how to accurately mill a rectangle with rounded corners (no cnc)

[picclock]

I need to mill a rounded rectangular slot 3.5mm wide x 1.3mm deep in a piece of aluminium. The corner radius is 60mm.

I have a rotary table but its the positioning and alignment that's tricky. My mill has DRO's so I'm thinking maybe work out the XY coords. for the curve and advance the table accordingly. Need to do this in the next few days so I have a little thinking time (I need it all .. .).

The pro's and cons are these.

With the rotary table I can get smooth curves easily, but getting them in the correct place is not so easy. Assuming I cut my 4 curved corners, I can then try to align the work on the table so that the horizontal and vertical cuts seamlessly intersect the quadrants. Aligning the endmill to the centre of the end of the quadrant is likely to be tricky.
 
Using calculated tables will give a more stepped approach, even with a large number of iterations. The potential for errors whilst doing this manually is very high, and I'm not that confident that I can do it 4 times without significant errors. I suspect that it will need at least 100 coords to get the curve to a reasonably smooth profile.

Can't think of any other ways at the moment. Jigging the rotary table will get the corners more or less correct, but I need a technique to align the endmill to join up the quadrants.

Whichever way I go I will have a couple of practise runs.

Any help much appreciated.

Best Regards

picclock

[Lew_Merrick_PE]

I need to mill a rounded rectangular slot 3.5mm wide x 1.3mm deep in a piece of aluminium. The corner radius is 60mm.

Pic,

Let's try this one again.  You need to mill a slot 3.5 mm X 1.3 mm with an R60 mm corner radius?  Are you sure that isn't an R0.6 mm?  (Understand that I have never once had my fingers go dyslexic on me -- 50,000 times maybe, but never once!)  A 60 mm radius is a 120 mm diameter.  A 0.6 mm radius is a 1.2 mm diameter.  I sure hope that the aluminum is 7075 or 2024 and not 6061 for this as 6061 would be a RPITA to mill with that small an endmill (slot drill).

So, if my interpretation is correct, I would start by milling with a (say) ø0.8 mm endmill as short as required to make the full slot depth to make a 3 mm long start and then switch over to a ø1.2 mm endmill to finish it.  Both endmills would be as short as possible.  Use lots of lubricant/coolant.  Keep the air-blast handy to clear the chips out of the slot.

Please note my initial question.  Something does not sound right in your description.  I may well have misinterpreted the question.

[John Rudd]

Lew,

Your reply was exactly my initial thought, hence my reluctance to reply to the question...
I think Picclock has got a measurement wrong...

Milling the slot to the dimensions stated isnt a problem if you have  X/Y dros...Set your initial starting point and then move according to the co-ordinates..

Milling with a .6mm cutter needs a lot of rotational speed......if you can get a cutter that small!!

[DaveH]

Picclock,

Me thinks a little drawing maybe useful.

DaveH

[Jasonb]

I think he wants to mill  a slot to form a rectangle with rounded corners, bit like the O ring groove in Robs hit & miss build.



So its 4 straignt cuts joined with a quater circle of 60mm radius. Its a case of very carefully locating the centre of teh 4 corners onto teh rotary table axis and then joining them with 4 staright grooves, just a lot of carefull setting up using edge finders and dti.

J

[picclock]

@ Jason B

That's a very impressive picture of the hit and miss engine plate, and pretty much spot on. I assume that Robs hit & miss part was made on a CNC machine.

Yeah, I screwed up on the radius dimension  . I was thinking about a method to do this with reasonable accuracy.

The work is 128mm x231 x 15mm. The slot dimensions are 3mm wide x 1.35mm deep. The corners are radiused 10mm (centre of slot) and the overall size of the rounded rectangle is 120x224 (centre of slot). I know these are correct cos I took them from the drawing 

But seriously,  I assume that Robs hit & miss part was made on a CNC machine.

The best I can come up with is do the 4 corner quadrants, getting the dimension to the quadrant edges using an edge finder and my DRO. Then once off the rotary table use the same edge finder to align the part and join the straight bits.

For stuff like this CNC rules 

Many Thanks

picclock
 

[David Jupp]

You could perhaps adapt the principle of a copy lathe - make up a suitably sized template from sheet steel (file the corners) - mount this on the milling table, offset from the stock and a short distance above the table.  Arrange an (adjustable) follower rigidly mounted to mill structure or bench.   Manually drive the table so that the follower stays in contact with the template all the way round whilst cutting the groove.

Will still take quite a bit of care, but I think it should work.

[picclock]

@ David
That's a neat way to transfer complex shapes. It may come to that if I can't work anything else out.

My current thinking is that the problem is aligning the mill centre with the slot centre. If I can do that accurately then I can ignore the errors that occur when I measure to the edge and set from the edge. So really I need to be able to centre on the slot. My Starret edge finder would work well except that its a 6mm tip.

So going down this route, a mod to the Starret edge finder, drill a hole in the end and superglue in a piece of 2mm silver steel, or obtain another edge finder with a smaller tip.  Not sure how the Starret would behave with a smaller diameter driving it, but at least it's a non destructive mod.

picclock

[Jasonb]

Why do you need to locate from the centre of the slot? Locate the straight slots from one edge (4mm one way, 3.5mm the other)and the pivot point of the 10mm corners can be located from two adjacent edges (9mm and 8.5mm)

[picclock]

@ JasonB

I figured that if I located both edges then I can use the DRO to get the centre of the slot, and any inaccuracies in the measuring technique or tool will be cancelled out.

picclock

[Lew_Merrick_PE]

Picclock,

Ah, so I did totally misunderstand the question.  How about this approach: (1) Drill & ream (4) holes at the center of the R60 mm corner radius points of sufficient depth to get the tip of a test indicator to read.  (2) Mill the straight portions of the o-ring groove at the appropriate 60 mm set-over from the holes (using stops set to the CL positions of the Item 1 reamed holes).  Repeat four times.  (3) Set-up your rotary table and set your zero to its center.  Position the part to the rotary table dialing in on the (Item 1) reamed holes.  Rotary mill each quadrant.  Repeat four times.

Remember that the index holes do not need to go through the part, merely deep enough for you to get the ball-edge of your test indicator to ride below the edge where the index hole meets the surface.

???

[Jasonb]

Or drill the 4 holes in the bottom of the plate and have a spigot in the rotary table for them to locate over.

J