Why will this not finish the sides?

[bmarkeyJ44PH]

Why will no process I try actually finish the sides, the flat vertical sides?   I've tried Z level and parallel. I've tried limiting it to 90 degrees only.  Every time it just does something completely weird and ignores the sides.   What am I missing here? 

( SU's 1-3 are just to get the part to the condition it was given to me. )

Thank you

[Muppett1]

I see it finishing the side that you have indicated with the geometry line.  What is it that you want it to do other than that?  Are you trying to zlev fin the model face on the inside?

[Muppett1]

I was able to quickly get this far by unchecking all the check surfs, and checking "don't roll over the edges at all" tab, but did have to go back and check one check surf to keep from rolling over one surface.  Hope this helps.

[Muppett1]

And then by unchecking the slope constraints was able to get this outcome.

[bmarkeyJ44PH]

I'm not really following here, sorry.   If you remove all the check surfaces it just demolishes the model, at least from what I'm seeing.  My general practice usually is, all surfaces are check surfaces if they're not a surface being cut.

The initial Z level rough is great.   Its face_152 and face_138 that are causing me issues. 

The side2 was just a test.  It wouldn't work anyway as the bottom is not flat.

I'll just have to keep playing with it until I get something acceptable.    Thank you for the help.

[Muppett1]

Please see in message 3 the directive of selecting "don't roll over edges at all" under the "edges tab".  I used this selection in conjunction with selecting 1 check surf to achieve the side milling you see in the last picture.  I understand it will not work b/c of the internal sharp corners, but this does allow toolpath to be applied to each wall surface.

[bmarkeyJ44PH]

I understand what you're saying but if I remove all check surfaces, select don't roll over edges, it mangles the part.

See sm2, that is the zlev finish with a single check surface, and don't roll over. (pic 1 is no check surfaces)   While it finished the bottom very nicely, the two vertical walls are not finished. It doesn't go out all the way to the edge as it should.  It stops like its using the bottom as a boundry.

If a face is selected as a part surface, then why doesn't it cut it?   Nothing gets cut if I only choose those two faces and not the bottom. 

At this point, all I'm trying to accomplish is to finish those two sides down to the curved bottom, which should take all of about 30 seconds. 

Again, sorry, this is so frustrating, trying to use SM'ing to do something and being unable to get there.

[Muppett1]

I think I had the same result as you show in SM1 when I checked the "don't roll over edges at all" tab, then I had to select that one surface where the damage to the model shows as a check surface to get, what looked to be conformed toolpath?

[Muppett1]

This is another way I've tackled these type configurations as well, as I have machined a part that is very similar to what you are showing.  I selected the entire model to process TP on, and then created a containment boundary to contain the TP within it's constraints.  This is what it looked like when I finished.  I turned off the slope restrictions, selected cut to full limits (not using the "don't roll over edges at all" button), and created the containment as the exact width of the slot to keep the center of the tool from going past the edge of the part at that point.  Hope this helps.

[bmarkeyJ44PH]

So I think we're going to redesign the part because nothing seems to work via surface milling.

With the previous suggestions it either tried to dive into the edges as it began each pass, or had a bunch of weird vertical retractions.

The closest I've gotten is an isoline finishing that does almost what we need it to, yet again it ignores the vertical walls.  If it would simply do one more pass and finish the walls, it would be perfect.  Whatever combination of black magic and hoodoo gets me there I can't figure.  You can see in the attached picture that its running parallel with x and following the contour of the bottom which is exactly what I'm looking for, but in the second picture you can see the remaining wall stock near the bottom.  That small bit of maroon between the vertical sides and the bottom.  ( The rest was taken off with a side feature since that just worked.)   This was done with the "select the entire model and limit by a curve" method, since that seems to work better than selecting the actual surfaces. 

Does surface milling only want to deal with the tip of whatever tool is being used, or does it take into account side milling also?   I'm wondering if I'm expecting an outcome that can't happen.   It seems like there should be an easier way to say mill these vertical walls but follow the bottom contour.

[Muppett1]

I would normally ask our engr dept if those sharp internal corners could have fillets, if nothing else for strength, but also for ease of machining.  I might also try creating untrimmed surfaces from those wide walls and try to machine them?  Other than these two thoughts I don't have much else.

[bmarkeyJ44PH]

Agreed with the sharp corners, that was never going to happen easily without another moveable axis.

I'll have to post my results later, I spent some time this weekend with that isoline.  After some hand editing I think I can make it work.   It seems it all just comes down to spending enough time with sm'ing in general to see what every setting does and how it affects the outcome.