A shaft arrives with a diameter result that passes inspection, yet another check near the shoulder gives a different answer. Before we adjust the machine, we ask a practical question: were both measurements taken at the same axial section? On a part with taper, a relief, or a short cylindrical land, a few millimetres of movement can change the feature being measured. When we develop a programme for a Horizontal Instant Vision Measuring Machine, we therefore define where the diameter belongs before deciding how many dimensions to collect.

The video shows a HanDing horizontal measurement application. It illustrates the equipment arrangement; the inspection time and measurable features of another shaft depend on its drawing, loading method and programme. We treat the section definition as part of that programme, rather than leaving it to the operator to remember.

Give each diameter a location that survives reloading

“Measure the straight part” is not a sufficiently clear instruction for a production check. We prefer a drawing-based description such as a specified distance from an identified end face or shoulder. The reference must be visible and suitable for the task. If a chamfer makes the end ambiguous, we agree how the reference is constructed before releasing the programme.

We also establish the shaft axis in the measurement view. A width measured obliquely across the silhouette is not the same construction as a width perpendicular to that axis. The operator should be able to reload the part and obtain the same section definition without dragging a measurement box to a convenient-looking position.

For a non-contact measuring machine, these rules concern the visible projected contour. They do not establish the complete three-dimensional shape of the shaft. We keep that distinction in the inspection instruction, especially when the drawing also specifies form or rotational characteristics.

A narrow section and a fitted land answer different questions

One programme may use a narrow band near a nominal axial position. Another may fit opposite edges over a longer cylindrical land and calculate their separation. Both can produce useful numbers, but they should not share an unexplained label such as “D1”. A longer fitting interval can smooth local edge variation while also hiding a gradual change along the land.

We document the section position, the width of the evaluation band, and the edge construction. We exclude chamfers, grooves and reliefs only where the drawing and inspection plan justify that exclusion. Removing inconvenient points after seeing the result changes the method and can conceal a process problem.

Our preferred report makes this visible: the feature name identifies the land, and the inspection drawing marks the axial location. Where the software report format cannot display all these details, a controlled setup sheet can carry them. The requirement is traceability of the method, not a particular screen layout.

Use two sections deliberately when checking taper

If the task is to compare two diameters, we specify two sections and their separation along the axis. For an ideal straight-sided cone, the diameter change divided by that axial separation expresses a taper ratio. That calculation assumes the measured features and reference axis match the intended geometry. It should not be applied automatically to a barrel-shaped, stepped or locally damaged profile.

Two passing diameter values also do not prove that every section between them is acceptable. We may add intermediate sections when the process risk warrants them. Conversely, a local burr should not silently become evidence of uniform taper. We examine the contour and keep the local condition distinct from the trend along the shaft.

A suitable Vision Measuring System can support a repeatable section-based inspection routine. We confirm the required constructions with sample parts and the intended software configuration before promising a reporting format.

Keep diameter, roundness and runout separate

A single backlit view does not establish roundness. Rotating a part and comparing projected widths may reveal orientation-dependent variation, but it does not automatically reproduce a roundness measurement. Runout additionally involves rotation relative to a defined reference axis. We do not rename a convenient optical width as either characteristic.

Where top features and side contours both matter, we can discuss a Vertical and Horizontal Integrated Instant Vision Measuring Machine. Access to two views still requires a clear definition for each reported result. Where the task needs selected features across a larger travel range, an Automatic Video Measuring Machine may be a more suitable starting point for evaluation.

Validate the section, not just the repeat button

For programme acceptance, we recommend unloading and reloading representative shafts, checking the located reference and confirming that the evaluation bands remain on the intended lands. We include parts with the permitted variation in shoulder position or overall length. This exposes a programme that measures the wrong section reliably.

When comparing results with another instrument, we first match the section, direction and feature definition. We then consider contact conditions, temperature and the measurement method. We retain evidence of the setup so a later disagreement can be investigated without reconstructing the inspection from memory.

At Dongguan City Handing Optical Instrument Co., Ltd., we help customers turn shaft drawings into practical optical inspection routines. Send us the drawing, marked measurement sections and representative samples at handing3d@163.com. We can review whether a horizontal system or another Video Measuring Machine fits the application. Any speed figures shown in promotional imagery are application-dependent; we establish the proposed cycle and measurement conditions during evaluation.