Hi all- As I recall, FilletEdge was cleaned to make revolves where possible but evidently FilletSrf not - I will add this to the pile. RH-75374 FilletSrf: Simplify to revolves
Above example is taken from this years fillet exercise I asked my students to do.
Same course / test as last year - old exercise see here.
if the surface is build as revolve the intersection works nice: green curve until (x).
if the surface is build as approximation by _filletSrf (what many students did) the intersection and trimming does not work in some cases: red curve not hitting (x) but an edge.
A well implemented test for a revolve / torus will give much better results for the next / dependent modelling steps an handle tolerance in a much more “economic” way. (my guess…)
hope this explains the background of my wish - kind regards - tom
thanks for some action in the field of fillets - and hopefully you can bring some early results maybe in some testcommand… “_testFilletSrfV9” “_testFilletEdgeV9”