A thickness and flatness detection method and device based on the transmission mode

Abstract

The application discloses a kind of thickness and flatness detection method and device based on opposite laser mode, including bottom plate, protective cover, double-shaft module, laser detection component, display component and control component, device whole is desktop type structure, its external dimensions is less than 500mm×600mm×500mm, double-shaft module uses independent separation type double-shaft module design, is composed of X-axis module, Y-axis module, both orthogonal linkage realizes full point scanning, Y-axis module is fixed on the top of bottom plate;X-axis module is vertically installed on the back plate of equipment, and the moving end is connected to the "C-shaped" support bracket, directly drives laser detection component to complete X scanning in left and right directions, and laser detection component uses up and down opposite laser detection architecture.The application realizes product thickness multi-point full-automatic high-precision detection by the cooperation of opposite laser measurement structure and double-shaft movement module, considers measurement accuracy, detection efficiency and operation safety, and completely solves the precision bottleneck and single-function problem of traditional scheme.

Description

Technical Field

[0001] This invention relates to the field of industrial manufacturing measurement technology, specifically to a method and apparatus for thickness and flatness detection based on a through-beam method. Background Technology

[0002] In the quality control of precision manufacturing, food packaging, and light industrial products, product thickness is one of the important technical indicators for measuring the stability and consistency of product quality. Especially for sheet-like, plate-like, or regularly shaped products, their thickness is easily affected by factors such as batch differences in raw materials and fluctuations in processing technology, resulting in a certain degree of dispersion. Measurement results from a single location are often insufficient to fully reflect the overall thickness of the product. Therefore, in practical applications, it is usually necessary to conduct a comprehensive evaluation of product thickness through multi-point measurements to improve the representativeness and reliability of the test results.

[0003] Currently, thickness measurement for the aforementioned products still primarily relies on manual methods, commonly employing contact measuring tools such as vernier calipers and micrometers for single-point or small-scale measurements. While these methods are simple to operate, they have significant shortcomings in practical use. Firstly, manual measurement is inefficient and cannot meet the demands of rapid testing for batch products. Secondly, measurement results are easily affected by human factors such as operator experience, measurement angle, and contact pressure, resulting in poor repeatability and consistency. Furthermore, contact measurements may cause scratches or indentations on the product surface during the testing process, making them unsuitable for products with high surface quality requirements.

[0004] With the development of non-contact inspection technology, thickness measurement methods based on laser displacement sensors are gradually being applied. Laser measurement has advantages such as high accuracy, fast response, and no mechanical contact with the measured object, and has good application prospects in the field of industrial inspection. Existing laser thickness detection methods mainly include single-sided measurement and double-sided measurement. Among them, the double-sided through-beam measurement method based on laser triangulation, by measuring the positions of both sides of the workpiece separately and calculating the thickness, can effectively reduce the influence of changes in the workpiece placement position on the measurement results, resulting in higher measurement accuracy.

[0005] However, existing laser thickness measurement equipment still has certain limitations in practical engineering applications. On the one hand, some equipment has a complex structure and large size, usually requiring fixed installation on production lines or dedicated testing platforms, making it unsuitable for desktop or laboratory environments. On the other hand, existing equipment mainly focuses on single-point measurement, lacking stable and precise motion mechanisms, making it difficult to achieve multi-point automatic scanning and detection, resulting in deficiencies in thickness consistency assessment. Furthermore, insufficient precision in the motion mechanism or installation structure can easily introduce repeatability errors at the detection points, thus affecting the overall detection accuracy. Therefore, there is an urgent need to design a thickness and flatness detection method and device based on a through-beam method to solve the above problems. Summary of the Invention

[0006] The purpose of this invention is to provide a thickness and flatness detection method and apparatus based on a through-beam method, so as to overcome the above-mentioned shortcomings in the prior art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A thickness and flatness detection device based on a through-beam method includes a base plate, a protective cover, a dual-axis module, a laser detection component, a display component, and a control component.

[0008] In a preferred embodiment of the present invention, the device is a desktop structure with dimensions less than 500mm × 600mm × 500mm.

[0009] In a preferred embodiment of the present invention, the dual-axis module adopts an independent and separate dual-axis module design, consisting of an X-axis module and a Y-axis module. The two modules move in a positive cross-linking manner to achieve full-point scanning. The Y-axis module is fixed above the base plate. The X-axis module is vertically installed on the back plate of the equipment, and the moving end is connected to a "C-shaped" support bracket to directly drive the laser detection component to complete the X-axis scanning in the left and right directions.

[0010] In a preferred embodiment of the present invention, the laser detection component adopts an up-and-down through-beam laser detection architecture.

[0011] A thickness and flatness detection method based on a through-beam method, using the aforementioned through-beam method thickness and flatness detection device, includes the following steps: S1. Mark the positioning references at multiple core detection points preset on the product surface; S2. Control the dual-axis motion module to drive the through-beam laser detection component to move sequentially to the area above each of the positioning references, and scan each area, simultaneously collecting the two-dimensional mechanical coordinates (X,Y) of each scanning point and the Z-axis height value measured by the upper and lower laser displacement sensors to form a three-dimensional coordinate point cloud (X,Y,Z). S3. For each core detection point, calculate the average height difference based on multiple sets of upper and lower laser height values ​​collected within its area to obtain the thickness characteristic value of that point. S4. Based on the thickness characteristic values ​​and planar coordinates of all core points, construct a spatial fitting plane for the product surface, and calculate the angle between the fitting plane and the theoretical reference plane as a thickness consistency evaluation index. S5. Extract all Z-axis height data from the three-dimensional coordinate point cloud data collected in step S2, construct an ideal reference plane for the side surface, and calculate the peak-valley range of the height deviation of each measurement point relative to the reference plane as an index of the flatness or warping of the side surface. S6 outputs the thickness at each point, thickness consistency index, and flatness / warping index.

[0012] In a preferred embodiment of the present invention, in step S1, the operator places the product to be inspected stably in the inspection area, selects 5 core inspection points on the product surface according to the product specifications and inspection requirements, and marks a cross reference at the center of each point to provide a reference for subsequent accurate scanning and positioning.

[0013] In a preferred embodiment of the present invention, in S2, the control component issues instructions to drive the dual-axis module to accurately position itself at each detection point; at a single point, with the cross reference as the core, high-frequency multi-point scanning and acquisition are carried out on the surrounding area according to a preset trajectory. During the scanning process, a deep fusion mechanism of hardware and software data is enabled: the control component collects the motion position commands of the X-axis and Y-axis modules or the encoder feedback values ​​in real time to generate the two-dimensional physical plane coordinates (X,Y) of each measurement point; and simultaneously reads the optical distance data of the upper and lower laser displacement sensors and maps it to a high-precision Z-axis normal height value. The control system aligns and deeply integrates the two-dimensional mechanical coordinates with the Z-axis optical data in real time to construct a spatial three-dimensional coordinate point cloud (X,Y,Z) containing product position and height information, providing accurate underlying data for subsequent algorithm processing.

[0014] In a preferred embodiment of the present invention, before calculating the average height difference in step S3, an outlier removal step is included.

[0015] In a preferred embodiment of the present invention, in S4, a spatial plane fitting algorithm is used to construct the spatial fitting plane.

[0016] In a preferred embodiment of the present invention, in S5, the ideal reference plane is constructed using the least squares method.

[0017] In the above technical solution, the present invention provides a thickness and flatness detection method and apparatus based on a through-beam method, which has the following advantages: (1) By coordinating the through-beam laser measurement structure and the dual-axis motion module, the product thickness can be fully automatically detected at multiple points with high precision, taking into account measurement accuracy, detection efficiency and operation safety, and completely solving the precision bottleneck and single function problem of traditional solutions.

[0018] (2) Compared with traditional contact measurement, single-sided laser ranging, and conventional through-beam measurement schemes: this invention; By eliminating the shortcomings of single-sided ranging that relies on a reference surface, this method uses top-and-bottom laser beams to directly collect data from the upper and lower surfaces of the product, converting the data into absolute thickness. This completely eliminates the cumulative errors caused by reference surface tolerances and mechanical deformation, resulting in a measurement error of ≤6μm for the entire machine, far exceeding the accuracy of existing solutions. Breaking through the limitations of traditional single-point measurement, this method uses a crosshair reference as the core to conduct multi-point sampling in local micro-areas. Through averaging processing, interference from surface scratches, particles, and roughness is eliminated, significantly improving the authenticity and robustness of single-point data and avoiding misjudgments.

[0019] The innovative integration of a five-point spatial fitting algorithm and single-sided data stripping technology allows for the simultaneous output of two core indicators—absolute thickness and single-sided flatness / warping—without increasing hardware costs, achieving a technological leap from "one-dimensional ranging" to "three-dimensional morphology evaluation" through a single scan.

[0020] With a desktop-sized design, it is suitable for deployment in various scenarios; equipped with a protective cover and emergency stop device, it achieves a safe and foolproof design; the operation process is simplified to "loading - marking start - viewing results - unloading", which greatly reduces the skill requirements of operators and the detection efficiency is far higher than that of manual and conventional equipment.

[0021] (3) The core breakthrough of this invention lies in the multi-dimensional detection capability of deep integration of software and hardware, rather than single hardware integration. It breaks the traditional reliance on the reference of single-sided ranging and adopts an orthogonal architecture of Y-bearing product stepping and X-axis driven through-beam laser scanning. By directly superimposing the physical data of the upper and lower paths, the cumulative error of mechanical tolerance is eliminated, and high-precision acquisition of absolute thickness is achieved. It is the first to create a cross-reference positioning + local micro-area high-frequency sampling mode. Through multi-point averaging processing, surface micro-areas are effectively filtered out. The original data of the thickness measurement scan is deeply reused, and the single-sided laser height value is peeled off to construct a reference plane. The flatness / warping is calculated by the peak-valley range, so that the thickness measurement and morphology detection are completed simultaneously without additional time consumption. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0023] Figure 1This is a schematic diagram of the structure of an embodiment of a thickness and flatness detection method and apparatus based on a through-beam method according to the present invention; Figure 2 This is a structural schematic diagram of an embodiment of a thickness and flatness detection method and apparatus based on a through-beam method according to the present invention.

[0024] 1. Base plate; 2. Protective cover; 3. Dual-axis module; 4. Laser detection assembly; 5. Control assembly. Detailed Implementation

[0025] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0026] The embodiments of this invention provide: like Figure 1 As shown, a thickness and flatness detection device based on a through-beam method includes a base plate 1, a protective cover 2, a dual-axis module 3, a laser detection component 4, a display component, and a control component 5. Each module has a clear division of labor and works in coordination: the base plate 1 serves as the support base for the entire machine, ensuring structural stability; the protective cover 2 encloses and seals the detection area, preventing personnel from accidentally touching the moving module and detection component during equipment operation, thus improving operational safety from a physical perspective.

[0027] In this embodiment, the device is a desktop structure with dimensions of less than 500mm×600mm×500mm. It is compact, reasonably laid out, and can be directly placed on a laboratory workbench or a sampling inspection station in the production site, making it suitable for flexible deployment in multiple scenarios. The base plate 1 is made of aluminum alloy plate with a thickness of not less than 15mm and is equipped with leveling feet to ensure the rigidity of the foundation. The static rigid deformation of the whole machine is controlled within 10μm.

[0028] In this embodiment, the dual-axis module 3 adopts an independent and separate dual-axis module design, consisting of an X-axis module and a Y-axis module. The two modules work together to achieve full-point scanning. The Y-axis module is fixed above the base plate and mainly supports the product under test and drives it to complete the Y-axis stepping displacement in the front-to-back direction. The X-axis module is vertically installed on the back plate of the equipment, and its moving end is connected to a "C-shaped" support bracket, directly driving the laser detection component to complete the X-axis scanning in the left-to-right direction. Through precise linkage of the two axes, fully automatic array scanning of preset points on the surface of the product under test can be achieved without manual intervention.

[0029] It should be noted that the Y-axis module uses a ball screw slide with a stroke of not less than 200mm and a repeatability of ≤±5μm, and is used to carry products weighing no more than 2kg; the X-axis module (6) uses a synchronous belt or ball screw slide with a stroke of not less than 300mm and a repeatability of ≤±3μm; the two are interpolated by a controller and the linkage positioning accuracy can reach ±8μm; the “C-type” support bracket is made of 7075 aluminum alloy as a whole and is subjected to stress relief annealing to ensure its connection rigidity and form and position tolerance, and the parallelism error of the upper and lower laser sensor mounting surfaces is ≤0.01mm.

[0030] In this embodiment, the laser detection component 4 adopts a top-and-bottom through-beam laser detection architecture. This completely eliminates the dependence on the bottom reference plane, thereby eliminating indirect calculation errors caused by uneven platform and mechanical tolerances from the measurement principle. It enables direct measurement of the absolute thickness of the product, rather than the relative thickness conversion of traditional single-sided distance measurement, thus ensuring measurement accuracy from the source.

[0031] It should be noted that both the upper and lower laser displacement sensors are non-contact sensors based on the principle of laser triangular reflection, with a measurement resolution ≤1μm, linearity ≤±0.05%FS, and sampling frequency ≥2kHz. The two are installed in an absolute opposite manner, that is, the emitting optical path and the receiving optical path are strictly coaxial and opposite, the object being measured is placed in the optical path, and the absolute position of its upper and lower surfaces is directly measured.

[0032] like Figure 2 As shown, a thickness and flatness detection method based on a through-beam method is used, employing a thickness and flatness detection device based on a through-beam method, and includes the following steps: S1. Mark the positioning references at multiple core detection points preset on the product surface; The operator places the product to be inspected stably in the inspection area, selects 5 core inspection points on the product surface according to the product specifications and inspection requirements, and marks a cross reference at the center of each point to provide a reference for subsequent accurate scanning and positioning.

[0033] S2. Control the dual-axis motion module to drive the through-beam laser detection component to move sequentially to the area above each positioning reference, and scan each area. Simultaneously collect the two-dimensional mechanical coordinates (X,Y) of each scanning point and the Z-axis height value measured by the upper and lower laser displacement sensors to form a three-dimensional coordinate point cloud (X,Y,Z). The control component issues commands to drive the dual-axis module to accurately position itself at each detection point; at a single point, using the crosshair reference as the core, it performs high-frequency multi-point scanning and acquisition of the surrounding area according to a preset trajectory; During the scanning process, a deep fusion mechanism of hardware and software data is enabled: the control component collects the motion position commands of the X-axis and Y-axis modules or the encoder feedback values ​​in real time to generate the two-dimensional physical plane coordinates (X,Y) of each measurement point; and simultaneously reads the optical distance data of the upper and lower laser displacement sensors and maps it to a high-precision Z-axis normal height value. The control system aligns and deeply integrates the two-dimensional mechanical coordinates with the Z-axis optical data in real time to construct a spatial three-dimensional coordinate point cloud (X,Y,Z) containing product position and height information, providing accurate underlying data for subsequent algorithm processing.

[0034] S3. For each core detection point, calculate the average height difference based on multiple sets of upper and lower laser height values ​​collected within its area to obtain the thickness characteristic value of that point. Before calculating the average height difference, there is also a step to remove outliers, filter out high-frequency noise interference caused by micro-roughness, local scratches, and tiny particles on the product surface, accurately extract the true thickness feature value of the detection point, and ensure the reliability of single-point data.

[0035] S4. Based on the thickness characteristic values ​​and planar coordinates of all core points, construct a spatial fitting plane for the product surface and calculate the angle between the fitting plane and the theoretical reference plane as a thickness consistency evaluation index; use a spatial plane fitting algorithm to construct a spatial fitting plane, quantitatively evaluate the overall thickness consistency and wedge deformation level of the product, and realize macroscopic quality judgment.

[0036] S5. From the 3D coordinate point cloud data collected in step S2, extract all Z-axis height data of the single-sided laser displacement sensor, construct an ideal reference plane for that side surface, and calculate the peak-to-valley range of the height deviation of each measurement point relative to the reference plane, as an index of the flatness or warpage of that side surface; the ideal reference plane is constructed using the least squares method. The system separately extracts the absolute Z-axis height data of the single-sided (upper or lower) laser sensor from the underlying 3D data, calculates the peak-to-valley range of the normal deviation of each measurement point, and simultaneously outputs the micro-flatness and warpage index of the single surface of the product.

[0037] S6. Output the thickness, thickness consistency index, and flatness / warping index at each point. After the entire process of testing is completed, the multi-dimensional results such as absolute thickness, thickness consistency angle, flatness / warping are displayed in real time on the display component, presenting the test results intuitively.

[0038] Work steps: First, the operator places the product to be tested (such as silicon wafers, glass, thin metal sheets, packaging materials, etc.) flat on the vacuum chuck or flexible stage of the Y-axis module. Based on the product size (e.g., a 150mm wafer, an A4-sized sheet of paper) and testing standards, a 5-point measurement mode (1 point at the center, 4 points near the four corners) is preset in the accompanying host computer software. The software controls a laser to project a crosshair laser mark above the stage. The operator manually or with visual assistance places the product in its approximate position. Then, the software drives the Y-axis module to precisely position the crosshair to the theoretical center of each of the 5 preset points. The operator then uses an erasable marker to mark the crosshair on the actual object, serving as the positioning reference for subsequent scanning. The positioning accuracy in this step depends on the motion module; the deviation between the crosshair reference and the actual theoretical point can be controlled within ±0.1mm.

[0039] Next, the control software initiates the automatic detection process. The system drives the X and Y axis modules to move the center of the laser detection component directly above the crosshair reference point, with a positioning accuracy equal to the repeatability of the motion system (±8μm). The system controls the X and Y axis modules to execute a 5mm×5mm grid scanning path centered on the crosshair reference point (e.g., a step of 0.2mm, acquiring a total of 26×26=676 points). Data is acquired briefly at each point (e.g., 5ms). At each point, the motion control card reads the physical coordinates (X,Y) fed back by the current encoder in real time, and simultaneously triggers and reads the height values ​​of the upper and lower sensors. The control system timestamps and binds these three sets of data (X,Y,Z), storing them in an array. For a micro-area scan of a single core point, it can be completed in approximately 3-4 seconds, generating a 3D point cloud containing surface topography information of that area.

[0040] Then, the data is processed through S3 (single-point thickness feature extraction, including outlier removal and mean calculation), S4 (thickness consistency assessment, including plane fitting and angle calculation), and S5 (single-sided flatness assessment, including reference plane components and peak-valley value calculation). Finally, all calculation results (absolute thickness values ​​at 5 points, thickness consistency angle, and upper surface flatness value) are displayed in real time via a display component. The software interface presents the results intuitively in the form of numbers and charts (such as thickness distribution contour maps and 3D topographic diagrams), and supports the generation of inspection reports containing raw data and judgment results.

[0041] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A thickness and flatness detection device based on a through-beam method, characterized in that, It includes a base plate (1), a protective cover (2), a dual-axis module (3), a laser detection component (4), a display component, and a control component (5).

2. The thickness and flatness detection device based on a through-beam method according to claim 1, characterized in that, The device is a desktop structure with dimensions less than 500mm × 600mm × 500mm.

3. The thickness and flatness detection device based on a through-beam method according to claim 1, characterized in that, The dual-axis module (3) adopts an independent and separate dual-axis module design, consisting of an X-axis module and a Y-axis module. The two are positively linked to achieve full-point scanning. The Y-axis module is fixed above the base plate; the X-axis module is vertically installed on the back plate of the equipment, and the moving end is connected to a "C-shaped" support bracket to directly drive the laser detection component to complete the X-axis scanning in the left and right directions.

4. The thickness and flatness detection device based on a through-beam method according to claim 1, characterized in that, The laser detection component (4) adopts an up-and-down through-beam laser detection architecture.

5. A method for thickness and flatness detection based on a through-beam method, using the thickness and flatness detection device based on a through-beam method as described in any one of claims 1-4, characterized in that, Includes the following steps: S1. Mark the positioning references at multiple core detection points preset on the product surface; S2. Control the dual-axis motion module to drive the through-beam laser detection component to move sequentially to the area above each of the positioning references, and scan each area, simultaneously collecting the two-dimensional mechanical coordinates (X,Y) of each scanning point and the Z-axis height value measured by the upper and lower laser displacement sensors to form a three-dimensional coordinate point cloud (X,Y,Z). S3. For each core detection point, calculate the average height difference based on multiple sets of upper and lower laser height values ​​collected within its area to obtain the thickness characteristic value of that point. S4. Based on the thickness characteristic values ​​and planar coordinates of all core points, construct a spatial fitting plane for the product surface, and calculate the angle between the fitting plane and the theoretical reference plane as a thickness consistency evaluation index. S5. Extract all Z-axis height data from the three-dimensional coordinate point cloud data collected in step S2, construct an ideal reference plane for the side surface, and calculate the peak-valley range of the height deviation of each measurement point relative to the reference plane as an index of the flatness or warping of the side surface. S6 outputs the thickness at each point, thickness consistency index, and flatness / warping index.

6. The thickness and flatness detection method based on a through-beam method according to claim 5, characterized in that, In S1, the operator places the product to be inspected stably in the inspection area, selects 5 core inspection points on the product surface according to the product specifications and inspection requirements, and marks a cross reference at the center of each point to provide a reference for subsequent accurate scanning and positioning.

7. The thickness and flatness detection method based on a through-beam method according to claim 5, characterized in that, In S2, the control component issues commands to drive the dual-axis module to accurately position itself at each detection point; at a single point, with the crosshair reference as the core, it performs high-frequency multi-point scanning and acquisition of the surrounding area according to a preset trajectory; During the scanning process, a deep fusion mechanism of hardware and software data is enabled: the control component collects the motion position commands of the X-axis and Y-axis modules or the encoder feedback values ​​in real time to generate the two-dimensional physical plane coordinates (X,Y) of each measurement point; and simultaneously reads the optical distance data of the upper and lower laser displacement sensors and maps it to a high-precision Z-axis normal height value. The control system aligns and deeply integrates the two-dimensional mechanical coordinates with the Z-axis optical data in real time to construct a spatial three-dimensional coordinate point cloud (X,Y,Z) containing product position and height information, providing accurate underlying data for subsequent algorithm processing.

8. The thickness and flatness detection method based on a through-beam method according to claim 5, characterized in that, In S3, before calculating the average height difference, there is also a step to remove outliers.

9. The thickness and flatness detection method based on a through-beam method according to claim 5, characterized in that, In S4, the spatial fitting plane is constructed using a spatial plane fitting algorithm.

10. The thickness and flatness detection method based on a through-beam method according to claim 5, characterized in that, In S5, the ideal reference plane is constructed using the least squares method.

Images