Mill unexpectedly cutting crooked

[AdrianRParticipant @adrianr18614]

Hi,

I think I have the answer to why but just wanted to check my reasoning is right.

I was squaring the edges on a piece of EN32B 100mm square and 6mm thick. All went well and the finish was fine, then I noticed the edges were not square to the face. i.e. the 6mm edge to the 100mm face. I worked out the mill had undercut the sides at about 2 degrees.

The setup was. The piece was held in the vice on parallels with the working side to the right. This meant swarf was thrown backwards and conventional cuts were from front to back. The 12mm 4 flute cutter was in an R8 collet and was running at 800RPM. I was taking 0.4mm to 0.5mm conventional cuts.

I know my SX3 can cut square as I had just finished making some parallels and used the same setup to trim the ends. But that was using a 14mm cutter and running at 1600RPM.

I had used the speed indicated in Tubal Cain's Model Engineer's Handbook. Group D in table A on page 5.14 to find the 1600RPM. But I was not sure about running so fast and had calculated 800RPM from the 80ft/min figure given on page 5.15. I have since re-read and think I have too high RPM.

Feed was about 140mm/min.

I checked the head trimming, about 0.015mm over 200mm. I checked the surface of the parallels in the vice and they were similar. So the workpiece was flat and the head square.

Then to find the problem I tried squaring the end of a 40x6mm bar.

Same result;

• Using conventional or climb cuts.
• Using a brand new different make cutter.
• For 0.1mm cuts, and less undercut under this.
• Repeating the same cut without moving the X removed 1/2 of the slope. I.E only cut the top 3mm of the bar.
• To rule out the R8 collet I tried an ER collet, but the same result.

I then tried changing the speed, running it at 1600RPM I get square cuts even with a 0.5mm cut.

I can understand that a tool is pushed away on a deep cut and hence the need for a spring cut, what stumps me is the undercut.

My hypothesis is that the cutting edge first contacts at the bottom of the piece. Then as it rotates the cutting edge engages higher up, but at the same time, the tool/head flexes away. When the flute is passing the piece the tool springs back. This vibration of the head causes the sloping face.

Does this sound right?

Adrian

[AdrianRParticipant @adrianr18614]

[John HinkleyParticipant @johnhinkley26699]

If I've understood your problem correctly, it could "simply" be that the head has slipped or been bumped out of tram. That would give the effect that you're describing. Have you checked it lately?

John

[pgk pgkParticipant @pgkpgk17461]

Hmmm… OP states he checked the tram after the episode.

Options I see: Using the bottom of the cutter rather than towards the top so more flex available. Unnecessary exposed spindle/slop. Conventional milling will tend to pull the cutter into the work and increase lifting of any free edge – so needs minimal overhang. Not held down well enough on the parallels (often the squared sides aren't quite square or the vice lifts the work off the parallel..tapped the work end down first then lifted when the other end tapped down?). Driving the cutter through too fast may aggravate all of the above..

pgk

[John HinkleyParticipant @johnhinkley26699]

pgk²,

So he did! I missed that on first read. I have no further questions, m'lud. I rest my case and my brain.

John

[Martin ConnellyParticipant @martinconnelly55370]

Use a 16mm or 20mm cutter. They are a lot stiffer than 12mm cutters. That will rule out the cutter deflecting. Keep small diameters for when you need a small diameter.

Martin C

[David George 1Participant @davidgeorge1]

O

[Anonymous]

Well, that's an interesting conundrum. Never come across it on any of my milling machines. Tool deflection does happen, but it's small, especially with the light cuts being taken. And, as the OP says, tool deflection would cause the bottom of the work to stick out, not vice versa.

Feeds and speeds don't seem unreasonable, although we don't know if the cutter is HSS or carbide. If the cut is unstable then it should generate noise/vibration – did it?

Where's SoD with his vibration monitor when you need him?

Andrew

[EmgeeParticipant @emgee]

Adrian

Suggest you try a cut across the same material but at the opposite side of the vice, check if you have any problem after 1 pass, if so do a finish pass to see if all is well. If the taper is opposite to the cuts made on the other end of the vice it would point to the tram is out.

Emgee

[blowlampParticipant @blowlamp]

It sounds like the workpiece is lifting as it's being cut, if I'm seeing the problem correctly.

Martin.

[Anonymous]

Posted by blowlamp on 29/06/2020 22:14:32:

It sounds like the workpiece is lifting as it's being cut………

I wondered about that too, or if the work was being curved upwards due to pressure from the vice. But the OP states he had the same effect on a 40×6 part, so presumably not?

Andrew

[blowlampParticipant @blowlamp]

Posted by Andrew Johnston on 29/06/2020 22:56:03:

Posted by blowlamp on 29/06/2020 22:14:32:

It sounds like the workpiece is lifting as it's being cut………

I wondered about that too, or if the work was being curved upwards due to pressure from the vice. But the OP states he had the same effect on a 40×6 part, so presumably not?

Andrew

The cutter helix could provide the pulling force, but there would need to be some slack somewhere or inadequate grip of the workpiece to give the movement.

Maybe bowed vise jaws or a slighlty barrel-shaped workpiece would allow the part to pivot up at its pinch-point in the jaws. Loose vise jaws, even?

Martin.

[AdrianRParticipant @adrianr18614]

I have done some more tests.

I checked for the work moving, it is rock solid. I also put a digital level on the head and on the work while cutting. They did not move.

I tried a 16mm cutter and see similar undercutting.

I changed from the right to the left side of the vice. It still undercuts towards the vice. ie the direction is reversed.

I tried 0.25mm, 0.5mm and 1mm cuts all gave the same undercut.

I tried using the top of the cutter instead of the bottom (tip) and the undercut disappeared. For both the 16 and 12mm cutters.

I repeated the top and bottom of the cutter using an ER collet instead of R8 collet. It cut square at the top of the cutter, so the distance from the spindle is not the issue, but the length of the cutter is.

Over the 6mm height, the undercut is 0.15mm, if I do a 0.15mm cut on an undercut edge it cuts square.

All a big puzzle, I will carry on with what I am making and report back if I find any cause.

[pgk pgkParticipant @pgkpgk17461]

Posted by AdrianR on 30/06/2020 18:39:16:

I have done some more tests.

…

I tried using the top of the cutter instead of the bottom (tip) and the undercut disappeared. For both the 16 and 12mm cutters.

Hmm(again),,
the implication is flex in the cutters themselves.. seems hard to believe unless some really ropey cheapo cutters (and the ropey cheap ones I use don't do that) unless yours are all long series.

It suggests a test needed with carbide as opposed to HSS or possible the side of an indexable cutter?? Perhaps still a possibility of the edge of the work lifting and the angle gauge not being sensitive. Have just enough overhang from vice for the cut and find a way to clamp it down in the vice as well?

Climb milling is the other solution if you're happy with that…

pgk

[EmgeeParticipant @emgee]

If you are cutting with used cutters the top of the flutes will no doubt be in better condition so more likely to cut cleanly.

Emgee

[Martin ConnellyParticipant @martinconnelly55370]

It does not seemed to have been mentioned yet but does it make difference if the spindle is clamped tight or if it is a bit loose?

Martin C

[William ChithamParticipant @williamchitham75949]

I'm an absolute beginner but it occurs to me that if the cutter isn't perfectly aligned with the spindle it would describe a cone which would create an undercut, for example if the cutter was bent. If it happens with different cutters then maybe there is some crud in the spindle nose that prevents the collet from seating properly or maybe even the spindle is bent?

William.

[AdrianRParticipant @adrianr18614]

All my tests were with the spindle, Z and X all clamped down. Have not tried it with a loose spindle.

I have discounted a bent cutter as I have tried 4 different cutters, 3 of them were different makes, one was brand new out of the sealed box.

Also doing two spring passes removes the undercut, so the tool must be parallel and running true.

I think the most likely cause is head vibration, the cutter gets forced away from the work while in contact and then springs back between cuts. Which is why at high RPM it cuts true, it does not have time to move between cuts and the head deflection is averaged out.

I had a look with a magnifying glass and the undercut is not straight but convex which could also be explained by the above movement.

[Author]

Posts