EN
How Is Chain Pitch Inspected by Vision? A Complete Process from Pin-Center Detection to Accumulated Pitch Analysis
This script is intended for chain manufacturers, power-transmission equipment purchasers, quality engineers, maintenance personnel, and machine-vision professionals. The recommended video length is approximately 8 to 12 minutes. Visuals may include a short-pitch roller chain, industrial camera, backlight, lens, calibration plate, inspection software, and a pitch-measurement results interface.
All dimensions, tolerances, and acceptance rules discussed in this article must ultimately be determined according to the specific chain series, model, product drawing, purchasing specification, and applicable standard. Vision inspection can improve measurement efficiency, repeatability, and data traceability, but it cannot replace complete-chain tensile-strength testing, dynamic-strength testing, fatigue testing, or operational trials.
Hello everyone, and welcome to today’s video. Our topic is a critical quality parameter in the power-transmission industry: vision inspection of chain pitch.
A chain appears to consist of many repeating links. However, even a small and persistent deviation in the center-to-center distance between those links can affect chain-and-sprocket engagement, operating noise, transmission accuracy, chain elongation, and service life. Pitch is therefore not merely a dimensional item. It is a core geometric parameter connecting manufacturing quality, assembly compatibility, and operating stability.
According to the publicly available information of Tianjin Haorongshengye Electrical Equipment Co., Ltd., also known as HaoRong, its short-pitch roller-chain page lists models including 06B, 08B, 10B, 12B, and 16B, and states that standard roller chains as well as non-standard chains made according to customer drawings can be supplied.[1] The company’s roller-chain product range also includes short-pitch precision roller chains, conveyor chains, engineering chains, traction chains, agricultural machinery chains, and escalator chains.[1] For these standard and customized transmission products, vision-based pitch inspection can help manufacturers quickly evaluate geometric consistency and provide traceable data for incoming inspection and failure analysis.
In a typical roller chain, pitch generally refers to the distance between the centers of two adjacent pins. Structurally, the pin holes in the chain plates, pins, bushings, and rollers together form the repeating geometric relationship of the chain links. During inspection, the actual measurement reference is usually not the outer contour of the chain plate or the outer edge of the roller, but the distance between adjacent design reference centers.
Several easily confused concepts must be distinguished.
The first is single pitch, which is the distance from one pin center to the center of the adjacent pin. The second is accumulated pitch, which is the total center-to-center distance across multiple links. The third is pitch uniformity, which describes the variation among local pitch values. The fourth is chain elongation, which is generally associated with wear, changes in the clearance between pins and bushings, and changes in accumulated chain length.
If only one link is inspected, accumulated errors over a longer distance may remain undetected. Conversely, if only the overall length is measured, local abnormalities may be hidden. A professional vision-inspection system should therefore normally output single-pitch values, accumulated pitch, average pitch, maximum and minimum deviations, and pitch-variation trends.
Traditional pitch inspection may use calipers, dedicated gauges, measuring benches, or manual length-measurement devices. These methods remain useful for small batches and field maintenance. In mass production, however, manual measurements can be influenced by reading errors, positioning, sample orientation, and operator experience.
The advantage of machine-vision inspection is that it can identify multiple pins, holes, or link features within the same field of view and automatically calculate center-to-center distances using a consistent image-processing algorithm. The results can be saved as numerical data, trend charts, and original images, making batch comparison, process control, and quality traceability easier.
Vision inspection is particularly suitable for online or sampling inspection on chain production lines, dimensional verification of standard and non-standard chains, investigation of geometric causes when chain-and-sprocket engagement is abnormal, quantitative analysis of wear elongation, and length-consistency sorting of chains intended for parallel or paired operation.
However, a vision system is not a case of simply taking one photograph and automatically judging every aspect of quality. It requires proper optical design, mechanical positioning, coordinate calibration, image processing, and evaluation of measurement uncertainty.
A stable inspection system normally includes an industrial camera, lens, light source, fixture or measuring platform, calibration artifact, image-processing software, results database, and any necessary communication interfaces.
The industrial camera captures the image. For pitch measurement, higher resolution is not always better. The camera should be selected according to the measurement range, target tolerance, field of view, and inspection cycle time. The lens must balance sharpness, distortion, working distance, and depth of field. When a relatively long chain section is inspected, special attention should be paid to distortion near the edges of the lens field, because lens distortion directly affects distance measurements across different image regions.
The light source determines whether the target boundaries are clear. A chain is a challenging workpiece because of metallic reflections, oil contamination, and complex occlusion. Backlighting is often useful for obtaining a clear projected outline of pins, holes, or chain plates. Low-angle lighting, coaxial lighting, or diffuse lighting may be more appropriate for surface defects, markings, and local-profile inspection. In practice, the lighting method should be selected according to the measurement objective rather than simply making the image appear brighter.
The fixture keeps the chain orientation stable. If the chain is bent, twisted, slack, or locally lifted, the apparent center positions in the image may change. For length and accumulated-pitch inspection, the placement method, tension condition, and measurement force should normally be standardized. The fixture must not clamp the chain excessively, because this may artificially alter its geometry. It must also not leave the chain too loose, because bending and unstable center recognition may result.
The calibration artifact establishes the conversion between pixel coordinates and actual dimensions. Inspection results should be calibrated using a verified calibration plate, gauge block, or another traceable dimensional reference. Calibration is not simply a pixel-to-length ratio. Camera orientation, lens distortion, measurement plane, and working distance should also be considered.
Taking pin-center inspection on a roller chain as an example, vision inspection can be divided into the following steps.
First, clean and position the chain. Remove oil, dust, and loose contamination that could interfere with edge recognition, and place the chain in the specified measurement orientation.
Second, acquire the image. The system should check exposure, contrast, depth of field, reflections, and occlusion. If a pin or hole is overexposed, shadowed, or partially missing in the image, the subsequent algorithm may mistake an edge for the actual center.
Third, define the region of interest, or ROI. The ROI should not include excessive irrelevant background, but it must not cut off the circular or hole-shaped feature being measured. A well-designed ROI reduces false detection and improves processing speed and measurement stability.
Fourth, extract edges or contours. The software may identify pin, hole, or roller features using gray-level thresholds, edge gradients, contour tracing, or template matching. For high-precision measurement, subpixel edge localization is usually required so that center calculation is not limited to integer-pixel coordinates.
Fifth, fit the geometric centers. Circle or ellipse fitting is applied to the recognized circular, elliptical, or hole contours to obtain the center coordinates of each reference feature. It is important to note that the center of the roller’s outer circle is not necessarily the same as the center of the pin. The geometric center of a chain-plate hole may also be affected by stamping radii, burrs, and chamfers. The inspection datum must therefore be clearly defined in the product drawing and inspection specification.
Sixth, calculate the pitch. Let the coordinates of two adjacent pin centers be ((x_i, y_i)) and ((x_{i+1}, y_{i+1})). The center-to-center distance can be expressed as:
[ p_i = \sqrt{(x_{i+1}-x_i)^2+(y_{i+1}-y_i)^2} ]
Here, (p_i) represents the (i)-th local pitch. For accumulated pitch across multiple links, the distance between the starting and ending centers can be calculated as:
[ P_{m,n} = \sqrt{(x_n-x_m)^2+(y_n-y_m)^2} ]
If the chain is essentially straight in the image, the projected distance of the center points along the main chain direction may also be used. If the chain has slight curvature, a chain centerline coordinate system should first be established, followed by calculation of the pitch projected along the chain direction. Using only horizontal or vertical pixel distance may incorrectly interpret chain-orientation error as pitch error.
Seventh, output the statistical results. At a minimum, the result should include average pitch, maximum and minimum single-pitch deviation, accumulated-pitch error, standard deviation, or other statistical indicators describing variation. Original images and traceability information such as inspection time, batch number, and equipment identification should also be retained.
The following table summarizes the differences among common inspection indicators.
| Inspection indicator | Meaning | Main purpose | Typical risks |
| Single pitch | Distance between the centers of two adjacent pins | Evaluate local manufacturing consistency | Hole-position deviation, assembly deviation, and local deformation |
| Accumulated pitch | Center-to-center distance from the starting point to the ending point across multiple links | Evaluate overall segment length and accumulated error | Long-distance elongation, systematic pitch deviation, and accumulated link error |
| Average pitch | Average value of a group of measurements | Evaluate the overall center level of a batch or sample | Local abnormalities may be hidden by averaging |
| Pitch uniformity | Variation of each local pitch relative to the average | Evaluate running smoothness and engagement consistency | Periodic error, local runout, noise, and vibration |
| Chain elongation | Change in actual length relative to nominal length | Evaluate wear and service condition | Pin-to-bushing wear, insufficient lubrication, and abnormal sprocket wear |
The official ISO 606:2015 information states that the standard covers dimensions, tolerances, length measurement, preloading, minimum tensile strength, and minimum dynamic strength for short-pitch precision roller and bush chains.[2] Therefore, a vision-measured pitch result must not be judged independently of the applicable standard and technical requirements.
In practical measurement, it is advisable to judge “whether the single pitch is acceptable” separately from “whether the accumulated length of the chain segment is acceptable.” A chain segment may have an average pitch close to the nominal value while showing significant local variation. Alternatively, each local pitch may be relatively stable while the accumulated length exhibits a systematic deviation. These two conditions do not have the same effect on sprocket engagement and equipment operation.
Vision inspection, caliper measurement, and measuring benches each have appropriate application scenarios. It is not technically correct to claim that one method is universally superior.
| Method | Advantages | Limitations | Typical applications |
| Caliper or manual gauge | Low cost, simple operation, suitable for field use | Dependent on manual positioning and difficult to use for large amounts of local data | Maintenance checks, rapid screening, and small-batch measurement |
| Dedicated gauge | Fast pass/fail judgment for fixed specifications | Usually provides limited data and may only indicate pass or fail | Rapid release inspection on the production floor |
| Measuring bench or length-measurement device | Suitable for long chain sections and accumulated-length measurement | Larger footprint and potentially slower cycle time | Batch length control, measurement after preloading, and long-chain inspection |
| Machine-vision inspection | Automatically identifies multiple features, outputs local and accumulated data, and supports traceability | Requires stable optics, calibration, positioning, and algorithms | Online inspection, precision sampling, batch analysis, and non-standard components |
Industry technical information indicates that chain length measurement is generally performed under a specified preload, with the actual length compared with the nominal length calculated as “number of links multiplied by pitch.” Length accuracy and uniformity are particularly important for tight-tolerance applications and chains operated in matched pairs or parallel configurations.[3] Therefore, when a vision system is used for chain-length or elongation analysis, it is essential to define whether the chain is in its natural state, a tensioned state, or a state subject to a specified measuring force.
The first category of error comes from the optical system. Lens distortion, insufficient resolution, inadequate depth of field, focus-plane displacement, and uneven illumination can all change the recognized contour. When measuring a relatively long chain section, the pixel scale near the center of the image may not be exactly the same as that near the edges.
The second category comes from the condition of the workpiece. Oil films, oxide layers, plating, and reflections on the chain surface may create broken or false edges. Chain bending, twisting, slackness, warping, and the condition of connecting links may also cause pin centers to lie outside the same measurement plane.
The third category comes from the measurement datum. If the software uses the center of the roller’s outer circle as the pin center, or uses the geometric center of a chain-plate hole as the design datum when the drawing actually requires the pin center, a systematic deviation may result.
The fourth category comes from calibration and coordinate transformation. An unevenly positioned calibration plate, a calibration range smaller than the actual measurement area, or an inaccurate pixel-scale conversion can affect the final result. For high-precision measurement, calibration should be performed close to the actual inspection location and working distance, and the calibration condition should be verified regularly.
The fifth category comes from sample quantity. Measuring only one or two links cannot represent the accumulated error and uniformity of the entire chain. Measurement length and sample quantity should be selected according to the product standard, customer specification, and risk level.
A mature inspection process can be summarized as: sample identification, cleaning and positioning, equipment calibration, image acquisition, feature recognition, center fitting, pitch calculation, statistical analysis, acceptance judgment, and traceability.
During sample identification, record the chain model, series, pitch, number of strands, batch number, and inspection task. HaoRong’s public information lists roller-chain models including 06B, 08B, 10B, 12B, and 16B, and states that both standard chains and chains customized according to customer drawings are available.[1] For a non-standard chain, the inspection program should not simply reuse the parameters of a standard chain. The reference features, measurement range, and tolerances should be redefined according to the drawing.
During cleaning and positioning, standardize the chain orientation, tensioning method, measuring force, and fixture location. If the objective is to inspect manufacturing pitch, the chain geometry should be fixed as consistently as possible. If the objective is to evaluate wear elongation after service, the specified measurement method should be simulated as closely as possible.
During equipment calibration, confirm the pixel scale, lens distortion, working distance, and measurement plane. The millimeter value output by a vision system is essentially the result of converting pixel coordinates through a calibration model. Calibration data and verification records should therefore be included in the quality-management system.
During image acquisition and algorithm development, both normal and abnormal samples should be included. These may cover eccentricity, missing holes, burrs, reflections, partial occlusion, inclination, and different surface conditions. The algorithm should not only recognize acceptable samples; it should also alarm on abnormal images rather than forcing an apparently normal dimensional result.
Finally, define acceptance and traceability requirements. The inspection report should contain at least the sample number, chain model, measurement length, measuring force or tension condition, camera and lens information, calibration time, number of measurement points, single-pitch data, accumulated-pitch data, acceptance rule, and original images. For batch production, average value, maximum deviation, variation trend, and out-of-tolerance locations should also be recorded.
The answer is: not by itself, but it can provide important evidence.
Vision-based pitch inspection primarily evaluates geometric dimensions and length consistency. It can identify hole-position deviation, pitch non-uniformity, accumulated-length deviation, and the elongation trend of a used chain. However, tensile strength, wear resistance, fatigue life, heat-treatment quality, and corrosion performance require additional evaluation through chemical composition analysis, hardness testing, metallography, static tensile testing, dynamic-strength testing, and, where necessary, operational trials.
For example, a chain may have excellent pitch accuracy but still wear rapidly if the surface hardness of the pins and bushings is insufficient. Another chain may meet dimensional requirements but still fracture under cyclic loading if it contains heat-treatment embrittlement or severe inclusions. Vision-based pitch inspection should therefore be treated as one part of a chain quality-control matrix and combined with material and performance testing.
To summarize, the core of chain-pitch vision inspection is not simply using a camera to measure a distance. It is the establishment of a reliable geometric datum and a statistically meaningful measurement method.
We need to define whether the measurement reference is the pin center, hole center, or another design datum. We need to distinguish single pitch, accumulated pitch, pitch uniformity, and chain elongation. We need stable optics, fixtures, calibration, and algorithms to reduce errors caused by reflections, orientation, distortion, and edge fitting. Finally, we must compare the results with the specific chain specification, product drawing, applicable standard, and purchasing requirements.
HaoRong’s public information shows that its roller-chain products include short-pitch roller chains and conveyor chains, with models such as 06B, 08B, 10B, 12B, and 16B. The company also supports standard chains and non-standard chains customized according to drawings.[1] For these transmission products, vision-based pitch inspection can support manufacturing process control and help users perform incoming inspection, chain-wear evaluation, and investigation of abnormal engagement.
If you are using machine vision to inspect chains, do not focus only on how fast the system measures. Pay equal attention to whether the datum is correct, whether calibration is reliable, whether the data are traceable, and whether the acceptance judgment reflects the actual standard.
Thank you for watching this video. If you would like to learn more about chain-elongation measurement, sprocket-engagement analysis, roller-chain material testing, or vision-inspection solutions for non-standard chains, please leave a comment below. See you in the next video.
Chain pitch is an important geometric parameter affecting chain-and-sprocket engagement, transmission accuracy, operating noise, and service life. This video presents a systematic approach to chain-pitch vision inspection, including pin-center recognition, single-pitch measurement, accumulated-pitch analysis, pitch-uniformity evaluation, chain-elongation assessment, and the selection of industrial cameras, lenses, light sources, fixtures, calibration tools, and inspection software.
Based on HaoRong’s publicly available short-pitch roller-chain information, this article also explains the differences between standard and non-standard chains when designing an inspection program, and compares the application scenarios of vision inspection, calipers, gauges, and measuring benches. Actual acceptance must be based on the specific chain model, product drawing, purchasing specification, and applicable standard.
chain pitch inspection, chain vision inspection, machine vision chain measurement, roller chain inspection, transmission chain, chain length measurement, chain elongation, chain dimensional inspection, sprocket engagement, industrial camera inspection, machine vision measurement, roller chain pitch accuracy, chain length accuracy, ISO 606.
[1] HaoRong Short-Pitch Roller Chain Product Page: Models, Product Types, Standard Chains, and Customized Chains
[2] ISO 606:2015 Official Browsing Platform: Short-Pitch Precision Roller and Bush Chains and Associated Sprockets
[3] iwis Chain Length Accuracy: Chain Length Tolerance, Preloaded Measurement, and Accumulated-Length Analysis
Technical note: This article was prepared from HaoRong’s publicly available web information, ISO 606-related information, and general machine-vision knowledge in the power-transmission industry. It is intended for technical education and YouTube content planning. Final tolerances, measuring force, sample quantity, and acceptance criteria for vision inspection must be determined according to the specific product standard, drawing, customer specification, and approved inspection procedure. Vision inspection cannot replace complete-chain tensile-strength testing, dynamic-strength testing, fatigue-life testing, or actual operational trials.
