Method for further processing thin glass and thin glass produced by such method

Abstract

A method for processing a thin glass is provided. The thin glass is subjected to a tensile stress .sub.app smaller than $1.15.Math.\mathrm{Min}\left({\stackrel{\_}{}}_a-{}_{a}0.4.Math.\left(1-\ln \left(\frac{A_{\mathrm{ref}}}{A_{\mathrm{App}}}\right)\right),{\stackrel{\_}{}}_e-{}_{e}0.4.Math.\left(1-\ln \left(\frac{L_{\mathrm{ref}}}{L_{\mathrm{App}}}\right)\right)\right),$
wherein .sub.a is a mean value of tensile stress at break for fractures in a surface of samples of the thin glass under bending stress, wherein .sub.a is a mean value of tensile stress at break for fractures emanating from an edge of the samples, wherein L.sub.ref is an edge length and A.sub.ref is a surface area of the samples, wherein .sub.e and .sub.a denote standard deviations of the mean values .sub.e and .sub.a, respectively, and wherein A.sub.app is a surface area of the thin glass, L.sub.app is a summated edge length of opposite edges of the thin glass, and is a predefined maximum fracture rate within a period of time of at least half a year.

Claims

1. A method for further processing a thin glass, comprising: subjecting the thin glass to a tensile stress .sub.app smaller than $1.15.Math.\mathrm{Min}\left({\stackrel{\_}{}}_a-{}_{a}0.4.Math.\left(1-\ln \left(\frac{A_{\mathrm{ref}}}{A_{\mathrm{App}}}\right)\right),{\stackrel{\_}{}}_e-{}_{e}0.4.Math.\left(1-\ln \left(\frac{L_{\mathrm{ref}}}{L_{\mathrm{App}}}\right)\right)\right),$ wherein .sub.a is a mean value of a tensile stress at break for fractures in a surface of samples of the thin glass under bending stress, wherein .sub.e is a mean value of a tensile stress at break for fractures emanating from an edge of the samples, wherein L.sub.ref is an edge length of the samples and A.sub.ref is a surface area of the samples, wherein .sub.e and .sub.a denote standard deviations of the mean values .sub.e and .sub.a, respectively, and wherein A.sub.app is a surface area of the thin glass, L.sub.app is a summated edge length of opposite edges of the thin glass, and is a predefined maximum fracture rate within a period of time of at least half a year.

2. The method as claimed in claim 1, wherein the predefined maximum fracture rate is 0.1 or less.

3. The method as claimed in claim 1, wherein with the tensile stress .sub.app smaller than $0.93.Math.\mathrm{Min}\left({\stackrel{\_}{}}_a-{}_{a}0.4.Math.\left(1-\ln \left(\frac{A_{\mathrm{ref}}}{A_{\mathrm{app}}}\right)\right),{\stackrel{\_}{}}_e-{}_{e}0.4.Math.\left(1-\ln \left(\frac{L_{\mathrm{ref}}}{L_{\mathrm{app}}}\right)\right)\right).$

4. The method as claimed in claim 1, further comprising bending the thin glass to a minimum bending radius R, wherein the minimum bending radius R is related to the tensile stress .sub.app as follows: ${}_{\mathrm{app}}=\frac{E}{1-v^2}\frac{t}{2R},$ wherein E is Young's modulus of the thin glass, t is the thickness of the thin glass, and is Poisson's ratio of the thin glass.

5. The method as claimed in claim 4, further comprising winding the thin glass into a roll, the thin glass comprising a glass ribbon.

6. The method as claimed in claim 5, wherein the minimum bending radius R is on an inner surface of the roll.

7. The method as claimed in claim 6, wherein the step of winding comprises winding a length of at least 100 meters into the roll.

8. The method as claimed in claim 1, wherein the thin glass has a thickness of less than 500 m.

9. The method as claimed in claim 1, wherein the thin glass has a thickness of not more than 350 m.

10. The method as claimed in claim 1, wherein the tensile stress is at least 21 MPa.

11. The method as claimed in claim 1, further comprising determining a maximum tensile stress of the thin glass from the mean values .sub.a, .sub.e and standard deviations .sub.e and .sub.a, and wherein the step of subjecting the thin glass to the tensile stress .sub.app does not exceed the maximum tensile stress.

12. The method as claimed in claim 1, further comprising determining the mean value .sub.a of the tensile stress at break for fractures and the mean value .sub.e of the tensile stress at break for fractures emanating from the edge by subjecting at least twenty samples of the thin glass to a tensile stress until break.

13. The method as claimed in claim 12, wherein the mean value .sub.a is determined by a breaking test in which a thin glass sample is fixed annularly and is loaded until break using a stamp that has a curved surface.

14. The method as claimed in claim 12, wherein the mean value .sub.e is determined by a bending test in which a thin glass sample is bent uniaxially until break.

15. The method as claimed in claim 1, further comprising storing the thin glass while subjected to the tensile stress .sub.app for a period of at least half a year.

16. A thin glass article comprising: thin glass subjected to a tensile stress .sub.app smaller than $1.15.Math.\mathrm{Min}\left({\stackrel{\_}{}}_a-{}_{a}0.4.Math.\left(1-\ln \left(\frac{A_{\mathrm{ref}}}{A_{\mathrm{App}}}\right)\right),{\stackrel{\_}{}}_e-{}_{e}0.4.Math.\left(1-\ln \left(\frac{L_{\mathrm{ref}}}{L_{\mathrm{App}}}\right)\right)\right),$ wherein .sub.a is a mean value of a tensile stress at break for fractures in a surface of samples of the thin glass under bending stress, wherein .sub.e is a mean value of a tensile stress at break for fractures emanating from an edge of the samples, wherein L.sub.ref is an edge length of the samples and A.sub.ref is a surface area of the samples, wherein .sub.e and .sub.a denote standard deviations of the mean values .sub.e and .sub.a, respectively, and wherein A.sub.app is a surface area of the thin glass, L.sub.app is a summated edge length of opposite edges of the thin glass, and is a maximum fracture rate of not more than 0.1 within a period of time of at least half a year.

17. The thin glass article as claimed in claim 16, wherein the thin glass is wound into a roll having a radius R on an inner surface of the roll that is related to the tensile stress .sub.app as follows: ${}_{\mathrm{app}}=\frac{E}{1-v^2}\frac{t}{2R},$ wherein E is Young's modulus, t is a thickness of the thin glass, and is Poisson's ratio of the thin glass.

18. The thin glass article as claimed in claim 16, wherein the tensile stress .sub.app smaller than $0.93.Math.\mathrm{Min}\left({\stackrel{\_}{}}_a-{}_{a}0.4.Math.\left(1-\ln \left(\frac{A_{\mathrm{ref}}}{A_{\mathrm{app}}}\right)\right),{\stackrel{\_}{}}_e-{}_{e}0.4.Math.\left(1-\ln \left(\frac{L_{\mathrm{ref}}}{L_{\mathrm{app}}}\right)\right)\right).$

19. The thin glass article as claimed in claim 16, wherein the thin glass is subject to a maximum tensile stress of 21 MPa.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a glass article in form of a coiled thin glass ribbon;

(2) FIG. 2 is a graph of fracture probabilities as a function of tensile stress;

(3) FIG. 3 illustrates the fracture probability of thin glasses as a function of time;

(4) FIG. 4 illustrates a setup for determining the mean value of tensile stress at break for fractures in the surface area of thin glass samples and the standard deviation thereof; and

(5) FIG. 5 illustrates a setup for determining the mean value of tensile stress at break for fractures emanating from the edge of a thin glass and the standard deviation thereof.

DETAILED DESCRIPTION

(6) FIG. 1 shows a preferred application of the invention. In this embodiment of the invention, a thin glass 1 is further processed by winding the thin glass 1 in form of a thin glass ribbon 2 into a roll 3. The two edges 22, 23, or more specifically the longitudinal edges of the thin glass ribbon 2 form the end faces of the roll 3. Optionally, the roll 3 may be wound around a mandrel, so that the inner surface of the roll 3 engages the outer surface of the mandrel.

(7) The thin glass 1 or the thin glass ribbon 2 in this form may subsequently be unwound from the roll 3 for further processing steps. This form of confectioning of thin glass 1 is particularly well suited for automated manufacturing processes, such as lamination onto electronic components or the manufacturing of displays.

(8) In order to protect the glass surfaces, another web material 7 may be wound together with the glass, as in the example shown in FIG. 1. This web material then separates the successive glass layers in the roll. Preferably, paper or a plastic material is used as the web material 7.

(9) If the manufacturing process is automated, it is important that the entire coiled thin glass ribbon 2 has no fracture and that the thin glass ribbon 2 is not severed upon automated unwinding. However, the thin glass 1 is bent when being wound. The bending involves a tensile stress to which one of the sides of the thin glass 1 is subjected. The smaller the bending radius, the greater is the tensile stress. The coiled thin glass ribbon 2 will have its smallest bending radius on the inner surface 31 of the roll 3.

(10) Now, some time may pass between the processing of the thin glass 1 by winding it into a roll 3 and the unwinding thereof in a further manufacturing process. Typically, the roll 3 will be stored for some time after having been completed. Also, transportation thereof takes time. It has been found that adverse fractures of the glass may even occur with a time delay after winding, as a result of the unilateral tensile stress generated during bending and despite of the small glass thickness.

(11) The invention now makes it possible to wind thin glass ribbons into rolls 3 which in terms of their inner radius are dimensioned such that with a high probability they will withstand a predetermined time period such as an average or maximum storage period without being damaged. This is generally true as well for other types of processing of the thin glass in which the thin glass is under tensile stress. Similar to the embodiment of the roll, the most frequent tensile stresses occurring in applications or in a processed glass article are caused by bending of the thin glass. According to a preferred embodiment of the invention, therefore, the further processing of the thin glass 1 comprises bending of the thin glass 1, wherein the minimum bending radius R is related to the tensile stress .sub.app as follows:

(12) $\begin{array}{cc}{}_{\mathrm{app}}=\frac{E}{1-v^2}\frac{t}{2R}.& \left(3\right)\end{array}$

(13) In this expression, E represents the Young's modulus, t is the thickness of the thin glass, and is the Poisson's ratio of the glass.

(14) The thickness t is preferably less than 500 m, more preferably not more than 350 m. Furthermore, it is generally preferred that the glass has at least a thickness of 5 m.

(15) For the bending radius which satisfies the condition of a maximum tensile stress .sub.app calculated according to expression (1), the following relationship between the bending radius and the tensile stress is obtained by a combination with equation (3):

(16) $\begin{array}{cc}R\frac{\frac{E}{1-v^2}.Math.t}{2.3.Math.\mathrm{Min}\left({\stackrel{\_}{}}_a-{}_{a}0.4.Math.\left(1-\ln \left(\frac{A_{\mathrm{ref}}}{A_{\mathrm{app}}}\right)\right),{\stackrel{\_}{}}_e-{}_{e}0.4\left(1-\ln \left(\frac{L_{\mathrm{ref}}}{L_{\mathrm{app}}}\right)\right)\right).}& \left(4\right)\end{array}$

(17) Similarly, from a combination of equation (3) with expression (2), the following relationship is obtained for the bending radius which achieves a low fracture probability during elongated periods of time:

(18) $\begin{array}{cc}R\frac{\frac{E}{1-v^2}.Math.t}{1.86.Math.\mathrm{Min}\left({\stackrel{\_}{}}_a-{}_{a}0.4.Math.\left(1-\ln \left(\frac{A_{\mathrm{ref}}}{A_{\mathrm{app}}}\right)\right),{\stackrel{\_}{}}_e-{}_{e}0.4\left(1-\ln \left(\frac{L_{\mathrm{ref}}}{L_{\mathrm{app}}}\right)\right)\right).}& \left(5\right)\end{array}$

(19) To name an exemplary embodiment of a type of glass suitable for a thin glass, an alkali-free borosilicate glass may be mentioned which comprises the following components, in weight percent:

(20) SiO.sub.2 61

(21) Al.sub.2O.sub.3 18

(22) B.sub.2O.sub.3 10

(23) CaO 5

(24) BaO 3

(25) MgO 3.

(26) This glass has a coefficient of thermal expansion of 3.2*10.sup.6 1/K, which is matched to silicon. The modulus of elasticity or Young's modulus amounts to E=74.8 GPa. Poisson's ratio is =0.238.

(27) In case of a glass article in form of a thin glass ribbon 2 coiled into a roll 3 as shown in FIG. 1, the minimum bending radius R of the thin glass ribbon 2 from which the maximum tensile stress .sub.app according to equation (3) results, will be located on the inner surface 31 of the roll 3. However, in order to be able to keep the roll small and easy to handle, bending radii are preferred for which the maximum tensile stress that occurs on the inner surface 31 is at least 21 MPa.

(28) However, application related cases in which the thin glass 1 is subjected to tensile forces along the sides or along the surfaces thereof are also conceivable. In this case, tensile stresses will arise on both sides and in the volume of the thin glass.

(29) Regardless of the form in which the tensile stresses occur after the further processing, one embodiment of the invention is aimed at a maximum fracture rate of 0.1 or less, preferably less than 0.05. With the predefined fracture rate , the maximum tensile stress .sub.app corresponding to this fracture rate may then be determined according to expressions (1) or (2) given above, and then, in case of a tensile stress caused by bending, the minimum bending radius may be determined based on expression (3) using this tensile stress value.

(30) Expressions (1) or (2) which indicate the maximum tensile stress corresponding to a predefined fracture probability within at least half a year (expression (1)) or more (expression (2)), further include the surface area of the thin glass and the edge length thereof. That means, the fracture probability scales with the size and shape of the thin glass article. This is important, since thin glasses may have a considerable large surface area, in particular in an intermediate product such as roll 3. For example, in case of roll 3 it is preferred that a thin glass ribbon 2 is wound which has a length of at least 10 meters, preferably at least 50 m, more preferably at least 100 meters. According to yet another embodiment, a thin glass ribbon having a length of up to 1000 meters is coiled up to keep the fracture probability low and at the same time keep the inner radius of the thin glass coil 3 small. Preferably, the width of the thin glass ribbon 2 or of the corresponding roll 3 is 20 centimeters or more. According to one exemplary embodiment, a thin glass roll 3 is produced from a thin glass ribbon 2 having a length of 100 m, a width of 20 cm, and a thickness of 50 m.

(31) With reference to FIG. 2 the effect of the scaling of the fracture probability is illustrated. FIG. 2 shows a graph of fracture probabilities determined by breaking tests as a function of the tensile stress. The measured values illustrated by solid marks and the corresponding regression line designated A were determined from breaking tests on samples of a surface area of 80 mm.sup.2. The values illustrated by open marks and the corresponding regression line B are obtained by scaling the measured values to a surface area of 625 mm.sup.2 which is subjected to the same load as the samples. As can be seen from the intersections of the two regression lines with the line drawn at a tensile stress of 66 MPa, for example, due to the larger surface area the fracture probability increases by about one order of magnitude. Although the breaking strength measurements are performed with a dynamic load, in particular an increasing load, these breakage tests permit to determine the fracture probability within a predefined period of time, in particular for long periods of time of at least half a year under a static load.

(32) FIG. 3 shows the fracture probability F (corresponding to the fracture rate in expressions (1) or (2)) as a function of lifetime t.sub.lifetime in years for glass articles having a scaled surface area of 625 mm.sup.2 and under a static tensile stress of 66 MPa. Lifetimes of half a year and longer correspond to the specified periods of time, for which expressions (1) or (2) are valid. Generally, without being limited to the exemplary embodiments, it is therefore contemplated according to one embodiment of the inventive method that the method for further processing a thin glass 1 additionally comprises storing of the thin glass in the state subjected to the tensile stress .sub.app for a period of at least half a year.

(33) FIG. 4 schematically shows a setup for determining the parameters .sub.a and .sub.a, i.e. the mean value of tensile stress at break for fractures in the surface area of thin glass samples and the standard deviation thereof. Besides this setup described below, alternative measurement setups are likewise possible.

(34) The measurement is based on a procedure in which the parameters .sub.a and .sub.a are determined by a breaking test in which a thin glass sample 10 is fixed annularly and is loaded with a stamp 12 that has a curved, preferably spherical stamp surface 120, until break. For this purpose, the thin glass sample 10 is placed on an annular, preferably circular support surface 13, as shown in FIG. 4, and is fixed. Stamp 12 presses upon the surface of the thin glass sample 10, preferably centered within the annular support surface 13, with a force F. The force is increased until the thin glass sample 10 breaks. In the setup shown in FIG. 4, a fracture will typically result in the surface area of the thin glass sample, because the tensile stress induced is greatest in the area of engagement of the stamp 12 against the surface of the thin glass sample. This engagement area accordingly defines the surface A.sub.ref of the sample 10.

(35) The breaking test is repeated several times. From the force exerted at the time of breakage, the tensile stress on the surface of the thin glass sample may then be calculated. From the measured values, the average tensile stress at break .sub.a and its standard deviation .sub.a are determined. For this purpose, it is possible to convert the individual force values into tensile stresses, and then to calculate the mean value and the standard deviation.

(36) For fractures emanating from an edge of the thin glass, other measurement setups are suitable, for example the measurement setup described below and shown in FIG. 5. Using this measurement setup, parameters .sub.e and .sub.e are then determined similarly.

(37) In order to obtain sufficiently trustworthy statistics for reliable determination of the maximum tensile stress of a glass article, according to one embodiment of the invention at least ten, preferably at least twenty, more preferably at least 30, most preferably at least 50 samples 10 of the thin glass 1 are subjected to a tensile stress until break to determine parameters .sub.a and .sub.a, as well as .sub.e and .sub.e. Accordingly, with respect to the exemplary embodiments shown in FIG. 4 and FIG. 5, at least 10, preferably at least 20, more preferably at least 30, most preferably at least 50 valid breaking tests are performed using the setup illustrated in FIG. 4, and also at least 10, preferably at least 20, more preferably at least 30, most preferably at least 50 breaking tests using the setup of FIG. 5 that will be described below.

(38) For this purpose, FIG. 5 shows a setup for determining the mean value of the tensile strength for fractures emanating from the edge of a thin glass and the standard deviation thereof. The procedure performed with the setup comprises to determine the parameters .sub.e and .sub.e by a bending test, in which a thin glass sample 10 is uniaxially bent until break. In the setup shown in FIG. 5, the thin glass sample 10 is clamped between two jaws 15, 16. Jaws 15, 16 are moved toward each other, so that the thin glass sample 10 is bent more and more. In contrast to the setup shown in FIG. 4, the bending is effected in only one direction. The minimum radius of curvature R.sub.min is in the middle between the two jaws. If, for example, the jaws are arranged at a slight angle to each other, the edge at which the jaws 15, 16 are closer to each other will be stressed more than the opposite edge. Accordingly, the minimum radius of curvature will be at this edge. It is also possible to evenly load both edges 22, 23.

(39) For determining the tensile stress at break on the edges, and for determining therefrom, after having tested several thin glass samples, the mean value .sub.e and the standard deviation .sub.e, there are several possibilities. According to one embodiment, the force F exerted to the jaws 15, 16 can be measured, and the tension in the thin glass sample 10 can be determined therefrom.

(40) The tensile stress at break may be determined even more easily by determining the minimum bending radius R.sub.min at break and determining the corresponding tensile stress on the edge from this value. Here, the tensile stress is inversely proportionally related to the bending radius.

(41) An exemplary embodiment for producing a roll of coiled thin glass will now be described.

(42) It is intended to roll a glass web into a roll, and the glass web has a length of 100 m, a width of 20 cm, and a thickness of 0.05 mm and is made of a borosilicate glass of the composition given above, with a Young's modulus of E=74.8 GPa and a Poisson's ratio of =0.238. Fracture probability should not exceed 1% (=0.01) during a storage period of one year. The core radius of the winding body is chosen according to equation (3). Strength measurements on samples give the values of .sub.a=421 MPa (mean value) and .sub.a=35 MPa (standard deviation) for the surface strength on the basis of normal distributions for a reference surface area of 121 mm.sup.2, and values of .sub.e=171 MPa (mean value) and .sub.e=16.9 MPa (standard deviation) for the edge strength for a reference length L.sub.ref of 2 mm. The setups described with reference to FIG. 4 and FIG. 5 can be used for this purpose. Methods for evaluating strength measurements can further be found, for example, in: K. Nattermann, Fracture Statistics in Strength of Glass-Basics and Test Procedures, Advanced Course of the International Commission on Glass and Research Association of the German Glass Industry, Frankfurt (2006), ISBN 3-9210-8947-6).

(43) With A.sub.app=0.2 m*100 m=20 m.sup.2, and L.sub.app=2*200 m=200 m then follows

(44) $\begin{array}{cc}{\stackrel{\_}{}}_a-{}_{a}0.4\left(1-\ln \left(\frac{A_{\mathrm{ref}}}{A_{\mathrm{app}}}\right)\right)=\left(421-350.4\left(1-\ln \left(\frac{121{10}^{-6}{m}^2}{20m^2}0.01\right)\right)\right)\mathrm{MPa}=175\mathrm{MPa}& \left(6\right)\\ {\stackrel{\_}{}}_e-{}_{e}0.4\left(1-\ln \left(\frac{L_{\mathrm{ref}}}{L_{\mathrm{app}}}\right)\right)=\left(171-16.9\left(1-\ln \left(\frac{2.Math.{10}^{-3}\mathrm{mm}}{200m}.Math.0.01\right)\right)\right)\mathrm{MPa}=55\mathrm{MPa}& \left(7\right)\end{array}$

(45) The edge strength is thus the decisive strength parameter for the dimensioning of the core of the roll.

(46) For the permissible bending stress, in this case, according to expression (2) or equation (5) it follows:
0.93.Math.Min(175 MPa,55 MPa)=0.93.Math.55 MPa=51 MPa.

(47) With

(48) $\frac{E}{1-v^2}=79.3.Math.{10}^3\mathrm{MPa}$
and t=0.05 mm, the minimum bending radius of the thin glass ribbon may then be calculated according to equation (3) to give:

(49) $R\frac{1}{2}\frac{79300\mathrm{MPa}}{51\mathrm{MPa}}0.05\mathrm{mm}=39\mathrm{mm}.$

(50) Now, one may round up to a next larger standard roll core diameter, i.e. D=80 mm, for example. With this bending radius or diameter, respectively, the minimum value of the tensile stress of 21 MPa preferred according to the invention is exceeded, so that on the one hand a compact roll is obtained, which on the other hand nevertheless has a low fracture probability.

(51) In addition to the setup shown in FIG. 5 and the discussed measuring method for determining statistical parameters .sub.e and .sub.e, other test methods are also conceivable to obtain very precise statistical values. In the setup shown in FIG. 5, the bending radius varies and is minimal in the center of the edge. Thus, the edge length L.sub.ref and the surface area A.sub.ref of samples 10 is smaller than the total edge length of opposite edges and the total surface area of the sample. The values of L.sub.ref=2 mm and A.sub.ref=121 mm.sup.2 as mentioned in the exemplary embodiment above are therefore smaller than the actual dimensions of the sample. With suitable measuring methods it is possible to increase the edge length and surface area of the sample which are subjected to the tensile stress. For determining surface strength, setups other than the arrangement shown in FIG. 4 are conceivable as well.

LIST OF REFERENCE NUMERALS

(52) 1 Thin glass 2 Thin glass ribbon 3 Roll 7 Web material 10 Thin glass sample 12 Stamp 13 Annular support surface 15, 16 Jaws 31 Inner surface of 3 22, 23 Edges 120 Stamp face