Hi José,

Thanks for posting your question on the Community forums!

Regarding a grid-based NURBS surface, I do not believe we have a native surface which takes in a non-uniform spread of points other than those you’ve listed (Radial/Toroidal NURBS). The closest surface we have to a similar functionality is the TrueFreeForm surface, which is a superposition of different surface types (such as Standard, Even Asphere, Zernike Standard Sag), in addition to supporting a grid-based deformation of the surface.

This grid-based deformation is optimizable, as different points can act as variables. So, you can get localized surface optimization while leaving the rest of the surface unaffected. It does rely on this grid-like structure, though, so you’re not really having a non-uniform sampling of points. You can find more information on this surface in the article Using the TrueFreeForm surface​ for grid based freeform optimization. Note that this surface type is only available to Premium edition Subscription licenses.

Regarding your crashing with the Radial NURBS surface, this is certainly unexpected. I attempted to optimize a lens system with the Radial NURBS surface and did not experience any crashing, though it may be a case where you have a certain setup or complex model giving you issues. Do you begin the optimization with a starting point of some kind (some initial definition of the sag)? Or do you begin with a planar NURBS surface and let OpticStudio find a path forward for the sag (that is, all Zn terms begin as zero)?

It might help if we could take a look at the file itself, though I understand if it needs to be modified for any IP reasons. You can also feel free to submit a case to us through your Zemax Account on Zemax.com.

Let us know how these thoughts work out for you!

Is the TrueFreeform surface the only option for designing a surface which has no symmetry whatsoever and incorporates an optimizable grid to help define surface points?

Hi @Chris.Johnston,

Thanks for your additional question here on the Zemax Community!

Yes -- TrueFreeForm is the only surface which supports optimization of a grid of surface points as opposed to varying parameters of a surface’s sag equation. If you do not have access to the TrueFreeForm surface, then you would need to leverage other surface types such as the XY Polynomial, Chebyshev Polynomial, etc., to design a freeform surface. While you wouldn’t be able to optimize a grid of sag points directly, you could try to leverage merit function operands like those that pull surface data along the surface (SSAG, SCUR, etc.) to aim for some particular behavior of the surface shape at distinct positions on the surface.

Just as a quick aside (in case it helps), we also recently launched a functionality called the Composite Surface -- Introduction to Composite Surface – Knowledgebase (zemax.com). This new feature allows for “stacking” different surface shapes and having a cumulative sag profile based on those stacked surfaces. While it’s not the same as a grid-based optimization, Composite Surface can still allow for pretty arbitrary surface shapes, as it supports a larger range of surfaces as compared to the TrueFreeForm. It just requires the user have a legacy Professional or Premium Subscription license type or one of the newer Ansys Zemax OS licenses.

Let us know how these thoughts work for you or if you have any more questions!

Angel, appreciate the quick response and suggestions!