On Mon, May 27, 2013 at 4:58 PM, Alan Fletcher <a...@well.com> wrote:
> > From: "Joshua Cude" <joshua.c...@gmail.com> > > Sent: Monday, May 27, 2013 2:41:34 PM > > > And just in case you're wondering how e effects the calculated power > > > > P = a . e . (T1^4 - T0^4) -- T1 actual, T0 ambient > > > > a e Tc Tk P > > area 18 1.00E-10 0.8 564.1 837.1 38.84 <=== lower "e" OVER-estimates the > power > > area 19 1.00E-10 1 496.6 769.6 34.52 > > area 20 1.00E-10 0.95 511.7 784.7 35.49 > > > You're right. I did that calculation too. But the reason they're not > > equal is because they use an effective exponent not equal to 4 when > > they calculate temperature. It's not clear what that effective > > exponent would be if the emissivity were set to 0.2, and so we don't > > know what the effect would be there. And in particular, we don't > > know what the effect would be if the emissivity depended on > > wavelength. The literature warns about poor accuracy in such cases. > > But it's NOT metal : it's metal-ceramic-paint. > AND the "blank" test was in the same temperature range as the "live" test. > > They checked it with a) DOTS of known emissivity and b) A thermocouple -- > giving results in reasonable agreement with the calorimeter. > > I'm talking about the December test, when a different paint was used. I don't think we know anything about the emissivity of that paint, nor it's dependence on wavelength. In the March test, the power estimate was better, though far from good, but the input was dodgier.
