A stepped shaft can produce different runout values when it is rotated between centers, supported on two journals, placed on V-blocks or located in an assembly-like fixture. The machine has not necessarily become inaccurate; each setup may establish a different reference axis and boundary condition.
Datum selection must therefore begin with the drawing and functional requirement, not with whichever machine support is easiest to build. The straightening system needs a repeatable way to realize the agreed reference, measure the controlled surfaces and apply correction without allowing support error, feature geometry or shaft sag to dominate the result.
Five References That Must Not Be Confused
| Reference | Mening | Hvorfor det betyder noget |
|---|---|---|
| Drawing datum | Datum or datum system specified in the approved technical documentation | Defines how the geometric requirement is interpreted |
| Datum feature | Real surface or feature identified to establish the datum | Contains manufacturing variation and must be contacted/evaluated appropriately |
| Functional reference | Surfaces or axis that locate the shaft in the real assembly | Helps confirm whether the drawing and inspection method protect function |
| Machine support/rotation reference | centre, rollers, journaler, V supports or fixture elements used during straightening | Determines how the shaft is physically held and rotated |
| Acceptance gauge reference | Datum realized by the customer or agreed reference measurement system | Controls correlation and final conformity decisions |
These references may be intentionally aligned, but they are not automatically identical. A machine roller touching a convenient diameter does not become the drawing datum merely because the part rotates smoothly.
Why Stepped Shafts Need a Datum Plan
Unlike a uniform round bar, a stepped shaft can include:
- multiple diameters with different tolerances and surface finishes;
- bearing journals that define the functional rotation axis;
- center holes used during turning or grinding;
- gear, splines, threads or worm profiles;
- skuldre, riller, keyways and radial holes;
- hollow sections, flanges or asymmetric ends;
- different stiffness and sag along its length.
If the control treats every diameter as a perfect continuation of one cylinder, it can misinterpret eccentric machining, lobing, feature interruption or datum error as shaft bending. It can then choose the wrong correction direction or press a surface that should be protected.
Start with the Controlled Characteristic
Before selecting supports, identify exactly what must meet the drawing.
| Karakteristisk | Datum Dependency | Typical Question for Straightening |
|---|---|---|
| Axis straightness or surface-line straightness | May be controlled without an external datum, depending on the specification | Is the machine measuring the same geometric characteristic or only rotational variation? |
| Cirkulær udløb | Requires rotation about a specified datum axis | Which datum feature(s) establish that axis? |
| Total udløb | Requires a datum axis and evaluation over the controlled surface | Which axial path/stations represent the specified surface? |
| Coaxial location of stepped features | Depends on the applicable drawing definition and datum | Which functional journals establish the reference axis? |
| Gear or spline functional runout | May need a master gear, ball probe, span measurement or another feature-specific method | Can a smooth OD probe represent the functional requirement? |
Runout is inherently related to a datum axis. Straightness and shop-floor TIR are not interchangeable with it. Review Skaftets rethed vs Runout vs TIR before approving the gauge concept.
Common Datum Candidates for Stepped Shafts
Center Holes
Center holes can reproduce the axis used during turning or grinding when they are specified, intact, clean and suitable for the required measurement. They can be useful for rotating the part without contacting finished journals.
They are not automatically the correct final functional datum. Damaged, contaminated, mismatched or process-only center holes can create a stable but irrelevant measurement.
Two Bearing Journals or a Common Journal Reference
Two functional journals often provide a strong candidate when they locate the shaft in bearings. The support or mathematical association method must be agreed because journal roundness, diameter difference, spacing and surface condition influence the realized axis.
Supporting directly on finished journals also introduces contact and marking risks. The tooling material, contact pressure, cleanliness and permitted support zones must be validated.
One Cylindrical Datum Feature
A single sufficiently long cylindrical feature can establish an axis when specified and measured appropriately. A short journal may not provide stable orientation by itself, while a long feature can still contain form error that affects the associated axis.
Gear, Spline or Thread Functional Features
Some shafts are accepted relative to a functional tooth or spline datum. A standard radial probe on tooth tips may not reproduce that requirement. Master gears, ball probes, span methods or other application-specific gauges may be needed.
The straightening machine can still use smooth journals for support, but the relationship between the machine reference and the functional gauge must be demonstrated.
V-Blocks or Rollers
V-blocks and rollers are support methods, not drawing datums by themselves. Their angle, spacing, contact diameter, rundhed, cleanliness and shaft sag affect the observed result. They can be appropriate when correlated to the required datum and controlled as part of the measurement method.
End Face or Shoulder
An end face or shoulder can control axial location and probe-station position, but it does not by itself establish a rotational datum axis. Face runout or burrs can also influence seating.
Datum Selection Decision Matrix
| Arbejdsemne / Krav | Strong Starting Candidate | Main Risk to Resolve |
|---|---|---|
| Shaft finished and assembled in two bearings | Two specified bearing journals/common datum | Journal form, marking and gauge correlation |
| Shaft ground between centers and accepted from center-hole axis | Specified center holes | Damage, cleanliness and whether centers remain the final design reference |
| Gear pitch runout relative to journals | Journal datum plus feature-specific gear measurement | Tooth-tip probing may not represent pitch/functional runout |
| Spline alignment relative to bearing seats | Bearing journals plus approved spline gauge | Angular indexing and functional-gauge correlation |
| Raw stepped shaft before finish grinding | Process datum defined for the manufacturing route | Allowance, temporary surfaces and change to final inspection datum |
| Long slender shaft with self-weight sag | Drawing datum plus controlled support/sag method | Machine result may be dominated by gravity and support position |
| Hollow or thin-wall stepped shaft | Functional datum with distributed low-contact support | Ovality/collapse and support-induced deformation |
The matrix identifies a starting concept. Final selection requires the drawing, processtadiet, assembly function and representative gauge-correlation evidence.
A Seven-Step Datum-Selection Workflow
1. Review the Drawing and Revision
Identify the controlled characteristic, datum feature symbols, datum sequence, tolerance, material condition and any drawing notes. Record the applicable standard and edition instead of relying on a shop-floor term such as “concentricity” or “TIR” without a definition.
2. Identify Functional Interfaces
Determine which journals, centre, gear, splines, faces or holes locate the shaft in its assembly and which surfaces drive noise, bearing load, sealing, meshing or downstream machining.
3. Separate Process and Final Datums
A shaft may be straightened before finish grinding using temporary or process surfaces. The machine datum for that stage may differ from final inspection. The relationship and remaining machining allowance must be explicit.
4. Select Support and Rotation Surfaces
Choose surfaces that can locate and rotate the workpiece repeatably without unacceptable marks. Define support spacing, contact geometry, axial location and anti-slip/drive method.
5. Select Measuring Stations and Excluded Zones
Map every controlled station, monitoring station and excluded feature. Avoid keyways, riller, radial holes, damaged areas and unsuitable tooth surfaces unless the measurement method is designed for them.
6. Correlate with the Customer Reference Gauge
Measure the same good, borderline and bad parts on both systems. Investigate systematic differences in datum realization, station position, rotation, filtering, rounding and part condition before setting acceptance limits.
7. Freeze the Method in the Recipe and Test Plan
Document the datum, støtter, sonder, station coordinates, feature masks, rotation and decision rule. Treat later changes as controlled engineering revisions, not operator preference.
Physical Datum Realization vs Mathematical Association
A datum is theoretically exact, while the real datum feature contains form error. The reference can be realized physically by contacts/fixtures or mathematically from measured data. Those approaches can produce different results if the association method, filters or contact points differ.
For an automatic straightening project, the quotation and validation plan should state whether the machine:
- physically rotates on centers or journals;
- supports on rollers/V-blocks while deriving an axis from selected probes;
- fits an axis mathematically from one or more datum features;
- uses a common datum from two separated journals;
- references a functional gear/spline gauge outside the machine;
- combines machine measurement with a separate acceptance gauge.
Do not use the phrase “measured from datum A” unless the method explains how datum feature A is actually realized.
Support Error, Sag and Apparent Runout
The observed probe signal can change even when the shaft itself has not changed.
| Error Source | Apparent Effect | Kontrollere |
|---|---|---|
| Dirt or burr on a support journal | Periodic lift or unstable seating | Cleaning, inspection and seating check |
| Journal roundness/form error | Runout signal at other stations | Datum-feature evaluation and gauge correlation |
| Unequal roller height or worn V support | Axis tilt and station-dependent error | Tooling calibration and maintenance |
| Shaft sag under self-weight | Long-wave deflection between supports | Defined orientation/support spacing and sag assessment |
| Excessive probe force | Local deflection on slender sections | Suitable measuring force and support plan |
| Keyway, spline, tooth or radial hole | Discontinuous or false peak signal | Feature mask or dedicated measurement method |
| Thermal or clamping condition difference | Customer/machine disagreement | Freeze part condition and free/clamped state |
| Axial station mismatch | Different result on tapered/stepped geometry | Drawing-based station coordinates and axial location control |
Repeating a bad setup can make it look precise. Repeatability must be evaluated together with datum relevance and correlation.
Datum Choice and the Correction Model
The correction algorithm uses the measured bend direction and magnitude. If the datum changes, the apparent high point, correction station and required over-bend can also change.
The support reference used during pressing may differ mechanically from the measurement datum, but the relationship must remain controlled. The press head should act only on approved zones, and the shaft must be returned to the same measurement reference after unloading.
For springback and released-part control, se Tilbagespringskompensation ved akselretning.
Measuring Featured Shaft Sections
Keyways and Grooves
Do not let the probe drop into the interruption and treat that event as shaft bend. Use a clean measuring track, angular exclusion or another validated method.
Gears
Outside-diameter runout is not automatically pitch or functional runout. Define whether the requirement uses tooth tips, a ball probe, master gear or another approved method.
Splines
Major diameter, minor diameter, flanks and functional spline gauges answer different questions. The selected measurement must match the drawing and assembly requirement.
Threads and Worm Profiles
Thread crests and flanks are not equivalent to a smooth journal. Support and probing should protect the profile and avoid false periodic signals.
Radial Holes
Holes can disturb both probe readings and local stiffness. Exclude them from ordinary OD probing unless the recipe explicitly handles their angular position.
Datum Correlation Test
A useful correlation study includes parts across the decision range, not only clearly good samples.
| Test Item | Evidence to Record |
|---|---|
| Part identity and condition | Model, revision, batch, temperature/cleanliness and process stage |
| Customer reference method | Armatur, datum realization, måler, stationer, filtering and rounding |
| Machine method | Understøtter, sonder, stationer, rotation and feature masks |
| Repeat readings | Same part reseated and remeasured on each system |
| Good/borderline/NOK coverage | Results across the real acceptance boundary |
| Systematic difference | Offset, station-specific difference or unstable setup |
| Decision rule | Which method controls acceptance and how disagreements are handled |
Correlation is not proved by one part reading the same value once. The study should expose reseating, operatør, station and feature effects relevant to production.
The video shows a rotating shaft in a measuring and correction station. It demonstrates the physical arrangement only. It does not identify the drawing datum, probe correlation, tolerance or final acceptance result for every stepped-shaft project.
Datum Definition Sheet for a Straightening Project
Freeze the following information before machine acceptance:
- drawing number, revision, characteristic and tolerance;
- applicable GPS/GD&T standard and edition;
- datum feature(s) and datum sequence/common-datum interpretation;
- functional and process-stage reference;
- support surfaces, stillinger, span and contact materials;
- rotation/drive method and axial stop;
- measuring surfaces, axial coordinates and probe force/range;
- excluded features and angular masks;
- free-state/clamped-state and sag-control condition;
- customer reference gauge and correlation results;
- final decision rule and borderline-part response;
- cleaning, tooling inspection and maintenance requirements;
- recipe/version control and change approval.
Sample Validation Before Final Machine Configuration
The sample set should cover the shortest/longest parts, smallest/largest support diameters, relevant center-hole conditions, highest expected sag, featured surfaces, material/heat-treatment range and borderline acceptance results.
Test at least the following risks:
- repeated loading and reseating on the machine datum;
- correlation to the customer gauge;
- feature masking and angular indexing;
- support marks and probe marks;
- correction followed by return to the same datum;
- multi-station before/after measurement;
- abnormal seating, wrong model and damaged datum response;
- changeover between workpiece families.
Use the Udretningsprøvetest og acceptvejledning to turn these checks into FAT/SAT criteria.
Information to Send for Datum Engineering
Angiv venligst:
- complete shaft drawing and revision;
- applicable GD&T/GPS standard and customer interpretation notes;
- assembly interfaces and functional bearing/gear/spline references;
- process stage and remaining machining allowance;
- customer gauge, fixture and inspection program details;
- current good/borderline/NOK sample results;
- all shaft diameters, lengths, center holes and featured zones;
- permitted support, measurement and correction surfaces;
- surface, marking, sag and measuring-force restrictions;
- representative samples for correlation and straightening trials.
Our engineering team can then define the machine support, rotational reference, probe stations, protected zones and correlation plan. Final acceptance should be based on the agreed datum realization and gauge evidence, not on a generic “runout” value.
Ofte stillede spørgsmål
Are machine supports the same as the drawing datum?
Ikke nødvendigvis. Supports physically hold the shaft. They establish the required datum only when their surfaces, geometry and evaluation method correctly realize the drawing requirement.
Are center holes always the best datum for a shaft?
Ingen. They may reproduce a machining axis, but the final functional datum may be bearing journals, a common datum or another specified feature. Their condition and drawing role must be confirmed.
Can two bearing journals establish one reference axis?
They can be a strong candidate when specified as a common or related datum, but the exact physical or mathematical realization and journal form effects must be defined.
Why do V-block and between-center readings differ?
They contact different features and establish different boundary conditions. Journal form, center-hole condition, support angle/spacing and shaft sag can all change the observed value.
Can the straightening machine use one datum while the customer uses another?
Only if the relationship is deliberate, repeatable and validated through gauge correlation. The project must state which method controls final acceptance and how disagreements are resolved.