Glass--gold mix--water as a function of mix to see the theoretical range and sensitivy. A smart way to test sensitivity might be to actually run one sim for bare gold, say vfrac from 0-.2, then run another sim of same increments from 0.1-2.1, that way the data sets can be subtracted to test the epsilon factor.
Could also consider doing bare, semi-coated, coated for the same vfrac increment range, and use the difference in these curves as a way of testing sensitivy. Careful, as composite equivalence method will be an issue.
Really ought to do
Glass--gold mix--water as a function of mix to see the theoretical range and sensitivy. A smart way to test sensitivity might be to actually run one sim for bare gold, say vfrac from 0-.2, then run another sim of same increments from 0.1-2.1, that way the data sets can be subtracted to test the epsilon factor.
Could also consider doing bare, semi-coated, coated for the same vfrac increment range, and use the difference in these curves as a way of testing sensitivy. Careful, as composite equivalence method will be an issue.