I am attempting to model an x-ray beamline in a Wolter Type III configuration where the beam strikes the mirrors at grazing incidence angles. The mirrors I’m modeling have curvature in one direction, but are flat in the other, so I believe I need to use biconic surfaces. But I am only able to get rays to trace through my system if I set the second mirror to be an ‘alternate even’ (or ‘alternate odd’) surface. If I change it to a biconic surface, then the rays no longer trace through the system. According to the single ray trace, when the rays reach the second biconic surface, they ‘miss’, but I believe this is only because Zemax is choosing the wrong side of the second mirror. Is there a way to force Zemax to choose the “alternate solution” (as it’s called in the documentation) for a biconic surface? Or is there some other way to get rays to trace through such a sysem?