Pocket 0.06 mm over on one wall and under on the other after G41 finishing
On a vertical machining centre, a rectangular pocket 60 x 40 x 12 mm deep in mild steel is roughed with a 12 mm four-flute HSS end mill, then finished with a climb-milling pass using G41 D01. The pocket measures 0.06 mm oversize on the +X wall and 0.06 mm undersize on the -X wall, but is correct in Y. The tool room measured the actual cutter at 11.88 mm diameter with a micrometer. The programmer entered D01 = 6.00 from the nominal diameter and has no tool-length or radius pre-setter. The operator ran the finish pass at the same feed as roughing and reports chatter marks on the floor. The machine has linear guideways and preloaded ball screws, with backlash under 0.005 mm.
1. Diagnose the wall error and give the correct D value with the reasoning.
Read the worked answer
The error pattern identifies the cause: it is equal and opposite on the two X walls (0.06 over on one, 0.06 under on the other) and zero in Y. A machine geometry or backlash fault would not be symmetric like that; the cutter is simply being offset sideways by 0.06 mm, which is exactly the error in the radius value entered.
D01 must be the actual cutter radius: 11.88 / 2 = 5.94 mm, not the nominal 6.00 mm. With 6.00 in the offset register, the control holds the tool centre 0.06 mm too far from the programmed contour, so the wall on the compensation side is cut oversize and the opposite wall undersize. Correct D01 to 5.94, check that the finish allowance was actually left in the program, and cut one part and measure before running the batch. Start measuring the cutter on the machine with a pre-setter rather than trusting the nominal size.
2. How do you decide between G41 and G42, and how would you verify the compensation is acting on the correct side without cutting metal?
Read the worked answer
G41 keeps the cutter to the left of the programmed path as you walk along it in the direction of travel; G42 keeps it to the right. So the choice depends on the path direction and on which side the wall being machined lies. For a pocket finished in a counter-clockwise path with the wall on the left, G41 is correct; for a boss machined in the same direction, G42 would be needed. If you swap G41 and G42, the size error swaps walls.
Verify without cutting: raise the Z offset by 20 mm and run the finish pass in single block, watching the tool centre line on the position display. The distance from the tool centre to the contour should equal the D value, 5.94 mm, on the correct side. Check that D01 is active in the block where you think it is, check the lead-in and lead-out moves for gouging, and only then cut a scrap block.
3. Explain why climb milling was specified for the finish, and what cutting parameters would you change to remove the chatter?
Read the worked answer
Climb milling is right here because the machine has preloaded ball screws and backlash under 0.005 mm. In climb milling the tooth enters at maximum chip thickness and exits at zero, so the cutting force pulls the work towards the cutter rather than lifting it, giving a better finish and longer edge life. Conventional milling rubs on entry, work-hardens the skin and gives a worse finish. Climb milling is only safe on a machine with negligible backlash, which this one has.
The chatter comes from using roughing feed on a light finishing cut. For a 12 mm HSS cutter in mild steel take about 25 m/min: n = 1000 x 25 / (3.142 x 12) = about 660 rpm. At 0.05 mm per tooth and four flutes, feed is roughly 130 mm/min, with a radial engagement of about 0.3 mm and the shortest sensible overhang. Use a sharp, preferably new, cutter for the finish, and check that the cutter is not rubbing because the radial depth is too small.