Machining center accuracy checks

⌖How to inspect

• Checking levelCommon

  ToolsPrecision level (Hurco: resolution 0.020 mm/m)

  ProcedureHurco: place the level on the table center in the X direction, move the table to nine positions (dividing the X and Y strokes into thirds) and read at each; then place it in the Y direction and read the same nine positions. Makino: the length of the bubble changes with temperature and vibration, so adjust the bubble length before measuring (see the level manufacturer's manual for how).

  TolerancesHurco 3-axis machines: no position may exceed 0.020 mm/m (if it does, call service)

  Okuma America calls level "the foundation of all other accuracy" and says to include it in the annual check because it changes as the foundation moves.

  ^{[25][26][27][28]}

• Foundation (floor)Common

  Tools—

  ProcedureExample from Makino's machine delivery procedure: floor thickness of 300 mm or more for machine weights under 4 t, and 400 mm or more for 4 t and over (thicker still for large machines). Rigidity goes with the cube of the floor thickness. Use reinforcing bar, and isolate the floor from the vibration of other machines. Do not span underground beams. Floor flatness ±5 mm per 3–5 m square. After construction, allow 4 weeks before placing the machine.

  Tolerances—

  One company's example. Follow the machine manufacturer's installation specification.

  ^[29](1 source, for reference)

• Temperature environmentCommon

  ToolsThermometer

  ProcedureKeep the temperature around the machine constant. Avoid places hit directly by doorways, direct sunlight, warm air near the ceiling, or air-conditioning airflow. Prevent condensation.

  TolerancesMakino's recommendation: room temperature 10–40°C (ideal 20±1°C), humidity 35–70% (no condensation), dust 0.3 mg/m³ or less, illuminance 150 lux or more.

  1 m of cast iron expands or contracts about 10.5 µm per 1°C change in temperature (Makino).

  ^[30][31](1 source, for reference)

• Standardize the warm-up conditions before measuring accuracyCommon

  Tools—

  ProcedureConditions under which Makino measures accuracy at the factory: power on at least 3 hours before measurement, run a fixed program for 30 minutes, then measure immediately. When measuring early in the morning, leave the power on from the previous day. The same conditions apply for inspection, witnessing, installation, and trial runs.

  Tolerances—

  Accuracy (especially squareness) changes greatly with installation conditions, so to compare with the accuracy at shipment, match the conditions (Makino).

  ^[32](1 source, for reference)

• Spindle taper runout and test bar runoutCommon

  ToolsDial indicator reading to 0.0025 mm (0.0001"), precision test bar

  ProcedurePress the emergency stop. Put the indicator on the taper surface and turn the spindle by hand. Then insert the test bar and read directly below the gauge line and 150 mm below it (Haas, monthly or as needed).

  TolerancesHaas values: taper TIR 0.005 mm or less, test bar at the gauge line 0.013 mm or less, 150 mm from the gauge line 0.025 mm or less.

  Makino distinguishes spindle runout accuracy as static accuracy and vibration measurement as dynamic accuracy. Haas's paid inspection (HFO) also includes a spindle runout (sweep) check. Tolerances are the machine manufacturer's.

  ^[33][27][28]

• Correction when the test indicator is set at an angleCommon

  ToolsLever-type dial indicator (test indicator)

  ProcedureIn principle, set the stylus perpendicular to the direction of measurement. If that is not possible, keep the angle as small as possible (Makino: within 20°); true value ≈ reading × correction factor.

  TolerancesCorrection factors: 10° = 0.98, 20° = 0.94, 30° = 0.86, 40° = 0.76, 50° = 0.64, 60° = 0.50

  Mounting it wrongly can damage the table (Makino).

  ^[34][35]

• Circular motion accuracy (ballbar test)Common

  ToolsBallbar (a telescoping displacement sensor with precision balls at both ends), magnetic sockets for the spindle side and the table side

  ProcedureMount the ballbar between the spindle and the table, and measure once clockwise (CW) and once counterclockwise (CCW) with G02/G03 circular interpolation. Add extra arc (such as 90° or 180°) before and after to allow for acceleration and deceleration. Standard radii are 100 and 150 mm (larger with extension bars, and 50 mm with a small-circle kit).

  TolerancesThe difference between the maximum and minimum radius of the circle is the "circular deviation" (ISO 230-4). The machine tolerance is in the standard itself (paywalled) and was not checked.

  The causes can be told apart from the distortion of the plot: backlash (steps), spikes at reversal, lateral play, cyclic error, servo gain difference (servo mismatch), scale difference between axes, straightness, squareness, and stick-slip. A report from Tokyo University of Agriculture and Technology describes the ballbar as a measuring instrument widely used for accuracy inspection of machining centers.

  ^{[36][18][37][15][14][12]}

• Machining test pieces (circle, diamond, square)Common

  ToolsTest piece, coordinate measuring machine (CMM)

  ProcedureISO 10791-7 specifies a test in which a test piece with circle, diamond, and square shapes is cut under set conditions and measured on a CMM. There are three sizes: 80, 160, and 320 mm square (M1_80, M1_160, M1_320).

  TolerancesExample of the 320 mm test piece of ISO 10791-7:2014 presented in a Renishaw white paper: cylindricity of the center hole 0.015 mm, circularity of the circle 0.020 mm, position of the four corner holes 0.050 mm, straightness of each side 0.015 mm, squareness and parallelism 0.020 mm.

  Renishaw says the ballbar can give an idea, before cutting, of whether the circle roundness and hole positions are likely to be within tolerance. However, cutting adds spindle deflection, vibration, tool change repeatability, and so on, so the results do not correspond directly.

  ^[18][16][17]

• Positioning accuracy and repeatabilityCommon

  Tools— (the text of the standard's test method was not checked)

  ProcedureTests follow JIS B 6336-4 / ISO 10791-4; general requirements follow JIS B 6190-2 / ISO 230-2.

  Tolerances— (not checked; the text is paywalled)

  THK lists these causes of ball screw positioning error: lead accuracy, axial clearance, axial rigidity of the feed screw system, thermal displacement from heat, and change in guide attitude during travel (pitching and yawing).

  ^[38][7][10]

• Backlash (lost motion when the feed direction reverses)Common

  ToolsBallbar, etc.

  ProcedureIn a ballbar plot it appears as a step along one axis (Renishaw). Ask the manufacturer or dealer to adjust and compensate (Okuma America says to consult the dealer for accuracy and geometry adjustment).

  Tolerances—

  THK: axial clearance does not affect positioning accuracy in one-way feed, but becomes backlash when the feed or load direction reverses.

  ^[38][18][26]

• Thermal deformationCommon

  Tools—

  ProcedureTests follow JIS B 6336-10 / ISO 10791-10 (thermal distortion test). For daily control, keep the room temperature stable and warm the machine up before machining or measuring.

  Tolerances—

  1 m of cast iron expands about 10.5 µm per 1°C (Makino). The Machine Tool Engineering Foundation (Japan) says it has generally been thought to take 2 hours or more to reach thermal stability.

  ^{[21][22][30][39]}

• Annual accuracy and geometry checkCommon

  ToolsPrecision measuring instruments such as a precision level, test bar, and dial indicator

  ProcedureOkuma America says to include in the annual check not only level but also bringing accuracy and geometry back to the factory shipment standard. Haas's paid inspection includes level, spindle runout (sweep), spindle orientation alignment, and a check of the tool changer, among other items.

  TolerancesMachine manufacturer's standard

  ^[26][28]

≡Related standards

• HorizontalJIS B 6336-1:2018 Machining centres — Test conditions — Part 1: Geometric accuracy of machines with horizontal spindle (horizontal Z-axis) / ISO 10791-1:2015 (JIS is a modified adoption = MOD)^[1][2]
• VerticalJIS B 6336-2:2025 Machining centres — Test conditions — Part 2: Geometric accuracy of machines with vertical spindle (vertical Z-axis) / ISO 10791-2:2023 (identical = IDT)The description of ISO 10791-2:2023 says it covers 3-axis machines up to X 5000 mm, Y 2000 mm, Z 2000 mm, with tests for rotary and tilting tables in an annex.^[3][4]
• CommonJIS B 6336-3:2000 Machining centres — Inspection conditions — Part 3: Static accuracy of machines with integral indexable or continuous universal heads (vertical Z-axis) / ISO 10791-3:1998 (IDT)^[5][6]
• CommonJIS B 6336-4:2000 Machining centres — Inspection conditions — Part 4: Positioning accuracy of linear and rotary axes / ISO 10791-4:1998 (IDT). General requirements for the test method: JIS B 6190-2:2016 / ISO 230-2:2014ISO 10791-4 remains the 1998 edition (checked against the ISO catalog).^{[7][8][9][10]}
• HorizontalJIS B 6336-5:2000 Machining centres — Inspection conditions — Part 5: Positioning accuracy of pallets / ISO 10791-5:1998 (IDT)^[11][6]
• CommonJIS B 6336-6:2018 Machining centres — Test conditions — Part 6: Accuracy of feeds, speeds and interpolations / ISO 10791-6:2014 (IDT). General requirements for circular motion tests (ballbar, etc.): JIS B 6190-4:2025 / ISO 230-4:2022^{[12][13][14][15]}
• CommonJIS B 6336-7:2018 Machining centres — Test conditions — Part 7: Accuracy of a finished test piece / ISO 10791-7:2014 (IDT)ISO has issued 10791-7:2020, and the JIS corresponds to the previous edition (edition mismatch).^[16][17][18]
• CommonJIS B 6336-8:2002 Machining centres — Inspection conditions — Part 8: Evaluation of contouring performance in the three coordinate planes / ISO 10791-8:2001 (IDT)^[19][6]
• CommonJIS B 6336-9:2002 Machining centres — Inspection conditions — Part 9: Evaluation of operating times of tool change and pallet change / ISO 10791-9:2001 (IDT)^[20][6]
• CommonJIS B 6336-10:2025 Machining centres — Test conditions — Part 10: Evaluation of thermal distortions / ISO 10791-10:2022 (IDT). General requirements for thermal distortion tests: JIS B 6190-3:2023 / ISO 230-3:2020^[21][22][23]
• CommonJIS B 6190 Test code for machine tools — general requirements (Part 1 geometric accuracy, Part 2 positioning accuracy, Part 3 thermal distortion, Part 4 circular motion accuracy). Each is identical to ISO 230-1, -2, -3, and -4.The title of JIS B 6336 is inconsistent: the 2000–2002 editions say "Inspection conditions" and the 2018 and later editions say "Test conditions".^{[24][9][23][14]}(1 source, for reference)

The standard texts are paid, so tolerances are written from sources such as in-house standards published by machine manufacturers. For the correspondence between standards, see the standards correspondence table as well.