I have been conducting numerous simulations of the expected behavior of a thermally controlled energy source such as the HotCat designed by Rossi. Now I have constructed a technique that can be utilized to characterize a design and determine many of its important parameters.
It would be advantageous to have an opportunity to retest the original device that was experimented upon by the recent third party team of scientists. They had the perfect test vehicle to use according to my plans. Perhaps MFMP will take time to perform the tests that I am suggesting. The characterization should begin by taking measurements upon the dummy reactor that contains no core fuel. I call this curve the device thermal design function and will refer to it as curve 1 in the remainder of this post. The procedure is to make an x-y plot of thermal input power versus device temperature. I prefer to place the temperature along the X axis and the input power along the Y axis. Careful measurement of the temperature of the device by means of a thermocouple and IR camera is required if accurate prediction of the final operation is desired. Points upon the curve are located by applying a calibrated input power to the heating elements of the unit and recording the final temperature of the device after it has stabilized. This process is repeated many times throughout the temperature range over which the device will operate. A smooth X-Y plot is the desired outcome of this procedure. The temperature axis needs to be in absolute dimensions such as degrees Kelvin. The final completed curve will be monotonic with rising temperature where the conduction and convection processes dominate the lower range while the radiation kicks in to dominate the high temperatures. Next a small quantity of fuel is added to the device. A limited amount is required to ensure that the test subject remains stable for the duration of the measurement procedure. Too much fuel inserted might well lead into thermal runaway or otherwise make the test difficult to complete. I refer to this curve as a system response function and shall refer to it as curve 2. Again, the same type of curve is generated with input power on the Y axis and absolute temperature along the X axis. Many pairs of X,Y data need to be measured so that a smooth curve can be generated as before over the expected operational temperature. Once these two curves are available you generate a third one which is the core power generation function that is referred to as curve 3. For each temperature point you read or calculate the power input from curve 1 and curve 2 along the Y axis of that particular curve. You will find that curve 2 will show a smaller value of input power than shown by curve 1 at each temperature point. This is because the power generated by the core is added to the input power in this procedure and it therefore takes less input power to achieve a desired temperature. So complete curve 3 by taking the difference in power readings between curve 1 and curve 2. This curve numbered 3 is a measure of the internally generated core power that is calculated for each temperature. At this point enough data has been collected in order to characterize the device at a given ambient temperature. A final device characterization curve can be generated by taking the curve number 3 and multiplying it by a factor proportional to the amount of mass inserted into the system. Twice as much core mass should generate approximately twice as much heat power at a given temperature. There may be interaction of some type that depends upon the amount of mass placed into the device, but a first order approximation is about as good as can be achieved without extensive measurements. An error in this determination can be corrected for by changing the amount of mass to obtain the desired results. We are not through quite this easily. After the multiplication is completed, you take curve 1 which was a measurement associated with the dummy device and subtract this latest curve 3 multiplied by a factor from it point by point at each temperature. After this new curve 4, which I call the device characterization curve, is generated the real magic begins to show up. If totally stable operation is desired you will note that the resulting curve 4 will be monotonic with temperature and never demonstrates a negative slope over any range of operational temperatures. This is of course true in the case of the dummy device and will remain that way until a sufficient amount of core mass is applied. Operation within this region can be done, but the COP will never be very reasonable since it appears to be limited to around 4 according to my simulations. Sufficient positive feedback must be provided in order to achieve a reasonable COP. For more appropriate COP the core mass factor is increased until the slope of the final curve 4 becomes negative at some critical temperature and input power combination. If the input power is increased until this critical power is applied the device goes into a negative resistance type of operation. Positive feedback of thermal power takes over and the temperature rapidly increases. An excellent design can recover from this situation provided the geometry is constructed to ensure that the amount of heat leaving the device by means of conduction, convection and radiation is always greater at every temperature than the amount of heat internally generated. Truly magnificent COP might be possible since the input can be throttled back while the device remains at the high temperature and hence high power generation condition. Once the input power is totally taken away the device can begin to cool back to room temperature. Other methods of active cooling might be applied in order to shut down the device if eliminating the input is not sufficient to stop the reactions. Earlier I found that operation within the negative resistance region was possible with a PWM drive input. This remains possible provided the temperature is prevented from reaching a level that either results in latch up of the device or thermal runaway. The key parameter needing to be controlled is the same as above. At all temperatures at which the device is operated the core generated power must be kept below the power that is lost by conduction, convection, and radiation. This implies that the input must be switched completely off at a controlled high temperature point and then reactivated as the device cools to a desired level. COP can be very reasonable with this form of operation and the earlier statements made my Rossi fit well within the simulation results. I remain concerned about how critical operation of this type appears for a long duration system. It is important to perform the same type of curve generation and procedures listed above using ambient temperature as a parameter if a complete understanding of the operation of these devices is desired. There is not much more that I can do beyond the present analysis until additional measured data is obtained. Perhaps the MFMP group can help to fill in the blanks. Dave
