no need of a total theory to control a complex system. and even with a total theory about a mechanisme, the real system is often very different, much complex and simple that the predicted system.
the engineer method is to learn the characteristic of the system, upfront (eg: pulse response) and eventually continually (adaptative control). the interest of the theory is to know the limit of what you have measured (knowing that above a given limit it can diverge, saturate, oscillate, dampen...). if you try to control externally a structurally unstable system like an auto-catalytic reaction, with irreversible saturation (like hot gaz+powder LENR), the slow global control is a bad idea, since the external inertia of the system is too much to allow quick enough feedback. anyway you have various solution, that you should merge as needed. one first method is to find a structural feed back at the local level. ther is some in nuclear reactors, but seems non for LENR (no resonance). the second one is thermal dissipation feedback, that cause heat transfer to increase when temperature difference increase, so having a fluid at the similar temperature of the target, with high conductivity(if teperature difference is too high, cconductivity have to be lower else the cooling is so strong that the reaction is stopped), will stabilise the temperature naturally, up to a point where autocatalysis overcame the heat conductivity. H2 seems to be a good cooling (dPproduced/dT>Rthermal). H2 seems to be a good conductivity fluid, and powder have a good transfer caracteristic. another solution is to play with the different thermal resistance for different timescale, linked to thermal inertia. at short term the inertia of the powder, or reactor is the dominant factor. at longer term the external control can have an effect. another factor is tha LENR seems to be multi-stage, so ther might be 2 or more simultaneous decay/autocatalysis that happens at different timescale (probably 300ns for LENR eruption, then few seconds minutes for radioactive decay) so if you send pulse to activate the reaction, you can estimate that the reaction is between a grain and a heavy thermal mass at reactor temperature, and mostly activate the initial "eruption". when the pulse stop, the heat can be spread in the global reactor and the cooling fluid. the radioactive decay is then the dominant factor, and can be controlled with external slow feedback (like by cooling fluid speed control, via the flux calorimeter of DGT))so the idea of Defkalion to use pulse witdt/frequency control, is coherent with the idea to avoid reaction divergence at short term, and allow slow control of heat/temperature so the next pulse will happen in a stable environment. measuring the heat production, will also allow to measure the efficiency of the previous pulse, and compute the next one to be just enough efficient, avoiding meltdown and shutdown. all that make me convinced they have a good control, even if they have no validated theory, but just a phenomenological model and few theoretical assumptions the Rossi apparatus, manual throttle, cool but uncontrolled fluid, uncontrolled flow, in his older experiment seems not compatible with a good control, except if he have an intrinsic stabilizing factor (that seems false) like fission reactors. It is logic that in those condition, whe COP and temperature are pushed, it melt down. by the way it is a bit amateur. the push buttons stability claimed by rossi last visitor is much more coherent. however the fact that he says that the reactor take 1 hour to start, is a bit strange, but maybe it is only the slow heating up to working point, dow slowly to avoid unstability. maybe DGT have found a quick method to heat the reactor (their secret first phase heating) up to a controlled temperature, that avoid startup, but accelerate heating. good control of a LENr reactor seems to imply an adaptative controller, and internal flow calorimetry (flow control, and temperature sensors) to control the reaction efficiency. all to say that controlling LENR for me seems classic engineer job. much easier than controlling a supersonic fighter at transonic speed. there are wagons of student trained to solve that kind of problems, and I was in the same train as them 20years ago. 2012/4/2 David Roberson <dlrober...@aol.com> > In my opinion it will be far easier to control the variations in > performance and parameters which determine these variations once a proven > theory of operation exists. Until that time we will be stumbling along at > less than ideal performance. > > Dave > > > -----Original Message----- > From: Axil Axil <janap...@gmail.com> > To: vortex-l <vortex-l@eskimo.com> > Sent: Sun, Apr 1, 2012 7:21 pm > Subject: [Vo]:Stabilizing the E-Cat > > Stabilizing the E-Cat > > It is a common belief among knowledgeable cold fusion pundits that both > Rossi and DGT face challenges in controlling the reactions in their > reactors. One important reason for this is process variation. > > Naturally occurring variations in the physical properties of the > materials that are responsive to the reaction may vary widely. These > physical variations will produce corresponding variability in the reaction > produced by the various ranges in material composition. > > This range of variation in the material will reduce both the performance > and controllability of the material based on how widely the variation > deviates from the optimum specification regardless whether this variance > falls below or rises above that specification. > > Specifically for E-Cat micro powder, isotopic content, particle size and > shape, inter-particle contact points between each particle or sets of > particles, hydrogen flow patterns around particles, and hydrogen based > particle heat transfer dynamics may all be important material variation > parameters. > > Prefabrication, characterization and testing of subunits are all > important enablers of quality control and optimization through > standardization. > > The experience in process control gleaned from the semiconductor > industry teach how designers using process control modalities run tens to > thousands of simulations to analyze how the outputs of a circuit will > behave according to the measured variability of the transistors for that > standardization > process. The measured criteria for transistors are recorded in model files > given to designers for simulating their circuits before simulation. > > In this semiconductor example, if the variance causes the measured or > simulated performance of a particular output metric (bandwidth, gain, rise > time, etc.) to fall below or rise above the specification for the > particular circuit or device it reduces the overall yield for that set of > devices. > > One possible design approach is to stabilize the micro-powder onto a > substrate to make the material more like a transistor. > > Weld the micro-powder onto a nickel nanowire of thin film to stabilize > each particle’s mechanical contact environment. This enables each wire to > be tested reliably against a specification. Then sets of wires typified by > a common performance profile can be grouped and configured in a thermal > circuit driven by precise computer control. > > In common service: Axil > > >