We believe we can explain at least part of the apparent excess energy Celani saw based on the nature of the gas as it changes pressure.
Here, again, is one of the runs published by Celani. Note that the excess power quickly rises to 10W during the loading phase, and then returns there again after charging the cell w/Hydrogen. Note how the power jumps up when the pressure drops occur, too.
We were puzzled by how our excess power calculation could be so far negative in some of our key runs. Then we were cautiously pleased by how the indicated excess power rose to be a couple watts positive over a few days. One very insightful commentator, Ascoli65, contributed the following analysis pointing out that if we took our equilibrated starting point as the baseline, like Celani did, we would have shown power levels that would have been a basic replication of what Celani reported.
In both cases, the power out calculation is based solely on the temperature on the outside of the glass, and in both cases, the excess power indicated rose as the pressure decreased. That made us wonder if that was a key relationship. We started to explore that on the last blog post.
Since then we undertook two more tests. First, we charged the cell with the H 75%/Ar 25% mixture to 8 bars and then stepped the pressure down in 1 bar steps so we could see the effect it was having on the temperatures in the cell. The second test was the same thing, but with Hydrogen gas, to match the second part of Celani's "06giu12" run. Below are the results and some ramifications of the results.
In the graph below it is very clear that the exterior glass temperature and the mica temperature rise immediately with each pressure decrease. The magnitude of the temperature change is very significant.
When we plotted the settled temperature rise over ambient vs pressure, we see an interesting curve appear. We added it to the same graph we had shown in the previous post. The long green curve is for straight Hydrogen. The long blue line is for the blended gas.
Note that we added an estimated point at 9 bar, which is close to where the post loading run in Hydrogen entitled "06giu12" started out. The general slope is comparable to the other little data sets at lower power levels. The fact that these last two tests show higher temps than the calibration data may be that these sensors run hotter because of the denser wrapping. It may still be attributable to the fact that we are using a wire that is potentially active. The leveling off at low pressures also may merit more study.
Below is another representation of the same data but divided by the P_in to get some sort of normalization. Again, the behavior at low pressures looks like it may be interesting.
The ramifications of this effect are heavy. To estimate them, I plugged the values for T_glass_out and T_ambient into the S-B equations that Ascoli65 cited from Celani. First, I used the blended gas data that I had at my disposal first. I went from 8 bar to 3 bar, mostly as an excercise.
And next in Hydrogen, from 9 Bar, where Celani's Hydrogen run phase started and then 3 bar, approximately where it ended.
As a preliminary result, it appears that this pressure related temperature change in Hydrogen could account for the vast majority of the demonstrated rise in power in Celani's graph above the 10 watt baseline that the run starts at. It is unclear to me how he established his base line for that experiment that resulted in showing approximately 10 watts during the loading in the blended gas, and at the start of the Hydrogen phase.
This explanation does not at all address the measured gamma rays coinciding with hot spots in the wire. It also does not address the test runs he mentions in the calorimeter.
I am assuming at this point that the pressure dependence is caused by the thermal conductivity of the gas changing. I have not yet been able to locate a good reference on that. Can anybody help with that? I was able to find a reference showing that the thermal conductivity of the Hydrogen increases almost 50% going from room temp to 200C. The concept is that as the thermal conductivity drops at lower pressures, less of the heat from the wires flows out of the cell through the metal flanges and the cooler parts of the glass near them.
Going forward, I would like to take an entirely new cell with Nichrome or even Iron wire in it and repeat the pressure vs. temperatures tests from 8 bar down, again, perhaps at multiple power levels. I would also like to test the gas range under 1 bar. In all tests like that, we will take thermal images of the cell at 8 bar and at 1 bar and compare the relative temperatures of the glass and the metal flanges.
And, of course, I would like to test a nice, new wire in a calorimeter. We're working on that.
I welcome your thoughts, as always.
Comments
I needs to spend some time learning much more or understanding more.
Thanks for magnificent info I was looking for this info
for my mission.
We specialize in placing text links on high quality websites.
Looking for four proven Internet business promotion methods.
do you still confirm that there is a big question mark on Celani NiWeek measures?
Celani said that direct heating was more efficient than indirect heating by the nichrome wire, but that it worked a little. is it coherent with the pressure artifact hypothesis?
Celani claims of Anomalous heat are 30%, is it compatible totally with the pressure artifact hypothesis?
given Celani various claims, is it possible that all be mistakes, or are there results from lab that looks correct anyway?
I second what observer has said. Is Celani aware of this?
This is crude illustration, but it took me about ten minutes. In about an hour of real-time chat we could create a much more accurate one, which anyone could contribute to to make more accurate still. Each individual element of the system would an easy to reference name that could be used in discussion, and every possible energy flow was clearly identified and labeled. This could widen participation, focus discussion, broaden perspectives and highlight relationships. All the physical details could be mapped right behind it. You're doing great engineering. How about applying engineering to this very valuable dialog?
Most of your power input goes to heat up the cell, not specifically the wire. So if the wire is a source of excess heat and you increase the number of wires, your signal to noise ratio will go up. Add enough wires and maybe the considerations you are currently bogged down in will go away.
I know the wires are difficult to make but it may be worth doing it.
[BEGIN QUOTE]
In my opinion the spread between different runs with different wires could be due to the fact that the thermocouples are partly heated by direct radiation and partly by conduction. The temperature and the emissivity of the wires and hence their spectrum varies from run to run. We could try to reduce the spread, at least on Glassout, in two ways:
1) Using a steel container pipe (or anything that is either reflective or opaque).
2) Meanwhile, we could try to reduce the spread by putting a radiation shield before the glass_out probe. I'm thinking about a little rectangle of metal sheet inside the tube in correspondence with the glass_out probe, which is external.
[END QUOTE]
However if you make an analysis which raises the pressure on the horizontal axis and the vertical axis the ratio between the power and the internal temperature of the gas, will discover a very interesting fact:
When operating in calibration is obtained a line to slope almost constant, while if it has abnormal emission line tends to be lower.
End of frist part
Is IR truly privileged spectrum in this analysis? Is it known that the wire does not emit energy at ultraviolet or radio wavelengths, or even visible ones? (I.e., does it glow?) It seems that any of the materials -- the quartz glass, the gas combinations, the wire coating itself -- will have differential transparency across the entire EM spectrum, making it really tough to measure *energy* in all its forms without a full spectral analysis of everything that comes out. Sounds like a daunting measurement, if not an impossible one. But maybe the physicists in this virtual room can rule these possibilities out.
@MFMP team:
I am very pleased with your methodical approach and thoughtful analysis. Let's be open to wherever this goes, even if its disappointing. It makes me proud to be a scientist. Keep up the good work.
i know a little about quartz, these crystals has all kind of strange things going on in them , some of the strange stuff may be the quartz it self, its been reported that under great psi it emits a current, so i would say you might want to try a ss pipe with maybe an glass liner and a observation site hole in the pipe.--
just my 2cents,
Then calibrate delta t to ambient with a range of different input powers to get a good feel for the temp vs power curve of the reactor.
Ie make it not see through. Maybe a glass inner lining with steel or similar outer tube if contamination from metals are a concern.
The only reason I think there is some glass absorption to IR wavelength effect here is that you see the glass temp increasing with wire temp at constant power when decreasing pressure. Ie hypothesis is: wire gets hotter, emits shorter wave IR which the glass more readily absorbs, thereby increasing the amount of heat it has to loose via convection since less IR portion of power is getting out as photons.
IMO this method of measuring excess heat with a clear glass tube looks pretty but there are far too many variables to reliably calculate p_out with glass temperature. If you want to do that you have to ensure the only means of heat loss out of the reactor is convection/radi ation/conductio n from the reactor itself as that will be largely independent of gas pressure and wire temperature, and only dependent on power out of the reactor.
Apparently you have seen both temperatures increase when the pressure drops. Theoretically this could indicate excess heat.
But, as Holmium has pointed out, it might be explained by the outer temperature sensor receiving more radiation from the inside and being locally heated.
Another explanation could be that the transparency of the glass is very sensitive to the wavelengths of the ir-radiation.
Doing the same test in calibration mode could clear out this matter.
clearskies.dk/.../...
I also despair that Celani could make such a basic error - if indeed this is the case - at the very least a control run of pressure loss would have been obvious.
RSS feed for comments to this post