Amid the software delays there has been quite a lot of down time for us to reflect, improve, and plan our next moves. Of the many bugs, one in particular left us high and dry. 

A mysterious short in the reef's communications Friday resulted in 4 fried RS485-USB converters, of which we needed 2: 1 for the reef and 1 for a brilliant idea in software testing.

To better prepare ourselves for software debacles such as this week, we are designating a computer with the sole purpose of testing software releases before they're installed on streaming or testing computers. Hopefully this improvement can allow us to avoid such trainwrecks in streaming data again. 

Speaking of streaming data - it's pretty hard not to notice it's been lacking for nearly a week. Though it was previously stated in a comment that our server was on the fritz, we now have identified that a bum hard drive in the tower is the cause of our plight. A new hard drive will be in on Wednesday. We're quite ready to move ahead with experiments, but don't want you to miss out and will wait to make experimental changes until then. 

Goodnight, Reactor 6

Recent innovations in the reef's power sensing only allow us for 4 power supplies, each powering 1 reactor. It's hard to imagine we could keep up with 4 concurrent powder tests, much less 5, and the odd one out easily lent itself to the cause as its powder media related to the R3 test we're closing in on. I took the opportunity to slow down and document this process to better acquaint you with the look, feel, and construction of our powder cells. 

There sits R6, stripped naked of its insulating jacket and cut from the reef H2 manifold.

The following images offer closer speculation on some of its parts. 

Resistive-heating element wrapped around the shell and secured with hose clamp. 

Separating the heater from the shell reveals the fragile aluminum retainer sitting between them.
The ruts cut in the retainer give us a solid place to position the shell thermocouples. Below it is shown why this is important.

The thermocouples sitting in the retainer are homemade. In order to electrically insulate the TC, it is sealed with a high-temperature ceramic over the welded ball of metal on its end. Frequently the ceramic is crushed between the shell and the heater, giving rise to the necessary aluminum retainer. Such is the case in this photo. The ceramic ball is nowhere to be seen as it has crumbled off. The newfound electrical conductivity may be the cause for so many of these things failing us, in addition to their fabrication from my inexperienced hands. I'm currently sorting out issues in R3's shell TC (see powder exp. log) as it has an imaginary threshold of accurate temperature sensing before it gives up and drops to room temp. 

At the top of the cell sits the aptly-named crows nest. The thermocouple junctions are seated in this hexagonal block, made of thermally-treated ash (?) wood. We had some of this stuff on hand from other projects, and found it a befitting material to sit above a hot test chamber. We have our resident wizard, Wayne, to thank for this handsome addition with snug junction seats and a place to hang our unruly copper passthrough wires - shown in black insulation.

This one is here simply to show the connection between crow's nest and reactor - an aluminum cylinder tightened with a small hose clamp. 

Again a simple look at the passthrough and interior TC wiring, sealed with epoxy. It is good to note again, however, that this is the weakest point in our design. We have made strides to improve the seal with scuffed 1/4" tubing and a longer column of cured epoxy, but it doesn't compare to the sheer strength that the other metal bits of the cell's construction have to offer. If we detect a leak, this is the place to check first.

And finally to the decommissioning! The softened features in this image are due to the fact that a layer of plexiglass separates the camera from the cell setup - resting comfortably in a homemade PVC stand for such a purpose. The small glass jar to the right is the container in which we keep the powder remains from every reactor, during both commissioning and decommissioning alike, in the case that we'll need to get a look at them later on. 

The aforementioned plexiglass is part of this sandblasting box, modified for our purposes. Inside we set an oxygen sensor along with necessary tools to do work in this sealed container. Pumping a steady flow of Ar gas into the box until the sensor turns to ~0.1% oxygen ensures us that our gloved hands won't burn off at the wrist from oxidizing Ni powder. The nervous expression on my face is because of the 250psig H2 I'm about to release from R6. We had ensured many safegaurds with the Ar atmosphere along with a damp rag covering the reactor valve in the case that powder came out with it. We also wipe off all tools and surfaces with this rag after the job is completed to reduce the chances of airborne nanoparticles. 

Here's a real look at the passthrough removed from the cell stem. The discoloration near the solenoid tip is a good sign that the coil is immersed in the powder. Also notice the tiny metal pin at the same end - our T_Powder TC. Further up inside the rod (around the middle of the coil) is the T_Int TC which primarily senses H2 temps. 

The copper seal between top and bottom flanges in the setup. These are never touched by bare hands and replaced every time the top flange has been fastened to the bottom of the cell. 

And here's the 21.6g powder mix removed from the reactor. Below is another image just to get a better look at it. Remember that this is 75% BaTiO3 and 25% 40nm Ni. 

Sorry the lighting is a little bleached. One handed photography is a tough job! I wanted plenty of exposure to highlight the color of the BaTiO3 in the mix. It's stark white (visible in a white spot in the largest clump)! This report http://www.dtic.mil/cgi-bin/GetTRDoc?AD=AD0403783 (pages 13-21) had informed us of the likelihood that our piezoelectric powder was most likely reduced to TiO2 and such evidence would display itself in the yellowing color of the material. At first glance, there seems to be none at all, indicating little or no reduction (which is a pretty good thing)! This guess is supported by the horrible H2 loading this reactor had when we ran it last year, and the pressure dropped very little over some months. Additionally we only heat our cells to 350°C at most, which is comfortable below the report's 500°C minimum testing temperature. 

But as indicated, this is a first glance, and more testing is needed. Are there any (non-spectrometer-inclusive) ways anyone knows to identify its contents? This would also need to be done in the Ar box, probably, as I'm not too keen on introducing the Ni to atmosphere unless in very minute quantities. 

In other news. . .

We've also fabricated a railing for blast shields around each cell. Thanks for the tip off and concern for our safety, everybody!

Hopefully we can greet you with data by Wednesday!