FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

HUG Concentric Calorimeter

Keep up to date with the experiments using HUGs flexible Concentric Calorimeter right here.
Please feel free to add your comments to the blog entries.

LIVE DATA | Experiment Log | Latest Video | RSS feed of blog posts | RSS feed of all comments

50 meters of Constantan in one cell! [UPDATE#5 - Pressure Bottoms Out

on .

We are ready for some fun.  We shoved 50 meters of bare constantan into a 12CC cell, calibrated to 500C, and we're almost ready to start loading it.  The entire details are in this document: 50m Wire Test

And we just kicked off the pre-loading off gassing step in the protocol.  Give me your thoughts!


UPDATE 6-28 17:30

The cell was run last night under a passive vacuum at 500C. The valve was closed before the cell was heated and the gauge read 1 mBar.  Overnight a small amount of gas seemed to come out of the hot wire but the pressure sensor saw some 1 mBar oscillations of pressure with about a 20 to 30 minute period.  

This morning we opened the valve to expose the hot wire to active vacuum for a while.  Then we cooled the cell, loaded to 10 bar hydrogen and then started heating looking for where the wires would start to absorb hydrogen.  We scanned between 120 and 250C, roughly.  We didn't see any notable resistance drop or absorption, but we did see a very puzzling pressure oscillation ~ 40 mBar, 4 minute period.  It's so regular it is hard to imagine it as a wire related phenomenon, yet we don't have anything we know of making any sort of 4 minute signal, especially one so nicely sinusoidal.  Suggestions?

UPDATE 18:09 -- MYSTERY SOLVED - it is the PID control loop for the Active insulation causing a temperature to change in the pipe.  At low pressures it is not noticeable on the pressure gauge.  At this pressure it is much more noticeable.  It will average out over several minutes, though.  We are going to tweak the control loop to minimize the oscillations.


Since we didn't see any noticeable loading up to this point, though, we are going to continue raising the temperature to see if we can trigger any.


So the reactor is now at the 1st phonon resonance temp of Cu63 (the target element) as put forward by one of our followers in his freely published document you can review here


He suggested his research showed that if we hit the 489 degrees C in pressured hydrogen atmosphere, over time the Ni62 would become Cu63.

From the text:-

"Precision heating of nickel powder to 489°C or 859°C resonance with copper isotope Cu63 in the presence of pressurized H2 gas will target and maximize conversion of nickel into stable copper. The nuclear fusion reaction that produces excess heat can be specifically targeted by precision heating of the reaction to 620°C or 966°C resonance with meta-stable copper isomer Cu68m"

Now just look at what is going on in the cell!

You might note that the 2nd resonance temp is ballpark the December rossi temp where the cell melted down and the second is ballpark the 2nd COP 5.7 test... the 3rd test at a lower temp (in the range of our Celani experiments) but much lower COP.

Food for thought?

UPDATE#3 - Resistance dropping?

UPDATE 4 - Temperatures overshooting and other fun stuff


The last 24 hours have been extremely interesting as our bare constantan started demonstrating the same behavior that Ed Pell reported seeing with a very similar experiment in the last few days.  In addition to all the insanely useful analysis in the comments, I would like to add the following 24 chart and point out several of the features I found interesting.

Overshooting Temperature

This graph shows the pressure being refreshed to almost 10 bar while the cell was cold (note, the calorimeter runs warm, so "cold" is relative).  Then the power is turned on to 37W and the temperature in the cell rises to 511.4C, then proceeds to settle down to 499.1, or so, after a while.  This cell did not do that when operating near 400C a few days ago, and did not do that with the wires in place before hydrogen was added.  Each subsequent step since then has displayed the same behavior, though the size of the overshoot was decreasing.  The measured output power from the calorimeter showed the same basic shape, but only with an amplitude of about 60 mW, which is not beyond the confidence limits

Behavior of the Pressure

The drop in pressure is highly unlikely to be leakage, as far as we can tell, given that the pressure drop mostly stops when the cell is cooled down, and in the last few cycles, has actually increased.  If it is not leaking out, it is probably really absorbing, and our best guess at the moment is that it has absorbed ~40% by moles of Ni in the wire (spreadsheet) Figures are rough because of uncertainty about actual volume of cell after the wire and fiberglass sheathing is inserted.  The rate of absorption changes at 6:53, also.  It is after that change that the pressure starts rising while the cell is turned off.


The resistance is starting to be interesting, also.  In the graph, the resistance value rises when the temperature overshoots, too, possibly indicating that the higher temperature is localized at one spot on the wire that is getting significantly hotter.  Then it develops little sudden jumps.  These values are small, but clearly above the noise levels.  In the latest data (so new it isn't on this screen capture, the resistance is rising after a rather larger little jump in the resistance.  

Another interesting thing is that the resistance of the wire does not come back very close to the level it was before the first heating cycle started.  And even before the first cycle pictures, the resistance rose slightly as the hydrogen was loaded.



The meaning of it all?  I have no idea.  BUT, the overshooting is extremely interesting.  A big thank you to all the people following and contributing to the analysis of these events.  Especially, we'd like to thank Ecco, who we suspect is really a whole team posting under the same name because of how much he contributes :)

Our plan is to just keep letting it cycle and load as we watch.  Should make for an interesting holiday weekend here in the US.


 One more graph to leave you with, tonight:

Note how the rate of pressure drop increased tremendously in the second set of cycling.  We also upped the current through the active wires to 1 amp in between the cycle sets.  When we were running at 500 at the 12:00 on the 29th we also saw the faster rate, though, so it appears temperature dependent.



UPDATE 5 - Pressure Bottomed Out at 0.23 Bar

It certainly appears that the wire is continuing to absorb the Hydrogen even below atmospheric pressure.  We allowed the cell to bottom out at roughly 0.23 bar.  After that, the resistance slowly increased and the wire seemed to be releasing gas slowly and then reabsorbing it quickly every so often.  Curious, huh?  Funky stuff going on in that wire as it interacts with Hydrogen. 

After observing the minimum pressure for a while, we decided to refresh the pressure and try to achieve higher loading ratios.  We are currently at approximately 62% H to Ni.  We do not know what the Cu in the wire will do regarding Hydrogen.


Add comment

Security code


0 #166 Jeannine 2023-09-08 02:45
This post is really a pleasant one it helps new web people, who are wishing in favor of blogging.

My web blog ... discuss: gpsites.win/
0 #165 Ecco 2013-08-03 09:32
Another graph showing the apparently increased sensitivity to ambient temperature changes of the CTC cell in its new location:

0 #164 Ecco 2013-08-02 16:49
>Also, because the system is so low impedance (0.75 ohms) and we aren't doing a real 4 wire resistance and voltage measurement, undoubtedly, some of the measured input power is actually making heat in the copper feed wires outside the calorimeter.

So I guess this the reason why increasing power to the wire bundle while keeping the total power constant was decreasing temperatures, right?

If with this setup some power gets wasted outside the reactor, this would also prevent even at a low level (for example, something like 10W inside/20W outside 0W inside/30W outside) cycling power while keeping the total constant to create an active internal H2 flux.

It would be interesting to know if in CTC #2 which currently has a simpler wire setup, this would be doable or there would still be similar issues.
0 #163 Ryan Hunt 2013-08-02 16:26
Yes, it appears that low of a voltage (~4 volts) the power supply can't push that much current. I guess we had never run it at that condition before. At higher voltage it can do a full 6 amps.
Also, because the system is so low impedance (0.75 ohms) and we aren't doing a real 4 wire resistance and voltage measurement, undoubtedly, some of the measured input power is actually making heat in the copper feed wires outside the calorimeter. This test, therefore, has turned out rather un-useful.
It wasn't hard to do, though, and was worth a try.
0 #162 Ecco 2013-08-02 16:08
So, I seem to understand that the power supply can't output much more than 4.0A, and since the wire has a rather low resistance (0.75-0.80 Ohm) this means that the amount of power that can be applied to it is limited, am I correct?

This would also prevent the on-site 50m wire bundle oxidation protocol I previously suggested.

Too bad, I was looking forward to seeing this experiment. Do you have anything planned on this regard? (for example using the same total wire length with a lower amount of wires).
0 #161 Ecco 2013-08-02 12:42
This chart shows what's going on with the 50m constantan wire bundle:

Larger version here: i.imgur.com/l46rttc.png

Either a glitch/artifact due to the unstable electrical characteristics of the wire, or something major is happening.

Something happened when external heater wire power became equal to 50m constantan wire bundle power
0 #160 Ecco 2013-08-02 11:47
While I was searching information about CuNi alloys oxidation in air, I stumbled upon this very interesting page:


Basically, below 650°C almost no oxidation happens on the wire in air. Increasing local temperatures to 950-1000°C speeds the oxidation process significantly. However above 1050°C the oxide layer sinters, sticks to the core, and maybe this is not what we want.

What's interesting is that on the core wire copper atoms tend migrating to the surface during the oxidation process, forming Cu2O. This interesting for the test I suggested as CuO is very vulnerable to hydrogen attack at high temperature (>500°C). This should be able to produce plenty of defects, craters and cracks.

After the current test is finished (although I would like to perform it once again at a higher H2 pressure and in 1W or 1.5W steps, but only if the issues I highlighted in my previous comment get sorted out), I propose this protocol:

- Take out power from the cell, vacuum out hydrogen.
- Inject standard air. Can you inject compressed air?
- Apply power to the 50m wire bundle so that it becomes glowing hot. It should glow orange-yellow. This could be performed without removing the wire from the stick.
- Keep applying high power to the wire bundle for 3-4 hours.
- Vacuum air from the cell. Don't apply a deep vacuum, and don't do it for too long.
- Inject H2 at the usual level (8 bar?)
- Apply heat to the external heating wire so that internal temperatures increase to at least 500°C
- Keep heating for 24-48 hours, observe any change in wire performance and characteristics.
- Apply a relatively deep vacuum under high temperature (500°C) for 8-12 hours to remove residues and oxides
- Inject H2 again at the usual level (8 bar?)
- Apply power at varying levels without increasing temperatures above 400-450 °C
0 #159 Ecco 2013-08-02 09:22
Something went wrong with the 50m wire bundle after power to it got increased to 15W:


I think the power supply couldn't manage to keep up with it anymore and started feeding noise, which disrupts input power readings among other things.

Still, it would be really interesting if input power on the wire bundle really was around 3W (as of writing). If that were the case, then it would mean that a significant amount of excess power (>20%) started getting produced. I doubt it is, however.
0 #158 Ecco 2013-08-01 18:38
It looks like live data stream to the public got broken about one hour ago. All US experiments at HugLab are being affected by this issue.
0 #157 Ecco 2013-08-01 16:39
@Ryan Hunt: it appears that the control algorithm for the current test isn't working correctly. Total input power is lower than 30W as there's less than 0.5W on the 50m wire bundle, unlike as expected for second step.
0 #156 Ryan Hunt 2013-08-01 15:55
Thanks for pointing that out, Ecco. We'll dig into that.
0 #155 Ecco 2013-08-01 15:48
@Ryan Hunt: thanks for the update.

By the way, I noticed that the CTC cell appears to be more sensitive to ambient temperature changes in the new location than it used to be. Check out output power values here:

+1 #154 Ryan Hunt 2013-08-01 15:34
I think those all sound like good tests, Ecco. It will give us something to do while we write up the results so far in a compiled report.

We are also adding two more of these type of calorimeters to our lab. One is in place being calibrated and the other is under construction. We will make a blog post after we test the first new one asking for help compiling a test plan to make use of all 3.
+1 #153 Ecco 2013-07-31 19:43
@MFMP: seeing that you seem to be out of ideas for the 50m constantan wire (it's been idling unpowered for a while), what about playing around with it with some of the ideas I suggested a while back? They wouldn't require much effort to perform, I think.

EDIT: a completely effortless test I would like to see performed is to start with 0W on the active wire, 30W on the heating wire and then increase the former AND decrease the latter in discrete steps over time so that the total input power remains 30W while getting gradually shifted to the active wire. This is to check out if there will be changes in calorimeter performance depending on the heat source location and if 50m wire bundle performance over time improves as more electrical input gets routed to it.

Other tests include the oxidation of the 50m wire bundle and heating under high temperature H2 to create potentially active defects on its surface through hydrogen attack/H2-induc ed oxide reduction.
+1 #152 Ecco 2013-07-29 21:03
@Malachi: current on the wire bundle used to be 1A / 0.75V (0.75W)

I'm referring to this entry in the experimental log:
7/29/2013 15:41:53 I just unplugged the ctc for a moment. Reset the current to 0.5 amp across the 50m wire.
Please disregard if you lowered the input on purpose.
+1 #151 Ron B 2013-07-27 14:18
Sorry not to be more clear but the values listed were in Kelvin (K).. so the 450K was 176C. It would be nice to know the grain size for sure.

I would be cool to set a rare earth magnet close to the cells, also to put some Polaroid film close by too with a bit of metal in from of it to check for x-rays.
0 #150 btbbass 2013-07-27 09:59
@Ron B
From what I was just reading the Debye temp is lower with nano size particles.

If this is true, Defkalion could have given us also the dimension of their nanoparticles.

450F are 232C, but during DFK demo they indicated 179C = 354F

Maybe a material expert could also infer Nickel grain size?
0 #149 Ron B 2013-07-27 03:28
From what I was just reading the Debye temp is lower with nano size particles.

Also, no one really has talked much about superparamagnet ism. Accordingly, when an element is transitioning into this state magnetization can randomly flip direction under the influence of temperature.
It seems this should generate electrical current.
Could that affect our resistance measurements?
0 #148 Ron B 2013-07-27 02:28
Debye temperatures for common elements
Element in (K)
Al 428
Au 165
Cd 209
Cr 630
Cu 343
Fe 470
Ga 320
Hf 252
Hg 71.9
In 108
Nb 275
Ni 450
Pb 105
Sn 200
Ti 420
V 380
Zn 327
0 #147 Ecco 2013-07-26 00:42
@MFMP (Malachi): after power in the lab went out it seems that you forgot to set constantan wire bundle input power to the previous setting of about 0.75W. It's now ~0.185W.

However, even though total input power has decreased by ~0.57W, output power has only decreased by about 0.25-0.30W. This gives a tiny excess heat which is apparently above the confidence interval.

How could this be possible? Could you try restoring power to the wire bundle to 0.75W to see if anything changes?

EDIT: upon slightly deeper analysis of the available data it appears that the outer tube is now overheating less than it used to. This might explain the (apparently?) higher computed excess power. I would suggest again (I did a few weeks ago) to back input power down a bit in order to decrease temperatures so that the external tube won't heat under load and understate any excess power.

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