Glass cover plate internal stress detection system and method based on optical coherence tomography

Abstract

The application relates to the technical field of glass cover plate internal stress detection, and particularly discloses a glass cover plate internal stress detection system and method based on optical coherence tomography, wherein wide-spectrum light output by a sweep frequency light source is divided into reference light and detection light, the detection light is subjected to polarization modulation and then is incident to the glass cover plate, internal returned reflected light and the reference light are interfered to generate an interference signal; interference signal is collected, and spectral interference data of orthogonal polarization components are separated, a complex spectral matrix of each scanning point is constructed; inverse Fourier transform is performed on the complex spectral matrix to extract phase information, a change rate of a phase difference with a wave number is calculated to obtain group delay data; according to a corresponding relationship between the group delay data and birefringence, birefringence values of various depth positions are determined; the birefringence values are converted into stress tensors by using stress optical constants, and three-dimensional stress distribution is reconstructed; and the application realizes high-precision nondestructive detection of the internal stress of the glass cover plate.

Description

Technical Field

[0001] This invention relates to the field of internal stress detection technology for glass cover plates, and specifically to a system and method for internal stress detection of glass cover plates based on optical coherence tomography. Background Technology

[0002] As a crucial protective component, glass covers are widely used in smartphones, tablets, automotive displays, and display modules integrating smart sensors, leading to increasingly stringent requirements for their mechanical strength and reliability. During processes such as hot bending, chemical strengthening, and bonding assembly, residual stress is generated within the glass cover. Uneven distribution of this residual stress or localized stress concentration can cause microcracks or even breakage during use, severely impacting product yield and lifespan. Therefore, non-destructive, high-precision quantitative detection of the internal stress distribution of glass covers is of great significance for process optimization and quality control.

[0003] This invention addresses the problem of signal attenuation in the R-corner region of hot-bent curved glass cover plates, where drastic curvature changes make it difficult for probe light to be incident perpendicularly. Simultaneously, residual stress concentration in this region causes a phase delay of over 2π for the orthogonal polarization components, resulting in severe encapsulation. Existing phase unwrapping algorithms are prone to integration errors due to local phase abrupt changes when performing spatial path integration under low signal-to-noise ratio conditions, making it impossible to distinguish between real stress jumps and noise artifacts, ultimately leading to the failure of edge microcrack risk assessment. This invention provides a detection method based on wavelength tuning of a swept-frequency light source, replacing traditional spatial phase unfolding. By calculating the rate of change of phase difference with wavenumber, group delay data is directly obtained, fundamentally avoiding the dependence of phase encapsulation on unwrapping algorithms and achieving accurate reconstruction of stress in the curved edge region. Summary of the Invention

[0004] The purpose of this invention is to provide a system and method for detecting internal stress in glass cover plates based on optical coherence tomography, so as to solve the problems mentioned above.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A method for detecting internal stress in glass cover plates based on optical coherence tomography includes the following steps:

[0007] S1: The broadband light output from the sweep frequency light source is divided into reference light and probe light. The probe light is polarized and modulated before being incident on the glass cover plate. The reflected light returning from different depths inside the glass cover plate is collected, and the reflected light interferes with the reference light to generate an interference signal.

[0008] S2: Acquire interference signals and separate the spectral interference data of orthogonal polarization components, and construct the complex spectral matrix of each scanning point at different wavelengths;

[0009] S3: Perform an inverse Fourier transform on each depth position in the complex spectral matrix to extract the phase information of the orthogonal polarization components, and calculate the rate of change of the phase difference with the wavenumber to obtain the group delay data.

[0010] S4: Determine the birefringence values ​​at various depths inside the glass cover plate based on the correspondence between group delay data and birefringence.

[0011] S5: Using the stress optical constant, the birefringence value is converted into a stress tensor to reconstruct the three-dimensional stress distribution inside the glass cover.

[0012] As a further aspect of the present invention: the generation of the interference signal specifically includes:

[0013] The swept frequency light source outputs broadband light, which is then divided into reference light and probe light;

[0014] The probe light is polarized and modulated before being incident on the glass cover plate. The reflected light returned from each layer inside the glass cover plate is collected sequentially along the depth direction.

[0015] The reflected light and the reference light are combined, and wavelength tuning is used to separate the interference signals of different wavelengths in the time domain.

[0016] By acquiring the separated interference signals through balanced detection, an interference signal containing orthogonal polarization components and depth information is generated.

[0017] As a further aspect of the present invention: S2 specifically includes:

[0018] The interference signals are rearranged into a spectral sequence according to wavelength order;

[0019] For each wavelength point in the spectral sequence, the temporal peak position of the interference signal is detected, and the truncation window is dynamically adjusted according to the temporal peak position to extract the effective interference segment corresponding to each depth from the spectral sequence;

[0020] The effective interference segment is subjected to polarization beam splitting to separate two mutually orthogonal polarization components.

[0021] The two orthogonal polarization components are filled into the row and column positions of the complex spectral matrix according to the scanning point and wavelength, respectively, to form the complex spectral matrix corresponding to each scanning point.

[0022] As a further aspect of the present invention: the extraction of effective interference segments corresponding to each depth from the spectral sequence specifically includes:

[0023] The starting point of the interference signal is determined based on the peak position in the time domain. The first half of the interference signal from the starting point to the peak position is spliced ​​together with the second half after the peak position to form a symmetrical interception window centered on the peak position.

[0024] The amplitude fluctuation of the interference signal at each wavelength point near the peak is measured, and the length of the symmetrical interception window is dynamically adjusted according to the amplitude fluctuation to match the window coverage with the coherence length of the interference signal.

[0025] The data segment corresponding to the depth is extracted from the spectral sequence according to the adjusted symmetrical truncation window and output as the effective interference segment corresponding to the depth.

[0026] As a further aspect of the present invention: S3 specifically includes:

[0027] An inverse Fourier transform is applied to the spectral data at each depth position in the complex spectral matrix to obtain a complex signal in the depth direction. The principal phase values ​​of the orthogonal polarization components are then extracted from the complex signal in the depth direction.

[0028] The phase principal values ​​of adjacent wavenumbers are compared point by point along the wavenumber direction. The positions where the phase principal values ​​jump are detected. The phase principal values ​​are then corrected in segments based on the jump positions to obtain continuous phase absolute values.

[0029] Linearly fit the continuous absolute phase values ​​to the corresponding wavenumbers, calculate the slope of the linearly fitted line, and use the slope as the rate of change of the phase difference of the orthogonal polarization components with the wavenumber. The output is the group delay data.

[0030] As a further aspect of the present invention: the step of segmenting and correcting the principal phase value based on the transition position to obtain continuous absolute phase values ​​specifically includes:

[0031] Read the phase principal values ​​corresponding to two adjacent wavenumbers sequentially along the wavenumber direction, calculate the absolute value of the difference between adjacent phase principal values, compare the absolute value of the difference with a preset threshold, and mark the position where the absolute value of the difference exceeds the preset threshold as the jump position;

[0032] Based on the principal phase value of the first wavenumber, the phase values ​​are accumulated segment by segment starting from the initial wavenumber. Whenever a transition position is reached, the principal phase value of all subsequent wavenumbers is increased or decreased by a fixed period based on the direction of the transition, so that the phase connection between adjacent segments is continuous.

[0033] The phase principal value after periodic correction is mapped to the actual physical wavelength of the current wavenumber to generate a continuous absolute phase value for linear fitting.

[0034] As a further aspect of the present invention: S4 specifically includes:

[0035] Extract discrete numerical sequences that vary with depth from group delay data;

[0036] The dispersion curve of the glass cover material is measured within the bandwidth of the swept frequency light source. The discrete numerical sequence is compensated point by point according to the dispersion curve to eliminate the contribution of the material's own dispersion to the group delay.

[0037] The compensated discrete numerical sequence is multiplied by a preset conversion factor to convert it into birefringence values ​​corresponding to each depth position.

[0038] As a further aspect of the present invention: S5 specifically includes:

[0039] Based on the birefringence values ​​at each depth position, the birefringence values ​​at different depths at the same lateral position are arranged in depth order to form the birefringence depth distribution curve for the corresponding lateral position.

[0040] Traverse all scanning points along the transverse scanning direction to obtain the birefringence depth distribution curve of each scanning point. Cross-compare the birefringence depth distribution curves of adjacent scanning points and determine the principal stress direction based on the correlation between the curves.

[0041] By inputting the principal stress direction and the corresponding birefringence value into the stress optical constant, the normal stress component along the principal stress direction and the shear stress component perpendicular to the principal stress direction are calculated respectively. All stress components are then stitched together in three dimensions according to the spatial position of the scanning point to generate the three-dimensional stress distribution inside the glass cover.

[0042] As a further aspect of the present invention: the step of cross-comparing the birefringence depth distribution curves of adjacent scanning points and determining the principal stress direction based on the correlation between the curves specifically includes:

[0043] Select the birefringence depth distribution curve of the current scanning point as the reference curve, select the birefringence depth distribution curve of the adjacent scanning point as the comparison curve, slide the comparison curve point by point along the depth direction and calculate the degree of coincidence with the reference curve, and record the degree of coincidence at each sliding position.

[0044] Extract the sliding direction and sliding distance corresponding to the peak value of the overlap ratio, and use the sliding direction as a candidate value of the principal stress direction between the current scanning point and the adjacent scanning point;

[0045] Traverse all adjacent scan points around the current scan point to obtain multiple candidate values ​​of principal stress directions, and output the candidate value of principal stress direction with the highest frequency as the principal stress direction of the current scan point.

[0046] An internal stress detection system for glass covers based on optical coherence tomography includes:

[0047] The interference signal generation module divides the broadband light output by the frequency sweep light source into reference light and probe light. After polarization modulation, the probe light is incident on the glass cover plate, and the reflected light returning from different depths inside the glass cover plate is collected. The reflected light interferes with the reference light to generate an interference signal.

[0048] The complex spectral matrix construction module acquires interference signals and separates the spectral interference data of orthogonal polarization components, constructing a complex spectral matrix for each scanning point at different wavelengths;

[0049] The group delay data extraction module performs an inverse Fourier transform on each depth position in the complex spectral matrix to extract the phase information of the orthogonal polarization components and calculates the rate of change of the phase difference with wavenumber to obtain the group delay data.

[0050] The birefringence value determination module determines the birefringence value at each depth position inside the glass cover plate based on the correspondence between group delay data and birefringence.

[0051] The three-dimensional stress distribution reconstruction module uses the stress optical constant to convert birefringence values ​​into stress tensors, thereby reconstructing the three-dimensional stress distribution inside the glass cover.

[0052] The beneficial effects of this invention are:

[0053] (1) This invention employs a swept-frequency light source combined with wavelength tuning technology. By calculating the rate of change of the phase difference of the orthogonal polarization components with wavenumber, group delay data is obtained. Wavelength variation is used to replace traditional spatial phase unfolding, effectively avoiding the phase wrapping problem caused by residual stress concentration at the edge of hot-bent glass. This method can accurately extract birefringence values ​​in the curved surface R-angle region with extremely low signal-to-noise ratio, eliminating the pseudo-steps and discontinuous jumps caused by path integral errors in traditional algorithms, and realizing a reliable assessment of the risk of microcracks at the edge of curved glass cover plates.

[0054] (2) This invention constructs a complex spectral matrix and performs phase continuity correction along the wavenumber direction. Combined with material dispersion compensation and correlation analysis of birefringence curves of adjacent scanning points, it can accurately determine the principal stress direction at each depth. This method subtracts the material's own dispersion contribution point by point, converts the compensated group delay into birefringence value, and then calculates the normal stress and shear stress components by combining the stress optical constant. Finally, it reconstructs three-dimensional distribution data containing stress magnitude and direction, providing a complete quantitative basis for the process optimization and strength verification of glass cover plates. Attached Figure Description

[0055] The invention will now be further described with reference to the accompanying drawings.

[0056] Figure 1 This is a flowchart of the method of the present invention;

[0057] Figure 2 This is a system block diagram of the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Please see Figure 1 As shown, this invention is a method for detecting internal stress in glass cover plates based on optical coherence tomography, comprising the following steps:

[0060] S1: The broadband light output from the sweep frequency light source is divided into reference light and probe light. The probe light is polarized and modulated before being incident on the glass cover plate. The reflected light returning from different depths inside the glass cover plate is collected, and the reflected light interferes with the reference light to generate an interference signal.

[0061] S2: Acquire interference signals and separate the spectral interference data of orthogonal polarization components, and construct the complex spectral matrix of each scanning point at different wavelengths;

[0062] S3: Perform an inverse Fourier transform on each depth position in the complex spectral matrix to extract the phase information of the orthogonal polarization components, and calculate the rate of change of the phase difference with the wavenumber to obtain the group delay data.

[0063] S4: Determine the birefringence values ​​at various depths inside the glass cover plate based on the correspondence between group delay data and birefringence.

[0064] S5: Using the stress optical constant, the birefringence value is converted into a stress tensor to reconstruct the three-dimensional stress distribution inside the glass cover.

[0065] In S1, the broadband light output from the swept frequency source is divided into reference light and probe light. The probe light is polarized and modulated before being incident on the glass cover plate. Reflected light returning from different depths inside the glass cover plate is collected, and the reflected light interferes with the reference light to generate an interference signal. Specifically, this includes:

[0066] First, a swept-frequency light source is configured as the probe light output device, which outputs broadband light. The broadband light is split into two paths by a first fiber coupler. One path serves as the reference light entering the reference arm optical path, and the other path serves as the probe light entering the sample arm optical path. After entering the sample arm, the probe light is first polarized by a polarization controller and a linear polarizer to adjust the probe light to a linear polarization state. The adjusted probe light is then incident on the surface of the glass cover plate through a focusing objective lens.

[0067] Secondly, a two-dimensional scanning device consisting of galvanometers controls the incident point of the probe light on the surface of the glass cover. Simultaneously, a reference mirror in the reference arm moves stepwise along the optical axis to achieve depth scanning. When the probe light is focused to a certain depth inside the glass cover, the reflected light generated by the refractive index change at that depth returns along the original optical path, passing through the focusing objective and polarization controller before entering the first fiber coupler. During this process, the reference mirror position is moved sequentially from shallow to deep along the depth direction, thereby collecting the reflected light returned from each interface layer inside the glass cover in turn.

[0068] Next, the collected reflected light from each layer and the reference light are simultaneously input into the second fiber coupler for beam combining and interference. Since the output wavelength of the swept-frequency light source changes linearly with time, light of different wavelengths arrives at the detector at staggered intervals, causing the interference signals corresponding to different wavelengths to be naturally separated in the time domain. Specifically, when the swept-frequency light source completes a full scan of the wavelength from the initial value to the final value, the intensity of the interference signal received by the detector changes with time; this change curve contains the interference information at different wavelengths.

[0069] Finally, the combined interference light is input to a balanced detector for photoelectric conversion. The balanced detector consists of two photodiodes with identical performance parameters and a differential amplifier. The two photodiodes receive the positive and negative outputs of the interference signal, respectively, and the differential amplifier subtracts the two signals to output the result, thereby eliminating light source intensity noise and common-mode interference. The electrical signal output from the balanced detector is synchronously sampled by a data acquisition card and converted into a digitized interference signal. This interference signal contains information about the orthogonal polarization components and corresponds one-to-one with the reflection positions at various depths inside the glass cover, thus generating the interference signal for subsequent processing.

[0070] In S2, interference signals are acquired and spectral interference data of orthogonal polarization components are separated. Complex spectral matrices for each scanning point at different wavelengths are constructed, specifically including:

[0071] First, the raw interference signal data output from step one is input into a data processing unit. This unit receives the digital signal acquired by the balanced detector and converted from analog to digital, where each sampling point corresponds to a specific wavelength. Since the output wavelength of the swept-frequency light source varies linearly with time, the data processing unit rearranges the interference signal in ascending order of wavelength according to the sweep sequence of the light source, forming a spectral sequence with wavelength as the abscissa and interference intensity as the ordinate. This spectral sequence covers all wavelength points within one complete scan cycle of the swept-frequency light source, with each wavelength point corresponding to an interference intensity value.

[0072] Secondly, for the rearranged spectral sequence, the data processing unit detects the temporal peak position of the interference signal point by wavelength. The specific detection method is as follows: for each wavelength point, a series of sampling points acquired along the depth direction at that wavelength are extracted to form a one-dimensional depth signal; the maximum value point in this one-dimensional depth signal is found, and the depth position corresponding to the maximum value point is recorded as the temporal peak position of that wavelength point. After completing the peak detection for all wavelength points, a set of peak position data varying with wavelength is obtained.

[0073] Next, the spectral sequence is truncated based on the peak position data to extract the effective interference segments corresponding to each depth. The truncating window is constructed as follows: for each depth position, data points of the same length are truncated forward and backward, centered on the peak position at that depth. The window length is determined based on the coherence length of the interference signal, specifically by measuring the amplitude fluctuation at each wavelength point near the peak. The amplitude fluctuation is measured as follows: taking the interference intensity values ​​of 10 wavelength points before and after the peak position, and calculating the standard deviation of these intensity values; comparing the standard deviation with a preset fluctuation threshold, and gradually expanding the window range when the standard deviation is less than the fluctuation threshold until the standard deviation reaches the fluctuation threshold. The window length at this point is the truncating length matching the coherence length. The fluctuation threshold is set as follows: a segment of background noise without a sample is collected, the standard deviation of the background noise is calculated, and three times the standard deviation of the background noise is used as the fluctuation threshold.

[0074] Then, the captured effective interference segment undergoes polarization beam splitting. Since the interference signal contains the superposition information of two orthogonal polarization components, the data processing unit separates them using digital signal processing. Specifically, the data of the effective interference segment is input into two parallel processing channels. One channel retains the original signal as the first orthogonal polarization component, while the other channel performs a 90-degree phase shift on the original signal using a Hilbert transform to obtain the second orthogonal polarization component. These two components are mathematically orthogonal, corresponding to the light intensity information in two orthogonal polarization directions.

[0075] Finally, the two separated orthogonal polarization components are arranged according to the spatial order of the scanning points and the increasing order of the wavelengths to construct a complex spectral matrix. Specifically, the scanning points are used as row indices and the wavelengths as column indices. Data from the first polarization component is filled into the real part, and data from the second polarization component is filled into the imaginary part, thus forming a complex spectral matrix corresponding to each scanning point. Each element of this matrix is ​​a complex number, with the real part representing the intensity of one polarization component and the imaginary part representing the intensity of the other orthogonal polarization component, fully preserving the polarization interference information of each scanning point at different wavelengths.

[0076] In S3, an inverse Fourier transform is performed on each depth position in the complex spectral matrix to extract the phase information of the orthogonal polarization components, and the rate of change of phase difference with wavenumber is calculated to obtain group delay data, specifically including:

[0077] First, a complex spectral matrix is ​​received, with scan points as rows and wavelengths as columns, where each element is a complex number containing both real and imaginary parts. For each scan point, the data processing unit extracts all the complex data corresponding to that point along the wavelength direction, forming a spectral sequence for that point. An inverse Fourier transform is applied to the spectral sequence, specifically implemented using an inverse fast Fourier transform algorithm, converting the spectral domain data into depth domain data to obtain a complex signal in the depth direction. Each depth position of this complex signal corresponds to a complex number, and the argument of this complex number is the principal phase value of the orthogonal polarization component at that depth. The range of the principal phase value is restricted to between negative and positive pi, i.e., between -180 degrees and +180 degrees.

[0078] Secondly, the data processing unit performs continuous correction on the principal phase value along the wavenumber direction. The wavenumber is obtained by converting the wavelength by dividing 2π by the current wavelength. For each depth position, the data processing unit reads the principal phase value corresponding to each wavenumber at that depth in ascending order of wavenumber. Starting from the second wavenumber, the difference between the current wavenumber's principal phase value and the previous wavenumber's principal phase value is calculated, and the absolute value of this difference is taken. This absolute value is compared with a preset jump threshold. The jump threshold is set as follows: an interference signal of a stress-free glass sample is acquired, its principal phase value is extracted using the same steps, the average difference between adjacent wavenumber principal phase values ​​is calculated, and five times this average value is taken as the jump threshold. If the absolute value of the difference exceeds the jump threshold, the position is marked as a jump position, and the direction of change of the current wavenumber's principal phase value relative to the previous wavenumber's principal phase value is recorded, i.e., whether the difference is positive or negative.

[0079] Next, the data processing unit performs segmented correction on the principal phase values ​​after the marked transition positions. The principal phase value corresponding to the first wavenumber is used as the starting reference value. Starting from the first wavenumber, the system iterates backward. When encountering the first transition position, the direction of change at that transition position is determined: if the direction is positive, the phase value corresponding to 2π is subtracted from the principal phase value of all wavenumbers after the current transition position; if the direction is negative, the phase value corresponding to 2π is added to the principal phase value of all wavenumbers after the current transition position. After correcting the first transition position, the corrected principal phase value of the current wavenumber is used as the new reference, and the system continues to iterate backward. When encountering the next transition position, the phase value corresponding to 2π is again added or subtracted from the principal phase value of all subsequent wavenumbers based on the direction of change. This process is repeated until all wavenumbers have been traversed, obtaining the continuous absolute phase value corresponding to each wavenumber. This continuous absolute phase value is no longer limited by the positive or negative value of pi and can truly reflect the cumulative change in phase with wavenumbers.

[0080] Finally, the data processing unit performs linear fitting between the corrected continuous phase absolute values ​​and the corresponding wavenumbers. The linear fitting is achieved using the least squares method: first, the average value of all wavenumbers and the average value of all continuous phase absolute values ​​are calculated; then, the difference between each wavenumber and the average wavenumber, and the difference between each continuous phase absolute value and the average phase are calculated; these wavenumber differences are multiplied by their corresponding phase differences, summed, and divided by the sum of the squares of the wavenumber differences to obtain the slope of the fitted line. This slope value represents the rate of change of the phase difference of the orthogonal polarization components with respect to the wavenumber. The data processing unit outputs this slope value as the group delay data.

[0081] In S4, based on the correspondence between group delay data and birefringence, the birefringence values ​​at various depths inside the glass cover are determined, specifically including:

[0082] First, group delay data is received. This group delay data forms a discrete numerical sequence that varies with depth, with depth as the independent variable and group delay value as the dependent variable. The data processing unit extracts the group delay value corresponding to each depth position in ascending order of depth and arranges these group delay values ​​into a one-dimensional array, which serves as the input data for subsequent compensation processing. The sampling interval in the depth direction is determined by the wavelength scanning range of the swept-frequency light source and the resolution of the inverse Fourier transform, specifically half the reciprocal of the wavelength scanning range.

[0083] Secondly, the data processing unit performs material dispersion compensation on the extracted discrete numerical sequence. Before testing the glass cover, the dispersion curve of the glass material under test is measured within the bandwidth of the swept-frequency light source. The method for measuring the dispersion curve is as follows: a standard glass sample of the same material as the glass cover under test and without internal stress is taken, and its group delay data is collected. This group delay data is used as the background group delay caused by the material's own dispersion and recorded as the dispersion compensation reference curve. In the actual test, for each depth position, the group delay value is subtracted from the group delay value of the corresponding depth in the dispersion compensation reference curve to obtain the net group delay value after deducting the influence of material dispersion. After performing the subtraction operation point by point, a set of compensated discrete numerical sequences contributed only by stress birefringence is obtained.

[0084] Finally, the data processing unit converts the compensated discrete numerical sequence into birefringence values. The conversion process is based on the physical relationship between group delay and birefringence: the group delay equals the birefringence value multiplied by the optical path length corresponding to the current depth, then divided by the speed of light. The optical path length is determined by multiplying the refractive index of the glass cover by the depth value. In actual calculations, each value in the compensated discrete numerical sequence is first divided by the refractive index of the glass cover, then by the speed of light in a vacuum, to obtain a preliminary birefringence value. Since the speed of light and refractive index involved in the calculation are constants, these constants can be combined into a conversion coefficient. The specific value of the conversion coefficient is calibrated as follows: a standard stress sample with a known birefringence value is taken, and its compensated group delay value is measured following the same steps. The known birefringence value is divided by this compensated group delay value to obtain the actual value of the conversion coefficient. Each value in the compensated discrete numerical sequence is multiplied by this conversion coefficient to obtain the birefringence value corresponding to each depth position. The data processing unit arranges the calculated birefringence values ​​in depth order and outputs the birefringence depth distribution data for subsequent stress tensor reconstruction.

[0085] In S5, the birefringence value is converted into a stress tensor using the stress optical constant, and the three-dimensional stress distribution inside the glass cover is reconstructed, specifically including:

[0086] First, the birefringence values ​​at each depth location are received. For each lateral scan point, the birefringence values ​​at all depth locations below that point are arranged in ascending order of depth, forming a curve with depth as the independent variable and birefringence values ​​as the dependent variable. This curve is called the birefringence depth distribution curve for that lateral point. The sampling interval for the depth coordinates is determined by the parameters of the sweep frequency light source and is on the order of micrometers. After constructing the curves for all scan points, a set of birefringence depth distribution curves covering the entire detection area of ​​the glass cover is obtained.

[0087] Secondly, the data processing unit traverses all scanning points along the transverse scanning direction, performing a stress principal direction determination operation for each current scanning point. Specifically, the birefringence depth distribution curve of the current scanning point is selected as the reference curve, and the birefringence depth distribution curves of the eight adjacent scanning points (upper, lower, left, right, and four diagonal directions) are selected as comparison curves. For each adjacent direction, the comparison curve is slid point-by-point relative to the reference curve along the depth direction. At each depth step, the overlap ratio of the two curves at the current sliding position is calculated. The overlap ratio is calculated using the normalized cross-correlation coefficient, and its formula is as follows:

[0088] ;

[0089] in, Indicates the sliding distance as The degree of overlap at the time The value range is from the negative maximum sliding distance to the positive maximum sliding distance, and is taken as 1 / 3 of the total number of points in the depth direction; Indicates the reference curve at the 1st Birefringence values ​​at a depth point; This represents the arithmetic mean of the birefringence values ​​at all depth points of the reference curve. This indicates the comparison curve at the th The birefringence value at a depth point, when If the point is outside the range of the comparison curve, it will not be included in the calculation. This represents the arithmetic mean of the birefringence values ​​at depth points in the part of the comparison curve that are involved in the calculation. This indicates the number of depth points where the reference curve and the comparison curve coincide at the current sliding position. This value is calculated using this formula. The closer the value is to 1, the higher the similarity in shape between the two curves at that sliding position.

[0090] Then, the data processing unit records each sliding position. Corresponding overlap and find all Chinese envoy Sliding distance to reach maximum value The sliding direction corresponding to the maximum overlap (i.e. The positive or negative sign of the value is used as a candidate value for the principal stress direction between the current scan point and the adjacent point. Specifically, if... A positive value indicates that the comparison curve has slid downwards relative to the reference curve, and the principal stress direction points towards that adjacent point; if... A negative value indicates upward sliding, meaning the principal stress direction is opposite to the adjacent point. After comparing all eight adjacent directions of the current scan point, a set of candidate principal stress directions is obtained. The direction with the highest frequency in this set of candidate values ​​is determined as the principal stress direction of the current scan point. If multiple directions have the same frequency, the average of these directions is taken as the final principal stress direction.

[0091] Next, the data processing unit calculates the stress tensor components based on the determined principal stress directions and corresponding birefringence values. The stress optical constant is an inherent property of the glass material, obtained through pre-calibration, and is denoted as... For each scan point, the birefringence value at that point... This reflects the principal stress difference at that location. Considering the principal stress directions, the birefringence value is decomposed into normal stress components along the principal stress directions. and shear stress components perpendicular to the principal stress direction The specific calculation method is as follows: the normal stress component equals the birefringence value multiplied by the stress optical constant, i.e. The shear stress component is numerically equal to half the normal stress component, that is... This is because, for isotropic materials, the maximum shear stress is equal to half the difference between the principal stresses. The above calculation process is performed point by point and depth by depth to obtain the normal stress and shear stress values ​​for each voxel.

[0092] Finally, the data processing unit performs three-dimensional stitching of the stress components from all scanned points according to their spatial locations. The spatial location is uniquely determined by the transverse scan coordinates (X and Y directions) and the depth coordinates (Z direction). For each spatial coordinate point, the calculated normal stress and shear stress components are stored in their corresponding positions in a three-dimensional array, forming two independent three-dimensional stress component matrices. Merging these two matrices yields the complete three-dimensional stress distribution data inside the glass cover. This data can be displayed as a pseudo-color image or exported as a numerical table for subsequent process analysis and quality assessment. This completes the entire reconstruction process from the original interference signal to the three-dimensional stress distribution inside the glass cover.

[0093] Please see Figure 2 As shown, the internal stress detection system for glass cover plates based on optical coherence tomography includes:

[0094] The interference signal generation module divides the broadband light output by the frequency sweep light source into reference light and probe light. After polarization modulation, the probe light is incident on the glass cover plate, and the reflected light returning from different depths inside the glass cover plate is collected. The reflected light interferes with the reference light to generate an interference signal.

[0095] The complex spectral matrix construction module acquires interference signals and separates the spectral interference data of orthogonal polarization components, constructing a complex spectral matrix for each scanning point at different wavelengths;

[0096] The group delay data extraction module performs an inverse Fourier transform on each depth position in the complex spectral matrix to extract the phase information of the orthogonal polarization components and calculates the rate of change of the phase difference with wavenumber to obtain the group delay data.

[0097] The birefringence value determination module determines the birefringence value at each depth position inside the glass cover plate based on the correspondence between group delay data and birefringence.

[0098] The three-dimensional stress distribution reconstruction module uses the stress optical constant to convert birefringence values ​​into stress tensors, thereby reconstructing the three-dimensional stress distribution inside the glass cover.

[0099] The working principle of this invention is as follows: The broadband light output from the swept frequency light source is divided into a reference light and a probe light. The probe light is polarized and modulated before being incident on a glass cover plate. The reflected light returning from different depths inside the glass cover plate is collected and interfered with the reference light to generate an interference signal. The interference signal is acquired and the spectral interference data of the orthogonal polarization components are separated to construct a complex spectral matrix for each scanning point at different wavelengths. An inverse Fourier transform is performed on each depth position in the complex spectral matrix to extract the phase information of the orthogonal polarization components and calculate the rate of change of the phase difference with wavenumber to obtain group delay data. The birefringence value at each depth position is determined according to the correspondence between the group delay data and birefringence. Finally, the birefringence value is converted into a stress tensor using the stress optical constant to reconstruct the three-dimensional stress distribution inside the glass cover plate.

[0100] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for detecting internal stress in a glass cover plate based on optical coherence tomography, characterized in that, Includes the following steps: S1: The broadband light output from the sweep frequency light source is divided into reference light and probe light. The probe light is polarized and modulated before being incident on the glass cover plate. The reflected light returning from different depths inside the glass cover plate is collected, and the reflected light interferes with the reference light to generate an interference signal. S2: Acquire interference signals and separate the spectral interference data of orthogonal polarization components, and construct the complex spectral matrix of each scanning point at different wavelengths; S3: Perform an inverse Fourier transform on each depth position in the complex spectral matrix to extract the phase information of the orthogonal polarization components, and calculate the rate of change of the phase difference with the wavenumber to obtain the group delay data. S4: Determine the birefringence values ​​at various depths inside the glass cover plate based on the correspondence between group delay data and birefringence. S5: Using the stress optical constant, the birefringence value is converted into a stress tensor to reconstruct the three-dimensional stress distribution inside the glass cover.

2. The method for detecting internal stress of a glass cover plate based on optical coherence tomography according to claim 1, characterized in that, The generation of the interference signal specifically includes: The swept frequency light source outputs broadband light, which is then divided into reference light and probe light; The probe light is polarized and modulated before being incident on the glass cover plate. The reflected light returned from each layer inside the glass cover plate is collected sequentially along the depth direction. The reflected light and the reference light are combined, and wavelength tuning is used to separate the interference signals of different wavelengths in the time domain. By acquiring the separated interference signals through balanced detection, an interference signal containing orthogonal polarization components and depth information is generated.

3. The method for detecting internal stress of a glass cover plate based on optical coherence tomography according to claim 1, characterized in that, S2 specifically includes: The interference signals are rearranged into a spectral sequence according to wavelength order; For each wavelength point in the spectral sequence, the temporal peak position of the interference signal is detected, and the truncation window is dynamically adjusted according to the temporal peak position to extract the effective interference segment corresponding to each depth from the spectral sequence; The effective interference segment is subjected to polarization beam splitting to separate two mutually orthogonal polarization components. The two orthogonal polarization components are filled into the row and column positions of the complex spectral matrix according to the scanning point and wavelength, respectively, to form the complex spectral matrix corresponding to each scanning point.

4. The method for detecting internal stress of a glass cover plate based on optical coherence tomography according to claim 3, characterized in that, The extraction of effective interference segments corresponding to each depth from the spectral sequence specifically includes: The starting point of the interference signal is determined based on the peak position in the time domain. The first half of the interference signal from the starting point to the peak position is spliced ​​together with the second half after the peak position to form a symmetrical interception window centered on the peak position. The amplitude fluctuation of the interference signal at each wavelength point near the peak is measured, and the length of the symmetrical interception window is dynamically adjusted according to the amplitude fluctuation to match the window coverage with the coherence length of the interference signal. The data segment corresponding to the depth is extracted from the spectral sequence according to the adjusted symmetrical truncation window and output as the effective interference segment corresponding to the depth.

5. The method for detecting internal stress of a glass cover plate based on optical coherence tomography according to claim 1, characterized in that, S3 specifically includes: An inverse Fourier transform is applied to the spectral data at each depth position in the complex spectral matrix to obtain a complex signal in the depth direction. The principal phase values ​​of the orthogonal polarization components are then extracted from the complex signal in the depth direction. The phase principal values ​​of adjacent wavenumbers are compared point by point along the wavenumber direction. The positions where the phase principal values ​​jump are detected. The phase principal values ​​are then corrected in segments based on the jump positions to obtain continuous phase absolute values. Linearly fit the continuous absolute phase values ​​to the corresponding wavenumbers, calculate the slope of the linearly fitted line, and use the slope as the rate of change of the phase difference of the orthogonal polarization components with the wavenumber. The output is the group delay data.

6. The method for detecting internal stress of a glass cover plate based on optical coherence tomography according to claim 5, characterized in that, The step of segmenting and correcting the principal phase value based on the transition position to obtain continuous absolute phase values ​​specifically includes: Read the phase principal values ​​corresponding to two adjacent wavenumbers sequentially along the wavenumber direction, calculate the absolute value of the difference between adjacent phase principal values, compare the absolute value of the difference with a preset threshold, and mark the position where the absolute value of the difference exceeds the preset threshold as the jump position; Based on the principal phase value of the first wavenumber, the phase values ​​are accumulated segment by segment starting from the initial wavenumber. Whenever a transition position is reached, the principal phase value of all subsequent wavenumbers is increased or decreased by a fixed period based on the direction of the transition, so that the phase connection between adjacent segments is continuous. The phase principal value after periodic correction is mapped to the actual physical wavelength of the current wavenumber to generate a continuous absolute phase value for linear fitting.

7. The method for detecting internal stress of a glass cover plate based on optical coherence tomography according to claim 1, characterized in that, S4 specifically includes: Extract discrete numerical sequences that vary with depth from group delay data; The dispersion curve of the glass cover material is measured within the bandwidth of the swept frequency light source. The discrete numerical sequence is compensated point by point according to the dispersion curve to eliminate the contribution of the material's own dispersion to the group delay. The compensated discrete numerical sequence is multiplied by a preset conversion factor to convert it into birefringence values ​​corresponding to each depth position.

8. The method for detecting internal stress of a glass cover plate based on optical coherence tomography according to claim 1, characterized in that, S5 specifically includes: Based on the birefringence values ​​at each depth position, the birefringence values ​​at different depths at the same lateral position are arranged in depth order to form the birefringence depth distribution curve for the corresponding lateral position. Traverse all scanning points along the transverse scanning direction to obtain the birefringence depth distribution curve of each scanning point. Cross-compare the birefringence depth distribution curves of adjacent scanning points and determine the principal stress direction based on the correlation between the curves. Input the principal stress direction and the corresponding birefringence value into the stress optical constant, calculate the normal stress component along the principal stress direction and the shear stress component perpendicular to the principal stress direction, and stitch all stress components together in three dimensions according to the spatial position of the scanning point to generate the three-dimensional stress distribution inside the glass cover.

9. The method for detecting internal stress of a glass cover plate based on optical coherence tomography according to claim 8, characterized in that, The step of cross-comparing the birefringence depth distribution curves of adjacent scanning points and determining the principal stress direction based on the correlation between the curves specifically includes: Select the birefringence depth distribution curve of the current scanning point as the reference curve, select the birefringence depth distribution curve of the adjacent scanning point as the comparison curve, slide the comparison curve along the depth direction point by point and calculate the degree of coincidence with the reference curve, and record the degree of coincidence at each sliding position. Extract the sliding direction and sliding distance corresponding to the peak value of the overlap ratio, and use the sliding direction as a candidate value of the principal stress direction between the current scanning point and the adjacent scanning point; Traverse all adjacent scan points around the current scan point to obtain multiple candidate values ​​of principal stress directions, and output the candidate value of principal stress direction with the highest frequency as the principal stress direction of the current scan point.

10. A glass cover plate internal stress detection system based on optical coherence tomography, characterized in that, The method for detecting internal stress in a glass cover plate based on optical coherence tomography as described in any one of claims 1-9 includes: The interference signal generation module divides the broadband light output by the frequency sweep light source into reference light and probe light. After polarization modulation, the probe light is incident on the glass cover plate, and the reflected light returning from different depths inside the glass cover plate is collected. The reflected light interferes with the reference light to generate an interference signal. The complex spectral matrix construction module acquires interference signals and separates the spectral interference data of orthogonal polarization components, constructing a complex spectral matrix for each scanning point at different wavelengths; The group delay data extraction module performs an inverse Fourier transform on each depth position in the complex spectral matrix to extract the phase information of the orthogonal polarization components and calculates the rate of change of the phase difference with wavenumber to obtain the group delay data. The birefringence value determination module determines the birefringence value at each depth position inside the glass cover plate based on the correspondence between group delay data and birefringence. The three-dimensional stress distribution reconstruction module uses the stress optical constant to convert birefringence values ​​into stress tensors, thereby reconstructing the three-dimensional stress distribution inside the glass cover.

Images