A slide table arrives with a measured repeatability and a machined datum. Neither survives being bolted to a surface that is not flat, or being pulled down with an impact wrench. The installation takes half a day; getting it wrong takes a week to diagnose.
1. Check the mounting surface before anything else
The machine base must be flat to 0.02 mm over the length of the slide and free of burrs, paint and swarf. Check it with a precision straight edge and a feeler gauge, or with a dial indicator on a surface gauge. A 0.05 mm hollow under the middle of the base will bend a 600 mm aluminium frame into a bow, and the rails will follow it.
Clean the surface, then stone it lightly to remove high spots. Do not file it. A file creates a low spot that no shim will recover, because the shim conforms to the file marks rather than to the frame.
2. Rough position and pre-tighten
Set the slide in place and start the bolts by hand. Bring them to finger-tight in a cross pattern, then to 50 % torque. At this stage the slide should still be able to move a few hundredths under light hand pressure — if it cannot, something is sitting under the base.
Torque values
Tighten in at least three passes, always in a cross or diagonal pattern. Over-torque distorts the base; under-torque lets the frame move under load.
- M6 × 1.0
- 10 N·m
- M8 × 1.25
- 18 N·m
- M10 × 1.5
- 30 N·m
- Pattern
- Cross, 3 passes: 50 % / 80 % / 100 %
3. Indicate straightness along the reference edge
Mount a dial indicator on the machine structure — not on the slide — and sweep the machined reference edge of the base along its full length. Target 0.01 mm total indicator reading for a precision axis, 0.02 mm for a general one. Correct by shimming at the mounting points, adding shim at the point where the indicator reads negative.
Do not align to the table top first. Straightness of the base governs everything downstream, and adjusting the table to compensate for a bent base preloads the carriages and shortens rail life.
4. Indicate the table face for flatness and squareness
Sweep the table top at three positions: near end, centre, far end. The reading should not vary by more than 0.01 mm on a precision table. Then check squareness to the machine axis with a granite square and an indicator; 0.02 mm over 200 mm is a reasonable target for a cross slide.
If the table is square but the base is not straight, you have a mounting problem. If the base is straight but the table is not flat, the frame has been distorted by torque and one bolt needs to come back 20 %. Diagnose in that order and you will not chase a phantom fault.
5. Check backlash before you connect the motor
With the coupling disconnected, turn the screw by hand in one direction until the table moves, mark the coupling, then reverse. Measured lost motion should match the figure quoted for the model. If it is larger, the nut preload has relaxed in transit or the coupling is not clamped — check the coupling first, it is the cheaper mistake.
6. Couple the motor without inducing misalignment
A bellows coupling tolerates 0.1–0.2 mm of radial misalignment. Beyond that it transmits a bending load into the screw bearing, which shows up months later as noise and vibration at one end of travel. Fit the motor, leave the coupling bolts loose, rotate by hand through two revolutions, then tighten. If the screw turns freely by hand at every position, alignment is good.
7. Lubricate, then run in
Grease all four blocks and the nut with lithium grease before the first move. Run the axis at 20 % speed for 30 minutes, then at full speed for 30 minutes, then re-check backlash and straightness. Thermal growth during run-in can move a 600 mm frame by 0.01–0.02 mm; re-shim after the axis has reached steady temperature, not before.
8. Commission the control against the real axis
Set soft limits 5 mm inside the hard stops. Tune the servo with the load attached, not free-running, because a slide that is stable unloaded will oscillate with 50 kg on it. Then run a 100-cycle repeatability test with an indicator at the working position — not at the end of travel, where the frame is stiffest and every axis looks good.
The five mistakes that cause most callouts
- Bolting to a painted surface. Paint creeps under load; the axis loses alignment in weeks.
- Aligning to the table instead of the base. Compensates a bent frame by preloading the carriages.
- Impact wrench. Distorts the base and strips the preload out of the nut in one pass.
- Testing repeatability at the end of travel. The stiffest point on the axis and the least informative.
- No run-in before measuring. Cold geometry is not operating geometry.
If an axis still misses tolerance after this procedure, the cause is usually the mounting structure rather than the slide. Send us the indicator readings at the three table positions and we can normally identify which of the five sources is responsible. See what factors affect precision for the error hierarchy.
Common questions
Yes. Check straightness and backlash after 100 hours of operation. Bolted joints settle, and a 0.005 mm correction then avoids a 0.03 mm problem later.