Alan, Thank You for voicing these concerns – I wanted to suggest the same focus on fine control of the drive because of the inherent latency in sensing thermal data. The suggestion of induction heat is excellent and I would even go so far as to recommend a PWM induction heater where different control algorithms could be tested over time – I think you need a control scheme just to approach the window so slowly that the latency can be tolerated without cell destruction.. Again I think robust heat sinking is required to establish this sort of tolerant environment where a sensor [pressure/ temp or both] near the source can capture thermal excursions as the heat is sucked past on it’s way out of the system. I also wonder if the fan cooling Bob is planning will be enough to avoid run away.
I wonder if the sci fi examples of reactions in a big fish tank are actually a better suggestion for safety and cooling control? I could see an upside down clear trough submerged in the tank with the reactor fixed inside just above the waterline. Cooling the tube would be quickly controlled by air pressure to vary the waterline inside the trough and submerge the reactor to different levels as part of the control loop. The other advantage being safety since you now have a reactor that is surrounded on all sides by water. Fran From: AlanG [mailto:[email protected]] Sent: Monday, March 02, 2015 8:47 PM To: [email protected] Subject: EXTERNAL: Re: [Vo]:Questions Raised by Parkhomov Experiment Failure Consider this issue in the context of Dave Roberson's analysis. Suppose there is a critical temperature where XH suddely appears in part of the fuel. That heat pulse could propagate rapidly through the fuel, causing the positive feedback Dave identified. Such a hot spot could increase pressure and stress on the tube very quickly, leading to failure. It seems likely that given a thermal time constant of several seconds or more, negative feedback control of the heater by an external thermocouple won't be fast enough to catch a runaway reaction. A pressure sensor would be much faster, but there would still be substantial delay as heat would continue to propagate through the tube into the core, after heater power is cut. Bob's fan-assisted convection cooling is a good safety measure, but will it be fast enough? I look forward with great anticipation to his experiment. This is a classic servo problem, and two other solutions come to mind. The system could run open-loop, with an algorithm that reduces heater power as the core approaches the critical temperature. Lots of research and data collection are needed to fine tune the control parameters, and those will vary between cores and fuel mixtures. What if the reactor used induction heating rather than a resistance coil. The heat is generated directly in the fuel mass, so the thermal response time is much faster. With a pressure sensor, maybe the control loop would be quick enough to correct for the positive feedback at the critical temperature. On 3/2/2015 10:57 AM, Bob Higgins wrote: Yes, this is why I was worried about Alexander's new design that puts an air gap and another ceramic around the reactor core - increasing its thermal resistance to the lower temperature air around it and/or the water in the calorimeter. He already showed that if he put alumina powder insulating the reactor (to lower the input power to get it high temp) that it failed catastrophically. I am working on a large water volume calorimeter in which to test my Parkhomov-like reactors. It will include a variable convection fan to change the thermal resistance between the reactor tube and the cool water in the surrounds. This convection will only be activated if the heater coil power is turned off and the temperature of the reactor continues to rise. The intent in this design is for the water to never reach boiling. Also, the calorimeter will be a good shield for any explosive shrapnel (primarily alumina shards). The calorimeter also provides a port to measure radiations with low mass density between the reactor core and the sensor. Here are 2 links to diagrams of the calorimeter I am building: https://drive.google.com/open?id=0B5Pc25a4cOM2R1RYemRlTEdZSEE&authuser=0 https://drive.google.com/open?id=0B5Pc25a4cOM2dWh5VXJFSU1uT1U&authuser=0 Here is the link to my planned ZDV plumbing that will allow me to measure the pressure, sample the gas product, and vent the system before opening the reactor post-experiment. https://drive.google.com/open?id=0B5Pc25a4cOM2WU9MR3hyQ2NIWkE&authuser=0 As Bob Greenyer likes to do, he has dubbed this system in ASCII as "{Garbage Can}". Bob On Mon, Mar 2, 2015 at 11:40 AM, Roarty, Francis X <[email protected]<mailto:[email protected]>> wrote: Thanks Bob, that is great information but I still have a nagging concern that thermal loading is more important than anyone is currently aware and that XH needs an environment that is robustly subtracting heat away from an otherwise self destructing cell to rise above the noise. IMHO researchers need to perform something equivalent to an isometric where they are vigorusly fighting their own heating effort via thermal loading and then repeatedly push the drive thru the threshold temp while slowly increasing the load..and …with luck..decreasing the drive [I think this what Rossi has been doing]. The stories about life after death, evaporating water and explosions where reactors were left leaning in a bucket of water may have created a thermal gradient centered about the waterline that eventually favored a particular area within the tube and powders with just the right properties to run away…. Shot gunning by accident. Fran
