Did the GlowStick GS3 see excess heat? [UPDATE#1 - significant discovery affecting replications]

Written by Robert Greenyer on 13 June 2015.

Alan analysed the data from the calibration, active and subsequent re-heat runs and whilst it appears at that some of the observed departure from calibration can be explained by wire binding and movement - not all can be see below.

For all the data and Notes, block diagram, schematics, bill of materials for the experiment go here

https://goo.gl/9P85Gs

Jack Cole commented on the experiment also compared the family of GS3 charts produced by followers Sanjeev, Ged and Ecco during the GS3 run with his own data, see this graph clip

that had a similar behaviour to one of his own experiments that he suspected of producing excess heat.

"Here is a chart from an experiment I did that showed apparent excess heat."

Notice the similarities...

He went on to say "The declining and increasing power above the Parkhomov zone make it less likely that moved or constricted coils are the explanation. The coil would have to be shifting about to create an effect like that."

It is too premature to call anything, a post tear-down calibration may help, the real clincher would be evidence of elemental or isotopic ratio changes in the ash. Here is what the plan is in that direction in Alan's own words:

"I'll be opening the GS3 reactor on Monday after checking the residual pressure. Then I'll pump it down to vacuum and leave it for a while to de-load the hydrogen. After reinstalling the core thermocouple I'll do a final calibration run. These steps will be done with the reactor in place on the test stand, without disturbing the thermocouples or heater coil.

As a final step I'll remove the fuel and divide it into samples for isotopic analysis."

Whatever the outcome, we have to thank Alan and Skip for a job very well done.

UPDATE#1 - significant discovery affecting replications

Further study of *GlowStick* GS3 reveals important finding that all replicators need to consider.

Over the past few weeks, Alan Goldwater has conducted a study with the GS3 cell following observation of some strange behaviour of the in-cell K-type thermocouple in the post run book-end calibrations.

We have been discussing it for some time and Alan has now produced a report summarising the findings.

https://goo.gl/guRhzl

Thanks again to Alan for some excellent research.

This has implications for choice of materials, failure modes of heaters where thermocouples are in proximity and also for potential stimulation through the reactor body and fuel - perhaps that may explain one of the reasons for 3 phase - large differentials between coils.

Anyhow we would be interested on your thoughts.

< Prev Next >

Add comment

Comments

12

0 #27 GlowFish 2015-07-17 11:13

@AlanG

Do you have a capacitance meter? It would be interesting to see what capacitive coupling there could be between the surface area of the coil and the surface of the fuel (with the tube being the insulator between them).

0 #26 Ecco 2015-07-17 04:05

@AlanG: thank you. That makes more sense now.

+1 #25 AlanG 2015-07-16 23:52

@Ecco
I have to make a minor correction to your statement. I measured 100 volts peak-to-peak at the thermocouple, with 100 volts RMS on the heater. The leakage voltage was thus only ~32 Vrms. The measurement was made with a 10x scope probe (10 megohm input impedance), so the current was ~4 uA.

0 #24 Ecco 2015-07-16 12:06

@Robert Greenyer: it certainly is interesting! With the assumption that on a faithfully performed replication maximizing this effect a substantial amount of power from the heater wire will directly discharge inside the tube at high temperature, everything matters, even using a flat ribbon wire. I'm really thinking that the current leakage is the main key point for excess heat here.

If Alan Goldwater was able to see pretty much the same voltage as the resistor with no significant voltage drop, this was probably already happening almost as much as it could and perhaps a way to optimize it would be ensuring perfect electrical contact and that the tube isn't also leaking current elsewhere.

I ~~probably~~ most likely need my calculations to be checked out, but with a 2*10^7 ohm-cm volume resistivity @ 1000°C for Coorstek AD-998 alumina, the voltage drop along its thickness should have been significant.

As I've written on lenr-forum, I don't think the entire tube became almost as conducting as metal in his case, but rather that perhaps lithium and nickel managed with time, pressure and temperature to diffuse into the tube and create in a few spots short circuiting paths with the resistor wire at a sufficiently high temperature and voltage.

0 #23 Robert Greenyer 2015-07-16 11:28

@Ecco

Actually, I think you will find that that was niChrome flat resistance wire from a toaster!

But this wire is interesting

0 #22 Ecco 2015-07-15 10:42

@MFMP / AlanG: wouldn't using a ribbon resistor wire improve electrical contact with the inner tube and thus also the current leak effect at high temperature? Thinking about it, it's what Parkhomov also used last December!

i.imgur.com/lm4TauG.jpg

i.imgur.com/c8Gozst.jpg

0 #21 Ecco 2015-07-07 04:55

@MFMP: while I was searching up information for something not exactly related with Rossi I ended up on JONP and found this totally by chance:

www.journal-of-nuclear-physics .com/.../

Quote:

Robertgennick
February 12th, 2015 at 10:31 AM

Dr Rossi:
Can you inform the teams that are trying to replicate your results which is the treatment you do on alumina to make the reactors?
Thank you,
R.

Quote:

Andrea Rossi
February 12th, 2015 at 11:42 AM

Robertgennick:
Sorry, this is one of the most important know hows related to the operation of the E-Cat.
Warm Regards,
A.R.

Misdirection? I can't say, but it does seem to hint that the alumina tubes he uses are pretreated (doped), which is incidentally what I was speculating lately.

EDIT: this also makes me question if Inconel can be "doped" too, as Rossi once said. Was he trying to cover up the treatment performed on the alumina tube or did he literally meant that? Could it be something done to the passivating Al2O3 layer to improve high temperature electrical conductivity? Would it work without negatively affecting its reliability? (and, would it actually make any difference?)

EDIT2: another big question: can we safely assume that the emissivity of an alumina tube doped from the inside (with lithium and/or one or more transition metals) would remain the same as a plain, untreated one?

0 #20 Ecco 2015-07-06 05:07

@AlanG: it might be of interest to check out how diffusion is performed in the semiconductor industry. It's one of the processes commonly used to dope semiconductors for improving their electronic properties. This could be more relevant than it seems at first as Rossi worked on thermoelectric generators in previous years and must have learned at least a trick or two from that experience.

Doped mullite will be even more conductive at high temperatures.

EDIT: semi-random links on the subject which could be interesting or useful:
ece.umd.edu/.../...
personal.cityu.edu.hk/.../...

+2 #19 AlanG 2015-07-06 04:33

First, regarding the switch to mullite tubes, this is the most obvious difference between the GS2/GS3 and Parkhomov's reactors. It may have resulted in the difference in pressure between the two, by permeation and/or reaction of hydrogen with the SiO2 in the tubes he uses.

Further, current flow through the tube body/fuel interface might promote diffusion if the tube is at all permeable. Something like electrophoresis , ionic transport of the liquid metal into the tube surface, along with lots of H+ ions.

For the mullite GS4 cell at 1200 C, I expect to see 10's of milliamperes leakage current. I've ordered a 2 kVA isolation transformer for the heater power circuit. With this improvement, GS4 can still use thermocouples mounted on alumina coil covers, backed up by an IR pyrometer.

0 #18 Ecco 2015-07-05 17:59

AlanG said on Facebook:
facebook.com/.../...

Quote:

Ecco, The used cell shows a light grey coating on the inside of the fuel chamber. There is no measurable conductivity of the coating (>40 megohms/cm). The coating is a surface deposit, easily removed by light touch, and the broken edge shows no sign of visible diffusion into the bulk alumina.

i.imgur.com/bczPagc.jpg

I guess I need to drop my alumina tube doping hypothesis and... better check out my calculations.

Still, I think there might be merit in verifying this again with a new, unused tube just to be 100% sure that usage isn't somehow affecting the resistivity of the ceramic tube. Perhaps before the GS4 testing round begins?

EDIT: when asked if he thought there were differences from the composite photo I made, he answered:

Quote:

Yes, I see significant differences. The bang! tube was from a different source, and crystal grains and diffusion are visible in the shard photos. The GS3 tube is from a different source and seems to be much denser and stronger stuff. The supplier described it as 99.8% Al2O3, fully sintered. Our decision to try mullite is a move away from such an inert material. Have a look at the silane synthesis reaction for some insight.

Which also means that 1) there might still be hope for my hypothesis (just not for my calculations) and 2) that I wasn't imagining things. Either way, using mullite does sound like a good idea.

+1 #17 Ecco 2015-07-05 08:44

I posted this on FB earlier today, reposting it here so it won't get lost.

@AlanG: I'm suspecting that the active runs ended up doping the ceramic tube, improving its semiconducting properties significantly. Recall post-experiment photos (like those from the bang! experiment) showing a layer of lithium-nickel metal on the inner surface of these tubes and visible diffusion through quite some of their overall thickness. Yesterday I tried calculating the possible resistance through an alumina tube (as tested) and if resistivity values were as expected for alumina, even though it decreases by several orders of magnitude with temperature it still wouldn't get low enough for this effect to show up as you measured. So, there must be something else involved and it's debatable whether mullite alone would improve it. This is why I asked you if the tube was new or used. Of course, I could be missing/misunde rstanding something here, and I would be relieved if you could check that out yourself and correct me.

EDIT: related bonus picture. Follow the URL below to enlarge.

Much larger: i.imgur.com/VFdM8vW.jpg

EDIT2: I think this was posted some time back:

sciencedirect.com/.../...

Quote:

Lithium metal heated in α-Al2O3 crucibles in vacuo to 800 K reacts completely with the crucible material. The products formed by this exothermal reaction are identified as the phases Li9Al4, α/β-Li5AlO4 and LiAlO2 by means of X-ray powder diffraction analysis. Chemical analysis of the dark brown powders and comparison with the ternary Li-Al-O system confirm the results of the structural analysis. Two new equilibrium tie-lines in the Li-Al-O phase diagram are proposed. Nitrogen is not involved in the reaction.

After several runs the alumina tube might have reacted with significant amounts of Li.

+1 #16 Ecco 2015-07-04 14:44

EDIT
@Robert Greenyer: I'm not up to the task unfortunately.

To clarify, I did come up with something which also seemed to agree with AlanG's measurements, but upon closer inspection the way I calculated those values didn't make much sense, and since I was starting to get confused, I took the graph down.

As an aside, have a look at these links:

coorstek.com/.../...
(might be useful for reference)

books.google.com/.../

i.imgur.com/YxYx447.png
(10^4 Ω⋅cm @ 1400°C)

+1 #15 Robert Greenyer 2015-07-04 12:39

@Ecco

Re Paschen's law, maybe. Perhaps we could again consider using Stoyan's sparker again. However, Piantelli in his patent talks about an electrostatic field, not necessarily a discharge.

Would you be able to create a spreadsheet for evaluating various ceramics volume resistivity/pot ential current through a ceramic. Also for breakdown for hydrogen with diameter/thickn ess/material of core.

@Glowfish

The passivation of Kanthal is to stop further oxidation driven degradation of the wire. The layer that forms is Al2O3 so it will make no real difference when wire is wrapped onto Alumina tubes.

+1 #14 Ecco 2015-07-04 11:30

@MFMP: given that you want input current from the coil to leak mostly or possibly primarily inside the reactor tube and that the ceramic tube becomes conductive at high temperatures, do you believe it might be a good idea to electrically insulate and decouple it as much as possible from ground or other sinks? With this premise, thermal camera measurements instead of thermocouples start making much sense. Internal pressure sensing might also start being a hindrance.

Do you also think or speculate that a 220V maximum input voltage for the heating coil could improve chances of success compared to using 110V? What was the maximum input voltage in Parkhomov's case, in his first reportedly successful experiment? (EDIT: 220V, I believe)

And what about that of the original Lugano experiment?

Could Paschen's law be involved here?

+1 #13 Ecco 2015-07-03 17:00

@Glowfish: I agree that doing it properly with electrodes would likely be best in the short term and probably better from a controllability and reproducibility standpoint, but I believe that problems with them degrading with hydrogen exposure and heat might occur over time, in addition of making the task of maintaining the cell gas tight non-trivial. Hydrogen escaping might turn out to not be too much of a problem (in fact, I believe that H2 pressure should be kept low, ideally significantly below atmospheric pressure), as long as uncontrolled amounts of air aren't leaking in, with particular emphasis on "uncontrolled".

Doing it with the heating coil also seems more in the spirit of the original replication and experiment. This effect appears to be strong enough that it's just asking to get incorporated into current ongoing replication efforts, at least in my opinion.

To me it looks like MFMP are really looking forward to seeing next experiments getting performed with mullite instead of pure alumina:
Quote:

MFMP will be using mullite reactor tubes for the next series of experiments. Data from the supplier (Coorstek) shows resistivity declining even faster than described above

+1 #12 GlowFish 2015-07-03 14:17

@Ecco

Passing an electric current through the fuel to encourage ionization would be interesting experiment but then do it "properly" with a controlled isolated source attached to either end of the reactor tube via electrodes (steel? nickel?) inserted into the fuel. Relying on a leaky coil to fuel through slightly conductive alumina doesn't sound like a good way to do it. I am not sure that the leakage is a "good" thing in of itself. To me the report portrayed it as an issue causing measurement errors, ground faults etc.

Regarding using the alumina as a resistor/heater component, any power applied to the tube might be shorted/shunted by the fuel, thus voiding any heating benefit.

EDIT: Unless I have missed something in the report?

+1 #11 Ecco 2015-07-03 08:58

@GlowFish: you're making an extremely good point about the insulating oxide layer, but I strongly disagree about adding a gap between the coil and the reactor tube. The ultimate point of AlanG's latest update is that they actually want to maximize this effect! Also see my previous two comments here as for why one would want to do this.

@AlanG: here are volume resistivity values for this fully stabilized zirconia material from CoorsTek:

Larger: i.imgur.com/AoCvQMN.png
Source: coorstek.com/.../...

It seems significantly better than mullite as far as volume resistivity goes, but I wonder about other physical properties. By the way, today I learned that plain zirconia undergoes harmful phase transformations as its temperature increases and needs to be "stabilized" with dopants in order to be more useful. So, for high temperature operation and especially as a reactor tube, one probably needs the "fully" stabilized grade. "Partially" stabilized zirconia also exists.

coorstek.com/.../zirconia.php

EDIT: I've made this graph using data from CoorsTek. I believe the reactor tubes previously used were made of 99.8% purity AD-998 alumina:

i.imgur.com/fv6WQoG.png

+1 #10 GlowFish 2015-07-03 08:45

Out of interest sake, the heater coil was pre-burned to make an insulating oxide layer. What impedance is this oxide layer? As Alan pointed out, an isolation transformer would be needed to prevents the ground currents flowing. Could the heater coils be manufactured slightly larger than the reactor tube and spaced away from the reactor tube walls using thin alumina rods? Hopefully an air gap could isolate the coil from the fuel and the probes. The rods could be supported with two rings at either end of the reactor as the cool alumina is still an insulator. Thoughts on that?

Secondly, that filter circuit would put an additional significant differential load of 200 Ohms on the thermocouple. Can the thermocouple still drive that kind of load?

+1 #9 Ecco 2015-07-03 06:47

i.imgur.com/fMcn1Pq.jpg

I'm starting to see similarities now.

Apparently they have "heavy water" in it. Why would they? It looks like this is rarely mentioned any more.
A ceramic structure (zirconium dioxide) or possibly a ceramic container for the active material.
Nano-micrometric Pd and Ni as active material.

Doesn't this reminds my speculation? Parkhomov tubes and Lugano Dog Bones could be "MACRORs".

EDIT: just realized that I got that picture from Vessela Nikolova's blog where a post about NANORs was published yesterday: ecat-thenewfire.com/.../...

Quote:

{...} NANOR excess heat generation is produced thanks to complicated polarization/transconduction phenomena, including an avalanche transconduction electrical breakdown through the ZrO2-NiD Nanostructured CF/LANR component, as explained by Hagelstein at ICCF-19.

{...}

As the voltage was increased to about 24V, the impedance suddenly decreased to very low values. It was shown theoretically that this sudden reduction can be attributed to an “avalanche effect” that is typical of the current–voltage behavior that occurs in Zener diodes.

EDIT2: a preliminary web search reveals that zirconium dioxide (ZrO2) has a lower volume resistivity than mullite. 10^10 Ohm•cm vs 10^14 Ohm•cm at ambient temperature. See here for example: accuratus.com/zirc.html

It's not clear how this would evolve as temperature increases however as I haven't been able to find high temperature data yet. I would guess it behaves similarly to other ceramic materials.

+1 #8 Ecco 2015-07-03 03:51

@Robert Greenyer: mind boggling... hydrogen ionization or electrolysis could be happening at an increasingly higher rate the more temperature is increased, due to this effect.

Did you check out the recently released paper by Lalik et al where excess heat was apparently detected during H2-O2 recombination on a Pd-Al2O3 catalyst?

pubs.acs.org/.../...

As preposterous as this may sound, if recombination in certain conditions causes excess heat, what if this was done in a closed loop? What if the E-Cat was actually an HHO device and worked through a gainful cyclic water splitting and recombination process? Food for thought. Also read this related comment by Bob Higgins on vortex-l here: goo.gl/KXiWNb

As for Parkhomov's, if the alumina tube becomes conductive with temperature, wouldn't this mean that having the coil embedded into it like he and others did would optimize this effect as it would better conduct electricity to the tube? By keeping a low internal pressure one could also use thinner alumina tube walls which would enhance this as well.

But again if my previous premise is correct, does this mean that the tube doesn't have to be perfectly sealed anyway, in order to allow some air to come in and pressure to stay low? I don't think that people reporting success had a perfect seal. Could this have been a coincidence?

EDIT: do these findings also imply that once heated enough the alumina tube itself could be used as a resistor? In that case you would have all the input current passing through it and in addition to added stimulation you could perform interesting measurements in real time. I'm probably missing something here, but this sounds so weird and intriguing at the same time. It would be like having a big NANOR and likewise a very good power supply would be needed to drive it due to greatly changing resistivity with temperature (avalanche breakdown).

12

Refresh comments list
RSS feed for comments to this post

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