> From: "Alan Fletcher" <[email protected]> > Hmmm ... they calculate the convection term as 0-900W, so they just > ignore it.(which gives a lower COP) > > I'm not sure if the convection would affect the wall temperature from > bottom to top. Probably not by much. Re-re-re-corrected. The DO use convection in the COP.
So the http://kisi.deu.edu.tr/aytunc.erek/Proje2011/konu7.pdf paper in particular is relevant. a) They do the experiment with two cylinders -- 4.8mm and 9mm -- much smaller than Rossi's 9cm b) They only vary the temperature from 10 to 50C above ambient c) The experimental Nusselt numbers fall in a band of -20% to 0% of simulated numbers and -20% to +20% of theoretical (I'm not sure how a 20% error in the Nusselt number would affect the calculated results) d) The temperature at the top was greater than at the bottom -- maximum difference 1.6C The difference in size and temperature between this and Rossi is so great that I'm not sure how it applies. At very least, I'd de-rate the Rossi convection number with a 20% change in the Nusselt number. Even so, if you discard convection entirely you still get a COP of 2.08 I think it would be very difficult to explain away ALL of the calculated COP.
