Method for precisely regulating stress of sodium-calcium-silicon c-type monolithic non-insulating fireproof glass
By constructing a thickness-based stress benchmark library and using laser speckle interferometry detection, combined with a closed-loop control method for wind pressure-stress regression coefficients, the problems of stress control relying on manual experience and lack of thickness-based control rules in existing technologies have been solved. This has enabled precise stress control of sodium-calcium-silicon Class C monolithic non-insulated fireproof glass, improving product quality and production stability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SICHUAN POLYFILL TECH CO LTD
- Filing Date
- 2026-03-31
- Publication Date
- 2026-06-23
Smart Images
Figure CN121933165B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of stress detection and analysis technology, and in particular to a method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulated fireproof glass. Background Technology
[0002] Sodium-calcium-silicon Class C monolayer non-insulated fireproof glass is a core load-bearing fireproof component in building fire compartments. Its mechanical stability and fire resistance integrity directly determine its safety protection performance in building fire scenarios. Thicknesses of 6mm, 8mm, 10mm, 12mm, 15mm, and 19mm are suitable for different building space requirements and have become the mainstream application products in the industry. Tempering is a key post-processing step that imparts fire resistance and mechanical strength to this type of glass. Precise control of surface compressive stress is a core technical point. The stress state directly affects the glass's impact resistance, spontaneous breakage risk, and structural stability at high temperatures. Deviating from a reasonable range can easily lead to insufficient fire resistance time or sudden breakage during service. In existing technologies, stress control of sodium-calcium-silicon Class C monolayer non-insulated fireproof glass generally relies on traditional tempering processes. A single or similar stress control target is used for glass of different thicknesses. Offline polarized stress meters are used for qualitative or semi-quantitative sampling inspection of the tempered finished product. Process personnel judge whether the tempering parameters need to be adjusted based on their experience. Only after adjustment and re-tempering can the results be verified.
[0003] The existing technology suffers from three core defects. First, the offline testing mode has inherent lag; the detection deviation of polarized stress meters generally exceeds 20%, making it impossible to capture dynamic stress changes in real time during tempering. Defective products must undergo secondary tempering, which weakens the internal structural integrity of the glass, increasing production costs and introducing new quality risks. Second, the existing technology lacks a differentiated stress index system based on thickness, applying a uniform control target to glass thicknesses from 6mm to 19mm. This results in insufficient stress on thick substrates and overload on thin substrates. Furthermore, the lack of edge stress optimization design leads to uneven stress distribution between the edges and center, easily causing stress concentration cracking, resulting in a product qualification rate of less than 85%. Third, even with some advanced technologies incorporating laser speckle interferometry to obtain a full-field stress distribution map, the process from detection results to wind pressure adjustment still relies on manual judgment by process personnel. This results in untimely control response and inaccurate control amounts, making it difficult to narrow the stress distribution deviation to within 15%. Summary of the Invention
[0004] This application provides a method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulated fireproof glass, which solves the problems in the prior art where stress control of sodium-calcium-silicon C-type monolithic non-insulated fireproof glass relies on manual experience judgment, lacks thickness-specific control rules, and suffers from continuous decay of control accuracy during long-term mass production. It improves the control accuracy of stress distribution across the entire surface of sodium-calcium-silicon C-type monolithic non-insulated fireproof glass of different thicknesses and the long-term control stability during industrial mass production.
[0005] This application provides a method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulating fireproof glass, the method comprising:
[0006] Step S1: Construct a stress benchmark library for different thicknesses based on the substrate thickness. The stress benchmark library for different thicknesses includes the target range of surface compressive stress and the wind pressure-stress regression coefficient corresponding to each substrate thickness.
[0007] Step S2: Obtain the full-field stress distribution map of the tempered substrate by laser speckle interferometry. Subtract the full-field stress distribution map from the target center value of the surface compressive stress of the corresponding thickness in the thickness stress reference library pixel by pixel to obtain the full-field stress deviation map.
[0008] Step S3: Divide the full-field stress deviation map into physical zones of the wind grating, multiply the average stress deviation of each zone by the wind pressure-stress regression coefficient to obtain the wind pressure compensation amount of each zone, and implement gradient cooling for each zone's wind grating accordingly.
[0009] Step S4: Obtain a second full-field stress distribution map through laser speckle interferometry. Subtract the second full-field stress distribution map from the target center value of the surface compressive stress to obtain a second full-field stress deviation map. Use the average ratio of the second stress deviation of each zone to the target center value of the surface compressive stress as the stress distribution deviation rate. When the stress distribution deviation rate does not exceed 8%, allow it to cool naturally to room temperature to obtain the finished product.
[0010] The technical solution provided in this application constructs a stress benchmark library for different thicknesses, including the target range of surface compressive stress and the wind pressure-stress regression coefficient, based on the substrate thickness. The stress control standards corresponding to the six thicknesses of substrates from 6mm to 19mm are stored independently, allowing for automatic matching of control criteria when switching production between products of different thicknesses. This fundamentally eliminates the problems of insufficient stress in thick substrates and overload in thin substrates caused by a uniform stress target in existing technologies. After obtaining the full-field stress distribution map through laser speckle interferometry, the map is subtracted pixel-by-pixel from the center value of the surface compressive stress target corresponding to the thickness in the stress benchmark library to obtain the full-field stress deviation map. This transforms the original reliance on subjective judgment by process engineers to determine "where the stress is too high or too low" into a quantitative deviation value covering every physical location on the entire board surface. The spatial integrity and numerical accuracy of the detection results provide reliable data input for subsequent zoned wind pressure compensation calculations. Based on this, the stress deviation map of the entire field is divided into physical zones according to the wind grid. The wind pressure compensation amount of each zone is obtained by multiplying the average stress deviation of each zone by the wind pressure-stress regression coefficient. This product calculation automatically converts the stress deviation into the wind pressure adjustment amount required for the corresponding zone, completely eliminating the black box of manual experience between the detection results and the control actions. Moreover, the monotonically decreasing law of the regression coefficient of each thickness ensures that the wind pressure compensation amount calculation results under different thickness specifications are accurately matched with the actual wind pressure-stress response characteristics of each thickness substrate. After verifying that the stress distribution deviation rate does not exceed 8% by laser speckle interferometry, the product can enter natural cooling. When the deviation rate exceeds 8%, the wind pressure compensation amount is recalculated based on the stress deviation map of the entire field again and the product returns to the gradient cooling cycle. This closed-loop verification mechanism ensures that the stress distribution of the entire plate surface of each batch of finished products is quantitatively verified before it can be released from the furnace, thus preventing the outflow of unqualified products from the system.
[0011] The wind pressure-stress regression coefficient, as the core algorithm feature of this invention, contributes to the precise stress control of Class C monolithic non-insulated fireproof soda-lime glass on two levels: First, the coefficient quantifies the physical response relationship between the cooling rate and surface compressive stress of soda-lime glass at different thicknesses, transforming stress control from qualitative empirical judgment to quantitative calculation based on physical laws. The algorithm input is the average stress deviation of each zone, and the output is the wind pressure compensation amount of each zone. A clear linear mapping relationship is established between the input and output through the thickness-specific regression coefficient, making the algorithm logically clear and operable. Second, combined with the online update mechanism of the recursive least squares method, the coefficient can be automatically corrected as the furnace condition drifts during long-term mass production, enabling the algorithm to have adaptive capabilities in the actual industrial production environment and avoiding the inherent defect of the fixed parameter model where the control accuracy continuously decreases after equipment aging. Attached Figure Description
[0012] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0013] Figure 1 This is a schematic diagram of an embodiment of the method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulated fireproof glass in this application.
[0014] Figure 2 This is a schematic diagram of the twenty-four zones of the full-field stress deviation spectrum in the embodiments of this application. Detailed Implementation
[0015] This application provides a method for precise stress control of a sodium-calcium-silicon Class C monolithic non-insulated fireproof glass. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0016] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulated fireproof glass in this application includes:
[0017] Step S1: Construct a stress benchmark library for different thicknesses based on substrate thickness. The stress benchmark library for different thicknesses includes the target range of surface compressive stress and the wind pressure-stress regression coefficient for each substrate thickness.
[0018] Specifically, the construction of the thickness-based stress benchmark library is based on the elastic mechanical properties of soda-lime-silicon glass. Due to the different heat conduction rates during the tempering and cooling process, the temperature gradient formed between the surface and the interior of substrates of different thicknesses is significantly different. The larger the thickness, the smaller the temperature gradient, and the lower the cooling air pressure required to form the same surface compressive stress. Therefore, each thickness of substrate must correspond to an independent surface compressive stress target range and air pressure-stress regression coefficient. The two are stored in the benchmark library in a one-to-one correspondence according to thickness, serving as the sole data source for the subsequent pixel-by-pixel subtraction calculation in step S2 and the product calculation in step S3.
[0019] Step S2: Obtain the full-field stress distribution map of the tempered substrate by laser speckle interferometry. Subtract the full-field stress distribution map from the target center value of the surface compressive stress of the corresponding thickness in the thickness stress reference library pixel by pixel to obtain the full-field stress deviation map.
[0020] Specifically, in the process of obtaining the full-field stress distribution map by laser speckle interferometry, the input of the digital image correlation algorithm is two frames of image data: a reference speckle image and a stressed speckle image. By performing correlation calculation on the gray-level distribution of the same pixels in the two frames, the displacement of each pixel before and after being stressed is derived, forming a full-field displacement field. The full-field displacement field is then converted into stress values by the generalized Hooke's law. The final output full-field stress distribution map shows the surface compressive stress value of each pixel at a physical location on the substrate, which directly corresponds to the pixel-by-pixel subtraction operation in step S2.
[0021] Step S3: Divide the stress deviation map of the whole field into physical zones of the wind grid, multiply the mean stress deviation of each zone by the wind pressure-stress regression coefficient to obtain the wind pressure compensation amount of each zone, and implement gradient cooling of each zone's wind grid accordingly.
[0022] Specifically, the physical meaning of the wind pressure-stress regression coefficient is: within the target stress range, the amount of wind pressure adjustment required to reduce the average stress deviation of a certain zone by one MPa, expressed as MPa wind pressure / MPa stress. This coefficient monotonically decreases with increasing thickness, reflecting the objective law that thicker glass is less sensitive to stress changes in wind pressure than thinner glass. When the average stress deviation of each zone is positive, the measured stress of that zone is higher than the target center value, the wind pressure compensation is positive, and the wind pressure of the corresponding zone must be reduced; when the average stress deviation of each zone is negative, the wind pressure compensation is negative, and the wind pressure of the corresponding zone must be increased, thereby achieving differentiated wind pressure control by zone.
[0023] Step S4: Obtain the full-field stress distribution map again through laser speckle interferometry. Subtract the full-field stress distribution map from the target center value of the surface compressive stress to obtain the full-field stress deviation map. Use the average ratio of the stress deviation of each zone to the target center value of the surface compressive stress as the stress distribution deviation rate. When the stress distribution deviation rate does not exceed 8%, allow it to cool naturally to room temperature to obtain the finished product.
[0024] Specifically, the stress distribution deviation rate is calculated by taking the average of the absolute values of the ratio of the stress deviation of each zone to the target center value of the surface compressive stress, which reflects the overall uniformity of the stress distribution across the entire plate after gradient cooling. When the stress distribution deviation rate exceeds 8%, the average value of each zone in the stress deviation map is substituted back into the wind pressure-stress regression coefficient to calculate a new round of wind pressure compensation, and the process returns to step S3 for repeated execution until the stress distribution deviation rate does not exceed 8%. This cyclic process does not require secondary heating and tempering; stress convergence is achieved solely through adjustments to cooling parameters.
[0025] In one specific embodiment, step S1 includes:
[0026] The target ranges of surface compressive stress for substrates with thicknesses from 6mm to 19mm are set sequentially according to the thickness gradient. The target ranges of surface compressive stress for each substrate thickness are: 130-155MPa for 6mm substrate, 155-180MPa for 8mm substrate, 180-205MPa for 10mm substrate, 195-230MPa for 12mm substrate, 230-260MPa for 15mm substrate, and 260-280MPa for 19mm substrate. The target ranges of surface compressive stress increase with the substrate thickness to form a continuous adaptive gradient, thus obtaining the target ranges of surface compressive stress for each substrate thickness.
[0027] The average of the upper and lower limits of the target range of surface compressive stress for substrates of different thicknesses is used as the center value of the target surface compressive stress to obtain the target parameters of compressive stress for each thickness.
[0028] Based on the differences in stress response of substrates of different thicknesses to wind pressure changes, and constrained by the monotonically decreasing wind pressure compensation amount corresponding to unit stress deviation as thickness increases, the wind pressure-stress regression coefficients corresponding to substrates of different thicknesses are calibrated. The target parameters of compressive stress by thickness and the wind pressure-stress regression coefficients are stored in a one-to-one correspondence according to thickness to obtain the stress benchmark library by thickness.
[0029] Specifically, the higher the surface compressive stress of soda-lime-silicon glass, the stronger its impact resistance and high-temperature structural stability. However, excessive stress increases the risk of spontaneous explosion, while insufficient stress leads to inadequate fire resistance time. Therefore, each thickness has an optimal stress range that matches its mechanical stiffness. By adding the upper and lower limits of the target surface compressive stress range corresponding to each thickness of the substrate and dividing by two, the target center value of the surface compressive stress for that thickness is obtained. This center value serves as the reference quantity for pixel-by-pixel subtraction in step S2, directly determining the sign and magnitude of the deviation value at each pixel in the full-field stress deviation map.
[0030] The calibration process for the wind pressure-stress regression coefficient is as follows: In a dual-chamber high-stress tempering furnace, for each thickness of sodium-calcium-silicon substrate, tempering experiments are conducted under multiple sets of different wind pressure settings within its corresponding target range of surface compressive stress. The measured surface compressive stress values corresponding to each wind pressure setting are recorded. The wind pressure-stress regression coefficient for that thickness is calculated by the ratio of the wind pressure change to the stress change. Since the internal and external temperature gradients are smaller during the cooling process of thick glass, and the influence of the cooling rate on surface stress is less than that of thin glass, the larger the thickness, the smaller the wind pressure-stress regression coefficient, that is, the smaller the wind pressure compensation required per unit stress deviation. This monotonically decreasing law serves as a verification constraint for the rationality of the calibration results.
[0031] The target center value of surface compressive stress corresponding to each substrate thickness and the wind pressure-stress regression coefficient are stored one-to-one according to the thickness index, forming a thickness-specific stress benchmark library. When switching thicknesses during production, the corresponding target center value of surface compressive stress is retrieved using the current substrate thickness as the index for pixel-by-pixel difference calculation in step S2, and the corresponding wind pressure-stress regression coefficient is retrieved for wind pressure compensation product calculation in step S3. The thickness indices of the two in the benchmark library must be strictly consistent and cannot be mixed across thicknesses; otherwise, the wind pressure compensation calculation result will not match the actual stress response.
[0032] In one specific embodiment, step S2, obtaining the full-field stress distribution spectrum of the tempered substrate via laser speckle interferometry, includes:
[0033] First, a reference speckle image of the tempered substrate under stress-free state is acquired. Then, a stress speckle image of the tempered substrate under stress state is acquired. Based on the digital image correlation algorithm, the displacement field of the reference speckle image and the stress speckle image is derived to obtain the full-field displacement field.
[0034] By inputting the full-field displacement field into the generalized Hooke's law of elasticity, stress field transformation is performed to obtain the full-field stress distribution spectrum.
[0035] Specifically, the baseline speckle image was acquired after the tempering furnace heating stage and before the application of cooling air pressure. At this time, the substrate was in a high-temperature homogenized state with no cooling stress on its surface. The speckle image formed by laser irradiation on the substrate surface under this state was used as a stress-free baseline reference. The stressed speckle image was acquired after the application of cooling air pressure, during the stress establishment stage on the substrate surface. The speckle image under this state was used as a comparison image after stress was applied. The two images are strictly corresponding in spatial coordinates, and the mapping relationship between pixel coordinates and the physical position of the substrate is consistent, ensuring the spatial accuracy of subsequent pixel-by-pixel displacement calculations.
[0036] The digital image correlation algorithm takes a reference speckle image and a stress speckle image as input. The reference speckle image is divided into several sub-regions according to spatial location. For each sub-region, the corresponding region with the highest gray-level distribution correlation is searched in the stress speckle image. The coordinate difference between the center points of the two corresponding regions is used as the displacement of the sub-region. The displacement of all sub-regions is spatially interpolated to obtain the full-field displacement field covering the entire surface of the substrate. Each position in the full-field displacement field contains a horizontal displacement component and a vertical displacement component. The two components together describe the deformation state of the position under cooling stress.
[0037] When converting the stress field using the generalized Hooke's law of elasticity, the horizontal and vertical displacement components at each location are partially derived with respect to the spatial coordinates to obtain the normal strain and shear strain components at that location. Then, the elastic modulus of soda-lime-silica glass (approximately 72 GPa) and Poisson's ratio (approximately 0.22) are used as material constants and substituted into the generalized Hooke's law to convert the strain components at each location into the corresponding stress components. Finally, each pixel in the output full-field stress distribution map corresponds to the surface compressive stress value at a physical location on the substrate surface. This value is directly used to calculate the difference between the pixel and the target center value of the surface compressive stress.
[0038] In one specific embodiment, step S3, dividing the full-field stress deviation map according to the physical partitions of the wind grating, includes:
[0039] The stress deviation map of the entire field is divided into 24 rectangular partitions in four vertical columns and six horizontal rows according to the physical arrangement of the wind grid. The arithmetic mean of the stress deviation values of all pixel positions in each partition is taken to obtain the mean stress deviation of each partition.
[0040] The average stress deviation of each zone is multiplied by the wind pressure-stress regression coefficient of the corresponding thickness in the thickness stress benchmark library to obtain the wind pressure compensation amount of each zone. The wind pressure compensation amounts of all twenty-four zones are summarized to obtain the wind pressure compensation instruction set.
[0041] Specifically, each pixel in the full-field stress deviation map corresponds to a stress deviation value. This value is obtained by subtracting the target center value of the surface compressive stress corresponding to the current thickness from the surface compressive stress value of the corresponding pixel in the full-field stress distribution map. A positive value indicates that the stress at that location is higher than the target center value, and a negative value indicates that it is lower than the target center value. The full-field stress deviation map is divided into six rows horizontally and four columns vertically according to the physical arrangement of the air grating, forming twenty-four rectangular partitions. The spatial range of each rectangular partition corresponds one-to-one with the physical coverage area of the corresponding air grating guide plate in the tempering furnace, ensuring precise matching between the application location of subsequent wind pressure compensation and the spatial location of the stress deviation. The stress deviation values of all pixels within each rectangular partition are summed and divided by the total number of pixels in that partition to obtain the average stress deviation of that partition. This average value represents the degree to which the partition as a whole deviates from the target center value of the surface compressive stress.
[0042] The wind pressure compensation for each zone is obtained by multiplying the average stress deviation of each zone by the wind pressure-stress regression coefficient corresponding to the current thickness. The sign of the wind pressure compensation is the same as the sign of the average stress deviation: when the average stress deviation of a zone is positive, the corresponding wind pressure compensation is positive, indicating that the wind pressure of the zone's air grating must be reduced by this compensation based on the base wind pressure reference to slow down the cooling rate of the zone and allow the stress to return to the target center value; when the average stress deviation of a zone is negative, the corresponding wind pressure compensation is negative, indicating that the absolute value of the compensation must be increased based on the base wind pressure reference to accelerate the cooling rate of the zone and allow the stress to return upward to the target range. The wind pressure compensation for all twenty-four zones is summarized in order of zone number to obtain a wind pressure compensation instruction set. Each instruction in this instruction set uniquely corresponds to a physical air grating zone, driving the corresponding zone's guide vane actuator to adjust the wind pressure to the actual executed wind pressure of each zone.
[0043] Figure 2 This is a schematic diagram of the twenty-four zones of the full-field stress deviation spectrum in the embodiments of this application. Figure 2 The diagram illustrates the spatial distribution of the average stress deviation in each of the twenty-four rectangular partitions (four vertical columns and six horizontal rows) after dividing the overall stress deviation map according to the physical arrangement of the wind grating. The value within each rectangular partition is the arithmetic mean of the stress deviation values at all pixel locations within that partition, expressed in MPa. A positive value indicates that the measured surface compressive stress in that partition is higher than the target center value, while a negative value indicates that the measured surface compressive stress in that partition is lower than the target center value. The grayscale scale on the right reflects the change in grayscale from dark to light as the average stress deviation value changes from negative to positive. The average stress deviation value of each partition is directly multiplied by the wind pressure-stress regression coefficient of the corresponding thickness to obtain the wind pressure compensation amount for each partition.
[0044] In one specific embodiment, step S3 involves multiplying the average stress deviation of each zone by the wind pressure-stress regression coefficient to obtain the wind pressure compensation amount for each zone, including:
[0045] The wind pressure-stress regression coefficient decreases monotonically with increasing substrate thickness. The wind pressure-stress regression coefficient corresponding to a 6mm substrate is greater than that corresponding to a 19mm substrate. The wind pressure compensation amount for each zone is obtained by multiplying the wind pressure-stress regression coefficient corresponding to the current production thickness with the mean stress deviation of each zone.
[0046] When the average stress deviation of each zone is positive, the wind pressure compensation amount of the corresponding zone is positive, and the wind pressure of the corresponding zone's wind grid is reduced by the wind pressure compensation amount on the basis of the basic wind pressure reference. When the average stress deviation of each zone is negative, the wind pressure compensation amount of the corresponding zone is negative, and the wind pressure of the corresponding zone's wind grid is increased by the absolute value of the wind pressure compensation amount on the basis of the basic wind pressure reference.
[0047] Specifically, the physical basis for the monotonically decreasing wind pressure-stress regression coefficient with increasing substrate thickness is as follows: During the tempering and cooling process of soda-lime silicon glass, the surface compressive stress originates from the difference in cooling rates between the surface and the interior. The thicker the substrate, the smaller the internal and external temperature gradient, and the smaller the change in the difference in cooling rates between the surface and the interior caused by the same change in wind pressure. Therefore, the thicker the substrate, the smaller the change in surface compressive stress caused by a unit change in wind pressure, and the smaller the corresponding wind pressure-stress regression coefficient. If the same regression coefficient is used uniformly for substrates of different thicknesses, the wind pressure compensation calculated for thick substrates will be systematically too large, leading to overcooling and a sudden drop in stress; the wind pressure compensation calculated for thin substrates will be systematically too small, leading to insufficient stress regulation. Both situations cause the wind pressure compensation in each zone to deviate from the actual required adjustment.
[0048] When multiplying the wind pressure-stress regression coefficient corresponding to the current production thickness by the average stress deviation of each zone, the product calculation is performed independently for each zone, without interference. This ensures that the wind pressure compensation for each zone is determined solely by its own average stress deviation and the current thickness regression coefficient. When the average stress deviation of a zone is positive, it indicates that the measured surface compressive stress of that zone is higher than the target center value of the surface compressive stress, and the corresponding wind pressure compensation is positive. The wind pressure of the wind grid in that zone is reduced by this compensation based on the basic wind pressure benchmark, which moderately slows down the cooling rate of that zone, reduces the temperature gradient between the surface and the interior, and causes the surface compressive stress to decrease and return to the target center value. When the average stress deviation of a zone is negative, it indicates that the measured surface compressive stress of that zone is lower than the target center value of the surface compressive stress, and the corresponding wind pressure compensation is negative. The wind pressure of the wind grid in that zone is increased by the absolute value of the wind pressure compensation based on the basic wind pressure benchmark, which accelerates the cooling rate of that zone, increases the temperature gradient between the surface and the interior, and causes the surface compressive stress to rise and return to the target center value. This achieves independent and precise wind pressure control for each of the twenty-four zones.
[0049] In one specific embodiment, step S3, accordingly implementing gradient cooling for each zone's air grates, includes:
[0050] The actual executed wind pressure of each zone is obtained by superimposing the corresponding zone wind pressure compensation amount in the wind pressure compensation instruction set on the base wind pressure corresponding to each thickness substrate as a reference.
[0051] The actual air pressure for each zone is applied to the corresponding zone's air grid in three stages: rapid cooling, intermediate cooling, and slow cooling. The rapid cooling stage has a temperature range of 630℃ to 450℃ and a cooling rate of 18-25℃ / min; the intermediate cooling stage has a temperature range of 450℃ to 350℃ and a cooling rate of 10-15℃ / min; and the slow cooling stage has a temperature range of 350℃ to 320℃ and a cooling rate of 6-8℃ / min. Forced air cooling is stopped when the temperature reaches 320℃.
[0052] Specifically, the base air pressure corresponding to each substrate thickness is the empirical center value of the air pressure when the tempering furnace is stably producing within the target stress range. The base air pressure corresponding to the 6mm substrate is 1.9MPa, 8mm substrate is 1.3MPa, 10mm substrate is 0.8MPa, 12mm substrate is 0.6MPa, 15mm substrate is 0.4MPa, and 19mm substrate is 0.25MPa. The air pressure compensation amount corresponding to each zone in the air pressure compensation instruction set is added to the base air pressure corresponding to the current thickness to obtain the actual executed air pressure for that zone. When the air pressure compensation amount is positive, the actual executed air pressure is lower than the base air pressure; when the air pressure compensation amount is negative, the actual executed air pressure is higher than the base air pressure. The actual executed air pressure of each zone must not be lower than the lower limit of the base air pressure range for the current thickness and must not exceed the upper limit. When the boundary is exceeded, the corresponding boundary value is used as the actual executed air pressure to prevent overcooling or insufficient cooling.
[0053] In the rapid cooling section, the actual applied air pressure for each zone is directly used as the applied air pressure for that zone's air grid. The temperature range of the rapid cooling section is 630℃ to 450℃, the cooling rate is 18-25℃ / min, and the duration accounts for 30%-35% of the total cooling time. During this stage, each zone maintains a differentiated actual applied air pressure state, rapidly establishing an initial compressive stress layer on the substrate surface while eliminating spatial inhomogeneities in the overall stress deviation spectrum. After entering the transition cooling section, the temperature range is 450℃ to 350℃. The air pressure for each zone is uniformly adjusted to 65% of the corresponding actual applied air pressure, the pressure distribution ratio of the upper and lower air grids is adjusted to 1:1, the cooling rate is 10-15℃ / min, and the duration accounts for 40%-45% of the total cooling time. After entering the slow cooling phase, the temperature range is 350℃ to 320℃. The air pressure of each zone is further adjusted to 35% of the corresponding actual air pressure. The pressure distribution ratio of the upper and lower air grilles is adjusted to 1:1.05, the cooling rate is 6-8℃ / min, and the duration accounts for 20%-25% of the total cooling time. When the temperature reaches 320℃, forced air cooling is stopped. The relative difference of the air pressure compensation of each zone remains unchanged throughout the three phases, ensuring that the zone-specific control effect continues until the end of cooling.
[0054] In one specific embodiment, in step S4, when the stress distribution deviation rate exceeds 8%, the stress deviation map of the whole field is divided into twenty-four zones in the same way as in step S3, and the arithmetic mean of each zone is taken to obtain the stress deviation mean of each zone again. The stress deviation mean of each zone is substituted into the wind pressure-stress regression coefficient to recalculate the wind pressure compensation amount of each zone, and the gradient cooling is performed again in step S3 until the stress distribution deviation rate does not exceed 8%. The product is then naturally cooled to room temperature to obtain the finished product.
[0055] Specifically, the second full-field stress distribution map is obtained by a second laser speckle interferometry detection after the gradient cooling is completed. Its data structure is completely consistent with the full-field stress distribution map, with each pixel corresponding to the surface compressive stress value at the same physical location on the substrate surface. Subtracting the surface compressive stress target center value corresponding to the current thickness from the surface compressive stress value of each pixel in the second full-field stress distribution map yields the second full-field stress deviation map. Using the same vertical four columns and horizontal six rows twenty-four partitioning method as the first partitioning, the second stress deviation values of all pixels in each partition are added together and divided by the total number of pixels in that partition to obtain the average second stress deviation of each partition. The absolute value of the average second stress deviation of each partition divided by the surface compressive stress target center value is then taken and the arithmetic mean of the twenty-four partitions is obtained to obtain the stress distribution deviation rate. When the stress distribution deviation rate exceeds 8%, it indicates that the stress distribution of the entire substrate surface has not yet converged to the target range after this round of gradient cooling.
[0056] The average stress deviation of each zone is used as the input for the new round of average stress deviation of each zone. The wind pressure-stress regression coefficient corresponding to the current thickness is multiplied by the average stress deviation of each zone for each zone to recalculate the new round of wind pressure compensation for each zone. The new round of wind pressure compensation for all twenty-four zones is summarized to form a new round of wind pressure compensation instruction set. The wind pressure compensation for each zone is superimposed on the basic wind pressure corresponding to each thickness of the substrate to obtain the actual wind pressure executed for each zone in the new round. The wind pressure is then applied in three stages: rapid cooling, transitional cooling, and slow cooling. The entire cycle does not require reheating and tempering of the substrate. Stress convergence is driven only by iterative adjustment of cooling parameters. After each round of gradient cooling is completed, laser speckle interferometry detection and stress distribution deviation rate calculation are performed again until the stress distribution deviation rate does not exceed 8% and the surface compressive stress values of all twenty-four zones fall within the target range of surface compressive stress corresponding to the current thickness. The substrate is then allowed to cool naturally to room temperature to obtain the finished product.
[0057] In one specific embodiment, step S1 further includes:
[0058] Edge stress optimization constraints are set for substrates of different thicknesses, with the surface compressive stress in the edge region of the substrate being 10MPa to 20MPa lower than that in the center region. The edge stress optimization constraints are then stored in the thickness-specific stress benchmark library.
[0059] In one specific embodiment, step S4 further includes:
[0060] When the stress distribution deviation rate does not exceed 8%, the average stress deviation of each zone, the wind pressure compensation of each zone, and the average stress deviation of each zone during the current regulation process are used as the regulation data triplet. The wind pressure-stress regression coefficient is updated online based on the recursive least squares method, and the updated wind pressure-stress regression coefficient is stored in the thickness stress benchmark library.
[0061] In one specific embodiment, in step S3, the basic wind pressure corresponding to each thickness of the substrate is: 1.8-2.0 MPa for 6mm substrate, 1.2-1.4 MPa for 8mm substrate, 0.7-0.9 MPa for 10mm substrate, 0.5-0.7 MPa for 12mm substrate, 0.3-0.5 MPa for 15mm substrate, and 0.2-0.3 MPa for 19mm substrate. The actual wind pressure of each zone is not lower than the lower limit of the basic wind pressure range of the corresponding thickness and does not exceed the upper limit of the basic wind pressure range of the corresponding thickness.
[0062] Specifically, in the edge stress optimization constraint, the edge region is defined as the annular region within 50mm of the edge around the substrate, and the central region is defined as the remaining plate surface area after removing the edge region. After obtaining the full-field stress distribution map again, the arithmetic mean of the surface compressive stress values of all pixels in the edge region and the arithmetic mean of the surface compressive stress values of all pixels in the central region are extracted respectively. The edge stress difference is obtained by subtracting the edge region mean from the central region mean. When the edge stress difference is not less than 10MPa and does not exceed 20MPa, the edge stress optimization constraint is considered satisfied. When the edge stress difference is less than 10MPa, it indicates that the stress distribution at the edge and center is too close, and the risk of edge stress concentration has not been effectively avoided. The edge zone wind pressure compensation should be appropriately increased in the next round of gradient cooling. When the edge stress difference exceeds 20MPa, it indicates that the edge cooling rate is too low. The edge zone wind pressure compensation should be appropriately reduced until the edge stress difference falls within the range of 10MPa to 20MPa.
[0063] When updating the wind pressure-stress regression coefficients online using the recursive least squares method, the control data triplet consists of three quantities: the mean stress deviation of each zone is the zone stress deviation before gradient cooling is implemented; the wind pressure compensation amount of each zone is the actual wind pressure adjustment amount applied in this round; and the mean stress deviation of each zone after gradient cooling is implemented. The measured wind pressure-stress response ratio for this round is obtained by dividing the wind pressure compensation amount of each zone by the mean stress deviation of the corresponding zone. The difference between this measured ratio and the currently stored wind pressure-stress regression coefficients is multiplied by the forgetting factor and added back to the current coefficients to obtain the updated wind pressure-stress regression coefficients. The forgetting factor ranges from 0.05 to 0.10. The smaller the value, the more conservative the update range, preventing drastic fluctuations in coefficients caused by abnormal data from a single furnace. The updated wind pressure-stress regression coefficients are stored in the thickness-based stress benchmark library according to the current thickness index, replacing the original coefficients, and are used for the product calculation of the wind pressure compensation amounts of each zone in subsequent batches.
[0064] The upper and lower limits of the base wind pressure range corresponding to each thickness of substrate are used as hard constraint boundaries for the actual wind pressure of each zone. After calculating the actual wind pressure of each zone by superimposing the wind pressure compensation amount on the base wind pressure benchmark, it is determined whether the value falls within the base wind pressure range of the corresponding thickness for each zone: if the actual wind pressure of a zone is lower than the lower limit of the base wind pressure range of the corresponding thickness, the lower limit value is used as the actual wind pressure of the zone; if it is higher than the upper limit of the base wind pressure range of the corresponding thickness, the upper limit value is used as the actual wind pressure of the zone. The actual wind pressure of each zone after boundary truncation is applied in three stages: rapid cooling section, transitional cooling section, and slow cooling section, to ensure that the wind pressure of each zone is always executed within the safe process window and to prevent local overcooling or insufficient cooling caused by excessive wind pressure compensation amount.
[0065] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulating fireproof glass, characterized in that, The method includes: Step S1: Construct a thickness-based stress benchmark library according to substrate thickness. This library includes the target range of surface compressive stress for each substrate thickness and the wind pressure-stress regression coefficient. Specifically, the target ranges of surface compressive stress for substrates with thicknesses from 6mm to 19mm are set sequentially according to thickness gradient. The target ranges for surface compressive stress for each substrate thickness are: 6mm substrate 130-155MPa, 8mm substrate 155-180MPa, 10mm substrate 180-205MPa, 12mm substrate 195-230MPa, 15mm substrate 230-260MPa, and 19mm substrate 260-280MPa. a. The surface compressive stress target range forms a continuous adaptive gradient as the substrate thickness increases, obtaining the surface compressive stress target range corresponding to each substrate thickness; the average of the upper and lower limits of the surface compressive stress target range corresponding to each substrate thickness is used as the center value of the surface compressive stress target, and the thickness-specific compressive stress target parameters are obtained; based on the difference in stress response of each substrate thickness to wind pressure changes, and with the constraint that the wind pressure compensation corresponding to the unit stress deviation decreases monotonically as the thickness increases, the wind pressure-stress regression coefficients corresponding to each substrate thickness are calibrated, and the thickness-specific compressive stress target parameters and the wind pressure-stress regression coefficients are stored in a one-to-one correspondence according to thickness, to obtain the thickness-specific stress benchmark library; Step S2: Obtain the full-field stress distribution map of the tempered substrate by laser speckle interferometry. Subtract the full-field stress distribution map from the target center value of the surface compressive stress of the corresponding thickness in the thickness stress reference library pixel by pixel to obtain the full-field stress deviation map. Step S3: Divide the full-field stress deviation map into physical zones of the wind grating, multiply the average stress deviation of each zone by the wind pressure-stress regression coefficient to obtain the wind pressure compensation amount of each zone, and implement gradient cooling for each zone's wind grating accordingly. Step S4: Obtain a second full-field stress distribution map through laser speckle interferometry. Subtract the second full-field stress distribution map from the target center value of the surface compressive stress to obtain a second full-field stress deviation map. Use the average ratio of the second stress deviation of each zone to the target center value of the surface compressive stress as the stress distribution deviation rate. When the stress distribution deviation rate does not exceed 8%, allow it to cool naturally to room temperature to obtain the finished product.
2. The method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulating fireproof glass according to claim 1, characterized in that, In step S2, the full-field stress distribution spectrum of the tempered substrate is obtained by laser speckle interferometry, including: First, a reference speckle image of the substrate under stress-free state after tempering is acquired. Then, a stress speckle image of the substrate under stress state after tempering is acquired. Based on the digital image correlation algorithm, the displacement field of the reference speckle image and the stress speckle image is derived to obtain the full-field displacement field. The full-field displacement field is input into the generalized Hooke's law of elasticity for stress field transformation to obtain the full-field stress distribution map.
3. The method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulating fireproof glass according to claim 1, characterized in that, In step S3, the full-field stress deviation map is divided according to the physical partitions of the wind grating, including: The full-field stress deviation map is divided into 24 rectangular partitions in four vertical columns and six horizontal rows according to the physical arrangement of the wind grid. The arithmetic mean of the stress deviation values of all pixel positions in each partition is taken to obtain the mean stress deviation of each partition. The average stress deviation of each zone is multiplied by the wind pressure-stress regression coefficient of the corresponding thickness in the thickness stress reference library to obtain the wind pressure compensation amount of each zone. The wind pressure compensation amounts of all twenty-four zones are summarized to obtain the wind pressure compensation instruction set.
4. The method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulating fireproof glass according to claim 3, characterized in that, In step S3, the wind pressure compensation amount for each zone is obtained by multiplying the average stress deviation of each zone by the wind pressure-stress regression coefficient, including: The wind pressure-stress regression coefficient decreases monotonically with increasing substrate thickness. The wind pressure-stress regression coefficient corresponding to a 6mm substrate is greater than that corresponding to a 19mm substrate. The wind pressure compensation amount for each zone is obtained by multiplying the wind pressure-stress regression coefficient corresponding to the current production thickness with the average stress deviation of each zone. When the average stress deviation of each zone is positive, the wind pressure compensation amount of the corresponding zone is positive, and the wind pressure of the corresponding zone's wind grid is reduced by the wind pressure compensation amount on the basis of the basic wind pressure reference. When the average stress deviation of each zone is negative, the wind pressure compensation amount of the corresponding zone is negative, and the wind pressure of the corresponding zone's wind grid is increased by the absolute value of the wind pressure compensation amount on the basis of the basic wind pressure reference.
5. The method for precise stress control of sodium-calcium-silicon Class C monolithic non-insulating fireproof glass according to claim 4, characterized in that, In step S3, gradient cooling is applied to each zone's air grates accordingly, including: The actual executed wind pressure of each zone is obtained by superimposing the corresponding zone wind pressure compensation amount in the wind pressure compensation instruction set on the base wind pressure corresponding to each thickness substrate as a reference. The actual air pressure for each zone is applied sequentially to the corresponding zone's air grid in three stages: rapid cooling, intermediate cooling, and slow cooling. The rapid cooling stage has a temperature range of 630℃ to 450℃ and a cooling rate of 18-25℃ / min; the intermediate cooling stage has a temperature range of 450℃ to 350℃ and a cooling rate of 10-15℃ / min; and the slow cooling stage has a temperature range of 350℃ to 320℃ and a cooling rate of 6-8℃ / min. Forced air cooling is stopped when the temperature reaches 320℃.
6. The method for precise stress control of sodium-calcium-silicon Class C monolithic non-insulating fireproof glass according to claim 1, characterized in that, In step S4, when the stress distribution deviation rate exceeds 8%, the arithmetic mean of each zone is taken according to the same twenty-four-zone division method as in step S3 to obtain the average stress deviation of each zone. The average stress deviation of each zone is substituted into the wind pressure-stress regression coefficient to recalculate the wind pressure compensation of each zone. Then, the gradient cooling is performed again in step S3 until the stress distribution deviation rate does not exceed 8%. The product is then naturally cooled to room temperature to obtain the finished product.
7. The method for precise stress control of sodium-calcium-silicon Class C monolithic non-insulating fireproof glass according to claim 1, characterized in that, Step S1 further includes: Edge stress optimization constraints are set for substrates of different thicknesses, with the surface compressive stress in the edge region of the substrate being 10MPa to 20MPa lower than that in the center region. The edge stress optimization constraints are then stored in the thickness-specific stress benchmark library.
8. The method for precise stress control of sodium-calcium-silicon C-type monolithic non-insulating fireproof glass according to claim 6, characterized in that, Step S4 further includes: When the stress distribution deviation rate does not exceed 8%, the average stress deviation of each zone, the wind pressure compensation of each zone, and the average stress deviation of each zone during the current regulation process are used as regulation data triplets. The wind pressure-stress regression coefficient is updated online based on the recursive least squares method, and the updated wind pressure-stress regression coefficient is stored in the thickness stress benchmark library.
9. The method for precise stress control of sodium-calcium-silicon Class C monolithic non-insulating fireproof glass according to claim 5, characterized in that, In step S3, the base wind pressure corresponding to each thickness of substrate is as follows: 6mm substrate 1.8-2.0MPa, 8mm substrate 1.2-1.4MPa, 10mm substrate 0.7-0.9MPa, 12mm substrate 0.5-0.7MPa, 15mm substrate 0.3-0.5MPa, and 19mm substrate 0.2-0.3MPa. The actual wind pressure of each zone shall not be lower than the lower limit of the base wind pressure range of the corresponding thickness and shall not exceed the upper limit of the base wind pressure range of the corresponding thickness.