FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

Does Celani wire split H2? - Update 4

am .

We have been asked to do a fundamental test of the Celani wire.  The details are in the experiment write up below and more details are in logs linked within that document.  We have been busy getting it started and clarifying things a bit before talking about it.  Here it is though.  Let us know what we're missing.

 


Update Oct 11, 2013

 

We are in the calibration phase as we prove out the validity of the apparatus.  We have seen a number of interesting surprises in the behavior of the cell.  Some we have figured out and fixed, but others we are working on.  

 

We have seen strange temperature curves each time we raised the power.  The cell had air and the commercial, oxide coated constantan control wire.  We found that the lid on the cell was not fully fastened on and we were seeing strange convection currents as the cell heated up.  Once tightened, that went away and the curves were much more predictable.



When running in a vacuum, we noticed that small variations in vacuum level caused temperature differences.  Presumably that is from variation of thermal conductivity.



And, we noticed gas leaks.  We tracked them to the threads on the pass-throughs.  While we have had extraordinarily good luck with the threads on the pressure sensors, apparently we have more to learn about threads in this case.  Perhaps it’s the aluminum and the stainless differences, or possibly, the mere fact that the aluminum is changing temperature.  We have tightened them further and we are about to test it again.

Despite the leak, we are seeing some interesting things.   The table below is from a couple short tests in 1 bar of air and in 1 mbar vacuum.

Gas:

Watts

Copper Wrapped Temp

Delta Temp

Air

1.25

42.39

5

Air

2.5

48.41

7

Air

5

90.18

30

Vacuum

1.25

71.1

5.6

Vacuum

2.5

106.2

10.9

Vacuum

5

168.3

14.2


Less gas allows the thermocouples to get hotter.  It’s probable that one of the thermocouples may be closer to the test wires, so a difference between the two thermocouples could be expected.  In fact, that is the main reason for the calibration.  Why does the platinum run so much hotter in air, though?  

 

For the very latest test info, you can check out our experiment notes at this link:  


2013-10-14 
 
We bent the thermocouples around so they are now on opposite sides.  We noticed a few things after switching the thermocouples around:
     First the Pt thermocouple was hotter in earlier runs because of the position.  In the test after we swapped the thermocouples, the Cu thermocouple became hotter.
     
     We could see that the temperatures of the constantan wire were not uniform across the wrapped wires.  It was localized on the right half of the screen, under the Pt thermocouple in the earlier runs and under the Cu thermocouple in the current configuration.  We are wondering about an wrapping/positioning method or configuration to minimize a position effect or evenly distribute the heat from a wire.  
 
     We also wonder if the thermocouples are shifting when the cell is hot.  They are held in place by spring tension alone.  Perhaps we could get an anchor for the thermocouples so that we can be certain they won't deform during hot periods.  This would also help if we ever wanted to swap the thermocouples and have them in the EXACT location as they were before.
 
     The last issue we see is that the Cu is oxidizing and turning black.  This will continue to affect the temperature reading.  It may affect the catalytic recombination as well.  Could there be another metal to use?  Should we replace the copper each time, or should we only run without oxygen?

Update 3: Oct 25, 2013
 
The data for this experiment is streaming, now in test FC0408 LENR Cam/ Hydrogen splitting.  Sorry we don't have the bandwidth to stream the thermal camera.  
We have achieved good stability and predictability with the cell, finally, not to mention good tightness, finally.  There was one interesting moment during the calibrations where the cell was operating steady at 15W with a commercial oxide coated constantan wire and the temperature of the thermocouples both dropped by roughly 80C.  That still has us puzzled.  
 
After seeing good, close, repeatable temperatures on the thernocouples for several runs in various gasses, we installed the Celani wire yesterday.  Last night we tested it in a vacuum and got very close readings between the two thermocouples.   that means the cell was reassembled well and the relationship between the wires and the thermocouples was not disturbed.  At this moment we are doing a test run in He.  The next step is try a run in Hydrogen and look for a temperature difference between the Platinum and the Copper wrapped thermocouples.  We'll see if we can get one in this weekend.  If not, probably first thing Monday.
 
 
Update 4: Oct 28, 2013
 
Over the weekend, we ran a 40 step test from 1 to 40 watts of input power.  The cell started at roughly 1 bar H2.  The resistance dropped well and then made an interesting rise and fall again.  
 
We found that the power supply on the water reservoir failed.  The event shows in the data.  In the test cell, it shows as a pressure rise because the entire cell warmed up as the water in the reservoir warmed up.  The water flow is only for making sure the cell walls and camera do not get too hot.
 
 
Regardless, the run was a negative result for monatomic hydrogen being present and preferentially recombining on the platinum.  The temperatures of both thermocouples were within about 4C the whole time with Copper consistently reading warmer than the platinum.  The dataset is attached.
Next, we will run a shorter test in about 0.1 bar H2 and see if that makes any difference because of mean free path being longer.  

Comments   

 
0 #25 bob 2013-10-16 22:43
@charlie There have been several discussions of the vectors for producing a false positive in the standard Celani glass cell. For example unless all the energy (include IR) leaving the cell is collected and properly thermalized, changes in wire emissivity between the calibration wire and the actual wire can produce a false positive.

There have also been discussions of the method used to calibrate and calculate excess heat. I much prefer the cleaner conduction calorimetry approach where the outgoing heat flux is proportional to the delta T across a thin (but poor) conductor. The best candidate we have for evaluating this approach is to instrument both sides of the water bucket wall in the water bucket test. I still think the water bucket test is the most unambiguous indication of genuine LENR heat yet.

A false positive can also come from H recombination if we have bursty periods of excess heat combined with highly transient experimental protocols. ie. some earlier input power is being "accumulated" as dissociated H rather than heat and not being "noticed" because it occurs well before the excess heat burst. One possible counter to this problem is to deal only with integrated/cumu lated energy quantities. eg. W-hr rather than W.

MFMP is taking steps to improve their calorimeters, but it is a very difficult science to master.
Quote
 
 
+1 #24 Robert Greenyer 2013-10-16 10:47
@Alan G

We have been advised to silver plate the copper.

@charlie tapp

That is part of the point of this experiment, to establish the potential material point yield significance from this effect. That is why having a material that is very good at catalysing recombination (platinum) on one thermocouple and another that is bad will give help us understand the spread.
Quote
 
 
0 #23 charlie tapp 2013-10-15 21:57
@bob do you have the calculation they use for determining exess? just wondering if the energy to make the initial hydrogen being used for the cells is accounted for that could seriously change a lot of things, if not the tiny amount of exess seen so far would be wrong, wouldnt it ? or are we not concerned with that ?
Quote
 
 
0 #22 AlanG 2013-10-15 21:01
MFMP wrote:
The last issue we see is that the Cu is oxidizing and turning black. This will continue to affect the temperature reading. It may affect the catalytic recombination as well. Could there be another metal to use? Should we replace the copper each time, or should we only run without oxygen?

In the Hokkaido experiments on H1 catalysis by glass (at 300c), selective gold plating was used to prevent recombination in the apparatus outside of the test vessel. Gold-plated copper wire is easy to source from jewelry supply outlets.
Quote
 
 
+2 #21 Walker 2013-10-15 17:53
Hi MFMP people.

It is always worth noting that the Fleischmann Pons effect if it is in the Nuclear forces range of energy output obeys e=m c^2 and that total amount of Uranium mass that was actually converted into energy in the Hiroshima bomb was just 0.6 of a gram.

Incidentally that e=m c^2 seems to pass by many people who consider the amount of atomic mass that may be altered in say the Rossi apparatus, especially the ones who expect to make a killing on the nickel market.

Not that I think Nickel is a major contributor to the energy output by its transmutation.

The amount of Tritium being reported as out put and then sucked back in leads me to think the Fleischmann Pons effect is a Hydrogen Isotope Cascade to Helium.

IMHO When an energy differential is applied to a hydrogen loaded crystal lattice of sufficient strength, causing the looser hydrogen electrons and atoms to oscillate building up as a synchronized wave with increasing energy, the geometry of cracks and uneven surfaces acts to lens electrons and maybe whole protons and neutrons between a rock and a hard place, forcing then inexorably together. Possibly two or three adjacent columns of hydrogen reaching surface discontinuities geometrically tuned into a Goldilocks zone to spit out hydrogen up the next isotope in the chain until tritium and then H4 and then Helium.

Test tritium in a working LENR to see if it was quicker at producing heat.
Quote
 
 
0 #20 Walker 2013-10-15 17:50
Hi MFMP people.

Remember when calculating the total external energy to include energy used to raise pressure during loading and energy used to create vacuum during unloading.

I know the unloading versus the initial loading up to atmospheric should have a net effect of zero so only loading above atmospheric is energy added but it is better to ere on the side of caution and count both as energy input.

This why a large undeniable effect from a sufficient quantity of loaded material is preferable, though as you are doing one has to creep up to positive results.
Quote
 
 
0 #19 dak 2013-10-15 17:19
Follow-up

One suggestion, if you believe that different metals will recombine H radicals with different efficiencies, then why not sheath two RTD’s or thermocouples with two metals and place them inside your tube reactor with a working wire (as well as measure the outside temperature near the inside thermocouples). If it is convection or IR radiation, then the thermocouples will agree (barring differences in emissivity, which will partially be compensated by the external thermocouples). You will need to deal with emissivity changes in your current set-up as emissivity depends on oxides on the surface which will change under hydrogen. This assumes that the metals chosen will truly provide different recombination rates.
Quote
 
 
0 #18 dak 2013-10-15 17:18
I am glad that you are taking this conventional explanation about Celani’s observations very seriously. I was hoping by now that the S&G cells would have generated more data and put a conventional explanation to rest.

The problem with the on-going experiment is you do not know the form of the radicals and it is an assumption that the two metals will have different recombination rates. Say for example, that you were not dealing with just a H. radical but some complex such as H3., H5., etc. Complexes are possible under these high pressures. An alternative explanation would be ions, but these are unlikely. These complexes may or may not recombine on Pt better than Cu (assuming both metals were perfectly clean, which they are not). The radicals need not even recombine with each other but with other molecules such as water or oxides and even these radicals need to be considered.

In your experiment, you are left with two possibilities. Either: (1) the metals are different temperatures, or (2) the metals are at the same temperature. If at the same temperature you will not know if your chose the correct metals, they were active, or the conditions were correct to make radicals. If the metals are different temperatures you will not know if these conditions reflect the actual cell or if emissivity plays a part. Thus, your results may not be as definitive as one would like.
Quote
 
 
+1 #17 bob 2013-10-15 13:34
@charlie The best common example of time shifting energy is a rechargable battery. If one's test apparatus didn't properly account for the energy input during charging one could easily get "fooled" into thinking that a transient burst of discharge energy at some later time was internally generated (eg. by LENR type processes). If your test apparatus takes account of charging energy you won't be fooled because no rechargable battery is 100% efficient. ie. you'll never get all the energy back out on discharge that was input during charge no matter how carefully you perform your test.

H2 dissociation/re combination or water vapor splitting/refor mation are examples of chemical processes not unlike those in the rechargable battery above. We have to design our calorimeters to accumulate all energy (input + internal chemical reactions + LENR). thermalize it and vector all the resulting heat flux out a well defined and calibratable pathway. eg. conduction through solid layer a la water bucket test. We also have to design our experimental protocols to trigger and sustain the LENR effect for longer than brief transients to eliminate the possibility of false time shifted chemical reaction positives.
Quote
 
 
0 #16 charlie tapp 2013-10-14 18:18
@bob wouldnt there be water vapor in the cell, or is it only recombinating h into h2 ? i sent ryan an old exerpt from a 1950's modern chemistry book i have, it explains the atomic hydrogen torch dr. langmuir invented. they explain recombination yielding not only heat in the forming of water vapor but additional heat equal to the dissociation of the h2 into h. mabee i read it wrong
Quote
 

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