METHOD FOR ASSOCIATING A MARKING WITH AN OBJECT

Abstract

Disclosed is a method for associating a marking with an object, including the following steps:—identifying the position of at least two different elements of the marking in relation to the marking and/or the object; and—measuring a relative distance between at least two identified elements; then—recording in a database the position of at least two identified elements, and the relative distance between the identified elements so that the position of two identified elements is correlated with the measurement relating to their distance.

Claims

1. A method for associating a marking with an object, wherein, when the marking includes several distinctive elements the association method implements the following steps: identifying the position of at least two distinctive elements of the marking, with respect to the marking and/or to the object; and measuring a relative distance between at least two identified elements; then recording in a database the position of at least two identified elements, and the relative distance between said identified elements, in such a way that the position of two identified elements is correlated to the measurement relating to their distance.

2. The method for associating a marking with an object according to claim 1, wherein, when the marking comprises several distinctive cavities, the association method implements the following steps: identifying the position of at least one cavity of the marking with respect to the marking and/or to the object; and measuring at least one intrinsic characteristic of an identified cavity; then recording in a database the position of at least one identified cavity, and at least one intrinsic characteristic of the identified cavity, in such a way that the position of the cavity is correlated with at least one of the cavity's intrinsic characteristics.

3. The method for associating a marking with an object according to claim 2, wherein an intrinsic characteristic of a cavity is measured from a light pattern, reflected or transmitted by the cavity.

4. The method for associating a marking with an object according to claim 3, wherein a same light pattern is used to measure the intrinsic characteristics of several cavities of the marking.

5. The method for associating a marking with an object according to claim 4, wherein the light pattern is oval in shape.

6. The method for associating a marking with an object according to claim 5, wherein, when the marking is present on a cylindrical or partially cylindrical face with respect to a revolution axis, the light pattern lights the cavities in such a way that the greatest dimension thereof is perpendicular or substantially perpendicular to the axis of revolution.

7. The method for associating a marking with an object according to claim 5, wherein an intrinsic characteristic of a cavity is measured from the width of the light pattern inscribed in the cavity.

8. The method for associating a marking with an object according to claim 2, wherein at least two cavities are made by a hot-marking technique.

9. The method for associating a marking with an object according to claim 8, wherein the depth of the cavities is lower than 100 μm.

10. The method for associating a marking with an object according to claim 8, wherein, at a lateral face of the object, the greatest dimension of the cavities is equal to or lower than 500 μm.

11. The method for associating a marking with an object according to claim 1, wherein the marking is present on a wall at least partially transparent of the object.

12. A device for associating and/or checking the position of a marking made by hot marking on a glass container, comprising: an optical detection system, configured to identify the position and the shape of a marking on the glass container; means for memorizing the marking position and shape identified by the optical detection system; and calculation means adapted to implement an association method according to claim 1.

13. The association and/or identification device according to claim 12, wherein the optical detection system is telecentric and wherein the optical device includes a light source adapted to project an oval light pattern.

14. The method for associating a marking with an object according to claim 4, wherein the light pattern is rectangular in shape.

15. The method for associating a marking with an object according to claim 8, wherein the depth of the cavities is lower than 50 μm.

16. The method for associating a marking with an object according to claim 8, wherein, at a lateral face of the object, the greatest dimension of the cavities is between 400 μm and 100 μm.

17. The association and/or identification device according to claim 12, wherein the optical detection system is telecentric and wherein the optical device includes a light source adapted to project a rectangular light pattern.

18. The method for associating a marking with an object according to claim 6, wherein an intrinsic characteristic of a cavity is measured from the width of the light pattern inscribed in the cavity.

19. The method for associating a marking with an object according to claim 9, wherein, at a lateral face of the object, the greatest dimension of the cavities is equal to or lower than 500 μm.

20. The method for associating a marking with an object according to claim 15, wherein, at a lateral face of the object, the greatest dimension of the cavities is equal to or lower than 500 μm.

Description

DESCRIPTION OF THE FIGURES

[0034] The following description with respect to the appended drawings, given by way of non-limitative examples, will permit a good understanding of what the invention consists in and of how it can be implemented:

[0035] FIG. 1 is a front view of an original glass bottle of cylindrical shape, including on a lateral face an original Datamatrix code;

[0036] FIG. 2 shows a method for measuring the position of the Datamatrix code on the glass bottle illustrated in [FIG. 1];

[0037] FIG. 3 shows a method for measuring the position of a copy of the Data matrix code present on the glass illustrated in [FIG. 2], present on a counterfeit glass bottle;

[0038] FIG. 4 shows an enlarged view of the Datamatrix code present on the bottle illustrated in [FIG. 1];

[0039] FIG. 5 shows an enlarged view of the Datamatrix code present on the bottle illustrated in [FIG. 3];

[0040] FIG. 6 shows a partial longitudinal cross-section of the bottle illustrated in [FIG. 1], at its Datamatrix code;

[0041] FIG. 7 shows an enlarged view of the Datamatrix code present on the bottle illustrated in [FIG. 1], lighted by a light strip;

[0042] FIG. 8 shows an enlarged view of the Datamatrix code present on the bottle illustrated in [FIG. 3], lighted by a light strip;

[0043] FIG. 9 is a diagram of a device according to the invention, for associating and/or checking the position of a marking made by hot marking on a glass container;

[0044] FIG. 10 shows a diaphragm belonging to the association and/or checking device illustrated in [FIG. 9];

[0045] FIG. 11 shows a diagram of an alternative device according to the invention, for associating and/or checking the position of a marking made by hot marking on a glass container.

DETAILED DESCRIPTION OF AN EXEMPLARY EMBODIMENT

[0046] As mentioned hereinabove, the traceability is a major challenge to guarantee the authenticity and quality of a product. This is in particular the case in the field of glass container recycling. Indeed, the risk of breaking a glass container increases as a function of the number of recycling cycles performed. It is hence useful that this number can be accurately quantified, in order to discard the glass containers that have a too high risk of breakage.

[0047] For that purpose, it is known to use a specific marking associated with each glass container. As illustrated in [FIG. 1], a glass container 2 can include, at a lateral face 4, a specific marking in the form of a high-density two-dimensional barcode, also called Data matrix code 6. A Data matrix code is in the form of a matrix consisted of juxtaposed points or squares. According to the example illustrated in [FIG. 1], the Datamatrix code is composed of four points A, B, C and D, forming a substantially rectangular pattern.

[0048] Nevertheless, the counterfeit glass containers can also include a Datamatrix code substantially identical to that present on an original glass container. That way, the introduction of these counterfeits into the recycling cycle biases the process of counting the number of recycling cycles described hereinabove. The invention aims to solve this technical problem by proposing a method making it possible to more surely associate a specific marking to a unique glass container.

[0049] According to a first embodiment of the invention, during a first step, the position of a Datamatrix code 6 present on a lateral face 4 of a glass container 2 is measured. The glass container 2 can for example be a cylindrical, flat-bottom bottle, as illustrated in [FIG. 1]. The bottom 8 of the bottle is then used as a landmark to measure the height of the Datamatrix code 6, and the revolution axis 10 of the bottle is used as a landmark to measure the inclination of said code on the bottle. Of course, these markers are mentioned by way of example and are not intended to limit the scope of the invention. Other landmark points or axes can hence be chosen, such as the bottle neck, an edge or any other distinctive feature. The height H of the Datamatrix code is evaluated by measuring the smallest distance between an arbitrarily chosen point of the Datamatrix code, in the present case point A, and the bottom 8 of the bottle. The inclination of the Datamatrix code is evaluated by measuring the angle 9 formed between the cylindrical axis 10 of the bottle and a straight line 11 passing through the centre of two points of the Datamatrix code, for example points A and B.

[0050] According to a second step, the Datamatrix code shape is identified from its contour, the arrangement of its constituent points and/or its size. Then, according to a third step, called association step, the Datamatrix code shape is associated in a database to both the height H and the inclination angle 9 measured during the first step. According to an alternative embodiment, the first and second steps may be inverted or be carried out simultaneously.

[0051] The association method can possibly include an intermediate step, consisting in identifying a specific information item from the Datamatrix code shape. This information item can then be substituted to the shape of the Datamatrix code 6 in the database. A specific information item of a Datamatrix code can indicate the site of production of glass container, its manufacturing mould and/or a series number associated with said container. Both the height H and the inclination angle 9 can then be associated in a database with either one of the information items mentioned hereinabove. For example, both the height H and the inclination angle 9 can be associated, in a database, with the site of production of the glass container.

[0052] The invention also relates to a method for checking the originality of a Datamatrix code 6′ present on a glass container 2′, as illustrated in [FIG. 3]. The checking method implements the first and second steps described hereinabove. During a third step, called checking step, both the height H and the inclination angle 9′ of the Datamatrix code 6′, measured on the glass container 2′, are compared with a height H and an inclination angle 9 previously recorded in the above-described database and associated with the shape or with an information item linked to the Datamatrix code 6. According to an alternative embodiment, the first and second steps may be inverted of be carried out simultaneously.

[0053] The above-described association and checking methods make it possible to more surely verify that the Datamatrix code 6 present on a glass container 2 is authentic. Indeed, the application of the making of a Datamatrix code always shows a certain inaccuracy when it is arranged on an object. That way, as illustrated in FIGS. 2 and 3, the reproduction of an identical Datamatrix code on two different glass containers 2 and 2′ is characterized on each glass container by slightly different height and inclination of the Datamatrix codes 6 and 6′. The association and checking methods according to the invention make it possible to detect very rapidly these differences in order to more surely identify that the glass container 24 is a counterfeit of the glass container 2, despite the fact that the Datamatrix code 6′ is substantially identical to the Datamatrix code 6. The counterfeiting glass container 2′ will then be identified by the invention as a counterfeit of the glass container 2 having to be discarded.

[0054] Of course, the accuracy of the measurements performed hereinabove must be equal to, preferably lower than, the uncertainty range during the application or the making of a Datamatrix code on a glass container. Now, this uncertainty depends on the technique used to stick or to make the Datamatrix code on a glass container. By way of non-limitative example, when the Datamatrix code is stuck on a glass container, its height and inclination are respectively measured with an accuracy equal to or lower than one hundredth of millimetre and an accuracy equal to or lower than one tenth of degree, respectively. When the Datamatrix code is made on a glass container, by a hot-marking technique, its height and inclination are measured with an accuracy equal to or lower than one hundredth of millimetre and an accuracy equal to or lower than one tenth of degree, respectively.

[0055] According to a second embodiment of an association method according to the invention, during a first step, the relative position between at least three points of a Datamatrix code 6 illustrated in [FIG. 4] is measured. According to the present example, the distance X1 between the centres of the points A and C and the distance Y1 between the centres of the points A and B are measured. During a second step, the shape of the Datamatrix code 6 is identified. Then, according to a third step, called association step, the shape of the Data matrix code is associated, in a database, to the measured distances X1 and Y1. According to an alternative embodiment, the first and second steps may be inverted or be carried out simultaneously. Of course, the shape of the Datamatrix code 6 can be associated with a higher number of distances measured between other points of the code.

[0056] The invention also relates to a second method for checking the originality of a Datamatrix code on a glass container, implementing the first and second steps of the second embodiment described hereinabove. During a third step, called checking step, the value of the measured distances X1′ and Y1′ between the points of the Datamatrix code 6′ is compared with the distance values X1 and Y1 recorded for the same points in the database established hereinabove, and associated with the information item of said Datamatrix code 6. According to an alternative embodiment, the first and second steps may be inverted or be carried out simultaneously.

[0057] The second association and checking methods make it possible to more surely verify that the Datamatrix code present on a glass container is authentic. Indeed, the application or the making of a point forming a Datamatrix cod always shows a certain inaccuracy. That way, as illustrated in [FIG. 5], the reproduction of a same Datamatrix code on a counterfeit glass container will be characterized by slightly different spacings between the same points of the Datamatrix code. The glass container 2′ will then be immediately identified by the invention as a counterfeit of the glass container 2, having to be discarded.

[0058] Of course, the measurements performed hereinabove must be equal to, preferably lower than, the uncertainty range for making the points of a Datamatrix code. Now, this uncertainty depends in particular on the technique used to stick or to make said points on a glass container. By way of non-limitative example, when the Datamatrix code points are printed on a glass container, their distance is measured with an accuracy equal to or lower than one tenth of millimetre, preferably with an accuracy equal to or lower than one hundredth of millimetre. When the Datamatrix code is made on a glass container, by a hot-marking technique, their distance is measured with an accuracy equal to or lower than one tenth of millimetre, preferably with an accuracy equal to or lower than one hundredth of millimetre.

[0059] By way of example, in a Datamatrix code composed of a grid of 16 points by 16 points, with a height and a width of 8 mm, the theoretical pitch between two points is of the order of 500 micrometres. The accuracy of the measurement according to the invention is equal to or lower than 10 micrometres, in such a way that the fluctuations of the Data matrix code manufacturing method as mentioned hereinabove can be measured. Usually, the measurements made to read a Datamatrix code are performed with an accuracy higher than or equal to 100 micrometres. In other words, the usual methods for reading Datamatrix codes do not measure the fluctuations of positioning of the Datamatrix points.

[0060] The invention also relates to a third method for associating a Datamatrix code with a unique glass container, the Datamatrix code being made by a hot-marking technique. As illustrated in [FIG. 6], the hot marking consists in making cavities 12A, 12B, 12C and 12D at a front face 14 of a glass container, by means of a LASER source, for example. As known, the cavity sizes vary as a function of the position and the inclination angle of the front face 14 with respect to the focus point of the LASER beam. For economical reasons, the hot marking is made when the glass containers 2 are conveyed from a station to another one. That way, each glass container has a substantially different position at the time of its marking, due to its displacement, with respect to the focusing point of the LASER beam. That is why, for a same point of a same Data matrix code, the position and the size of the cavity associated with said point, vary from one glass container to another, as illustrated in FIGS. 4 and 5.

[0061] FIGS. 4 and 6 show a front view and a longitudinal cross-section, respectively, of a Datamatrix code 6 made by a LASER source on a glass container 2 of cylindrical shape with respect to an axis of rotation 10. Theoretically, points A, B, C and D forming the Datamatrix code 6 should be of identical size. Now, for each cavity 12A and 12B, the diameter, the depth and the convexity vary for the above-mentioned reasons. The invention proposes to advantageously use these uncertainties linked to the hot-marking technique to more surely identify the originality of the Datamatrix code 6 on an original glass container 2, from at least one measurement of the image of light pattern reflected by a cavity composing the Datamatrix code.

[0062] For that purpose, according to a third embodiment of the invention for associating a Datamatrix code 6 with a unique glass container 2, during a first step, a light pattern is projected on the glass container 2 as a light strip 16. The light strip lights the glass container 2 is such a way that its greatest dimension is perpendicular or substantially perpendicular to the axis of rotation 10 of the glass container 2. The light strip 16 is sized in such a way as to light at least two points or two cavities of a Datamatrix code 6 present of a front face 14 of the bottle, preferably the whole Datamatrix code 6.

[0063] According to a second step, for at least one cavity, the width of the light strip 16 inscribed in said cavity is measured. By “width”, it is meant herein a dimension of the light strip along a direction parallel or substantially parallel to the axis of rotation 10 of the glass container 2. According to the present example, the width of the light strip is measured for the four cavities: YA, YB, YC, YD. According to a third step, the shape of the Datamatrix code 6 is identified. According to an alternative embodiment, the second and third steps may be inverted or be carried out simultaneously.

[0064] During a fourth step, called association step, to the shape of the Datamatrix code 6 is associated, in a database, at least one cavity A with the measured width YA of the light strip visible in said cavity.

[0065] The invention also relates to a third method for checking the originality of a Datamatrix code present on a glass container 2′, the Datamatrix code being made by a hot-marking technique. The third checking method implements the first, second and third steps described hereinabove, for measuring the width YA′ of a light strip 16 inscribed in at least one cavity A′ and identifying the shape of the Datamatrix code 6′. According to an alternative embodiment, the second and third steps may be inverted or be carried out simultaneously. During a fourth step, called checking step, the measured value of width YA′ of a light strip inscribed in a cavity A′ of the Datamatrix code 6′ is compared with a width YA of a light strip recorded in a database established hereinabove, corresponding to the same cavity A associated with the same shape of the Datamatrix code 6.

[0066] The third association and checking methods described hereabove make it possible to more surely verify that the Datamatrix code 6 made by a hot-marking technique on a glass container 2 is authentic. Indeed, as mentioned hereabove, the hot marking of a Datamatrix code always shows a certain inaccuracy during its making. That way, as illustrated in FIGS. 7 and 8, the reproduction of the Datamatrix code 6 on a counterfeit glass container 2′ is characterized by different cavity depth and/or shape between two points of the Datamatrix code 6.

[0067] Advantageously, the invention uses this intrinsic defect of hot marking to characterize several cavities of the Datamatrix code from the image of a same light pattern reflected by said cavities. Indeed, the size of the light strip reflected by each cavity vary substantially from one cavity to another, due to the difference of depth and/or shape of said cavities. As illustrated by FIGS. 7 and 8, different widths YA and YA′ can be observed for the light strips inscribed in cavities A and A′, representing a same point of an original and a counterfeit Datamatrix code, respectively. This difference of widths YA and YA′ mainly reflects, in the present case, depth difference between cavities A and A′. Now, as mentioned hereinabove, each cavity has a substantially unique depth due to the intrinsic uncertainties of the technique for hot marking glass bottles travelling on a conveyor. Thanks to the invention, it is hence possible to easily detect a low depth difference between two cavities representing the same point of an identical Datamatrix code. In other words, the invention makes it possible to rapidly and more surely ascertain the originality of a Datamatrix code, made by a hot-marking technique on a glass container, by simply and rapidly checking that the size of a light pattern reflected by at least one cavity of the code corresponds to the size previous recorded in a database.

[0068] Hence, advantageously, the association and checking methods according to the invention make it possible to very rapidly detect these width differences between the light strips inscribed in the cavities, in order to more surely identify that the glass container 2′ is a counterfeit of the glass container 2, despite the fact that the Datamatrix codes 6 and 6′ rigorously code the same information. The counterfeiting glass container 2′ will then be immediately identified by the invention as a counterfeit having to be discarded.

[0069] According to an alternative of the third association and/or checking method, a calibration step is implemented. This calibration step consists in measuring the depth of several cavities using a chromatic confocal optical profilometer. In other words, the calibration step makes it possible to perform a map of the depths of the cavities forming the marking. The calibration step can be implemented before or after the first step described hereinabove. Advantageously, the calibration step makes is possible to establish a calibration curve between the measured widths of the light strips and the real depths of the cavities that reflect said light strips. The measured cavity depths can also be used as a variable for coding the Datamatrix code, for example.

[0070] Of course, the measurements made during the third association and checking methods described hereabove must be performed substantially in the same conditions in order to compare the measurements made in the best conditions. Nevertheless, according to an alternative embodiment, relative variations of light strip widths between two cavities can be compared, in order to minimize the influence of the conditions of acquisition of the measurements between the association method and the checking method. This embodiment advantageously makes it possible to get rid of the measurement uncertainties due to the measuring devices themselves.

[0071] The invention also relates to a device 18 for associating and/or checking a Datamatrix code 6, made by a hot-marking technique on a glass container 2, adapted to implement the third association and/or checking method described hereinabove. An association and/or checking device 18 according to the invention is illustrated in [FIG. 9]. The device 18 includes a telecentric optical detection system 20, a light source 22 and a semi-reflective plate 24. The semi-reflective 24 is arranged in such a way as to allow a light beam 26 emitted by the light source 22 to light the surface of a glass container 2 present in the field depth 28 of the optical detection system 20. Of course, the semi-reflective plate 24 is configured in such a way as to allow the light reflected by the glass container 2 to be perceived by the optical detection system 20.

[0072] The optical detection system 20 is connected to a central unit 30. The central unit 30 comprises means for memorizing an association and/or checking method described hereinabove as well as a database. The central unit 30 also comprises calculation means adapted to implement the memorized method(s), in an automated manner or not. Potentially, the central unit 30 can include a display device allowing an operator to verify the image acquired by the optical detection system 20.

[0073] The device 18 also includes a diaphragm 32 delimiting a rectangular aperture 36. The diaphragm 32 is interposed between the light source 22 and the semi-reflective plate 24, in such a way that the light beam 26′ passing through the aperture 36 forms a light strip 16 adapted to light at least one portion of a Data matrix code 6 hot marked on a glass container 2.

[0074] The length D1 of the diaphragm 32 is chosen in such a way that the light strip 16 covers at least the length of the portion of space observed by the optical detection system 20. By “length of the portion of space”, it is meant a dimension of the space observed by the optical detection system 20, along a direction perpendicular or substantially perpendicular to the optical axis of detection of the optical detection system 20. By way of non-limitative example, the length D1 is comprised between 2 cm and 20 cm, preferably between 5 cm and 15 cm.

[0075] The width D2 of the diaphragm 32 is chosen in such a way that the image of the light beam 16 can be identified in at least one cavity of a Datamatrix code made on the surface of a glass container, the cavity reflecting the light beam 26′ emitted by the light source 22. In other words, the width of the diaphragm 32 must be sufficient for the optical detection system 20 to allow the central unit 30 to measure the width of the slit image, for example YA as shown in [FIG. 7], at said cavity A of the Datamatrix code 6 present on the glass container 2. By way of non-limitative example, the width D2 of the diaphragm 32 is between 0.5 cm and 5 cm, preferably between 1.25 cm and 3.5 cm.

[0076] According to the present example, the optical detection system 20 comprises a high-resolution camera provided with a 5-Megapixel optical sensor, with a macro lens allowing an observation field at 100 mm from the lens, of 20 mm by 20 mm, and a field depth of the order of 20 mm. The light source 22 includes several light-emitting diodes, configured to emit a light beam in a wavelength range included in the visible light spectrum. Preferably, the light source 22 projects a telecentric or substantially telecentric light beam.

[0077] As illustrated in [FIG. 4], the device 18 described hereinabove advantageously allows obtaining a high-contrast image with little geometric distortion, of a Datamatrix code 6 present on a lateral face 4 of a glass container 2 located in the field depth of the optical detection system 20.

[0078] When the Datamatrix code is present on a cylindrical front face 14, an orientation of the container is favoured in the field depth area 28 of the optical detection system 20, in such a way that the light strip 16 is perpendicular or substantially perpendicular to the axis of revolution 10 of said face. This embodiment advantageously makes it possible to perform measurements of the width of the light strip 16, which depends little or not on the radius of curvature of the front face 14 of the glass container 2, in a plane that is normal or substantially normal to the axis of revolution of said face.

[0079] According to an alternative embodiment illustrated in [FIG. 11], a device 18′ according to the invention can include an optical diopter 36 for focusing the light strip 16 at a Datamatrix code hot marked on a glass container. This embodiment is particularly advantageous when the Datamatrix code 6 is present on a convex or concave surface, to optimize the light flow and/or to obtain a homogeneous light flow, in order to limit the normal measurement aberrations and errors at any point of the curved surface, over at least the useful surface extend to be controlled. For example, a Fresnel lens 36 can be interposed between the diaphragm 32 and the glass container that is to be observed.

[0080] According to an alternative embodiment, the width D2 of the slit is adjusted as a function of the internal curvature of the cavity or cavities lighted by the light strip 16, in such a way that the width of the light strip observable by the optical detection system 20 can be measured by the central unit 30 in each cavity. It is to be noted that the hot marking of the Datamatrix points, with a LASER source, on a glass container, creates cavities whose depth and curvature vary. By way of non-limitative example, the depth of the cavities varies between 0 and 50 μm. That way, the inside of each cavity is characterized by a proper depth, reflecting the image of the light strip 16 with a different width. Hence, the width D2 of the diaphragm 32 is chosen in such a way that the edges of the diaphragm delimiting the width of the light strip 16 are both visible in at least two distinct cavities, preferably at least three distinct cavities. The width D2 of the diaphragm 32 is adjusted in such a way that the shallowest cavity of the Data matrix code reflects the image of a light strip that is the widest possible (see, for example, reference YA in [FIG. 7] and reference YD′ in [FIG. 8]). This choice is motivated in order to be able to observe the opposite edges of the light beam 16 reflected in each cavity. Indeed, the shallower a cavity, the larger its equivalent focal and the wider the generated image of the light slit (see [FIG. 7] and [FIG. 8]). It is to be noted that, for a same diameter of cavities at the front face 14 of the glass container 2, the depth of the cavities can vary, which translates into light slit images of different widths. The device according to the invention hence makes it possible to precisely make the distinction between cavities that are identical at the front face 14 but with different depths. In other words, the device according to the invention exploits the optical properties of each cavity to make the distinction between said cavities as a function of their depth.

[0081] By way of non-limitative example, for the cavity depth range mentioned hereinabove, a width D2 of the diaphragm 32 between 0.5 cm and 5 cm, preferably between 1.25 cm and 3.5 cm, is chosen when the front face 4 of the glass container 2 is positioned at a distance of the order of 60 mm from the diaphragm 32.