FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

Active Wires in the Chambers & Insulated Cell Details

Written by Ryan Hunt on .

Loading New Celani Wires

We loaded new Celani wires into each cell.  We are in the process of testing the first four calibration steps (so we don't overload the wire) in Helium so we can confirm that it behaves similarly to the bare wire.  Here are some pictures of the process of installing the wire.  

No, Malachi is not pretending to be a mime.  What you can barely see is one meter of 14 layer Celani wire.

Once we wrapped it onto cell 1.0, we trimmed off a small piece of excess.

From a 100 cm piece this is what we didn't need in the test cell.

Below is a shot of cell 1.1.  Note that this cell is designed to have the glass assembly separate easily from the rest of the unit.  In this picture, Malachi is scraping off the coating from the Celani with the edge of a knife blade in order to make good electrical contact with the end of it.

Cell 1.0 got a 14 layer wire

Cell 1.1 got a 2 layer wire - the same kind that ST Microelectronics showed success with.

Both wires ended up with close to 95 cm worth of wire, which means we have a small scrap to keep as a control and take microscope images of.  On both cells, they wrapped beautifully, just as expected.  It seemed to Malachi and I that the wires with fewer layers were less brittle and crispy feeling and wrapped on the support more like the bare Isotan.  The 2 layer wire also has a resistance of around 16 ohms compared to the 14 layer wire at closer to 18 ohms.  The first wire we used was 300 layers and just over 18 ohms.  So, it appears the wire gets more brittle and higher resistance with the first several layers and changes less with more and more layers.

Stainless Cell and Calorimeters

We now have 3 things that can be considered improved calorimeters compared to the first cell design.  

Stainless Cell with Insulation Jacket

We insulated a stainless steel cell on Monday.  The insulated cell will be able to handle high temperatures and pressures, and will give a much better signal to noise ratio.  The reason for this is that it will get much warmer with every watt of power in the cell.  That means instead of 48 or 100 watts to get it up to a high temperature, it should need 10 to 20 watts.  And if the Celani Wire produces a few watts of heat, the resulting temperature difference will be much larger and easier to measure.  We will measure across the insulation at 3 locations with thermocouples.  The result is this should be a pretty effective isoperibolic calorimeter.  

This is what it looks like.  Malachi carving the mineral wool insulation.

 The stainless cell in the insulation jacket.

 Now with plugs on both ends

We decided we needed an aluminum jacket to attach the thermocouples to so we could remove the cell to work on it. We wrapped the aluminum tube with 3 aluminum bands and inserted the last couple centimeters of the thermocouple tip between the band and the tube.  That should make great thermal contact.

 Here you can see the foil tape exterior and the aluminum tube interior.

 The external thermocouples were installed by coiling the tip up and taping it down under more foil tape.  They are located just outside the insulation form the internal sensors.  The stainless wall of the test cell and the aluminum sleeve should help even out the temperatures extremely well and give a pretty even temperature distribution around the cell and along the length of the cell.

 

When we calibrate this, we will operate it inside the air flow calorimeter.  This double checks the calorimetry and provides a wonderfully consistent ambient temperature and enough air flow to keep the surface of the insulation as close to a consistent temperature as possible.  This picture shows Malachi sizing up the space available for plumbing and wiring.

Here is an illustration of how it turned out.  Once we get to the right point, we will put this on the public data flow with these temperature sensors.

 

 

RTD-Heat Shunt Type Calorimeter

Last week a few ideas we had learned about calorimetry came together and we made a relatively quick and dirty test assembly to try out those ideas.  The idea is to make an insulated box with an inside and outside aluminum shell, insulation in between, and a few aluminum "heat shunt" fins as the dominant path for heat to flow from the inside skin to the outside skin.  The heat flow will be proportional to the temperature across the fin, so all we have to do is measure that temperature difference along each edge of each fin that will give us an indication of the heat flow.  It's similar in principle to the way a Seebek type calorimeter works by averaging the temperature difference in many, many places where the heat flows.

The fins and shell are made according to this highly technical diagram below.  There are 4 pieces.  One is highlighted in blue to make it distinguishable.

 Here they are in 0.020" inch (0.5mm) thick aluminum (0.5mm).  We also made an inside box to make that skin even more thermally conductive and a better structure.

Then we attach custom, linear RTDs to each edge of the fin.

 

Assemble the box with 1 inch (2.5 cm) thick polyisocyanurate foil faced insulation in the cavities, and hook together the RTDs

At the top and bottom are aluminum caps that help conduct the heat to the walls.  You can see the inside bottom cap here.

Our first calibration indicates it does generally work.  It has a long time constant - almost 2 hours to fully settle.  Ambient thermal noise has a pretty big effect.  We are learning that controlling the thermal environment is a huge issue for every version of calorimetry  have looked at.  

 Here are 3 improved calorimeters in one picture.  Just need more hours in the day in order to be able to fully qualify each one and do more tests with wires.  The exciting thing is that once we do get comfortable with them, it should allow us to test wires 90 to 95% faster than we can today.


If we see anything interesting in the vertical, dual cell tests this week, the path to better quantify it is getting clearer.

 

Add comment


Security code
Refresh

Comments   

 
0 #16 David Roberson 2013-01-10 17:51
Cell 1.1 seems to be reading lower than the calibration runs of the last few days by several degrees. I suggest that you perform the same step runs with the inactive heating wire to see if the two measurements agree.

This will help to determine why there is variation. And, it will prove once and for all that the major heat flow outward is convection instead of IR. It would be good to clarify this issue.
Quote
 
 
0 #15 Robert N. Hunt 2013-01-10 15:48
@Ryan
You may have connection problems to the active wire in cell 1.0 that is disapating heat at the connection at the end of the tube and not along the active wire. Open the cell and make sure the connections are good. If you are inputting 40W it should have the same temperature profile as the control wire unless the heat dissapation has changed. One off the wall possibility would be the IR emission of the active wire is different then the control wire and the heat is leaving as IR through the glass tube.
Quote
 
 
0 #14 Franco 2013-01-10 11:57
Hi,

did You check the performances of the Calorimeters (measurement of loss and efficiency) by means of a reference resistive wire applying a known input power?
If yes, could You show the data collected?

Regards
Quote
 
 
0 #13 Ryan Hunt 2013-01-10 05:48
We double checked the entire thermocouple data acquisition and from the raw voltage to the A/D readings to the thermocouple tables and the 10 degree lower temperatures on cell 1.0 in Helium with the Celani wire is real. That leaves the question of why it ran 10C hotter during the helium calibrations with the Isotan wire. The mica was off-gassing and making the black goo on the wire, but that chemical energy in the few milligrams of material involved could not have kept the cell 10C hotter for an hour and a half. Could it?
Anyway, we refeshed the helium and we're going to let it step through the cycle up to 40 watts again while we sleep and see what it did in the morning.
Quote
 
 
0 #12 Edwin Pell 2013-01-10 01:18
Beautiful work. I am excited to see the active wire results.
Quote
 
 
0 #11 Ryan Hunt 2013-01-10 00:20
We saw temperatures up to 10C lower on all 3 sensors in cell 1.0 and some weird resistance variation (14 layer wire). Cell 1.1 is showed a lower T_Glass_out and a generally decreasing resistance like it was loading somehow (2 layer wire). We are going to try to rule out instrumentation problems on the temperatures, but the resistance seems to be a real and curious thing. More later.
Quote
 
 
0 #10 Ecco 2013-01-09 23:43
@David Roberson: I mean, try closing and reopening the live data page.

Try also closing and reopening your browser window, then clearing the browser cache, just to be sure.
Quote
 
 
0 #9 David Roberson 2013-01-09 23:27
I tried the button but the data did not come up. I get the main screen but do not get a listing of the different cells to choose from.

I really need this live feed to help look for excess power with my time domain simulation.
Quote
 
 
0 #8 Martin 2013-01-09 23:06
Hello guys

I think you are moving in the right direction. Just a question on the isolationbox. Wouldn't a cylindical geometry be simpler (give rise to fewer possible artifacts)?

Also, for this to be stable, would you not have to keep the temperature of the outside wall constant?

Cheers
Quote
 
 
0 #7 Ecco 2013-01-09 21:58
@David Roberson: try reloading the live data page (F5).
Quote
 
 
0 #6 David Roberson 2013-01-09 21:54
The live data feed is not working for me. When will the feed resume?
Quote
 
 
0 #5 Ecco 2013-01-09 21:32
I thought there was a simpler air flow in the air flow calorimeter.

Speaking of something different, the v1.1 Cell with the Celani wire under Helium atmosphere appears to be already producing excess heat, doesn't it? I'm assuming that, as reported in the experiment log, correct values for P_Out and P_xs are already in place.

However, I seem to remember that using internal temperatures (macor temperatures in this case) to calculate output power wasn't regarded to be very safe (= prone to measurement artifacts), even by Celani. Is there any particular reason why you are not using glass temperatures instead? (T_GlassOut in particular).
Quote
 
 
0 #4 Ryan Hunt 2013-01-09 21:21
The air flows in the top, down past the device under test, and then back past the outside of the air channel and out the vents in the side walls.
Quote
 
 
0 #3 Ecco 2013-01-09 21:16
I was thinking a vertically mounted position, with the fan working as an air extractor on the top, so that you don't even have to take into account the tiny heat generated by it.
Quote
 
 
0 #2 Ryan Hunt 2013-01-09 21:08
It can be either way. Right now we are thinking vertical.
Quote
 
 
0 #1 Ecco 2013-01-09 20:17
Is the air calorimeter going to stand vertically in practice?
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