Alignment testing of machine tools

Geometric and practical tests of machine tools: instruments (dial gauge, test mandrel, spirit level, autocollimator, laser), lathe alignment tests, 0°/180° mandrel reversal, spirit-level straightness and slope calculations.

Drafted with Aria, reviewed by the AiCanCode.org team. Spotted an error? Use Give Feedback at the bottom of the page.

Why it matters

A machine tool can only make parts as accurate as its own geometry. If the lathe spindle axis is not parallel to the carriage travel, every turned shaft comes out tapered; if a drilling spindle is not square to the table, every hole is inclined. Alignment (geometric) tests are run when a machine is accepted from the maker, after installation or relocation, and after overhaul, and they are the basis of acceptance certificates.

Key ideas

Geometric tests and practical tests. Alignment testing has two parts:

  • Geometric (alignment) tests check the relative position and movement of the machine's elements with no cutting: straightness and levelling of slideways, flatness of tables, run-out of spindles, parallelism, perpendicularity (squareness) and coaxiality of axes and movements.
  • Practical (performance) tests machine standard test pieces under specified conditions and inspect them, e.g. turning a cylinder and checking taper and roundness, or facing a disc and checking flatness.

The tests, their sequence, the instruments and the permissible deviations were systematised by Georg Schlesinger and are now given in ISO 230-1 (general test code) and in machine-specific standards (Indian Standards exist for lathes, drilling, milling and other machines). The permissible deviations are always taken from the applicable standard; they are not derived from a formula.

Instruments used

  • Dial gauge (indicator) on a magnetic stand, usually mounted on the moving element (carriage, table) and touching a reference surface or test mandrel.
  • Test mandrel: a hardened, ground, accurately cylindrical bar with a taper shank that fits the spindle nose. It makes the spindle axis "visible" so that a dial gauge can measure it.
  • Spirit level: measures small inclinations to the horizontal. Its sensitivity is quoted as the slope per division, e.g. 0.02 mm/m per division (about 4 seconds of arc).
  • Straight edge, square and parallel blocks: references for straightness, squareness and parallelism.
  • Autocollimator and laser interferometer: for straightness over long beds, angular errors (pitch, yaw) and positioning accuracy of NC axes.

Typical geometric tests on a centre lathe

  1. Levelling of the bed in the longitudinal and transverse directions (spirit level).
  2. Straightness of carriage (saddle) movement in the horizontal plane (dial gauge against a mandrel held between centres or a straight edge).
  3. Spindle run-out: radial throw of the spindle nose and of a test mandrel near the nose and at about 300 mm from it; axial slip (float) and camming of the face.
  4. Parallelism of the spindle axis to carriage movement in the vertical and horizontal planes (mandrel in spindle, dial gauge on carriage).
  5. Parallelism of tailstock sleeve and taper bore to carriage movement; difference in height between headstock and tailstock centres (coaxiality).
  6. Pitch accuracy of the lead screw.

Dealing with mandrel run-out. A test mandrel is never perfectly concentric with the spindle axis. To separate the spindle-axis error from the mandrel's own eccentricity, readings are taken with the spindle in one position and again after rotating it through 180°. The mean of the two readings at each station locates the true spindle axis; half the difference gives the eccentricity at that station.

Direction tolerances. Many standards allow a deviation only in one sense. For example, the free end of a lathe mandrel may only rise or point towards the tool, because cutting forces and wear push it the other way in service. A test result outside the permitted sense is a failure even if its magnitude is small.

Other machines. On drilling machines the squareness of the spindle to the table is checked by sweeping a dial gauge mounted in the spindle around a circle on the table. On milling machines the tests cover flatness of the table, parallelism of the table to its traverse and to the spindle (horizontal machines), and squareness of the vertical spindle to the table.

Formulas

θ = h / L

  • θ = inclination or angular misalignment (rad); h = difference in height or dial-gauge deviation (mm); L = length over which it is measured (mm). Valid for the small angles met in machine tools (θ well below 0.01 rad), where tan θ ≈ θ.

h = n · S · L_b

  • Rise across one spirit-level base. n = bubble movement (divisions); S = sensitivity (mm/m per division, converted to a pure slope by dividing by 1000); L_b = length of the level's base (mm). Gives h in mm.

θ_div = l / R

  • Angle represented by one division of a spirit level. l = length of one division on the vial (mm); R = radius of curvature of the vial (mm).

e = TIR / 2

  • e = eccentricity of a rotating feature (mm); TIR = total indicator reading (maximum minus minimum) during one revolution (mm).

d_axis = (r₀ + r₁₈₀) / 2

  • Position of the true spindle axis at one station from readings r₀ and r₁₈₀ (mm) taken with the spindle at 0° and 180°.

Conversion: 1 rad = 206 265 seconds of arc; 1 mm/m = 0.001 rad ≈ 206 seconds

Worked examples

Example 1 (standard): straightness of a lathe bed with a spirit level Given: spirit level of sensitivity 0.02 mm/m per division, base length 250 mm. It is stepped end-to-end along a 1.25 m bed. Bubble readings at the five stations: +1, +2, 0, −1, +1 divisions. Find the straightness error relative to the line joining the end points.

  1. Rise per division over one base: h = n · S · L_b = 1 × (0.02/1000) × 250 = 0.005 mm.
  2. Rise at each step: +0.005, +0.010, 0, −0.005, +0.005 mm.
  3. Cumulative heights at stations 0 to 5: 0, 0.005, 0.015, 0.015, 0.010, 0.015 mm.
  4. Line joining ends rises 0.015 mm over 5 steps, i.e. 0.003 mm per step. Its heights: 0, 0.003, 0.006, 0.009, 0.012, 0.015 mm.
  5. Deviations from that line: 0, +0.002, +0.009, +0.006, −0.002, 0 mm.
  6. Straightness error = maximum − minimum = 0.009 − (−0.002).

Straightness error = 0.011 mm over 1.25 m. (The bed is convex in its middle part.)

Example 2 (GATE level): parallelism of the spindle axis using a test mandrel Given: a mandrel in a lathe spindle; dial gauge on the carriage touching the top of the mandrel. Readings at station A (near the nose) and station B (300 mm further out), with the spindle at 0° and then turned through 180°:

  • A: r₀ = 0.000 mm, r₁₈₀ = +0.020 mm
  • B: r₀ = +0.030 mm, r₁₈₀ = +0.010 mm

Find the spindle-axis deviation over 300 mm, its slope and the mandrel eccentricity at each station.

  1. True axis at A: d_A = (0.000 + 0.020)/2 = 0.010 mm.
  2. True axis at B: d_B = (0.030 + 0.010)/2 = 0.020 mm.
  3. Deviation over 300 mm: d_B − d_A = 0.010 mm (the free end is higher).
  4. Slope: θ = h / L = 0.010 / 300 = 3.33 × 10⁻⁵ rad, i.e. 0.033 mm/m or about 6.9 seconds of arc.
  5. Eccentricity: TIR at A = 0.020 mm, so e_A = 0.010 mm; TIR at B = 0.020 mm, so e_B = 0.010 mm.

Spindle axis rises 0.010 mm over 300 mm (0.033 mm/m); mandrel eccentricity 0.010 mm. Whether this passes depends on the permissible deviation and its permitted direction in the applicable standard.

Common mistakes

  • Treating a single mandrel reading as the spindle-axis position; without the 180° reversal, mandrel eccentricity is mistaken for misalignment.
  • Taking eccentricity as equal to TIR instead of TIR/2.
  • Forgetting to convert spirit-level sensitivity in mm/m to a slope before multiplying by base length, giving answers 1000 times too large.
  • Adding spirit-level readings without regard to sign, or referring straightness to the first station instead of the line joining the end points (or the minimum-zone line).
  • Ignoring the permitted direction of a deviation in the test chart.
  • Testing a machine that is not levelled first or not at operating temperature.

For GATE PI

Questions are mostly conceptual: which instrument is used for a given test, what a test mandrel or spirit level measures, the difference between geometric and practical tests, and run-out versus eccentricity. Numericals are short: spirit-level sensitivity and height differences, slope from dial-gauge readings over a length, eccentricity from TIR, and combining 0°/180° readings. Practise unit conversions between mm/m, radians and seconds of arc.

Quick check

  1. What is the difference between a geometric test and a practical test of a machine tool?
  2. A dial gauge shows a TIR of 0.03 mm on a rotating mandrel. What is the eccentricity?
  3. A spirit level of 0.02 mm/m per division, base 200 mm, moves 2 divisions. What is the height difference across the base?
  4. Why is the spindle rotated through 180° during a mandrel test?

Answers: 1. Geometric tests check the machine's geometry and movements without cutting; practical tests machine and inspect a test piece. 2. 0.015 mm. 3. 0.008 mm. 4. To cancel the mandrel's own eccentricity; the mean of the two readings locates the true spindle axis.

Interview questions

All Metrology, Quality and Reliability interview questions

Try answering each one aloud before you open it.

  1. 1.What is alignment testing in the context of machine tools?Concept

    Alignment (geometric) testing checks the geometry of a machine tool without cutting: straightness and level of slideways, flatness of tables, spindle run-out, and the parallelism, squareness and coaxiality of axes and movements. Each test is done with a specified instrument and set-up and compared with a permissible deviation from the test code (ISO 230-1 and machine-specific standards). It is done at acceptance, after installation or relocation, and after overhaul, and is complemented by practical tests on machined test pieces.

  2. 2.Explain why alignment testing is important for machine tools.Concept

    The machine copies its own geometric errors onto the workpiece: a spindle axis not parallel to carriage travel turns a taper, a drill spindle out of square drills inclined holes, a non-flat table gives non-flat milled faces. Alignment tests detect these errors before they cause scrap, verify that a new or overhauled machine meets the standard before it is accepted, and give a baseline so that wear can be tracked over time.

  3. 3.What are some common methods used for alignment testing of machine tools?Concept

    The basic kit is a dial gauge on a magnetic stand, a test mandrel that represents the spindle axis, a spirit level for levelling and straightness, a straight edge and a square. Run-out is measured with a dial gauge on a rotating mandrel; parallelism by traversing a dial gauge on the carriage along the mandrel; squareness by sweeping a spindle-mounted dial gauge over the table. For long beds and NC axes, autocollimators and laser interferometers measure straightness, pitch and yaw, and positioning accuracy.

  4. 4.How does misalignment affect the reliability of a machine tool?Application

    Misalignment loads bearings, guideways and couplings unevenly: an out-of-line spindle or lead screw produces extra radial loads and edge contact, so wear and heating rise and bearing life falls sharply. Uneven guideway wear then increases the misalignment further, so geometric errors grow with time. Periodic alignment checks catch this early and are part of preventive maintenance.

  5. 5.Why is laser alignment preferred over traditional methods in some cases?Application

    A laser interferometer or laser alignment system keeps its accuracy over long distances where straight edges become impractically long and heavy, and it measures linear positioning, straightness, pitch and yaw in one set-up. It records data electronically, so errors can be mapped along the axis and fed into the CNC's compensation tables. The trade-offs are cost, sensitivity to air temperature and turbulence (which must be compensated), and set-up skill; dial gauge and mandrel tests remain standard for spindle checks.

  6. 6.Describe a scenario where alignment testing would be critical in a manufacturing process.Application

    When a new CNC lathe is installed to turn crankshaft journals with tight cylindricity tolerances, the bed must be levelled and the parallelism of the spindle axis to carriage travel checked with a test mandrel before production. A small inclination of the spindle axis in the horizontal plane would turn every journal with a taper, so the machine is not accepted until the tests are within the standard's limits and a practical test piece confirms it.

  7. 7.A dial indicator traversed along a test mandrel shows a deviation of 0.02 mm over 500 mm. Express the misalignment as a slope and as an angle.Numerical

    Slope = 0.02 / 500 = 4 × 10⁻⁵, i.e. 0.04 mm per metre. As an angle this is 4 × 10⁻⁵ rad, about 8.3 seconds of arc. Before accepting it as spindle-axis error, the reading should be repeated with the spindle turned through 180° and averaged to remove the mandrel's own eccentricity.

  8. 8.A laser alignment system shows a deviation of 0.1 mm over a distance of 2 m. What is the angular misalignment?Numerical

    For small angles θ ≈ deviation / distance = 0.1 mm / 2000 mm = 5 × 10⁻⁵ rad. That is 0.05 mm/m, or about 10.3 seconds of arc (1 rad = 206 265 seconds).

Finished this topic? Mark it so your progress, study plan and readiness keep up.

Stuck on something here?

PreviousCoordinate measuring machinesNext Quality concepts and cost of quality