Keep in mind the possibility that the value of n depends on the wavelength, and therefore presumably on the final calculated temperature, and so an iterative procedure may be needed. In other words, the comparison will not be between 2 emissivities for the same n, but for different n's, and the company literature does not give the method of determining n.
And of course, none of this takes account of surfaces that are not grey bodies. The obvious solution would have been to use thermocouples in the December run as well, but they didn't. On Mon, May 27, 2013 at 3:22 PM, Alan Fletcher <a...@well.com> wrote: > I'm putting the Optris calculations into a spreadsheet -- the following is > documentation of the formulae used in readable form > > From the Optris "IR Basics" documentation (Page 7) > > > From the actual object temperature (To) and ambient (Ta) > > To Actual temperature e > emmisivity > Ta Actual ambient C a > constant in the calorimeter > Tp Temperature of pyrometer > n exponent -- depends on wavelenght > > U = C *( e*To^n + (1- e)*Ta^n - Tp^n) > > Measured temperature reported by the calorimeter > > Tm Measured temperature > Note : the optris equation uses the same > symbol for Tm and To --- > so it seems to be self-referential > > Tm = root(n,(U - C*Ta^n + C*e*Ta^n + C*Tp^n)/ C*e) > Note that a lot of the "e" cancel out, > leaving 1/e terms > root(n,val) can be computed as > power(val,1/n) -- some languages have problems with this > > Pm Total power calculated from Tm > > Pm = a * e * ( Tm^4 - Ta^4) > > I'll put these equations into the spreadsheet and see what happens for > various "n" (wavelength) and "e" emmisivity > >
