The US team has begun the loading phase in Cell A. We just loaded 3 Bar of H2 into the cell and turned the power on to 30 Watts. It will stay that way for a day or so while the initial absorption happens. Then we'll refresh the H2 and cycle it 6 hours on, 1 hour off. We are excited to see if our wire can achieve a similar (or better) resistance change compared to Mathieu's. The race is on and it's down to the wires!
UPDATE 21-06-13 - More troubles than I can shake a stick at
The US cell's loading has been plagued by several minor difficulties that make the data less than attractive to look at. First, we hadn't turned up the power on the active wire far enough to get a clean baseline resistance reading. Then the temperature with the hydrogen in the cell only barely got into the loading range. Then when went to turn it up, we found we had to adjust our max power limit (a hardware circuit we had installed after the last time we toasted a wire). When we adjusted that, we bumped a ground wire on the power meter, causing the reading to go wile and the power output on the NiCr wire to go to max. In that process, our output port controlling the Celani wire power was toasted, and we had to move that to a different output pin, reconfigure the software, and reload. After that, over lunch we had a power blink, which reset the power outputs. Then we found that the second tier software power limit was getting in the way of achieving full power on NiCr wire and had to reset that. Finally, we thought we had it all going and then one of the instrumentation boards started blinking out every so often and dragging the other one down with it. When they reset, the power outputs reset also, and the cell cools down. We are still troubleshooting that one. -- Update-- That was found to be a power wire issue caused by the data acquisition boards being powered from different sources. Paul and Malachi found it by watching the power chip on the board heating from 40C to 150C before it blinked out. Turns out it has been wired poorly for 8 months and has worked ok till now. Pretty resilient boards.
That is all pretty crazy for a system that has basically been up and operating for a couple months as we troubleshot the ambient environment and performed the calibrations.
While the data is not the cleanest, we are definitely seeing a drop in resistance from roughly 16.2 to 15.1, or so. Not as big a change as Mathieu saw.
The next step is to begin the cycling (35W, 3 bar H2, 6 hours on, 1 hour off) for most of the weekend. Since Mathieu has other obligations for the weekend, we should be caught up on Monday and ready to move forward nearly simultaneously.
Here are some pictures from today:
A photo of the wires in the cell. The Celani wire is looking very light and coppery colored. The NiCr wire is looking almost blue-grey.
Below: Wes has the interns helping set up to measure something on the powder reactors while Paul and Wes troubleshoot the latest instrumentation mystery.
B
Below: Paul and Malachi watch for an intermittent problem with the thermal camera.
UPDATE #2 SEM Pics now in Ignite Gallery
After some much-needed (and time-consuming) maintenance, our SEM was available for a closer look at what we're testing on in the US1.3 experiment. How timely!
Head over to the Celani V1.3 production gallery to get a glance.
There're also pictures of the metamorphology produced by adding acetone to the CTC Celani wire in the gallery of all things related to the hydrocarbon anomaly we've been experiencing.
Notice how poor our resolution is? I did a LOT of doctoring to amp-up the contrast and clarity of the resulting images. We are in desperately short supply of an experienced SEM technician since we lost ours to a walkabout in Australia 
. It seems nano-fine tuning of a 20-year-old JEOL SEM is more of an art than the technical manual lets on. If you or someone you know could be of any help in working with an SEM, drop us a line! We're exceedingly good at utilizing remote instruction and could really use the remaining 295,000X resolution this machine is capable of. It's quite dated but that's no excuse for the limits we're experiencing.
Comments
i.imgur.com/WEso0m2.png
US Cell B
i.imgur.com/SluQEAD.png
Whatever it's causing this, it's affecting US Cell A slightly more than US cell B.
It also shows something very interesting (see red arrow). For some reason external temperature has increased a bit, even though pressure is decreasing and convection toward the glass tube should be decreasing. Furthermore, it doesn't seem this increase is correlated with changes in the data I can see.
Could this show excess heat generation as the wire gets loaded more deeply? Or is it just an artifact (it's a very small change after all) due to something else?
Wearing back my skeptical hat, have there been changes in wire appearance/colo r or apparent glass tube transparency in the past hours?
EDIT: the control cell unfortunately is currently not powered, but its ext.temps seem to show a similar pattern. So maybe it's an artifact. Or somehow one cell is affecting the other one.
A day later... Wow that is beautiful. Looks like temp cycling is not needed.
i.imgur.com/FuRBjXA.png
Does this ring any bell?
It does to me!
BTW: US Cell A with 44W of heating power is currently behaving as I previously written. 45W will improve loading further.
Then, as time passed, hydrogen started getting heated too, which improved absorption in the active wire noticeably.
If this makes sense to you, perhaps it would be useful to heat the hydrogen atmosphere more efficiently rather than the wire itself, for example by keeping the gas actively moving around inside the cell. Things could get really interesting.
Just a hunch.
From all experiments seen so far, it looks that wires respond faster to heat than hydrogen absorption.
What I think happened is that when you increased power all at once, wire resistance increased noticeably, but chances are that if you had let it run, after a certain time/threshold you would have seen resistance decreasing in a way similar to when you load a new Celani wire for the first time.
By the way, would it harm the cells if you temporarily disabled forced air cooling, to increase temperatures without increasing input power?
Yes the gas leak is quite obvious.
@Edwin
Good question, but it does point to the importance of having multiple experiments.
@Maxwell61
The passing of Sergio Focardi saddens us greatly, his contribution will never be forgotten and hopefully we can all here play a role in realising the potential of the technology he helped advance.
There is also a correlation between ambient temp and power input. I hope the wire is still in working condition.
EU is still showing a good resistance drop, not flattend so far.
I assumed that input power was set to 35W before the spike (like the EU cell). If it got increased just after that however, then that perfectly explains the difference in temperatures and pressure afterwards. Sorry for pressing you on that.
I guess I didn't pay attention enough. But I think I can tell you some of the reasons why:
- Spikes in data make changes harder to see. It would be cool to have an option to filter them (or to not make the autoscaling algorithm account for them) without having to use the recently implemented zoom function.
- Wa_NiCr(W) in the live data viewer has a very pale cyan color that is very hard to read on certain displays.
In comment #10, the before and after difference is accounted for by the power to the heating wire changing from 32.27 (the max hardware limit to protect the wires) to 35.0 watts (the set point after we adjusted the hardware limit up to almost 40 W).
Thank you for your awesome attention to the data, though. I wish the process had been smoother for us and easier to understand.
i.imgur.com/UNymp2s.png
Was a global offset applied or did temperatures and pressure actually suddently increase for some reason?
RSS feed for comments to this post