I wrote:
> 1. Put a tap in the hose to draw off samples periodically. To get an > accurate temperature, you draw off 1 L into a Dewar (a thermos bottle), stir > vigorously and insert several thermocouples and thermometers. > Toss out the first liter and fill it again. Maybe put the Dewar into a large bucket and let the water run into it and overflow for a few minutes. Bring the body of the Dewar itself up to the same temperature as the water. I don't use an actual Dewar, since they are expensive. I use a thermos like this: http://www.target.com/p/Thermos-Stainless-Steel-Briefcase-Bottle/-/A-10318103 Only mine was free, from Office Depot. I have done this often with two thermocouples and three red liquid thermometers. Believe me, you get the same temperature on all 5 instruments to within 1°, and you get the right answer. Not precise, but accurate. There is no 2°C discrepancy. You can be darn sure of the Delta T, because you compare each thermometer to itself, when placed in a sample of tap water. Red liquid ones cost $25. See: http://www.omega.com/Temperature/pdf/GT-RL_THERMOMETERS.pdf If Rossi had done this we would know with absolute certainty what the inlet and outlet temperatures are to within 1°C. At that flow rate, that is comes to +/- 750 W accuracy. That's nothing to write home about. You can do better with electronic instruments and of course he should have used electronic instruments. You can trust any electronic computer based instrument to about 0.1°C, even the low quality ones. Assuming you are smarter than a boiled sheep and you remember to calibrate it and test it with an ordinary electric heater. My point is, with 5 handheld instruments costing ~$200 total, you could do this test and get data with such assurance that only a scientific illiterate would argue with it. (Let's say someone who believes you can store up 30 MJ once and then release them twice.) You would have proof that 3 hours into the heat after death the power was 6.3 kW +/- 0.8 kW. Measure it manually 4 times an hour, add in the computer log which has much better precision and thousands more data points, and Bob's your uncle. - Jed
