ETGT is a good start.  Make it large enough to prevent negative ET all along the tilted lens.

Another way is to use NORD, the normal distance to the next surface.  Use several (X,Y) values along where edge interference can occur, and that will keep it from intersecting the next surface.

@Mike.Jones , 

Thanks for the reply. I have tried both of your suggestions without success. The operands show the nominal air gap, but don’t seem to update when the first lens is tilted. I must be missing something. Any further suggestions will be greatly appreciated.

Update:

If I remove the surface curvature, the NORD operand look to work as expected, but not the ETVA/ETGT operands…

Maybe this is a clue to finding a solution?

@John.Hygelund:

One option is to just employ simple geometry.  It takes a few MF operands to do the calculation, which makes it slightly tedious but otherwise straightforward. For a rotation by theta, the trailing air gap at the edge of the lens changes by dZ = R tan(theta), where R is the mechanical semi-diameter and theta is assumed to be a reasonably small angle.

Here is an example before rotation,

and then after a 5-degree rotation,

Regards,

Jeff

Thank you @Jeff.Wilde,

Your example is much appreciated. My naive assumption was there would be a built in merit function that would account for edge thickness with coordinate break rotations… But I can certainly use your approach from my task. Thanks again.