There are a few reasons as to why a geometry error might be issued when working in purely Non-Sequential ZEMAX.  Sometimes an improper placement of a source, or the mis-use of the “Inside Of” flag will generate geometry errors.  To demonstrate such possibilities, the following sample file will be used, which is available for download on the very last page (Summary and References) of this article.

FILE:  "Revised LED_model_Geometry Error.ZMX"

Note:  This file is a revision of the “led_model.zmx” file located in the ZEMAX > Samples > Non-Sequential > Sources directory.  A few changes were made to demonstrate the errors associated with the misplacement of the NSC sources.

Once you have downloaded the file, open the file from within ZEMAX.  This file uses some of the purely Non-Sequential capabilities in ZEMAX to model two LEDs illuminating onto a single detector.  Note that upon opening the file, ZEMAX automatically performs a NS ray trace.

Currently, the Ignore Errors option is turned ON to prevent the geometry error dialog from continually being issued.  Even from the Shaded Model plot, it appears as if the flux on the detector plane is strikingly too circular, almost as if the rays from the topmost LED are not being accounted for.  Also take note that the Total Power reported on the Detector Viewer is roughly 5mW, which is one quarter of the combined total power from the two sources.

To see how much power is actually lost due to errors, let’s re-run the Ray Trace/Detector Control (which performs a Monte Carlo Ray Trace).  To open the Ray Trace/Detector Control dialog, select “Analysis > Detectors > Ray Trace/Detector Control” from the main menu, or click the “Dcl” button on the button bar.  Select “Clear Detectors ->” and check the following options in the dialog.

Click on “Trace” and let the ray trace complete.  Note that your results may be slightly different due to the random generation of rays!  Nevertheless, you should obtain similar results.

Once the ray trace is complete, ZEMAX will report the lost energy due to thresholds as well as energy lost due to errors.  The lost energy due to errors is roughly 10mW, which is extremely significant since it serves as roughly half of the combined power from the two sources.  It is critical that these errors be addressed immediately so that accurate results may be obtained.