Producing Concave Surfaces With A Vertical Mill

[randyc]

I posted a method of producing small ball configurations on a vertical mill in a previous thread regarding various setups for a rotary table. This method might be useful to a few people (telescope guys, maybe). I'm not sure that this is the correct place to post this but I couldn't think of a more appropriate one...

This is an experiment that I tried a few years ago following another interesting experiment to learn how to produce large concave surfaces on a lathe or a vertical mill.

Tilting the head of the mill (or the rotary table) to a specific angle and adjusting the “stickout” of a flycutter, one can produce a fairly precise concave surface. Errors in the radius caused by runout in the mill spindle or rotab bearing are averaged out with multiple passes.



The flycutter is SLOWLY advanced into the work with minimal DOC and the rotary table is then rotated through a full turn(s) until the full depth is nearly reached. Final cut should be multiple revolutions to average errors.



It is an exercise in descriptive geometry to determine the starting parameters. One of these days I’ll make a spreadsheet to determine them. I suspect that making a CAD layout would be lots faster

PS I'm too impulsive about spelling and construction otherwise I wouldn't have to edit !

 

[Billh50]

There is a formula for cutting a large radius such as 32" in a 3" wide part with a 6" cutter as an example. The formula tells you how much to tip the head of the mill. I would assume if you used that formula and put your cutter at it's lowest point on center of a round piece it would create a concave the same way when turning the part in a rotary table.
The formula is this. Divide 1/2 radius of the cutter by the radius to cut. The answer will be the Tangent of the angle to tip the head of the mill.

What you are really cutting is the bottom of a parabola.

 

[Billh50]

example of what will be cut.

 

[randyc]

We'll have to agree to disagree. The flycutter, rotating in a circle cannot produce a parabola as the work is also rotated in a circle - this can only generate a spherical surface when set up as shown.

The concave surface is a generated one just as the ball produced on a vertical mill was generated (mentioned above, referencing the post). In fact, except for the cutting tool orientation, the two processes are identical.

PS: I neglected to mention that your sketch shows the spindle axis and the rotary table (and the work) axis as being coincident. That's not the case, see the first photo.

 

[chips&more]

I use the mill and rotary head all the time to turn round balls on the mill. Can’t remember the last time I did it on the lathe? You must have the spindle datum center in line with the rotary head datum center or you will not produce a round ball. Rather, it will look like Stewie’s head! IMHO the concave machining example in this thread will produce a concave segment. And if the example always has a perfect center with no tit, then it will be a true arc of a circle…Nice Job, Dave.

 

[randyc]

...You must have the spindle datum center in line with the rotary head datum center or you will not produce a round ball. Rather, it will look like Stewie’s head! ...

Hi Dave, I'm probably misinterpreting your statement but to produce a ball in a vertical mill the spindle axis and rotary table axis can't be coincident. Consider the following sketch, which I posted in the "Rotary Table Tricks and Tips":


As you can see, the angle between the rotary table axis and the spindle is 105 degrees (15 degrees off vertical) in this case. This isn't particularly critical, the angle is mainly determined by the desired dimension of the "neck". In the following photo, I used 15 degrees offset:



If the axis of rotary table and mill spindle were coincident, then the vertical mill would be a vertical lathe unless I am not understanding your statement.

Cheers and thanks for your comments !

 

[chips&more]

I'm probably misinterpreting

You misinterpreted, but in fact your drawings show the two datums in line as I explained! I did not say in line with the arc of the boring head, I said in line with the datum center of the boring head/spindle. Sorry, typing and explaining are not my first love...Dave.

 

[Billh50]

I was talking about the concave part not the ball. It will appear as a sphere because you are only seeing the lower part of the parabola. But unless the cutter is at 90 degree to the part you are cutting a rotating parabola. If you take a flat round disc. Hold it in front of your eyes and slowly tilt it away from you. you will see the parabola start to appear at edges until the disc is 90 degrees when it becomes a circle again. But tipping the head or part the radius will become smaller and smaller causing more of a dish. That is the parabola effect.

 

[randyc]

I was talking about the concave part not the ball. It will appear as a sphere because you are only seeing the lower part of the parabola. But unless the cutter is at 90 degree to the part you are cutting a rotating parabola. If you take a flat round disc. Hold it in front of your eyes and slowly tilt it away from you. you will see the parabola start to appear at edges until the disc is 90 degrees when it becomes a circle again. But tipping the head or part the radius will become smaller and smaller causing more of a dish. That is the parabola effect.

I see your point about rotating the disc but it would be elliptical, not parabolic. I definitely agree that the cutter axis should be at 90 degrees and the correction is appreciated !

 

[randyc]

You misinterpreted, but in fact your drawings show the two datums in line as I explained! I did not say in line with the arc of the boring head, I said in line with the datum center of the boring head/spindle. Sorry, typing and explaining are not my first love...Dave.

Dave, my memory is not so good anymore. I thought that you posted a sketch and the axis of the sketch was labeled indicating that the spindle and the rotary table were coincident. Maybe I'm thinking of something else. Anyway we're in agreement now