What I don't know about TIR is enough to probably keep me from saying anything here but heck, I won't let that stop me. I believe that the refractive index as well as critical angle of the material used will either qualify or disqualify the material as viable as an optic. As I understand it, the outside surface of the the optic is designed with a "pseudo" parabolic response in mind. That is to say that light from the LED will be reflected internally from this surface and exit in a close to parallel path with light hitting other areas of this surface.
In the optics I have looked at, the initial entrance surface to the optic is a straight walled tube. At the point of entrance, the light is initially bent as a function of the refractive index of the material. When the light contacts the outer surface of the optic, we want it to remain in the optic (angle of incidence > critical angle) and we want it to continue, exiting out of the front of the optic (angle of incidence < critical angle) with a final bend due to the index of refraction such that the light's final coarse is parallel to the axis of the optic.
If my understanding of the physics is even close here, I believe that the index of refraction of the material will dictate the shape of entrance, reflection and exit surfaces. Further, I suspect that there is likely only a certain range of refractive indices (sp) that would be viable. The Photon must first enter, then bounce and then exit. Point of entry will dictate point of bounce which then dictates point of exit. The photon's course is bent at points of entrance and exit as a function of the refractive index of the material. The photon will bounce internally with angle of incidence equal to angle of reflection provided that the angle of incidence is greater than the critical angle. If the angle of incidence is less than the critical angle, the photon will pass through the optic without bouncing.
With a great level of uncertainty I think what I have stated above is on track. With a great level of certainty, I state that optics are complicated! /ubbthreads/images/graemlins/grin.gif
- Don
BTW, I have pondered the flat exit surfaces of both the NX05 and the 30 mm. Now to me, flat is a "constant" curve. What happens if we alter that curve either by making it concave or convex? The NX01 which is favored by many has a complex exit surface.