Referencing to help module: See “Specifying Coatings on Surfaces”

OpticStudio interprets the coating definition using one of four rules:

If the surface specified is a boundary going from air to glass, the coating layer order is interpreted exactly as specified in the coating file. The incident media is air, then the outermost layer is listed first (at the top) of the coating definition, then the next layer, etc., with the substrate being the glass type on the surface. The definition of the coating should not include the substrate index or material definition. The term glass here means the glass type is not "MIRROR", and not blank (which is treated as unity index for air). Gradient index lenses are considered glass.

If the surface specified is a boundary between air and air, or glass and glass, the coating is also interpreted exactly as for air to glass, with the appropriate calculations done for the incident and substrate media.

If the surface specified is a boundary going from glass to air, then the order of the layers is automatically reversed, so that the coating is the same as if it had been applied going from air to glass. Therefore, a coating defined by ALHHS, would be interpreted as SHHLA if the boundary goes from glass to air.

If the surface type is a mirror, then the coating definition must include the substrate index. The last layer in the coating definition is then assumed to be a semi-infinite thickness of substrate material.

Thanks for that, but it doesn’t answer either of my questions.

Hi James,

As a quick answer to your questions (someone else could add some additional details):

1. Coatings do not affect the Snell’s Law aspect of ray tracing (both sequential and non-sequential).  Rather, coatings only affect the e-field component of the ray, thus affecting polarization, phase, & intensity.  For visualization purposes, you can think of OpticStudio performing 2 distinct ray traces, the first one that ignores coatings and simply uses Snell’s Law to trace from Object to Image and the second one to go back and calculate the modified e-field at each interface.  Therefore, coatings will not affect lateral or axial shifts for rays in a system.  For non-sequential mode, since the default glue distance is 1e-6 and coating layers are typically 1e-5, you can physically model the dielectric stack up as separate Objects in the NCE, but this would be extremely difficult to keep accurate if you’re changing/optimizing your model.
2. Since coatings do not have Thermal Coefficient data, they cannot model dispersion changes (wavelength drift) at different temperatures.  Only materials defined in the Materials Catalog with the D0, D1, D2, E0, E1, Ltk values can model wavelength dependency at different temperatures.  Since thermal modeling is done in the Multi-Config Editor already at discrete TEMP and PRES values, you can simply define a new coating at each TEMP and use the COTN operand to change the coating on a surface.

Makes sense in both cases, Michael.  Much obliged!