Research conducted in recent years in the area of metrology is increasingly based on the use of non-contact measurement methods [
1,
2]. Gears occupy a very important area in the machine industry. The quality of the produced mechanical gears and their appropriate service life depend on their performance. Research on the implementation of non-contact measurements using optics is conducted mainly for applications in the aviation industry [
3,
4] and automotive industry [
5]. For many years, work has also been underway on improving non-contact measurement techniques by developing new methods of using the light spectrum in order to increase the accuracy of measuring devices [
6,
7]. The proposals of conceptual measurement methods are widely known, which consist in using a source of structured light directed at the measured gear and performing scanning using cameras [
8,
9,
10]. The light falling on the measured gear undergoes deformation, which is observed by cameras. Then, the obtained image is analyzed by a computer program, which, by combining scans taken from many directions, calculates a 3D surface model of the measured gear. A commonly used scanner that uses structured light is the GOM ATOS II scanner. As a result of the measurement, it is possible to reproduce the shape of the measured gear wheel into an editable digital model. Measuring elementary deviations of the gear wheel requires generating a digital model from a point cloud (counted in millions), which is then subjected to appropriate processing using CAD software. An example of this type of software dedicated to the GOM ATOS optical scanner is GOM Inspekt. It provides the ability to compare the measured model of the actual gear wheel with the nominal model as a result of transforming two models into a common coordinate system. Then, as a result of using model matching tools, based on the construction databases of a given gear wheel, the effect of a visual map is obtained presenting the exceedance of nominal dimensions in the form of a color scale. However, this operation does not provide numerical values of elementary deviations of the profile, tooth line and pitches, and run-out; therefore further geometric processing of the model system is required. In the work presented in [
11], a method for measuring the tooth profile based on the use of an additional cross-section plane supporting the measurement of profile parameters is graphically presented. Then, in the separated cross-section plane, a tooth outline is generated, which includes the adjusted outline from the measured (real) model and from the nominal model (obtained from the CAD program). The appropriately adjusted outline on the left and right side of the tooth allows for determining the total deviation of the outline on the designated measuring section of the tooth profile. The researchers also presented a method for measuring the total deviation of the tooth line in a similar (graphic) way. The additional plane allows for the isolation of the measured section, which is subject to assessment in accordance with the measurement guidelines included in the adopted standards for gear wheels, e.g., DIN 3961/62. In the work presented in [
12], there is a non-contact method for measuring gears using optics to determine the accuracy of gears obtained as a result of production using a rapid prototyping method. This publication discusses the possibilities of using optical measurements to determine the geometric accuracy of gear wheel castings produced in the rapid prototyping process. The tested gear wheel prototype was made and cast from an aluminum alloy. Coordinate optical measurement methods and a GOM scanner were used to test the accuracy of the geometry of gears produced by casting. The obtained measurement results are highly reliable, because the accuracy of castings in the rapid prototyping method is within the limits of the scanner’s measurement capabilities. A similar method was described in [
13], for measuring the performance properties of polymer gears using coordinate measurement methods. The measurements were performed using the ATOS II Triple Scan optical system. The main imperfection of the measurement method using an optical scanner is obtaining a discontinuous measurement surface. This problem was described in [
14], where a method for reconstructing the actual surface from a point cloud was proposed. The researchers used 3D modeling to align the point cloud, which is processed to recreate actual tooth surfaces. A very significant aspect in industrial metrology is obtaining the appropriate measurement accuracy, as well as determining the measurement uncertainty estimate [
15]. Publication [
16] presents the results of accuracy measurements using a laser triangulation method. The article examines the potential of triangulation and confocal–chromatic sensors for measuring gears. The sources of deviations, such as the angle of inclination between the nominal to the tooth surface and to the sensor axis, the variable surface curvature and the topography of the gear surface, were analyzed. Measurements were carried out on the side surface of a straight-toothed gear wheel and it was shown that optical sensors have the potential to measure the shape of gears, especially confocal–chromatic sensors, which can achieve a measurement uncertainty of less than 10 µm. The geometric shape of gears is another issue for which an appropriate measurement method should be selected. Publication [
17] presents the possibility of using the ATOS II optical scanner for measuring aircraft bevel gears. The research presented in the article was carried out based on a non-contact method of measuring bevel gears using a 3D optical scanner for preliminary and quick verification of correctness of execution. Deviations of all individual measurement points are calculated in relation to the nominal value of the geometry. Due to their number, deviations are visualized in the form of a color map. Such an image shows the critical points of the measured gear, which should be carefully analyzed using other—more accurate—methods. Study [
18] presents the advantages of using a vision system in metrology, by using a vision-based gear profile measurement system. Thanks to the integration of the camera system with the measuring equipment, accurate registration and subsequent analysis of measurement results are possible. This system has the ability to record videos and save image frames in the JPEG format during the measurement, which allows their later opening and analysis in offline mode. The vision-based inspection system presented in that paper was designed mainly for measuring surface errors of various types of gear wheels. A lot of information about the geometric structure of the gear tooth surface can be obtained by analyzing the surface topography. In paper [
19], an experimental optical approach to assessing the deflection of the gear tooth during meshing is presented, which is crucial for understanding the wear and fatigue resistance of gears made of polymers. The features depend on factors such as working load, speed, temperature and lubrication. The proposed approach is an alternative to numerical analyses, such as the finite element method (FEM), and uses image recording from high-resolution cameras and image processing methods.
Scientific publications also include other methods of measuring gear tooth profiles using incoherently structured light. Publication [
20] presents the use of the incoherent linearly structured light method for precise measurement of gear tooth profiles. Inconsistent light, unlike coherent light used in lasers, helps reduce speckle noise, which is a common problem in laser measurements. As a result of the conducted experimental studies, it was shown that the incoherent linearly structured light method provides higher measurement resolution and is less susceptible to speckle noise, compared to traditional laser methods. The tooth profile error for the involute standard measured using the incoherent linearly structured light method was ±2.2 μm. The obtained level of accuracy of the method allows its industrial application, where precise measurements of gear tooth profiles are required. Publication [
21] presents 3D measurements of gears using a linear laser enabling quick acquisition of full 3D data of the tooth surface. That method, called LL3DMG, allows for the representation of the complex 3D topography of the tooth surface, including the size and modification of the gears, and compensates for the limitations of traditional measurement techniques that rely on a limited number of points on the tooth surface. In publication [
22], the focus was on 3D measurement of gears based on linear light sensors. Measured 3D point cloud data were used to calculate the profile error and then compared with the results obtained from traditional contact measurements obtained using a Klingelnberg P26. The results obtained proved the agreement between measurements with a structured light sensor and reference measurements, which allows to conclude that, using a 3D point cloud measurement system, it is possible to perform fast and accurate measurements of gears, which is an innovative system for measuring involute for specified accuracy of gears.
This article presents the results of research conducted on the measurement of gear wheel profiles using the non-contact optical method and the contact method. The article is organized as follows: after the Introduction, in
Section 2, the gear wheels are characterized and the measurement methods used in the research are described. Then, in
Section 3, the measurement results are presented. In
Section 4, the obtained results are commented on. In
Section 5, the main conclusions from the research are presented.