Michael,

Thank you for your rapid reply. See attached. I found that fragments are longest when viewing the planar face along normal (though difficult to see at this angle) and shortest when viewed at grazing incidence to the planar face.

I'll try trim on Monday.