FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

Welcome

The Martin Fleischmann Memorial Project is a group dedicated to researching Low Energy Nuclear Reactions (often referred to as LENR) while sharing all procedures, data, and results openly online. We rely on comments from online contributors to aid us in developing our experiments and contemplating the results. We invite everyone to participate in our discussions, which take place in the comments of our experiment posts. These links can be seen along the right-hand side of this page. Please browse around and give us your feedback. We look forward to seeing you around Quantum Heat.

Join us and become part of the project. Become one of the active commenters, who question our work and suggest next steps.

Or, if you are an experimenter, talk to us about becoming an affiliated lab and doing your work in a Live Open Science manner.

Error
  • Error loading feed data

The US team has reconfigured their V1.3 Celani type cells to test in a pure differential mode between cell A and cell B.  Mathieu had done the same with the EU cells just before ICCF.  This configuration will allow us to set the cells to any temperature and pressure and power configuration and simply look for a large temperature difference between the active cell and the passive cell.  The US cells are now operating.  The EU cells will start up soon after some troubleshooting in the pass-throughs.

The reconfiguration in the US was made by taking the "used" Celani wire from cell A, taking a couple centimeter sample off of it, and placing the rest into Cell B in place of the heater wire.  Then we replaced that wire in cell A with a fresh nichrome wire.  That means cell B has 2 Celani wires and Cell A has 2 nichrome wires.  The cells are plumbed together  and the connection between them will be open so that they always see the same gas and pressure conditions.  

 US Cell Configuration

Cell A NiCr1 Previous NiCr wire 
NiCr2 New NiCr wire
Cell B CuNi1 Celani Wire previously loaded in Cell A.  The Resistance for this wire is Rb_CuNi1 in test "Cell US1.3B"
CuNi2

Celani Wire that had remained unloaded in Cell B during the V2.0 Protocol tests. The Resistance for this wire is Rb_CuNi1 in test "Cell US1.3B"

We will be watching this wire load first.  

 

Loading:  Start at 5 bar H2 and 70C, raise the power in 2.5 Watt steps and wait for the resistance to level off at each step.  

First 24 Hours of loading test:

We started the low power first step of this test of this yesterday. As the power was turned on and then turned up a couple hours later, the CuNi2 wire showed a negative temperature coefficient by dropping in resistance while the CuNi1 wire, having already been loaded, increases , instead.

 

During the previous tests we did not see anything more than a watt or two, and because of ambient changes, limited power, vacuum variation, and other unknowns that may have been occurring, we could not call that test a definite excess energy.  With the flexibility that running in differential mode gives us, we should be able to test to a higher wire temperature and see if we just weren't getting it hot enough, or if our vacuum was too strong.  The tradeoff is that our calibrations will not be accurate at all since they were prepared in 1 mBar vacuum and only up to 25 or 30 watts.  The really nice thing is that abandoning calibrations mean we should be able to swap out these wires in cell B with new ones every week till we find one that shows enough signal to be compelling.

 


Update 8/16/13

Every once in a while I wish I was faking the whole thing and only going live for a few hours at a time so I could show some exciting results and be pretty sure things would look perfect.  Yesterday we made a mistake and it was, thankfully, caught by Ecco, our most consistent comment contributor.  When we reconfigured the wires in the cells, we accidentally moved the NiCr wire from cell A instead of the Celani Wire.  Malachi came in late last night to correct the problem.  (Note:  this does explain why the "loaded wire" in the graph above was showing us a positive temperature coefficient :-/ )

Ecco noticed what happened as he was scouring the data and noticed that one of the wires in Cell A was decreasing in resistance significantly.  The graph below is for Cell A.

Looking at the graph above, it is clear that the blue resistance line is in the range of the Celani wire instead of the NiCr.  What amazed me when I saw it, though, is that the "loading" started after a few hours at 70C mica temperature.  I did not expect to see resistance dropping at that temperature.  The knee on the slope about 09:00 when nothing else seems to be changing except the pressure is interesting.  This is the same wire that we spent a lot of time trying to load after the first few runs and thought we were seeing it get less active because it just wouldn't seem to load much, anymore.  What is going on inside this wire?  

You can keep watching this wire with us, but you'll have to look in cell B, now.  If we make more mistakes, let us know.  We want to make many mistakes and fix them quickly so we make the most rapid progress.


UPDATE 2 - 8/19 

Over the weekend, we continued the loading of the active wires in Cell B.  The plan was to up the temperature and wait for the resistance to level off.  Since were weren't going to be around to watch, though, we gave the cells 12 hours per step.  As you can see from the graph below, the rate of resistance drop didn't seem to change much over the weekend.  Now we will leave the power steady till it does.  The rate of resistance drop at such a low temperature is something pretty interesting to me.  

 

Since Hydrogen seems to be percolating into the new wire and the old wire was already mostly loaded, I thought I would see if there was any temperature difference, yet.

There is, but it was the control cell that was consistently slightly hotter.  Interesting, huh?  It should make any positive looking result a little more believable, anyway.

 

Comments   

 
0 #144 Pearline04 2017-10-29 10:40
I think your website needs some fresh articles.
Writing manually takes a lot of time, but there is tool for this time
consuming task, search for: ssundee advices unlimited content for your blog
Quote
 
 
+3 #143 Edwin Pell 2013-09-04 03:44
I would be interested to point an IR camera on the wire. Are there hot spots? Would it look like the SPAWAR IR video?
Quote
 
 
+4 #142 Edwin Pell 2013-09-03 22:21
Wire #2 in cell B is getting interesting. On the 2nd it was not just breathing it was panting (I have two dogs). Lots of flux but I think we are still at very low loading.

Quote
 
 
+1 #141 Edwin Pell 2013-09-03 15:20
Ecco, we are having a communications problem. I think the flux is the amount of hydrogen added per unit time. I think R/R0 is a measure of total loaded hydrogen. So the change in R/R0 per unit time is a measure of the change in loaded hydrogen caused by addition or subtraction per unit time.
Quote
 
 
0 #140 Ecco 2013-09-03 11:24
@Edwin Pell: I don't know where you got the notion that the flux needed is the hydrogen adsorption rate change. Maybe you have more information than I do, but as far as I understand, the flux actually needed is that of the hydrogen gas as a whole on the active material surface. This is also described in the 2010 Piantelli patent where it's suggested that the gas should be continuously blown on it.

http://www.rexresearch.com/piantelli/piantelli.htm#WO2010058288

Quote:
...Advantageously, during said step of bringing hydrogen into contact with said clusters, the hydrogen flows with a speed less than 3 m/s. Said hydrogen flows preferably according to a direction that is substantially parallel to the surface of said clusters. In such condition, the hits between the hydrogen molecules and the metal substrate occur according to small impact angles, which assist the adsorption on the surface of the clusters and prevents re- emission phenomena in the subsequent steps of H- ions formation. Advantageously, said step of creating an active core by hydrogen adsorption into said clusters is carried out at a temperature that is close to a temperature at which a sliding of the reticular planes of the transition metal, said temperature at which a sliding occurs is set between the respective temperatures that correspond to the absorption peaks alpha and beta...
As for why only recently (ICCF18) this idea is being deemed as very important for excess heat generation, despite having been in the public domain for a long time, I don't know. Perhaps it might be a good idea to give a more detailed look at Piantelli LENR patents for other useful information.

EDIT: like for example: Quote:
In particular, said step of triggering {includes} a thermal shock ... caused by a flow of {hydrogen}, which has a predetermined temperature that is lower than the active core temperature ...
Quote
 
 
0 #139 Ron B 2013-09-03 00:43
@Malachi,
I'd talked about keeping a constant pressure on the cell by having a reserve tank and regulator on the line. Is that just impossible? I only ask because I would love to see changes over time that were a result of only the hydrogen being absorbed. The constant power disruptions for refill add a whole new level of complexity in analyzing the data. If pressure and power were constant then it would be easier to see excess power over time.

Flash suppressors can be used in-line to address safety issues.
Quote
 
 
+2 #138 Edwin Pell 2013-09-02 18:10
Ecco, I think what you say about loading and temperature is true.

It depends if we think excess heat depends on total loading or on the time rate of change of total loading (flux).

If total loading we should be getting lots of excess heat now. We are getting some but not the maximum. ??? I am interested to see where excess heat goes with time at this applied power level.

[addition]
the above was written in haste. I expect NiH behaves like PdD that is
excess heat =k*((loading - loading threshold)^2)*abs(flux)
that is both flux and loading are required.

[addition]
The last 24 hours of data are showing fascinating correlations between R/R0 changes and temperatures changes.
Quote
 
 
0 #137 Ecco 2013-09-02 16:38
@Edwin Pell: a much simpler explanation for the resistance change rate falling down is that H2 loading in metals has a finite limit and overall follows a logarithmic trend that is initially accelerated by temperature (over a limited range) and pressure. But if the material is already almost completely saturated with hydrogen, no matter the temperature or the pressure, loading is inevitably going to slow down and ultimately stop progressing.

In the end, I'm not getting why this is supposed to be a problem. It's exactly what should be expected. The resistance of CuNi Wire #1 with power applied is currently at 13.53 Ohm; when power will be removed it will likely fall down to 12.7-12.8 Ohm which is a good result, and close (or even better than) to the lowest value achieved during the first loading under the V2.0 vacuum protocol.
Quote
 
 
0 #136 Edwin Pell 2013-09-02 16:08
The big picture in terms of flux. Powering down and then stepping down and the stepping back up changed the flux from average 0.0020 to average 0.0003. It really killed the flux.



[addition]
as of the 3rd we are real close to zero net flux

Quote
 
 
+1 #135 btbbass 2013-09-02 15:14
@Robert Greenyer, @Ryan Hunt, @All who works on this, even if behind the scenes.

Thank you for you're great work, of course! Never forgot that you are putting some extraordinary effort in this project.

What I was saying (maybe too "rudely" :-) ) was in fact to be sure of what pass outside MFMP, in order not to expose your hard work to easy criticism.

Unfortunately I had not time to analyse those data, but they seem too beautiful to be true :-) Finger crossed!!
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

MFMP Facebook Feed