FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

Encouraging Results [UPDATE#1 - More interesting data]

Written by Ryan Hunt on .

Good Data

While both cells in the dual cell test have been very slowly increasing their P_xs indicators over the last several days, last night's test was particularly encouraging.

We decided to give the cell tiny little power pertubations to see what would happen as we changed current in the wire.  In the graph below we can see the power in (stepped blue line) and the power out.  Those are 45 minutes steps, which we know isn't quite enough time to let it come to equilibrium, so the rising difference between them might be explainable by that.  But once the power in holds constant, the temperatures in the cell increase.  That is VERY interesting.  Then when we tweaked the power down another step, whatever was holding the temperature up seemed to stop. 

The resistance over that time was like this for this cell:

Note the slight change of slope concurrent with the power_out rising between 07:00 and 09:00.  And then it drops off again as the current is decreased.

 

In cell 1.1 we had a similar story, only more dramatic.

Ont he second step down, the temperatures in the cell all rose.  The P_xs hit a new high of almost 4 W.  Then, again, when the power was held constant, the power_out rose again.  I'm open to suggestions on how to interpret this, but I'm seriously leaning towards it being the first sparks of the New Fire ever seen in this part of the world.  The resistance for this wire had a similar feature to the other wire's resistance.  Notice the resistance rise after 17:00 just when the power output rose?  Then, notice the change of slope from 06:00 to 09:00?  Curious, huh?  And, of course, the obvious fact that the resistance continues to generally decrease means the wire must still be changing dynamically somehow.  Celani believes that the wires absorb other gasses including Helium.  Perhaps Helium was absorbed and Hydrogen is slowly displacing it.

 

Is It Real Excess Energy?

I'm pretty encouraged.  Here's why:

  • The confidence limits for each cell at these powers are nice and tight as explained in this blog entry.  
  • Cell 1.0@105W has a confidence limit of 0.65W (5 min average)
  • Cell 1.1@103W has a confidence limit of 1.2W (5 min average)
  • Nothing in the environment was changing that we could see in the data.  
  • The pressure measurements corroborated an overall cell temperature rise.

Of course, I still have to be cautious, and here's why:

  • We are running both wires when we calibrated with only one.  But the lower powew performance with both wires fell reassuringly close to the calibration line.  
  • We are running at total power beyond the calibration. But not by many watts.  The shape of the curve is pretty smooth and predictable, though.
  • The pressure is still changing.  But the fast temperature rise events and then falls cannot be explained by this.  Additionally, many of our calibration runs were done over many days and had similar pressure drops.  When we did 0.5 bar starting runs, the results were also extremely close to the 1.0 bar runs, so the slight drift here shouldn't be ignored, but still doesn't seem to answer much.

So good progress!  We'll see how the data looks in the coming days.  I implore everyone to help scour the data, the assumptions, and the calculations for any flaws.  When we do declare that the Age of the New Fire is here, I want to do so with confidence.

In other news, we have rebuilt are doing a first heating on the insulated stainless cell, again, tonight.  This time it has a 2 amp fuse and full instrumentation on it.  We know, so far, that the thing has a very long time constant of four or 5 hours.  That'll be fun to calibrate.  It'll give is great averages, but not so great info on fast events.


UPDATE#1 - More interesting data

The last 34 hours pressure vs PXs on Cell 1.1 (Indian time on graph)

Interesting_34_Hours

 

Add comment


Security code
Refresh

Comments   

 
0 #87 Chuck 2013-01-25 20:50
#122, @CharlieTapp:

To the best of my knowledge, constantan wire is used primarily in theromocouples. Most common is iron-constantan and copper-constant an. Related nickel-based TC wires are Chromel and Alumel (both more than 90 percent nickel).

Since constantan is only about 45 percent nickel and the reaction expected is H-Ni, perhaps a more nickel-rich alloy might give stronger results. I'm not entirely clear on what part copper plays in the overall picture.

In fact, constantan theromocouples are used in industry where hydrogen is used as a reducing atmosphere, such as a continuous galvanize line in the steel industry, not to mention in the production of hydrogen itself from methane.

So it's not as if the combination of hydrogen and constantan at elevated temperatures has never happened before.
Quote
 
 
0 #86 Ron B 2013-01-25 00:39
Ged,
It would seem possible to a 1 minute (Nyquist limit ) resolution with the existing data.

A nice digital storage scope on a few inputs/outputs would really be interesting.

Would you like to borrow one?
Quote
 
 
0 #85 Ged 2013-01-25 00:11
@Robert,

Really looking forward to this update. The data has been interesting so far, a lot of very good stuff developing.

I wonder if we can figure out the impulse response of the control system; as that would give a very powerful mathematical tool for detecting excess heat (or any other perturbation of the system's function).
Quote
 
 
0 #84 Ecco 2013-01-24 23:04
I guess I was wrong with my idea of MFMP's plans for today :lol:
(referring to the cooling off step started a while ago)
Quote
 
 
0 #83 Ecco 2013-01-24 22:28
@Robert Greenyer: ok, I guess we'll have to wait and see.

Anyway, regarding the cells currently being tested in the US: if it doesn't interfere with David Roberson's test (since the last power step has lasted more than 90 minutes and I feel the US team is going to leave it like that overnight), I wonder if would be possible to explore higher temperatures by raising input power to the active wire at like 50-52 watts (like last time) and to the heater wire at 85-90 watts or more if glass temperatures and the wire allow (the heater wire definitely should be able to take the load).

That's for testing the possible positive temperature feedback effect that the wire is supposed to have. At some point temperatures should rise significantly more than what input power alone would suggest.
Quote
 
 
0 #82 Robert Greenyer 2013-01-24 20:54
@Ecco / All

I think most of these questions are going to become redundant in the near future. We will be preparing a post over the next few days that will explain what I mean.
Quote
 
 
0 #81 Ecco 2013-01-24 20:29
Of course, what I proposed would need new calibrations, etc.
I'm not sure if this is feasible with wires that are already loaded.
Quote
 
 
0 #80 Ryan Hunt 2013-01-24 19:36
The blue channel power supply is independent of everything else. The T_ambient is a thermocouple and is likely being affected by the door (from the outdoors at -21C) opening briefly before and after the adjustment. If you look carefully at the voltage, it is also affected the ambient, but only in the tenths of millivolts.
Quote
 
 
0 #79 Ron B 2013-01-24 19:29
Is the power for the blue channel wire also the same power supply that's feeding the sensor power for the thermistors?
The reason I ask is that the ambient temp is following the power transitions with big drops in ambient just when the power is ramped up. It would seem that the request for current on the blue channel is coupling into the ambient sensor.
Quote
 
 
0 #78 Ryan Hunt 2013-01-24 18:55
The cells do have hydrogen in them. It will be interesting to see if the P_xs rises above zero at all now that we are re-warming it.
Quote
 
 
0 #77 Ron B 2013-01-24 18:43
During the cal testing that's ongoing, is the cell filled with hydrogen? It would seem that if excess power is generated then that will be the cal values for no excess power.
At the temps we're seeing now, it would be expected that excess power could be generated.
Quote
 
 
0 #76 Ryan Hunt 2013-01-24 17:04
Must be just the air flow disruption. I don't know what else might account for that.
Quote
 
 
0 #75 David Roberson 2013-01-24 16:58
Do you guys have an explanation as to why the T_GlassOut waveform for cell 1.0 begins rising before the Pin has been applied? The Power input has began at 11:34 and is stable within about a minute, but the outer glass temperature has already begun its rise by about a degree or more ahead of this time.

Could this be due to the motion of people in the room stirring up the air currents as they prepare to adjust the power?

I just noticed that the curve for cell 1.1 also has an unusual behavior just prior to the input power setting.

These effects should show up in my simulation error report. The curves will most likely not be a super match at the initial edges.
Quote
 
 
0 #74 Ecco 2013-01-24 15:25
@Ron B: personally, not having followed closely what happened to the EU Cell (also because of lack of communication by the EU testing team), I can't say whether the excess heat reported is actual or apparent. I seem to remember having read that the P_Out/Xs calculation formula hasn't been set in place correctly yet, but I'm not 100% sure.

By the way, power has just started being applied to the heater wires on the US cells. It's going to be ramped up in three 1.5 hour long steps.
Quote
 
 
0 #73 Ron B 2013-01-24 15:15
I see the Euro cell is back in action and really doing great. With power only on the active wire it can reach the critical temps and showing quite a bit of excess power. Excellent!
Quote
 
 
0 #72 Ecco 2013-01-24 11:07
I hoped too to see power applied to the heater wire immediately after the stepped test on the active wire; not only because of anticipation but also because I think it is in your best interest to minimize the effects of decreasing internal pressure over time on temperature readings.
Quote
 
 
0 #71 David Roberson 2013-01-24 01:53
Is the plan to keep the power at the Celani maximum for the remainder of the evening? I was hoping that we would begin the power steps with the heating wire soon after completing the Celani maximum level run.

So far the curve fit is very good for the Celani wire.

Thanks.
Quote
 
 
0 #70 Dieter Seeliger 2013-01-23 20:56
@Ryan,
your fast as light !
Perfect !!!

Dieter
Quote
 
 
+1 #69 Ryan Hunt 2013-01-23 20:31
@ Dieter - Column E is now UTC in the experiment log.
Quote
 
 
0 #68 Dieter Seeliger 2013-01-23 19:47
@Ryan,
is there any chance to note all the timestamps in UTC.
It`s a mess to calculate back all the local time messages or events in the live view, log and comments.

BR Dieter
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