Finding the right assistance is one of the wonderful aspects of community involvement.
I would be glad to assist in looking at DOE when the appropriate time arises! It would be great to contribute in a constructive way albeit small compared to what others are doing!
I will be even MORE glad when positive and repetitive results are realized! That will be a day to raise the glass and smile indeed. 
Just to make sure you all notice... the latest graphs from can are on his cloud storage folder: https://mega.nz/#F!XpQwQIrL!UWg9jMHh4I__eK33-d5x8Q
I will be copying these to his folder on my Google Drive as well. They may not get posted frequently in this message list inline with the messages as before.
Yes, besides nothing interesting seems to be occurring yet so to me it feels a bit redundant to upload them here too while they're being regularly uploaded there.
In the meanwhile I tried arranging the graphs into a different multipage document format (which I have not yet applied to those currently being saved in the shared folder); I don't know if this is more useful for others. It's intended to be viewed in full screen with one page per screen.
can ,
I like the new format.
We are just about to head into the more interesting region. Operation at 1200°C will surely reduce the life of my heater coil, so I have avoided it as much as possible previously. What is probably worse than 1200°C is an overshoot when reaching that temperature. I haven't lost a single heater coil yet though.
If prolonged high temperature operation will reduce the life of the heater coil, why about half of the experiment time is >= 1150 °C? If the objective is seeing if any effect arises at a high enough temperature with the current conditions, couldn't those temperatures be reached quicker? This is among the things I previously wondered when I asked about the rationale behind the experiment schedule. It might be worth noting that from Parkhomov's latest email posted by Alan Smith, it seems that in his case after a dwell time at 400°C temperatures are "smoothly increased" to "1200-1300°C"; it doesn't sound like it's a very long process.
On a different subject: I attached a slightly different version of the previous graph, which is still not the same as those being uploaded in the shared folder on mega.nz.
When I was conducting NiH experiments, I developed software based custom PID algorithms. You could set how fast you want the temperature to rise between each temp level. This virtually eliminates overshoots and and excessive input power that damages the heater coil. I also used a step up and step down algorithm to repeatedly take it up and down in temperature ranges. There is a way of running it this way that allows the reactor to serve as its own control. I can explain that more if interested. I used a windows desktop app in combination with an Arduino board for running these experiments.
Thanks. I have implemented the PID control in Labview. I find that the coefficients need to change between low and high temperatures due to the change in loading at high temperature. It is tuned to be lower overshoot at higher temperatures right now. I could also make it so that it even less likely to overshoot at high temp by adjusting the coefficients. I could even make the coefficients a function of temperature, but just haven't had the time to get to that. I am only really worried about the upcoming sharp steps between 900°C to 1200°C because I haven't done it before.
Yes, you are right in the observation that I am spending a lot of my time >1100°C. In the end, that is where the action always shows up. Sometimes it takes some time there to show up - and stimulus. But XH never shows at lower temperature with this fuel system. While the temperature change from 400-1150°C is not a ramp, it is in steps with some soak - particularly at 680-700°C where the fuel is melting. In the end, it is not far off the 2°C/minute mark for this run. I can't imagine that conforming to a linear ramp is actually required and it is fraught with measurement inaccuracy to try to force to that ramp. Using the steps with settling allows that check for XH at each temp.
BTW, I am seeing some bursts show up in the GM detector. I wonder if it is low energy gamma or x-ray. The GM tube will pick that up, but the scintillator may not. There are more peaks showing up in the scintillator, but not as prominent. May just be my imagination.
Also, I like the newest chart best.
BTW, I am seeing some bursts show up in the GM detector. I wonder if it is low energy gamma or x-ray. The GM tube will pick that up, but the scintillator may not. There are more peaks showing up in the scintillator, but not as prominent. May just be my imagination.
I'm hoping you'll find time to do that extended null run at some point.
If the slow ramp has to do with the decomposition of the ionic hydride as others have suggested, I suspect it would not be useful to keep further increasing temperatures slowly after there's nothing left to decompose.
I've checked the larger graphs but I'm not sure if I'm seeing gamma count bursts. I've attached below the latest one (page 3).
(Besides what I'm still regularly uploading automatically on the shared folder on mega.nz)
I am only really worried about the upcoming sharp steps between 900°C to 1200°C because I haven't done it before.
Bob,
The temperature overshoot per se is not the biggest danger. Depending on where you are measuring the temperature, the input power spike during the run up will cause the heater coil to become much hotter than where you are measuring until the temperature equalizes. I found this is where wires most often melt.
Jack
Yes, you are right in the observation that I am spending a lot of my time >1100°C. In the end, that is where the action always shows up.
BobHiggins : Do you have no possiblity to alter the pressure inside the reactor? And what would be the lowest you can get (in Torr). The former mfp signal usually showed up after a pressure reduction at medium Temp.
In the latest graph, the lines going horizontally through the ROI and GM counts are a great idea. It looks like both the counts are going up on average. We will have to take a close look at the spectra. While I wish for XH, the data doesn't seem to show any.
Yes, I understand exactly what you are saying. In the case of my heater, the wire is pretty well coupled to the inner alumina tube, which I hope buffers it somewhat, but at minimum the coil temperature will have a bump at least equal to that measured (probably more as you say). Until I can reasonably fix this (software), I can limit the steps in the script. I wanted to try some fast steps here to see if it stimulated activity, despite the danger to the coil. If the coil dies, it is just some work - not a tragedy.
I can at any time lower the pressure of the H2 being generated within the cell, but at the moment I have no means to control the H2 pressure to increase. I can program the back pressure regulation to a higher value and not hold it low in the face of internal H2 release, but I cannot force it to higher pressure. Regarding the lowest vacuum that I can achieve, it will probably be about 15 Torr by the time you take into account the continuous residual release of a small amount of H2 and the fact that it is being pulled through a 10 micron orifice. In this last couple of hours, the pressure was reduced to about 25 Torr at ~09:00UT and remained around this pressure with the temperature at 1200°C for one hour. Then the pressure was allowed to rise as it will because the back pressure regulator was changed to 90PSIA - basically meaning it wouldn't pull out any more gas.
We are now heading into the less interesting (probably) cool-down cycle that will last for about 10 more hours. The cool-down plateaus allow for a check of the behavior of the system and the thermocouple at the end of the experiment. If there is no XH, the settled control power for each step down should be about the same as on the way up.
I should have metallic Li arriving early next week - perhaps in time for the next experiment. Also, I am working on a large square multilayer copper coil that will have an ID large enough to go around the whole K-26 insulated reactor assembly. With this coil, I plan to be able to at least manually stimulate a strong magnetic field - DC, and probably damped sinusoid pulses (charged capacitor switched to the coil). I have a power supply that will allow me to put 20A of DC through the coil. I think I will make a switch detector so that even if the pulses are stimulated manually, they will be logged in the DAQ for reference with the other readings.
Please keep the suggestions coming.
There are two places where I go particularly slow - around 180°C and around 680°C - where LiAlH4 decomposition and LiH & Al melting occur. In these regions I have a long soak to allow the gas to evolve and/or melting to occur. Beyond this, I have been trying to limit the steps to about 2°C per hour on average. But at each step, I want to have time for the temperature control to completely settle so the system can be compared against the calibration. Do you believe that other temperature regions need a slow traverse as well? Or are you saying that you believe that outside these regions that the temperature rise should be faster?
In the latest plots (Thanks!), the gamma seems to have increased where the pressure was low.
BobHiggins (#415)
I've added the horizontal lines to make any change in count rate trend more visible. Specifically, the horizontal lines are placed at the count rate average values for the displayed data. From longer term data (see attached file) however there don't appear to be significant changes overall.
BobHiggins (#417)
I am indeed saying that outside the regions where the ionic hydride decomposes the temperature rise might probably be faster, but this is only my hypothesis; I have not been provided detailed information on this regard yet. I do not think that temperatures necessarily have to increase with a linear ramp, it's probably more important that in general the hydride decomposes slowly to take advantage for longer that the nascent H atoms will at least for a short period of time be either in an excited state or ionized. So, the rise may for example be executed in practice with small constant steps. Again, my opinion.
I tried with Firefox, Microsoft Edge, Chrome and an external pdf viewer, and it seems to work fine on my PC (using Windows 10 64 bit).
Maybe the resolution is too high for older PCs? (I recall you mentioned you have Windows XP installed)
I will download it and check it in Acroreader outside the browser (Firefox on Win7). The laptop I am using to take the gamma spectra is an XP laptop. The main control computer is Win7-64b 8 core 4GHz 16GB ram.
Have a look at your cloud file hcletch_new_20170505_145304.pdf in the region between 11:00UT and 12:00UT. Look how the averaged ROI and GM counts rise above the baseline.
