How can accurate Flute Geometry be generated in Alibre ?
- Thread starter barryhc
- Start date
If the "runout" is a simple reflection of the shape of the grinding wheel (it's not withdrawn under load) then surely one could position its profile relative to the final axis position of the wheel and Revolve Cut.
However, this still leave the problem that a "2D" profile swept through "3D" space cannot simulate, directly, the action of a 3D solid (grinding wheel) doing the same thing in the physical world.
Are either/both of these statements comprehensible/ correct?
Re:
>This is quite true, and is my first theoretical solution. In some instances ( especially when helix angle is close to 90 deg. ), this approach will actually solve the problem. Not all flutes however are cut with a grinding wheel. Many of them, in the case of indexable tools, are cut using regular, corner rounding, or ball nose endmills.
Let's avoid any "withdrawal" considerations for now. We might get somewhere faster as a result.
Quite true Miles. I think some progress could be made at least for the grinding wheel type ( And possibly the end mills as well ) if some very careful work was done to prepare the "shape" ahead of time, and then either a "modified normalcy option" other than "rigid", or "strictly normal", was applied to the "sweep or whatever".
Does that last statement make any sense to you ?
Thanks > Barry
MilesH said:
>This is quite true, and is my first theoretical solution. In some instances ( especially when helix angle is close to 90 deg. ), this approach will actually solve the problem. Not all flutes however are cut with a grinding wheel. Many of them, in the case of indexable tools, are cut using regular, corner rounding, or ball nose endmills.
Let's avoid any "withdrawal" considerations for now. We might get somewhere faster as a result.
Quite true Miles. I think some progress could be made at least for the grinding wheel type ( And possibly the end mills as well ) if some very careful work was done to prepare the "shape" ahead of time, and then either a "modified normalcy option" other than "rigid", or "strictly normal", was applied to the "sweep or whatever".
Does that last statement make any sense to you ?
Thanks > Barry
Re:
I understand what you're proposing but, honestly, I haven't a clue whether one could achieve anything along these lines, or not. I suspect there's no substitiute for working back from the math of sweeping a solid - in which case....
Are you aiming to use the model to actually control the cutting, or what?
barryhc said:
I understand what you're proposing but, honestly, I haven't a clue whether one could achieve anything along these lines, or not. I suspect there's no substitiute for working back from the math of sweeping a solid - in which case....
Are you aiming to use the model to actually control the cutting, or what?
swertel
Alibre Super User
Although a flute is a good example for a real world problem, perhaps we can restart this discussion with simpler geometry - a slot cut by a ball end mill.
The slot will have a edge radius at the bottom surface and a spherical radius at the long ends. How do I cut a slot with a radius on the long ends, radius on the bottom edges, and the blended spherical radii of the two combined?
Technique:
I cut a rectangular slot into the solid.
I use a fillet feature on the two long edges of the bottom surface.
I then do a revolve cut on the end face of the slot, the 2D profile matches the cross-section of the end mill, rotated about the top edge.
I mirror this feature symmetric about the midpoint of the slot.
4 Steps and 4 Features to make one solid feature.
Of course, taking a ball-end mill shape cutting tool and protruding along the length of my slot would be a nice feature, instead. But how close does that come to having to code a CAM package?
Also, to save myself the future work, I could create the slot as a feature library and just insert it as a grouped feature in another model.
Well, the point was to definitely clarify the geometry, and that has been done. I understand it would be a feature enhancement at the very least if it were to exist.
I haven't had the to time to "dig-in-hard" on this one yet, as I'm already drowning in another "one", but I may play with it soon. Still, I'm a neophyte with Parametric modeling so "other's opinions" are like gold to me.
I think some progress might still be made with a very thoroughly ( even painfully ) developed "Shape", even if it has to be used in a "normal" or "rigid" fashion.
The "in which case" might take "forever" at my age, so I still look forward to some sort of usable solution, even if it becomes complex.
And No, the information is not so much for driving the tool in a machine, as it is to simply design the tools in the first place. Eventual stress analysis, actual "space" used, etc..
> Barry
The interesting thing is that CAM packages already do this kind of work without difficulty ( at least Gibbs does ), but you can't currently get CAD data out of the CAM package.
I'm certainly not trying to create CAM functionality in the CAD Package. I still have to design the tools first however, and their geometry is not trivial.
For me, this is an R & D project for tools that don't exist, not a drafting session for copies of tools that already exist.
Thanks very much for the input. > Barry
Edit: This is not a reply to your last post Miles, I'm still working on that.
According to what was said in a discussion we had earlier, about drills, the specification of the profile of the flute is defined normal to the axis of the helix, not normal to the helical path. There is a way to project from one to the other, of course. Don't know if this helps... This is all foreign territory, for me..... still, it keeps the brain ticking over....
A
Anonymous
Guest
I'm not sure I understand exactly what this discussion is about but I have posted a JPEG in the binaries called twist drill. Is this what you trying to do ?
Yeah, it is a bit strange....
Anyway, partly the discussion is about the difference between the geometry obtained by sweeping a planar profile through '3D' space and that obtained by sweeping a solid through '3D' space. The latter being what you're inevitably stuck with in the real world... Does this make sense?
Nice drill model. Did you manage to model the relief behind the cutting edge?
A
Anonymous
Guest
Sorta makes sense. The drill model is not mine. It is an Inventor iPart that is available on the internet here
http://www.cbliss.com/inventor/Parts/Tools/index.htm
Re:
Indeed..
Yes, I saw your question, and the response this week.
I can't see how you could do it in Rhino, in the way that you require (working back from a given cutting/grinding tool form)........... Post it as a naive question in the Rhino newsgroup, maybe? :wink:
moyesboy said:
Indeed..
Yes, I saw your question, and the response this week.
I can't see how you could do it in Rhino, in the way that you require (working back from a given cutting/grinding tool form)........... Post it as a naive question in the Rhino newsgroup, maybe? :wink: