A system and method for detecting roundness error of a shaft

Abstract

The application belongs to the technical field of size detection, and discloses a shaft roundness error detection system and method, which comprises: an air floating support seat for supporting the lower end surface of a shaft to be detected; a rotating driving mechanism for driving the shaft to be detected to rotate around its own axis; a positioning mechanism for transversely positioning the shaft to be detected; a measuring system comprising a plurality of measuring units arranged at intervals in the up-down direction, wherein each measuring unit comprises an interferometer and a measuring head; the measuring head is used for contacting the circumferential surface of the shaft to be detected, so as to produce displacement with the profile change of the shaft to be detected; the interferometer is used for converting the displacement of the measuring head into an interference fringe image and outputting; and a host computer is used for calculating the roundness error of the shaft to be detected at a plurality of different cross sections according to the interference fringe image, so as to calculate the overall roundness error of the shaft to be detected; thus, a feasible technical scheme is provided for accurate and efficient detection of large shafts.

Description

Technical Field

[0001] This application relates to the field of dimensional inspection technology, and more specifically, to a system and method for detecting the roundness error of shaft parts. Background Technology

[0002] Currently, the measurement of roundness error of shaft components mainly relies on high-precision roundness meters. Although roundness meters can provide extremely high sampling and evaluation accuracy, their high cost and equipment size limitations make them primarily suitable for measuring the roundness error of small and medium-sized parts. This results in the limitations of existing roundness meters being particularly pronounced when dealing with large, heavy shaft components, especially those requiring sub-micron level roundness error accuracy.

[0003] Specifically, for special shaft components such as those with large diameters and long shafts, the systematic error of the roundness measuring instrument has a significantly increased impact on measurement accuracy, and its limited measurement space cannot accommodate such large workpieces. This results in a lack of a universally applicable method for detecting the roundness error of large shaft components that can meet high-precision requirements in actual production and inspection. Existing technologies are insufficient to effectively solve the problem of roundness error measurement for large-size, high-precision shaft components, especially under sub-micron level accuracy requirements. How to achieve accurate and efficient inspection of special shaft components such as those with large diameters and long shafts has become an urgent technical challenge to be solved.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] The purpose of this application is to provide a system and method for detecting the roundness error of shaft components, which aims to solve the problem that there is a lack of universally applicable and high-precision roundness error detection methods for large and heavy shaft components, especially shaft components that require sub-micron level roundness error accuracy.

[0006] In a first aspect, this application provides a roundness error detection system for shaft components, comprising:

[0007] Air-bearing support base, used to support the lower end face of the shaft component to be tested;

[0008] A rotation drive mechanism is disposed above the air-bearing support base and is used to drive the shaft under test to rotate around its own axis from the upper end of the shaft under test.

[0009] The positioning mechanism includes a V-shaped positioning component and a pre-tightening component arranged opposite to each other. The V-shaped positioning component has a V-shaped positioning groove, and the pre-tightening component is used to provide pre-tightening pressure to the shaft component to be tested, so that the circumferential surface of the shaft component to be tested fits against the two positioning groove surfaces of the V-shaped positioning groove.

[0010] The measurement system includes multiple measurement units arranged at intervals along the vertical direction. Each measurement unit includes an interferometer and a probe that is linked to a movable mirror of the interferometer. The probe is used to contact the circumferential surface of the shaft-like component under test to generate displacement as the contour of the shaft-like component under test changes. The interferometer is used to convert the displacement of the probe into an interference fringe image and output it.

[0011] The host computer is used to calculate the roundness error of the shaft under test at multiple different cross sections based on the interference fringe images output by each of the measurement units, and to calculate the overall roundness error of the shaft under test by combining the roundness errors of each cross section.

[0012] Secondly, this application provides a method for detecting the roundness error of shaft parts, based on the shaft part roundness error detection system described above, including the following steps:

[0013] A1. After installing the shaft to be tested in a vertical position on the shaft to be tested roundness error detection system, drive the shaft to be tested to rotate one revolution, and collect interference fringe images at multiple different cross sections of the shaft to be tested in real time.

[0014] A2. Preprocess the interference fringe image; the preprocessing includes grayscale conversion, dynamic noise reduction, and fringe sharpening.

[0015] A3. Extract the pixel positions of the zeroth-order and secondary fringes in each preprocessed interference fringe image;

[0016] A4. For each cross section, based on the pixel positions of the zero-order fringe and the secondary fringe, calculate the zero-order fringe pixel displacement and the secondary fringe pixel displacement corresponding to each two adjacent frames of interference fringe images, and verify and correct the zero-order fringe pixel displacement based on the secondary fringe pixel displacement.

[0017] A5. For each cross section, calculate the actual displacement of the corresponding probe based on the zero-order fringe pixel displacement after each verification and correction, and obtain the actual displacement sequence of each cross section;

[0018] A6. Based on the actual displacement sequence, combined with the actual apex angle of the V-shaped positioning groove and the preset error transmission matrix corresponding to the actual apex angle, the roundness error of each cross section is calculated using the three-point method and the minimum area method.

[0019] A7. Calculate the overall roundness error of the shaft component under test by combining the roundness errors of each cross section.

[0020] Beneficial Effects: This application provides a roundness error detection system and method for shaft components. By employing an air-bearing support to reduce rotational friction, a rotation drive mechanism to drive the shaft component to rotate from the top, a V-shaped positioning component and a pre-tightening component to achieve light-contact positioning, and an interferometric measurement system with multiple measurement units, it effectively solves the problem of roundness error detection for large and heavy shaft components in the prior art. The system converts the probe displacement into an interference fringe image using an interferometer, and the roundness error is calculated by a host computer, achieving sub-micron level high-precision roundness error detection for shaft components at multiple different cross-sections. Furthermore, the system can comprehensively calculate the overall roundness error by integrating the roundness errors of each cross-section, overcoming the limitations of traditional roundness meters when dealing with large-size, high-precision shaft components, and providing a feasible technical solution for the accurate and efficient detection of large shaft components. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the structure of a roundness error detection system for shaft parts provided in this application.

[0022] Figure 2 A top view of the V-shaped positioning component.

[0023] Figure 3 This is a schematic diagram of the pre-tightening assembly.

[0024] Figure 4 This is a top view of a pure couple shift fork transmission mechanism.

[0025] Figure 5 This is a schematic diagram of the measurement unit.

[0026] Figure 6 A flowchart of a method for detecting roundness error of shaft parts provided in this application.

[0027] Labeling Explanation: 1. Air-bearing support; 2. Rotation drive mechanism; 201. Motor; 202. Reducer; 203. Torque sensor; 204. Pure torque shift fork transmission mechanism; 2041. Rotating component; 2042. Shift fork; 2043. Pulley; 2044. Belt; 3. V-shaped positioning assembly; 301. V-shaped positioning groove; 302. Positioning groove surface; 303. Fixed seat; 304. Swing block; 305. Swing drive device; 4. Pre-tightening assembly; 401. Abutting component; 402. Pre-tightening spring; 403. Pressing component; 404. Pressure sensor; 405. Adjusting component; 406. Fixed screw; 5. Measuring unit; 501. Interferometer; 5011. Light source; 5012. Image sensor; 5013. Beam splitter; 5014. Fixed reflector; 5015. Movable reflector; 502. Probe; 6. Host computer; 90. Shaft to be measured. Detailed Implementation

[0028] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0029] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0030] Please refer to Figures 1-5 A roundness error detection system for shaft components, as described in some embodiments of this application, includes:

[0031] Air-bearing support 1 is used to support the lower end face of the shaft component 90 to be tested;

[0032] Rotation drive mechanism 2 is disposed above the air flotation support 1 and is used to drive the shaft component 90 to be tested to rotate around its own axis from the upper end of the shaft component 90 to be tested.

[0033] The positioning mechanism includes a V-shaped positioning component 3 and a pre-tightening component 4 arranged opposite to each other. The V-shaped positioning component 3 has a V-shaped positioning groove 301. The pre-tightening component 4 is used to provide pre-tightening pressure to the shaft component 90 to be tested, so that the circumferential surface of the shaft component 90 to be tested fits against the two positioning groove surfaces 302 of the V-shaped positioning groove 301.

[0034] The measurement system includes multiple measurement units 5 arranged at intervals along the vertical direction. Each measurement unit 5 includes an interferometer 501 and a probe 502 linked to a movable reflector 5015 of the interferometer 501. The probe 502 is used to contact the circumferential surface of the shaft-like component 90 to be measured, so as to generate displacement as the contour of the shaft-like component 90 changes. The interferometer 501 is used to convert the displacement of the probe 502 into an interference fringe image and output it.

[0035] The host computer 6 is used to calculate the roundness error of the shaft component 90 under test at multiple different cross sections based on the interference fringe images output by each of the measurement units 5, and to calculate the overall roundness error of the shaft component 90 under test by combining the roundness errors of each cross section.

[0036] The core of the shaft roundness error detection system of this application lies in achieving high-precision, non-contact or light-contact roundness error measurement of large shafts through the coordinated work of various components.

[0037] Specifically, the air-bearing support 1 can be implemented in various ways. For example, a porous air-bearing bearing can be used, which has uniformly distributed micropores inside. When high-pressure gas is ejected through these micropores, a uniform gas film is formed between the support surface and the shaft component 90 to be measured, thereby providing stable air-bearing support. Another implementation method is to use a throttling orifice type air-bearing bearing, which controls the gas flow rate and pressure by setting multiple throttling orifices on the support surface to achieve the air-bearing effect. The air-bearing support 1 can significantly reduce the friction of the shaft component 90 to be measured during rotation, ensuring its smooth and free rotation, laying the foundation for subsequent accurate measurements.

[0038] The rotation drive mechanism 2 can take various forms. For example, it can be directly driven by a motor 201, with the output shaft of the motor 201 connected to the upper end of the shaft component 90 to be measured via a coupling. Alternatively, it can be driven by a belt or gear, with a reduction mechanism adjusting the speed and torque of the motor 201 to accommodate the rotation requirements of shaft components 90 of different sizes and weights. The driving method of the rotation drive mechanism 2 should ensure good coaxiality and stability of the shaft component 90 during rotation, avoiding the introduction of additional measurement errors.

[0039] In the positioning mechanism, the V-shaped positioning component 3 can be composed of a single, integrally machined V-shaped block with a fixed angle for its V-groove. Alternatively, it can consist of two independent positioning blocks, with the V-shaped positioning groove 301 formed by adjusting the distance and angle between these two blocks. The pre-tightening component 4 can be spring-loaded, for example, by using an adjustable spring mechanism to press the abutment 401 against the circumferential surface of the shaft component 90 to be measured, thereby providing pre-tightening pressure. The pre-tightening component 4 can also be driven by a pneumatic or hydraulic cylinder, applying pre-tightening force by precisely controlling the pneumatic or hydraulic pressure. The function of the positioning mechanism is to provide stable three-point positioning for the shaft component 90 to be measured, ensuring the accuracy of its position during measurement and eliminating minor displacements caused by the shaft component's own weight or driving force.

[0040] In the measurement system, the interferometer 501 converts the displacement of the probe 502 into an interference fringe image and outputs it. The probe 502 can be a contact probe 502, which contacts the surface of the shaft component 90 to be measured. Its displacement is transmitted to the movable mirror 5015 of the interferometer 501 through a mechanical structure. The interferometer 501 can be an Ubbel interferometer, Michelson interferometer, or Fabry-Perot interferometer. The light beam emitted by the light source 5011 interferes after passing through the beam splitter 5013, the fixed mirror 5014, and the movable mirror 5015. The interference fringe image is acquired and output by the image sensor 5012. Multiple measurement units 5 are arranged at intervals along the vertical direction, which can realize synchronous or quasi-synchronous measurement of the roundness error at different cross-sections of the shaft component, improving measurement efficiency and comprehensiveness.

[0041] The host computer 6 can run specially developed measurement software, which integrates functions such as image processing, interference fringe analysis, displacement calculation, and roundness error calculation. For example, the host computer 6 can process the interference fringe image using a Fourier transform-based method to extract fringe phase information and then calculate the displacement of the probe 502. Regarding roundness error calculation, the host computer 6 can employ various evaluation methods such as the three-point method, the least squares method, and the minimum region method, selecting the appropriate algorithm based on actual needs. The host computer 6 can also perform weighted averaging or statistical analysis of the roundness errors of multiple cross-sections to obtain the overall roundness error of the shaft component 90 under test and generate a detailed inspection report.

[0042] The shaft roundness error detection system of this application forms a complete detection process through the coordinated operation of the above-mentioned components. First, the shaft 90 to be tested is placed on the air-bearing support 1, and its lower end face is stably supported. The V-shaped positioning component 3 and the pre-tightening component 4 of the positioning mechanism perform three-point positioning of the shaft to ensure its positional stability during the measurement process. The rotation drive mechanism 2 drives the shaft to rotate uniformly around its own axis from the upper end. During the rotation of the shaft, the probes 502 of multiple measuring units 5 contact (or perform non-contact measurement) with the circumferential surface of the shaft, and generate displacement as the contour of the shaft changes. These displacements are converted into interference fringe images by the interferometer 501 and transmitted to the host computer 6 in real time. After receiving the interference fringe images of each measuring unit 5, the host computer 6 performs image processing, displacement calculation, and roundness error calculation. The specific process can be referred to the steps of the shaft roundness error detection method described later. The entire process realizes automated and efficient detection of high-precision roundness errors of large shafts.

[0043] Compared to traditional roundness testers, the shaft roundness error detection system of this application has significant advantages and innovations. Traditional roundness testers are limited by their structure and measurement range, making it difficult to perform high-precision measurements on large and heavy shafts. This application, by employing an air-bearing support 1, effectively reduces the frictional resistance of large shafts during rotation, enabling them to rotate smoothly and providing a foundation for high-precision measurement. The positioning mechanism uses a V-shaped positioning component 3 and a pre-tightening component 4 to achieve stable three-point positioning of the shaft, avoiding deformation and errors that may be introduced by traditional clamping methods. The measurement system employs multiple measurement units 5, combined with the linkage between the probe 502 and the movable reflector 5015, to achieve non-contact or light-contact high-precision displacement measurement of changes in the circumferential contour of the shaft, with a measurement accuracy far exceeding that of traditional contact sensors. Furthermore, the host computer 6, through advanced image processing and roundness error calculation algorithms, can accurately calculate the roundness error of each cross-section and synthesize the overall roundness error. This modular and high-precision design enables the system to adapt to the inspection needs of large shaft components with different sizes and precision requirements, effectively solving the problem of high-precision roundness error measurement of large shaft components in existing technologies, and has significant technological progress and practical value.

[0044] In some preferred embodiments, see Figure 2 The V-shaped positioning component 3 includes a fixed base 303, two swing blocks 304, and two swing drive devices 305. The two swing blocks 304 are swingably disposed on the side of the fixed base 303 near the pre-tightening component 4. Each side of the two swing blocks 304 that is close to each other is provided with a positioning groove surface 302, and the V-shaped positioning groove 301 is formed between the two positioning groove surfaces 302. The two swing drive devices 305 are respectively used to drive the two swing blocks 304 to swing, so as to adjust the included angle between the two positioning groove surfaces 302.

[0045] The fixed base 303 serves as the base of the V-shaped positioning assembly 3, providing mounting and support for other components. Two swing blocks 304 are designed to swing around their respective swing axes on the fixed base 303, with the swing axes typically perpendicular to the axis of the shaft component 90 to be tested. Each swing block 304 has a positioning groove 302, which are positioned opposite each other to form a V-shaped positioning groove 301 for supporting the shaft component 90 to be tested. The swing drive device 305 is used to precisely control the swing angle of the two swing blocks 304, thereby adjusting the angle between the two positioning grooves 302. For example, the swing drive device 305 can be a miniature electric actuator, with its two ends hinged to the fixed base 303 and the swing blocks 304 respectively. By controlling the extension and retraction of the miniature electric actuator, the swing blocks 304 can be driven to swing, thereby changing the angle between the positioning grooves 302. In a preferred embodiment, the included angle between the positioning groove surfaces 302 can be adjusted within the range of 30° to 90° to accommodate shafts of different diameters and shapes.

[0046] The solution in this application introduces a swingable swing block 304 and a swing drive device 305, allowing the included angle of the V-shaped positioning groove 301 to be flexibly adjusted according to actual needs. When it is necessary to inspect shafts of different diameters or with specific geometric features, the swing drive device 305 can drive the two swing blocks 304 to swing, thereby changing the included angle between the two positioning groove surfaces 302. This adjustability ensures that the circumferential surface of the shaft 90 under test can always be in close contact with the two positioning groove surfaces 302 of the V-shaped positioning groove 301, achieving stable positioning even when the diameter of the shaft varies greatly. By precisely adjusting the included angle, positioning accuracy can be optimized, measurement errors caused by improper positioning can be reduced, thereby ensuring the accuracy of the measurement results.

[0047] Preferably, the two positioning groove surfaces 302 of the V-shaped positioning groove 301 may be coated with polytetrafluoroethylene.

[0048] Specifically, the polytetrafluoroethylene (PTFE) coating is a polymeric material coating with excellent low coefficient of friction, wear resistance, chemical stability, and non-adhesion. Its purpose is to form a protective lubrication interface between the circumferential surface of the shaft component 90 to be tested and the positioning groove surface 302 of the V-shaped positioning groove 301, thereby significantly reducing the friction between them. In practical applications, this coating can be applied to the positioning groove surface 302 through various methods such as spraying, dipping, or sintering, ensuring a uniform and firm coating.

[0049] The solution of this application involves applying a polytetrafluoroethylene (PTFE) coating to the positioning groove surface 302 of the V-shaped positioning groove 301. This significantly reduces the contact friction between the circumferential surface of the shaft component 90 and the positioning groove surface 302 during rotation. Due to the inherent low-friction properties of PTFE, the resistance of the shaft component 90 during rotation within the positioning groove is effectively reduced, ensuring that the shaft component 90 rotates at a smoother and more uniform speed. Furthermore, the coating prevents scratches or wear on the surface of the shaft component 90 due to friction, protecting the integrity of the workpiece.

[0050] In some implementations, see Figure 3 The pre-tightening assembly 4 includes an abutment 401, a pre-tightening spring 402, a clamping member 403, a pressure sensor 404, an adjusting member 405, and a fixing screw 406. The clamping member 403 is slidably sleeved on the end of the fixing screw 406 near the V-shaped positioning assembly 3. The adjusting member 405 is disposed on the side of the clamping member 403 away from the V-shaped positioning assembly 3 and is threadedly connected to the fixing screw 406. The pressure sensor 404 is disposed between the clamping member 403 and the adjusting member 405. The abutment 401 is disposed on the side of the clamping member 403 near the V-shaped positioning assembly 3. The pre-tightening spring 402 is connected between the clamping member 403 and the abutment 401. The pre-tightening spring 402 is used to provide elastic force to press the abutment 401 against the circumferential surface of the shaft-like part 90 to be tested.

[0051] Specifically, the abutment 401 is used to directly contact the circumferential surface of the shaft component 90 to be tested. Its contact surface is preferably spherical to reduce the contact area, lower contact stress, and accommodate any minor unevenness that may exist on the surface of the shaft component 90. A preload spring 402 is positioned between the clamping member 403 and the abutment 401, providing an adjustable elastic force to stably press the abutment 401 against the circumferential surface of the shaft component 90. The clamping member 403 slides on the fixing screw 406, serving as a support and force transmission component for the preload spring 402 and the abutment 401. An adjusting member 405 is threadedly connected to the fixing screw 406. By rotating the adjusting member 405, the position of the clamping member 403 on the fixing screw 406 can be changed, thereby compressing or releasing the preload spring 402, and thus adjusting the preload pressure applied by the abutment 401 to the shaft component 90 to be tested. A pressure sensor 404 is positioned between the clamping member 403 and the adjusting member 405 to monitor the preload pressure generated by the preload spring 402 in real time and feed the pressure data back to the host computer 6 or other control units. The fixing screw 406 provides structural support and guidance for the entire preload assembly 4, ensuring stable and reliable sliding of the clamping member 403 and threaded connection of the adjusting member 405.

[0052] This application's solution introduces components such as a pressure sensor 404, an adjusting component 405, and a preload spring 402, making the application and monitoring of preload pressure a controllable process. Specifically, when the shaft component 90 to be tested is placed in the V-shaped positioning groove 301, rotating the adjusting component 405 drives the clamping component 403 to move along the fixed screw 406, thereby compressing the preload spring 402 and causing the abutment component 401 to press against the circumferential surface of the shaft component 90 with a certain elastic force. During this process, the pressure sensor 404 collects preload pressure data in real time and transmits it to the host computer 6. The host computer 6 determines whether the current preload pressure is within the ideal range based on the preset "light contact positioning" requirement. If the pressure is too high or too low, the operator can fine-tune the adjusting component 405 according to the feedback from the host computer 6 until the preload pressure reaches the optimal state. Thus, it ensures that the shaft component 90 to be tested is stably positioned while avoiding deformation caused by excessive pressure and preventing unstable positioning caused by insufficient pressure.

[0053] In some implementations, see Figure 1 The rotation drive mechanism 2 includes a motor 201, a reducer 202, a torque sensor 203, and a pure torque shift fork transmission mechanism 204, which are connected in sequence from top to bottom.

[0054] The motor 201 provides the initial rotational power to drive the entire mechanism. The reducer 202 is connected below the motor 201; its function is to reduce the output speed of the motor 201 while increasing the output torque to meet the stable rotation requirements of the shaft component 90 under low speed and high torque. The torque sensor 203 is located below the reducer 202 and is used to monitor and provide feedback on the torque output by the drive mechanism in real time, thereby ensuring that the applied driving force is a pure torque and avoiding the introduction of additional radial force or bending moment.

[0055] A pure torque shift fork drive mechanism 204 is positioned below the torque sensor 203, for connecting to the top of the shaft component 90 to be measured, and for applying a pure torque drive to it. (Reference) Figure 1 , Figure 4 The pure torque shift fork transmission mechanism 204 specifically includes a rotating member 2041 for connecting to the top of the shaft 90 to be tested, and a shift fork 2042 disposed above the rotating member 2041. Two rotatable pulleys 2043 are arranged side-by-side at the upper end of the rotating member 2041, and a belt 2044 is sleeved between the two pulleys 2043. The two forks of the shift fork 2042 respectively abut against the outer surfaces of two belt segments located between the two pulleys 2043 on the belt 2044. Driven by the motor 201, the shift fork 2042 rotates and applies a pair of forces—one in opposite direction, equal in magnitude, and with non-coincident lines of action—to the belt 2044 through its two forks, thereby driving the shaft 90 to rotate stably around its own axis.

[0056] The solution in this application achieves stable and pure rotational drive for shaft components through the aforementioned rotational drive mechanism 2. Specifically, the motor 201 provides power, which is reduced and amplified by the reducer 202, enabling the shaft component 90 under test to rotate stably at the required low speed. The torque sensor 203 monitors the torque output by the drive mechanism in real time, and its data can be used for feedback control to ensure that the driving force is always a pure torque, thereby effectively avoiding interference from radial force or bending moment caused by impure driving force on the shaft component. The pure torque shift fork transmission mechanism 204, through its unique structure, that is, the two forks of the shift fork 2042 apply a pair of torques to the belt 2044, converting the driving force into pure rotational torque, which acts directly on the shaft component 90 under test, thereby minimizing non-axial disturbances that may be introduced during the driving process. This design ensures that the shaft component 90 under test is only subjected to pure rotational torque during rotation, and its axial position and orientation remain highly stable, providing ideal reference conditions for subsequent interferometric measurements.

[0057] In some preferred embodiments, the motor 201 can be a stepper motor to achieve precise speed control and positioning. The reducer 202 is preferably a harmonic reducer, which features a high reduction ratio, high transmission accuracy, and zero backlash, better meeting the needs of precision measurement. For example, the reduction ratio of the reducer 202 can be selected in the range of 100:1 to 500:1 to adapt to the common speed requirements of the workpiece under test, ranging from 0.5 r / min to 5 r / min. Furthermore, the torque sensor 203 can monitor the driving torque in real time and feed the monitoring data back to the control system to ensure stable output of the pure torque, further improving the accuracy and stability of the drive.

[0058] In some implementations, see Figure 5 The interferometer 501 includes a light source 5011, an interference optical path assembly, and an image sensor 5012. The interference optical path assembly includes a beam splitter 5013, a fixed reflector 5014, and a movable reflector 5015.

[0059] In this scheme, the light source 5011 provides a stable beam, which is the basis for interferometric measurement. The beam splitter 5013 in the interferometric optical path assembly splits the beam from the light source 5011 into two beams: one beam is directed towards the fixed reflector 5014, forming the reference optical path; the other beam is directed towards the movable reflector 5015, forming the measurement optical path. The movable reflector 5015 is linked to the probe 502. When the probe 502 moves due to changes in the contour of the shaft, the movable reflector 5015 also moves accordingly, thereby changing the optical path of the measurement optical path. The fixed reflector 5014 maintains a constant optical path, providing a stable reference beam. The two beams reconverge at the beam splitter 5013, and interference fringes are generated due to the optical path difference. The image sensor 5012 captures and outputs these interference fringe images. By analyzing these interference fringe images, the displacement of the probe 502 can be accurately calculated, and the roundness error of the shaft can be derived. This design ensures that the interferometer 501 can stably and accurately convert the minute displacement of the probe 502 into a quantifiable interference fringe image, providing high-precision raw data for subsequent roundness error calculation, thus effectively solving the problem of performance limitations caused by the unclear specific structure of the interferometer 501.

[0060] The image sensor 5012 can be a CMOS image sensor. A CMOS image sensor can convert light signals into electrical signals, thereby forming a digital image.

[0061] In some preferred embodiments, the light source 5011 includes two switchable lasers, namely a helium-neon laser and a semiconductor laser.

[0062] Specifically, the light source 5011 is configured to include two different types of lasers: a helium-neon laser and a semiconductor laser, which can be switched according to actual detection needs. The helium-neon laser typically features wavelength stability and good coherence, providing the light source 5011 required for high-precision interferometric measurements. The semiconductor laser, on the other hand, offers advantages such as small size, fast response speed, and relatively low cost, making it suitable for scenarios with high detection efficiency requirements. By providing two switchable lasers, the system can flexibly select the appropriate light source 5011 for different measurement scenarios, balancing measurement accuracy and detection efficiency.

[0063] This application's solution effectively addresses the limitation of a single light source 5011 in meeting varying accuracy and efficiency requirements by introducing two switchable lasers—a helium-neon laser and a semiconductor laser—into the light source 5011 of the interferometer 501. When high-precision (e.g., sub-micron level measurement accuracy) shaft component inspection is required, the system can switch to the helium-neon laser, utilizing its excellent coherence and wavelength stability to ensure the clarity of the interference fringe image and the accuracy of the measurement results. Conversely, in scenarios with relatively lower measurement accuracy requirements but higher inspection efficiency, such as batch inspection, the system can switch to the semiconductor laser, leveraging its fast response and high repeatability to significantly improve inspection speed. This switchable light source 5011 configuration allows the system to flexibly adjust the type of light source 5011 according to the characteristics of the shaft component 90 under test and the specific requirements of the inspection task, thereby optimizing overall inspection performance.

[0064] refer to Figure 6 This application provides a method for detecting the roundness error of shaft parts, based on the shaft part roundness error detection system described above, including the following steps:

[0065] A1. After installing the shaft component 90 to be tested in a vertical position on the shaft component roundness error detection system, drive the shaft component 90 to be tested to rotate one revolution, and collect interference fringe images at multiple different cross-sections of the shaft component 90 to be tested in real time.

[0066] A2. Preprocess the interference fringe image; the preprocessing includes grayscale conversion, dynamic noise reduction, and fringe sharpening.

[0067] A3. Extract the pixel positions of the zeroth-order and secondary fringes in each preprocessed interference fringe image;

[0068] A4. For each cross section, based on the pixel positions of the zero-order fringe and the secondary fringe, calculate the zero-order fringe pixel displacement and the secondary fringe pixel displacement corresponding to each two adjacent frames of interference fringe images, and verify and correct the zero-order fringe pixel displacement based on the secondary fringe pixel displacement.

[0069] A5. For each cross section, calculate the actual displacement of the corresponding probe 502 based on the zero-order fringe pixel displacement after each verification and correction, and obtain the actual displacement sequence of each cross section;

[0070] A6. Based on the actual displacement sequence, combined with the actual groove apex angle of the V-shaped positioning groove 301 (i.e., the included angle between the two positioning groove surfaces 302) and the preset error transmission matrix corresponding to the actual groove apex angle, the roundness error of each cross section is calculated using the three-point method and the minimum area method.

[0071] A7. Calculate the overall roundness error of the shaft component 90 under test by taking into account the roundness errors of each cross section.

[0072] Specifically, in step A1, the shaft component 90 to be tested is mounted vertically in the aforementioned shaft component roundness error detection system (e.g., Figure 1 (As shown). The system utilizes an air-bearing support 1 to support the lower end face of the shaft component 90 under test, and drives the shaft component 90 to rotate around its own axis from the upper end via a rotation drive mechanism 2. A positioning mechanism ensures the stable positioning of the shaft component 90 under test during rotation. Multiple measurement units 5 in the measurement system convert the displacement of their respective probes 502 into interference fringe images in real time and output them, thereby obtaining interference fringe images at multiple different cross-sections.

[0073] In step A2, the acquired interference fringe image is preprocessed. This preprocessing aims to optimize image quality, providing a clear and accurate data foundation for subsequent fringe feature extraction. Specifically, the preprocessing includes grayscale conversion, which converts the color image to grayscale to simplify the image data and highlight brightness information; dynamic noise reduction, used to eliminate random noise and interference in the image and improve the clarity of the fringes; and fringe sharpening, to enhance the edges and details of the interference fringes, making them easier to identify and extract.

[0074] In step A3, the pixel positions of the zero-order and secondary fringes are extracted from the preprocessed interference fringe image. The zero-order fringe is usually the brightest fringe in the interference image, representing the position where the optical path difference is zero or an integer multiple of the wavelength. The secondary fringes are the adjacent fringes on both sides of the zero-order fringe. Accurately extracting the pixel positions of these fringes is crucial for the subsequent calculation of the displacement of the probe 502.

[0075] In step A4, for each cross-section, the zero-order fringe pixel displacement and secondary fringe pixel displacement corresponding to every two adjacent frames of interference fringe images are calculated. During the rotation of the shaft component 90 under test, the probe 502 will experience displacement due to changes in its contour; this displacement is reflected in the interference fringe image as fringe movement. By comparing the fringe pixel positions of adjacent frames, the fringe pixel displacement can be obtained. Furthermore, verifying and correcting the zero-order fringe pixel displacement based on the secondary fringe pixel displacement can effectively improve the accuracy and robustness of the displacement calculation, reducing errors caused by noise or abnormal conditions.

[0076] In step A5, for each cross-section, the actual displacement of the corresponding probe 502 is calculated based on the zero-order fringe pixel displacement after verification and correction, thus obtaining the actual displacement sequence for each cross-section. The actual displacement of the probe 502 is a direct quantity reflecting the change in the surface contour of the shaft-like part 90 under test. For example, the actual displacement of the probe 502 can be calculated using the formula S'=(L / T)*S, where S' is the actual displacement of the probe 502, L is the zero-order fringe pixel displacement after verification and correction, T is the preset fringe period, which is equal to the pixel size corresponding to the wavelength of the laser beam of the light source 5011, and S is the pre-calibration coefficient.

[0077] In step A6, based on the actual displacement sequence, combined with the actual apex angle of the V-shaped positioning groove 301 and the preset error transfer matrix corresponding to the actual apex angle, the roundness error of each cross-section is calculated using the three-point method and the minimum region method. The actual apex angle of the V-shaped positioning groove 301 refers to the angle between the two positioning groove surfaces 302, which directly affects the measurement results. The preset error transfer matrix is ​​used to convert the displacement of the probe 502 into profile error. The three-point method is a commonly used method for calculating roundness error, which determines roundness using data from three measurement points. The minimum region method is a method used to determine the least squares circle or the minimum enclosing circle, thereby calculating the roundness error. For example, the actual apex angle can be obtained first to retrieve the corresponding preset error transfer matrix. Then, based on the actual displacement sequence, the actual apex angle, and the preset error transfer matrix, the three-point method is used to calculate the profile error sequence of each cross-section (the profile error sequence includes the profile error corresponding to the acquisition position of each frame of interference fringe image). Finally, based on the profile error sequence of each cross-section, the minimum region method is used to calculate the roundness error of each cross-section.

[0078] In step A7, the overall roundness error of the shaft component 90 under test is calculated by combining the roundness errors of each cross section. This typically involves weighted averaging or other statistical processing of the roundness errors at different cross sections to obtain an index that can comprehensively reflect the overall roundness quality of the shaft component 90 under test.

[0079] The solution proposed in this application combines the aforementioned shaft roundness error detection system with a sophisticated data processing and error calculation method to achieve high-precision detection of the roundness error of shaft components. Specifically, the detection system first ensures that the shaft component 90 under test rotates in a stable and controlled state through an air-bearing support 1, a rotation drive mechanism 2, and a positioning mechanism. Then, the interferometer 501 and probe 502 in the measurement system acquire interference fringe images at multiple cross-sections of the shaft component 90 under test in real time and non-contactly. These raw images contain displacement information regarding the surface contour changes of the shaft component 90 under test, which forms the basis for subsequent calculations.

[0080] Subsequently, this method transforms these raw images into accurate data usable for roundness error calculation through a series of data processing steps. First, preprocessing steps (grayscale conversion, dynamic noise reduction, and fringe sharpening) effectively improve image quality, laying the foundation for subsequent fringe feature extraction. Next, by extracting the pixel positions of zero-order and secondary fringes and verifying and correcting the zero-order fringe pixel displacement based on the secondary fringe pixel displacement, this scheme overcomes the weakness of traditional methods where single-fringe identification is susceptible to noise interference, significantly improving the accuracy and robustness of the 502-pixel displacement calculation of the probe. This verification and correction mechanism ensures that the displacement data extracted from the interference fringe image is more reliable, thus avoiding error accumulation caused by inaccurate raw data.

[0081] Furthermore, the zero-order fringe pixel displacement after verification and correction is converted into the actual displacement of the probe 502. This conversion process takes into account the working principle and calibration parameters of the interferometer 501, ensuring the physical accuracy of the displacement data. Finally, in the roundness error calculation stage, this scheme combines the actual groove apex angle of the V-shaped positioning groove 301 and the preset error transfer matrix, using the three-point method and the minimum area method to effectively eliminate the measurement error caused by the V-block positioning, and accurately calculate the roundness error of each cross-section from the actual displacement sequence of the probe 502. By integrating the roundness errors of each cross-section, the overall roundness error of the shaft component 90 under test is finally obtained, thus providing a comprehensive and accurate roundness evaluation result.

[0082] Preferably, step A3 includes:

[0083] A301. Based on the Otsu threshold segmentation algorithm, the background region and the fringe region in the preprocessed interference fringe image are determined, and the gray value of the background region is set to zero to obtain the zeroed interference fringe image;

[0084] A302. Based on the extension column direction of the fringe region, calculate the sum of gray values ​​of each column of pixels in the interference fringe image after setting to zero, and extract the column coordinates corresponding to the maximum value of the sum of gray values ​​as the pixel position of the zero-order fringe.

[0085] A303. Using the pixel position of the zero-level stripe as the center, search for the local extreme points of the sum of the gray values ​​on both sides, and combine with the preset stripe period to determine the pixel positions of the ±1 level secondary stripes.

[0086] A304. The pixel positions of zero-order and secondary stripes are corrected by Kalman filtering.

[0087] Specifically, in step A301, the Otsu threshold segmentation algorithm is used to automatically determine the optimal threshold in the interference fringe image, thereby clearly dividing the image into background and fringe regions. By setting the grayscale value of the background region to zero, the interference of background noise can be effectively eliminated, highlighting the fringe information and providing a clean data foundation for subsequent fringe position extraction.

[0088] In step A302, in the interference fringe image where the background grayscale value is set to zero, the sum of the grayscale values ​​of each column of pixels is calculated along the extension column direction of the fringe region. Since the zero-order fringe usually has the highest intensity or the widest width, the sum of the grayscale values ​​of the corresponding column of pixels will reach its maximum value. Therefore, by extracting the column coordinates corresponding to this maximum value, the pixel position of the zero-order fringe can be accurately located.

[0089] In practical applications, in step A303, a search is performed to both sides of the determined pixel position of the zero-order fringe to identify local extrema of the sum of gray values. These local extrema correspond to the positions of secondary fringes. Combined with a preset fringe period, the two local extrema closest to the zero-order fringe can be accurately selected as the pixel positions of the +1 and -1 order secondary fringes, respectively. The preset fringe period can be calibrated based on the wavelength of the light source 5011 and the optical system parameters of the interferometer 501.

[0090] Further, in step A304, Kalman filtering is applied to correct the pixel positions of the extracted zero-order and secondary fringes. Kalman filtering is an optimal estimation algorithm that can effectively handle noisy dynamic system data. Through prediction and update mechanisms, it compensates for random and systematic errors that may exist during the measurement process, such as positioning deviations caused by mechanical vibration, airflow disturbances, or sensor noise, thereby improving the accuracy and stability of the fringe pixel positions.

[0091] This application's solution effectively addresses the accuracy and robustness issues in traditional interference fringe pixel location extraction by introducing a series of image processing and filtering techniques. Specifically, the Otsu threshold segmentation algorithm adaptively separates interference fringes from complex backgrounds, ensuring the purity of subsequent data processing. By calculating the sum of column pixel grayscale values ​​and finding the maximum value, the zeroth-order fringe with the strongest energy can be accurately identified, avoiding errors caused by subjective judgment or simple thresholds. Subsequently, based on the zeroth-order fringe position and a preset fringe period, local extrema are searched, enabling the systematic location of secondary fringes and ensuring correct identification of fringe orders. Finally, the introduction of a Kalman filter allows the estimation of fringe pixel positions to dynamically adapt to environmental changes and measurement noise. By fusing historical data and current measurements, the fringe positions are corrected in real time and optimally, significantly improving the accuracy and anti-interference capability of fringe localization.

[0092] In some implementations, step A4 includes:

[0093] A401. For each cross section, calculate the difference in pixel position of the zero-order fringe, the difference in pixel position of the +1-order secondary fringe, and the difference in pixel position of the -1-order secondary fringe corresponding to each two adjacent frames of interference fringe images, and obtain the corresponding zero-order fringe pixel displacement, +1-order secondary fringe pixel displacement, and -1-order secondary fringe pixel displacement.

[0094] A402. Calculate the absolute difference between the zero-level fringe pixel displacement and the corresponding +1 level secondary fringe pixel displacement, and record it as the first absolute difference. Also calculate the absolute difference between the zero-level fringe pixel displacement and the corresponding -1 level secondary fringe pixel displacement, and record it as the second absolute difference.

[0095] A403. If both the first absolute difference and the second absolute difference are not greater than a preset threshold, then the zero-level stripe pixel displacement is determined to be valid; otherwise, the zero-level stripe pixel displacement is determined to be invalid and anomaly processing is performed.

[0096] Specifically, in step A401, for each cross-section of the shaft component 90 under test, the pixel displacement of the zero-order fringe can be obtained by comparing the pixel positions of the zero-order fringe in two consecutive frames of interference fringe images (specifically, by subtracting the pixel position of the zero-order fringe in the previous frame from the pixel position of the zero-order fringe in the later frame of the two consecutive frames of interference fringe images). Similarly, by comparing the pixel positions of the +1 and -1 order secondary fringes in two consecutive frames of images, the pixel displacements of the +1 and -1 order secondary fringes can be obtained, respectively. These pixel displacements reflect the relative movement of the probe 502 within the corresponding time interval.

[0097] In step A402, the first absolute difference refers to the absolute value of the difference between the zero-order fringe pixel displacement and the +1-order secondary fringe pixel displacement, and the second absolute difference refers to the absolute value of the difference between the zero-order fringe pixel displacement and the -1-order secondary fringe pixel displacement. These absolute differences are used to quantify the degree of consistency between the zero-order fringe pixel displacement and the secondary fringe pixel displacement.

[0098] In practical applications, in step A403, the validity of the zero-level fringe pixel displacement can be determined by comparing the first absolute difference and the second absolute difference with a preset threshold. If neither absolute difference is greater than the preset threshold, the zero-level fringe pixel displacement is considered reliable and valid. Conversely, if either absolute difference is greater than the preset threshold, it indicates that the zero-level fringe pixel displacement may be abnormal and requires anomaly handling. Specific methods for anomaly handling may include discarding the zero-level fringe pixel displacement of that frame, or using the average of the valid zero-level fringe pixel displacements of the previous few frames (e.g., the first three frames) as a substitute to smooth the data and reduce the impact of outliers on subsequent calculations.

[0099] This application's solution verifies the validity of the zero-order fringe pixel displacement by introducing the secondary fringe pixel displacement as a reference for the zero-order fringe pixel displacement. Ideally, due to the periodicity of the interference fringes, the relative positional relationship between the zero-order and secondary fringes should remain stable, thus their pixel displacements should have a high degree of numerical consistency. When the zero-order fringe pixel displacement deviates due to noise or interference, the absolute difference between it and the secondary fringe pixel displacement will increase significantly. By setting a reasonable preset threshold, these abnormal zero-order fringe pixel displacements can be effectively identified. This verification mechanism can promptly detect and process inaccurate measurement data, thereby avoiding its negative impact on subsequent calculations of the actual displacement of the probe 502 and the final roundness error detection results.

[0100] Preferably, after step A403, the following step may be included:

[0101] A404. Acquire environmental data and correct the zero-order stripe pixel displacement using a preset compensation correction model.

[0102] Specifically, environmental data can be understood as external physical quantities that affect the stability and accuracy of the measurement system, such as temperature and vibration data. Temperature data can be acquired in real time by placing temperature sensors at key parts of the detection system (such as the interferometer 501 and the vicinity of the shaft under test 90); vibration data can be monitored by installing accelerometers or vibration sensors on the system base or key support structures. After being acquired in real time, this environmental data is input into a pre-set compensation and correction model. This compensation and correction model is a mathematical model based on experimental data or theoretical analysis, used to describe the relationship between environmental data and zero-order fringe pixel displacement deviation. This model can be obtained in advance through calibration experiments, for example, by measuring known displacements under different temperature or vibration conditions and recording the corresponding pixel displacement deviations, thereby establishing a mapping relationship between environmental parameters and deviations. Its purpose is to quantify the impact of environmental factors on the measurement results and perform corresponding compensation to eliminate or reduce errors caused by these external disturbances.

[0103] The proposed solution, after verifying and correcting the zero-order fringe pixel displacement, further incorporates environmental data and combines it with a compensation correction model for further correction, effectively addressing the impact of environmental factors on measurement accuracy. Specifically, when environmental conditions such as temperature or vibration change, these changes are collected in real time as environmental data. This environmental data is then input into a pre-established compensation correction model, which calculates the potential deviation of the zero-order fringe pixel displacement based on the current environmental conditions. It is precisely this dynamic compensation mechanism based on actual environmental conditions that allows for precise adjustment of the original zero-order fringe pixel displacement, thereby eliminating systematic errors introduced by environmental factors and ensuring the accuracy of subsequent displacement calculations of the probe 502.

[0104] In some preferred embodiments, steps A1-A6 are repeated multiple times to obtain multiple roundness errors for each cross section;

[0105] Step A7 includes:

[0106] A701. Calculate the mean of all the roundness errors of each cross section to obtain the effective roundness error of each cross section;

[0107] A702. Based on the preset weights of each cross section, perform a weighted sum or weighted average sum operation on the effective roundness errors of each cross section to obtain the overall roundness error of the shaft component 90 to be tested.

[0108] Specifically, repeating steps A1 to A6 multiple times refers to performing multiple complete measurement cycles on the same shaft component 90 under the same measurement conditions. Each cycle begins with the drive rotation of the shaft component 90 (which may further include an installation step), image acquisition, preprocessing, pixel position extraction, displacement calculation and correction, and finally, the calculation of the roundness error of each cross-section. Through this repeated measurement, a series of independent roundness error measurements can be obtained for each cross-section. As a preferred implementation, the number of repetitions can be no less than 5 times. This is based on statistical principles, increasing the sample size to improve the statistical significance and reliability of the measurement results.

[0109] Step A701 aims to eliminate or significantly reduce the impact of random errors by arithmetically averaging the multiple roundness errors obtained for each cross-section. The mean calculation smooths out any instantaneous fluctuations or outliers that may occur in a single measurement, resulting in a more stable and accurate effective roundness error.

[0110] In practical applications, step A702 further considers the importance of different cross-sections of the shaft component 90 under test, or the ease of measurement. Preset weights can be set based on the shaft component's design requirements, functional criticality, machining accuracy requirements, or measurement experience. For example, cross-sections closer to the center of the shaft component 90 can be assigned higher weights to ensure their roundness error plays a more significant role in the overall evaluation, thereby reducing the impact of deformation caused by external forces on the bottom and top on the accuracy of the test results. Through weighted summation or weighted average summation, the influence of roundness errors of different cross-sections on the overall quality can be comprehensively reflected, resulting in a more representative and practically valuable overall roundness error.

[0111] The solution presented in this application effectively addresses the random errors and uncertainties that may exist in a single measurement by repeatedly executing the measurement process and statistically processing the results. Specifically, when steps A1 to A6 are repeated multiple times, the random errors generated in each measurement exhibit a certain random distribution characteristic. By averaging the roundness errors obtained from these multiple measurements (step A701), these random errors statistically cancel each other out, making the calculated effective roundness error closer to the true roundness error of the shaft component 90 under test, significantly improving the measurement accuracy and reliability of the roundness error of a single cross-section. Furthermore, when calculating the overall roundness error of the shaft component 90 under test, considering that different cross-sections may have different importance or different degrees of influence on the overall performance of the shaft component, this application introduces a preset weight (step A702). By performing a weighted sum or weighted average sum operation on the effective roundness errors of each cross-section, greater attention can be paid to the roundness error of key cross-sections, so that the final overall roundness error can more accurately reflect the performance requirements of the shaft component 90 under test in practical applications, avoiding the underestimation of errors in key areas that may be caused by simple averaging.

[0112] Preferably, before step A701, the following step may be included:

[0113] A700. For each cross section, perform a t-test on the contour error sequence corresponding to each roundness error to identify contour error sequences with systematic errors and remove the corresponding roundness errors.

[0114] Specifically, the t-test is a statistical hypothesis testing method primarily used to assess whether there is a significant difference between the means of one or more samples. In this application, the t-test is applied to perform statistical analysis on multiple contour error sequences obtained for each cross-section. Each roundness error corresponds to a contour error sequence, which records the contour deviation data of the tested shaft part 90 at multiple circumferential positions at a specific cross-section. By performing t-tests on these contour error sequences, it can be determined whether a particular sequence is statistically significantly different from other sequences or the expected error-free state. For example, a significance level can be set; if the t-test result of a particular sequence shows that its difference from the remaining sequences exceeds this significance level, then the sequence is considered to potentially contain systematic errors.

[0115] Identifying contour error sequences with systematic errors involves using a t-test to determine whether a particular contour error sequence is statistically significantly different from the overall distribution or the pre-defined normal range of other sequences. Such a significant difference may indicate systematic bias or sporadic anomalies in the measurement process.

[0116] In practical applications, removing the corresponding roundness error means that once a contour error sequence is identified as having a systematic error by a t-test, its corresponding roundness error will be excluded from subsequent mean calculations. This measure aims to prevent outlier data from negatively impacting the final roundness error calculation result, thereby ensuring that the calculated effective roundness error is more accurate and reliable.

[0117] The proposed solution employs a t-test to statistically evaluate the contour error sequence obtained from multiple measurements of each cross-section. When a contour error sequence becomes abnormal due to systematic errors or occasional interference, its data distribution will differ significantly from the normal sequence. The t-test can sensitively capture such statistical differences, thereby effectively identifying these contour error sequences with systematic errors. Once an abnormal sequence is identified, its corresponding roundness error is eliminated, avoiding interference from these inaccurate data in subsequent mean calculations. It is precisely because of this abnormal data identification and elimination mechanism based on statistical principles that the effective roundness error of each cross-section calculated in the end can more accurately reflect the actual roundness of the shaft component 90 being measured, significantly improving the accuracy and reliability of the measurement results.

[0118] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A system for detecting roundness error of shaft components, characterized in that, include: Air-bearing support (1) is used to support the lower end face of the shaft component (90) to be tested; A rotation drive mechanism (2) is disposed above the air flotation support (1) and is used to drive the shaft to be tested (90) to rotate around its own axis from the upper end of the shaft to be tested (90); The positioning mechanism includes a V-shaped positioning component (3) and a pre-tightening component (4) arranged opposite to each other. The V-shaped positioning component (3) has a V-shaped positioning groove (301), and the pre-tightening component (4) is used to provide pre-tightening pressure to the shaft part (90) to be tested, so that the circumferential surface of the shaft part (90) to be tested fits with the two positioning groove surfaces (302) of the V-shaped positioning groove (301). The measurement system includes multiple measurement units (5) arranged at intervals along the vertical direction. Each measurement unit (5) includes an interferometer (501) and a probe (502) linked to a movable mirror (5015) of the interferometer (501). The probe (502) is used to contact the circumferential surface of the shaft component (90) to be measured, so as to generate displacement as the contour of the shaft component (90) changes. The interferometer (501) is used to convert the displacement of the probe (502) into an interference fringe image and output it. The host computer (6) is used to calculate the roundness error of the shaft component (90) under test at multiple different cross sections based on the interference fringe images output by each of the measurement units (5), and to calculate the overall roundness error of the shaft component (90) under test by combining the roundness errors of each cross section. The V-shaped positioning component (3) includes a fixed base (303), two swing blocks (304), and two swing drive devices (305). The two swing blocks (304) are swing-mounted on the side of the fixed base (303) near the pre-tightening component (4). The two swing blocks (304) are provided with positioning groove surfaces (302) on the side of each other. The V-shaped positioning groove (301) is formed between the two positioning groove surfaces (302). The two swing drive devices (305) are used to drive the two swing blocks (304) to swing, so as to adjust the included angle between the two positioning groove surfaces (302).

2. The roundness error detection system for shaft parts according to claim 1, characterized in that, The two positioning groove surfaces (302) of the V-shaped positioning groove (301) are coated with polytetrafluoroethylene.

3. The roundness error detection system for shaft parts according to claim 1, characterized in that, The pretensioning assembly (4) includes a stopper (401), a pretensioning spring (402), a clamping member (403), a pressure sensor (404), an adjusting member (405), and a fixing screw (406). The clamping member (403) is slidably sleeved on one end of the fixed screw (406) near the V-shaped positioning component (3). The adjusting member (405) is disposed on the side of the clamping member (403) away from the V-shaped positioning component (3) and threadedly connected to the fixed screw (406). The pressure sensor (404) is disposed between the clamping member (403) and the adjusting member (405). The abutting member (401) is disposed on the side of the clamping member (403) near the V-shaped positioning component (3). The preload spring (402) is connected between the clamping member (403) and the abutting member (401). The preload spring (402) is used to provide elastic force to press the abutting member (401) against the circumferential surface of the shaft part (90) to be tested.

4. The roundness error detection system for shaft parts according to claim 1, characterized in that, The interferometer (501) includes a light source (5011), an interference optical path assembly, and an image sensor (5012). The interference optical path assembly includes a beam splitter (5013), a fixed reflector (5014), and a movable reflector (5015).

5. The roundness error detection system for shaft parts according to claim 4, characterized in that, The light source (5011) includes two switchable lasers, namely a helium-neon laser and a semiconductor laser.

6. A method for detecting roundness error of shaft-type parts, characterized in that, The roundness error detection system for shaft components according to any one of claims 1-5 includes the following steps: A1. After installing the shaft component (90) to be tested in a vertical position on the shaft component roundness error detection system, drive the shaft component (90) to be tested to rotate one revolution, and collect interference fringe images at multiple different cross sections of the shaft component (90) to be tested in real time. A2. Preprocess the interference fringe image; the preprocessing includes grayscale conversion, dynamic noise reduction, and fringe sharpening. A3. Extract the pixel positions of the zeroth-order and secondary fringes in each preprocessed interference fringe image; A4. For each cross section, based on the pixel positions of the zero-order fringe and the secondary fringe, calculate the zero-order fringe pixel displacement and the secondary fringe pixel displacement corresponding to each two adjacent frames of interference fringe images, and verify and correct the zero-order fringe pixel displacement based on the secondary fringe pixel displacement. A5. For each cross section, calculate the actual displacement of the corresponding probe (502) based on the zero-order fringe pixel displacement after each verification and correction, and obtain the actual displacement sequence of each cross section; A6. Based on the actual displacement sequence, combined with the actual groove apex angle of the V-shaped positioning groove (301) and the preset error transmission matrix corresponding to the actual groove apex angle, the roundness error of each cross section is calculated using the three-point method and the minimum area method. A7. Calculate the overall roundness error of the shaft component (90) to be tested by taking into account the roundness errors of each cross section.

7. The method for detecting roundness error of shaft parts according to claim 6, characterized in that, Step A3 includes: A301. Based on the Otsu threshold segmentation algorithm, the background region and the fringe region in the preprocessed interference fringe image are determined, and the gray value of the background region is set to zero to obtain the zeroed interference fringe image; A302. Based on the extension column direction of the fringe region, calculate the sum of gray values ​​of each column of pixels in the interference fringe image after setting to zero, and extract the column coordinates corresponding to the maximum value of the sum of gray values ​​as the pixel position of the zero-order fringe. A303. Using the pixel position of the zero-level stripe as the center, search for the local extreme points of the sum of the gray values ​​on both sides, and combine with the preset stripe period to determine the pixel positions of the ±1 level secondary stripes. A304. The pixel positions of zero-order and secondary stripes are corrected by Kalman filtering.

8. The method for detecting roundness error of shaft parts according to claim 7, characterized in that, Step A4 includes: A401. For each cross section, calculate the difference in pixel position of the zero-order fringe, the difference in pixel position of the +1-order secondary fringe, and the difference in pixel position of the -1-order secondary fringe corresponding to each two adjacent frames of interference fringe images, and obtain the corresponding zero-order fringe pixel displacement, +1-order secondary fringe pixel displacement, and -1-order secondary fringe pixel displacement. A402. Calculate the absolute difference between the zero-level fringe pixel displacement and the corresponding +1 level secondary fringe pixel displacement, and record it as the first absolute difference. Also calculate the absolute difference between the zero-level fringe pixel displacement and the corresponding -1 level secondary fringe pixel displacement, and record it as the second absolute difference. A403. If both the first absolute difference and the second absolute difference are not greater than a preset threshold, then the zero-level stripe pixel displacement is determined to be valid; otherwise, the zero-level stripe pixel displacement is determined to be invalid and anomaly processing is performed.

9. The method for detecting roundness error of shaft parts according to claim 6, characterized in that, Repeat steps A1-A6 multiple times to obtain multiple roundness errors for each cross section; Step A7 includes: A701. Calculate the mean of all the roundness errors of each cross section to obtain the effective roundness error of each cross section; A702. Based on the preset weights of each cross section, perform a weighted sum or weighted average sum operation on the effective roundness errors of each cross section to obtain the overall roundness error of the shaft component (90) to be tested.

Images