After a long delay as we failed to troubleshoot the water flow calorimeter, the Multi-wire test has been installed in a concentric tube calorimeter. It is now heating up and the first of the 3 wires is starting to drop resistance.
This is the test with 3 lengths of different Celani wires in a LENR-stick test cell. All the details are in the protocol document here: Protocol: Multi-Wire Test
Previous blog post: Multi-wire test commencing -Update4
You can follow the data on Test FC0405 LENR Stick: Multi-wire and FC0403 CTC #2: Air Jacket
From Malachi:
We are starting to see the 270L wire loading. The internal temperature is ~208C. The interesting thing is that neither of the other wires (350L and 400L) are dropping in resistance yet. Could lower numbers of layers correlate to a lower loading temperature? This test will be an interesting one in the coming weeks!
The other two wires (a 350 layer and a 400 layer) are actually increasing over time at this temperature.
Comments
I agree that more loading is needed. Currently the wires are set up only for resistance measurements. It would require another power supply and this is something I also want to see on the wires. This could happen very soon.
What is next for the multi-wire test? Does anyone have an idea?
We've run it up and down through our temperature range, tried to deload the wires and we've tried to reload them.
They don't seem to be decreasing any more, in terms of resistance.
It looks like each wire has its own optimal loading temperature range. That of the 270L wire appears to be rather wide (in one of the latest blogpost updates it was observed to decrease resistance at an internal temperature of 208°C while other ones didn't yet).
This also means that there's an optimal H2 desorption temperature range too, depending on wire characteristics , which might have some interesting practical implications.
The calibrations look sound. We are going to increase the hydrogen pressure inside the cell. Then we will try to load the wires as far as we can.
But I guess we'll see in a while.
I will set the input power to 35 watts and see if it still holds.
y = a + bx + cx^2
Fitting target of lowest sum of squared absolute error = 8.8798690830743515E-03
a = 3.7115236254221067E+01
b = -3.2643234519158354E+03
c = 1.2266525881480571E+04
This is the new fit equation. We can plug it in and see how well it fits. If it does have a positive bias, then we will perform extra calibration cycles, in a decreasing fashion perhaps.
Hopefully, perhaps with the aid of the quicker power cycling, temperatures will keep decreasing during the next calibrations, showing that what was observed during the active runs under hydrogen and this prolonged vacuum run was actually real.
i.imgur.com/dPDL7vs.png
Weird, isn'it? I tried adjusting data to the previous runs at a lower outer tube temperature, with a second order polynomial curve, and this is the result:
i.imgur.com/GBbTK4M.png
I wouldn't take this too much seriously... but it's interesting data nevertheless.
i.imgur.com/4VBPjF4.png
As a side note, it appears that wire resistance is increasing over time noticeably faster than before, after increasing temperatures (besides the immediate increase due to higher temperatures / PTC behavior of the wires at this stage).
The inside thermocouples are free floating. They could be shifting when the wires (separated by fiberglass sheathing) get hot and deform. Just a thought, but it could explain the difference in temperature rises between runs or power levels.
Outer tube temperatures haven't budged yet though, which is a good thing.
i.imgur.com/lbpI3gR.png
With the point being: once input power is high enough, if there really is exponentially increasing excess heat, it will be very noticeable. It shouldn't be visible only through internal temperatures.
Don't take them too much seriously, they are a moving target and will get adjusted as new data comes in :)
By the way, it was so fun toying with the idea that I tried further imagining / figuring out how temperature would progress after 44W, with the hope and assumptions I mentioned. Under that scenario, at 58W of input power the cell would exceed 1000°C of internal temperature, although I haven't taken into account that it would likely go into thermal runaway somewhere before that.
Here's the full data set if you really want to play with it:
Pin [W] T_Int1 [°C]
34.000 486.30
36.000 509.00
38.000 534.00
40.000 562.00
42.000 593.00
44.000 627.10
46.000 665.00
48.000 707.50
50.000 755.00
52.000 808.25
54.000 868.00
56.000 935.00
58.000 1010.00
i.imgur.com/PHF1efJ.png
At low power it takes more effort for the cell to increase the temperature of the inner tube relatively to the outer tube. Normally, the lower the temperature is, the lower the effort required should be.
This was the same graph during the first sweep under vacuum and no cooling hose added:
i.imgur.com/lmZacu9.png
Another, likely related strange thing is that at 0W of input power (+ a few milliwatts applied to the wires) the outer tube has a slightly higher temperature than the inner tube. This caused calculated output power to be negative (around -0.5W).
i.imgur.com/Vugp8J2.png
More data points needed, especially at lower input power, in order to make sure that a single percentage is enough to adjust data to the new outer tube temperature. I feel it isn't because I would have expected the calculated excess power at 30W to be more than that.
However, I wonder if adjusting them with a fixed value to the average output power curve of the previous three active runs at 46°C, would work. If it will, the resulting output power curve at a higher outer tube temperature shouldn't diverge significantly or at all from them (at least for the input power range where outer tube temperature wasn't affected).
I think this would be an interesting test to try, if you don't have anything else planned for this cell for the time being. I think it's a good trade-off between turning the temperature compensation algorithm off and using only at mid and low input power.
RSS feed for comments to this post