To discus the 'science' behind the dispute between Rossi and Industrial Heat

@randombit0

I am glad that we are slowly narrowing in on your misconception here. I hope the precise answers below help?

Quote from randombit0

Mr Clarke ! You are mixing properties of the body and of the instrument calling that a definition of s property of the body!

No, I' never said band emissivity was a property only of the alumina - it is the definition of emissivity needed for thermography calculations with a sensor that defines the band... Of course it depends on the instrument!

Quote from randombit0

Are you writing your own science ?

This is a rhetorical flourish with no meaning? Which bit of science do you think I'm writing?

Quote from randombit0

Have you ever realized also that every pixel has his OWN B(V) ?

Now who is mixing things up? But I think you are confused. The bolometer sensitivity B(v) is the same for each pixel. What you mean is that the band (effective) emissivity eb(T) that you need to put in to the camera to get a temperature out can be varied pixel by pixel and this will (should) be in a case where eb depends on temperature. Note that v = optical frequency, completely different from T = surface temperature

Quote from randombit0

Calibration files serve just the purpose to eliminate the need to knowing the B(v) !

You never normally need to know B(v) because you calibrate eb(T) by measuring it, with the camera, at the operating temperature. That is Thermography 101 and what you are told to do in every thermography handbook. The authors here only did it at low temperatures (where they say they found errors in the book values), but used the "book emissivity value" for the high temperatures - not realising that book "total emissivity" is different from "band - or - measured by Optris" - emissivity.

Quote from randombit0

I think you are hopeless ! But try to read this page:en.wikipedia.org/wiki/Emissivity
You will read that the values are referred to Total Hemispherical Emissivity.

Correct (but I remain hopeful)

Quote from randombit0

And if you read the big amount of literature and Emissivity Tables available you will see that they all have very similar numbers even for alumina.

Correct. Because they are all giving total emissivity. Some of these tables actually say it is total emissivity. Some do not bother to say it and just call it emissivity.

Quote from randombit0

so there is NOT a detector dependency on the numbers and nowhere is written "use this numbers ONLY with model XXX" !

That is correct, because no-one expects thermographers to use book emissivity values. It is as you point out not in general possible because the value you would need varies over the IR spectrum and depends on the camera. And it is contrary to what all the thermography manuals tell you to do.

You will find that nowhere in the thermography literature does it tell anyone ever to use total emissivity (without a calibration check) to determine temperature of some surface. They do say you should use emissivity for this purpose - meaning of course emissivity as determined by the specific camera. They say you should always do this calibration at temperature. If you do, then there is no problem.

One more thing. When the surface is a grey body - which is mostly true - book total emissivity, and band emissivity, are the same. That is because there is no large spectral variation in emissivity. Alumina is a very specially bad case that breaks this. The Lugano testers were unlucky that they had the wrong material.

So - this is a subtle mistake. You need to understand quite a bit before you get to the bottom of it. You should thank Bob Higgins and GSVIT for understanding this issue - I merely did the calculations (and corrected one of Bob's incorrect assumptions where he did not follow through his ideas to their conclusion).

Last word - an industry help page:
http://www.infratec.co.uk/ther…pectral-thermography.html

Quote

Emissivity values of materials greatly vary throughout the infrared spectrum. Concentrating the thermographic measurements on spectral ranges with an explicitly high or low emissivity can lead to detection of special effects.

The effect here was very very special!

Quote from Thomas Clarke

Alumina is a very specially bad case that breaks this. The Lugano testers were unlucky that they had the wrong material.

And for reasons known only to the authors and Rossi, black refractory paint wasn't used. (And thermocouples weren't used for calibration. And calorimetry wasn't used.)

Quote from Eric Walker

Quote from Thomas Clarke: “Alumina is a very specially bad case that breaks this. The Lugano testers were unlucky that they had the wrong material.”
And for reasons known only to the authors and Rossi, black refractory paint wasn't used. (And…

Yes, I find the bad methodology difficult to understand too. They invent a whole new (and incorrect) way of doing thermography!

Quote from Thomas Clarke

The bolometer sensitivity B(v) is the same for each pixel

No Mr. Clarke they are NOT ! Anyone (even students) with some experience in laboratory known that when you have an array of detectors they have different characteristics. If you want an uniform response from the whole you need a careful factory calibration. This is true for every electronic component.

Quote from Thomas Clarke

That is correct, because no-one expects thermographers to use book emissivity values.

Come ON ! Be serious ! What the tables are compiled for ?
ANY manual give you a table with reference emissivity values to be used
when you can't measure emissivity directly. The Authors did this measure for the alumina pipes, obtaining exactly the reference (total) emissivity.

Also you are completely ignoring the fact that in that case we are interested to measure ENERGY radiated and not temperature and Energy has a weak dependence on emissivity.

Quote from Thomas Clarke

The effect here was very very special!

Oh yes you Googled a page on a completely different type of measure and are trying to mess up thinks.
The page you linked refers to a narrow band thermography, achieved using filters in order to isolate the signal of the sample from the background. The equipment used is quite special and techniques and goals differ from our case.

Quote from Eric Walker

And for reasons known only to the authors and Rossi, black refractory paint wasn't used.

Black refractory paint should be VERY special to resist to that range of temperature, using paint on a porous surface introduce a HUGE series of problems because emissivity can change depending on the thickness of paint, how much solvent is still in it ( e.g. changing emissivity with time).
Also any real expert known that not everything that look black in visible light is a good IR emitter.
TiO2 dots are white in visible light but they are "black" in IR.

Quote from Thomas Clarke

They invent a whole new (and incorrect) way of doing thermography!

No MrClarke they have NOT invented nothing. Stop Bulling.

Quote

Come ON ! Be serious ! What the tables are compiled for ?

Total emissivity tables are used mostly to work out radiation from a body (for which they of course work well). If you only want heat loss you don't care whether it is near or far infra-red photons...

Quote

ANY manual give you a table with reference emissivity values to be used when you can't measure emissivity directly.

That is true, And in some cases these will be temperature dependent. Have you found the optris manual figures for alumina? Or any other IR camera figures? There will be quite some variation depending on which sensor is used....

Quote

The Authors did this measure for the alumina pipes, obtaining exactly the reference (total) emissivity.

Really? If you look at the report it says that they had to correct the book (total) emissivity to match the observations.

Quote

We therefore took the same emissivity trend found in the literature as reference; but, by applying emissivity reference dots along the rods, we were able to adapt that curve to this specific type of alumina, by directly measuring local emissivity in places close to the reference dots

And, remember, they cannot do this except at low temperatures...

Look at their example (the dot and adjacent rectangle). They get +2.5C from the "literature" emissivity value of 0.7. But, the rectangle is positioned downstream of the dot along the rod temperature gradient so the real discrepancy will be larger.

Also, if you have paid attention to the alumina total emissivity graph you can see that at low temperatures where most of the radiant power is in the far infra-red it matches the band emissivity pretty well, and the effect of an emissivity error is nonlinear with temperature. The large error is at high temperatures where total emissivity goes way down.

Quote

Also you are completely ignoring the fact that in that case we are interested to measure ENERGY radiated and not temperature and Energy has a weak dependence on emissivity.

I think you would benefit from actually reading my comment, where you will see I ignore nothing. We are indeed interested in the power (not strictly energy, but I know what you mean) radiated which depends (modulo errors paradigmnoia will tell you about) on the surface temperature raised to the 4th power and the total emissivity at that temperature. Finally, a use for the total emissivity graph! It is this 4th order dependence that means a temperature miscalculation due to the wrong Optris emissivity can have such a dramatic effect on the power out.

Quote

Oh yes you Googled a page on a completely different type of measure and are trying to mess up thinks.The page you linked refers to a narrow band thermography, achieved using filters in order to isolate the signal of the sample from the background. The equipment used is quite special and techniques and goals differ from our case.

Of course I did! you seem to ignore the fact that emissivity can vary with frequency and therefore you cannot use total emissivity to determine the actual emissivity seen by a camera. Do you seriously want to tell me this effect is negligible in alumina when averaged over 8u-13u? If so lets have at it and post the alumina spectral emissivity graph and check 8-13um vs other frequencies. I know what this looks like

I agree that as I pointed out many materials are pretty flat measured over 8-13u, but not alumina!

Now, will you agree with me and the thermography industry that:

Quote

Emissivity values of materials greatly vary throughout the infrared spectrum. Concentrating the thermographic measurements on spectral ranges with an explicitly high or low emissivity can lead to detection of special effects.

and therefore you cannot safely use book values of total emissivity as a proxy for the emissivity seen by an IR camera sensor? Nor reckon low temperature calibration is safe when the properties at high temperatures are so very different?

I recommend Bob Higgins, or Alan Fletcher (as here)
http://lenr.qumbu.com/blackbody_141027A.php

if you don't understand this issue. More pretty pictures than I have patience to postI

Thank you @Clarke and @randombit0 for this discussion.

Tom is very impressive by his continuous patience to answer questions again and again and always trying to go straight to the point.
Randombit0 posts are refreshing in this debate and she seems fully convinced by her argumentations on the lugano reports.

However, instead of having an infinite discussion, it would be nice to go straight to the core of the issue:

Randombit0, do you agree that Alumina is a special case in the sense that its spectral emissivity is far from flat and this strongly affects calculation of the Lugano report (as described by Tom in his report)?
If not, could you please point directly Tom (and us) to the exact error in his reasoning?

Thank you again to both of you.

Quote from randombit0

Also any real expert known that not everything that look black in visible light is a good IR emitter.
TiO2 dots are white in visible light but they are "black" in IR.

Exactly, just like alumina. And "black" means emissivity=1. So you were agreeing all the time: all's well that ends well.

Or, how does one find the correct temperature of the same hot material using an IR camera with a filter, and without a filter?

Clearly the output power will be the same, but the camera will see two different things, depending on the filter or lack thereof.
The camera ε adjustment will be different to correct for the filter, but the material total IR output ε isn't changed due to the filter.
The Optris 160, due to its spectral range, is effectively filtered 100% both below 7.5 microns, and above 13 microns.
A short wave IR camera will need a whole new camera ε adjustment to measure correctly the temperature of the same material, but that won't change the output IR energy or ε associated with radiant output.

Quote from Thomas Clarke

Not sure where you are going with this, but between dummy and active (first 10 days - the bit Wyttenbach likes): dummy = 450C = 720K active = 710C = 980K A factor of 1.36 (for all the equation Kelvin is the relevant scale to use, not Centigrade).

One bone = one full Forum page filled - just to explain that Clarks law can be applied to a sole ohm'sch heated E-cat! What a pity for a man like Rossi who is working for ten years... I feel tear drops coming ! But..

How can a pure Ohm heater bring a large external Rod up to more than doubling (130--> 300W) the heat emission?? The Rod Temperatures are correctly measured approved by Clarks statement and self referring law! Heat conduction is a fairly linear business. It works driven by delta T law. Heat transferrate( CAP--> rod) = watts emiteddummy/ (Tcapdummy – Touterorddummy) = 130/ (330 -150) = 0.7 W/K ; valid for steady state and correct according to clarks ...

If the rods are 50C warmer inside what we don't know, but which is reasonable, then 1W/K is also plausible. But higher values would shrink the needed delta between T Rod/Cap to drive the heat flow.
To dissipate 2.3 Times more energy, the T rods must increase by about 30% 150C → 270C Now we calculate the TCap need of a cap to get the Wattage of the Test Rod watts emitedtest/ ( Touter Captest - Touterordtest) = 300/ (Tx - 540) = 0.7
With a heat flow of 0.7 W/K: → T outercap*1 searched = 300/0.7 + 540 = 940K measured 810K .. makes no sense

With higher internal heat flow of 1W/K
→ T outercap*1 searched = 300/1 + 540 = 840K measured 810K —>OK!

Conclusion: The T measurements of the E-cat Caps were correct... Strange???

But this may be as wrong as all the other things somebody else calculated, because it's Lugano a no quality test...

Thomas will be pleased to know that some limited thermocouple data was captured by IH observers during the Lugano test. That information is in the hands of the Uppsala folks as part of their present review of the report.

Dewey:

You might want to make sure that they read, properly, my comment? It is a good summary of the issues and while it leaves many question marks, some of which they can fill, they will be better off reading it than not. IH can pay me consultancy if they want me to go discuss the finer points of it with them I don't volunteer that lightly, and am not looking for an extra job, but I'd do it well.

Tom

Thomas - thank you for that offer. I think that everyone on the IH side is pleased with the technical understanding of all things Rossi at present. The PR war is a bit of a distraction but that to will pass. I'll see if I can get you an advance look at the Penon report but cannot make any promises on that.

You continue to amaze me with your ability to cut to the chase with only tidbits of information and all in the face of enormous harassment. You really don't need to waste any more time addressing the two-bit teardown artist. You're a solid, positively contributing citizen of the LENR community.

Thank you for caring,
Dewey

Quote from andrea.s

<a href="https://www.lenr-forum.com/forum/index.php/User/1578-randombit0/">@randombit0</a>

Why thermocouples weren't used is explained in a deeply unsatisfactory way: was the reading too low and attributed to the poor contact?

there is no need to derive emissivity from books if one calibrates, which was not done in the Lugano test in the high temperatures range.
Calibration can be done with stickers at lower temperatures, and with thermocouples at hotter…

I hope swedes have watched MFMP Dog Bone 5.2 videos. Reason, why thermocouples weren't used, can be learned (and was replicated) from there. thermocouple readings were not reliable in higher temperatures, since connection to body started give up. Bob Greenyer can enlighten more if needed.

Argon - first you actually have to have higher temperatures for certain thermocouples to become problematic. The TCs were good enough for Rossi's feedback / reactor control management and the data should have been preserved and made available for use in the output calculations. That was an intentional decision by Rossi. Fortunately, some TC data is available for reference and consideration by certain investigators.

In an earlier post I had speculated that
1) during experiment setup the Lugano testers had fed 3kW into the dogbone (would fit the Joule heating data and fig.5 of the report) but COP was <1 if relying on thermocouple reading.
2) decided to discard the result attributing the low temperature to a poor thermocouple contact, and resorted to uncalibrated optical thermography which gave a favorable result due to the mistaken Optris setting, allowing the experiment to run with lower input power (900W) for the same apparent temperature (around 1400°C, whereas the thermocouple would have shown roughly 1000°C or less).

I am very curious to know if this matches the scenario the Uppsala professors described.

@andrea.s,

The problem(s) with putting 3 kW into the reactor is:
That requires another error, in that they report 1/3 that much power going in. (Flipped clamp, etc.)
At 3 kW, it would not be unlikely that the reactor actually could reach 1400 °C

What I mean is exactly that they had 1400°C from the thermocouple with 3kW in, which implies COP=1.

Then when removing the thermocouple or discarding its reading, they could back off to 900W, in which case the camera read 1400°C and the thermocouple (if present) likely less than 1000°C. Since the apparent COP was satisfactory they could use this condition, which is less stressing, for the 1 month endurance test.

Gotcha.
Then in this scenario, they simply do not calibrate the Optris while at or on the way up to from 380°C to 1400°C. Plausible.

I do hope we get to find out more about what they did, and were thinking at the time.
The long-awaited update and Q&A session is long overdue.

@andrea.s
The remaining problem with the 1400°C external temperature is that the internal temperature will be higher, maybe 1500°C, perhaps more.

The end of the skinny thermocouple wire (see the image I posted earlier) would probably fail. Type K melts at 1380°C, and will usually fail before that in small diameters.
Possibly even the coil windings might melt. Kanthal A1 melts at 1500°C, and is not recommended to be used above 1400°C or the oxide protection layer fails. Inconel wire melts at a lower temperature; I think 1430°C is the maximum for any Inconel alloy.

(Maybe a very short burst of 1400°C external could be done, but would be very sketchy)