David Roberson <[email protected]> wrote: I suspect that it is not easy to simulate the actual heat exchanger and > environment of the Rossi test. A true test would require an exact copy of > the one he used, but that is not going to be possible. >
I think a true test is one that addresses the specific physical question without extraneous stuff such as air pockets or the efficiency of the heat exchanger. I think a simplified test is better. It is better to test the air pocket hypothesis separately. The idea of using a cold water copper pipe and blow torch with two TCs is > about as good as we will get. > I was kidding about that. A blow torch is too hot. The temperature difference is too extreme. It is around 2000°C. I think a heater around 100°C would be better. Maybe a copper pipe running through boiling water. Not as hot as Rossi's steam, but pretty hot. Maybe a gas grill flame some distance from the pipe would be a good approach. A compromise. That is with cold water running through a pipe being heated. How about hot water running through a pipe being cooled down with ice? That's a lot easier to arrange. > Try to obtain a copper pipe with the largest OD/ID ratio as possible since > the manifold appears to be composed of very thick metal. > I think that would confuse the issue somewhat. Actually, you need a steel pipe, from the looks of it. That has less conductivity. A copper pipe can be thinner and have the same conductivity, or better. Anyway, you can simulate one or the other by moving the TC. You can test at multiple points to make a profile. Rossi's flow rate is 10.6 L/min. which I cannot achieve with a small copper pipe. Remember, we are not trying to determine the efficiency of this measurement technique. In my kitchen there is only a 1°C difference between the pipe surface and the fluid temperature. Suppose in isolation we tested a thick pipe and a large nut, and we found there is a 5°C difference with Rossi's pipe. It is much cooler on the surface. That would mean he is losing much of the heat. His actual results are much better than he thinks. That would be a mistake that *reduces* the estimate of the heat. We don't care about such mistakes. Even if his estimate is only half of the actual that does not matter. (I suspect it is roughly half, especially given all heat that radiates from the reactor vessel.) The only thing we are concerned about here are potential mistakes that incorrectly *increase* the estimate of anomalous heat. So the only thing we want here is some measure of the heat conducted directly by the pipe, to see if it can be a significant fraction of the heat convected by the water in the pipe. If it is 5%, that is not worth bothering about. If I can measure it at all with an ordinary TC meter I will be surprised. With laboratory grade calorimetry such as McKubre does, a 5% difference would stand out like a Broadway Marquee. With the kind of calorimetry they use in factories, you would never notice it, and nobody would care if they did. - Jed
