FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

New generation Celani wire experiments. UPDATE2: Measuring the flow

Written by mat on .

It is with great pleasure that we finally have the opportunity to test the latest generation of Celani wires.

These are the electrically/joule heat cycled, knotted, at chemically treated wires that are sheathed and bound.

Mathieu has worked hard to prepare everything for this series of experiments, which will include testing in the standard active and control dual Celani cells and the now calibrated and instrumented MFMP Mass Flow Calorimeter (MFC). The Celani cells will predominantly be looking for signs of Gamma emissions, the MFC will be looking for any signs of excess heat.

You can get up to speed and keep up to date by following the live document published below. The live data will be published into this public folder.

Thank you to all our generous donors over the past two years that have made construction of the MFC possible. Thank you also to National Instruments for provision of the DAQ and LabView License.



UPDATE: Understanding the Flow-Control

I keep saying to people it is too early to say, but we have passed the P&F milestone of skepticism.

History keeps repeating, in March 1989, Stanley Pons and Martin Fleischmann did exactly what we are doing with Live Open Science. They opened the door of public knowledge about their discovery to the general public, and soon after the crowd was rejecting what was brilliant piece of calorimetry. Much better than what we are dealing with right now by the way. Their system was so precise that solutions of specific differential equation were necessary but never done at their time! They truly revolutionized calorimetry at the same time.


To get back to our matter, there is much difficulty to get everything explained in a single comment lost in the discussion. So here I am.


In flow calorimetry, flow is as much important as reading temperature. And only scientists who are trying to resolve the hundredth of a degree with precision and repeatability are able to tell you how hard is that. Well, it is the same for flow control.

I got in contact with Arik El-Boher from university of Missouri in June 2015 on matters about flow control because I knew this is an important part of the system. His technician and himself were very much encouraging about the use of a Bronkhorst L30. One plot is better than tons of discussions. There is the graph they sent me:

...just look at the numbers on the time scale (X-scale), I think it is in minutes.


A quick quote on phone with their available commercial in Paris gave me the public price of such equipment. This was way over any possible affordability for the project. So I wrote to them and other manufacturers, competitor on the market. It was about the mass-flow calorimeter, the project, my work, our field of research, Live Open Science and how we can work together. Even thought I had two replies, their answer was the one positive.

After few exchange and discussions we agreed on satisfactory terms. That would imply spending time on counseling to their engineers. This is ongoing, and a very fruitful collaboration is born.

The system is not a pump, the moving part is a valve placed behind a flow-meter. You have to supply pressure to the system to make it working. To do this I used the integrated pump inside the cryostat to supply sufficient pressure. Additionally, it measures the flow rate using the thermal expansion of the fluid. Precision is 1% but repeatability is 0.2%. Repeatability is the characteristic here. If we want to have the right comparison between calibrations and the test, this plays a key role.

When I received the system, it was simple and straightforward to put in operation. Supplying 24V to the logic board, then connecting RS232 and running the provided VI showed me how much this controller is the Roll-Royce of flow control. Variations were just imperceptible on the front panel of the program. Then I implemented key parts of the virtual instrument in the main program that runs the calorimeter. Within the program I had to increase the number of decimals to 10^-5 to see any decimals in the readings.

The tricky part was to make sure the normally closed valve wouldn’t completely close when the program stops or crashes. This way the heat doesn’t accumulate in the calorimeter and boils off if for an unknown reason the power supply keep delivering power, or worse “heat after death”.  :)

I had knowledge of my future lab’s situation only few days earlier, and I was not able to push tests and characterization of the flow-controller further on. But I kept in mind it was necessary to have other means of back-checking that crucial part of the system. So I did the calibrations using the reading coming straight from the logic board through RS232.

After the calibrations and during the loading, it was possible to modify the program, so I added the analog voltage reading to my DAQ, in place of the previous flow-meter used.

Using the integrated 4-wire resistance reading from the DMM was also a great improvement done while the wire was loading hydrogen. And yes, it loads hydrogen, because we saw evidence of pressure lowering while the resistance changes, while the temperature of the chamber changes. Everything correlates and will be published in a paper. Actually, this is already an exciting part of this test because we had so many questions and wonders on how we should interpret the resistance changes. This was the main question we came back with from ICCF18.


Back to the flow.

When the DAQ got the flow-metering input, I quickly set up the voltage input, but also quickly put a scale to the measurement. This DAQ is programmed with DAQmx from National Instruements’ LabVIEW and it requires use of a scale where you just put the signal-in type and amplitude and the scaled output.

The objective was not to provide proper measurements; it was in the first time necessary to check the correlation in variations and amplitude between power output and the 10V ouput. Secondly, it would provide backward information for later calculations. Hence this values are wrong, because not properly scaled-up, but they hold the pattern of a variation.

The DAQ provides noisy signals and need to be check-up. At the last minute, it was necessary to change the pressure gauge form an older one used for Celani’s tube cells that, I think, perturbates other signals on the DAQ. In the future, it will be necessary to use something like a cold cathode gauge, able to resolve pressure between 10^-7 mbar to 1.5 bar since I won’t work with higher positive pressure again.

To summarize, I am only trustful of the RS-232 reading coming from the program than the DAQ measurement. And this for three strong reasons:

  1. I tested the VI, saw it working and extracted the code from it, it works.
  2. I didn’t properly set the DAQ, hence readings are unscaled and should be disregarded
  3. The DAQ might be damaged or gets perturbations from the pressure gauge.

Additionally, there is no reason to consider using the values of DAQ flow-meter for a direct calculation, simply because they were not used for the calibration. During calibration, only the RS-232 input was connected, so it was reading the value of the flow-controller coming straight from the logic board.

If the measurements from the RS-232 during our test are wrong, so do the calibration values. Repeatability is very good on this instrument, so the systematic error would have take place on the power output calculation of the calibration.

“It is not because you don’t look at something that there is no problem”. That is why I added an additional 10V reading. I took the risk of showing wrong values to our followers, but I conservatively wanted to store as much data as possible. In case when I check the flow-meter after the test, the graduated flask or the scale shows me a wrong quantity of mass displaced during the measurement.

With all the surprise to show a positive excess energy value on the first experiment, I must say that if something is wrong, it will be assessed and it will be characterized properly after the test. Further update will share the details of these verifications.

I didn’t do this experiment because I was 100% confident in the calorimeter otherwise if would have take me 3 more months of my personal time; I did it because I was confident in its capabilities to characterize the material and assess indication of energy within the given time I have. This is the reason why I am confident, but very prudent.

Thank you for taking the time to read that.


UPDATE2 : Measuring the flow

After an issue with flow measurement was questioned by one of our follower, between the different instruments readings, additional verification of the flow control setup are done locally on Feb 6th. Three measurements of the flow using a graduated flask and a scale shows results of 130,05 g/min, 127,63 g/min and 129.44 g/min. This is lower than the requested value of 150 g/min.

Finding the origin of the issue was not obvious at first. But when I remembered the switch behind the cryostat head, this hinted the origin of the problem. This switch is setting the flowrate of the cryostat pump. Since this pump provides the pressure necessary for the flow-controler to control the flow, it was the best candidate to be the cause of the problem. Because it was on the lower flow setting it was not providing sufficient pressure to the controller. Hence explaining also the disparities on the analog readout, mainly because Kv factor of the valve was not calibrated to the condition imposed by the setup. Flipping it to the higher setting corrected the problem within few minutes, bringing it back up the requested command of 150 ml/min within a ±1ml/min variations on the analog output.

To keep going with the current run and avoiding disparities with our current calibration, the average value of the analog measurements is used from a very large sample population. The result is 128.177 ml/min. The scale applied to the analog output is verified and should give proper results, variations are within a tenth of a percent from our previous measurements. However, with a standard deviation of 4.86 ml/min, this value lowers by 3.8% our confidence of the current run to previous calibrations. We will have to wait new a calibration, that will be published after the test, to have a better assessment of the results.

On a personal note, this proved me not to trust the program and the given tools. I did log the analog out for this very reason, but it was not showing the right calculation since the beginning of the test. The only conclusion is to avoid making wrong or awkward assumptions until we have ruled all the potential problem that are potentially occuring on the system. As I said before, I am very surprised to have positive results on the first run. It is very likely later calibrations cancel any positive results.


0 #88 jeff morriss 2016-03-09 09:18
Dust contamination

Many comments have mentioned dust contamination as something that might yield false positives, so I ran a quick experiment which consisted of turning on the blower fan in the heating unit in the basement. No heat, just the fan pulling air through the filter. After approximately 1 hour I opened the blower cabinet and measured an enormous level of radiation, almost 10 mrem/hr at the filter. Now things get even stranger. About six hours later the radiation level at the air filter had returned to background level of 0.03 - 0.05 mrem/hr. To make sure that this reading was not a fluke I wiped the TV screen (another place where dust collects) with a paper towel and measured a similar rad level on the towel. Just as before, after ~ 6 hours the level had returned to normal.

Here is what I believe is happening but I will not be able to prove it until gamma spectrometer measurements become available. Radon and its decay products may be entering the basement and sticking to dust particles. Most of the time the dust is fairly dilute, but under certain conditions (fans, filters) the dust can become concentrated. That would account for the monotonically increasing radiation levels observed in the airflow calorimeter. Dust was being pulled into the chamber at a faster rate than the Rn decay products were decaying. Once the fan was turned off no new Rn entered the chamber (or furnace air filter) and the radiation level returned to normal. The tricky part is that the half lives of Rn decay products have half lives measured in minutes, so 6 hours is many half lives. That would explain the apparent "activation" that I reported earlier. It was most likely radioactive dust with short lived isotopes of Po, Bi, Pb that I was observing. Look up the Rn decay diagram; half lives and decay products are listed. It would also explain why testing of dust on the floor and in a vacuum cleaner not used for weeks would return background levels.

Bottom line: I may not have observed LENR and almost certainly did not observe neutron activation. However, other Celani-type experiments have reported radiation and excess heat, so the experimental approach is still worth pursuing.

Note on the high radiation level measured on the air filter. The GM detector was enclosed in a polyethylene bag to prevent dust contamination, so some alphas would be stopped. Inserting 3 mil thick Al foil reduced the level by ~50% indicating that some of the radiation was alphas and betas. Inserting a 1.75 cm lead shield in front of the detector stopped almost all radiation. So it is likely that alphas, betas, and gammas were all present.
0 #87 EccoEcco 2016-03-09 08:23
Jeff's results may have actually been for the most part caused by radioactive dust contamination from Radon gas decay products accumulating. Read his comment on LENR-Forum. A scintillation detector with a spectrometer will allow to tell more of what's going on.

It's not clear why these contamination anomalies haven't been reported before with his previous null results with Celani and Parkhomov replications using the same air flow calorimeter.

Similar contaminated dust issues might also be affecting GlowStick-type experiments in unexpected ways, so methods for preventing it with good certainty, given that the elevated radiation measured in GS5.2 was not that much above background, will have to be found.

Here's a report of contaminated dust issues from a third party website, which now sounds more plausible (some have noted that since it's a commercial website selling products for measuring air quality, the report might have been "enhanced"):

Radioactive Dust - Decay of Radon daughter products

222Rn decay chain:

0 #86 EccoEcco 2016-03-07 11:45
Jeff Morriss is now reporting a 20x increase in radiation level over background (with a geiger counter) after a further run with his flash-reduced Ni wire in a 5 Torr H2 atmosphere. He's also observing the nuclear activation of elements, dust and equipment nearby the reactor. Intriguing and worrying at the same time.

At this level this should be quite easily measurable with a scintillation detector. Since this radiation persists over time even after the reactor is turned off, it would also be quite convenient to measure.
0 #85 JustaGuy 2016-03-05 21:26
@jeff morriss (#81)

Your results are encouraging. We are still on track for a Mid March Delivery of the Cooled CdTe Detector ... An additional NaI Detector loan is in the works and could come through by the end of next week, at the earliest. I need to move some equipment over there, also ...

There is a comment I'd like to make, concerning Radon Gas Daughter Radioactivity and recent comments made by Jones Beene about it, which are important to acknowledge/investigate.

Alan/I have detected what I believe is enhancement of the Radon Gas Daughter Gamma Ray Peaks, especially during rapid cooling and at room temperature. I have consulted a few people about this and have not gotten any affirmation, but I believe what happens, is that the daughter particulates tend to condense on the heated components on the cell when they cool, causing a peak to suddenly show up in a matter of several minutes, definitely disappearing in their enhanced intensity after about an hour or so.

Radon Gas Level Measurement is typically done by collecting radioactive daughters on a cloth while presenting fan-driven air towards the cloth in front of the detector, to make a reading. There are other ways to do it, but that is one of the common techniques.

You may want to make some attempt to remove these settled radioactive progeny by taking a wipe or blowing air from an air can, etc. and see if the Geiger Measurement suddenly changes. I realize your cell is rather hot and that would be difficult to do, safely... In our case the radioactive species can condense right on the NaI(Tl) Scintillator Aluminum Shield/Cover and should be wiped to remove them. We never confirmed that, though.

Just something to consider/contem plate during the measurements. Since you also see activity during cell heating periods, I would say that this is not the case and is a good sign, unless they're strangely condensing on your Geiger Pancake Tube, etc.

... I'll keep you informed here if we suddenly make progress on the NaI Loaner. I'm willing to go pick it up if I can get a ride to Stanford from South San Jose.

- Mark
0 #84 EccoEcco 2016-03-05 15:10
@jeff morriss: Thanks, this one works. However, it lacks images.

Once you get a more powerful power supply I would try repeating the entire process at a higher pressure like you previously planned. The devil in the details; the low pressure used for both the reduction and loading process could be important and is the sort of thing that can make apparently successful replications frustratingly non-replicable.
0 #82 EccoEcco 2016-03-05 09:32
Thread about jeff's experiment on LENR-Forum:

@jeff morriss: it looks like your document on Google Drive requires viewing permission.
+1 #81 jeff morriss 2016-03-05 09:29
Celani_type Replication report.

Radiation levels many times above background and displaying an exponential decay profile were observed on two consecutive runs using H2 loaded Ni as sub-atmospheric pressures and 300-500C temperatures. See URL below. https://drive.google.com/file/d/0B1YNaBYpwLw0dk4zSGhWX21yRnc/view
0 #80 jeff morriss 2016-03-03 10:22
[quote name="JustaGuy"]@#75

The experiment is repeatable, at least with this particular piece of Ni wire. Tonight I was able to reproduce the radiation signature observed last night. Furthermore, the output of the GM detector is now connected to a DAQ module, so a radiation vs. time plot can be generated. The radiation appears to drop off logarithmically with a time constant of approx 1 hour, reaching background is about 3 hours. I'll put a chunk of Al next to the cell and measure before/after radiation levels of the Al. That will demonstrate whether I'm seeing LENR activity or neutron activation in the Al. BTW I really would benefit from getting hold of an energy sensitive detector.
0 #79 EccoEcco 2016-03-03 07:01
@jeff morriss: ok, so with 7T you actually meant 7 Torr.

For what it's worth - it might or might not be relevant to your findings - Leif Holmlid also generally uses very low pressures with his experiments (significantly less than 100 mbar, more often in the sub-millibar range), which he performs with K:Fe2O3 porous catalysts. He's observed muon emission with his experiments last year (see here and here). If muon capture occurs within a material, a proton can be turned into a neutron and unstable elements will form without necessarily involving free neutrons.
Interestingly once his catalyst (called "emitter") starts emitting muons this effect lasts for a few hours after gas is admitted, which resembles what you're reporting.

Holmlid and Ólafsson speculate that muon emission might be inherent in working LENR experiments, so it will be interesting to find out if this is actually what is going on in your case.

Add comment

Here is your generous contributions so far towards our $500,000 target, thanks everyone! : $45,020   Please Donate
See the current state of our booked costs here