How to Prevent Datum Shift on the WJ-800 | Clamping Force, B-Axis Accuracy and Probe Checks

On the WJ-800 machining center, datum shift is a major cause of out-of-tolerance workpieces in precision batch production.

The main causes are insufficient clamping force, workpiece lift, B-axis thermal mispositioning, and probe contamination. Together, these issues account for over 80% of datum-related quality incidents.

With controlled clamping force, contact point inspection, anti-lift support, B-axis warm-up and verification, plus probe cleaning and calibration, datum shift risk can be controlled within ±0.01 mm. This supports Class IT8 and tighter tolerance requirements.

Lock the Workpiece

Set Clamp Force

When machining heavy mold blanks on the WJ-800, clamping force directly determines whether the workpiece stays stable during cutting.

In one aerospace aluminum structural component project, a clamping pressure of 0.8 MPa under a 3 mm cutting depth and 800 mm/min feed rate caused about 0.08 mm of lift. This was enough to exceed Class IT8 tolerance limits.

Clamping force must be calculated from cutting force, not set by experience.

After the initial clamping force is set, verify it with a dial indicator during a dry run.

1. Mount the dial indicator magnetic base on the machine table.

2. Touch the workpiece top surface with the indicator tip.

3. Run the rapid dry-cut path in MDI mode with the Z-axis raised to a safe height.

4. Check whether the reading stays within ±0.01 mm.

5. If the reading is out of tolerance, increase system pressure by 0.2–0.3 MPa and test again.

For angled clamping at B-axis 90° positioning, the effective pressure plate force is usually only 60–70% of the total clamping force.

Increase system pressure by 1.4–1.6× for compensation, and measure at the initial pressing stage rather than after the value has stabilized.

Check Contact Points

The condition of the fixture-to-workpiece contact points is a hidden risk for machining accuracy.

In a WJ-800 stainless steel impeller project, the reference face looked normal, but a local protrusion of about 0.03 mm was present. It was caused by coolant crystallization left from the previous job.

After clamping, this created a gap between the reference face and fixture locator block. The actual cutting depth became 0.03 mm deeper than programmed, which caused tool edge chipping.

A clean-looking reference face can still create a measurable datum error.

Coolant and chip residue are the main contact point contaminants.

1. Before each workpiece changeover, blow off loose chips with compressed air at ≤0.3 MPa.

2. Wipe locating surfaces with a lint-free cloth dipped in IPA, with purity ≥99.5%.

3. Finish with a dry lint-free cloth.

4. Wipe in one direction along the reference locator axis.

5. Do not push contaminants from one end of the locator to the other.

Stop Part Lift

Workpiece lift caused by cutting force is the most common datum shift trigger.

During WJ-800 machining of 6061-T6 aluminum alloy, milling forces can generate a vertical component of up to 500–800 N. If clamping force is not enough, lift displacement builds up layer by layer.

This causes the actual cutting depth in each pass to drift away from the programmed value.

In one actual case, after rough machining a box-type component, the Z-axis datum shifted 0.05 mm per layer. After three consecutive layers, the total offset reached 0.15 mm.

The cause was insufficient clamping force and vibration-induced micro-lift after each cut.

The core anti-lift method is to restrain the workpiece from below, not only press it from above.

Serrated-face clamping pressure plates create mechanical interlock.

They reduce horizontal slip and vertical lift under cutting vibration, cutting workpiece lift by 60–70% compared with flat plates.

In an aerospace structural job shop, we personally experienced a typical datum shift batch scrap case.

After 20 impeller blanks were machined continuously, impellers No. 11–20 showed a systematic 0.06 mm excess wall thickness.

The investigation found that after impeller No. 10, the operator changed the coolant supplier. The new coolant had a higher pH and caused micro-corrosion on the fixture locator block surface.

This reduced the effective contact point height by 0.06 mm. After switching back to the original coolant supplier and re-grinding the locator blocks, the problem disappeared.

Anti-lift control also requires matching cutting parameters with clamping force.

When cutting depth suddenly changes, such as from 2 mm to 5 mm, cutting force can increase by 150–200%. A previously safe clamping setting may become unsafe immediately.

1. Insert an M00 pause before deep-cut tool change areas.

2. Ask the operator to confirm seating with a feeler gauge.

3. Make sure the workpiece is flush against the fixture locator before continuing.

For thin-walled parts with wall thickness ≤3 mm, lift risk is severe.

For a thin-walled aluminum housing with a 2.5 mm wall and 3 mm cutting depth, the cutting force is about 300 N. Reducing clamping pressure to 0.4 MPa and adding internal supporting ribs can remove elastic deformation by making the wall section self-supported instead of force-constrained by clamping pressure.

Anti-lift inspection should be completed during dry running before formal machining.

1. Run the first-pass cutting path in single-block mode.

2. Use half of the planned cutting depth.

3. Measure the Z reference face distance before and after cutting with a depth micrometer.

4. The difference should not exceed 0.01 mm.

Check B-Axis Accuracy

Warm Up Axis

The B-axis rotary indexing axis is a core positioning axis on the WJ-800 horizontal machining center. Its rotation accuracy directly affects workpiece coordinate system accuracy.

The B-axis commonly uses a worm-gear pair, torque motor, or indexed rotary table structure. At cold start, backlash and elastic deformation can be 0.02–0.05 mm larger than in the thermally stabilized state^[1].

In actual projects, B-axis positioning error was +0.03 mm when cold. After thermal stabilization, it dropped to +0.008 mm.

This difference alone can decide whether a precision part passes or fails.

During WJ-800 machining of 6061-T6 aluminum alloy, we encountered a case where the Z-axis datum shifted 0.05 mm per layer during rough machining of a box-type component.

After three consecutive layers, the total datum shift reached 0.15 mm. The root cause was insufficient clamping force combined with vibration-induced workpiece micro-lift after each cut.

Adjusting clamping pressure from 0.8 MPa to 1.2 MPa and installing bottom support screws eliminated the datum drift.

Do not start precision machining before the B-axis reaches a stable thermal state.

Check Zero Position

The B-axis zero position sets the reference datum for the workpiece coordinate system.

Its accuracy controls the precision of all following machining positions.

B-axis zero is defined by a zero dog on the rotary bearing housing and a proximity switch on the main spindle housing.

Each time the B-axis unclamps and re-clamps, the zero switch trigger position may have a small random error. Error within ±0.01 mm is normal; beyond this, the dog position needs adjustment.

1. Mount a reference bar in the spindle.

2. Use a bar length of ≥200 mm and radial runout ≤0.003 mm.

3. Rotate the B-axis to the 0° position.

4. Use a dial indicator to measure reference bar radial runout.

5. Rotate the spindle one full revolution.

6. The maximum minus minimum dial indicator reading should be ≤0.015 mm.

If the result exceeds 0.015 mm, loosen the B-axis clamping mechanism without powering off.

Then manually push the B-axis to the mechanical stop limit position, re-clamp, and measure again.

If runout does not converge to within 0.015 mm after three repetitions, inspect the zero dog mounting face parallelism and proximity switch trigger voltage.

Another possible zero shift source is the Hirth coupling.

At specific angles, such as 0° and 90°, the B-axis may rely on the Hirth coupling for positioning. Tooth face wear or debris inside the coupling can reduce positioning accuracy.

If repeatability at specific angles, such as 45° or 90°, is much worse than at other angles, clean the Hirth coupling and test again.

Measure Axis Error

B-axis positioning error includes positioning accuracy, repeatability, homing accuracy, and backlash.

Laser interferometer measurement is one of the most precise methods used in the industry.

The WJ-800 factory accuracy target is around ±0.005 mm/500 mm.

After years of use, actual error may expand to ±0.015–0.020 mm.

When measuring the B-axis with a laser interferometer^[2], fix the laser head on the machine base and mount the reflector on the B-axis rotating component.

Select angular measurement mode. Output the result in arcseconds, then convert it to linear error according to the B-axis rotation radius.

The ideal error curve should be smooth.

If the curve changes sharply at specific angles, those points often indicate Hirth coupling mesh positions.

1. Divide the axis into angular intervals, such as 0°–15° and 15°–30°.

2. Enter the corresponding error compensation value into the CNC system parameters.

3. Use backlash or pitch error compensation depending on the error source.

4. Use bidirectional pitch error compensation if the WJ-800 CNC system supports it.

5. Use no fewer than 11 compensation data points, covering 0°, 9°, 18° through 90°.

6. Measure again after compensation.

7. Confirm that all points converge to within ±0.008 mm.

Verify Probe Setup

Clean Probe Tip

The probe stylus is part of a precision measurement system.

It works through mechanical triggering or non-contact induction, and contamination on the stylus tip directly affects trigger stability.

The WJ-800 commonly uses Renishaw OMP60^[3] or Heidenhain TS series touch-trigger probes.

The allowable stylus ball form deviation is usually ≤0.001 mm. Coolant, aluminum chips, steel chips, and oil are the main contamination sources.

We once found that probe repeatability had degraded to 0.015 mm after a late-night shift change.

This was five times worse than the normal value. After half an hour of troubleshooting, the cause was traced to coolant splash entering the probe protective guard during tool cleaning on the previous shift.

After disassembling, cleaning, and reassembling the probe guard with a new seal ring, repeatability returned to 0.003 mm.

A dirty probe tip can turn a good datum into a wrong offset.

1. Use cotton swabs and IPA with purity ≥99.5%.

2. First blow off loose chips around the stylus tip with compressed air at ≤0.3 MPa.

3. Blow in the direction opposite to the stylus tip to avoid pushing debris into the probe body.

4. Wipe the stylus ball surface with an IPA-damped cotton swab.

5. Use a single-direction rotary motion for two passes.

6. Finish with a dry cotton swab.

7. Confirm that no liquid droplets or fibers remain.

Do not blow compressed air directly at the probe housing seams.

This can force coolant into the probe internal circuitry.

If the seal ring is aged or damaged, coolant can enter the probe guard.

Replace seal rings with original manufacturer parts. Standard nitrile rubber O-rings can swell after long coolant exposure, which weakens sealing performance.

Calibrate the Probe

Probe calibration builds the exact relationship between the stylus radius compensation value and the trigger position.

The WJ-800 CNC system calculates workpiece coordinates from probe trigger signals, so any stylus radius calibration error directly affects all measurement results.

For example, if the actual stylus radius is 3.000 mm but the system compensation value is 3.005 mm, all measurement results will have a constant +0.005 mm positive deviation.

This can be fatal for Class IT7 tolerance grades.

1. Mount the master ball in the spindle at a fixed position.

2. Move the probe to contact the master ball from X, Y, and Z directions.

3. Record trigger coordinate values.

4. Let the system calculate the stylus radius compensation.

5. Use the formula: stylus radius compensation = master ball radius − contact point coordinate offset from master ball center.

Heavy cutting vibration can transmit through the spindle and probe mount.

This may cause micro-movement of the probe mounting face, so recalibration is required.

After calibration, run a verification program and measure the same master ball again.

The 3D measurement result should deviate from the master ball nominal value by no more than 0.003 mm.

Run Probe Checks

The probe check program is the final gate for confirming that the probe, CNC system, and mechanical interface are working normally.

Running the probe check before formal machining helps detect probe failure or calibration failure before batch nonconformance occurs.

In the master ball measurement test, the probe follows a preset path and contacts a master ball with known diameter and position.

The system compares measured values with the master ball nominal value. If the deviation exceeds ±0.010 mm, the system triggers an alarm.

• This test can detect probe trigger threshold drift.

• It can detect trigger delay caused by stylus contamination.

• It can detect sampling error in the CNC measurement module.

• Trigger threshold drift may be caused by coolant temperature changes affecting trigger circuit parameters.

The reference point return test verifies probe accuracy when returning to machine origin.

1. Move the probe along X, Y, and Z directions to their reference points.

2. Use hard stops or optical reference switch trigger positions as the reference points.

3. Record return coordinates.

4. Compare them with previously recorded values.

5. Run 5 repetitions.

6. The maximum difference should be ≤0.002 mm.

If the deviation increases progressively, such as adding 0.001–0.002 mm after each return, the reference switch or encoder origin may have a problem.

Datum shift prevention depends on proactive control, not post-incident detection.

1. Before each workpiece changeover, inspect contact points and record the result.

2. Warm up the B-axis for at least 15 minutes per shift, then verify repeatability by measurement.

3. Clean the probe every 2 hours and recalibrate it every 8 hours.

4. Create a probe check record archive.

5. If check deviations show a systematic trend, stop the machine immediately and investigate.

Strictly applying these four measures can reduce out-of-tolerance issues caused by datum shift by over 90%.