JP7516422B2 - Edge strength test method and device - Google Patents

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the benefit of priority under 35 U.S.C. § 119 to U.S. Provisional Patent Application No. 62/848,091, filed May 15, 2019, U.S. Provisional Patent Application No. 62/852,677, filed May 24, 2019, and U.S. Provisional Patent Application No. 62/959,559, filed January 10, 2020, the contents of each of which are relied upon and incorporated by reference in their entirety into this application.

The present disclosure generally relates to an apparatus for testing glasses and/or glass ceramics and a method for testing glasses and/or glass ceramics.

High performance display devices such as liquid crystal displays (LCDs) and plasma displays are commonly used in a variety of electronic devices such as mobile phones, laptops, electronic tablets, televisions, and computer monitors. Current commercially available display devices may employ one or more precision glass sheets, for example as electronic circuit components or as color filters, to name a few applications. The primary technology for making such precision glass substrates is the fusion draw process developed by Corning Incorporated and described, for example, in U.S. Pat. Nos. 5,399,623 and 5,433,363, which are incorporated herein by reference in their entireties, although the embodiments described herein are applicable to any forming process, including slot draw, redraw, float, and the like.

For each of these applications, the glass sheets are typically cut to size and then the resulting sharp edges of the glass sheets are beveled by grinding and/or polishing. Cutting, edge machining, grinding, and other processing steps can introduce defects, such as chips or cracks, into the surfaces and edges of the glass sheets. These defects can act as a source of fracture and reduce the strength of the sheet, especially if the glass is bent such that the defects are under tensile stress. The presence of these defects can be of concern because display devices undergo some bending. Flexible display devices, by their very nature, can generate significant stresses in one or more display substrates during the manufacturing process or during use. Thus, residual defects that may be present in the glass sheets can cause the glass sheets to crack when the glass sheets are subjected to a sufficiently large stress. Because typical display manufacturing involves cutting glass to form individual displays, and cutting is known to create numerous flaws in the glass along the cut edges, glass substrate-based flexible display devices may have an unacceptably high probability of fracture after the glass sheet becomes a final product.

Attempts to reduce flaws on the edges of glass sheets include laser cutting, grinding, polishing, etc., all of which are aimed at removing or minimizing the flaws formed when cutting glass sheets to size. However, many of these approaches are unsatisfactory because the techniques are unable to remove flaws to the size required for the expected stresses, or because the techniques are difficult to apply to such thin glass sheets (less than about 0.4 mm thick). Acid etching of the glass edges can also be used, but this can degrade the display device disposed on the substrate. Thus, some flaws will continue to exist in glass sheets after manufacture, especially at the edges of the sheets, and there is a need in the industry to accurately test the edge strength of such glass sheets and panels or laminate structures using such glass sheets in order to screen out glass sheets that have less than the desired edge strength.

U.S. Pat. No. 3,338,696 U.S. Pat. No. 3,682,609

In conventional inspection methods, over 100 full-time employees are dedicated to performing V4PTB measurements on glass edge samples. Even with this significant manpower, only a small percentage of the total production can be tested. This leads to quality leakage in the form of defective products reaching customers due to infrequent testing. Also, with nearly all of the allocated resources being directed to maintaining quality requirements, there is little or no opportunity to perform process optimization studies that would help improve product quality. Thus, conventional methods create a precarious situation where defective products may be produced, but there is no time to remove the products or identify ways to resolve the issues before they are shipped. However, the exemplary embodiments provide a dramatic reduction in the amount of time spent controlling edge quality, a dramatic increase in the total glass tested relative to the glass produced, a dramatic increase in the percentage of perimeters tested, and a means for simultaneous process feedback to be used in pursuit of product improvements.

An exemplary embodiment is described that is directed to a method for continuously measuring the breaking strength of a glass edge by subjecting the edge to stress such that the stress at locations spaced from the edge is significantly lower than the breaking strength at each location. Moreover, the exemplary embodiment can be used to apply substantially the same tensile stress to both sides of an edge during measurement. Moreover, the exemplary embodiment provides a continuous, high speed nature that increases the processing speed by at least 30 times, and increases the size of the number of sample edges to be tested and the size of the cluster of glass sheets surrounded by the sample edges to be tested by at least 3 times. In this way, such an increase in statistical sampling can ensure a reduction in defect leakage to customers, and can be processed in an online configuration.

In some embodiments, an apparatus for testing edge strength of a sheet of material is disclosed. The apparatus includes: a plurality of assemblies configured to selectively apply a three-point bend load along an edge of the sheet of material in a test area of the apparatus, the plurality of assemblies being capable of establishing a loaded condition on the sheet of material in the test area by applying the three-point bend load and an unloaded condition on the sheet of material in the test area by not applying the three-point bend load. The apparatus also includes a detection mechanism configured to optically measure a strain of the sheet of material in the test area when the sheet of material is in the unloaded condition and when the sheet of material is in the loaded condition. The loaded condition is generated by the three-point bend load. The apparatus also includes a processor configured to determine a stress in the sheet of material based on the measured strain.

In another embodiment of the apparatus for testing a sheet of material, the apparatus can include a plurality of assemblies configured to continuously advance the sheet of material through a test area of the apparatus along an edge of the sheet. The plurality of assemblies are further configured to selectively apply a three-point bending load to the edge of the sheet of material passing through the test area. The apparatus also includes a detection mechanism configured to optically measure a strain on a surface of the sheet passing through the test area, where the strain is generated by the applied load. The apparatus also includes a processor for determining a stress in the sheet passing through the test area due to the applied load, where the stress is determined based on the measured strain. The load applied to the test area generates a strain on the surface of the test area as the load bends the test area of the sheet of material.

A method of testing edge strength of a sheet of material along an edge of the sheet is also disclosed. The method includes: applying a surface pattern of visual markers onto a surface of the sheet in an area of interest along the edge of the sheet; acquiring a first optical image of the surface of the sheet in the area of interest without applying any three-point bend load to the area of interest; applying a three-point bend load to the area of interest along the edge of the sheet; acquiring a second optical image of the surface of the sheet in the area of interest while applying the three-point bend load to the area of interest; and determining a stress on the surface of the area of interest along the edge of the sheet due to application of the three-point bend load based on the first optical image and the second optical image.

According to some embodiments, another method is disclosed for testing edge strength of a sheet of material along an edge of the sheet using a testing apparatus with a test area. The method includes: applying a surface pattern of visual markers on a surface of the sheet in an area of interest along the edge of the sheet; continuously advancing the sheet of material through the testing apparatus, the edge of the sheet advancing through the test area while a three-point bend load is applied to a portion of the sheet passing through the test area; acquiring a first optical image of the surface of the sheet in the area of interest while the edge of the sheet advances continuously through the test area, but before the area of interest reaches the test area; acquiring a second optical image of the surface of the sheet in the area of interest as the area of interest advances into the test area while applying the three-point bend load; and determining a stress on the surface of the area of interest along the edge of the sheet due to application of the three-point bend load based on the first optical image and the second optical image.

Additional features and advantages of the present disclosure are described in the Detailed Description below, some of which will be readily apparent to those skilled in the art from the Detailed Description or will be recognized by practicing the methods described herein, including the Detailed Description below, the claims, and the accompanying drawings.

It should be understood that both the above Summary of the Invention and the following Detailed Description of the Invention present various embodiments of the present disclosure and are intended to provide an overview or framework for understanding the nature and features of the claimed subject matter. The accompanying drawings are included to provide a further understanding of the present disclosure and are incorporated in and constitute a part of this specification. The drawings illustrate various embodiments of the present disclosure and, together with the description, serve to explain the principles and operation of the present disclosure.

The following Detailed Description is best understood when read in conjunction with the following drawings, in which like structures are designated with like reference numerals wherever possible.

FIG. 1 is a schematic diagram of a standard edge strength measurement system (ESMS) hardware setup for performing a three-point bend test on the edges of a glass sheet for edge strength. FIG. 1 is a schematic diagram of a standard edge strength measurement system (ESMS) hardware setup for performing a three-point bend test on the edges of a glass sheet for edge strength. FIG. 1 is a schematic diagram of a standard edge strength measurement system (ESMS) hardware setup for performing a four-point bend test on the edges of a glass sheet for edge strength. FIG. 1 is a schematic diagram of a standard edge strength measurement system (ESMS) hardware setup for performing a four-point bend test on the edges of a glass sheet for edge strength. Schematic diagram of a three-point bending test apparatus for testing the edge strength of test samples. Plot of load-stress variation on tension side of test panel due to change in bond position between panels Plot of load-stress variation on tension side of test panel with change in roller engagement position FIG. 1 is a schematic diagram of the basic principles of the digital image correlation (DIC) process used in the edge strength testing method described herein. 1 is a plot of stress versus time showing a repeatability study of DIC at the same location during loading cycles. Stress distribution along the top edge during three-point bending Overall load-stress curves of test samples in the DIC repeatability study FIG. 1 is an isometric view of an embodiment of an ESMS disclosed herein. A side view of the ESMS of FIG. 5B is a close-up view of the arched member of the device of FIG. 5A engaging the sheet being tested; FIG. 4 and FIG. 5A are top views of the device shown in FIG. FIG. 1 is a close-up view of one configuration of an assembly for bending a test sheet with three points in contact with the edge of the test sheet to test edge strength, according to an embodiment of the present disclosure. Strain field of the sample obtained at a load of 6 N Chart showing maximum stresses from tests in static mode Comparison of stress field results generated using the digital image correlation (DIC) apparatus disclosed herein with stress field results generated using finite element analysis (FEA). 3 is a chart of maximum stress versus displacement calculated using the digital image correlation apparatus disclosed herein and using finite element analysis. FIG. 1 is a schematic diagram showing the dynamic mode of the device disclosed herein. Schematic showing a series of areas imaged as a sheet of material advances through multiple assemblies without a load. Schematic showing a series of areas imaged as a sheet of material advances through multiple assemblies under an applied load. Schematic diagram showing two methods of using DIC to measure stress on a test sheet: measuring the stress on the top surface of the test sheet and measuring the stress on the vertical edge surface. 3D Stress Plots on the Edge and Top Surfaces of Glass Test Sheets Digital image correlation strain map for 11 N load with 60 steps between Distortion maps obtained from the same test with only one step, thereby ensuring that the distortion maps are identical In the upper part of the figure, a first schematic diagram of a sample arrangement for use in calibrating an ESMS is shown, and in the lower part, a top view of the test sheet shown in the above schematic diagram is shown. At the top of the figure, a second schematic diagram of the arrangement of samples for use in the calibration of an ESMS is shown, and at the bottom, a top view of the oblique sheet shown in the previous schematic diagram is shown. At the top of the figure, a third schematic diagram of a tilted sample arrangement for use in calibrating an ESMS is shown, and at the bottom, a top view of the sheet shown in the previous schematic diagram is shown. At the top of the figure, a fourth schematic diagram of an inclined sample arrangement for use in ESMS calibration is shown, and at the bottom, a top view of the oblique sheet shown in the previous schematic diagram is shown. 1 is a schematic diagram of a belt roller that can be used in the embodiments disclosed herein; Schematic diagram of a ball roller that can be used in the embodiments disclosed herein. 1A-1C are schematic side and top views of a camera configuration for monitoring both the horizontal top surface of a test sheet along the edges and the vertical end surface of the test sheet, according to another embodiment of the DIC system of the present disclosure; Schematic diagram showing the position of the load cell relative to the arched member and the position of the test sheet 204. 1 is a plot of applied load versus sample edge position showing the improved control of load achieved by the sum of the improvements to the ESMS disclosed herein; 1 is a plot of load versus time measured for ten test sample sheets at a relatively fast speed of 10 mm/sec as the test sample sheet 204 moves over the rollers. 1 is a plot of load versus time measured for ten test sample sheets at a relatively slow speed of 1 mm/sec as the sample sheet 204 moves over the rollers. FIG. 1 is a diagram of an ESMS testing system according to an embodiment, showing a test sheet mounted on a sample stage. 1 is a close-up detail of the desired roller engagement location along the edge of the test sheet 204 at a distance D from the edge E of the lower roller 218. FIG. 13 illustrates the positioning of the test sheet 204 at the start of a testing procedure using the ESMS such that the leading edge E _L of the test sheet rests on the apex of the lower roller 218. FIG. 2 is a diagram of an example test sheet 204, identifying the fly-in zone, measurement length, and fly-out zone. A plot showing an example load profile starting from an initial period during the descent of the upper rollers 220a, 220b, continuing through the arrival zone, measurement length, and take-off zone. FIG. 2 is a schematic diagram of an example test sheet 204 locating new cracks detected; FIG. 1 is a schematic diagram illustrating a Phantom Motor technique applied to improve control of loads in an ESMS, according to an embodiment of the present disclosure. 13 is a plot of measured load readings obtained without correction using the phantom motor technique. 1 is a plot of measured load readings obtained using the phantom motor technique; Graph showing the effect of roller diameter/polymer thickness on the load-stress relationship analyzed by finite element analysis. 1 is a plot of stress (MPa) versus load (N) comparing two groups of panels of two different thicknesses, 0.2 mm and 0.3 mm, measured by DIC and modeled by FEA. 1 is a plot of stress (MPa) versus load (N) showing modeled load-stress relationships for panels having different total thicknesses; 1 is a plot of normalized load (N) versus stress variation showing the normalized stress variation for varying panel thickness; Plots of force over time measured by dynamic ESMS for glass sheets of 0.1, 0.15, 0.2, and 0.5 mm thickness. Plots of displacement over time measured by dynamic ESMS for glass sheets of 0.1, 0.15, 0.2, and 0.5 mm thickness. Schematic diagram of a test panel/sheet showing the distance d from the center of the edge of the test panel/sheet to the test location Plot showing the effect of roller engagement on load-stress relationships as determined by FEA and DIC measurements. 30A is a photograph showing an example of an off-apex break of a test panel/sheet during edge strength testing; 30B is a plot of normalized stress versus distance from the roller apex for the off-apex break of the test panel/sheet. ESMS stress profile plot Stress vs. probability plot showing the strength distribution of batches of panels analyzed with the "Simplified" or "Right" method. Schematic showing the nomenclature of different panel surfaces Schematic of the propagation of a pre-existing crack on the long edge that prevented measurement of the short edge Stress profile on Surface 1 and Surface 3 when the gap between the sealant and the glass edge is 0 mm Stress profile on Surface 1 and Surface 3 when the distance between the sealant and the glass edge is 2 mm Schematic of a machine learning neural network for identifying cracks in test panels/sheets Schematic diagram of the panel and its design parameters FIG. 1 is a schematic diagram of a terminated and non-terminated edge of a panel engaged with a roller in an ESMS of the present disclosure. 1 is a flowchart of a method according to an embodiment of the present disclosure; 1 is a flow chart of a method according to another embodiment of the present disclosure;

In the manufacture of glass sheets for display applications, for example, glass sheets cut to size are subjected to finishing processes, such as grinding and/or polishing to remove sharp edges and create chamfered edges. During these finishing, handling, or other manipulation steps, residual stresses may remain at the edges of the glass sheets. Such residual stresses and improper handling may result in the glass sheets being broken at a later time. If breaks occur before the glass sheets are shipped to a customer, they reduce glass production yields and interrupt the glass manufacturing process. If breaks occur in the possession of a customer, who is typically a manufacturer of display devices using the glass sheets, they will cause yield losses and interruptions to the customer's manufacturing process. Regardless of when such later breaks occur, they are undesirable.

Therefore, glass sheets are tested for edge strength at the glass manufacturing plant before they are shipped to customers. The traditional edge strength testing method is the vertical four point bending (V4PTB) method. V4PTB testing is performed on small specimen samples of glass cut from the glass sheets at the end of production. These specimens have dimensions of approximately 150 mm x 10 mm and are tested individually. Because the glass sheets are broken to cut the specimens, only a small number of glass sheets are tested, and it is assumed that this small number of glass sheets is statistically representative of the set of glass sheets produced.

Testing of sample specimens with the V4PTB is a manually intensive process that requires approximately one day to process multiple samples from one glass sheet, so only a small number of glass sheets are tested, for example on the order of approximately 3 sheets for every 22,000 sheets produced. Furthermore, such methods cannot evaluate laminated structures or panels. These shortcomings can result in significant leakage in the form of defective products reaching customers.

Corning Incorporated's U.S. Patent Application Serial No. 15/557,991 (published as U.S. Patent Publication No. 2018/0073967) discloses an edge strength measurement system (ESMS) and related methods for non-destructively testing the edge strength of glass sheets without cutting test specimens. The ESMS system performs three-point or four-point bend tests directly on the edges of the glass sheets, allowing testing during production without the need to cut the glass sheets, thus allowing a larger proportion of the manufactured glass sheet population to be tested.

1A-1D are schematic diagrams of some examples of existing ESMS system hardware configurations for performing three-point or four-point bend testing on the edges of a glass sheet for edge strength. FIGS. 1A and 1B are perspective and plan views, respectively, of an exemplary configuration that allows three-point bend testing on the edge 205 of a glass sheet 204. Roller assemblies 207b and 213b with three associated rollers 215, one roller on the top side 226 (see FIG. 1E) of the glass sheet 204 and two rollers on the bottom side of the glass sheet 204, provide a three-point bend testing configuration. Roller assemblies 207c and 213c with three associated rollers 215, two rollers on the top side of the glass sheet 204 and one roller on the bottom side of the glass sheet 204, provide a three-point bend testing configuration. As shown in FIG. 1B, a vertical load is applied to the top side with rollers to bend the glass sheet 204. As shown in FIG. 1D, the roller 215 and glass sheet 204 are positioned so that the roller 215 receives the edge 205 of the glass sheet 204 between the upper and lower rollers and applies a normal load force to the edge 205.

1C is a side view of an exemplary configuration that allows four-point bend testing on the edge 205 of the glass sheet 204. Roller assemblies 207d and 213d with four associated rollers 215, two on the top side of the glass sheet 204 and two on the bottom side of the glass sheet 204, provide a four-point bend testing configuration. A vertical load that bends the glass sheet 204 is applied to the top side using roller assembly 207d.

ESMS systems can be useful in testing glass used in the manufacture of flat panel displays because such glass sheets must meet stringent surface quality requirements in the display area, which is often the central portion of the glass sheet away from the edges. Contact with the usable or "quality" area of the glass sheet can cause surface defects that render the glass sheet unusable. The exemplary configuration of roller 215 arrangement shown in Figures 1A-1D preserves surface quality on both sides of glass sheet 204.

The ESMS system can utilize test sheet materials other than the glass sheet 204 shown in the illustrated example. For example, the ESMS system can be used to test the strength of laminated structures (also called panels) of other rigid or semi-rigid materials. The laminated structure or panel can include multiple glass sheets laminated with one or more intermediate polymer layers, or in alternative embodiments, can also include structures having thin film transistor glass substrates and color filter glass substrates with one or more films between them or adjacent to one or both of the substrates. For purposes of this disclosure regarding the ESMS system and methods of using the ESMS system, the terms "sheet(s)" and "panel(s)" are used interchangeably. Unless otherwise distinguished, whether one or more sheets or one or more panels are being tested by the ESMS system does not fundamentally change the edge strength test method using the ESMS system.

The structure of one example of a panel that can be tested with the ESMS is described below. Panels with different designs can be tested. Figure 38 shows a schematic diagram of a panel, which specifies several design parameters. W and L are the width and length of the panel, respectively. (1) is the protruding length of the termination edge, (2) is the distance between the sealant and the panel edge, (3) is the sealant width, and (4) and (5) are the CF glass thickness and the TFT glass thickness, respectively.

There are some priorities for panel dimensions and sealant properties to perform testing with the ESMS. They are as follows: Small rectangular panels with length (L) and/or width (W) of 600 mm or less can be tested. Larger panels can also be measured. Preferably, the protruding length (1) of the terminal edge is longer than the roller engagement, which may be 0.5-2 mm. Preferably, the distance (2) between the sealant and the panel edge is less than 1 mm. Preferably, the sealant width (3) is 500 μm or more. Preferably, the sealant has a thickness of ≦10 μm and a Young's modulus of ≧1 Gpa. Preferably, the interfacial bond strength between the brittle material and the seal should be sufficient so that delamination does not occur during edge strength testing. Preferably, the total thickness of the CF glass thickness (4) + TFT glass thickness (5) is ≧0.2 mm, but ≦0.5 mm. The CF glass thickness (4) and the TFT glass thickness (5) can be the same or different. Preferably, the display layer deposited on the TFT surface has negligible mechanical stiffness.

Both non-terminated and terminated edges of the panel can be tested. Figure 39 shows a side view and a top view of a panel engaged with rollers at both the non-terminated and terminated edges. The terminated edge of the panel is where the TFT glass protrudes relative to the CF glass edge.

One of the drawbacks of the ESMS system described in U.S. Patent Publication 2018/0073967 referenced above is that the stress measurements during edge strength testing are performed using load cells attached to the upper roller assemblies 207b, 207c that are used to apply the test load to the glass sheet. This is an indirect method of detecting edge stress. A more direct method of measuring the stress experienced by the glass sheet being tested is desired.

Previous techniques for measuring edge stress and strength of ultra-thin glass sheets using a three-point bend configuration have produced highly variable results. These results had a significant level of variability based on a variety of factors, including but not limited to: (a) pre-test breakage; (b) residue or debris in the system; (c) product variability (e.g., panel thickness, adhesive properties/location variations), and (d) alignment. These sources of variability have resulted in inaccurate strength measurements and unexpected product failures in the field. Figures 2 and 3 show how changes in adhesive location and roller engagement lead to variability in stress values. Figure 2 is a plot of the load-stress variation on the tension side of a test panel with changes in adhesive location between panels. Figure 3 is a plot of the load-stress variation on the tension side of a test panel with changes in roller engagement location. An additional source of variability is that the conventional three-point bend configuration obtains results based on a model of the glass sheet, rather than a direct measurement of strain on the surface of the area of the glass sheet in question.

In contrast, the measurement apparatus and method disclosed herein provides real-time information on the stress distribution within the sheet 204 being tested using digital image correlation (DIC) full-field imaging techniques. Using DIC, the strain on the surface of the sheet 204 can be measured simultaneously in both in-plane orthogonal directions as a function of time. Stress can then be determined from the measured strain, which can generate a stress field map of the sheet. The direct measurement techniques disclosed herein are more accurate than conventional techniques and avoid the need to derive new calibration equations and models for every redesign or reconfiguration of the glass sheet/panel.

Examples of uses of the ESMS described herein include, but are not limited to: (a) calibrating existing edge strength testers that do not have DIC capabilities; (b) investigating failure modes of new glass sheet/panel designs when the test system has the flexibility to vary various test parameters such as roller engagement, sheet/panel angle, roller diameter, load application speed, etc.; and (c) optically performing direct strain measurements during dynamic and continuous edge strength testing suitable for manufacturing plants.

Examples of benefits and advantages of the devices described herein include, but are not limited to:
Direct full-field (optical) distortion measurement on ultra-thin glass panels;
Optical imaging for detection along with full-field stress mapping provides insight into failure modes;
Real-time stress visualization during dynamic and continuous edge testing, suitable for deployment in manufacturing plants;
- Improved analytical equations for obtaining stresses from measured strains;
A new material handling assembly that provides enhanced optical visibility to enable the above-mentioned stress measurements;
Ability to act as a calibration bench for existing edge strength testing devices;
Ability to vary test parameters such as roller diameter, engagement area, roller thickness, etc. to develop new panel designs;
Accurate stress measurement with an accuracy of ±10% for the measured stress.

With reference to Figures 4-7, 9B, and 9C, an apparatus 200 for testing edge strength of a sheet 204 of material is disclosed. The structure of the apparatus 200 described herein is an exemplary embodiment configured to optically measure the strain of the sheet 204 when the sheet is in an unloaded state and when the sheet 204 is in a loaded state. The loaded state is generated by a three-point bending load.

The apparatus 200 includes a plurality of assemblies 206a, 206b configured to selectively apply a three-point bending load along the edge of the sheet of material 204 in the test area 208. By "selectively applying a three-point bending load" it is meant that the plurality of assemblies 206a, 206b are configured to apply or not apply a three-point bending load as required and to control the amount of load applied. The apparatus 200 also includes a detection mechanism configured to optically measure the strain of the sheet of material 204 in the test area 208 when the sheet of material 204 is in an unloaded state and when the sheet of material 204 is in a loaded state. The strain in the loaded state is produced by the applied three-point bending load. The apparatus 200 also includes a processor 214 (see FIG. 9C) for determining the stress in the sheet 204 based on the measured strain. The applied three-point bending load bends the sheet 204, so that the three-point bending load applied to the sheet 204 produces a strain on the surface of the sheet 204.

The plurality of assemblies 206a, 206b (see FIG. 7) is comprised of two opposing assemblies 206a and 206b, the first of which 206a includes a single arched member 218 for engaging a first side of the sheet 204. The second of which 206b includes two spaced apart arched members 220a, 220b for engaging a second side 226 of the sheet 204 opposite the first side. The two spaced apart arched members 220a, 220b define a test area 208 between the two spaced apart arched members.

The arched members 218, 220a, and 220b may be rotatably mounted rollers, belt rollers, bearing rollers, or fixed bushings, as shown in FIG. 9A. Fixed bushing (e.g., non-rotating) structures are shown in FIGS. 4 and 5B. The fixed bushings are configured to have an arched surface formed from a material with a low coefficient of friction (e.g., fluoropolymer resin). In a preferred embodiment of the test apparatus 200 of the present disclosure incorporating DIC functionality, the two spaced apart arched members 220a and 220b are fixed bushing structures for engaging the second side 226 of the sheet 204, as shown in FIGS. 4 and 5B. These non-rotating bushing structures are configured to have a narrower width than the rotating roller embodiment, so that the bushing structures provide more space between the two arched members while still keeping the two contact points on the surface of the test sheet 204 the same distance apart.

In some embodiments, the detection mechanism comprises a first optical system 222 for determining the distortion on the surface of the test sheet 204 in the test area 208. The first optical system 222 determines the distortion using a DIC technique, which is described in more detail below, but essentially, DIC determines the distortion by taking two images of the area of interest of the test sheet 204, a first image obtained in an unloaded state and a second image obtained with the area of interest of the test sheet in a loaded state, i.e. under a three-point bending load in the context of the present disclosure. The surface of the area of interest is provided with several visual markers, so that when comparing the two images, any displacement of the visual markers in the loaded image caused by distortion (deformation) can be detected.

As mentioned above, the sheet 204 may be a brittle material such as glass or glass ceramic. The three-point bending load is sufficient to deform the sheet 204 of material. As shown in FIGS. 4-5B and 7, in some embodiments, the plurality of assemblies 206a, 206b are comprised of: a first assembly 206a positioned below the sheet 204 with a single arched member 218 for engaging a first side (lower side) of the sheet 204; and a second assembly 206b positioned above the sheet 204 with two arched members 220a, 220b for engaging a second side (upper side/upper side) 226 (see FIGS. 11A, 11B) of the sheet 204 opposite the first side. The two arched members 220a, 220b are separated by a minimum spacing S to provide the optical system with a clear view of the test area 208. The actual dimensions of the minimum spacing S will depend on the particular type of camera and the size of the test area 208, which is determined by the particular arrangement of the arched members 220a, 220b, and 218.

The arched members 218, 220a, and 220b are preferably structures that contact the sheet 204 while the arched members and the sheet move relative to one another, providing a low friction contact surface. Some examples of such arched members are cylindrical rollers, such as those shown in Figures 7, 9A-9C, and 11A-11D, and bearing rollers, such as those shown in Figure 13.

In a belt roller embodiment such as that shown in FIG. 12, the arched members 218, 220a, 220b may be belt rollers 240c, 240a, 240b. In this belt roller embodiment, the belts 240a, 240b, 240c extend around the drive shafts 242a, 242b, 242c and the arched tensioners 244a, 244b, 244c. In some embodiments, the arched tensioners may be cylindrical rollers or static arched tensioners (e.g., polished metal fingers with an arched portion that contacts the belt). In a bearing roller embodiment such as that shown in FIG. 13, the arched members 218, 220a, 220b are bearing rollers 246a, 246b, 246c. In such an embodiment, the bearing rollers 246a, 246b, 246c are formed from roller balls 248a, 248b, 248c disposed and retained in sockets 250a, 250b, 250c.

In some embodiments, as shown in FIG. 9A, a single arched member 218 is aligned vertically between two spaced apart arched members 220a, 220b. As used herein, "longitudinally" refers to the direction of movement of the sheet 204 through the testing apparatus 200 (e.g., the machine direction).

9B, in some embodiments, the optical system 222 comprises a first set of stereo (i.e., paired) CCD cameras 222a, 222b positioned above the second side 226 of the sheet 204 to inspect the surface of the sheet 204 in the test area 208 between the two spaced apart arched members 220a, 220b. The test area 208 is the area between the arched members 220a, 220b under the two cameras 222a, 222b, where the two cameras 222a, 222b are intended to take digital images for optical measurements. In other words, the test area 208 represents the field of view of the two cameras 222a, 222b. The test area 208 is also the area where the three arched members 220a, 220b, and 218 can apply a three-point bend test load. The test area 208 thus represents the target area of the test apparatus for applying the three-point bend test to the test sample. The test sheet 204 to be tested must be brought into the testing area 208 for edge strength testing. A pair of cameras 222a, 222b.

In some embodiments, the testing apparatus 200 can be operated in a dynamic mode. In the dynamic mode, an edge strength test is performed on the sheet 204 by passing the sheet 204 through a test area 208 of the testing apparatus 200. Alternatively, the testing apparatus 200 can be configured to perform an edge strength test on the sheet 204 by holding the sheet 204 stationary and moving the testing apparatus across the sheet 204.

The apparatus 200 can include a plurality of assemblies 206a, 206b configured to continuously advance the sheet of material through a test area 208 along the edge of the sheet 204. The plurality of assemblies 206a, 206b are further configured to apply a three-point bending load to the edge of the sheet of material passing through the test area 208. The apparatus 200 also includes a detection mechanism configured to optically measure strain on a surface of the sheet passing through the test area 208, where strain is generated by the applied three-point bending load. The apparatus 200 also includes a processor 214 for determining a stress in the sheet 204 passing through the test area 208 due to the applied load. The stress is determined based on the measured strain. The load applied to the test area 208 generates strain on the surface of the test area 208 as the applied three-point bending load bends the sheet 204 in the test area 208.

Digital Image Correlation (DIC)
The DIC functionality of the test apparatus 200 will now be described in more detail with reference to Figures 9B and 9C. DIC involves point tracking. It is a full-field optical technique for acquiring displacement and thereby strain. In some embodiments, a surface pattern 230 is applied to the surface of the second side 226 as a visual marker to facilitate strain measurement in both unloaded and loaded conditions. In a loaded condition measurement mode, a load applied to the sheet 204 in the test area 208 deforms the surface of the second side 226 of the sheet 204, which deformation distorts and displaces the surface pattern 230 on the second side 226 of the sheet 204. The surface pattern 230 can be an array of dots or points applied on the surface being monitored, which will act as a visual marker for measuring the strain of the sheet 204. The array of dots or points can be either a regular or random pattern. The images of the surface pattern 230 on the sheet 204 taken in the unloaded state act as reference points for determining the displacement of these reference points caused by the subsequent distortion of the surface pattern 230 by the application of a three-point bending load. A first set of stereo cameras 222a, 222b in the optical system 222 observes the distortion of the surface pattern 230 and the image data can be used to determine the amount of distortion of the sheet 204.

To determine the amount of distortion in the sheet 204 from the distortion of the surface pattern 230, the amount of distortion must be quantified. This is accomplished by comparing an image having the distorted surface pattern 230 on the top surface 226 of the sheet 204 with a reference/registered image of the second side 226 in an unloaded state where the surface pattern 230 is undistorted.

Regardless of whether the device 200 is operating in a static or dynamic mode, an image of the surface pattern 230 in an unloaded state is recorded using the optical system 222 as a reference or baseline state. The reference image is used to register the original positions of various points of the surface pattern 230. The new positions of the registered points are then tracked in the loaded state image. If the top surface 226 is distorted by the applied load stress, many of the registered points on the top surface 226 will be displaced from their original positions. By correlating the registered points in the two image data, the processor 214 can determine and quantify the deformation, i.e., distortion. Algorithms for performing such digital image correlation are known in the art and further detailed discussion is not necessary here.

The surface pattern 230 may be printed on the top surface 226, or may be projected as an image onto the top surface 226, or may be sprayed as a random pattern onto the top surface 226, or may be etched onto the top surface 226 using a chemical process, or may be applied using a pre-fabricated thin film having a pattern on the top surface 226. Nanoparticles may be sprayed onto the top surface 226 using an airbrush. The pattern 230 may be a speckle pattern or a grid pattern.

In some embodiments, the surface pattern 230 can consist of a coating of a mixture or blend of black and white dots to form a speckle pattern. Stereo cameras 222a, 222b and/or 224a, 224b with appropriate lenses are used to record images of the surface pattern 230 during testing. The coating of the speckle pattern 230 can be made using an oil-based black and white spray that produces a fine mist resulting in speckles with each speckle covering 5-7 μm. This paint layer is reliably less than a few micrometers and therefore does not affect the results. A white light LED lamp is used to illuminate the area of interest for optical inspection with the stereo cameras.

There are two methods for measuring edge strength using DIC. One measures the strain on the top surface of the sample, and the other measures the strain on the edge surface. The diagram in FIG. 9D identifies the top surface 226 and the edge surface 204e on the test sheet 204. The edge strength measured by both approaches should be the same since the stress is continuous at the edge boundary. This is illustrated in FIG. 9E, which shows a 3D stress plot on the edge and top surfaces of the test sheet. The gradient in the image, represented by the difference in shading darkness, indicates that the stress on the top and edge surfaces is continuous. The shading darkness of the gradient represents different stress levels, but the actual stress values are not important for the purposes of this disclosure.

Alternative methods of speckle coating and implementing dynamic DIC thereof
Coating the test sheet 204 with a visible speckle dot surface pattern 230 would render the glass sheet unusable for consumer applications even if it passed the edge strength test screening. To avoid this, invisible speckles may be used that are visible only to the dedicated optics employed by the device (e.g., ultraviolet light, fluorescent imaging, infrared light, or another invisible portion of the electromagnetic spectrum). Alternatively, a laser can be used to project speckle dots onto the test sheet 204, which can be tracked by a camera during testing. The dot pattern may be random.

Exemplary embodiments have been described that are directed to a method for continuously measuring the breaking strength of a glass edge by placing only the edge under stress such that the stress at locations spaced from the edge is significantly lower than the breaking strength at each location. Moreover, exemplary embodiments can be used to apply substantially the same tensile stress to both sides of an edge during measurement. Although one method for providing this continuous stress has been detailed (e.g., opposing offset rollers), the claims appended hereto are not intended to be so limited, as acoustic and/or infrared energy (both coherent and incoherent) can also be used for the same purpose of inducing stress at the edge of a glass sheet. For example, focused ultrasound can be used to induce stress at the glass edge from which measurements can be obtained using the apparatus and methods of the present disclosure. Additionally, IR irradiation (in a spectrum that allows for significant absorption by the respective glass material) using a laser or other means can also be used to induce stress at the glass edge from which exemplary measurements can be obtained using the apparatus and methods of the present disclosure. Moreover, the exemplary embodiment provides a continuous high speed nature that increases the processing speed by at least 30 times, and increases the amount of edges to be tested, as well as the size of the sheets to be processed and tested, by at least 3 times over conventional methods. In this way, such increased statistical sampling can ensure a reduction in defect leakage to customers, and can be handled in an on-line configuration.

[Upper surface DIC]
The basis of the optical methodology of DIC is shown in Figure 3A. By comparing images at each time interval (during increasing load or before and during load application), the displacement of the speckle dots can be tracked. This provides a displacement map, which can be differentiated to obtain the strain. Using this 3D DIC technique, a 3D strain field can be obtained. Using the two in-plane strain fields (x and y directions), the bending stress can be calculated using Equation (1):

where E is the elastic modulus, εxx is the strain along the bending direction (x), εyy is the strain along the other in-plane principal axis (y), and ν is the Poisson's ratio. Conventional stress transformation practice ignores the biaxiality of the stress state and uses Hooke's law for uniaxial stress conditions. This provides less information than the DIC technique described herein and oversimplifies the results.

[Vertical end face DIC]
Referring to FIG. 14, a series of three-point bending tests were performed on the test sheets 204 while monitoring the vertical edge 204e of each test sheet 204 using DIC. In these tests, the test sheets 204 had a laminated structure, such as a display panel. The thickness of the test sheets was about 200 μm. A speckle pattern 230 was created on the vertical edge 204e of the test sheets 204. When testing certain types of test sheets 204, such as those having a laminated structure, it may be more desirable to perform optical monitoring on the vertical edge 204e of the test sheets 204 rather than on the top surface (horizontal surface) of the test sheets 204.

14 shows a schematic diagram of an embodiment of a camera configuration for the optical system 222 of the test apparatus 200 that enables DIC to optically measure strain along the vertical edge 204e during a three-point bend test according to an embodiment. In this embodiment of the test apparatus 200, a second set of at least one camera 222c, 222d is provided that targets the vertical edge 204e of the test sheet 204. The at least one camera 222c, 222d targets the vertical edge 204e of the test area 208 between the two spaced apart arched members 220a, 220b. In a static mode of operation of the test apparatus 200, the at least one camera 222c, 222d can be used to capture optical images of the vertical edge 204e in both the unloaded and loaded states. Although a pair of cameras 222c, 222d are shown in FIG. 14 observing the vertical edge 204e of the test area 208, a single camera can be used to observe the vertical edge 204e because the vertical edge 204e is flat even under the three-point bending load. Because the vertical edge 204e is flat, a 3D view using two cameras is not necessary.

14 also shows a second set of at least one camera 224c, 224d aimed at the vertical edge 204e but away from the test area 208. Similar to the pair of cameras 224a, 224b, the second set of at least one camera 224c, 224d can be used to obtain optical images of the unloaded vertical edge 204e when the test apparatus 200 is operating in a dynamic mode. Again, although a pair of cameras 224c, 224d is shown in FIG. 14, a single camera can be used to view the vertical edge 204e. The processor 214 is connected to the cameras 222c, 222d, 224c, and 224d and can control the cameras and can receive and process optical images from the cameras.

We developed two methods to create fine speckle patterns. In the first method, we used a fine brush with liquid paint to create a layer of paint on the edge. Then, lens paper was used to create a pattern on the edge. The lens paper has a microscale fine random pattern, where areas with fibers absorb the paint and areas without fibers leave the paint on the edge. This technique provides a good speckle pattern for imaging. Another technique to create fine speckle patterns is to use nanoparticles to make the speckle pattern. However, agglomeration affects the quality of the speckle pattern. The quality is not very reliable and not very stable.

Figure 3B shows a repeatability study of DIC at the same location during the load cycle. It shows that the repeatability of DIC on the edge is stable. As can be seen in the unloaded state, the accuracy of DIC is ±10 MPa. Using edge DIC, it was observed that the failure mode of the panel is mostly due to stress concentration. Figure 3C shows the stress distribution along the top edge during three-point bending. The individual curves are the stress distribution along the top edge at different loads. Figure 3C shows the stress distribution from the load application to the point where the test sheet 204 breaks. From these curves, it can be seen that defects along the edge affect the stress distribution. Figure 3D shows the overall load-stress curve of the test sample. Note that the edge DIC is performed at high magnification. The resolution of one pixel is about 0.5 μm, and the stress is localized at the apex location. The variation is due to edge defects, accuracy of the apex location, and sample variations such as thickness, penetration, layer differences, sealing stage, etc.

In some embodiments, a conventional acrylic paint can be used to form an ultra-thin (a few micrometers) coating of the speckle surface pattern 230 on the glass sheet 204. The paint has a modulus of elasticity at least one order of magnitude lower than the glass sheet 204, and its thickness is two orders of magnitude smaller than the glass sheet 204, so the effect of the pattern on measurements is negligible. The use of such surface patterns in digital image correlation is known in the art and further discussion is not necessary here.

Top and edge DIC can be performed statically or dynamically. Static DIC measures the load-stress correlation at a fixed location along the edge of the test sheet 204. Dynamic DIC measures the load-stress correlation along the entire length of the edge of the test sheet 204. Static DIC is easy to implement but can only measure one location at a time. Dynamic DIC is more complex to implement but can measure the entire length of the edge of the test sheet 204. Static DIC works best for samples such as monolithic glass where the correlation does not change with edge position. Dynamic DIC works best for samples such as panels with laminated structures where the correlation can change.

[ESMS Static Mode - Edge Strength Tester with Real Time Stress Measurement]
In a static mode embodiment, a first set of stereo cameras 222a, 222b are used to incrementally obtain both unloaded and loaded measurements. A surface pattern 230 of visual markers is provided on the top surface 226 of the test sheet 204. The test sheet 204 is incrementally moved through a test area 208 of the testing apparatus, stopping when a particular area of the sheet 204 to be tested enters the test area 208 so that unloaded and loaded measurements can be performed.

In this example, four different test target areas (areas 1, 2, 3, and 4) along the edge of the sheet 204 are placed in the test area 208 by moving the sheet 204 in a right-to-left motion as shown in FIG. 9B. To test test target area 1, which is at the leftmost portion of the lower edge of the sheet 204, the test target area 1 is placed in the test area 208. To test test target area 2, which is slightly to the right of area 1, the sheet 204 is moved to the left in the direction of the arrow in FIG. 9B. Similar steps are repeated for test target areas 3 and 4. This initial step is performed without any vertical load being applied by the upper arched members 220a, 220b. For the three-point bending test, the applied load is increased until a predetermined load is reached or until failure of the sheet 204 is detected. The first optical system 222 includes at least two cameras 222a, 222b for optically inspecting and detecting deformation of the sheet in the unloaded and loaded measurements.

The loading conditions of the test apparatus for each of the test target areas 1-4 are shown in FIG. 9C. In the loading conditions, a vertical load is applied by either the two upper arched members 220a, 220b or by the single lower arched member 218. Either way, the applied load bends the sheet between the arched members. The four loading conditions corresponding to areas 1-4 are labeled as deformations 1, 2, 3, and 4, respectively.

In static mode, unloaded and loaded measurements can be performed consecutively for each of the four test regions. Alternatively, unloaded measurements for the four regions 1-4 can be performed first, followed by returning the sheet 204 to the initial position, after which loaded measurements can be performed for each of deformations 1, 2, 3, and 4 by moving the sheet 204 through the test region 208.

Three-point bend based edge strength testing techniques and DIC optical methods, combined with appropriate design modifications, can provide real-time stress measurements while testing the edge strength of ultra-thin monolithic or laminated glass structures (e.g., sheets of material).

Any form of optical measurement, including DIC capabilities, requires a clear and direct optical view of the area of interest on the test sheet 204 in the test area 208 for measurement. In previous systems (e.g., U.S. Patent Application Publication No. 2018-0073967), there was no direct stress measurement, and these designs did not allow a clear optical path to the area of interest because the rollers blocked the path. In accordance with this disclosure of an embodiment of a test fixture 200 with DIC implemented, the arched members 220a, 220b are modified to allow a clear optical path to the area of interest for the camera. See Figures 4-7, 9A-9C, and 11A-11D. This modification, and the resulting improvement, is also present in the test fixture 200, regardless of whether the fixture is operated in static or dynamic DIC mode. The main design considerations were:
The upper rollers (arch-like members) 220a, 220b were modified from wheels to partial sections, specifically, the width of said partial sections was minimized to obtain maximum light path while maintaining the required contact area of the entire rollers against the glass sheet 204. See Figures 5A and 5B.

- To study the effect of engagement, we have incorporated a miniature micrometer level adjuster 232 for x, y, z axis control to allow precise alignment of the glass to the roller/upper partial section. See Figures 5A and 5B.

- To study the effect of skew, we have incorporated a miniature micrometer level adjuster 234 for axial control to allow precise alignment of the glass to the roller/upper partial section. See Figures 5A and 5B.

- To ensure initial test alignment, the two halves of the fixture and the partial section were all secured together with dowels 236a, 236b. See Figures 5A and 5B.

With a clear light path to the test area 208 available, an optical system was designed employing DIC techniques to perform real-time direct stress measurements. Two 4 megapixel commercial monochrome cameras fitted with a dedicated lens system were used to observe the 7 mm x 7 mm test area 208 near the edge of the glass sheet 204. The sample was coated with a random black and white dot surface pattern 230 such that the dots were approximately 5-7 pixels in size and the coating thickness did not exceed a few micrometers. A stereo camera system was used to record a series of images (150-200 pairs of images at 4 frames per second) during the edge test experiment. The series of images were correlated to obtain the strain field, and then the stress was obtained using equation (1). An example of the stress field obtained at a 6 N load step is shown in Figure 8C, along with the history of the maximum stress throughout the test. It is believed that this type of stress visualization/measurement has not been performed before. Correlation of these results with numerical simulations (e.g., finite element analysis) confirms good agreement as shown in Figure 8C.

With reference to flow chart 500 of FIG. 40, an example method for testing edge strength of a sheet of material along an edge of the sheet of material 204 can be outlined as follows: The method includes: (a) applying a surface pattern 230 of visual markers to a surface 226 (second side) of the sheet 204 in an area of interest along the edge of the sheet 204 (see box 510); (b) acquiring a first optical image of the surface of the sheet 204 in an area of interest along the edge of the sheet without applying any three-point bending load to the area of interest. (c) applying a three-point bending load to the region of interest along the edge of the sheet (see Box 530); (d) acquiring a second optical image of the surface of the region of interest along the edge of the sheet while applying the three-point bending load to the region of interest (see Box 540); and (e) determining the stress on the surface of the region of interest along the edge of the sheet due to the application of the three-point bending load based on the first optical image and the second optical image (see Box 550).

In some embodiments of the method of flowchart 500, step (e) of determining the stress on the surface of the region of interest along the edge of the sheet includes: measuring the strain of the sheet of material 204 in the region of interest in an unloaded state based on the first optical image; measuring the strain of the sheet of material in the region of interest in a loaded state where a three-point bending load is applied based on the second optical image; and determining the strain of the sheet induced by the applied three-point bending load by comparing the measured strain in the unloaded state to the measured strain in the loaded state.

In some embodiments of the method, measuring the strain of the sheet of material in the region of interest under load based on the second optical image includes determining a displacement of the visual marker 230 in the second optical image compared to the position of the visual marker in the first optical image. In some embodiments, determining the stress on the surface of the region of interest along the edge of the sheet includes calculating the stress required to generate the measured strain of the sheet of material in the region of interest.

In some embodiments of the method, applying a surface pattern of visual markers includes printing, coating, spraying, etching, attaching, or projecting an image onto the surface of the sheet.

In some embodiments of the above method, the first optical image and the second optical image are acquired using an optical system that includes at least one camera 222a, 222b.

[Dynamic Mode - Edge Strength Tester with Real Time Stress Measurement]
9C, in some embodiments, the apparatus 200 can operate in a dynamic mode, in which the apparatus 200 can obtain both unloaded and loaded measurements while continuously moving the test sheet 204 through the testing apparatus 200 without pausing. In a preferred embodiment, both the unloaded and loaded measurements are performed during a single pass of the test sheet 204 through the testing apparatus 200. To accomplish this, the optical system 222 can further comprise a second set of stereo cameras 224a, 224b positioned in front of the test area 208 to obtain unloaded measurements by recording images of the test sheet 204 without any three-point bending load being applied.

A second set of stereo cameras 224a, 224b are positioned above the second (upper) side 226 of the test sheet 204 so that they view the section of the test sheet 204 in front of the test area 208 as the test sheet 204 passes before moving into the test area 208, allowing the second set of cameras 224a, 224b to record an image of said section of the test sheet 204 in an unloaded state. This is illustrated by the leftmost view of Variation 1 in FIG. 9C.

In the illustrated three-point bend test configuration, imparted by arched members 220a, 220b, and 218, the second side 226 of the test sheet 204 is under tension as the sheet passes under the upper arched members 220a, 220b and over the lower arched member 218. In some embodiments, the arched members 218, 220a, 220b are rollers, as shown in Figures 9A, 9B, and 9C.

A method for testing the edge strength of a sheet of material 204 along the edge of the sheet using the testing apparatus 200 operating in dynamic mode can be summarized using the flowchart 600 of Figure 41. The method includes: (a) applying a surface pattern of visual markers 230 to a surface of the sheet in a region of interest along an edge of the sheet 204 (see box 610); (b) continuously advancing a sheet of material through the testing apparatus 200, advancing the edge of the sheet through the test area 208 while applying a three-point bend load to the portion of the sheet passing through the test area 208 (see box 620); (c) acquiring a first optical image of the surface of the sheet in the region of interest while continuously advancing the edge of the sheet through the test area 208, but before the region of interest reaches the test area (see box 630); (d) acquiring a second optical image of the surface of the region of interest of the sheet as the region of interest advances into the test area while applying the three-point bend load (see box 640); and (e) determining the stress on the surface of the region of interest along the edge of the sheet due to the application of the three-point bend load based on the first optical image and the second optical image (see box 650).

In some embodiments of the method summarized in flowchart 600, the steps of determining the stress on the surface of the region of interest along the edge of the sheet 204 include: measuring the strain of the sheet of material in the region of interest in an unloaded state based on the first optical image; measuring the strain of the sheet of material in the region of interest in a loaded state where a three-point bending load is applied based on the second optical image; and determining the strain of the sheet induced by the applied three-point bending load by comparing the measured strain in the unloaded state to the measured strain in the loaded state.

In some embodiments of the method of flowchart 600, the step of measuring the strain of the sheet of material in the region of interest under a load based on the second optical image includes determining the displacement of the visual marker 230 in the second optical image compared to the position of the visual marker in the first optical image.

In some embodiments of the method, determining the stress on the surface of the region of interest along the edge of the sheet includes calculating the stress required to produce the measured distortion of the sheet of material in the region of interest.

In some embodiments of the method of flowchart 600, the first optical image and the second optical image are acquired by an optical system that includes at least two cameras 222a, 222b, 224a, 224b. In some embodiments, applying the surface pattern 230 of visual markers includes printing, coating, spraying, etching, attaching, or projecting an image onto the surface of the sheet.

In some embodiments of the above method, the stress is determined in at least two dimensions. The two-dimensional stress is displayed as a surface plot. FIG. 8A shows the two-dimensional surface strain on a sheet when loaded by three-point contact. 216 indicates the peak strain location where the stress is high, and 217 indicates the far-field location where the stress is low. FIG. 8B is an exemplary plot showing the applied stress versus the maximum stress on a test sheet measured multiple times. And FIG. 8C is the two-dimensional surface stress field on a sheet when loaded by three-point contact. The plot on the left is measured by DIC, and the plot on the right is modeled by FEA. FIG. 8D shows the normalized stress value as a function of the vertical distance from the peak stress location toward the inside of the sheet (indicated by the black arrow). The stress increases as a function of the applied juice.

Apart from being a static calibration unit, certain embodiments of the system may also be used in a manufacturing environment (or otherwise) for continuous edge testing in a dynamic mode. As described throughout this specification, a direct optical path to the surface of the material being tested facilitates the more accurate direct measurement of distortion on which the present invention relies.

The ability to perform measurements in dynamic mode is a design improvement over existing edge measurement techniques and also provides the capability of direct stress measurement in real time. In some embodiments, the above configuration can be modified by switching the partial rollers (e.g., arched members) with three fully functional rollers that can capture a continuously fed sample as shown in Figures 9A-9C and 11A-11D. Such a system would have the rollers in a position such that the incoming glass edge would be continuously subjected to the required bending/loading. In such a system, the spacing between the rollers and the diameter of the rollers are controlled to provide a minimum spacing S between the two rollers 220a, 220b so that there is a clear light path for the camera to inspect.

9B-9C illustrate a methodology for achieving real-time stress measurements in a dynamic type edge strength testing setup as described above. The edge strength tester needs to pre-inspect the edge before testing (the sample passes through rollers without any load/bending) so that existing fractures can be detected. If such fractures are not detected, the system is malfunctioning or over-estimating the edge strength. During such an inspection, the stereo cameras 222a, 222b can take a series of pictures along the length of the edge as shown in FIG. 9B. Then, on a second pass while a load is being applied to the sample edge, the cameras can take another set of images as shown in FIG. 9C at exactly the same positions on the edge as the first series of images, and by pairing and correlating corresponding images from the same positions (i.e. by correlating the same numbered images in FIGS. 9B and 9C), the strain field and thus the stress can be obtained as described above.

Alternatively, dynamic measurements can be performed in a single pass using an embodiment of the test apparatus 200 with two sets of stereo cameras 222a, 222b and 224a, 224b. An example of such an embodiment is shown in the left-most schematic diagram of FIG. 9C. The system includes a first set of stereo cameras 222a, 222b and a second set of stereo cameras 224a, 224b. In such a system, both sets of stereo cameras are connected to a processor 214, which controls the cameras 222a, 222b, 224a, 224b and processes image data from the cameras. The system can also include a display 228 for displaying any analytical information regarding the measured stresses in the sheet 204 and for displaying images of objects viewed by the cameras.

An obvious question that arises is whether the algorithm and image correlation can obtain the strain from zero stress to peak stress without using intermediate images (150-200 images for static testing). Figures 10A and 10B show results obtained from a proof test of this concept, which confirm that accurate results can be obtained without intermediate images. The stress distribution map shown in Figure 10A was obtained by correlating 150-200 images taken during a static test at a load of 11 N. Meanwhile, Figure 10B shows the stress distribution map obtained by correlating the first image with the 200th image without any intermediate images. These two stress distribution maps are identical, indicating that the absence of an intermediate step does not affect the results. Thus, the present embodiment describes a dynamic system of edge strength measurement on ultra-thin monolithic and laminated glass samples with real-time stress visualization.

The test apparatus 200 described herein can be used to implement the above method. This refers to a dynamic DIC method. Dynamic DIC refers to obtaining strain measurements in real time while three points "roll" along the edge. In such a scenario, the strain must be obtained after the no-load measurement.

An example of a dynamic method for testing the edge strength of a sheet of material 204 along the edge of the sheet using a testing apparatus 200 having a testing area 208 can be outlined as follows: The method includes: providing a surface pattern 230 of visual markers on a surface 226 of the sheet 204, including the edge of the sheet; continuously advancing a sheet of material through a testing apparatus, advancing the edge of the sheet through a test area 208 of the testing apparatus while applying a three-point bend load to the portion of the sheet passing through the test area; acquiring a first optical image of the surface in the area of interest along the edge of the sheet while continuously advancing the edge of the sheet through the test area 208 of the testing apparatus, but before the area of interest reaches the test area 208; acquiring a second optical image of the surface 226 of the area of interest of the sheet as the area of interest advances within the test area 208 while applying the three-point bend load; and determining the stress on the surface of the area of interest along the edge of the sheet due to the application of the three-point bend load by comparing the first optical image with the second optical image.

In some embodiments, the stress can be determined in a time-resolved domain, i.e. a history of the stress evolution is obtained. This can be true both in a static mode, where measurements are taken at a position along the edge of the test sheet 204 while the test sheet 204 is stationary, and in a dynamic mode, where measurements are taken along the length of the edge of the test sheet 204 while the test sheet 204 is continuously fed through the testing apparatus 200 under the application of a predefined load.

It should be noted that although some embodiments are described with respect to a sheet 204 formed of glass, the claims appended hereto are not so limited, as the test apparatus 200 described herein can be used to receive or house and analyze laminate structures or panels. Suitable laminate structures include multiple glass sheets with one or more intermediate polymer layers, or in alternative embodiments, structures having thin film transistor glass substrates and color filter glass substrates with one or more films between or adjacent to one or both of these substrates. Thus, when referring to a sheet 204 or glass sheet herein, the reference may also apply to glass, glass ceramic, plastic, as well as laminate structures and other panels. For simplicity, only the sheet 204 is referred to herein.

The test apparatus 200 can test sheets of various sizes. For example, the test sheet 204 can have length/width dimensions of about 5 mm/5 mm to about 100 mm/100 mm, about 600 mm/600 mm, about 1000 mm/1000 mm, about 2300 mm/2600 mm, about 4000 mm/4000 mm, and all subranges therebetween. The glass sheets in the panel or laminated structure may also have length/width dimensions of about 5 mm/5 mm to about 100 mm/100 mm, about 600 mm/600 mm, about 2300 mm/2600 mm, about 4000 mm/4000 mm, and all subranges therebetween. Additionally, adjacent glass sheets in a panel or laminated structure may have different length/width dimensions, allowing one sheet to overlap the other and on one or more sides of the sheets. Exemplary glass thicknesses for testing of single glass sheets or individual glass sheets included in a panel or laminate structure can be from less than 0.1 mm (e.g., as little as 10 micrometers) to greater than 5 mm, 0.1 mm to 3 mm, 0.4 mm to 2 mm, 0.5 mm to 1 mm, 0.5 mm to 0.7 mm thick.

The table 202 may include multiple drive mechanisms configured to move the glass sheet 204 to a predetermined position to initiate a measurement cycle or to continuously advance the glass sheet 204 through the test area 208 for dynamic mode testing.

A predetermined portion of the edge of the glass sheet 204 can be tested. The width of the predetermined portion can be from about 1 mm to about 5 mm, from about 1.5 mm to about 3.5 mm, from about 2 mm to about 3 mm, and all subranges therebetween. Since the primary purpose of testing the glass sheet 204 is to test the strength of the sheet material along its edge, it is generally desirable to test an area as close to the edge as possible. To this end, for a glass sheet having the dimensions described above, engaging the last 2 mm of the surface of the glass sheet 204 along its edge with the arched members 220a, 220b, and 218 generally ensures that the stress concentration is at the glass sheet edge. This configuration is also beneficial in minimizing any opportunity for trapping foreign particles (e.g., dust) between the glass sheet 204 and the arched members 220a, 220b, and 218 that could introduce undesirable surface cracks. In embodiments in which sheet 204 is a panel or laminate structure for which edge strength is being measured and adjacent glass sheets within the panel or laminate structure are different (e.g., one or more edges of the structure have overlapping features), the predetermined portion is measured for the smaller of the glass sheets within the structure (i.e., the sheet that does not overlap).

In some embodiments, the arched members 220a, 220b, and 218 in each or any assembly 206 can be made of a compliant material to minimize the risk of causing fracture of the glass sheet 204 during non-destructive testing (e.g., not evaluating maximum stress). The arched members or rollers can be selected to be sufficiently compliant while providing a long life to minimize maintenance and downtime, and sufficient friction to allow the rollers to roll freely over the glass surface. Exemplary arched member materials include hardened steel rollers, steel rollers, urethane rollers, polyetheretherketone (PEEK) rollers, Shore 80 urethane rollers, polycarbonate (PC) rollers (e.g., Lexan, etc.), high density polyethylene (HPDE) rollers, Shore 90 urethane rollers, urethane coated rollers, etc. Exemplary urethane rollers can also be employed to reduce rolling noise that may contaminate any signals, feedback, etc. used by the system. Additionally, urethane or urethane coated rollers can be used to accommodate debris in the path of the roller and ensure that there are no internal stress concentrations in the y-direction stress profile. In embodiments used to measure edge strength of panels and laminate structures, it has been found that less compliant rollers (e.g., PC, HPDE, etc.) are required to achieve adequate edge strength test results.

Exemplary dimensions for each arched member or roller may vary depending on the particular embodiment of the present subject matter. For example, roller dimensions may be 5 mm to 15 mm outside diameter (OD), 7 mm to 12 mm OD, 9 mm to 10 mm OD. In some embodiments, exemplary roller dimensions may be approximately 9 mm OD so that the stress is applied almost entirely to the corners of the glass sheet (this is important since many customer issues occur in this area). Exemplary systems may also traverse the glass edge at speeds of 5 mm/sec to 500 mm/sec or more, or 2 mm/sec to 400 mm/sec or more. Exemplary systems are robust and can be used on glass having thicknesses as low as 0.1 mm to 1 mm, depending on the durability of the polymeric material used for the rollers.

In some embodiments, a high speed closed loop stress control mechanism can be employed to detect cracks and ensure that the applied stress is within a predetermined target, e.g., a target of 2 MPa. For example, a single arched member 218 can be used to apply a load to the glass sheet 204, which can send a load cell signal to a high speed controller (not shown) that continuously monitors for cracks. The load cell signal can also be used to control the load applied while traversing the edge at a predetermined speed (e.g., 5 mm/sec to 100 mm/sec or more).

The use of the edge strength testing apparatus 200 is not limited to the examination of edge features alone. It is contemplated that embodiments may examine surface features as well to perform defect screening on the production line. For example, any features on the surface of a glass sheet, such as particle contamination and/or visible types of surface defects such as dents, chips, or scratches, may be employed with embodiments of the present subject matter. However, in such embodiments, rather than providing an intensity distribution, these embodiments would utilize a size, shape, and/or depth distribution, i.e., dimensional metrics, of the surface defects. Exemplary, non-limiting surface features include the near surface region (e.g., about 20 mm in from the edge) and the border region (where the surface meets the edge), as well as any size, shape, or depth characteristics of the surface defects. Such dimensional metrics may be used alone or in conjunction with strength metrics obtained from edge features.

In some embodiments, the tested length may span the entire edge of the glass sheet, or the test may be performed on one or more portions of the edge of the glass sheet. Thus, the tested length may range from as little as about 1-5 mm to as much as about 2600 mm, 3000 mm, 4000 mm or more, depending on the length of the glass edge.

3-Point Bend Based Edge Strength Testing Figures 4-7B, 9A-9C, and 11A-11D show a schematic diagram of an edge strength testing apparatus 200 using a 3-point bend method. In the destructive failure testing mode, a test sample glass panel 204 is placed under increasing loads until failure and the peak load at the point of failure is recorded. This peak load is mapped to stress based on a calibration curve previously empirically generated from strain gauges. This technique provides strain/stress along two dimensions (x and y axes), which is a significant improvement over the current practice of only using strain along the bending direction by strain gauges.

[Calibration of the optical system (camera)]
With reference to Figures 11A-11D, the calibration of the system is completed before performing the actual test on the glass panel. By performing this operation, the image correlation software knows the angle and distance at which the camera is positioned relative to the test sample. This helps to translate the movement of the dot pattern with respect to the image pixels into physical dimensions in 3D space. This calibration step is illustrated diagrammatically in Figures 11A-11D. Figure 11A shows the relative positions of the camera (222a, 222b), rollers, and test panel. The upper roller is relatively far away from the panel since the test has not yet started. The test panel is then switched to a predefined pattern (known to the image correlation software) printed on a flat surface. The camera takes a series of pictures of this pattern, where the pattern is pivoted in 3D space while remaining within the focal range of the camera, as shown in Figures 11B-11D. Using the series of images and appropriate software, a calibration file can be created, which is used during the actual test on the panel to obtain the distortion.

[Static test configuration capabilities]
The static mode has the following design features that help it act as a calibration device (or benchmark device) for studying panel/laminate designs along with other parametric studies:
a) Interchangeable rollers (roller material and diameter).
b) Ability to change the feed angle, which helps to address edge testing of non-rectangular display panels.
c) Adjustment of the roller assembly closer to or further away from the edge being tested, allowing testing of smaller panels (mobile devices) as well as inspecting specific areas of the edge.
d) The roller can be replaced with a partial roller profile, which increases the visible area, especially near the contact point, and allows the study of the stress distribution in that area without additional stress from the roller.

This incorporates alignment techniques such as alignment pins 236a, 236b that precisely align the various components, which is very important for accurate testing.

[Improving Stress Measurement Accuracy in Non-DIC ESMS Tests]
Load Cell Location In the existing ESMS, the load cell is positioned adjacent to the upper roller assembly. The inventors have discovered that improved control of the load can be achieved by moving the load cell below the lower roller 218. Figure 15 is a schematic showing the new location of the load cell. With this new configuration, no load cell drift was observed.

Speed Control The stress on the sheet 204 is a function of the load applied to the sheet 204. Thus, the accuracy of the stress measurement depends on the accuracy of the load measurement. In non-DIC ESMS testing, where the test sample sheet 204 is continuously fed into the test area 208 of the tester, one factor for improving accuracy is the speed at which the test sheet 204 is moved through the test area 208. Transport of the test sheet 204 is achieved by using rollers in the arched members 218, 220a, 220b. Thus, the material of the rollers can affect the friction between the rollers and the test sheet, which can affect the ability of the rollers to move the test sheet 204. Compliant rollers can expand and contract as they rotate relative to the test sheet 204, causing fluctuations in the load applied by the lower roller 218. When the test sheet is moving quickly through the rollers, the ESMS test system 200 has little time to adjust to the load fluctuations. The inventors investigated the effect of the speed at which the test sheet 204 moves over the rollers on the load control using a commercially available load frame Instron. The results are shown in the plots of Figures 17 and 18. The data was generated using 10 test specimens. The plot in Figure 17 shows the load measured over time at a speed of 10 mm/sec. The plot in Figure 18 shows the load measured over time at a speed of 1 mm/sec. Reducing the speed from 10 mm/sec to 1 mm/sec significantly improved the load control. Load control also depends on the speed and accuracy with which the ESMS test system software can adjust to load variations. Thus, the optimum speed must be selected taking these factors into consideration. With the improved ESMS configuration of the present invention, the inventors were able to achieve good load control at speeds between 1 mm/sec and 40 mm/sec. Using a speed of 5 mm/sec, for 0.2 mm monolithic glass, we were able to achieve load variations of less than ±0.1 N with a target of 2 N. This corresponds to a stress variation of 100 ±5 MPa. Faster speeds would result in shorter test times, but some maximum limit is necessary to avoid compromising the ability to control the load.

Polymer Coating on Rollers In embodiments where the arched members 218, 220a, 220b are rollers, the rollers preferably have a polymer coating on the rolling surface that contacts the test sheet 204 to aid in the movement of the test sheet 204. Because polymers are viscoelastic, the material may expand and contract while rolling. Thus, the inventors expected that a thinner polymer coating on the rollers would provide better control of the pressure applied to the test sheet 204. However, unexpectedly, the inventors discovered that a thicker polymer coating provided better control of the load and less variation in the applied load. The overall roller diameter (roller bearings + polymer coating) was 9 mm. The polymer coating thickness was 1.32 mm.

The inventors further investigated the effect of roller diameter and polymer coating thickness on the load-stress correlation to understand potential variability in dynamic measurements. Modeling results showed that for a given roller inner diameter L2 and outer diameter L1, the change in correlation is negligible up to 400 MPa when the roller inner diameter is varied from 4 mm to 14 mm and the polymer thickness is varied from 0.5 mm to 2.5 mm. Figure 24 is a plot of the load-stress correlation as a function of roller outer diameter L1/roller inner diameter L2 as determined by FEA. The correlation does not vary up to 400 MPa for a wide range of L1 and L2. From this study, the inventors could see that the potential variation in the load-stress correlation is minimal for dynamic measurements.

[Software Improvements]
Improvements were made to the control software of the ESMS system 200 to fully automate the operation of the system to make the testing operation more reliable. The system targets small test sheets/panels, i.e., ∼200x130x0.2mm. Test samples with a thickness of only 0.2mm require extremely accurate control of the load to accurately measure stress since even small changes in load induce relatively large changes in stress. Therefore, improvements had to be made to the control software by improving the control loop to maintain better control of the load. This control software enabled the ESMS to perform the following automated procedure: One of the initial steps in the edge strength test procedure is the alignment of the test sheet 204. The desired roller engagement is at a distance D (typically at about 2mm) from the edge E of the test sheet 204. This roller engagement position is shown in Figure 19B.

Before edge strength testing can begin, several pre-scanning routines are performed. Such routines include generating a thickness profile of the edge of the test sheet 204 to be tested. Referring to FIG. 19A, the length of the test sheet 204 is slowly moved past a pair of thickness sensors 410a, 410b that measure the thickness of the test sheet 204 to generate a thickness profile of the edge of the test sheet 204 to be tested. This thickness data allows for accurate post-measurement analysis of the failure stress if the test sheet 204 is cracked. This routine also accounts for the leading edge E _L and trailing edge E _T of the test sheet 204, as well as the length of the sample sheet, so that actual dimensions rather than specification dimensions can be used during edge strength testing. The test sheet 204 is positioned such that the leading edge E _L is positioned above the apex of the lower roller 218 at the start of the edge strength testing procedure. FIG. 19C illustrates this configuration.

The ESMS then drives the test sheet 204 in the direction indicated by arrow M until the apex of the lower roller 218 is aligned with the starting offset position, which refers to a horizontal edge alignment as shown in FIG. 19C. FIG. 20 is an illustration of an example test sheet 204 identifying the arrival zone, measurement length, and take-off zone. To begin the edge strength test procedure, the ESMS slowly drives the upper roller assembly 220a, 220b downward toward the top surface of the test sheet 204 until a small force is detected, signifying that the upper rollers 220a, 220b have made contact with the test sheet 204.

The load application force and the velocity of the glass test sheet are then synchronously increased through the arrival zone over the specified arrival distance within a specified time frame. The load application force is maintained along the measurement length until the take-off zone is reached. The length of the take-off zone is defined by the take-off distance, and the load application force and the velocity of the glass test sheet are decreased to complete the test procedure. FIG. 21 is a plot showing an example of a load profile starting from an initial period during the descent of the upper rollers 220a, 220b, continuing through the arrival zone, the measurement length, and the take-off zone.

Crack Detection Under the updated control software, the ESMS performs the following crack detection procedure: The user specifies a threshold value around the target load. If the threshold value is breached, the ESMS immediately disengages the rollers 218, 220a, 220b and stops the test sheet 204. The position of the lower roller 218 where the test sheet 204 is under the applied stress (i.e., the apex of the roller 218) is defined as the new crack location. If enough space is left along the edge of the test sheet 204, taking into account the user-defined crack offset, the edge strength test continues. Figure 22 is a schematic diagram of the test sheet 204 showing the location of the new crack detected. The crack offset refers to the distance the rollers are moved away from the detected crack location before they can re-engage the edge of the test sheet to resume the dynamic test. If the crack offset is too short, the crack will propagate and interfere with the measurement. If the crack offset is too large, it may waste the testable edge as a physical asset. According to the modeled magnitude and shape of the stress field, we used a minimum distance of 25 mm as the crack offset.

After re-engaging the test sheet 204, the ESMS loops through all of the user-defined load application forces as long as a sufficient length of the edge area of the test sheet remains under test. The ESMS applies a constant force as it traverses along the edge. When failure occurs, the strength of the test sheet 204 can be between zero and the stress corresponding to the applied force. Therefore, to accurately determine the strength of the sheet, multiple load steps must be applied in increments to better understand the lower limit of strength. The user can define these force steps based on the level of measurement accuracy the user desires. The system must loop through all of the user-defined load steps until failure occurs.

[Load Control] A phantom motor technique (also known as a cascading servo loop) was incorporated into the ESMS control software to control the three-point bending stress load applied to the test sheet 204 by the rollers 218, 220a, 220b. The phantom motor technique was applied to the analog input of the load cell. This technique allowed the ESMS to read the input signal of the load cell and make changes to the moving sample stage 202 to control the load being applied to the test sheet 204 in real time.

Figure 23A is a schematic diagram showing the phantom motor technique. The Delta Tau Power PMAC motion controller can be configured to allow the output of one independent servo loop to be used as the input of another servo loop. This allows the system to combine the functionality of both loops in a single actuator and allows engineers to make the most of the controller's fine tuning parameters for a device that was not specifically intended to control motion. In this case, the load cell's output was integrated into the motion system such that its analog output was directed to the output of a servo loop in the controller. This is known as the "inner loop." The output of the inner loop is then fed to the servo loop that actually controls the physical motion of the actuator to command the physical motion. By scaling the relationship between the two loops and the motor fine tuning parameters provided by the Delta Tau system, an optimal balance is achieved that allows the motion controller to apply the necessary corrections in response to real-time deviations from the target load, which aids in the application of a stable load across the sample.

Figures 23B and 23C show plots of measured load readings from a load cell versus time comparing load cell readings from one test run. The plot in Figure 23B is for measured load readings obtained without the phantom motor technique correction, and the plot in Figure 23C is for measured load readings obtained with the phantom motor technique. The plot in Figure 23C shows that the variation in the load cell readings is significantly reduced when the phantom motor technique is implemented.

Method for estimating the strength of a batch of panels
According to some embodiments, a method is disclosed for measuring the average strength of multiple individually manufactured panels having the same design but manufacturing tolerances. There can be two ways to measure the strength of multiple panels to obtain the average strength of a batch:
- Directly measure the strength of every individual panel using the optical distortion measurement technique disclosed in this invention;
- Using the nominal load-stress relationship, determined by measuring several samples using the optical strain measurement technique described above, estimate the strength of all remaining panels that make up the batch.

The first method has the advantage of being more accurate than the second method, but can be time consuming since all samples must be measured using optical distortion measurement techniques, which have multiple process steps and can be relatively time consuming. The second method has a relatively short processing time since the load can be used to directly estimate the strength, but the same nominal correlation is employed for panels with specific manufacturing tolerances, resulting in error bars in the estimation. Nevertheless, in practice, the second method is preferred since the advantage of the shorter test time outweighs the additional accuracy gained by utilizing the first method. The second method is described in further detail herein.

A nominal load-stress correlation can be obtained by performing multiple optical distortion measurements (i.e., DIC) on multiple panels using either a static or dynamic DIC approach. A larger number of measurements will provide a more accurate estimate of the nominal correlation. As the number of measurements increases, the error bars also increase due to the manufacturing tolerances of the panels.

For example, Figure 25 shows a plot of stress versus load for two different batches of panels with average thicknesses of 0.2 mm and 0.3 mm, measured by DIC and modeled by FEA. The DIC measurement results have variability due to the manufacturing tolerances of the panels, but the FEA model only shows a single result because the nominal panel design parameters were used in the calculation.

To minimize the error from using the nominal load-stress correlation for all panels, a correction factor can be applied to the nominal load-stress correlation for each panel based on the measured variability of each panel.

[Panel correction coefficient]
Thickness - For example, if a nominal load-stress correlation derived from a batch of panels with an average thickness of 0.3 mm is applied to a panel with a thickness of 0.29 mm, the stress may be underestimated. Therefore, a correction factor can be applied based on the measured thickness of 0.29 mm. In this case, the correction factor is a function of the measured panel thickness and the average thickness of the panel batch. The mathematical form of the correction factor can be derived from sensitivity studies performed by DIC and/or FEA modeling.

The effect of panel manufacturing tolerances can be difficult to characterize using experimental methods such as DIC when the tolerances are small. In such cases, FEA modeling can be used to understand the sensitivity and derive correction factors. Figure 26A shows the modeled load-stress relationship for panels with different thicknesses of 0.1, 0.2, 0.21, 0.22, and 0.25 mm, and Figure 26B shows the normalized change in stress versus load (N) for these different thicknesses. Using these graphs and measuring the variation in thickness for each panel, appropriate correction factors can be applied to the nominal load-stress relationship.

Similarly, correction factors may be applied to other panel design parameters that have specific manufacturing tolerances.

[Measurement process correction factor]
Measurement location-measurement process may also involve measurement variability that should be taken into account as a correction factor. For example, Figures 27A and 27B show force and displacement curves versus time for glass sheets with thicknesses of 0.1, 0.15, 0.2, and 0.5 mm when tested with a dynamic ESMS at constant load. Although the force remained constant for all thicknesses, the displacement curves showed significantly different behavior as the roller approached the corner, as shown by the increasing behavior over time. This indicates that the stiffness at the corner decreases as the glass thickness decreases, which implies that the load-stress correlation must be corrected for measurement locations along the edge. Referring to Figure 28, a measurement location d along the edge of a panel 204 is measured from the center C of the edge of the panel to the center of the test area 208, which is marked with an "X" in Figure 28.

Roller engagement - Variability in roller engagement can also be considered as a correction factor. Roller engagement refers to the engagement length between the roller and the panel edge. Roller engagement is typically set to a nominal value, but can vary depending on the edge straightness and alignment process. Figure 29 shows an example of the effect of roller engagement on the load-stress correlation. When the engagement varies between 0.7mm and 1.7mm, this had a 10% effect on the load-stress correlation.

Off-Apex Failure (Static ESMS only) - For static ESMS, the off-apex failure location must be considered as a correction factor for accurate stress measurements since the stress drops rapidly away from the apex of the roller. Figure 30A is a photograph of an example of an off-apex failure on a test panel, and Figure 30B is a plot of the stress profile for the off-apex failure location.

Mathematically, a correction factor can be applied to the nominal load-stress relationship, for example as follows:
Stress=F(N)×C1(t)×C2(x)×C3(R)×C4(L)×…
where F is the nominal load-stress relationship and C1-C4 are correction factors that take into account each of the individual variables: thickness (t), measurement location (x), roller engagement (R), and off-apex failure location (L).

Error Bars - Variability in certain panel design parameters or measurement process that cannot be easily measured for practical reasons can be accounted for as error bars. For example, variability in sealant width, position, or thickness of the individual pieces of glass that make up the panel can be accounted for as error bars. Additionally, variability in the measurement process such as errors from load cells, operators, etc. can also be accounted for as error bars. The final load-stress relationship that accounts for both the correction factors and error bars is shown below:
Stress = F(N) x C1(t) x C2(x) x C3(R) x C4(L) x ... ± error bar [Method to estimate the probability of panel failure]
[1. Derivation of strength distribution based on ESMS and prediction of product reliability]
Traditional methods for predicting the reliability of glass in a particular application include deriving a strength distribution based on strength tests, and then scaling the probability of failure from a given reference size to the entire production size. This approach has been used successfully for many years, but it relies on the strength distribution being referenced to a given size. Traditional means of testing glass strength, such as four-point bending and ring-on-ring, performed on glass specimens with sufficient thickness, have the advantage that the stress field in the tested area is constant. Therefore, when deriving a strength distribution from this data, the resulting distribution is referenced to the tested area (inside the inner ring in the ring-on-ring test, and inside the inner knife in the four-point bending test). Using these distributions for reliability prediction is correct in this case, since the reference size of the strength distribution relative to the product size is known, and therefore the probability of failure can be easily scaled to the product size using the Weibull scaling (or closure) property. To use this approach, it is necessary to know the reference size to which the Weibull strength distribution applies.

Various new testing approaches, such as ESMS, are being used to measure the strength of thin glass. Although ESMS is suitable for evaluating thin glass edge strength, it does not have a constant stress field. In static ESMS testing, the stress field is symmetric around the center roller and is shown in Figure 31, which shows the stress along the edge of the panel/sheet, where the location and stress of the failure site are shown as stars. A common mistake is to generate a strength distribution based only on the stress at the failure location. The problems with this approach are as follows:
- Because the stress is maximum at the center roller and decreases monotonically away from this location, the resulting strength is determined by the weakest point among the "center ± distance from break to center". The distribution is therefore that of the weakest flaw among the break. Since the location of the weakest flaw varies, the tested length from which the distribution is obtained also varies. The distribution is therefore not based on a specific length.

The length of the edge that can withstand stresses higher than the failure stress must be considered as this can have a dramatic effect on the strength distribution.

As a result, a distribution based solely on the stress at the break cannot be used to predict product reliability. Therefore, special methods are required to derive a robust strength distribution. To properly determine the strength distribution for an ESMS, the effects of length and stress must be deconvoluted and the distribution must be referenced to a specific unit length. A methodology was developed to achieve this. This methodology includes: defining a short reference length (or reference area in the case of surface reliability); and identifying the specific reference length (or area) that broke, the stress at break, and the stress experienced by the specific reference length (or area) that survived. The resulting data is then statistically analyzed to derive a robust strength distribution for a known length (or area). This distribution can then be used for reliability prediction purposes. Details of this approach are not provided here.

"Simplified method" for dynamic ESMS data analysis
For dynamic ESMS, we also implemented a simplified method to derive the strength distribution by using the first failure data. This simplified method assumes that the stress profile is constant for dynamic ESMS over the entire test area. Figure 32 shows a stress vs. probability plot showing the strength distribution of a batch of panels analyzed by the "simplified method" or the "special" method outlined in the above section. There is no significant difference between the results from both methods because at each step load, the stresses in the measured area over the entire test length are nearly the same and the step stresses are small. This allowed us to use traditional Weibull statistics in dynamic ESMS data analysis without relying on the "special" method.

[Panel damage mode]
Panel failure modes refer to the various ways or locations where a panel can fail when tested with an ESMS. In general, it is preferable to have a consistent or predictable failure location on a panel in order to classify the failure. However, due to the complexity of panel construction and flaw collection, it can be difficult to identify the failure location and understand the exact failure stress. Here we provide examples of failure modes found in panels and methods to estimate the failure stress of a panel.

Edge failure - Edge failure is when the edge of surface 1 breaks under tension (Figure 33). Edge failure of surface 1 under tension is the failure mode intended to be obtained from the ESMS test. However, to obtain this failure mode, it is necessary for the two pieces of glass that make up the panel to collectively behave like a single piece of glass. If the two pieces of glass that make up the panel behave like two separate pieces of glass, edge failure may also occur at the edge of surface 3, complicating the identification of the fracture location. The behavior of the panel as a single piece of glass can be achieved by ensuring that the sealant that bonds the two pieces of glass is sufficiently cured and is sufficiently wide and close enough to the panel edge to be measured. For example, the sealant must not delaminate due to the shear stress applied by the test, the modulus of the sealant must be sufficiently high, typically greater than 1 GPa, and the sealant must adequately bond the area around the edge where the load is applied.

This behavior of the panel with respect to the sealant properties can be better explained by the change in the stress profile on surfaces 1 and 3 with the variation of the sealant-edge distance. Figure 35 shows the stress profile on surfaces 1 and 3 when the gap between the sealant and the glass edge is 0 mm. The sealant is about 1 mm wide and has a stiffness of 3 Gpa. The peak stress is at the edge of surface 1 and the edge stress on surface 3 is less than half. This implies that failure will occur only on surface 1. Figure 36 shows the stress profile on surfaces 1 and 3 when the distance between the sealant and the glass edge is increased to 2 mm. The peak stress is now present on both the edges of surfaces 1 and 3 as the sealant has moved away from the load application area. The probability of failure at the edge of surface 3 is the same in this case as for surface 1.

In some cases, it may be preferable to test both edges of surfaces 1 and 3 simultaneously. In such cases, a panel may be designed in which the sealant does not have the above-mentioned properties. As shown in Figure 36, the failure stress for both edges will be calculated based on the same correlation since the glass behaves like two identical pieces of glass.

Surface failure - Surface failure is a fracture that occurs on the surface of surface 1 under tension near the edge, typically within 3 mm of the edge (Figure 33). If the panel behaves like two separate pieces of glass, failure may also occur on the surface of surface 3. As with edge failure, if it is preferred to test both surfaces 1 and 3 simultaneously, panels may be designed in which the sealant does not have the properties described above and the stresses estimated using the same correlation.

Failure from compressive side surface - Failure from compressive side surface refers to failure at the edge or surface of surface 4 due to the small tensile stress applied by the two outer rollers (Figure 33). Due to the balance of forces, the tensile stress under each outer roller is half the stress applied by the center roller. Even though the stress on surface 4 is relatively small, failure can still occur at the edge or surface of surface 4 if a large flaw is present. This can be used to test for flaws on both surfaces 1 and 4 simultaneously, although it is still generally desirable to have failure occur only on surface 1. This potentially eliminates the need to use two sets of rollers to test both the A and B sides.

Adjacent crack-adjacent crack occurs when a pre-existing crack extends into the load application area (Figure 34). In effect, the pre-existing crack propagates into the adjacent test area and subsequently initiates failure. This failure mode often occurs when the test areas are too close together. The way to mitigate this failure mode is to ensure that there is sufficient distance between adjacent fractures and between test locations.
Gauge Failure Modes Gauge failure modes describe the manner in which a gauge may fail to detect a breach on a panel or may falsely detect a breach on a panel.

Missed break detection - Missed break detection refers to when the ESMS fails to detect a break on a panel. The inventors have found that this often occurs when the fracture energy from the crack is small relative to the sensitivity of the load cell. In some cases, the crack may remain bonded by the underlying sealant that bonds the two pieces of glass together.

False break detection - A false break detection refers to when the ESMS detects a break when in fact no break has occurred. The inventors have found that this often occurs when the load cell is too sensitive to noise from the environment.

Machine Learning to Identify Fractures In crack detection in ESMS, panel samples are captured using a camera with different illuminations and magnifications. It is difficult to identify manually created features to define panel cracks from image frames. The captured video samples have most image frames as glass without cracks. From the collected data, there are only a limited number of image frames of broken glass. This problem is treated as anomaly detection using unsupervised learning. In other words, the machine learning model is trained using normal images to find common features. Then, an encoder and a decoder are used to detect outliers and treat the outlier images as abnormal images. In this case, the images with cracks will be detected as abnormal images. Figure 37 shows the machine learning model.

There are three convolutional neural networks. Two of the networks, G and D, are adversarially unsupervised learning with training data that consists only of the target class. Specifically, G learns to reconstruct positive samples and tries to fool detector D using feature weights z and G(z). D, on the other hand, learns to distinguish original (positive) samples from reconstructed samples. In this way, D can be used to distinguish between positive and novel classes since it only learns concepts characterized by the space of all positive samples. There is also another neural network E, which is trained to mirror the behavior of G, which is able to extract accurate feature weights from images. G, on the other hand, learns to efficiently reconstruct positive samples and acts as a decimator (or informally a distorter) for negative (i.e. novel) samples, since it cannot accurately reconstruct the input for them. In the testing phase, D acts as the actual novelty detector, while G improves the detector's performance by fully reconstructing the positive, or target, samples and decimating (or distorting) any given negative, or novel, sample.

For our ESMS case, if the input image is x (image), this will go into neural network E to produce feature weights E(x). Define Z equal to E(x), then Z will be passed through neural network G to become G(z) (image). If (G(z),z) and (x,E(x)) are passed through neural network D and achieve similar scores, and (G(z),z) is passed through D and achieves a higher score, then the input x is normal/positive; otherwise it is abnormal/negative.

The embodiments and functional operations described herein may be implemented in digital electronic circuitry, or computer software, firmware, or hardware, or a combination of one or more of these, including the structures disclosed herein and structural equivalents thereof. The embodiments described herein may be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a tangible program carrier for execution by or control of the operation of a data processing apparatus. The tangible program carrier may be a computer-readable medium. The computer-readable medium may be a machine-readable storage device, a machine-readable storage substrate, a memory device, or a combination of one or more of these.

The term "processor" or "controller" can encompass all apparatus, devices, and machines for processing data, including, for example, a programmable processor, a computer, or multiple processors or computers. A processor can include hardware as well as code that establishes an execution environment for the computer program in question, such as code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of these.

A computer program (also known as a program, software, software application, script, or code) can be written in any type of programming language, including compiled or interpreted languages, and can be deployed as a standalone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a single file in a file system. A program can be stored in part of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), or in multiple coordinated files (e.g., a file that contains one or more modules, subprograms, or code portions). A computer program can be deployed to be executed on one computer or on multiple computers, either at one location or distributed across multiple locations and interconnected by a communications network.

The processes described herein may be performed by one or more programmable processors executing one or more computer programs to perform functions by manipulating input data and generating output. The processes and logic flows may also be performed by, and an apparatus may be implemented as, special purpose logic circuitry, for example an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit).

Processors suitable for executing a computer program include, by way of example, both general purpose and special purpose microprocessors, as well as any one or more processors of any kind of digital computer. Typically, a processor receives instructions and data from a read-only memory or a random access memory, or both. The essential elements of a computer are at least one processor for executing instructions, and one or more data memory devices for storing instructions and data. Typically, a computer will include one or more mass storage devices (e.g., magnetic, magneto-optical, or optical disks) for storing data, or will be operatively coupled to receive data from and transmit data to the mass storage devices, or both. However, a computer need not have these devices. Additionally, a computer can be embedded in another device, a mobile phone, a personal digital assistant (PDA), to name a few.

Computer-readable media suitable for storing computer program instructions and data include all forms of data memory, including non-volatile memory, media and memory devices, such as semiconductor memory devices (e.g., EPROM, EEPROM and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and CD-ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.

To provide for user interaction, as shown in the drawings included herein, the embodiments described herein can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, etc.) for displaying information to a user, and a keyboard and pointing device (e.g., a mouse or trackball or touch screen) for providing input to the computer by the user. Other types of devices can also be used to provide for user interaction. For example, input from the user can be received in any form, including acoustic, speech, or tactile input.

The embodiments described herein can be implemented in a computing system that includes a back-end component (e.g., a data server), or includes a middleware component (e.g., an application server), or includes a front-end component (e.g., a client computer having a graphical user interface or web browser through which a user can interact with an implementation of the subject matter described herein), or includes any combination of one or more of the above-mentioned back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN) and a wide area network (WAN) (e.g., the Internet).

A computing system may include clients and servers. Clients and servers are generally remote from each other and typically interact through a communication network. The relationship of client and server is created by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

It will be understood that various embodiments of the present disclosure may involve specific features, elements, or steps that are described in connection with a particular embodiment. It will also be understood that certain features, elements, or steps, although described in connection with a particular embodiment, may be interchanged or combined with alternative embodiments in combinations or permutations not illustrated.

It should also be understood that, as used herein, the terms "the," "a," or "an" mean "at least one" and should not be limited to "only one," unless expressly stated otherwise. Thus, for example, reference to "a component" includes examples having two or more such components, unless the context clearly indicates otherwise.

As used herein, ranges may be expressed as from "about" one particular value and/or to "about" another particular value. When such a range is expressed, examples include from the one particular value and/or to the other particular value. Similarly, when values are expressed as approximations, by use of the antecedent "about," it will be understood that the particular value forms another aspect. It will be further understood that the endpoints of each of the ranges are significant both in relation to the other endpoint, and independently of the other endpoint.

As used herein, the terms "substantially," "substantially," and variations thereof are intended to describe that a described feature is equal or approximately equal to a value or description. Additionally, "substantially similar" is intended to refer to two values being equal or approximately equal. In some embodiments, "substantially similar" can refer to values within about 10% of each other, e.g., within about 5% of each other, or within about 2% of each other.

Unless expressly stated otherwise, it is not intended that any method described herein be construed as requiring that its steps be performed in a particular order. Thus, if a method claim does not actually recite the order in which its steps are to be followed, or it is not specifically stated in the claim or description that the steps are to be limited to a particular order, no particular order is intended to be inferred.

Although various features, elements, or steps of a particular embodiment may be disclosed using the transitional phrase "comprising," it should be understood that alternative embodiments are also implied, including those that may be described using the transitional phrase "consisting of" or "consisting essentially of." Thus, for example, alternative embodiments implied for a device comprising A+B+C include embodiments in which the device consists of A+B+C, and embodiments in which the device consists essentially of A+B+C.

It will be understood by those skilled in the art that various modifications and variations can be made to the present disclosure without departing from the spirit and scope of the present disclosure. Since modifications, combinations, subcombinations and variations of the embodiments of the present disclosure that incorporate the spirit and content of the present disclosure may occur to those skilled in the art, the present disclosure is to be construed as including all that are within the scope of the appended claims and their equivalents.

The following describes preferred embodiments of the present invention.

EMBODIMENT 1
1. An apparatus for testing edge strength of sheets of material, the apparatus comprising:
a plurality of assemblies configured to selectively apply a three-point bend load along an edge of the sheet of material in a test area of the apparatus, the plurality of assemblies being capable of establishing a loaded condition on the sheet of material in the test area by applying the three-point bend load and of establishing an unloaded condition on the sheet of material in the test area by not applying the three-point bend load;
11. An apparatus comprising: a detection mechanism configured to optically measure a strain in the sheet of material in the test area when the sheet of material is in the unloaded state and when the sheet of material is in the loaded state, the strain in the loaded state being produced by the three-point bending load; and a processor configured to determine a stress in the sheet of material based on the measured strain.

EMBODIMENT 2
the plurality of assemblies comprises two opposing assemblies, a first of the two opposing assemblies comprising a single arched member for engaging a first side of the sheet, and a second of the two opposing assemblies comprising two spaced apart arched members for engaging a second side of the sheet opposite the first side;
2. The apparatus of embodiment 1, wherein the two spaced apart arcuate members define the test region between the two spaced apart arcuate members.

EMBODIMENT 3
3. The apparatus of embodiment 2, wherein the single arcuate member is vertically aligned between the two spaced apart arcuate members.

EMBODIMENT 4
the detection mechanism includes a first pair of cameras positioned to capture one or more images of a surface of the second side of the sheet within the test area between the two spaced apart arcuate members;
2. The apparatus of claim 1, wherein the detection mechanism optically measures the distortion in the sheet of material from an image of the surface of the second side of the sheet acquired with a first pair of cameras.

EMBODIMENT 5
3. The apparatus of embodiment 2, wherein the arcuate member is a non-rotating bushing, a cylindrical roller, a belt roller, or a bearing roller.

EMBODIMENT 6
2. The apparatus of embodiment 1, wherein a pattern provided on the surface of the sheet in the test area facilitates optical measurement of the strain in the sheet of material.

EMBODIMENT 7
5. The apparatus of embodiment 4, wherein the detection mechanism further comprises a second pair of cameras positioned to capture images of the surface of the second side of the sheet in an area other than the test area.

EMBODIMENT 8
the detection mechanism includes at least one camera positioned to capture one or more images of a vertical edge of the sheet within the test area between the two spaced apart arcuate members;
2. The apparatus of claim 1, wherein the detection mechanism optically measures the distortion in the sheet of material from the image of the vertical edge of the sheet acquired with the at least one camera.

EMBODIMENT 9
9. The apparatus of embodiment 8, wherein the detection mechanism further comprises at least one additional camera positioned to capture images of the vertical edge surface of the sheet in an area other than the test area.

EMBODIMENT 10
1. A method for testing edge strength of a sheet of material along an edge of said sheet, said method comprising:
applying a surface pattern of visual markers onto a surface of the sheet in an area of interest along the edge of the sheet;
obtaining a first optical image of the surface of the sheet in the region of interest without applying any three-point bending load to the region of interest;
applying a three-point bending load to the area of interest along the edge of the sheet;
acquiring a second optical image of the surface of the sheet at the area of interest while applying the three-point bending load to the area of interest; and determining a stress on the surface of the area of interest along the edge of the sheet due to application of the three-point bending load based on the first optical image and the second optical image.

EMBODIMENT 11
The step of determining stress on the surface of the area of interest along the edge of the sheet comprises:
measuring a strain in the sheet of material in the region of interest in an unloaded state based on the first optical image;
11. The method of claim 10, comprising: measuring a strain of the sheet of material in the region of interest in a loaded state where the three-point bending load is applied based on the second optical image; and determining the strain of the sheet induced by the applied three-point bending load by comparing the measured strain in the unloaded state with the measured strain in the loaded state.

EMBODIMENT 12
The step of measuring the strain of the sheet of material in the region of interest under the applied load based on the second optical image comprises:
12. The method of claim 11, comprising determining a displacement of the visual marker in the second optical image compared to a position of the visual marker in the first optical image.

EMBODIMENT 13
The step of determining the stress on the surface of the area of interest along the edge of the sheet comprises:
13. The method of embodiment 12, comprising calculating the stress required to produce the measured strain of the sheet of material in the region of interest.

EMBODIMENT 14
11. The method of claim 10, wherein the step of applying the surface pattern of the visual markers comprises printing, coating, spraying, etching, pasting, or projecting an image onto the surface of the sheet.

EMBODIMENT 15
11. The method of embodiment 10, wherein the first optical image and the second optical image are acquired using an optical system including at least one camera.

EMBODIMENT 16
1. A method for testing edge strength of a sheet of material along an edge of said sheet using a testing device having a test area, said method comprising:
applying a surface pattern of visual markers onto a surface of the sheet in an area of interest along the edge of the sheet;
continuously advancing the sheet of material through the testing apparatus, the edge of the sheet advancing through the testing area while a three-point bending load is applied to the portion of the sheet passing through the testing area;
acquiring a first optical image of the surface of the sheet at the area of interest while the edge of the sheet continues to advance through the test area, but before the area of interest reaches the test area;
acquiring a second optical image of the surface of the area of interest of the sheet as the area of interest advances into the test area while applying the three-point bend load; and determining a stress on the surface of the area of interest along the edge of the sheet due to application of the three-point bend load based on the first optical image and the second optical image.

EMBODIMENT 17
The step of determining stress on the surface of the area of interest along the edge of the sheet comprises:
measuring a strain in the sheet of material in the region of interest in an unloaded state based on the first optical image;
17. The method of claim 16, comprising: measuring a strain of the sheet of material in the region of interest in a loaded state where the three-point bending load is applied based on the second optical image; and determining the strain of the sheet induced by the applied three-point bending load by comparing the measured strain in the unloaded state with the measured strain in the loaded state.

EMBODIMENT 18
The step of measuring the strain of the sheet of material in the region of interest under the applied load based on the second optical image comprises:
18. The method of embodiment 17, comprising determining a displacement of the visual marker in the second optical image compared to a position of the visual marker in the first optical image.

EMBODIMENT 19
20. The method of claim 18, wherein determining the stress on the surface of the area of interest along the edge of the sheet comprises calculating the stress required to produce the measured distortion of the sheet of material in the area of interest.

EMBODIMENT 20
17. The method of embodiment 16, wherein the first optical image and the second optical image are acquired by an optical system comprising a pair of stereo cameras.

EMBODIMENT 21
17. The method of claim 16, wherein the step of applying the surface pattern of the visual markers comprises printing, coating, spraying, etching, pasting, or projecting an image onto the surface of the sheet.

200 Apparatus, test device 202 Table, sample stage 204 Test sheet, test sample sheet, glass sheet, sheet, sheet of material 204e Edge, vertical edge 205 Edge 206 Assembly 206a Assembly, first assembly 206b Assembly, second assembly 207b, 207c, 207d, 213b, 213c, 213d Roller assembly 208 Test area 214 Processor 215 Roller 216 Peak strain position 217 Far field position 218 Lower roller, lower roller, arched member, roller 220a, 220b Upper roller, upper roller, arched member, roller 222 First optical system, optical system 222a, 222b CCD camera, camera, stereo camera 222c, 222d Camera 224a, 224b Stereo camera, camera 224c, 224d Camera 226 Upper surface, second side, top surface, surface of sheet 204 228 Display 230 Surface pattern, speckle dot surface pattern, speckle pattern 232, 234 Micro micrometer level adjuster 236a, 236b Dowel, alignment pin 240a, 240b, 240c Belt roller, belt 242a, 242b, 242c Drive shaft 244a, 244b, 244c Arched tensioner 246a, 246b, 246c Bearing roller 248a, 248b, 248c Roller ball 250a, 250b, 250c Socket 410a, 410b Thickness sensor

Claims (15)

1. An apparatus for testing edge strength of sheets of material, the apparatus comprising:
a plurality of assemblies configured to selectively apply a three-point bend load along an edge of the sheet of material in a test area of the apparatus, the plurality of assemblies being capable of establishing a loaded condition on the sheet of material in the test area by applying the three-point bend load and of establishing an unloaded condition on the sheet of material in the test area by not applying the three-point bend load;
a detection mechanism configured to optically measure a strain in the sheet of material in the test area when the sheet of material is in the unloaded state and when the sheet of material is in the loaded state, the strain in the loaded state being produced by the three-point bending load; and a processor configured to determine a stress in the sheet of material based on the measured strain ,
The device, wherein the material is glass . the plurality of assemblies comprises two opposing assemblies, a first of the two opposing assemblies comprising a single arched member for engaging a first side of the sheet, and a second of the two opposing assemblies comprising two spaced apart arched members for engaging a second side of the sheet opposite the first side;
The apparatus of claim 1 , wherein the two spaced apart arcuate members define the test area between the two spaced apart arcuate members. The device of claim 2, wherein the single arched member is vertically aligned between the two spaced apart arched members. the detection mechanism comprises a first pair of cameras positioned to capture one or more images of a surface of the second side of the sheet within the test area between the two spaced apart arcuate members;
3. The apparatus of claim 2, wherein the detection mechanism optically measures the distortion in the sheet of material from the image of the surface of the second side of the sheet acquired with the first pair of cameras. The device of claim 2, wherein the arched member is a non-rotating bush, a cylindrical roller, a belt roller, or a bearing roller. The apparatus of claim 4 , wherein a pattern provided on the surface of the sheet in the test area facilitates optical measurement of the strain in the sheet of material. The apparatus of claim 4, wherein the detection mechanism further comprises a second pair of cameras positioned to capture images of the surface of the second side of the sheet in an area other than the test area. the detection mechanism comprises at least one camera positioned to capture one or more images of a vertical edge of the sheet within the test area between the two spaced apart arcuate members;
The apparatus of claim 2 , wherein the detection mechanism optically measures the distortion in the sheet of material from the image of the vertical edge of the sheet acquired with the at least one camera. The apparatus of claim 8, wherein the detection mechanism further comprises at least one additional camera positioned to capture images of the vertical edge surface of the sheet in an area other than the test area. 1. A method for testing edge strength of a sheet of material along an edge of said sheet, said method comprising:
applying a surface pattern of visual markers onto a surface of the sheet in an area of interest along the edge of the sheet;
acquiring a first optical image of the surface of the sheet in the region of interest without applying any three-point bending load to the region of interest;
applying a three-point bending load to the area of interest along the edge of the sheet;
acquiring a second optical image of the surface of the sheet at the area of interest while applying the three-point bending load to the area of interest; and determining a stress on the surface of the area of interest along the edge of the sheet due to application of the three-point bending load based on the first optical image and the second optical image .
The method according to claim 1, wherein the material is glass . The step of determining stress on the surface of the region of interest along the edge of the sheet comprises:
measuring a strain of the sheet of material in the region of interest in an unloaded state based on the first optical image;
11. The method of claim 10, comprising: measuring a strain in the sheet of material in the region of interest in a loaded state where the three-point bending load is applied based on the second optical image; and determining the strain in the sheet induced by the applied three-point bending load by comparing the measured strain in the unloaded state to the measured strain in the loaded state. The step of measuring the strain of the sheet of material in the region of interest under the applied load based on the second optical image includes:
The method of claim 11 , comprising determining a displacement of the visual marker in the second optical image compared to a position of the visual marker in the first optical image. The step of determining the stress on the surface of the region of interest along the edge of the sheet comprises:
The method of claim 12 , comprising calculating the stress required to produce the measured strain of the sheet of material in the region of interest. The method of claim 10, wherein the step of applying the surface pattern of the visual markers comprises printing, coating, spraying, etching, attaching, or projecting an image onto the surface of the sheet. The method of claim 10 , wherein the first optical image and the second optical image are acquired using an optical system that includes at least one camera.

Images