*Basically, after 2.4 hours you should either ignore the rest of the data, or use a much more complex modeling method which takes into account the lag.* This is exactly what we have already done. You will read it soon. So please publish the original file so that Dave and other can check our results. We could discuss in detail the matter at ICCF-19.
*Many thanks* 2015-01-10 18:28 GMT+01:00 Jed Rothwell <jedrothw...@gmail.com>: > Gigi DiMarco <gdmgdms...@gmail.com> wrote: > > >> just as an example, in the missing file, in the row 989 which corresponds >> to 24131.191 seconds the room temperature is 18.78 °C and the water >> temperature is 21.90. Doing some mathematics we get that the temperature >> difference is 3.13 °C that appears to be higher than what you say. >> > > That is because the ambient temperature is falling rapidly. The reactor + > Dewar are well insulated so they retain heat for a long time. It takes > several hours for them to "catch up" to ambient. They finally do catch up > early in the morning, except they remain slightly warmer because of the > heat from the pump. > > That is why I do not want to use this data. The rapid, large fall in > ambient greatly confuses the issue. It has confused you. > > You have confused a temperature difference from heat with a temperature > difference caused by falling ambient. This is why it is important in > calorimetry to keep ambient stable and not to try to use data when ambient > is changing rapidly and with a big temperature change, either up, or down. > > That is the main reason I deleted this data. I plan to put back the first > several hours only, to prevent this kind of confusion. Actually, I hope to > replace the whole spreadsheet with new data when Mizuno improves the > heating and airconditioning to eliminate these ambient fluctuations. > > > >> Why did you choose 1.4 hours? >> > > I did not choose 1.4 hours. I meant to say that the temperature rise from > the pump heat stabilizes at 1.4 hours (as shown in Fig. 19). I also meant > to say that the ambient was fairly stable for the first ~2.4 hours that > day, and we should only be looking at those first 2.4 hours. That is why I > cut off the graph. When ambient starts to fall, the reactor and Dewar lag > behind, and it becomes impossible to see the effect of the pump heat or to > derive the constant for Newton's law of cooling. > > Basically, after 2.4 hours you should either ignore the rest of the data, > or use a much more complex modeling method which takes into account the lag. > >
