Test Cell #2 - Calibration, Data, and Getting ready to fly.
Over the weekend, Celani Test Cell 2 ran through a complete calibration cycle with Helium. We don't necessarily plan to use this data directly, but it is good practice and proves out the hardware and software:
Channel 1 - (blue wires) -12 points
Channel 2 - (red wires) - 12 points
Both Channel 1 & 2 - 12 points
I will post the excel files as soon as somebody asks. I'm just not sure the best way to post them, yet. Or are there people watching this interested in the raw CSV text files, too? Let me know in the comments.
The 12 points are loaded towards the low and high ends. We set it like that when we did a test of the linearity of the power supply and just haven't decided to change it, yet.
At the end, when both channels were on, the cell topped out at near 430C and held it about seven hours till morning. I would call that a successful stress test. Looks kinda dramatic in this picture, doesn't it?
During this calibration, the T_ambient sensor was too close to the safety shield and measured way higher than it should have. I would re-run the calibration if Mathieu and Nic in weren't itching to do their own European calibrations so soon.
Another interesting and concerning thing we noticed is some hazy condensation on the inside of the glass near the cooler flanges. Our best guess is that this is either a silicone binder from the mica or possibly some oils from the machining fluid that soaked into the mica. We haven't decided what to do about it, yet.
None the less, we decided to pull the whole kit together on a table for a nice photo shoot. We are putting the final touches onto the laptop for remote access, doing a short training video, getting to work on the webcam, and making the generic gas/vacuum manifold. Despite the crunch to get it all done, it is really rewarding to see the whole thing coming together.
Sorry it isn't totally focused. You can expect better tomorrow.
Comments
In order to avoid (or reduce) that, would be useful:
a) using taller lateral supports that increase the vertical distance between the cell and the horizontal support;
b) using not metallic supports or, alternatively, build them with thin rods instead of "full" metal;
c) the same for the horizontal support. It shouldn't be metallic or should be covered with materials that don't accumulate heat or that don't reflect heat toward the cell;
d) performing calibrations for long periods of time (many days) in order to estimate if the problem exists.
In particular long time calibrations are very important to see if there is a thermal drift and to estimate the influence of the room temperatures in the day/night cycle.
Ecco, for the time being we are focused on just replicating a simple experiemnt with hydrogen and will leave the derivative experiments like playing with dueterium to the researchers we will hopefully inspire.
Also, with Deuterium, he detected X-ray burst activity after the loading phase, when applying power. That will be extremely interesting to see and check out.
As example, if 50 watts is applied to the inactive wire, does the active one get heated as though it had 10 watts applied?
I would like to attempt to separate hydrogen gas temperature versus wire temperature effects on the excess power.
By the way, if I understood correctly, both NiCr and Active Constantan wires are in the reactor right now?
Is it ok to apply that much power directly to the active wire? It might not be the case of Helium (which I think you are using right now), but Celani limits input power to 30 watts when using Argon gas due to sintering problems when applying high power to the active wire under such gas (used by him in 100% concentration during the calibration phase).
In general it looks like you're using a much higher input power than what Celani has shown in his slides so far (where the limit was 48W. He performs calibration at steps of 5, 15, 30 and 48 watts)
Please disregard if you're not using an active wire yet inside the reactor.
Are you able to produce a graph of reactor temperatures at a constant input temperature over many hours?
For the sake of avoiding controversies, next time you will perform a calibration run also try making each calibration step completely stabilized in temperatures at least in the short term.
RSS feed for comments to this post