MFMP: Automated experiment with Ni-LiAlH

  • BobHiggins has just started a fully automated experiment:

    The GO button has just been clicked for an automated Ni + LiAlH4 experiment. Whilst Bob [Higgins] will always be able to intervene, but if he doesn’t, it will take 6 days.

    The experiment contains 1.02g of pre-processed AH-50 carbonyl Ni powder and 0.11g of LiAlH4. That was all that he could fit in the 2″ end of my smaller ceramic tube.

    There is a very slow temperature rise and the system will not be >600C until late tomorrow. The script includes soaking and evacuation cycles to clean out residual water vapor and O2. Bob will be putting the data files, photos, and and videos in the following shared Google drive folder.

    Maybe BobHiggins can tell us something more about it here :)

  • The experiment was aborted last night due to a leak, apparently at the reactor tube metal seal interface. I will be reworking that today. The experiment did not get very far into heating - only to 140C. I will rework the reactor tube seal today and re-install. Hopefully the leak will have been resolved and the test can proceed as planned.

    The experiment uses AH-50 Ni powder that has been processed to remove oxide prior to mixing with 10% by weight LiAlH4. It is a Parkhomov-like reaction that is insulated in the K-26 bricks to improve the sensitivity to excess heat.

    When the experiment is re-started, there will be a new folder for link to the data.

  • IMO it wouldn't work - and if yes, than just by accident. Try to use magnetic field, high-frequency plasma, electric impulses...

    BTW The magnetron (microwave oven) could heat the sealed pipe reactor and to create some plasma inside it. This is what I'm going to try myself - watch the nuclear shrooms..

  • The pressure is computed from columns 12 and 13 in the data. Column 13 is the voltage across the sensor and column 12 is the sensor voltage. The equation applied to this is pressure = (100/(0.9*C13))*(C12-0.05*C13) and the result is in PSI absolute. The result in the beginning is the local atmospheric pressure of about 10.6 PSIA. The sensor was working properly, but the computer controlled back pressure regulation was not working, in part because there was a leak at the reactor tube. Circuitry problem was fixed while running, but the leak was not fixable without taking the system apart.

  • To make sure: are columns 12 and 13 respectively "Pr (V)" and "PrS (V)" ?

    With the equation above and those columns I get this result, with pressure starting at 15 psiA, which seems about right but not exactly what you're reporting.

  • You will find some pressure variations. I was trying in later hours to repair the circuit on the fly including fixing the USB interface back pressure regulator circuit (while live!) and then the plumbing afterwards. After I got the regulator working as it should, I discovered that the residual leak was too big to fix while the experiment was running. So I aborted the experiment to disassemble the reactor tube is repair its seal before the fuel was damaged.

  • With the equation in comment #8 I can confirm that pressure starts at 10.6 psiA.

    Since you're starting a new experiment with new data files in a new shared folder I would suggest adding a column with directly the computed pressure value.

  • Those .csv files are created automatically by Labview. I don't know how to add the equations into the .csv files. I like to leave the original raw data in the .csv files. Normally when I am processing the data, I read that into excel and add those columns and others. For example column 11 is the proxy for the line voltage and I have an equation for converting it to ACrms. What I should do is put a prototype excel file into the folder as well into which the .csv data can be copied and which will have the other columns.

    I am just 1 person and there is a lot to do. I had to learn to program in Labview and write several drivers to make this all work together. It has been a long haul.

  • I can batch-process the .csv files semi-automatically without using a spreadsheet like Excel, so I don't really have a problem with having to calculate the actual pressure or other parameters. However, having to hunt for the proper equations to interpret the data correctly could make the data less accessible for most people, especially if they're not following the experiments too closely.

  • Bob,

    My guesses why Ni - LiAlH4 experiments are challenging to reproduce.

    1) Every specific brand/type of nickel probably requires a different degree of pre-processing. A prep routine for brand/type A won't work (or be sub-optimal) for brand/type B or even for the same brand/type that has been exposed to atmosphere (after repeated bottle openings.

    2) Related to number one, it may take a few or even dozens of processing attempts to find the optimum pre-treatment for a sample of nickel powder/wire. Most people probably give up too quickly. Then they switch to another brand/type of nickel or add an extra additive before figuring out how to pre-treat the original fuel. Eventually, they may produce significant excess heat by mostly luck but they won't have a solid understanding of the parameters required. Baking in atmosphere (for intentional oxidation), vacuum degassing, flushing with hydrogen (chemical reduction), surface treatments: there are many durations, temperatures, pressures, and other variables to test. The work is hard and tedious.

    3) On a practical note, Me356 seemed to focus on cycles of vacuuming and flushing with hydrogen until his nickel wire could breath so much hydrogen (in a very short time period) the resistance would go up or down 40%. I've also read and heard (even privately) of multiple accounts where extended, deep vacuuming under heat was critical for the effect to show up. The problem, as you've pointed out in the past, is the sintering of the nickel powder. If the degassing is done at a low enough temperature so the nickel powder doesn't sinter, the diffusion rate of oxygen and CO2 out of the lattice will be reduced.

    4) A lot of commonly available LiAlH4 is of very poor quality. Me356 claimed his Alfa Aesar 97% LiAlH4 seemed to emit twenty times the hydrogen as the LiAlH4 he had been using. Also, a few minutes of exposure to atmosphere can allow LiAlH4 to absorb a LOT of moisture that could result in oxygen going into the reactor. LiAlH4 also bonds with nitrogen in the atmosphere. If someone had a high quality professional lab and was qualified to do so (no one except a trained expert should try this) they could dissolve and purify a lower quality LiAlH4 into a 99.9% pure sample. I also wonder if by controlling this process if the particle size of the LiAlH4 could be made smaller.

  • @TDIU

    I agree with most of what you said. However, don't expect Ni of any kind to absorb (volumetric) much H2 - most of the H2 is adsorbed onto the surface. A properly activated Ni surface will catalytically break the H2 bonds into H+ and H- ions adsorbed onto the surface (re. Piantelli). The surface type 2D (foil, wire) is much easier to clean and activate than the carbonyl powder, but the carbonyl powder will have much higher surface area. I think the hydroxide formation common in the LiAlH4 (darkening) is quickly dealt with in vacuum cycles during the low temperature breakdown. The hydroxide forms H2O vapor and is vacuumed out in a properly designed heating sequence. LiAlH4 is LiAlH4 - I don't understand the comment about some being better than others. If it is not LiAlH4, it is not going to decompose when it is supposed to. Somebody must have gotten some oregano instead of the real stuff. But, you can be sure if it comes from Alfa that it is the real stuff.

    Ultimately, I want to connect my UH vacuum system to the reactor and sample the gas during the experiments. I want to use the RGA to see how the gas composition evolves. Eventually I want to get an ultra-high dispersion light gas RGA that will discriminate between the D2 vs the 4He and the HD vs the 3He. These high dispersion RGAs exist in the $15k range new (they just have a high frequency quadrapole mass analyzer - not that much design change).

  • can

    Note that the most important temperature is not shown in your graph. That is column 8, the k-type thermocouple temperature of the reactor tube. At the end, that temperature was 140C. This is the temperature that is regulated by software PID control. In my script, I tell the system for example to H[old] at 140C and the DC voltage is changed under accelerated PID control to go to that temperature and regulate. But the temperature that is being regulated is the temperature of the reactor tube.

    Column 15 is the ROI (Region Of Interest) counts from the gamma spectrometer - essentially any count coming in at any energy makes a count here. Think of it as a very sensitive gamma count. The spectrum is an integration over 5000 seconds, but the ROI counts will show if the energy was a burst or was constant.

    Nice work on plotting the other data.

    I am heading down to the lab to re-install the repaired reactor tube. I will post the other equations for line voltage, heater voltage, and heater current.

  • If Rossi can get his QuarkX to work in a plasma, does fuel preparation have any impact on the LENR reaction? R. Mills in his SunCell does not do any fuel prep in that reactor...just get a dirty plasma going and the plasma takes care of itself. Are you'll still chasing your tails here?