I continue to plod along on a simulation of prospective E-cat designs
to fit the 6 Oct 2011 Rossi test results. I have simulated various
combinations of materials for thermal storage and have found that a
couple slabs of ordinary Portland cement with a heating resistor
sandwiched between them seems to fit the properties of the E-cat
fairly well in terms of heat storage dynamics. All but the slab ends
can be insulated with vermiculite. With a lot of experimentation (by
simulation) a much better fit can probably be obtained, using mixed
materials, but what I have now is very simple and looks like it will
be fairly good once control dynamics are added. It is most notable
that attempting to simulate results from a black box using rational
and credible designs is far more difficult than simulating
prospective designs for a new construction of some kind. In the
latter case control of the design is available and all is known. In
the former case the degrees of freedom for the black box contents
provide increased orders of magnitude of difficulty. Given the
unreliability of the data, it is perhaps a nonsensical thing to
attempt, and has been given far to more effort than is justified.
I have only been working on the basics. I have not yet begun to
fully explore the dynamics of water volume fluctuations and the
possible dynamic controls suggested in my review at:
http://www.mtaonline.net/~hheffner/Rossi6Oct2011Review.pdf
Some sample graphs of ouput data, corresponding to Graph 2 and Graph
5 are shown here:
http://www.mtaonline.net/~hheffner/Graph2S.png
corresponds to Graph 2 at:
http://www.mtaonline.net/~hheffner/Graph2.png
and:
http://www.mtaonline.net/~hheffner/Graph5S.png
corresponds to graph 5 here:
http://www.mtaonline.net/~hheffner/Graph5.png
The difference between old graphs and the new corresponding graphs is
that the outputs in the new graphs were calculated using only the
input data, not using the experiment output data at all.
I also produce a dynamic temperature profile of one of two internal
slabs of cement assumed in the simulation to exist in the E-cat
(their profiles are symmetric):
http://www.mtaonline.net/~hheffner/Graph6S.png
This profile will look very different if dynamic control is used to
control timing of when heat is released from the slabs, e.g. when
water is admitted to the inside of the E-cat case to absorb the slab
heat, or when slabs are joined under pressure to transmit heat better.
Some sample text output from a run is shown here:
http://www.mtaonline.net/~hheffner/RptR4
Graph 6S that demonstrates the FEA part of the simulation. It shows
the largest thermal gradient at the water side of a cement slab
between times T300 (min) and T330, well after the power cutoff at
T281, and thus the largest thermal output after or before the power
was cut off. This corresponds to the nice bump in the power out
curve between T300 and T330 in Graph 2S. I think the power out peak
will be pushed further to the right once I get logic in the program
to reduce water access to the heat in the slab in proportion to the
power supplied. The heat released prior to the main power cut-off
at T281 can be reduced if dynamic controls are in place and thus the
curve prior to T281 will be reduced to look more like Graph 2, and
the heat after death power curve will be shifted to the right and
increased in magnitude. Also, the slab, or possibly one of two
slabs, of material can fully and uniformly come up to temperature,
essentially doubling the thermal storage. In a realistic device the
water interface should be metallic, at least a thin layer, but this
has little effect on the overall thermal dynamics.
I have been looking for a very small normally open valve with a
(current) proportional response. It only needs to control a flow of
about 4 ml per second, or about 0.24 liters per minute. I think I
could make one using a Taco Power Head zone valve actuator
(discarding the metal case), but I'll bet there is something
available off the shelf. It would have to be small and able to work
in up to maybe 120°C heat, using only about 7 watts of power. Here
is the best I've found, but it would need to be smaller and cover a
smaller flow rate, and doesn't need to handle such extreme pressures:
http://www.hydraforce.com/proport/Prop_html/
2-380-1_PV08-31/2-380-1_PV08-31.htm
I like the operating temperature!
"Temperature: -40° to 100°C (-40° to 212°F) with Buna seals; -26° to
204°C (-15° to 400°F) with Fluorocarbon seals"
These are interesting, but too large and not heat tolerant:
http://www.aamextras.com/main/PrdMenu/Valves/TAC-Zone-ELEC.pdf
http://cgproducts.johnsoncontrols.com/CAT_PDF/1900191.pdf
The AB12 model looks interesting:
http://www.webbersupply.com/Webberpdf/section5ws_89_104.pdf
This company is interesting:
http://www.ascovalve.com/
They have a miniature valve line too.
http://www.ascovalve.com/Applications/News/NewsASCOMiniatureValve.aspx
Interesting that the pulse code modulation PCM variable valves
operate at kHz frequency.
I suspect I'm looking in the wrong places for the kind of valves
desired.
If I can get a good realistic computer simulation then it might be
fun to build a physical E-cat simulator using that design. OTOH, it
would be a much better use of time to resume other experimentation.
I slightly changed the format of my web site, but it is all the same
old stuff. I thought it felt better starting off with my haiku
regarding cold fusion:
http://www.mtaonline.net/~hheffner/
Best regards,
Horace Heffner
http://www.mtaonline.net/~hheffner/
