lugano reactor to have been operated at 1400C?