Dave--
Your model of the Mizuno test set up is what I would say is well validated.
Validation of models over the range of their application is good engineering
practice. However, it is frequently a chore which is neglected in research
activities because it points to embarrassing short comings in theories or
hypotheses. Researchers have been known to avoid data that is problematic.
Bob Cook
----- Original Message -----
From: David Roberson
To: vortex-l@eskimo.com
Sent: Friday, January 30, 2015 3:26 PM
Subject: Re: [Vo]:Alternate Calculation and Calibration Method for Mizuno
Report
My model is able to take modest ambient variations into account and subtract
their effect from the measured coolant temperature. This is not too difficult
since the external time constant is so large. It also is able to handle pump
leakage power very well. It turns out that the ambient can be broken into two
parts. One is the average that is applied to the system during the period of
interest and the second part is the variation from that average. The pump
power leakage is added to the constant component of the ambient to end up with
a step response that operates through the thermal resistance as it charges the
thermal capacitance.
If you recall from simple electrical theory, the voltage step response of a
series resistance leading to a parallel capacitance is an exponential function
of the type V(t) = V(0) * (1 - e^(t/rc)). Here the thermal resistance and
capacitance is what we know, and of course the time is real time. The V(0) is
the final temperature (voltage in a real rc) that the capacitance reaches
(coolant temperature) when many time constants has elapsed. It so happens in
this system that V(0) is determined by taking the constant average value of the
ambient temperature and adding it to the constant pump power multiplied by the
system's thermal resistance.
It should be understood that we can adjust the pump power term to trim slight
unknowns associated with those two constants. For example Jed mentions that
the thermometers are not exact or perfectly match. In that case we adjust the
equivalent pump power term slightly to compensate for the error. After this
adjustment excellent agreement is seen over a wide range of temperatures.
The variation with ambient over time, determined earlier by subtraction from
the mean, is then treated as a signal that is applied to the series resistance
to parallel capacitor network. I used a single integrating term to handle this
process. The time steps are those given in the report exactly as posted. In
this particular system the effect of this noise is modest when compared to the
large step response generated by the constant average ambient temperature
input. If I recall it amounts to less than .2 degrees C at most points in
time and is handled properly by the model.
Keep in mind that my simulation very accurately follows the real life model.
If it were poorly constructed that would not happen. The time domain signal
due to the input heating pulse is clearly visible along with any excess power
that may be generated. If it works this well with the amount of ambient
variation presently measured then it is going to be quite a powerful tool once
that is reduced.
Jed, you really should take a step back and realize that the excess power
numbers that you reported are not correct. As I pointed out, the heat energy
contained within the thermal capacitance of the test system leaked out enough
during the evening hours when the heating system was turned off to invalidate
the measurements taken too soon afterwards. This is obviously displayed within
the figures that you posted. The water and or cell temperatures were already
rising before any of the pulses were applied to the device. This was certainly
due to the fact that the step in average ambient temperature plus its pump
addition is starting to drive the system temperature towards a temperature that
is significantly higher than the initial coolant temperature at the beginning
of the test period.
Also, as I have pointed out several times, I can see each of the individual
pulses and measure the change in temperature they generate. This can be done
on a very short time frame before a significant amount of energy can leak away
from the thermal capacitance. The numbers as reported in November would be
seen as a huge increase that I could not miss. And, you need to realize that
most of the excess power you calculated will evaporate as soon as the ambient
is held constant. My model is much better than you appear to believe.
It is important that the we ultimately determine the true behavior of the
LENR systems that we are testing. That is all that I am attempting to do at
this time and if I make an error then it is great for someone to point it out
to me so that I can learn from the experience. I assume that everyone else
shares that desire.
Dave
-----Original Message-----
From: Jed Rothwell <jedrothw...@gmail.com>
To: vortex-l <vortex-l@eskimo.com>
Sent: Fri, Jan 30, 2015 3:02 pm
Subject: Re: [Vo]:Alternate Calculation and Calibration Method for Mizuno
Report
Bob Cook <frobertc...@hotmail.com> wrote:
All measurements should be accomplished with as much precision as possible,
since adiabatic calorimetry is not possible without adiabatic conditions.
You mean it is not perfectly insulated. No system is. When the level of heat
is very small, such as the heat from the pump, the system soon reaches a
terminal temperature.
This instrument works for sustained power levels of ~2 W to ~30 W. The heat
from the pump is too low to measure with confidence using this instrument. No
calorimeter works well at any power level. It is not possible to make adiabatic
conditions for any temperature or power level.
As Dave has indicated the heat lost of the pump to the ambient is not
adiabatic and of significant amplitude relative to excess heat generation over
time.
Yes. It is not adiabatic. If it were adiabatic, the water would not come to a
terminal temperature. It would not record 14 hours at 17.85°C average, and the
next 14 after that at 17.81°C. That is remarkably stable. Fluctuations from
ambient air temperature hardly affect it.
With such a low power level it soon converts into an isoperibolic system.
I tried to estimate the pump power based on the difference between the
ambient and the terminal temperature. I find this is not possible because
the ambient temperature is unstable and it varies from one place to another.
The air is being moved around by fans and room heaters. I see from my lab notes
that when I placed the two Omega handheld thermocouple probes in different
locations they often measured air temperature differences of 0.3°C and
sometimes more. The Omega typically measured air temperatures close to the
reactor at about 0.3°C lower than the reading from Mizuno's ambient
thermocouple. So there's really no telling how much cooler the air was than the
cooling water. It is somewhere between 0.6 and 1.2°C, I believe. That is more
of a guess than a measurement.
If I were trying to do calorimetry based on the difference between the
cooling water and the air, the answer would be inaccurate to the point of being
useless. Fortunately, I need only compare the cooling water at the start of the
test to the end of the test.
- Jed
