At 10:57 AM 1/24/2012, Jed Rothwell wrote:
The problem with flow calorimetry with this system is that the
working fluid is not water but high temperature glycol or something
similar. You could measure the temperature of the fluid, but you
can't just run it through the machine and dump it down the drain. So
the starting temperature will rise. That is, the glycol reservoir
temperature will rise.
The commercial unit has a primary glycol cooling loop, a heat
exchanger, and a secondary water cooling loop. That's complicated!
You can do calorimetry on it, of course. But that's a lot of
equipment. It is a large mass of material, with many things
happening in it, pumps pumping and whatnot. The skeptics would have
a field day. Alan Fletcher could think of dozens of ways to fake
that. For a scientific test, especially in the first round, I prefer
the "naked" reactor approach with isoperibolic calorimetry
Keep it simple.
As far as I can tell, isoperibolic (I haven't found a formal
definition of the term yet -- what the heck IS a peribole?)
calorimetry assumes that the entire system being tested is fully
enclosed in the calorimeter.
How do you ensure that the SINGLE internal/external thermometers (on
the walls of the kernel) are representative of the temperatures as a
whole? (See my two-heating-resistor fake)
Particularly, since the heating resistor and "thermalization zone"
are presumably in different locations?
We haven't even seen a diagram of the single-kernel hyperion. Is it
tube-like, with radial symmetry?
