Mazak QT 20 Chatter: Turret Rigidity Comes Before Feed

OD surfaces chatter on every part while ID cuts remain stable. The number that matters is dynamic stiffness at the cutting edge: when cutting force deflects the tool, turret, slide, or workpiece far enough to regenerate a waviness on the next revolution, vibration sustains itself. Unchanged cutting data shifts the first diagnostic priority from feed and speed to turret seating, indexing hardware, tool orientation, and axis lost motion.

Cutting-force and stiffness mechanism

Chatter is a force-versus-stiffness problem. An OD tool and an ID boring tool load the machine through different toolholders, overhangs, contact directions, and turret stations. Stable boring therefore does not prove that the turret is rigid in every cutting direction. It does show that the spindle, workholding, and machine structure can complete at least one class of cut without visible chatter.

The machine is 13 years old, its turret encoder was replaced three weeks before the report, and OD chatter began during the last week. That timing makes the turret the first subsystem to inspect, but it does not make the encoder the direct cause. An encoder primarily establishes or reports turret position. Mechanical seating still depends on the indexing mechanism, shaft, coupling, locking action, and clean contact between the curvic coupling halves.

Heat also changes a backlash measurement. Ball screw, nut, bearings, and machine structure move as they reach operating temperature. Warm the machine before measuring so the recorded reversal error represents normal running conditions rather than a cold transient.

Diagnostic approach comparison

Approach	What it tests	Evidence that favors it	Limitation
Turret mechanical inspection	Coupling contact, locking rigidity, pins, bolts, indexing shaft, and contamination	OD-only chatter, recent turret encoder work, and possible rocking during indexing	Requires safe access and qualified mechanical inspection
Cutting-feed backlash test	Lost motion under a commanded G01 reversal	Chatter may follow axis reversal error or mechanical wear	Compensation cannot restore stiffness to loose or damaged hardware
Rapid backlash test	Lost motion under a G00 reversal at 100% rapid	Separates rapid behavior from cutting-feed behavior	Uses a different compensation value from cutting feed
Encoder/home-position check	Turret orientation and tool centerline after encoder replacement	The symptom appeared after encoder work	Correct home position does not prove full curvic engagement or mechanical rigidity

Inspect turret seating and indexing behavior first. This route addresses faults that backlash parameters cannot correct: loose bolts, sheared pins, a cracked indexing shaft or coupling, excessive sludge between coupling faces, and incomplete locking. Measure backlash after the turret passes the mechanical inspection, or sooner if an indicator test can be performed without disturbing the turret.

Symptom-to-cause decision path

Observation	Likely diagnostic direction	Next check
All OD tools chatter, but ID boring is stable	Load-direction stiffness, OD tool seating, or turret coupling contact	Compare multiple OD stations and inspect turret lockup
One OD station chatters	Local holder, insert, station, or seating problem	Move a known stable tool or holder to another station and repeat the cut
Turret rocks heavily before locking	Indexing shaft, coupling, or locking mechanism wear or damage	Observe indexing without cutting, then inspect the indexing hardware
Drills or boring tools are off centerline	Turret home or orientation error	Check tool centerline and turret position after the encoder replacement
Indicator shows reversal error during G01	Cutting-feed backlash	Average measurements at three axis locations
G00 and G01 results differ	Separate rapid and feed compensation behavior	Record and correct the applicable parameter independently

Begin with a controlled comparison using the same material, workholding, and established cutting data. Determine whether every OD station produces the same pattern. Check whether all drills and boring tools remain on centerline. Index the turret while watching for abnormal rocking before lockup, and check the locked turret for repeatable seating rather than relying only on the displayed station number.

Turret coupling and locking inspection

Isolate machine energy before opening the turret. Do not place hands near an indexing or locking mechanism until the machine is under the site’s energy-control procedure. Internal inspection belongs with personnel qualified to service the turret.

1. Record the relationship between the encoder replacement and the first rejected part: encoder work three weeks earlier and chatter beginning during the last week leave a gap that must be checked against machine usage and any intervening alarms or impacts.
2. Index through the turret stations and observe motion immediately before lock. Heavy rocking points toward the shaft, coupling, or indexing mechanism rather than cutting data.
3. Check the curvic coupling halves for full, clean contact. Remove excessive sludge that can hold the faces apart, following the machine service procedure.
4. Inspect accessible fasteners and locating hardware for loose bolts or sheared pins. Use manufacturer torque and assembly instructions; no torque value is established here.
5. Inspect the indexing shaft and coupling for cracking or looseness. Small movement at this point can become measurable cutting-edge displacement under an OD cutting load.
6. Confirm that the turret locks consistently at each station and that drills and boring tools remain on centerline.

A home-position correction may align the turret logically while damaged hardware still moves under load. Conversely, solid mechanical lockup with tools off centerline directs attention back to encoder alignment or turret home setup. Treat those as separate checks.

Backlash measurement conditions

Backlash must be measured under defined motion modes because this control uses separate compensation for rapid and cutting feed. Manual pulse generator motion does not select those behaviors reliably for this test; run the moves in MDI or automatic mode.

Quantity or condition	Required value	Where to read or apply it
Machine temperature	Normal warmed-up condition	Complete the normal warm-up before mounting the indicator
Measurement locations	Three axis locations, then average	Dial indicator readings
Cutting-feed test	G01 at 10 in/min	MDI command and indicator
Rapid test	G00 at 100% rapid	MDI command and indicator
Common move distance	2 in or 50 mm	Commanded incremental move
Initial directional move	At least one ball-screw pitch	Axis command before zeroing the indicator
Compensation entry	Micron increments	Applicable rapid or feed backlash parameter
Example compensation	10 ≈ 0.0004 in; 20 ≈ 0.0008 in	Parameter value after measurement and sign verification

Back up the current parameter values before changing them. The rapid and feed parameter identifiers are control-specific and are not provided here; read them from the machine parameter documentation. A value of 10 represents 10 microns, which converts to approximately 0.000394 in, matching the stated approximation of 0.0004 in.

MDI backlash procedure

1. Warm the machine and position the axis at the first of three test locations.
2. Record the existing rapid and cutting-feed backlash values. Set the applicable compensation to zero for the measurement only when authorized by the machine procedure.
3. Mount an indicator rigidly and align it with the tested axis. Confirm that all commanded travel is clear of the chuck, tooling, tailstock, and hard limits.
4. Move in one direction by at least the pitch of the ball screw. A common test distance is 2 in or 50 mm.
5. Set the indicator to zero on the trailing side of the motion.
6. Move an additional 2 in or 50 mm away from the indicator, then command the same distance back toward it. The remaining indicator reading is the reversal error at that location.
7. Repeat using G01 at 10 in/min and G00 at 100% rapid. Record the two results separately.
8. Repeat the complete measurement at two more axis locations and calculate the average for each motion mode.

The supplied rapid example is:

G00W2.;W2.;W-2.;M99

The supplied cutting-feed example is:

G01W2.F10.;W2.;W-2.;M99

These examples use W. Confirm from the control documentation that W commands the intended incremental axis on this machine, and convert the move values if the active units are metric. Prove each move with the indicator clear before placing it in contact.

Correction and verification

Apply mechanical correction before electronic compensation when the turret rocks, fails to seat, contains damaged hardware, or has contamination between coupling faces. Backlash compensation removes a predictable position difference at reversal; it cannot replace structural stiffness or hold damaged coupling components together.

If the turret is mechanically sound, enter the averaged cutting-feed result in the feed backlash parameter and the averaged rapid result in the rapid parameter. Confirm the required sign and entry convention in the parameter documentation. The stated scale is micron increments: 10 is about 0.0004 in and 20 is about 0.0008 in.

1. Repeat the indicator test at all three locations after correction.
2. Verify both G01 at 10 in/min and G00 at 100% rapid; improvement in only one mode means the other parameter or mechanical condition remains unresolved.
3. Index through all turret stations and check for consistent locking and tool centerline.
4. Run a controlled OD test cut using the established program, material, insert, holder, and workholding.
5. Compare surface finish across more than one OD station, then repeat a known stable ID operation to confirm that the correction did not introduce an orientation problem.

Accept the repair only when turret seating is repeatable, centerline checks pass, measured reversal error is controlled in both motion modes, and the OD chatter is absent under the previously stable cutting conditions.

Frequently asked questions