In the long week since my last post, we have been doing a lot of dry, boring calibration.  We have also done a lot of other work on calorimeters that I will now share.

A Calibration to be Proud of

In last week's episodes, our intrepid heroes had been battling the horrible, black, goo monster that came from the new mica pieces and collected on the Isotan wires in Cell 1.0.  We managed to cook it all out, as far as we can tell.  In the process we got enough data to form a good comparison baseline for the cell in Helium.  Once we load the Celani wire, we will run it in Helium to try to demonstrate that it is installed in a comparable manner to before.

Then we moved on to calibration runs in Hydrogen at 1.0 bar starting pressure.  The big question was whether the reconstructed cells with new glass, new wire supports, and all new wires inside would perform the same.  We let the hydrogen tests run over the weekend so we would have several runs to do statistical analysis with in case it proved to be different.  It turns out, to our relief, the cells performed extremely close to the way they did before.  This gives us a strong confidence in our cell's ability to behave in a very reproducible way.  Below is a comparison for cell 1.1 over the whole range. 

Here are comparisons for each cell in the range we are likely to run in.

On both cells, see how the new lines for each of the sensors fell very close to the original, and well within the confidence interval.  We decided it was close enough to combine the two sets of data and come up with a single averaged curve with confidence limits and derive a calibration equation from the most stable.  The result is a calibration that we are feeling really good about for this apparatus and it's ability to tell us we get more than a couple watts of excess energy from the Celani wire.

What this tells us is that over 14 calibration cycles, the temperature of the sensors averaged out to the lines indicated.  The variation in the data points (27 data points for each point from the ups and down steps) had a 99% confidence limit at the top and bottom dotted lines.  That means, as long as the system isn't altered some other way, temperatures above the dotted line would strongly indicate that there was excess energy being created.  We will be plotting all 3 temperatures vs input power as we watch this.

Here is all the data in spreadsheets.

From these graphs, it appears the uncertainty in power output will be in the 1 to 2 watt range.  Any opinions?  Who wants the data?

Next Step:

Load the Celani Wire into the cell (in process) and pull a vacuum on in before loading with Helium for a comparison run.  Then load with Hydrogen and attempt to trigger excess energy.

And more blog entries to come about the insulated stainless steel cell we are about to commission inside the air flow calorimeter.