How to model 4th axis milling cut

[DakotaLarry]

I would like some step-by-step help on how to model a cut to be done on a 4th axis mill. The work piece is cylindrical. I want to bring an end-mill into the cylinder from one end aligned perpendicular with the axis of the cylinder. As theworkpiece travels 1/4" forward, it will rotate 1/4 turn (90 degrees). I imagine I should be using sweep but when I get into the "menu" I get lost. If someone could please take the time toexplain step by step I would very much appreciate it.

 

[jeff4136]

2008-04-19_120853_helical_end_mill_cut--wf2--.zip


It's easier to do than to explain, I think.
Not exactly what you were describing but should give you the idea.
If you can't read WF2 or if you have trouble putting the pieces together, holler back.

 

[SRINIVASANIYER1]

Jeff,


I think the cut will be as shown in the pic with the demarkation of flat surface being along the red line.


It would do well for dakotalarry to define a slot cam mechanism and check for interferance and the desired motion profile.

 

[jeff4136]

That may be true, Srini.


I chewed on that a little while and decided that the planar cutter end
would always be tangent to groove bottom in the plane defined by cutter
axis and work piece axis without regard to the ratio of axial and
rotational displacement. That may have been a mistake. (There are two
very small areas on the groove that caused me to doubt the idea but
time prohibited further examination.)


I have, in the past and without real comprehension or resolution,
considered the surface(s) that would be created by starting with only
rotation and ending with only axial displacement (or vise versa) with
transitions of varying combinations of displacement component. The groove
bottom will (presumably) be cylindrical where there is no axial
displacement and planar where there is only axial displacement. It seems
reasonable to assume that the ratio will affect the surface making it
something between cylindrical and planar. I can't reconcile that with the
idea expressed in paragraph above (or can I?) and sweeping a line normal
to a center to center trajectory at groove depth may be the correct
answer.


I thought I'd tried that but may be confusing the results with using a
solid Thicken / Cut which I once thought correct (satisfies the conditions
just above) but trying to put the planar faces at the groove ends
contradicted. Or, I may have discounted the idea because I assumed it
wouldn't satisfy the (possibly incorrect) tangent in perpendicular axis
plane assumption.


I'm open to ideas / explanations / examples or will just play with the
shapes some more when time permits. I've given up on the Great Mysteries
of life but this is one of the little things I'd like to thoroughly
understand before I go completely senile. ;^)

 

[wsylvester]

one option,


create curves by equation.. one a lil' more than your cylinder radius, the other to the depth you desire


boundry blend the 2 curves.. then thicken


points at the end of your curves,to create a perp. plane to make a circle extrudes<cut> at the ends..


it does pass the mechanism interference test


and you can get all fancy and define prams to contol the equation so you can adjust from outside the equation

 

[SRINIVASANIYER1]

I've given up on the Great Mysteries of life but this is one of the little things I'd like to thoroughly understand before I go completely senile. ;^)

had a hearty laugh after a long time. This topic interested me because I had lost a few bunches of hair while working on a Globoidal cam groove.


For the above problem, I used the mechanism approach, created a motion envelope (quality level 10, file size 50 MB) and then performed a boolean operation on the cylinder. The end mill does follow the groove correctly, but then even at a quality level 10, the quality of cut is very bad that it gives interference all along the path. However, the path is similar to what I have described above.


Imagine an end mill plunged into the cylinder. The axis of th end mill is constrained to remain normal to the surface. After taking the plunge, it will move along the path. maintaining normalcy to the path. Will try to post a file a little later.

 

[SRINIVASANIYER1]

As for creating the curve, if the curve is a true helix,perhaps a wrap command would also work.

 

[wsylvester]

jeff I think that you may be the one who suggested this method...but maybe I'm the one going senile

 

[SRINIVASANIYER1]

Attaching a few Interesting pics using the mechanism approach (small surfaces) to give an idea of how the surface would be and its deviation from our perception of a helical cut.

 

[jeff4136]

I logged on to post and thought I was going to have to look back for the topic,
Srini.


I ~think~ if someone wanted to be picky about the groove bottom geometry it
would look something like the attached (layer Features-Grp_02).
2008-05-01_072442_helical_grv--wf2--.zip
I'm stumped trying to duplicate the imported surfaces in Pro/E. They are arcs
swept along the edges of Var Sect Sweep 8. The section plane remains tangent to
a cylindrical face. Ideas anyone? When we get this'n whipped we can start on
a "J" profile. ;^)

 

[wsylvester]

jeff... it was a j-profile... it's me going senile.. no doubt..





Srini.. on your patches coming thru.. what's your model accurcy?


I changed the default to .0001 and .. I get some feathering thru .. but worse interfence is...


(0.000213368 INCH^3)

 

[SRINIVASANIYER1]

Model accuracy was 0.00012 which I later changed to 0.0001. Interference 0.000213368 Inch^3... let me convert to mm^3I am a metric man

. Oh that is 3.496475071552^3. quite big.


In my view the profile will not follow the theoretical cylinder because the leading edge at the start will become the trailing edge at the end with a circular cutter.


If you use a Parting tool kind of cutter (akin square thread cutting on a lathe) perhaps will follow the theoretical cylinder, the cutting edge being parallel to the axis of rotation.


I am waiting for some more inputs from other users...

 

[jeff4136]

Well, I dunno. It needs some techique-al clean up (it's hanging together
by the skin of its teeth) and I can NOT model it entirely in Pro/E but I
think it's the real deal. Maybe someday a 'real people' will come along
and help us out. ;^) If anyone can spot something that looks amiss don't
be shy.
2008-05-09_071029_j_groove--wf2--.zip



10-22-2008 - A version that's had the geometry cleaned up a bit ...


2008-10-22_231547_200805101059.prt.zip
Edited by: jeff4136