roughness question

Reply

replied on Thu, Jan 7 2010 8:57 PM

The RMS roughness has an statistical value that the roughness average
or surface area data don't have. The RMS roughness is nothing but the
standard deviation of the height values contained in your AFM image.
So it is a well-known statistical quantity that is used everywhere.
Ideally, the RMS value tells you that around a 60% of the points of
the surface have a height between average-roughness/2 and average
+roughness/2, so it gives you a direct information on how smooth is
your surface. The advantage with respect to the max-min height
difference is that the influence of stray values or outlier points is
minimized (one point in a million has a very small effect). However,
there are some caveats if you care about the nitty-gritty statistical
details (for instance, the RMS value is representative only when your
height distribution function can be fitted to a gaussian. From a
practical point of view, if your surface does not have huge steps or
artifacts you don't have to worry). Roughness Average and surface area
data are more ambiguous concepts and sometimes implementation-
dependent. Surface area data gives you an idea of how convoluted is
your surface, but even though I have carried out some hardcore
statistical analysis of AFM images I have never found it of much use.
Maybe others can give you more details.