All this can be avoided when Mizuno would always run a control at the same temperature.
Of course he would waste some electricity, but any discussion would immediately stop.
The other thing we recommend and do: Do a proper calibration run with a heater that goes up to 3000W (at least 1500W ). There is not need to use the same heater for the calibration as long as the external reactor geometry is the same.
it is an astounding result that will likely soon be replicated by multiple parties
is anonymous at one of those parties?
does he have an invite?
Display MoreTHHuxley and Robert Bryant,
You're both word fighting/arguing over how much excess heat R20 makes when you both agree that R20 makes the output air hotter by 13C vs 3C for the control on the same 50 Watts input from the presented data in figure 5.
Irregardless, it is an astounding result that will likely soon be replicated by multiple parties. Hotter MUST be excess power from the reaction (assuming the thermal characteristics of the control vs. the active reactor in the calorimeter are substantially similar). Sustained for sufficient time, it must be non-chemical excess energy. That is a revolutionary result. Only excess power makes things HOTTER if the other thermal characteristics are the same. It must be non-chemical, i.e. LENR.
A better calorimeter can and will be built later by other scientists and engineers looking to exactly quantify the LENR effect. Arguing over the first calorimeter's accuracy/inaccuracy when the results is clearly HOTTER is like not seeing the forest for the weeds.
If you continue investing your time in this technical disagreement, can you both please tone it down and make it less personal. (I.e. eliminate personal and personal possessive pronouns when critiquing the other's work.) Thank you.
OK - so I agree that if R20 is replicated all this is boring, LENR (or something equiv) exists, the world changes.
Otherwise:
We are not arguing about R20. We are doing R19 calculations.
It is not boring if the aim is to understand how Mizuno got those results: however if the default position that it is likely LENR there is less motivation to do this. LENR is no good for curiosity because it closes of any questions: pretty well any observed behaviour can make sense if LENR exists, so you don't need to puzzle it out.
What might be relevant is that, having conceded completely to RB re the radiation transfer issue, the R19 paper figures for how the calorimeter works now don't make sense.
That BTW is what at its best peer review does; points out things in papers that don't make sense. I'm sure Mizuno can resolve this. If R20 works as billed he probably would not bother, and in any case would publish another paper detailing R20 results.
We will see?
LENR is no good for curiosity because it closes of any questions: pretty well any observed behaviour can make sense if LENR exists, so you don't need to puzzle it out.
What is most interesting in the R20 configuration for me is the possible delocalised magnetoelectic interaction
of some of the 30 or so metal isotopes in the mesh and 316ss reactor wall.
58 Ni , 60 Ni , 61 Ni , 62 Ni and 64 Ni
102Pd, 104Pd, 105Pd, 106Pd, 108Pd, and 110Pd,
50Cr, 52Cr, 53Cr, and 54Cr
Fe-54, Fe-56, Fe-57, and Fe-58.
Mo 92, 94, 95, 96, 97, 98, and 100
55Mn 59Co Rh103?? Ag109??
How important are the low lying gamma states of isotopes
eg Mo 97
Many other questions.
optimum time temperature profile?
effect of oxygen?
are getters superior to vacuum cleanising of impurities?
10-400 keV gamma spectrum?
protium versus deuterium?
neutron emission level? how is heat emitted without neutrons?
is the energetic exchange by photonchannel or other channel,,? magnetoelectic effect?
He3 He4 production?
correlation with heat output?
effect of deuterium gas on the surface structure of the mesh and the wall..microcracks..stress..?
burnishing versus electrodeposition ?
effect of burnishing with metastable Ag107/109 with lowlying gamma states
others..niobium93..ytrrium scandium ytterbium, hafnium?
effect on radioisotopes..K40 U235 U238 Cs147? Co60 
Co57 )
is decay by alphabeta or other channel
transmutation of key isotopes? effect on overall performance?
etc etc.
Ad nauseum calorimetry discussions have started... Please, someone do a dummy (without the nickel mesh) versus active reactor replication with both reactors operating simultaneously with sheath heaters in series and with a tube connecting them together (same interior gas and pressure)!!! With the expected COP, difference of temperature between reactors will confirm excess heat (or not...).
Display MoreWhat is most interesting in the R20 configuration for me is the possible delocalised magnetoelectic interaction
of some of the 30 or so metal isotopes in the mesh and 316ss reactor wall.
58 Ni , 60 Ni , 61 Ni , 62 Ni and 64 Ni
102Pd, 104Pd, 105Pd, 106Pd, 108Pd, and 110Pd,
50Cr, 52Cr, 53Cr, and 54Cr
Fe-54, Fe-56, Fe-57, and Fe-58.
Mo 92, 94, 95, 96, 97, 98, and 100
55Mn 59Co Rh103?? Ag109??
How important are the low lying gamma states of isotopes
eg Mo 97
Many other questions.
optimum time temperature profile?
effect of oxygen?
are getters superior to vacuum cleanising of impurities?
10-400 keV gamma spectrum?
protium versus deuterium?
neutron emission level? how is heat emitted without neutrons?
is the energetic exchange by photonchannel or other channel,,? magnetoelectic effect?
He3 He4 production?
correlation with heat output?
effect of deuterium gas on the surface structure of the mesh and the wall..microcracks..stress..?
burnishing versus electrodeposition ?
effect of burnishing with metastable Ag107/109 with lowlying gamma states
others..niobium93..ytrrium scandium ytterbium, hafnium?
effect on radioisotopes..K40 U235 U238 Cs147? Co60
is decay by alphabeta or other channel
transmutation of key isotopes? effect on overall performance?
etc etc.
These are all interesting questions. And that gets to my point.
Because LENR (as a hypothesis) is so broad, it stimulates an enormous number of questions, as you have shown. The problem is that there are too many of these. Only one or two of them, if LENR is real, will end up relevant. But then to make progress we need somehow to knock out the others, and that is difficult.
You can, like axil here, enjoy open speculation where things are put together without concern for plausibility or evidence. I don't see that process as being curious, though perhaps others would.
What I mean by curiosity is the desire to understand something that was previously a mystery. It can be pursued when there are unanswered questions, as long as those are precise enough to admit a procedure for investigation and answering.
Scientific hypotheses are predictive. They can therefore be tested, and aspects (free parameters or whatever) unknown can be estimated. When predictions fail the hypothesis can be discarded.
The test of LENR - as explanatory science - is whether it does make such predictions that allow parameters to be estimated, or the hypothesis as a whole to be rejected as not likely.
W's SO4 ideas similarly can be judged, and if they are science no doubt that will happen. Thus far I've not seen experimentally testable predictions different from the well developed existing theories (rather than duplication of known results).
You can have cast iron experimental evidence that existing theories break, without a new hypothesis. A replicable excess enthalpy experiment well above chemical with no understood nuclear activity would be such. LENR would be one possible explanation for it, if the details could be worked out for a plausible LENR hypothesis that was scientific. But even without those details the fact of extraordianary results is still a big big find and one that would energise the scientific community.
If Mizuno results (R19 or R20) are as billed and replicable, they would be such an experiment. Based on them (if they are shown to be replicable) many of RB's questions above could be tested, one by one, also hypotheses to explain the results could be developed. At the end, you might have LENR, or hydrinos, or Ubba Ubba Wuk force, or whatever.
All this can be avoided when Mizuno would always run a control at the same temperature.
Of course he would waste some electricity, but any discussion would immediately stop.
The other thing we recommend and do: Do a proper calibration run with a heater that goes up to 3000W (at least 1500W ). There is not need to use the same heater for the calibration as long as the external reactor geometry is the same.
This is exactly what I plan to do and thereby avoid all of the long and quite tedious discussions.
Then the next thing is to take the generated heat and create electricity in sufficient amounts to both run the experiment and power an external light bulb.
But then to make progress we need somehow to knock out the others, and that is difficult.
We have made progress
The questions that THHnew has raised whether genuine or ingenuous have been knocked out.
20% error on an anemometer velocity traverse?????
I was doing these thirty years ago with pitot tubes and inclined manometers
No way would 20% error be allowable.
Got any more questions?
Please be specific why people should not replicate R20?
Calculations please.
Make sure you use the correct formulas otherwise you will make egregious errors like this
"
That is because at the temperatures I did this calculation (380C reactor vs 80C wall) the re-radiation is\\
less that 10% of the radiation due to the T^4 factor for relatively small gaps.
No 'fringe science ' polemic please.
We have made progress
The questions that THHnew has raised whether genuine or ingenuous have been knocked out.
20% error on an anemometer velocity traverse?????
I was doing these thirty years ago with pitot tubes and inclined manometers
No way would 20% error be allowable.
Got any more questions?
RB: I don't think you get it.
The point of all my questions about the Mizuno results is that such extraordinary results need exceptionally clear methodology, and cross-checks.
Let me also point out: for everyone else. If you ignore completely the R19 results, and take the sample R20 results as a one-off observation that is extraordinary and clearly of commercial interest, you can ignore all this.
If you go by R20 results ignore 80% of the paper - the calorimetry does not matter and that one R20 calibration versus active comparison is enough.
Personally I think the R20 results are billed as preliminary and it is not fair to Mizuno to treat them as anything else. All that would be needed is a heater power miscalculation, possible because the R20 has a different heater design from the previous reactor, and that results no longer holds. But, I can understand others having a different view.
So: R20 enthusiasts stop reading here.
Take this question about the anenometer readings. There have been a number of issues any one of which would (rightly) be questioned in peer review: and would be tightened up, or mentioned as unresolved with appropriate bounds, by paper authors who wanted the best possible paper.
Let us look at these one by one:
1. This was poorly presented in the draft paper. I'm sure that before any publication Mizuno would want to make clear the methodology, that the actual output power is calculated based on calibration of the blower. As long as the calorimeter geometry and blower specification do not chnage between calibration (several years ago?) and measurements this is not a problem. Readers would have no problem understanding this is the calibration and setup details were clearly stated: blower type, airspeed measurement methods (e.g. where is the anenometer, what diameter tube is it placed in, how far from blower. Without these details the blower speed could be calibrated under quite different conditions from the active runs. That matters now we know that the air speed was not directly measured.
2. This is just a matter of more care writing up the paper: assuming that detailed enough logbook data exists to know the answer! With Mizuno's help the work could easily be redone if needed.
3. Normally I'd say this is a level of detail not required. However it most certainly is required here, because of 4.
4. This is a real problem. We have two ways to calculate total airflow that do not agree, and a 24% difference between the turbulent flow estimate and the assumption of constant air velocity profile. The anenometer results cannot be trusted to measure accurately the edge flow because of lack of detail on anenometer geometry (how long is the hot wire, which way is it oriented, from what reference point on the wire is the (clarified by Jed, but not yet written in paper) dimensions of 30mm from centre of a 66mm tube, and finally the fact that the flow velocity profile will be chnaged by insertion of the anenometer and the "slow areas" close to the edge may then suffer additional turbulence and no longer be slow.
More investigation might clarify the velocity profile issue. For example, turbulence introduced by the blower might mean that the flow after the blower had a flatter velocity profile than would be the case further down a tube - where the standard profile would dominate. The problem is that this is a guess, and until the matter is resolved we (properly) have an unknown correction factor of between 0 at -24% on the airflow estimation. A conservative bound might put that as +5% - -28%, although with more precise data (e.g. conditions under which all measurements are made) that could be tightened. I'm allowing 5% for combination of measurement tolerances and difference in conditions between blower calibration and active runs. Actually, that is a guess, it is really difficult to bound these things without more specific data from Mizuno, but I'm fairly confident in it.
So what does this one point (airflow estimation errors) mean for the paper results?
It makes the absolute measurements less clear. There is still an apparent absolute excess of 1.48 * 0.722 = 1.068. However that is too close to other not yet bounded tolerances.
THUS ABSOLUTE R19 POWER MEASUREMENTS IN THIS PAPER DO NOT CONFIRM EXCESS HEAT BECAUSE AIRFLOW CANNOT BE WELL BOUNDED
That is not necessarily a problem. The control measurements also show extraordinary results, and the comparison of control reactor with active reactor does not depend on the airflow measurement tolerance.
So all we need in this case is a clear statement of exactly under what conditions the control versus active measurements were made, so we can bound any differences that could be a result of these different conditions. We do not have clarity here yet. The possible differences are:
I agree with RB that calculating thermal characteristics here is very complex. These can alter results by modulating the amount of heat lost from the calorimeter box differently between control and active runs. That would be bounded by the calorimeter efficiency: 77% as estimated in the paper. Unfortunately the airflow bound means that those efficiency figures vary by between +5 and -30%. That means without more data comparison of control and active runs does not prove excess heat if the above changes make big differences in calorimeter efficiency.
Most of these differences will likely not be significant, some will be significant: it is difficult to bound this because we have not enough information about timing and conditions of control runs.
This is taking into account just one (the most obvious) of the uncertainties here.
Other uncertainties:
RTD heating due to conduction from bracket (not known, probably no significant error but a check would be good)
Air temperature change through test combined with time constant of reactor and calorimeter (difficult to bound, but probably bounded at less than +/- 10%).
Context
In any experiment with normal results you would not worry so much about these things. You would reckon that the chances of all these uncertainties going in the same direction is small, that most of these uncertainties, when investigated, will prove to be illusory.
However, when highly unexpected results are shown, caution is required. In this case resolving these uncertainties would not be very difficult. The way I'd do it would be to ignore the absolute measurements (too many variables and work to tie all down) and concentrate on better bounding of the differences between control and active R19 data. That could quite easily be done.
The problem in this paper, as in much LENR work, is that two separate "almost good enough" indications are taken as reinforcing each other. Thus the absolute measurements are almost good enough. Equally, they are not needed if the control measurements are under the same conditions as the active, that comparison alone is almost good enough. But, individually, these two types of measurement, absolute, or control vs active, can be criticised. And when results are unexpected that means everything must be revisited and checked.
Another (LENR typical) conflation of two almost good enough results. R20 alone is well beyond calorimetry. If real it can be proven by anyone with a hand over the reactor. But this is a single result, billed as sample. R19 is documented over many observations, with a lot of collected data and much investigation to determine accuracy. Unfortunately, as we see above, you can't bound the results in a way that removed possibility of errors without further observations, or further clarification of exact conditions of the given observations. I'd caution viewing R19 as proven due to R20, or R20 as proven due to R19.
THH
I agree with RB that calculating thermal characteristics here is very complex
Especially when THHnew uses the totally inappropriate formulas
That's why I ask for calculations rather than words
1448 words in the last post... no calculations
Please show CALCULATIONS of errors otherwise its only words
among other bloopers.
THHnew has lost credibility with
That is because at the temperatures I did this calculation (380C reactor vs 80C wall) the re-radiation is\\
less that 10% of the radiation due to the T^4 factor for relatively small gaps.
The actual effect of re radiation is 40%..
If THHnew doesn't know how to do this calculation I can educate him
CALCULATIONS PLEASE.
Please show calculations of 20 % error in the velocity traverse
RB above says "no ways could this be a 20% error". But it easily could if the anenometer housing was creating turbulence that made airflow over the anenometer less close to the slower close to boundary flow
pls state the diameter assumed
please state the velocity profile assumed.
CALCULATIONS PLEASE.
BTW THHnew you have a bad habit of misquoting
"No way would 20% error be allowable" referred to my tests thirty years ago
Please make your velocity traverse calculations intelligible.
If possible draw a diagram of the velocity profile.
Please show calculations of 20 % error in the velocity traverse
pls state the diameter assumed
please state the velocity profile assumed.
CALCULATIONS PLEASE.
RB - please give me the information I've asked for: how does the insertion of the anenometer affect air velocity profile, what is its geometry, and we could start to do that, but as i'm sure you will agree it is not easy to work out airflow in complex turbulent systems.
Do you agree that for normal turbulent airflow in a pipe Re=12,000 the average velocity (integrated over cross-section) and peak (middle of pipe) velocity are related by a factor of approx 20%? It could be up to 24%, the various approximations here differ a bit. I have previously posted the calculations from many textbooks that show this. If you doubt it please say so, and we will revisit that.
So: what, in your opinion, gives? Is this flow just so turbulent, due to the blower perhaps, that normal turbulent velocity profile does not apply? And how do we know that is the case from the given measurements when the anenometer insertion surely changes the airflow near the edges?
I'm not saying it is likely to be 24% in error. If I had to guess, I'd say 0% - 20%, allowing the possibility that the blower makes a very non-standard profile. Why are you so certain here?
I'm not saying it is likely to be 24% in error. If I had to guess, I'd say 0% - 20%,
Now you are backing down...0% - 20%,???
Justify your 24%.
You don't need Reynolds
Use the data ,,,you have already been told the diameter.
If you don't know how I will show you how..
I might get time to do it in the next week.
Its boring though..fluid mechanics 101.
For R20 .. much more interesting is the
cooperative effect btw the 30 isotopes
RB:
I truly think you have a conceptual error here in the way you approach this.
I'm saying that the information in the paper does not bound the airflow. Specifically that the anenometer measurements do not bound flow near the edge, because of two errors (insertion alters airflow, anenometer geometry means that airflow measured is not exactly at position stated).
For example, we only need the edge 3mm of the airflow to be very low, and we have a 20% error.
Unless you can bound these errors - you have given no information about what the airflow is. Not just 24% error.
I'm proposing, as a possible "sensible" bound, the expected velocity profile of air in a tube at this speed, which gives around 24%. I agree, it is not definitive, I could include the anenometer tolerance, the fact that bound should always be outside calculated, and make this 30%. And it is not "backing down" to say that you don't expect flow to be close to its bound - that is just statistics.
The point is that you are not bothering to give an error bound on this figure. You are saying it is precisely accurate (obviously wrong). Now, what error bound do you think is appropriate from the airflow, and how is that broken up into the various errors?
I truly think you have a conceptual error here in the way you approach this.
You are the one with conceptual errors.
That is because at the temperatures I did this calculation (380C reactor vs 80C wall) the re-radiation is\\
less that 10% of the radiation due to the T^4 factor for relatively small gaps.
show calculation please.
I need to check your conceptual errors.
I'll give you a week to come up with a calculation of 24% error.
then you can check with my calculation.
For example, we only need the edge 3mm of the airflow to be very low, and we have a 20% error
This is wrong wrong wrong... and its not 24%????
show your assumed velocity profile and how this compares with the data
SHOW CALCULATIONS.....
if you can't do this
I will show you in August
1) the Shanetsu insulation used in this paper is rather poor relative to the state of the art insulation (Vacuum insulated panels). This product (https://www.turvac.eu/0/Products/WhatisVIP.aspx) gives extremely low thermal conductivity (3~10x the product Mizuno used) (3,5mW/mK), with just 20 mm thickness, U value less than 0,22 W/(m2K) can be reached. This would allow more precise calibrations at higher temperatures and more capture of the reactor heat in the air flow.
2) for replicators with sufficient funding a mass flow meter such as this (https://www.sierrainstruments.com/products/oem-probes.html) would help put to bed about 70% of the discussions here. This device provides 1% accuracy in air mass flow and 0.2% in repeatability. This slight modification would leave zero doubt about velocity profiles and turbulent flow, etc. since it measures mass flow not velocity. Also the turndown rate is 1000:1 and dynamic range is 0 to 20,000 SFPM so this device should allow replicators to measure a higher power level at higher temperatures accurately all the way up to 3000W.
If someone is successful replicating exactly as in the paper, if it were me, I would improve the air mass flow measurements and improve insulation with VIP technology and run the experiment again.
1) the Shanetsu insulation used in this paper is rather poor relative to the state of the art insulation (Vacuum insulated panels). This product (https://www.turvac.eu/0/Products/WhatisVIP.aspx) gives extremely low thermal conductivity (3~10x the product Mizuno used) (3,5mW/mK), with just 20 mm thickness, U value less than 0,22 W/(m2K) can be reached. This would allow more precise calibrations at higher temperatures and more capture of the reactor heat in the air flow.
2) for replicators with sufficient funding a mass flow meter such as this (https://www.sierrainstruments.com/products/oem-probes.html) would help put to bed about 70% of the discussions here. This device provides 1% accuracy in air mass flow and 0.2% in repeatability. This slight modification would leave zero doubt about velocity profiles and turbulent flow, etc. since it measures mass flow not velocity. Also the turndown rate is 1000:1 and dynamic range is 0 to 20,000 SFPM so this device should allow replicators to measure a higher power level at higher temperatures accurately all the way up to 3000W.
If someone is successful replicating exactly as in the paper, if it were me, I would improve the air mass flow measurements and improve insulation with VIP technology and run the experiment again.
Never loose sight of the fact Dr. Mizuno is doing it with several budget and facility constraints. He is doing as best as he can and his reaching out to others to replicate is precisely because he needs other to confirm his results.
RB: I've laid out the possible error mechanisms in the anenometer traverse data. You have not engaged with them. The graph above shows middle of pipe air velocity vs blower power. You have not even bounded the error on this graph (which i agree is quite small). The issue here is how does airflow relate to middle of pipe air velocity.
Let me reiterate: Mizuno supposes the air velocity everywhere is the same as what is measured in the middle of the pipe, when that can't be true (the edge velocity must be 0). The question is how much is it not true. The anenometer measurements are not reliable near the edge
Given the uncertainty there, and your inability to put bounds on these errors, I am going with standard turbulent flow velocity profile: 24% would be a decent bound here, on the high side of the various estimates from textbooks.
I have given the calculation for the 24% bound before: it comes from any textbook on turbulent flow as follows:
Re for the given pipe (66mm) and speed 3 m/s and air temperature 40C (all units S.I.)
https://www.engineersedge.com/…c_and_kinematic_14483.htm
rho = 1.127
kv = 1.702 X 10^-5
https://www.engineeringtoolbox…eynolds-number-d_237.html
Re= 11633
For turbulent flow Re = 11633 we need the velocity profile - and then to integrate this over area to get the average velocity. Textbooks do it, and produce approximations for the number we want, which is average velocity / peak velocity.
One thing you need to know, this is turbulent flow so all these numbers are stationary time averages (obviously). Average / peak means average over cross section of tune / peak in middle of tube.
https://jingweizhu.weebly.com/…bulent_flow_modelling.pdf
Page 8:
Here f = 0.0304
sqrt(f) = 0.174
from (39) ratio between max and av velocity is 1 +1.33sqrt(f) = 1 + 0.23 = 1.23
From which I get my 20% reduction figure.
As I said, there are different approximations so this figure is +/- a bit, and for a decent bound I therefore take 24% (look back on this thread for the approximation that gives this higher figure).
As I've indicated there is quite some uncertainty in what these bounds should be. Doing it properly we need to add in all the errors in the measurements.
NOW - before you cavil at these figures let me point out - I DONT KNOW THEY ARE CORRECT.
But they are better than your results, where YOU WILL NOT PUT ERROR BOUNDS ON THE DATA IN THE PAPER.
Conceptual error 1: thinking that empirical measurements without error bounds prove anything
Conceptual error 2: thinking that error bounds are exact figures
For example, my bound of +5 - -30% may be too large. Maybe +2 - -21% is a possible better bound. But, to justify this tighter bound, you have to engage with the full error analysis here, and show that a tighter bound is proven.
