Share Email Print

Proceedings Paper

Incorporating skew into RMS surface roughness probability distributions
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Characterizing surface roughness is important for predicting optical performance. Typically, this is accomplished by taking multiple statistically independent measurements and averaging. But, this approach assumes that the statistical distribution of the roughness has a Gaussian (normal) probability distribution. Our analysis shows that this assumption is wrong. Real data acquired from two different sets of telescope optics indicates that roughness of highly polished surfaces is skewed and is best described by a largest extreme value probability (LEV) distribution. Assuming a normal distribution and simply averaging overestimates the most probable surface roughness and could result in the expenditure of unnecessary polishing effort.

Paper Details

Date Published: 7 September 2013
PDF: 7 pages
Proc. SPIE 8838, Optical Manufacturing and Testing X, 883814 (7 September 2013); doi: 10.1117/12.2024883
Show Author Affiliations
Mark T. Stahl, NASA Marshall Space Flight Ctr. (United States)
H. Philip Stahl, NASA Marshall Space Flight Ctr. (United States)

Published in SPIE Proceedings Vol. 8838:
Optical Manufacturing and Testing X
Oliver W. Fähnle; Ray Williamson; Dae Wook Kim, Editor(s)

© SPIE. Terms of Use
Back to Top