Not to much of rounding on the groove and it will be fine. As long as the ring can squish.
If the slit is thin ([21.2*flow rate of system / slit width (the smallest one)] is more than 100, try and get a good sf near the exit as it will improve perfromance slightly but a relatively poor finish on the end of the channels is fine. What ever you find that works best.
I would highly recommend making a reducer insert that changes the nozzle diameter slowly if possible as it will save on head loss alot. Admitedly the geometary is awful but it will help. Getting a bar then milling an angle in it. Need to punch some numbers to get what a long enough section would be but any insert would help. Picking numbers out of the air, maybe 10 to 20% head loss recovery. Its a similar thing with the exit sections a nice curve section on the end of the mill (as much as possible) will save some head loss, but its probably too late now.
I could always run it through cfd with dxf file if you want. I figure its relatively cfd able.
|