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.