Hello,

I am simulating a multimode grin fiber (100/125) spliced to the SMF-28 fiber. The grin fiber will function as a collimator. I used non-sequential cylinder volume as grin fiber, selected the  grin9.dll and set the index (shown in the attached). But I think the  layout is not correct. If I increase the radius of cylinder volume, at this time it is like a grin lens, i find layout is more accurate. So I wonder why the grin material is not working when I use small size (fiber size) and if this not working, how could I simulate grin fiber and have collimated ray?

Thanks