I'm trying to make a cut that a ball end mill would do over a given profile, keeping the end mill vertical. I was able to get the path I wanted but I had to sweep a sphere first and then a cylinder separately. I can't figure out how to make this work doing a single sweep. A sample Inventor 2024 file is attached.
Hi
The problem is a poorly defined path. All non-track elements should be set as construction.
Kacper Suchomski
I edited the sketch so everything except the path is construction. The solid sweep does not fail but the orientation of the cutting solid changes, which is not what I'm looking for. Can you share a working example?
You can set a fixed orientation in the dialogbox. Remember to define the path through the center point of the sphere in this case. Otherwise, the path will be followed by the vertex of the sphere (arc), instead of tangency.
Kacper Suchomski
I cannot get a single solid toolbody to behave as I wish. The fixed orientation option fails to create the feature. I recreated both the toolbody and sweep path, neither of which seemed to affect it. The start of the sweep path is at the center of the spherical end of the tool body.
I see harmful lines in the sketch in your screenshots.
Please read my comments carefully.
Kacper Suchomski
Could you elaborate please, your comment is unhelpful.
The sphere of the toolbody is centered at the top of the vertical line.
A must be centered relative to the horizontal line. The vertical line must be construction.
Kacper Suchomski
BTW. Why don't you like the variable orientation?
Kacper Suchomski
@pball, I think I was able to get the expected result by making the vertical 0.250 line in the path a construction like as shown, so that that the selected path shows 3 Curves in the Path selection of the dialog. I think this is what kacper.suchomski was referring to as well.
Hope that helps,
Curtis
PS. The path radius is too small for fixed-position machining. If you increase the radius, it might work.
But from a geometrical point of view, it's more convenient to model with a variable orientation. This is a sphere - it's the same size in every direction and leaves the same trace.
Kacper Suchomski
Yes, you're right, that's what I meant.
I'm not at the computer right now, but that's exactly what I meant in my first comments.
However, I later learned that the user absolutely wants to keep the tool vertical. And that's where the problems start.
If we set a constant tool direction, we need the path to pass through the center of the ball - otherwise the path will not be tangent to the obtained geometry (as was done in the example you showed).
For example, if the path is set to the apex of the ball (with a constant tool position) in the bevel area it will mill deeper than the path - because the determinant will be the apex, and the tool has its owndiameter.
In real machining, preventing this phenomenon is called tool compensation. Most CAM programs have such an algorithm. But here we have no compensation - we have pure geometry - and we have to approach it consciously.
Kacper Suchomski
Thanks @kacper.suchomski !
I read your P.S. note and then had a closer look at what @pball had wrote and realized I was not keeping the tool vertical.
@kacper.suchomski wrote:
If we set a constant tool direction, we need the path to pass through the center of the ball - otherwise the path will not be tangent to the obtained geometry (as was done in the example you showed).
^ I think this is the part I was missing. I'll try and have another look here shortly.
per @kacper.suchomski's advice, when I set the path to be to the center of the tool, and changed the radius to be slightly larger than the tool diameter, then I was able to get the the fixed orientation solution to work.
I raised the path up to be center of the tool, but lowering the tool to be on the center of the path worked also.
( I have seldom used the sweep solid option since it was introduced, so this was all a learning / re-learning experience for me
)
I feel like there are few things to unpack here. I reached the same conclusion as @Curtis_Waguespack earlier with adjusting the position of the toolbody and removing the vertical line from the path. That's what V4 of my file attached below does. This generates the path I want but feels like a work around for something that should work with that vertical line as part of the path.
In the original file I posted with two separate solid sweeps, if you suppress sweep 1 which is the sphere it is obvious that the sweep of the cylinder toolbody correctly follows the path including the vertical line at the start. However if you add the ball end to that cylinder toolbody the sweep behaves differently. I'm beginning to feel that this is a bug. @johnsonshiue would you be able to look at this and give any input?
@kacper.suchomskias shown in one of Curtis's examples in this particular exercise even if the toolbody doesn't remain vertical, the same end result is achieved. My insistence on having the toolbody remain vertical is for broader applications of this where it is important to the resulting geometry, for example the picture above showing a regular end mill only moving in 2 axis.
P.S.
After some more screwing around it seems the radius in the sweep profile is the issue with using the fixed orientation option when the toolbody has the ball end. If the sweep radius is less than the ball end radius it fails. I'm curious if that should be a limitation in the CAD world.
There are several factors that contribute to this:
Kacper Suchomski
It's already over 2,5 hours after the New Year in my house
, and the explanation I tried to describe earlier appeared on my computer.
And to remember not to create self-intersecting surfaces/edges. @Curtis_Waguespack showed in his posts that this can be achieved by increasing the radius within tolerances, for example.
Another consideration is to extend the path in case of constant tool direction to avoid ball stamp at the end of machining. This can be parameterized using trigonometry.
All these issues need to be kept in mind and considered in the design during drafting to generate healthy geometry.
Kacper Suchomski
Can't find what you're looking for? Ask the community or share your knowledge.