FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

EU Cell Calibration Procedure

Written by Mathieu Valat on .

For the EU cell, I slightly changed the way the calibrations are being done, taking into account what we have learnt from Ryan's team during their first few weeks.

First, I am using a constantan wire kindly provided by Nicolas, it has the same metallic composition as the active ones, it is the same size (0.2mm diam. 1m long), but it is coated with an (unknown yet) oxide layer, insulating and preventing hydrogen uptake. This is very important, Francesco Celani told us, even bare constantan wire absorbs small quantities of hydrogen.

After a few tests, the drop in resistivity in this specific constantan wire was only 2.4% compared to the 18%+ we had with the active wire ran by the US lab. Moreover, this drop occurred at very low temperature, where the active wire dropped at 170°C. To me, it looks pretty convincing that no hydrogen was loaded into the wire. Maybe I am wrong. Let me know your thoughts.

So the cell calibrations will be performed in the same conditions as with the active wire (75%H2 + 25%Ar @3.5, 2 and 1 bar because it will leak through overtime) with a range of power that is close to the operating powers (28W, 38, 43, 47, 52, 58, 63, 69, 73, 78, 83, 88, 92 and 100W, the last one is not necessary...). After one hour of temperature stabilisation we sample one point that will be used to plot the polynomial cuve used as a baseline for calculation of excess energy. The choice of which Tc to use is boggling since T_in is influenced by the gas mixture. We tend to use T_out but it is highly influenced by ambient temperature...
T_well is the last option, but it is broken on Cell#2.

I did another one this afternoon using the same gas mixture at 1 bar pressure for each power step.

This gives a set of 3 curves for a gas mixture. 1 for constant pressure. It is also worthwhile to use 50-50 mixture because the hydrogen is the most inclined to leak through the cell. All this curves will help to interpolate and anticipate experimentally the future behaviour of the cell.

Then we have to consider low pressure case. So I use 0.5b of pure H2, I also do 1b of pure H2 because I might try this way at the end. However the cell will leak...air inside, hence it is also necessary to do a 50-50:This email address is being protected from spambots. You need JavaScript enabled to view it. ; and acquiring data during that time too to detect endothermic/exothermic reactions.

The question of the leakage is related to the fact that we are doing replication as much close to what Francesco was doing in NI week and ICCF. We also want to run the cell "until it dies".

But that is not sufficient for the calibrations, as we have to do calibrations for third parties with activated wire (as a reminder, we plan on shipping the cell to universities). Calibrations runs with He under 3.5b, 2, 1b and 0.5b will be helpful for future users.
When this will be the necessary, she can uses Francesco's technique of heating the wire under vacuum that can help releasing H for the wire's lattice. It can also be lethal for the wire too if too much power is injected in (especially in vacuum because of the absence of convection). 1.9A is the maximum current that Francesco recommends.
The use of oil-free vacuum pump is mandatory because of possible back-streaming of oil inside the cell reaching the wire.

Finally, we have the loading phase that require to use of 3b of pure H2 then rise temperature of absorption: ~170°C.
Then the gaz mixture is introduced and the power is immediately turned on to reach 48W operating power.

I have already done the first batch of calibrations with the mixture, taking advantage of the programmed calibration script, I did ran the cell 24/7. I will make my best to reach operational run as soon as I am done with all these.

Add comment


Security code
Refresh

Comments   

 
0 #17 Alberto F. De Souza 2012-12-11 03:07
To avoid problems with pressure and gas composition, the cells could also be connected in series with tubes and receive the gases in series at the same time. Before power in, valves between cells (open during gas load) could be used to isolate then to avoid heat transfer via gas.

I believe this approach would easier and cheaper than flow calorimetry...
Quote
 
 
0 #16 Alberto F. De Souza 2012-12-11 03:06
In this case, all cells would receive the same amount of power and should show about the same internal and external temperatures. Unless of course, as many would expect (including me!), the active cells show excess heat. In this case, considering the amount of excess heat Celani saw (~12W of excess heat from 48W of input heat), it would be easy to spot the temperature differences and, therefore, prove that the excess heat came from some mysterious reaction (LERN?) in the cells with an active wire in it.

By using two active and two inactive cells, one could rule out other factors if the inactive cells show the same internal and external temperatures and the active cells, at the same time, show significantly higher temperatures in accordance with Celani's colorimetric formulation.
Quote
 
 
0 #15 Mathieu Valat 2012-12-05 14:14
Al, the problem we have is coming from the triggering temperature of the effect (>250°C). With a water flow calorimeter we have to put insolation in the between and that fade the ability to see short events. Seebeck is the most relevant option.
Quote
 
 
0 #14 Dieter Seeliger 2012-12-03 15:12
I agree with Al also !
Quote
 
 
0 #13 Rats 2012-12-03 03:07
I agree with Al as well.
Quote
 
 
0 #12 Sanjeev 2012-12-03 00:48
Constantan having a property of keeping its resistance constant with rise of temperature, may not have an ideal constant resistance and the resistance will probably go up and down from an average value with temperature.

So I guess the result is normal. Now the question comes , whether the R/Ro of 0.8 for Celani wire has anything to do with H2 adsorption and how to show it conclusively.
Quote
 
 
+2 #11 Ron B 2012-12-02 20:24
@AlexRa
That was my idea and while it might help sort out anomalies it's not without possible side effects itself. Since putting gas under pressure into the cell is going to have a delta effect on the temperature. I still think it's worth a try though.

I didn't get much feedback about the question of the location of the thermocouple on the glass (is it on the top of the glass? How's it attached to the glass? Is it possible to move it to the bottom of the glass or to turn the entire cell upside down just to get some more information?
Quote
 
 
+1 #10 123star 2012-12-02 19:01
I concur with Al of course, however it would still be interesting to fully characterize the possible artifacts (for now).
Quote
 
 
0 #9 Pekka Janhunen 2012-12-02 18:59
The impedance anomaly has started again, and at the same T_mica=210 as before. Whatever causes it, it seems to be temperature controlled rather than pressure controlled. Interesting to see if it stops again at T_mica=230..240 .
Quote
 
 
+4 #8 Al Potenza 2012-12-02 17:41
If you build a simple liquid cooled flow calorimeter around the existing cell, you won't have to worry about ANY of the thermal properties of the contents.

And while I recognize that the supply of wire is limited, the more of it you can get and use at one time, the better your signal to noise ratio will be.

You have done an excellent demonstration of why spot temperature measurements are rarely a good substitute for formal calorimetry in LENR/CF experiments.

I respectfully suggest that these two ideas (doing calorimetry and increasing the quantity of active elements in the system) should be the thrust of future efforts.
Quote
 
 
+1 #7 Pekka Janhunen 2012-12-02 17:30
A completely different avenue would be to leave out electric heating and instead put the whole reactor inside a furnace. Then any substantial and prolonged increase of temperature near the wire would be indicative of LENR and there would be no need to model heat transfer.
Quote
 
 
0 #6 Robert Greenyer 2012-12-02 09:50
@Alexra

The idea of having a much larger same pressure reservoir was noted and may be tried at some point.

@Pekka

This has been identified as being annoying internally and a fix is proposed that would mean that there would be a toggle to switch between UT and local time.
Quote
 
 
0 #5 Pekka Janhunen 2012-12-02 08:24
I notice that there is an unexpected impedance anomaly in the EU cell calibration run which started at 1.4 bars (4 hours ago).
(By the way: it would be more convenient if hugnetview would operate on UT time instead of local time.)
Quote
 
 
0 #4 AlexRa 2012-12-02 07:48
Regarding the resistivity changes: it would be interesting to compare the R=f(T) behavoir of the Celani wire with that of a regular (not oxide covered) constantan wire, to see to what degree the change of resistivity is caused by absorption of hydrogen into the volume of the wire, as opposite to possible surface effects in Celani's nano-coating.

Is the thickness of "nanostructured " skin of Celani's wire known (and accordingly, its % in the cross-section of the wire)? If the "skin" is very thin, it should not be able to affect wire resistance in any significant manner, unless something very unusual is happening in this layer.
Quote
 
 
+1 #3 AlexRa 2012-12-02 07:32
As recent updates suggest, the pressure instability during long-duration experiments does affect the temperature measurements significantly, possibly invalidating any excess-energy indications.

A while ago some other reader suggested in the comments a simple way to stabilize the pressure. I haven't seen any follow-up on that idea, so assumed it was overlooked by the MFMP team. Might be a good idea to have a second look at it now.

The idea was to stabilize the gaz pressure in the cell by keeping the cell permanently connected to a large-volume "buffer" gas tank (both kept at the same pressure, obviously). That would drastically increase the total pressurized volume, so any leak you might have in the cell would affect the pressure much less.
Quote
 
 
0 #2 David Roberson 2012-12-02 03:03
I performed a curve fit on the power change that happened on 11/30/2012 at 22:20 according to my download from the data source. The blue power was increased at that point so that the total power input was 82.8 watts. The red source was only used prior to that point at 48.2 watts. Temperature at T_GlassOut transitioned from 122.9 to 167.2 degrees C.

An extremely good fit was obtained with my time parameter beginning at the power transistion for t=0 seconds. The fit envolved using three exponentials. The first and second are related by a factor of 2. I assumed distortion due to quadratic nonlinearity for this choice. The third exponential is most likely due to a delay of some kind and is of short duration. The curve fit equation is as follows:

T_GlassOut=114.7+52.01*(1-e^-.003532*t)+4.91*e^-.0071*t+3.39*e^-.034*t.

Values rounded to save space.
Quote
 
 
0 #1 Arnaud 2012-12-01 23:44
If I was you, I would not try the 50% - 50% H2 / Air ... except if you want to trty what happens in case of chemical explosion.
50% - 50% mix is an explosive composition.

The temperature ignition of H2 and Air mixed is around 500°C. (See wikipedia) The heating wire might come easily above 500°C. If iginition occurs then a lot of chemical energy released in a short time (explosion).
Quote
 

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