MIZUNO REPLICATION AND MATERIALS ONLY

Wyttenbach the form of energy in and the form of energy out makes a difference. Surely you understand enough about thermodynamics to see why this is critically important.

Quote from Daniel_G

I have said this many times before but we use electrical input for physics experiments because electrical power is easy to measure.

And because you can turn electric heating up, or turn it down. In a practical, commercial reactor you might need electric heating, but you could greatly reduce it by insulating the reactor. You would not want to insulate an experimental reactor for several reasons. Mainly:

1. You cannot control insulation. If the cell starts to overheat you cannot instantaneously remove insulation, whereas you can turn off the electric heating. A commercial reactor would be fully under control. (It wouldn't be allowed, otherwise.)
2. You need heating to calibrate with electric heating in any case, so you might as well use it to heat the cell. Electric heat is not noise, so it does not reduce the accuracy of calorimetry.
3. Insulation slows down heat release and makes the calorimetry take too long to register a given power level.

Quote from Bruce__H

Resistive heating. Take a resistor and put some current through it.

I am not saying that this is a mechanism that explains your results. I am saying that if you replaced the nickel mesh with a resistive heater so configured as to turn on only when the temperature of the reactor increases, you would have a sort of physical model replicating some of the properties of what you believe is LENR heating in your system.

I see. After rereading your statement, I see I misunderstood your point.

Quote from Bruce__H

The internal heating depends sigmoidally on temperature. Internal heating is low at low temperatures and goes to a maximum at high temperatures. This is roughly what I think you (and Mizuno and Rothwell) are suggesting for your LENR mechanism.

I am trying to understand your model. The Exh effect is exponentially related to temperature, based on our data. There is no asymptote in our data. Model the gain according to Mizuno's exponential function and run this simulation again and then see what happens. Our Exh effect increases exponentially without flattening. You can artificially try to model something like this by adjusting the Amax, Slope and T ½ by to 200, 0.03 and 500 for example and the bump goes away. We need to find a better model of our Exh effect. I don't recall any paper from Mizuno showing the effect is sigmoidal.

Can you explain to me why the assumption for a sigmoidal gain was made?

Quote from Daniel_G

Can you explain to me why the assumption for a sigmoidal gain was made?

Sure. I originally came up with this model as a way to think about what was going on in the system Russ George and Alan Smith were working with. From Alan's description it sounded like the dependence of rate of excess heat production on temperature was sigmoidal (actually, more complicated, almost like a staircase). Also, it seemed to me unphysical to have a process that just produces more and more heat at an ever increasing pace forever as temperature increases. A sigmoid function is the simplest way of doing that so that is what I went with.

I will rejig my spreadsheet to use an Arrhenius function with an activation energy, Ea, of 0.165 eV/K/atom which is the figure given in the 2017 paper by Mizuno. The Arrhenius function itself is sigmoidal (just look at the function and ask yourself what happens when T becomes large) but I think that you are working in the low-temperature region of the thing where the effects of the upper asymptote are not much felt.

The insistence of trying to make observations fit a preconceived model and not letting them guide your thinking towards a higher level of understanding has always struck me as a kind of self defeating intellectual arrogance.

You can’t force nature to fit your limited understanding of it, you have to observe nature and try to understand what it does.

Quote from Bruce__H

Sure. I originally came up with this model as a way to think about what was going on in the system Russ George and Alan Smith were working with. From Alan's description it sounded like the dependence of rate of excess heat production on temperature was sigmoidal (actually, more complicated, almost like a staircase). Also, it seemed to me unphysical to have a process that just produces more and more heat at an ever increasing pace forever as temperature increases. A sigmoid function is the simplest way of doing that so that is what I went with.

I will rejig my spreadsheet to use an Arrhenius function with an activation energy, Ea, of 0.165 eV/K/atom which is the figure given in the 2017 paper by Mizuno. The Arrhenius function itself is sigmoidal (just look at the function and ask yourself what happens when T becomes large) but I think that you are working in the low-temperature region of the thing where the effects of the upper asymptote are not much felt.

Again I don’t understand your choice of data for modeling. Stick to what we know. Excess heat is exponentially related to temperature. That’s empirical. Then see if your model fits the data. I think it will.

Quote from Curbina

The insistence of trying to make observations fit a preconceived model and not letting them guide your thinking towards a higher level of understanding has always struck me as a kind of self defeating intellectual arrogance.

You can’t force nature to fit your limited understanding of it, you have to observe nature and try to understand what it does.

Because at some point, no matter the theory or how the special circumstances occur, at some point the “rubber meets the road” and the heat acts upon our known universe in the usual way.

Quote from Paradigmnoia

Because at some point, no matter the theory or how the special circumstances occur, at some point the “rubber meets the road” and the heat acts upon our known universe in the usual way.

Exactly.

For instance, the extremely simplistic model I have been proposing is mostly about how a single thermal mass heats up in response to external and internal heating. The only thing specialized about it is that LENR heating is said to be temperature sensitive in such a way as to create a positive feedback -- higher temperatures evoke more intense heating which, in turn causes temperatures to rise even more. So I put that into the model in a very simple way without positing any particular mechanism. This now allows one to think about what should happen when a big lump of metal or ceramic (which has well-known physics) has a source of temperature-sensitive heating inside it.

Quote from Bruce__H

Exactly.

For instance, the extremely simplistic model I have been proposing is mostly about how a single thermal mass heats up in response to external and internal heating. The only thing specialized about it is that LENR heating is said to be temperature sensitive in such a way as to create a positive feedback -- higher temperatures evoke more intense heating which, in turn causes temperatures to rise even more. So I put that into the model in a very simple way without positing any particular mechanism. This now allows one to think about what should happen when a big lump of metal or ceramic (which has well-known physics) has a source of temperature-sensitive heating inside it.

Slightly on a tangent: I tested simulated excess heat in ceramic tubes using parallel Joule heaters. Destructive failure was the primary result.
(Mizuno’s cylinders are much tougher and don’t get so hot)

Quote from Daniel_G

Again I don’t understand your choice of data for modeling. Stick to what we know. Excess heat is exponentially related to temperature. That’s empirical. Then see if your model fits the data. I think it will.

I think it will too. I have incorporated Arrhenius functional temperature-dependence into my simple model using the activation energy Mizuno found empirically in his 2017 paper. I am finding that, indeed, his observations turn out to all be from the extreme low-temperature end of the function where the temperature-dependence is almost exponential. For some parts of parameter space, the heating curve shows little apparent inflection (although the heating is non-exponential) but for others the inflection is very distinct and that is where the sigmoidal nature of the activation function come into play.

I still have to check things over to make sure all is correct. I will post the spreadsheet within 24 hours.

Quote from Paradigmnoia

Slightly on a tangent: I tested simulated excess heat in ceramic tubes using parallel Joule heaters. Destructive failure was the primary result.
(Mizuno’s cylinders are much tougher and don’t get so hot)

Yes. The suggested LENR heating behaviour should create a positive feedback -- so runaway heating to destruction is what you expect under a wide variety of parameters. One of the most instructive things that Daniel_G could do in his system would be to install a temperature-sensitive joule heater (using a thermocouple-controlled current source?) into his existing reactor housing just to find out how his system should react if his LENR reaction has the properties he thinks it has.

Quote from Bruce__H

Yes. The suggested LENR heating behaviour should create a positive feedback -- so runaway heating to destruction is what you expect under a wide variety of parameters. One of the most instructive things that Daniel_G could do in his system would be to install a temperature-sensitive joule heater (using a thermocouple-controlled current source?) into his existing reactor housing just to find out how his system should react if his LENR reaction has the properties he thinks it has.

Thermal shock was the biggest problem. Ceramics don’t like surprises at high temperatures. An imbedded wire suddenly increasing in temperature was fatal to many tubes. Inserted but not embedded J type lamps and joule heaters fared better, although the lamps often violently exploded so I stopped using them.

Quote from Bruce__H

Yes. The suggested LENR heating behaviour should create a positive feedback -- so runaway heating to destruction is what you expect under a wide variety of parameters. One of the most instructive things that Daniel_G could do in his system would be to install a temperature-sensitive joule heater (using a thermocouple-controlled current source?) into his existing reactor housing just to find out how his system should react if his LENR reaction has the properties he thinks it has.

I would love to see how the calorimeter reacts to a 100 W incandescent (oven) lamp being switched on an off, maybe for an hour at a time.

I’m still not sure I see the utility in all of this. I am currently working on designs to produce kW levels of heat. If I get that, I don’t see why I should be concerned about inflection points. Do we agree if I’m getting 10kW if usable excess heat that inflection points won’t matter?

Quote from Daniel_G

I’m still not sure I see the utility in all of this. I am currently working on designs to produce kW levels of heat. If I get that, I don’t see why I should be concerned about inflection points. Do we agree if I’m getting 10kW if usable excess heat that inflection points won’t matter?

The inflections will be way more important then.

Quote from Paradigmnoia

The inflections will be way more important then.

Correct.

The inflection point marks a threshold beyond which a regime of temperature instability takes over. Go beyond the threshold and positive feedback drives you more or less uncontrollably up to a maximum temperature set by either meltdown of the LENR-sustaining metal mesh or arrival at the asymptote part of the activation curve. I would advocate going beyond that threshold as immensely persuasive evidence that you have what you say and in that case it is nice to know where the threshold is. If one doesn't want the thing to melt down, however, it is still useful to know where the threshold is so as to avoid it.

I think producing lots of heat without any input is more persuasive. Even Wyttenbach would be impressed then...