What's it about?
Optical designers assume a mathematically derived statistical distribution of the relevant design parameters for their Monte Carlo tolerancing simulation. Presented are measured distributions using lens manufacturing data to better inform the decision-making process.
Generating optical tolerances in lens design software such as Code V or Zemax begins by modeling a perturbed optical system to capture the effect of imultaneous changes in lens radius, center thickness, and other critical parameters in amounts matching the lens manufacturer’s stated capabilities. The analyst would assume a statistical distribution of the relevant parameters for the Monte Carlo simulation. In some cases, the assumed distributions do not resemble the distributions that occur in lens manufacturing. For example, optical designers assume a center thickness tolerance that is symmetric about the nominal value, but lens manufacturing brings parts to tolerance, not nominal. The best option would use actual manufacturing data to build a statistical distribution for the Monte Carlo simulation. Understanding true distributions will have other beneficial consequences. The designer or manufacturer could easily predict the cost of changing a particular tolerance (as a result of lost yield) for a given manufacturing process...