Hello to the collective: I hope it is not too soon to begin discussions that are of the "heavy" nature. We will be better off now with a reduction in the number of extraneous posts unless the guilty parties decide to change names and reappear.
Back to work: The intrinsic delay of the internal heating process of the LENR devices is not well defined as far as I have determined. I have asked for information concerning the P(T) function from both Rossi and Defkalion to no avail. They either want to keep this information secret for trade reasons or have not measured it. In either case, I have not received any guidance. I would hope that a calibration curve and testing of the devices would be conducted before they were released for sale, where the control mechanism is designed to compensate for these effects. So far I have not seen any data that helps me to know how much margin is available between operation temperature and meltdown levels. It is hoped that the thermal heat capacity of the core region, which includes the large metal framework and coolant, in Defkalion's case, will slow down the rate of change of the temperature at the outside of the active cores. This slowdown coupled with a feed forward plan that is obtained by experience could hopefully save the day. Defkalion seems to monitor enough parameters to allow them the ability to operate in the quasi unstable mode. Here I use the expression quasi unstable mode to refer to operation where the core supplies virtually all of the heat that keeps itself active at the desired power output level. In my estimate, the ideal operation is when the input heating power is reduced to approximately zero and the coolant flow rate and temperature are carefully controlled to extract the LENR generated power which is then heat exchanged with the outside loops. The spice model that I have been using is mainly intended to expose open loop instabilities. I have been attempting to determine how difficult it is to have a large ratio of output to input with real world like transfer functions. It might be possible to add the feedback networks to simulate a system, but that has not been done at this time. I am amazed at how unstable LENR systems with nonlinear transfer equations can become even when the LENR power is small relative to the drive power. It is a good thing that LENR reactors do not behave like their fission brothers. It is a lot easier to clean up molten nickel than radioactive materials. Dave P.S. I have played around with my model and was amazed to see that it actually showed that a gain of approximately 6 for (output power)/(input power) was obtained with a third order functional relationship of P(T). In this particular mode, I allowed the device to go into an unstable state by overdrive at the input followed by zero drive level. The drop in drive to zero quenched the instability and caused the output power to fall rather slowly. After the output fell to a level that was still above the drive, I reapplied the full drive. Again the simulated LENR output rose upward just as before and was quenched by drive reduction as during the first cycle. This cyclic behavior seemed to be repeatable and the amazing thing about it was that the drive duty cycle was as Rossi has stated at 25%. Now I am not sure whether or not my model action was a coincidence, but it does add credibility to Rossi's claim. Further playing around with this spice model is definitely in the works. I need to see if similar behavior appears when the P(T) function is modified. -----Original Message----- From: Alain Sepeda <[email protected]> To: vortex-l <[email protected]> Sent: Mon, Jan 23, 2012 3:44 am Subject: Re: [Vo]: Stabilization of the reactor (was: Opponents should please go away and form your own group) Interesting work. as far as I know the biggest problem to stabilize, even if we have a negative feed back, will be to compensate intrinsic delay, like the thermal delay... in nuclear reactor early theory ignored delayed neutrons, and in theory it would have exploded... hopefully nature was nice, and the neutronic loop was damped a little. not so nice for LENR it seems. even if the coolant is very strongly cooling the reactor, evacuating heat massively for even one degree, a delay of say 1 second in heat propagation can let a small reactor get melted, especially because of micro-powder. Maybe this explain why there are so many difficulties to raise COP above 2 in labs. maybe also it explain whay powder is only micrometer... a compromise. maybe research should focus on stability, and not power. Research should try to design intrinsic fast negative feedback to go further . but even if the feedback is intrinsic it can be too slow to avoid random melting above some point. It is funny that everybody is spinning about quantum physic and video analysis of calorimetry, and that key explanation of results and behaviors might be around basic "control engineering". 2012/1/23 David Roberson <[email protected]> You have shown an excellent example of the problem and Defkalion's solution. The model that I am working on suggests that it is the slope of the Power Output versus Temperature curve of the core device multiplied by the thermal resistance that defines stability. I have modeled many different power functions with a simple spice model and they always go unstable when the product of these functions is greater than or equal to 1. Also, the incremental gain of the system follows the normal feedback relationship of: gain=1/(1+af). Here af is the product of dP/dT and R(thermal impedance). With a linear Power Output versus Temperature curve, you can achieve any desired COP by adjusting the thermal resistance of the core path. A gain of 6 can be set with ease. Nonlinear functions, such as third order relations, are very unstable due to the rapid change in slope versus temperature. My model suggests that there is little hope of significant controlled gain with these functions without being able to drain heat away from the core rapidly. In all of the non linear cases I have modeled, the (power output)/(power input) is barely above 2 in the best design. Of course I refer to the situation where the linear term is much smaller than the high order term. An exponential curve as in P(T) =a* EXP(b*T) allowed the excess power to be about 40% above the drive before instability occurred. Dave -----Original Message----- From: Robert Lynn <[email protected]> To: vortex-l <[email protected]> Sent: Sun, Jan 22, 2012 4:21 pm Subject: Re: [Vo]:Opponents should please go away and form your own group I am pretty sure Rossi's stability and control problem stems from relying on heat transfer through a large temperature differential (low Watts per degree) to cool a reaction that has a positive temperature coefficient (ie gets more powerful with increasing temperature). The simple fix is to use high temp coolants with greater Watts per degree heat transfer rates via fins or coolant tubes through the reactor etc. As an example (using numbers plucked from air) say you are using a heat transfer setup that removes 10W per degree of temperature difference with water at 100°C and Ni powder that produces 4kW at 500°C. This 400°C temperature difference results in 4kW of heat transfer. Now assume that your power output doubles to 8kW when the Ni powder temperature rises to 600°C (ie positive temperature coefficient). Unfortunates the 500°C temperature differential only increases your heat transfer rate to 5kW. Result is uncontrollable thermal runaway; E-cat go boom (or melt). Now instead if you use a coolant at 450°C and heat transfer setup that removes 80W of heat for every degree of temperature difference then with Ni temprature of 500°C you get 4kW of power and 4kW of heat transfer as before, but at 600°C you get 12kW of heat transfer from only 8kW of heat output. Result being that the reaction can no longer run away or increase above 4kW 500°C. Most engineers would quickly see this consequence from the nature of increasing power with temperature (apparent from almost all reports of gas-nanopowder LENR). I think Defaflion got it straight away with their high temp coolant, but I don't think Rossi did (or does?) as he has persisted in using low pressure water as a coolant. On 22 January 2012 20:30, David Roberson <[email protected]> wrote: It is strange that anyone would want an unproved and expensive device. I suspect that Rossi thinks that he can work with NI and stabilize the thing, and this may be true. My personal opinion is that some serious engineering work will be required to make the system safe and repeatable. If I were Rossi, I would be looking into a method of core cooling that is active and powerful. The core itself probably should be operating in the thermal run away mode to get the COP into an acceptable range while the cooling needs to be able to prevent additional heat energy from resulting in much higher core internal temperature. The approach used by Defkalion appears to address my issues. Their design includes a very tight thermal control of the core region by the 6 coolant paths. To startup, they would reduce the coolant flow to a minimum allowing the electrical heater to easily raise the core temperature. Once the core reaches an unstable temperature, it will begin to heat rapidly on its own. At that point the coolant flow rate can be increased to absorb the excess heat and achieve the final desired operating temperature. All of the heat energy required to keep the device operating would now be supplied by the core. The overall COP at this point is infinite in the core itself, but the control and pump energy drains would make the net COP as specified. Rossi may not understand the problems that he faces in this task. Actually, no one may really know at this point. The model I am using is quite simple, but makes sense to me. That is no proof that it is accurate however. Why would we expect Rossi to reveal to us his major problems? Most engineers assume that the problems will be defeated sooner or later and see no reason to air the dirty laundry. He has a positive outlook and has overcome many obstacles in his life and these issues appear minor in comparison. To him, the solutions most likely seem just a little way off. Dave -----Original Message----- Name of poster deleted..... On Sun, Jan 22, 2012 at 11:55 AM, David Roberson <[email protected]> wrote: Mary, there are serious problems with Rossi's demonstrations that we are all aware of. It is apparent to me that he has a very difficult problem trying to maintain stability of the power output and I have been doing some interesting simulation that tends to support this claim. The October 6 test data shows a clear fingerprint of LENR heat production which I hope to explain soon. All of the other models that I have seen thus far do not respond in a manner that comes even close to explaining the anomaly. These models have been based upon energy storage and release from a large mass of material inside the smaller cube. A better explanation for the curve can be obtained by assuming that a large peak of excess thermal energy is released at the end of the drive cycle due to an inherently unstable thermal run away process that is quenched just before it becomes unstoppable. If so, shouldn't Rossi be telling us that? Do you think he told his anonymous customer who supposedly bought **13** "power plants" consisting of some 600+ individual modular units? Do you believe there is such a customer? Is someone really that dense? What would 13 such things possibly be used for? I would hope that a calibration curve and testing of the devices would be conducted before they were released for sale, where the control mechanism is designed to compensate for these effects. So far I have not seen any data that helps me to know how much margin is available between operation temperature and meltdown levels. It is hoped that the thermal heat capacity of the core region, which includes the large metal framework and coolant, in Defkalion's case, will slow down the rate of change of the temperature at the outside of the active cores. This slowdown coupled with a feed forward plan that is obtained by experience could hopefully save the day. Defkalion seems to monitor enough parameters to allow them the ability to operate in the quasi unstable mode. Here I use the expression quasi unstable mode to refer to operation where the core supplies virtually all of the heat that keeps itself active at the desired power output level. In my estimate, the ideal operation is when the input heating power is reduced to approximately zero and the coolant flow rate and temperature are carefully controlled to extract the LENR generated power which is then heat exchanged with the outside loops. The spice model that I have been using is mainly intended to expose open loop instabilities. I have been attempting to determine how difficult it is to have a large ratio of output to input with real world like transfer functions. It might be possible to add the feedback networks to simulate a system, but that has not been done at this time. I am amazed at how unstable LENR systems with nonlinear transfer equations can become even when the LENR power is small relative to the drive power. It is a good thing that LENR reactors do not behave like their fission brothers. It is a lot easier to clean up molten nickel than radioactive materials. Dave P.S. I have played around with my model and was amazed to see that it actually showed that a gain of approximately 6 for (output power)/(input power) was obtained with a third order functional relationship of P(T). In this particular mode, I allowed the device to go into an unstable state by overdrive at the input followed by zero drive level. The drop in drive to zero quenched the instability and caused the output power to fall rather slowly. After the output fell to a level that was still above the drive, I reapplied the full drive. Again the simulated LENR output rose upward just as before and was quenched by drive reduction as during the first cycle. This cyclic behavior seemed to be repeatable and the amazing thing about it was that the drive duty cycle was as Rossi has stated at 25%. Now I am not sure whether or not my model action was a coincidence, but it does add credibility to Rossi's claim. Further playing around with this spice model is definitely in the works. I need to see if similar behavior appears when the P(T) function is modified. -----Original Message----- From: Alain Sepeda <[email protected]> To: vortex-l <[email protected]> Sent: Mon, Jan 23, 2012 3:44 am Subject: Re: [Vo]: Stabilization of the reactor (was: Opponents should please go away and form your own group) Interesting work. as far as I know the biggest problem to stabilize, even if we have a negative feed back, will be to compensate intrinsic delay, like the thermal delay... in nuclear reactor early theory ignored delayed neutrons, and in theory it would have exploded... hopefully nature was nice, and the neutronic loop was damped a little. not so nice for LENR it seems. even if the coolant is very strongly cooling the reactor, evacuating heat massively for even one degree, a delay of say 1 second in heat propagation can let a small reactor get melted, especially because of micro-powder. Maybe this explain why there are so many difficulties to raise COP above 2 in labs. maybe also it explain whay powder is only micrometer... a compromise. maybe research should focus on stability, and not power. Research should try to design intrinsic fast negative feedback to go further . but even if the feedback is intrinsic it can be too slow to avoid random melting above some point. It is funny that everybody is spinning about quantum physic and video analysis of calorimetry, and that key explanation of results and behaviors might be around basic "control engineering". 2012/1/23 David Roberson <[email protected]> You have shown an excellent example of the problem and Defkalion's solution. The model that I am working on suggests that it is the slope of the Power Output versus Temperature curve of the core device multiplied by the thermal resistance that defines stability. I have modeled many different power functions with a simple spice model and they always go unstable when the product of these functions is greater than or equal to 1. Also, the incremental gain of the system follows the normal feedback relationship of: gain=1/(1+af). Here af is the product of dP/dT and R(thermal impedance). With a linear Power Output versus Temperature curve, you can achieve any desired COP by adjusting the thermal resistance of the core path. A gain of 6 can be set with ease. Nonlinear functions, such as third order relations, are very unstable due to the rapid change in slope versus temperature. My model suggests that there is little hope of significant controlled gain with these functions without being able to drain heat away from the core rapidly. In all of the non linear cases I have modeled, the (power output)/(power input) is barely above 2 in the best design. Of course I refer to the situation where the linear term is much smaller than the high order term. An exponential curve as in P(T) =a* EXP(b*T) allowed the excess power to be about 40% above the drive before instability occurred. Dave -----Original Message----- From: Robert Lynn <[email protected]> To: vortex-l <[email protected]> Sent: Sun, Jan 22, 2012 4:21 pm Subject: Re: [Vo]:Opponents should please go away and form your own group I am pretty sure Rossi's stability and control problem stems from relying on heat transfer through a large temperature differential (low Watts per degree) to cool a reaction that has a positive temperature coefficient (ie gets more powerful with increasing temperature). The simple fix is to use high temp coolants with greater Watts per degree heat transfer rates via fins or coolant tubes through the reactor etc. As an example (using numbers plucked from air) say you are using a heat transfer setup that removes 10W per degree of temperature difference with water at 100°C and Ni powder that produces 4kW at 500°C. This 400°C temperature difference results in 4kW of heat transfer. Now assume that your power output doubles to 8kW when the Ni powder temperature rises to 600°C (ie positive temperature coefficient). Unfortunates the 500°C temperature differential only increases your heat transfer rate to 5kW. Result is uncontrollable thermal runaway; E-cat go boom (or melt). Now instead if you use a coolant at 450°C and heat transfer setup that removes 80W of heat for every degree of temperature difference then with Ni temprature of 500°C you get 4kW of power and 4kW of heat transfer as before, but at 600°C you get 12kW of heat transfer from only 8kW of heat output. Result being that the reaction can no longer run away or increase above 4kW 500°C. Most engineers would quickly see this consequence from the nature of increasing power with temperature (apparent from almost all reports of gas-nanopowder LENR). I think Defaflion got it straight away with their high temp coolant, but I don't think Rossi did (or does?) as he has persisted in using low pressure water as a coolant. On 22 January 2012 20:30, David Roberson <[email protected]> wrote: It is strange that anyone would want an unproved and expensive device. I suspect that Rossi thinks that he can work with NI and stabilize the thing, and this may be true. My personal opinion is that some serious engineering work will be required to make the system safe and repeatable. If I were Rossi, I would be looking into a method of core cooling that is active and powerful. The core itself probably should be operating in the thermal run away mode to get the COP into an acceptable range while the cooling needs to be able to prevent additional heat energy from resulting in much higher core internal temperature. The approach used by Defkalion appears to address my issues. Their design includes a very tight thermal control of the core region by the 6 coolant paths. To startup, they would reduce the coolant flow to a minimum allowing the electrical heater to easily raise the core temperature. Once the core reaches an unstable temperature, it will begin to heat rapidly on its own. At that point the coolant flow rate can be increased to absorb the excess heat and achieve the final desired operating temperature. All of the heat energy required to keep the device operating would now be supplied by the core. The overall COP at this point is infinite in the core itself, but the control and pump energy drains would make the net COP as specified. Rossi may not understand the problems that he faces in this task. Actually, no one may really know at this point. The model I am using is quite simple, but makes sense to me. That is no proof that it is accurate however. Why would we expect Rossi to reveal to us his major problems? Most engineers assume that the problems will be defeated sooner or later and see no reason to air the dirty laundry. He has a positive outlook and has overcome many obstacles in his life and these issues appear minor in comparison. To him, the solutions most likely seem just a little way off. Dave -----Original Message----- Name of poster deleted..... On Sun, Jan 22, 2012 at 11:55 AM, David Roberson <[email protected]> wrote: Mary, there are serious problems with Rossi's demonstrations that we are all aware of. It is apparent to me that he has a very difficult problem trying to maintain stability of the power output and I have been doing some interesting simulation that tends to support this claim. The October 6 test data shows a clear fingerprint of LENR heat production which I hope to explain soon. All of the other models that I have seen thus far do not respond in a manner that comes even close to explaining the anomaly. These models have been based upon energy storage and release from a large mass of material inside the smaller cube. A better explanation for the curve can be obtained by assuming that a large peak of excess thermal energy is released at the end of the drive cycle due to an inherently unstable thermal run away process that is quenched just before it becomes unstoppable. If so, shouldn't Rossi be telling us that? Do you think he told his anonymous customer who supposedly bought **13** "power plants" consisting of some 600+ individual modular units? Do you believe there is such a customer? Is someone really that dense? What would 13 such things possibly be used for?
