WO2008000936A1

WO2008000936A1 - Installation for producing flat glass, comprising a stress measuring device, and method for operating a flat glass annealing lehr

Info

Publication number: WO2008000936A1
Application number: PCT/FR2007/001030
Authority: WO
Inventors: Wolf Stefan Kuhn
Original assignee: Fives Stein
Priority date: 2006-06-27
Filing date: 2007-06-21
Publication date: 2008-01-03
Also published as: EP2032973A1; RU2438116C2; CN101467029A; RU2009102503A; FR2902881B1; CN101467029B; FR2902881A1

Abstract

The invention relates to an installation for producing flat glass, comprising a melting and refining furnace followed by a device for forming a flat glass ribbon and a lehr (K). The inventive installation comprises a device (G) for the non-contact measurement of the stresses in the glass ribbon on the production line, said device being integrated into the lehr (K).

Description

PRODUCTION PLANT OF FLAT GLASS WITH EQUIPMENT OF MEASUREMENT STRESS, AND DRIVING METHOD ¹ A ANNEALING LEHR OF FLAT GLASS.

The present invention relates to a flat glass production plant comprising a melting and refining furnace followed by a device for forming a flat glass ribbon and an outrigger, and comprising measuring equipment in line and without contact of the stresses in the glass ribbon

A flat glass annealing lehr is a tunnel kiln equipped with heating and cooling means for following a controlled cooling thermal cycle with a glass ribbon. It is placed downstream of the tin bath for a production line according to the float process, or downstream of the melting and forming furnace for a laminated glass production line.

The first critical phase of the cooling cycle of the flat glass strip is located in the areas of the lehr where the glass is in a viscoelastic state. Cooling induces thermal gradients and stresses. To limit the creation of permanent stresses (also called residuals) and to allow their relaxation, the beginning of the cooling is carried out at a reduced rate to allow a 'annealing' of the glass. A level of permanent stress that is too high causes problems in the subsequent processing of the glass such as cutting. Once this annealing around the transition temperature is completed, the second critical phase of the cooling cycle begins, where the goal is to cool the glass quickly to limit the length of the lehr. Since the glass is now in the solid state, thermal gradients during this cooling induce so-called temporary constraints. However, after the first phase of cooling, the permanent stresses are always present in the glass. Total stress refers to the combination of permanent and temporary constraints.

During cooling, temperature gradients in the thickness of the glass ribbon result in so-called laminated stresses, shown in Fig.1 of the accompanying drawings. The cooler surface layers are in isotropic plane expansion state and the inner layers in isotropic plane compression state. The temperature gradients in the plane of a ribbon also cause plane but membrane-like stresses, also known as shape constraints. In the case of gradients only over the width of a long-length ribbon (see Fig. 1), the orientation of the membrane stresses is parallel to the length of the ribbon. These constraints in the length of the ribbon are compressive or extensive as a function of the temperature profile across the width of the ribbon. In a lehr, one tries to maintain the edges of the tape in slight compression because of their greater fragility. These thermal stresses disappear at the end of cooling with the disappearance of thermal gradients, hence their characterization as a temporary constraint.

We can easily understand the creation of permanent constraints. A ribbon of glass solidifies first on both sides. The heart is still plastic at this moment. It solidifies late and tends to compact it further. This is no longer possible because the heart is attached to both solidified faces. The heart then undergoes extensive stress during solidification. Consequently, the two already solidified faces are simultaneously under compressive stress. The ribbon thus shows a permanent constraint with a central zone under extension and two zones under compression. The permanent stresses in the areas are isotropic in the plane of the sheet. The level of permanent stress is intimately related to the cooling rate during the first cooling phase. A high cooling rate, especially for thick glass, induces high permanent stresses and causes problems of cutting the glass panels. Asymmetric cooling of the upper and lower faces offsets the stress profile in the thickness and causes deformation of the ribbon.

The same phenomenon of solidification of zones as a function of the cooling rate appears on the width of the ribbon. Controlling the cooling speed of the edges of the ribbon presents a particularly difficult problem, especially for thick ribbons. Adjusting the cooling during the first phase of cooling should provide a slight compression of the edges and a small extension of the center of the ribbon to reduce the problems of breaking the ribbon.

We recall here that in the absence of external forces, the inner forces must be in equilibrium (Cauchy principle). This is valid for a ribbon of glass in an outrigger if we ignore the transport force by the rollers and the gravity.

Glass is a fragile material. It breaks under the effect of traction and perpendicular to the direction of the normal stress of extension which would come to reign there. Glass does not respond to shear stress by ductile deformation. It is therefore necessary to identify the directions and amplitudes of the main constraints to evaluate the risk of rupture. The determination of the main stresses makes it possible to eliminate the shear components in the stress tensor. The shape of the glass ribbon already makes it possible to reduce the dimensions to the plane case in 2D (two dimensions). In addition, for a ribbon of infinite length and thermal gradients only over the width of the ribbon, the stress component over the length of the ribbon and the main stress coincide. A measure of the shear stress to find the main stress is no longer necessary.

A particularity of the glass material concerns the sensitivity of its surface to rupture under extension stress. A glass manufactured using the float process with a surface without macroscopic defects can already break from 50MPa traction on the surface. Surface defects can further lower this limit.

When cooling a glass ribbon in a lehr it is therefore important to finely control the stress profiles in the thickness and width of the ribbon. The determination of the main stresses on the surface of the ribbon would make it possible to anticipate excessive stresses and this on the entire surface of the ribbon in the lehr. A measurement of the constraints must therefore make it possible to determine the total principal stresses at any point of the ribbon in the lehr to control the risk of breakage but must also make it possible to determine the permanent stresses. From the solidification of the ribbon around the solidification temperature, the permanent and temporary stresses are always present simultaneously in a lehr. A simultaneous determination of the stresses and temperature profiles makes it possible to separate the temporary stress in the total stress to obtain the permanent stress. Another way to determine the permanent stress is to measure the stress in an area of the lehr without cooling and therefore where the temperature of the ribbon is homogeneous. This is particularly interesting for the profiles in the thickness because the measurement of the temperature profile in the thickness is less obvious than measuring the temperature profile across the ribbon width. The condition of the thermal homogeneity of the ribbon is, of course, given at the end of the lehr if the ambient temperature and the temperature of the ribbon are close. The zone of interest for the stress measurements on the flat-plate drying equipment thus extends between the temperature of transitions with the beginning of the establishment of the stresses until the position of the thermal homogenization of the ribbon after the lehr.

In general, an unsuitable cooling of the ribbon in the lehr may lead to an excessive level of glass stress detrimental to the quality of the glass and its completion:

1. In the lehr

Beyond a critical value of traction, the total stress with its main component in the glass causes the breaking of the ribbon.

In addition, significant stresses in the glass can induce deformation of the ribbon perpendicular to the plane of the ribbon. This can seriously disrupt the transport of the ribbon in the lehr This type of deformation is also troublesome if it is permanent and is found in the glass panels after cutting.

2. In the cutting section, downstream of the leech

Important permanent stresses in the ribbon make it difficult, if not impossible, a clean cut of the glass. Controlling and maintaining an acceptable level of stress throughout the annealing process and cooling the glass ribbon is therefore a major concern for the manufacturer.

To simplify the following discussion, we adopt a constraint coordinate system in accordance with the geometric shape of the ribbon (see Fig 1). The directions of the main stresses can for particular conditions coincide with these geometrical constraints of the ribbon.

Thus, a component σ _x is defined in the direction of the length of the strip, a component σ _y in the direction of the width of the strip and a component σ _z in a direction perpendicular to the strip. The components σ _x and σ _y present the plane stresses, stratified stress fields for a thin sheet. On the production lines, the lateral profile over the width (Fig. 2) is measured after the lehr, which traces the component σ _x integrated on the thickness of the ribbon, thus a constraint membrane. The lateral profile is not measured in the lehr according to the state of the art. It is however this constraint that can reach high values for a bad adjustment of the cooling on the width and to cause the breaking of the ribbon. The membrane component σ _y integrated on the thickness of the tape is also not measured in the state-of-the-art closures. This constraint is nevertheless not negligible in certain cases (example: the thermal profile over the width is not constant over the length of the ribbon)

The vertical profile traces the evolution of the plane stress σ _x or σ _y on the thickness of the ribbon (FIG. 2). For a small cut sample of the ribbon, the components σ _x and σ _y are equivalent because the contribution of the constraints of shape or membranes is eliminated. Due to the off-line measurement of the vertical profile, only a permanent constraint is measured. The vertical profile of the total stresses σ _x or σ _y is not measured within the state-of-the-art layouts. It is however this constraint which can reach the high and critical values in particular for a thick glass.

The component σ _z is absent in the case of a thin sheet. But near the edges of a float glass ribbon or in contact with the transport rollers, this component appears. No measure of this constraint is made on the ribbon according to the state of the art.

According to the state of the art, there are different non-contact measurement methods of ribbon constraints, in particular those described below.

An optical method often used for the measurement of the stresses of the glass resides in the analysis of the polarization of a beam of light after its passage through glass. It is based on a property related to the photoelasticity of glass that is characterized by a directional change in the refractive index of light in the presence of a constraint.

On a flat glass ribbon, the measurement is made with a beam of light that passes through the ribbon mainly perpendicular to its surface. Beam polarization is analyzed after a single or double pass through the ribbon thickness. This method requires a precise adjustment of the light beam with optical elements above and below the ribbon. Constraints performed online

(on-line) according to this method are carried out at a significant distance downstream of the lehr to reduce the influence of the temporary constraint on the measurement. This measure is therefore aimed at permanent membrane stress (see US Pat. No. 4,619,681 or DE 1,202,028). They allow the characterization of an average value for σ _x by integrating the constraints on the thickness of the sheet. Thus, these measurements characterize only the membrane stress in the direction length of the ribbon. In particular, they make it possible to check whether the ribbon edges are under compression because they are particularly sensitive to rupture under extension stress. This method also supposes the absence of a component σ _y to identify the component σ _x as the main constraint (the measure of the delay of a perpendicular beam only makes it possible to measure a difference of two components).

The inertia related to the position of the equipment on the line does not allow a quick adjustment of the operating parameters of the lehr during the first phase of ribbon cooling. In addition, the method gives no information as to the temporary constraints that prevail in the lehr. An implementation of this type of instrument in the drying room faces two difficulties: 1. The mechanical, optical and electronic components are not adapted to a hot environment with temperatures up to 600 ⁰ C. 2. The radiation Thermal glass ribbon severely disrupts the detection of polarized light by the optical system. Other methods based on the analysis of the polarization of light after passing through glass are used for the laboratory measurement of the vertical stress profile in the thickness of a sample.

Due to the cutting of the sample, the membrane stresses are more or less relaxed. We thus measure the vertical profile of the permanent plane stresses σ _x or σ _y with the equivalence of the two values because of the isotropy. The integration on a vertical profile of such a sample gives practically a zero stress due to the absence of membrane stress. The vertical profile reflects the level and performance of the glass annealing. This measurement on a cut sample only makes it possible to determine the permanent stress profile at the position on the ribbon where the sample was taken.

Another method of measuring the vertical profile of permanent stress lies in the analysis of the light scattered by a ray of light or a laser beam. This method exists in two variants: A) Analysis of the scattered and polarized light which then passes through volumes under constraints

B) Analysis of intensity of light scattered by a polarized beam crossing volumes under constraints. Both variants preferably use the effect of elastic scattering of light, called "Rayleigh scattering" or, alternatively, the effect of the interaction of light with phonons, called "Brillouin scattering" or

Raman ". These methods will be discussed in more detail later in this document.

The diffused light method always requires the coupling of a beam in the glass either on the edges of a sample or on the surface using a prism placed on the surface of the glass according to the state of the art . In US 2003/0076487 a coupling of the beam by means of a refraction grating enables the light to be efficiently guided in the glass through the prism-free surface. The refraction grating is created by a local heating of the glass using a laser. However, this method is not applicable on a hot ribbon scrolling. Local heating is also detrimental to the measurement of low thermal stresses as present in the annealed glass. The method mainly targets tempered glass with a much higher level of stress.

Another variant of the diffused light method is proposed in DE 10161914 C1. A disadvantage of the scattered light method is the loss of intensity of the beam which passes through an angle grazing glass especially for a tinted glass. The cited patent provides compensation for this loss of intensity by a gray key. However, the prism for a good coupling of the light across the surface of the sample is still necessary which limits the application of this method to offline (off-line).

All sample measurement methods generate significant inertia related to the time required for sample collection and feedback from the laboratory. In addition, they do not provide any information on the temporary constraints like membranes or plane stresses with variation in the thickness. Temporary stress measurements automatically involve a measurement within the lehr. An indirect approach is proposed in US Patent 6,796,144.

According to this patent a measurement of temperature in the thickness of a glass plate is carried out by analysis of the light emitted by photoluminescence of an area in the thickness. With this method, the vertical temperature profile is determined by several measurements. From this temperature profile, the vertical profile of temporary stress is calculated. This method gives no information on the permanent stress and the membrane stresses superimposed on the vertical profile of temporary stress during the cooling of the glass in a lehr. The solutions implemented according to the state of the art for the determination of the stress level in the glass ribbon are unsatisfactory because, in particular:

They do not make it possible to directly control the annealing process for continuously obtaining glass having good characteristics after passage in the lehr.

They do not make it possible to measure the vertical profile of plane stress in the lehr.

It does not make it possible to measure the membrane stress profile in the lehr.

They do not make it possible to measure at the same time the membrane stress profiles on the ribbon width and the vertical profile of the plane stresses.

They do not make it possible to determine the total stress and its principal directions at any point in the lehr to anticipate conditions of breakage of the ribbon or the curvature of the ribbon.

The main difficulties of measuring stress on a moving glass ribbon reside in the requirement of non-destructive measurement, without mechanical contact, on the glass at high temperature, in a warm environment and without disturbing the cooling of the ribbon. the set of main components of constraints.

To provide a solution to these problems, a flat glass production installation, according to the invention, comprising a melting and refining furnace followed by a device for forming a flat glass ribbon and a stringer, and comprising in-line measuring equipment and non-contacting constraints in the glass ribbon, is characterized in that the on-line stress measuring equipment is implanted in the lehr.

The on-line measurement, directly in the lehr and without contact, of the total stress in the glass makes it possible to quickly adjust the operating parameters of the lehr so that the total stress level remains in all point below a specified value.

Online measurement directly in the drying room and without contact also allows to determine the zone of solidification of the ribbon with the establishment of the permanent stresses The measurement in line directly in the lehr and without contact in combination with a measurement of vertical profile or Lateral temperature of the ribbon makes it possible to deduce the contributions of the temporary constraint to the total stress and to deduce from it the permanent constraint. The measuring equipment according to the invention comprises a light emitter which directs a light beam on the glass ribbon, and a light receiving and analyzing means scattered in different directions of the space resulting from the interaction of the light. beam with glass. The measuring equipment includes optical components and signal processing means for measuring and analyzing light scattering with sufficient sensitivity to eliminate optical accessories, including prisms placed on the glass surface.

Advantageously, the equipment includes a CCD camera for measuring and analyzing the scattered light.

It is thus possible to measure the stresses in the flat glass by characterizing the scattered light with sufficient precision for the analysis of the temporary and permanent stresses typical for the glass ribbon in a stringer. Another advantage of this equipment lies in the possibility of simultaneously measuring the profile of the membrane stresses and the profile of the stresses in the thickness of the glass.

The complete information on the temporary and permanent stress components allows to adjust the cooling of the lehr to avoid breakage and better control the level of permanent stress. It is thus possible to exploit this information to control the lehr with an automatic system including measurement, signal processing and motorized cooling adjustment.

According to a preferred embodiment of the invention, the measuring equipment makes it possible to combine the characterization of the vertical and lateral stress profile by measuring the stress component σ _x . These two profiles correspond to the measurements currently performed 'cold' after the lehr or samples in the laboratory. The measurement according to the invention also makes it possible to evaluate the temporary and total stress during cooling. The measuring equipment of the vertical and lateral profile is advantageously placed in each cooling zone of the lehr to individually control the stress generated by each zone. The measurement of the stress makes it possible in particular to control the rate of cooling over the width of each zone, the upper and lower cooling and the total rate of cooling.

According to another embodiment, the measuring equipment makes it possible to combine a measurement of the vertical and longitudinal profiles. this allows to identify in particular the evolution of the permanent and temporary stress in the thickness over the length of the lehr. If this measurement is repeated at several positions on the width of the ribbon, it is also possible to establish the lateral membrane stress profiles.

According to another exemplary embodiment, these profiles are measured at many positions of the ribbon which makes it possible to establish a stress mapping along the entire length of the ribbon in the lehr. This makes it possible to identify the places in the glass ribbon with a high total stress.

In the presence of a component σ _y of the membrane stress, the principal stress is no longer oriented parallel to the stress σ _x . There is therefore simultaneous presence of both components. The principle of photoelasticity makes it possible to measure only differences in the stresses perpendicular to the observation beam. However, we manage to separate the contributions of the different constraints by a repetition of the measurement in different orientations. To find the direction and amplitudes of the main stresses, the measuring system can rotate. Another way to go back to the main constraints is to analyze the optical signal at different angles to the normal on the glass plate.

Another embodiment aims to check the level of stress at particular points such as the ribbon support on the rollers and the change in thickness near the edge of the ribbon. However, the principle of photoelasticity makes it possible to measure only differences in the stresses perpendicular to the observation beam. By comparing the measurements at close positions but without and with the component σ _z, its value can be traced. According to another exemplary embodiment, the orientation of the observation beam is varied to find the principal 3D constraints which contain the contribution of the vertical component σ ₂ .

The measuring equipment may comprise a light source on one side of the ribbon and an optics for analysis on the opposite side, or preferably on the same side as the light source. The light source may be above or below the ribbon. The emitter and receiver of the measuring equipment can be cooled for placement in a relatively high temperature area of the lehr. The measuring equipment is completed by a control station, data processing provided by the measuring equipment, and visualization of constraints. An optional interface allows the link between this control station and the control unit of the lehr. An installation equipped with measuring equipment with an optical system advantageously comprises means for filtering the incoming rays in the optical system of the measuring equipment in order to eliminate the thermal radiation detrimental to the accuracy of the measurement.

The optical system may include a sighting tube and an optical chopper provided after the sighting tube.

The invention also relates to equipment for measuring in-line and non-contacting stresses in a glass ribbon in a lehr, characterized in that it comprises a light emitter which directs a light beam on the glass ribbon, and receiving means and analyzing scattered light in different directions of space resulting from the interaction of the beam with the glass.

Preferably, the on-line measurement equipment uses the effect of elastic scattering of light, called "Rayleigh scattering" or the effect of the interaction of light with phonons, called "Brillouin or Raman scattering". The equipment advantageously comprises a CCD camera for the measurement and analysis of the scattered light.

In a variant of the equipment, only the membrane stress is measured by passage and direct analysis of a polarized beam according to the conventional method of photoelasticity. The signal analyzer advantageously comprises means for filtering the incoming rays in the optical system of the measuring equipment in order to eliminate the thermal radiation detrimental to the accuracy of the measurement.

In a variant of the equipment, a measurement is made of the lateral and / or longitudinal surface temperature profiles of the ribbon by measuring means known as pyrometers. Other methods based on the volume emission of radiation make it possible to go back to the temperature profile in the thickness of the ribbon. It is thus possible to separate the temporary and permanent stress in the measure of the total stress.

The invention also consists in a method for driving a flat glass annealing lehr, characterized in that a continuous measurement of the stress of a glass ribbon is carried out by measuring equipment installed in the lehr , and is used to automatically adjust operating parameters of the outrigger via a control loop.

Advantageously, according to the annealing lehr method of the invention, a combination of the leech control system and the stress measuring equipment is provided to enable rapid adjustment of the operating parameters of the invention. the lehr so that the total stress level remains below a determined value that makes it possible to avoid breakage of the glass or deformations of the ribbon perpendicular to the plane of the ribbon and that the level of permanent stress remains below a determined value allowing the subsequent treatment of the glass.

On the other hand, according to the method of the invention, the stress measurements can be made according to the width of the glass ribbon, and can be used for adjusting the distribution of the heating over the width of the ribbon and / or the adjustment the distribution of cooling over the width of the ribbon.

Preferably, according to the method of the invention, a mathematical model of operation of the lehr is established and used to define the optimum instructions to be applied to the lehr, according to the measurements made, in order to obtain the temperature level and desired stress.

The invention consists, apart from the arrangements described above, in a certain number of other arrangements which will be more explicitly discussed hereinafter with regard to exemplary embodiments described in detail with reference to the appended drawings, but which are in no way limiting. On these drawings:

Fig. 1 schematically illustrates a glass ribbon and the directions of the profiles and constraints. Fig. 2 is a diagram illustrating a possible variation of the stresses along the y direction of the width of the glass ribbon.

Fig. 3 is a diagram illustrating a possible variation of the constraints in the vertical direction z in the glass ribbon.

Fig. 4 is a schematic side view of a flat glass production plant.

Fig. 5 is a partial schematic vertical section, on a larger scale, of an outrigger according to the invention comprising an example of implementation of the stress measurement system. Fig. 6 is a schematic view from above with respect to FIG.

Fig. 7 is a diagram of measuring means and means of control of the lehr.

Fig.8 is a perspective diagram of a measuring equipment positioned above the glass ribbon, and

Fig.9 is a schematic elevational view of a receiver means and scattered light analysis.

Referring to Fig.1 of the drawings, schematically shown a ribbon of glass 1, located in a horizontal plane, which progresses in the direction of the arrow S, parallel to the longitudinal edges of the ribbon. The vertical line V in dashes indicates the direction of the thickness of the ribbon. The horizontal line L in dashes indicates the lateral direction y of the width of the ribbon, orthogonal to the arrow S. The orientation of the stresses in the glass ribbon can be defined in three orthogonal directions, namely a component σ _x in the direction x the length of the ribbon, a component σ _y in the y direction of the width and a component σ _z in the z direction of the thickness. In particular cases, the orientation of a principal stress may coincide with one of the geometric orientations. FIG. 2 shows the distribution of the stress, plotted on the ordinate, positive for a tension (tension) and negative for a compression, according to the position of a point in the y direction of the width L plotted on the abscissa. According to the example of Fig.2 the longitudinal edges of the ribbon are in compression while the intermediate zone is in tension. Fig. 3 shows a possible distribution of the temporary stress in the thickness of the ribbon. Constraints are plotted on the abscissa, with positive values for stress and negative for compressive stress. The position of a given point of the glass ribbon along the thickness is plotted on the ordinate axis. From Fig. 3, it appears that the upper face and the lower face of the ribbon are in tension while the zone located mid-thickness is in compression.

. This vertical profile makes it possible in particular to quantify the tensile or compressive stress at each point of the thickness and in particular at the two surfaces of the ribbon. FIG. 4 schematically illustrates a flat glass production plant comprising a station P for preparing and charging the raw materials, a furnace H for melting and refining, a forming device J for glass sheet, a K-stretcher and an outlet section M including the cutting and conditioning of the glass.

The K-channel is composed of different successive zones traditionally defined as follows: Zone AO: Optional entry zone for a particular treatment, Zone A: Pre-conditioning zone, Zone B: Annealing zone, Zone C: Indirect cooling zone, Zone D: Tempered direct cooling zone, Zones E and F: Final direct cooling zones, and last zones of the lehr.

In zones AO, A ₁ B and C, the control of the cooling of the glass is obtained by radiative exchanges with cold parts, commonly called heat exchangers, or heating elements, whereas in zones D, E and F the cooling is carried out by convection blown air.

According to the invention, a non-contact measuring device G of the stresses in the glass ribbon is implanted in the length of the stringer, in characteristic zones of the annealing process, for example towards the end of the slow cooling zones A, B and C or towards the end of the rapid cooling zones E and F. It is also possible to place several measuring devices at different characteristic points along the length of the lehr at the zones A, B, C, D, E and F.

An exemplary embodiment is shown in FIG. 5, with a glass ribbon 1 running on transport rollers 2 inside the chamber 3 of the Ranch K equipped with a cooling system of the glass 4 by radiation or convection. The strain measurement system 5 comprises a cooled housing 6 equipped with openings for the optical system 7, the latter being able to be equipped with a thermal protection device such as an air sweep or the closure of optical windows. The entire equipment G of stress measurement is supported by a mechanical support 8 mounted on a scroll device 9 and a manual or automatic transverse displacement system 11 to cover the entire width of the glass ribbon. The system makes it possible to characterize the vertical stress profile in the thickness of the glass and the membrane stress profile at any point along the width of the strip.

An optional rotation device 10 makes it possible to orient the optical system 7 in the direction corresponding to the width of the ribbon or that corresponding to the length of the ribbon. Rotation identifies direction and amplitude of the principal stresses from the measurements of the components σ _x and σ _y and intermediate of the stress in the glass.

The measuring equipment G comprises a light emitter which directs a light beam on the glass ribbon, and a light receiving and analyzing means scattered in different directions of the space resulting from the interaction of the beam with the light. glass. A more complete description of a measuring equipment is given later with reference to FIGS. 8 and 9.

The measuring equipment G preferably uses the effect of the elastic scattering of light, called "Rayleigh scattering" or, alternatively, the effect of the interaction of light with phonons, called "Brillouin or Raman scattering". ".

The optical components and the signal processing methods are provided for the implementation of the measurement system of the light diffused in the lehr. Advantageously, the measuring equipment comprises a combination of optical elements in a single support of limited size and on one side of the ribbon, which allows a precise and reliable adjustment of the optical components. The measurement and analysis of the scattered light can be performed with a CCD camera. Thus any variation of the relative position of the measuring system with respect to the ribbon can be sensed by the system and considered / exploited by the computer processing of the signal.

Alternatively, all of the stress measuring equipment is placed in a support mounted on a scrolling device and a manual or automatic longitudinal displacement system for covering a part or the entire length of the lehr so to raise the stress levels on the length of the ribbon.

The displacement system may consist of a horizontal servo-axis equipped with a carriage on which the measuring equipment is mounted.

Other embodiments of the present invention are possible, for example: 1. A stress measuring system placed on the underside of the glass ribbon.

2. A measuring system comprising an optical transmitter and receiver placed on the same side of the glass ribbon in two boxes and / or two separate separate supports. 3. A measuring system consisting of a transmitter on one side of the ribbon and a receiver on the other side.

4. A measurement system comprising a transmitter and a receiver placed on one side of the ribbon and a reflector placed on the other side. 5. A stress measurement system completed by a tape temperature measurement system.

The measuring equipment is adapted to the temperature level in the drying room at the point where the measurement is made. For example, it will be integrated in a cooled box in order to keep all its components at temperature levels compatible with their proper operation.

Additional measuring equipment may be placed downstream of the lehr. The information delivered by the measuring equipment can be used by the operators of the installation to manually adjust the operating parameters of the lehr.

According to another exemplary embodiment, the measurements of the stress components, in particular σ _x and their profiles, in particular in the vertical and lateral direction, can be displayed for the information of the operator of the lehr to enable him to confirm the setting of the heating and cooling distributions operated on the lehr. It is also possible to record the values, in particular the permanent stress, for example in the form of curves, in particular for monitoring the quality of the product.

Preferably, the information delivered by the measurement equipment G is used by an installation control system to automatically adjust the operating parameters of the lehr, via a control loop, in particular for adjusting the heating and cooling of the glass in the running direction of the ribbon and its perpendicular direction.

The regulation loop can be advantageously completed by a physical model of the annealing of the glass which, from the measurements made in one section of the lehr, allows the calculation of the instructions of the different zones upstream and downstream of the measurement section. , for heating and cooling the glass ribbon at each step of the glass annealing process.

FIG. 7 schematically represents different embodiments of the loop control loop of the K-channel from the information supplied by the equipment G for measuring the stresses.

One or more optional measurement points 12, with measuring equipment G, may be provided for the measurement of the stresses. The information of the measurement points is sent to a processing station 13, stress analysis and control. Station 13 sends instructions to a leech control unit 14. This control unit 14 sends instructions to a control cabinet 15 for various equipment such as fans, electric heating, position control of the valves. One can provide a unit 16 in which is stored a physical model of annealing glass. The information from the analysis station 13 is then sent to the unit 16 for comparison with the model and output instructions to the control unit 14. The unit 16 can furthermore receive results from means 17 of measurements. complementary parameters of the glass ribbon, for example the temperature.

Stress analysis by the diffuse light method is discussed in more detail with reference to Fig.8 and 9.

An ordinary ray of light that passes through a glass sample is generally not 100% transmitted due to diffusion in the sample. This diffusion can be considered as a secondary vibration of the matrix excited by the main ray. It is reflected by scattered light propagating radially from the main beam in a plane y-z perpendicular to the x direction of the beam. The observation of this scattered light shows that it is polarized in the plane y-z in a sample without constraints. The analysis of the polarity of this scattered light makes it possible to measure the stresses in a sample because the field of stresses in the glass changes the polarization thereof.

According to this measurement principle, a non-polarized, preferably monochromatic light source is used, and it is the sample that polarizes the scattered light according to its stress level. Part of the scattered light coming out of the glass is picked up by an analysis system comprising an analyzer and an optical sensor (CCD camera or photomultiplier) to measure the polarization of the outgoing signal.

This measurement principle is also based on the analysis of the light scattered during the passage of a ray of light through a sample, but using polarized light as the source of incident radiation. The stress field present in the sample causes a change in the polarization of the light along its path in the sample. This leads to a spatial modulation of the intensity of the scattered light as a function of the orientation of the polarization of the main beam. Since the dipoles of the matrix vibrate parallel or perpendicular to an observer in the yz plane, only the positions with a polarization perpendicular to the direction of observation effectively emit scattered light in the direction of observation. The fringes obtained, observed by an optical sensor (CCD camera or photomultiplier), are directly representative of the stress level of the glass because they correspond to the alternation of the polarization of the source ray. The glass sample acts with this measuring principle as an analyzer.

The two principles of analysis A and B make it possible to characterize the stress in the sample. However, Principle B is more often used on laboratory samples because it is easier to use. An embodiment of a measurement system G according to principle B is described with reference to FIGS. 8 and 9, and then the few differences for a system developed according to principle A will be specified.

Light source In principle, any light source Q (Fig.8) with wavelengths included in the optical window of the glass can be used to create the incident beam of polarized light. However, some criteria can increase the performance of the system:

• A source of short wavelength will benefit from a greater amount of scattered light (more pronounced Rayleigh scattering effect),

• A monochromatic laser source will improve the signal-to-noise ratio and avoid dispersion effects,

• A laser source, which automatically emits an almost parallel and small diameter beam, will simplify a focus optics OfI

(Fig.8) disposed between the source Q and the glass ribbon

• A laser source can emit a polarized beam directly.

• The polarized beam is modulated to create a periodic shift of the phase. Each measurement point in the sample varies its light intensity scattered according to the offset period. This measurement increases the spatial resolution of the measurement.

A suitable source for Principle A is differentiated by an unpolarized light beam while it is polarized for Principle B.

Detection

A DT optical system for analyzing polarized light faces two requirements: 1. To be kept at a low temperature in the hot environment of a lehr,

2. Allow spectral filtering of the signal to eliminate heat radiation.

Thermal protection / filtering of the detection system The protection of the optical system DT from the heat in the lehr is effectively obtained by placing all its components in a chamber N (Fig.9) cooled with water with a sweep by a flow of air or nitrogen evacuated by the sighting tube T.

The filtering of incoming rays to eliminate heat radiation is carried out in several successive stages consisting of:

. A filter FL1 in soda-lime glass which suppresses the wavelengths higher than 2.7 μm,. A short pass IR filter FL2 which cuts the radiation from the wavelength to be analyzed (preferably in the visible).

. A 'long pass' filter (high pass) FL3 which makes it possible to eliminate the wavelengths lower than the wavelength to be analyzed. . A monochromatic filter FL4 of high precision (eg filtered bandwidth dλ <20 nm) adapted to the wavelength to be detected.

This succession of filters eliminates most of the thermal radiation detrimental to the accuracy of the measurement.

The signal is then focused by focusing optics Ofl2 on the detection system J, ie the CCD or CMOS sensor of a high sensitivity camera capable of detecting light of very low intensity.

(High sensor quantum efficiency, low "dark current" or "dark current").

The sensitivity of the detection system J can be further enhanced by the intermittency technique (involving a chopper), with the elimination of system-specific noise by an optical chopper or chopper, possibly combined with a intermittent source.

The signal obtained is then transmitted to a processing system TR to obtain the image of the spatial modulation of the light emitted by the sample and finally, the stress field in the glass corresponding to this distribution of light.

A detection system for equipment according to Principle A is designed according to the same criteria and includes the same succession of filters. Particular attention is given to the cooling of the filters which must be axisymmetric to avoid any creation of membrane stress in the filter material which would modify the polarization of the signal to be analyzed. Similarly, the quality of the filters is greater here to avoid that they induce a distortion of the polarization of the signal, because it is this polarization that is measured to characterize the level of stress. For this purpose, an ANL polarization analyzer (FIG. 9) is added with respect to a device according to principle B. The signal obtained is then transmitted to a processing system to obtain the image of the distribution of the polarized light from of the sample, then finally, the stress field in the glass corresponding to this distribution of the polarized light.

In the simplest configuration of stress analysis by scattered light, the beam enters the edge of a sample to cross it parallel to its surface. This method is not applicable to tempered glass because of the difficulty in taking samples. In this case (see FIG. 8), a variant is used: the incident ray M. penetrates the surface of the glass 1 with an inclination α relative to the surface of the glass. The length of the radius in the sample depends on this angle of inclination α, the refraction on the surface and the thickness of the sample. We try to keep a large length of the radius in the glass which allows to maintain a good spatial resolution.

A prism placed on the glass would obtain this incident beam at a grazing angle in the glass. It would avoid the reflection of the beam on the surface of the glass and allow the polarization of the beam to be maintained at the point of entry into the sample without a difference in intensity between its vertical component and its horizontal component as it would have been induced by the passage of the light beam through an interface between two different refractive index materials.

It is generally not possible to place a prism on a scrolling glass ribbon in a lehr, which makes measurement more difficult.

The incident ray _million. form, according to Fig.8, an angle α greater than 10 ° with respect to the surface of the glass to limit reflection losses. The beam then passes through the thickness of the glass at an angle of about 40 ° which limits the length of its path in the glass. The offset of the vertical and horizontal components of the polarized beam by the plane stresses in the glass becomes small. The operation of the static signal becomes sophisticated because it must take into account the coupling of the vertical component and of the polarized beam in the glass, the variation of the stress on the thickness, the angle of the 'horizontal' and 'vertical' component of the beam with respect to the plane stress (σ _x or σ _y ), the attenuation signal on his way out of the glass. The periodic modulation of the polarization of the incoming beam makes it possible to circumvent this sophisticated exploitation. It makes it possible to determine the phase shift between two neighboring points on the path of the beam and to deduce the average stress between these two points. Despite the unfavorable angle of the beam through glass, a good spatial resolution of the profile measurement in the thickness is maintained. The integration of the plane stress on the thickness then gives the membrane stress in the direction considered.

For the application of this method in a lehr, it is therefore essential to create a clean and undisturbed signal by parasitic rays. In the closed sections of the lehr, disturbance by ambient light is automatically excluded. For open sections, local obscuration is achieved, for example by curtains.

Disturbance by thermal radiation in the lehr is also to be considered. For a temperature of 600 ⁰ C at the entrance of a lehr, a "black body" environment would produce a total hemispheric energy flow of 33 kW / m ² . A camera with an optical aperture of 20mm diameter would thus receive a radiation heat flux of about 10W compared to the 0.5W of a class III laser used as a source. It is therefore essential to reduce the thermal radiation, hence the need to filter it with the FL1-FL4 system described above. At the output of the filtration system, the signal obtained has a spectral band limited to about 20 nm. The hemispherical energy flow of a black body at 600 ° C between 500 and 520nm has more than 2.1μW / m2 and between 400-420 nm more than 2.5 nW / m2. The fraction that will enter the optical aperture of the camera is even smaller. It corresponds to a flow of 0.8 pW for an aperture of 20mm in diameter.

The order of magnitude of the attenuation of visible light in soda-lime glass is 60 dB / km. It is a function of the wavelength and the quality of the glass. Over a length of 1mm in the sample there remains only 6x10 ⁵ dB / mm. From a polarized laser source, for a beam entering the glass having a power of 0.1 W, the amount lost by Rayleigh scattering on the first mm is 1.4μW. If it is assumed that 0.1% of this scattered light is captured, the detection system receives a flow of 1.4nW. Since this value is 3 times greater than the 0.8 pW of residual heat radiation flux after filtering, the signal can therefore be exploited properly.

The non-contact measuring equipment of the invention makes it possible, in particular, to measure all the three components σ _x , σ _y , σ _z , to determine the principal stresses, at any point of the ribbon in a flat glass drying rack, and thus to identify critical locations and optimize the cooling setting. It is possible, of course, to draw the two conventional profiles as 'vertical profile plane stress' and 'membrane stress σ _x lateral profile'. In addition to the state of the art, these profiles are measured in the lehr and thus obtain a direct measurement of the total and temporary stress.

The measuring equipment allows the measurement of the components of the stresses in different orientations to deduce the orthogonal main stresses. Stress measurements made along the longitudinal direction of the glass ribbon are used to adjust the distribution of heating and / or cooling distribution over the length of the ribbon.

Claims

1. Flat glass production plant comprising a melting and refining furnace followed by a device for forming a flat glass ribbon and an outrigger, and comprising on-line measurement equipment and contact-free constraints in the glass ribbon, characterized in that the in-line measuring equipment (G) is implanted in the lehr (K), and comprises a light emitter (Q) which directs a light beam on the ribbon glass, a receiver and light-scattering means scattered in different directions of space resulting from the interaction of the beam with the glass and means for spectrally filtering the incoming rays in the optical system (DT) of the equipment (G) to eliminate thermal radiation detrimental to the accuracy of the measurement.

2. Installation according to claim 1, characterized in that the measuring equipment (G) uses the effect of the elastic scattering of light, called "Rayleigh scattering" or the effect of the interaction of light with the phonons, called "Brillouin scattering or Raman".

3. Installation according to claim 1 or 2, characterized in that it comprises a CCD camera for measuring and analyzing the scattered light.

4. Installation according to one of claims 1 to 3, characterized in that the measuring equipment allows the characterization of an elongated longitudinal scrolling glass ribbon for drawing the vertical profiles of the plane stresses and the lateral or longitudinal profiles of the membrane constraints at different positions in the lehr.

5. Installation according to one of claims 1 to 4, characterized in that the measuring equipment allows the measurement of the components of the constraints in different orientations to deduce orthogonal main stresses.

6. Installation according to claim 1, characterized in that it comprises a control system (13,14) which exploits the information delivered by the equipment (G) to automatically adjust the operating parameters of the outrigger via a control loop.

7. Installation according to claim 6, characterized in that the combination of the control system (13,14) and the measuring equipment (G) is provided to allow to quickly adjust the operating parameters of the lehr ( K) so that the total stress level remains lower than a determined value making it possible to avoid breakage of the glass or ribbon deformations perpendicular to the plane of the ribbon and that the level of permanent stress remains below a determined value allowing the treatment subsequent glass.

8. Installation according to claim 1, characterized in that a transverse displacement system (11) of the measuring equipment is used to measure the stress levels on the ribbon width.

9. Installation according to claim 8, characterized in that the stress measurements made according to the width of the glass ribbon are used for adjusting the heating distribution over the width of the ribbon and / or the adjustment of the distribution of the ribbon. cooling over the width of the ribbon.

10. Installation according to claim 1, characterized in that a longitudinal displacement system of the measuring equipment is used to measure the stress levels along the length of the tape.

11. Installation according to claim 10, characterized in that the stress measurements made in the longitudinal direction of the glass ribbon are used to adjust the distribution of heating and / or distribution of cooling along the length of the ribbon.

12. Installation according to one of the preceding claims, characterized in that a mathematical model (16) of operation of the furnace is established and used to define the optimum instructions to be applied to the lehr, according to the measurements made, in order to to obtain the desired level of stress.

13. Installation according to one of the preceding claims, characterized in that it comprises a cooling device of the measuring equipment. (G) for placement in a relatively high temperature area of the lehr.

14. Installation according to claim 1, wherein the optical system (DT) comprises a sighting tube, characterized in that it comprises an optical chopper after the sighting tube.

15. Equipment for a flat glass production plant according to any one of the preceding claims, for measurement in line and without contact, in the lehr, constraints in the glass ribbon, characterized in that it comprises a transmitter of light (Q) which directs a light beam on the glass ribbon, and receiver and analysis means (DT) of light scattered in different directions of space resulting from the interaction of the beam with the glass.

16. Equipment according to claim 15, characterized in that it uses the effect of the elastic scattering of light, called "Rayleigh scattering" or the effect of the interaction of light with phonons, called "Brillouin scattering". or Raman ".

17. Equipment according to claim 15 or 16, characterized in that it comprises a CCD camera for the measurement and analysis of the scattered light.

18. A method of driving a flat glass annealing lehr in a flat glass production plant according to any one of claims 1 to 14, characterized in that a continuous measurement of the stress of the glass ribbon is performed by measuring equipment installed in the lehr, and is used to automatically adjust the operating parameters of the lehr via a control loop.

19. A method according to claim 18, characterized in that a combination of the control system of the lehr and the equipment for measuring the stress is provided to allow the operating parameters of the lehr to be adjusted rapidly. so that the total stress level in the glass ribbon remains lower than a determined value making it possible to avoid breakage of the glass or deformations of the ribbon perpendicular to the plane of the ribbon, and that the level of permanent stress remains below a determined value allowing the subsequent treatment of the glass.

Method according to claim 18 or 19, characterized in that stress measurements are made according to the width of the glass ribbon and are used for adjusting the distribution of the heating over the width of the ribbon and / or the adjustment the distribution of cooling over the width of the ribbon.

21. The method of claim 18 or 19, characterized in that a mathematical model of operation of the lehr is established and used to define the optimum instructions to be applied to the lehr, according to the measurements made, in order to obtain the desired temperature level and stress.