Problem- Difference in Female Taper angle using Taper Cutter

[Buntron]

I have been having trouble with a female taper I am machining in titanium.

the angle of the taper is 4 deg (2 deg each side), +/- 2mins 30secs.

The pocket has two parallel sides with a 2x .170" rads connecting the sides at either end. the overall depth of the pocket is .660".

I am using a custom made end mill with a 4 deg taper ground onto the tool, with a 2mm rad.

the problem is the results. In plastic test parts,the angle of the parallel sides comes out at - 32.4 secs while the curved end comes out at 1min 19.2secs. this is still within the tolerance.

When cut in titanium however the difference in the angle ranges from + 28.8secs for the parallel side, to +5mins 9.6secs for the curved side.

Can anyone explain the difference between the angles?

P.S.
The amount of material left to be removed from the semi-finish pass is 0.25mm.

[rowbare]

You didn't mention the diameter of your cutter or whether it is carbide or HSS. Your cutter might be flexing. Have you tried doing a second (spring) pass to see if that improves the result? How rigid is your machine? Maybe the machine is flexing a bit.

You might also want to play with the amount of material you are leaving for the finish pass. Too much and you get cutter deflection, not enough and the material work hardens and is a bugger to cut.

I don't know if you use G-Wizard or not, but its feed and speed calculator will tell you how much cutter deflection you have.

good luck
bob

[christinandavid]

Hi Buntron, man you get some fussy jobs!

Are you using cutter comp? If not your cutter centreline is likely moving at a constant feedrate. This will mean that the the periphery of the cutter will actually be moving at a faster feedrate when machining the arcs, resulting in greater chipload. On some controls (may be dependant on parameter settings), the correct feedrate is calculated when using cutter compensation.

The formula for internal path is something like: -

Feed=Programmed Feed x [[Programmed Radius - (+ for external) Cutter Radius] / [Programmed Radius]]

In tight corners the feedrate will change significantly, so you will know if its active.

Of course, there is no way to correct the feedrate variations over the length of the taper - the best results would probably be gained using the major diameter as your tool offset.

DP

[Buntron]

The machine is a 6-year old Mori-Seiki NV5000. The machine and the fixturing are both very stable.

The cutter is a custom 2-flute carbide cutter from Sandvik, 8mm shank with a 2 deg taper giving a 7mm diameter with a 2mm rad on the corners. I have tried varying amounts of material for the finishing pass but it doesn't have any effect.

I have tried 2 methods. the first is using our CAM software to spiral cut the pocket. The second is to go to depth and run the cutter using G41 to give better control over the length and width of the pocket. with both methods, the angle is over top tolerance for the curved sides of the pocket.

I have dropped the feed by 50% and still there is no difference. I have made a few changes that I will be trying today.

Believe it or not, these jobs are easy. You should have seen the trouble we had turning an ellipse.

[christinandavid]

So the cutter radius is virtually the same size as the programmed radius? That would require roughly 10% feedrate when machining the rads
(%=100% x [[.170 -.150] / [.170]]).

The other thing to consider is the big increase in cutter engagement in the corners - you really need to use a tapered cutter for roughing and semi-finishing also to eliminate this. The other option would be to 'z-level' machine the pocket in small increments when roughing, to create the taper with hopefully a lot less flexing going on at the corners.

DP

[Buntron]

The roughing and semi-finish passes were spiral cut with a pitch of .3374mm.

The cutter diameter is 45% of the required rad size. I have slowed the feed rate down to 25mmpm and it has reduced the difference between the two angles, but its still 0.080 deg out of mid tolerance.

There should be some new tool holders being delivered in a few days that I am hoping will give improved results.