Jed
Why can’t you run calibrations with the resistance wire on joule heating at 50-300W levels, say every 50W or at the very least run one with an empty reactor but with 300W joule heating? That would remove a lot of doubt by having an accurate calibration at a level that the calorimeter is supposed to be measuring. Just a good practice to have calibration points outside the point of measurement.
Daniel
Why can’t you run calibrations with the resistance wire on joule heating at 50-300W levels, say every 50W or at the very least run one with an empty reactor but with 300W joule heating?
This has been done, many times.
(It isn't me, by the way. Mizuno conducted all of these experiments. My role has been to kibitz, ask stupid questions, and write the papers.)
Display MoreSome time ago here, I reported that the reactor surface temperature is much lower during calibrations than excess heat runs. That was a mistake. There are two columns in the spreadsheet:
G Control reactor surface temperature
K Reactor surface temperature [Ni-mesh reactor]
I got them mixed up. Both temperatures rise during calibrations and excess heat tests. Column G goes much higher during calibrations, K goes higher in excess heat tests. During excess heat tests, the Control reactor surface temperature is close to the outlet air temperature.
I will investigate this and report on it in more detail, probably in two weeks.
Any news about that? I know you said two weeks, but this seems to me like a show stopper. If the active reactor surface temperature is not significantly larger than the calibration, I believe there is a strong chance of no excess heat...
Any news about that? I know you said two weeks, but this seems to me like a show stopper. If the active reactor surface temperature is not significantly larger than the calibration, I believe there is a strong chance of no excess heat...
Is it surface temperature per watt of output power or surface temperature per input power. You will basically draw two very different conclusions depending on which is the case.
Any news about that? I know you said two weeks, but this seems to me like a show stopper. If the active reactor surface temperature is not significantly larger than the calibration, I believe there is a strong chance of no excess heat...
Alberto, Why do you think active surface temperature (in active tests) is not significantly higher than cal surface temperature (in cal tests)? I'd expect this.
However, as you point out, if the cal and active reactors have different surface area they cannot easily be compared. There is some evidence from photos this is the case...
Isnt the whole point of a calibration run to see how much electrical power is needed to produce the _same_ reactor and air flow temperatures as an experimental run? Eg with fuel it consumes say 100w and produces temperatures of XYZ. Then without fuel it takes 1000W to produce the same temperatures. Therefore the fuel must be responsible for producing say 900W.
Is that the method he used or do i need to read the papers again?
If the active reactor surface temperature is not significantly larger than the calibration, I believe there is a strong chance of no excess heat...
It will be another few weeks. I think it is higher than the calibration, but recent calibrations have been with a smaller, dedicated heater with less surface area than the active reactor. This is unlike previous tests.
It will be another few weeks. I think it is higher than the calibration, but recent calibrations have been with a smaller, dedicated heater with less surface area than the active reactor. This is unlike previous tests.
This is different from what is described in the paper if you consider that it refers to the previous paper for calibration methodology. It is pretty important to have the calibration fully described. We really need the nature and dimensions of the control heater and the active device. It's details like this that have the potential to provide a conventional explanation for the anomalous results. Better yet, pull the fuel from the active reactor and use it for a calibration.
Alberto, Why do you think active surface temperature (in active tests) is not significantly higher than cal surface temperature (in cal tests)? I'd expect this.
However, as you point out, if the cal and active reactors have different surface area they cannot easily be compared. There is some evidence from photos this is the case...
THHuxleynew, with a COP of 5 or more, I would expect that, if the active and calibration reactors are made of the same material and have the same dimensions, the active reactor (when active) would have a significantly higher surface temperature than that of the calibration reactor. If they have similar temperatures, we have COP ~= 1...
A thought experiment:
Take two incandescent light bulbs. A 40 watt clear and a 200 watt clear.
Measure the bulb envelope surface temperature by painting a spot onto each and using a cheap IR temperature gun.
Blow a fan across both.
See the complications?
It is undeniable that both get rid of all their waste heat without melting down.
Calculate the glass surface areas. Compare the spot temperatures. Relocate the fan. Repeat. Find bulbs with different surface areas. Use halogen and non-halogen bulbs. Measure again. Calculate again. Change fan speed. Measure inside a cardboard box. Post to YouTube.
My guess: The larger wattage bulb has more surface area and might have a cooler surface than the smallest lower wattage bulb. Like statistics, the numbers can be manipulated by scale and end points. If both bulbs have the same surface area, then the larger wattage bulb MUST have a higher surface temperature, if and only if, both bulbs see the same flow per surface area of air.
THHuxleynew, with a COP of 5 or more, I would expect that, if the active and calibration reactors are made of the same material and have the same dimensions, the active reactor (when active) would have a significantly higher surface temperature than that of the calibration reactor. If they have similar temperatures, we have COP ~= 1...
Agreed mostly - if Jed has spreadsheet data for the R20 test, but we have also to check reactor size, because surface area could vary by factor of 2, and position, because forced airflow could vary by factor of 2 or more.
That is the problem when controls are not exact controls...
Physical simulation costs a pittance.
Two bulbs, a fan, an IR gun. An enclosure. Some paint. Time.
Do you want to talk or do? Do you want to know or posit?
Put your stupidphone to use.
That is the problem when controls are not exact controls...
This is pure horse shit. Any flow calorimeter works fine with any sort of resistance heater. There is no "exact" control. The only issue would be the surface temperature of the reactor itself, which is not usually accounted for with a flow calorimeter.
I will not engage with this kind of nonsense, but for the benefit of other people, I want to point out it is nonsense. People should read about calorimetery.
This is pure horse shit. Any flow calorimeter works fine with any sort of resistance heater. There is no "exact" control. The only issue would be the surface temperature of the reactor itself, which is not usually accounted for with a flow calorimeter.
I will not engage with this kind of nonsense, but for the benefit of other people, I want to point out it is nonsense. People should read about calorimetery.
This is just not true Jed. I go back to a scenario that could explain the apparent excess heat.
1) Control run is performed. Because there is less surface area exposed to the moving air, relatively more heat is lost through conduction through the bottom of the calorimeter.
2) Experimental run with a bigger reactor is performed. Because there is more surface area, relatively more heat is collected by the air moving over it than is conducted out the bottom.
I agree that controls don't necessarily have to be exact, but you run a risk of unforeseen errors by having different sized control and experimental reactors.
Had he performed a similar type of experiment as he did with the previous reactor type, this would not have been an issue.
. I go back to a scenario that could explain the apparent excess heat.
Show calculations please Jack
Once again, I would like to suggest a setup for the replication of Misuno’s results that I believe is skeptic-proof. In this setup we would have two identical reactors operating side-by-side at the same time: one active and one dummy (mounted without the nickel meshes inside it). The sheath heaters of these reactors would have the same characteristics and would be connected in series and to a single power supply. The voltage between the terminals of reactors' heaters would be monitored during the experiments to show that the reactors would receive the same amount of power. Also, both reactors would be connected to the same deuterium gas source through a shared plumbing system, so that they would have the same pressure during the experiments. Finally, thermocouples would monitor the temperature in the external metal surface of both reactors. A significant temperature difference between the reactors would demonstrate that there is anomalous heat. Later, an inert gas could be used in place of deuterium to show that the external temperature is about the same now, even considering the difference between reactors (the active has nickel meshes inside and the dummy do not). If we have a large COP, as Misuno has had, this replication would save us from those ad nauseam debates about calorimetry. It is also cheaper than alternatives using a calorimeter.
I would like to suggest a setup for the replication of Misuno’s results that I believe is skeptic-proof
No setup is skeptic proof.
there are skeptics and skeptics..
No skeptic has explained the huge R20 output heat by any calculation of any "artefact"/
although they have tried.
The R20 setup as explained with a simple insulated calorimeter box is sufficient to replicate.
This is pure horse shit. Any flow calorimeter works fine with any sort of resistance heater. There is no "exact" control. The only issue would be the surface temperature of the reactor itself, which is not usually accounted for with a flow calorimeter.
I will not engage with this kind of nonsense, but for the benefit of other people, I want to point out it is nonsense. People should read about calorimetery.
We all agree there is no exact control. But calibration depends on the assumption that control and active systems have the same calorimeter efficiency. To the extent that breaks we have an error even after calibration.
Well, I give figures: the paper calculates that 22% of heat is lost through the calorimeter walls. The real amount could be higher than this due to 18% error in airflow measurement tolerance (34%).
That efficiency depends on the temperature of the inside of the walls, which depends on the position, size, shape and color of the reactors, all of which affect air flow or radiation.
Unless those things are controlled (making a much more exact control) we have errors due to these factors of an unknown magnitude (some fraction of the measured efficiency) due to differences between the control and active system.
THH
Display MoreNo setup is skeptic proof.
there are skeptics and skeptics..
No skeptic has explained the huge R20 output heat by any calculation of any "artefact"/
although they have tried.
The R20 setup as explained with a simple insulated calorimeter box is sufficient to replicate.
Of course. In fact with claimed R20 levels of power you don't need a box.
The R20 results in the paper comprise one sample datapoint, and anecdotal accounts. Ignoring the anecdotal evidence, the one datapoint needs to be validated, it is easy for such one-off tests to have stupid errors - I've even suggested one (ignoring the different characteristics of the new heater). Or perhaps some incorrect unit conversion. Or some quantity shown as measured that is in fact calculated. How can we know? In fairness to Mizuno we should not treat data presented as sample and preliminary as though it is definitive.
walls. The real amount could be higher than this due to 18% error in airflow measurement tolerance (34%).
THHnew's vague calculations of 18%, 34% are based on implausible assumptions.
I estimate much less . The velocity profile in the annular region btw 3 and 3.3 cm is not zero.
A stepwise change from ~ 4m/s to zero m/s is unrealistic.
