MFMP: Automated experiment with Ni-LiAlH

Quote from BobHiggins

@Alan,

Yes, but this is 1/16" OD tubing. It came from McMaster Carr. I just sent them a picture and asked them to send me new tubing and 2 new ferrule sets for my compression fittings. I am sure they will respond quickly. In the mean time I have just put some JB Weld over the holes. Hopefully that will work and tomorrow I an re-start the experiment. That is, heaven forbid, I don't find yet another leak!

I can't tell from the text here, but be careful with compression fittings. Swagelok is NOT compatible with generic fittings. It may work for awhile, but not as reliable as staying within one manufacturer. You have to call them, but the folks at McMaster are good at identifying the manufacturer for individual things that they sell.

Dan

Also, the magic lookup term is seamless hypodermic tubing. They sell in gauge sizes rather than inch or mm diameter. Here's one example: https://microgroup.com/product…tubing/hypodermic-tubing/

Most of the 1/16" compression fittings I am using are ZDV (Zero Dead Volume) from Valco. However, to attach to the reactor tube end, I am using a Yor-Lok 1/16" to 1/4" adapter. Yor-Lok ferrules are used in that Yor-Lok adapter. Yor-Lok claims they are 100% compatible with Swagelok, but it doesn't matter because it is a Yor-Lok compression adapter anyway. With the ferrules being so small, the compression forces are really high and these small ferrules almost always seal well. That's why I was confused when that tube assembly was leaking because I have not had the 1/16" ferrules leak before. It is what caused me to inspect the tubing where I found the small breaches in the tubing. As I recall, this tubing is rated to 9000 PSI - maybe that is with the built in relief holes

I can't argue with something that works. Just trying to be helpful if you saw a leak in the fitting.

Good news. I received the new tubing and replacement compression ferrules today and have replaced the tube with the holes in it. It looks like (fingers crossed) that my leaks are all fixed. I am reassembling the reactor and warming up the gamma spectrometer. I hope to be able to restart the experiment in a few hours. The data will be put in this URL as it comes out:

https://drive.google.com/drive…RwT0RFeGVjTjQ?usp=sharing

The experiment is now running, the first file drop in Google drive will be in about 30 minutes.

BobHiggins

It appears to work. Here follows an example of the graphs that will come as more data arrive (at what time interval?). I think I might have again some issues with timezones.

[attachment deleted]

EDIT: it appears that the "Time Stamp" column in the CSV data has a UTC timezone offset so no adjustment is necessary after all.

round2.pdf

can

Nice graphs! Thanks for sharing the load.

I have added the first spectrum integration in the folder for that. Note that when you plot the next .csv of the data, there will be a 5 second gap in the NaI scintillator counts while the spectrometer starts a new integration. During that time, the NaI ROI counts will read 0. It is a good marker.

BobHiggins

If I remember correctly the saved spectra have a bug where although the time appears to be in 24h format, it actually is in 12h format. Since there are no AM/PM abbreviations this makes automated date parsing and cataloguing problematic. Can you confirm this? I can already see for example that the start time for the .spu file that just arrived is "04:59:44" and the end time "06:23:06". Basing on your timezone these should be either 4:59:44 PM / 6:23:06 PM or 16:59:44 / 18:23:06.

I wonder if the authors of the program are receptive to bug reports... could you file one?

can

The times you have surmised are the correct times. I am going to check into this. I know it is not a computer clock time problem because before the start of the experiments I internet syncronized the times of the two computers. I will be happy to send a bug report to Spectrum Techniques. Let me look first to find out if there is any indication of AM/PM in the data.

can

I just looked in the file and you are correct - as far as I can tell, there is no AM/PM indicator. I will contact Spectrum Techniques about this. Fortunately, the file numbering is sequenced and if we had to thread out the time, we can count the markers in the time count data for the ROI.

can

I just emailed Spectrum Techniques about the issue. I should hear something tomorrow.

BobHiggins

Thanks. Actually, if they can fix that, there are also a couple other things which they could very easily fix/improve, like:

• Saving filenames with a reasonably long zero-padded sequential number, for example "GamSpec__000004.spu" instead of "GamSpec__4.spu". In this way alphabetical file ordering would also follow their chronological ordering.
• Adding an option or command line feature to convert many binary/proprietary .spu files to .csv at once. it's not a problem when you have only a few to convert, but when you have hundreds of thousands of files it's a very tedious job. (I'm still not doing it manually but the method I'm using is potentially prone to failure).

On a different subject, here's a graph of the data so far. Compared to the previous one I've limited the scale of the Y axis of the power graph to 20W in order to make any change in steady state applied input power more apparent, since it's being cycled up-down to a base value corresponding to a Tube temperature of 140°C. No interesting changes so far, however.

round2.pdf

can

Very nice set of graph compositions. Thanks again for this work. Soon I will have to sleep and there will be a 4 or so hours where the files don't get copied across to Google drive (I am doing it manually for now). After I get a few hours of sleep, I will copy the accumulated files to Google drive.

The long period low pressure peaking seen in the pressure is due to the vacuum pump being turned on and off. The faster sawtooth is the regulation.

This part up to now is the boring drying out sequence before the main decomposition of the LiAlH₄. However in the next 4 hours, the main phase of decomposition will begin as the temperature increases slowly to 200C. I don't expect anything particularly interesting to happen at this low temperature. The pressure will go up and will be allowed to go up to about 90 PSIA and then will be regulated to track Parkhomov's pressure vs. temperature.

Actually, I can cause the spectrometer to just save the files in .csv format during setup of its multiple integrations. I probably should have done that. It would be nice if the files were indexed with the leading zeros. I will mention that to them.

BobHiggins

You could make the directory where the files are a synced Google Drive folder (using the Google Drive App), so that as soon as they are created they will be automatically uploaded, but perhaps you don't locally have the same directory structure as the shared folder.

The .spu files are very convenient for quick viewing and analysis with the MCA software, the only problem is that manual conversion to .csv isn't feasible when there are a lot of files. I guess that Spectrum Techniques didn't think of use cases where so many files are created in a single session.

EDIT: here are the latest graphs with the latest data. I think I might be soon reaching the file size limit for LENR-Forum attachments:

round2-pdf-1490951575.pdf

Amazing work you're doing here, Bob.

May I ask you a question? I take it from your comments in other places that you agree that Rossi was engaged in some kind of fraud. Do you think he had anything with the 1MW plant? Do you think he ever had anything? I'm asking because you are working so hard to replicate his technology, and yet if he was defrauding IH, maybe he never had a working technology in the first place (which Jed Rothwell sort of implies). In which case, why the attempt at replication? I'm genuinely curious as to your views on this delicate, complicated and vital subject.

So are many people I expect. But in a more appropriate thread please. Alan.

By the way, If I understand correctly the file named Experiment_20170330_Fueled.txt in the shared folder, these will be the programmed reactor temperature changes for the next few days:

EDIT: inserted clearer graph.

can,

I thought about just having the Labview program write to the Google drive folder, but it doesn't write it all at once. It waits until it has buffered about 50 samples, then opens the file and appends to it. That way, if I have a glitch, I don't loose more than the last 50 samples (sample period 1.234 sec). When I have seen files put in the Google drive folder, there is a synchronization behavior that I was concerned about not being compatible with the appending. I plan to write a little utility that will just keep an eye on the folder and copy the files to Google drive when appending is done. This also will provide backup of the files as they are created.

Got a few hours of sleep ...

BobHiggins

From my past usage of Google Drive with frequently updated files, the app seems to work fine with them, but your mileage may vary. You shouldn't end up with partially uploaded files.

Here's the latest experiment graph with the almost-latest data:

round2-pdf-1490965917.pdf

It's nearing 2 MB so I may have to use a different format soon.

joshg,

Well, people are certainly polarized when it comes to Rossi. I have to admit that it was his early work that got me re-interested in cold fusion. I have never met Rossi and his behavior is that of a scoundrel. However, I have a great deal of respect for Focardi, and Focardi measured gamma emissions and believed Rossi had excess heat. Rossi's work sprung from what he had read of Piantelli. Piantelli has working Ni-H LENR ongoing and is a scientist's scientist. He lives by Galileo's rules for the scientific method. I have heard that one the US Naval Reasearch labs had one of Rossi's reactors (vintage unknown) and concluded that it produced excess heat (amount unknown). From what I have seen of his work, his engineering skills appear weak and his measurement skills are terrible. Even with all of the analysis I did on the Lugano experiment, I still came away with a conclusion that there was probably excess heat there in the same amount that Parkhomov has seen. In addition, Mizuno, Jiang, and Zhanghang have all published positive measurements of excess heat in Ni-H. I believe I have locked into a hypothesis of why some Ni-H experiments succeed and some fail. I am going to play with this hypothesis some, and if it looks like it has feet, I will write a paper and tell everyone.

So, I believe Ni-H is a LENR domain which is capable of excess heat. I believe Rossi probably demonstrated some excess heat in the past. Rossi's current experiments are so clouded with deception it is hard to know whether he still has excess heat. I note that Rossi seems to like the limelight. He talked a lot about what he was doing in the early days, but when he went commercial he was forced to say a lot less. Between the less description of his experiments he has provided, his poor science and engineering skills, his natural tendency for deception, it is hard to know if any of Rossi's current reactors produce excess heat and how much. But, that's OK, because there is still plenty of evidence now that Ni-H LENR real (despite naysayer Rothwell). Rossi is no longer needed to fan its fire.