A lot of this depends on the internal heat conductivity structure of the device not developing hot-spots that runaway. Is there a good model of this conductivity structure?
On Wed, Oct 9, 2013 at 1:42 PM, Robert Lynn <[email protected]>wrote: > That would be a very simple means of providing excellent high temperature > control. > > A coil of tube inside the reactor containing water with a pressurised cold > reservoir attached to one end to keep it filled with water and a pressure > relief valve at the other end to release steam above a certain pressure. > The pressure release setting could control the temperature very > accurately, at any point from 100°C to 6-700°C, and the steam from the > pressure relief valve could be sparged into a water tank for simple > calorimetry. > > The water filled tube would absorb relatively little energy until the > rector temperature rose above it's set point, at which point it would > absorb a huge amount of energy with small temperature increase. > > Would give safe reliable hot-cat operation without danger of thermal > run-away. > > > On 9 October 2013 16:18, <[email protected]> wrote: > >> On 9/10/2013 12:20 PM, David Roberson wrote: >> >> I finally got around to checking my ECAT model performance with active >> cooling control and the results were very interesting. First, I applied >> normal heating to the device which leads to thermal run away conditions if >> allowed to go beyond a certain critical core temperature. As expected, >> the model showed that the ECAT began a path toward melt down. >> >> The path to destruction would proceed even when the original drive is >> removed since the critical temperature was exceeded. Then, I allowed the >> model to continue heating for a period of time and applied the brakes >> quickly with a new load that withdraws a significant amount of extra heat >> power from the system. This might be possible in real life if for instance >> a phase change coolant were sprayed onto the core directly. The extra >> cooling must be applied continuously while the temperature drops and >> until an optimum trip temperature is reached. The turn around point must >> be above the normal critical temperature since no additional heating is >> required to achieve a condition where thermal run away can begin anew. >> >> The model demonstrated that this type of cycle could be repeated >> indefinitely while the total power being generated by the device is >> significantly above that safely obtained by heating control alone. I would >> assume that a cooling method similar to that suggested above would take >> less input power than the standard heating process that I have modeled >> earlier so this type of control would result in a higher COP for the system. >> >> The model also demonstrated that a continual application of the extra >> cooling resulted in the cores return to room temperature as desired. >> Actually, the powerful cooling was not required to be applied once the >> critical run away point was passed. In this case, the core would be >> subjected to positive feedback that forces it toward turn off by itself. >> >> It is evident that a hybrid type of control system that uses both power >> resistive heating as well as active cooling would perform the function. If >> both techniques were available it might be possible to keep the ECAT >> temperature very near the critical point in which case the COP would be >> extremely high. For this type of tight control to work the loading as well >> as all the other parameters which cause the critical temperature to vary >> must be kept under tight control. This might be possible. >> >> >> Good to see someone thinking! In fact I think the problem is really >> simple. (And there should be no "extremely high" for overunity devices - >> if they are not beyond infinity then you haven't got an engineer worth his >> salt working on it). There should be no need even for an active system. >> You simply need a passive system that presents a steeper load onset with >> temperature (cooling effect) than the (expected) exponential power increase >> with temperature produced by the reaction. >> >> With a few moments thought you can come up with a really simple one:- >> Put the reactor into a stainless thermos flask, cover it with water, and >> heat it up with its internal heater until the overpressure release of the >> flask starts to leak steam. Disconnect the power and it should >> self-regulate and hiss away for hours or days and there is your demo (until >> it runs out of coolant). If you want a continuous system, then you might >> need to put it on an electronic scale and organise a little pump to inject >> more coolant very slowly to just keep the weight constant. >> >> Consider a Thermos flask that is capable of keeping tea hot for maybe 6 >> hours. If we have a 2 liter flask and the temperature drops from 100 down >> to 70 degrees in 6 hours, then that represents an total heat loss of >> 4.2kJ/kg/K * 2kg * 30K => 250kJ or an average heat loss rate of >> 2.5*10^5J/(6h*3600s/h) => 12watts. If we do the calculation a little bit >> better and assume that it is an exponential decay which is heading for 20C >> after infinite time then we get a time constant of 12.8h, and the rate of >> heat loss at the start is simply the total heat divided by the time >> constant (4.2*10^3*2*(100-20)/(12.8*3600)) => 14.6watts. So this means >> that if we supply 15 watts of power by some means, then whatever is in the >> flask will eventually reach boiling point and stay there. >> >> So if the reaction can produce more than 15 watts of power while it is >> held at 100C, then it must remain at this temperature because as soon as it >> rises above 100C, the water will boil and steam will exit the flask >> providing an almost infinite heat sink capability (boiling water can easily >> consume kilowatts of power without allowing the temperature to rise >> significantly). If 100C is not hot enough then we can either pressurise it >> (with a simple overpressure valve), or go to a lot more trouble and use a >> higher boiling point fluid. The sort of silicone oil that they use in >> diffusion pumps comes to mind. You would need to have a closed system (at >> least not allow oxygen to get to it I think) but you could easily get >> boiling points around 300C which could be further raised by pressure. >> Mercury would be an excellent boiling point coolant for 360C and up. (It >> used to be used and worked very well in diff pumps before people got all >> paranoid about it). >> >> Rossi's "December Test" suggested a power generation capability of ~2kw >> while being held at a peak core temperature of approx 500C. The "March >> Test" suggested a power generation capability of 530w while being cycled >> around a temperature of approx 320C at the core. Provided the core reactor >> in the two cases is very similar, we can use these two points to solve >> Arrhenius' equation for the temperature dependence of the reaction rate. >> We get a pre-exponential factor of 1.6*10^5 and exponential (activation >> energy / R) factor of 3380. Arrhenius' equation thus suggests a power >> generation rate dependent on temperature given by Power = >> 1.6*10^5Exp-(3380/Temp in K). Using this equation to see what happens at >> 100C tells us that it should generate a power output of 18.6 watts. >> >> So Rossi's reactor should in fact work and generate steam continuously if >> simply put in a thermos flask and warmed up to 100C! If we were to >> pressurise a well insulated container to a reasonably easy to obtain >> pressure of say 10 atmospheres (giving a water boiling temperature of 180C) >> we would expect to be able to generate 92 watts in a continuous manner. If >> we were to set up a mercury cooling loop, then we could run it continuously >> at kilowatt levels with no active control and just a small pump to inject >> the condensed mercury back into the pressure chamber. >> > >
