EP2627995A1

EP2627995A1 - Method and apparatus for measuring the quality of a transparent tubular object

Info

Publication number: EP2627995A1
Application number: EP11817229.5A
Authority: EP
Inventors: Michele Zanda; Cosimo Antonio Prete
Original assignee: Universita di Pisa
Current assignee: Universita di Pisa
Priority date: 2010-10-08
Filing date: 2011-10-10
Publication date: 2013-08-21
Also published as: WO2012046136A1; ITPI20100110A1; IT1402103B1

Abstract

An apparatus for measuring the quality of a tubular body (1) and the geometric position of an alteration or defect or impurities comprises a supply means for causing at least one first radiation beam and a second radiation beam to pass through a point of the object, for example light radiation generated by respective light sources, so that the directions of said first and second beams form a predetermined angle. The radiation beams are subjected each to four refractions when they cross curved surfaces. If in the point a defect is present, the position of the point can be found between a position X and Y on the two video cameras, whereas for an extended defect two limit positions (X1 and X2) (Y1 and Y2) are defined. By detecting, from known positions, the first light beam and the second light beam after that they have crossed the object, in order to determine the light alteration that occurs moving the same by the first and second light beam, a first and a second alteration signals are provided in case the alteration exceeds a predetermined threshold and with the program means a geometric position of the alteration in the object is determined. If the defect is at the surface, it can represent a crack capable of triggering a fracture, causing the object to be extremely vulnerable and destined to an easy break under a minimum action.

Description

TITLE

METHOD AND APPARATUS FOR MEASURING THE QUALITY OF A TRANSPARENT TUBULAR OBJECT

DESCRIPTION

Scope of the invention

The present invention relates to a method for measuring the quality of an object that is transparent with respect to a predetermined technology of inspection, with tubular shape, for example a glass tube.

Furthermore, the invention relates to a device arranged to carry out this method for measuring the quality.

Background of the invention - Technical problems

Inspection devices are known used for monitoring the production process and controlling the quality of products. In particular, such devices are used, for example, for tubes or cylinders of glass used as semi-finished products for manufacturing many products in the pharmaceutical field, such as vials, syringes etc.

It is therefore particularly relevant, in this field, to make a cylinder or a glass tube without defects and of very high quality, in particular to obtain syringes containing for example single-dose drugs/vaccines. In the latter case, in fact, the drugs have to be protected from biological contamination. In case of surface defects such as open air bubbles that form a rifling in the inner wall of the glass tube, a contamination could occur with the substance that could cause damages directed to the user. The surface air bubble can, in fact, connect the outer environment with the content of the syringe, bypassing the piston.

The defects to inspect are in most cases cavities created by air bubbles, formed during the hardening process of the glass, but they can be also fragments stuck to the glass when manipulating the product. In case of cavities, is important to determine whether such defects are a closed cavity or an open cavity. In particular, the closed cavity is formed completely in the thickness of the glass, whereas the open cavity is located near the inner or external surface of the glass tube and faces directly towards the inner/outer space. Hereafter, with the term defects either cavities or impurities are intended.

Some devices are based on a video camera technology, like WO0077499, WO9834096 or EP1816466, or a laser technology equipped with a software of image analysis, as disclosed in US4136779 or in US4168907.

For example, in US4483615 a system is described of inspection of glass tubes based on a linear video camera for checking surface defects present on the surface of the glass, causing the tube to rotate about its own axis. The system highlights only whether a defect is present or not, and it is not, therefore, capable of determining the position of the defect in the tube wall.

In WO2006039192 a device is described for inspection of defects or dust in a transparent object. In particular, the device comprises at least one light source and at least one array of sensitive elements, and the object is arranged therebetween, in order to determine a differential value of the light that crosses the object with respect the overall incident light.

In JP20091749 8 a video camera inspection device is described associated with a mirror system arranged to measure defects present in a plane transparent object. The above described device tests plane glass, for checking substantially powder and/or false defects.

However, the above described systems of inspection cannot check the position, the kind and the features of the defects in the glass, and then cannot detect the shape and three-dimensional position of the defect, nor its size, with the consequence that in the production lines some products are rejected by the presence of the defect as such, without taking into account additional features in an intelligent way. If the rejection step were more selective, the waste could be exploited better. In fact, elongated air bubbles can be inside the thickness of the tube and so they are not problematic for the pharmaceutical field, since they do not create for example a leakage line for a piston of a syringe. However, determining only one information of depth of the defect and not also the three-dimensional shape can make problems: if such closed elongated bubbles are of relevant size and/or close within the inner wall, their geometric features could be in any case dangerous: owing to mechanical or thermal stress they can become open lines on the inner wall and then bypass the piston of the syringe.

In other words, the above described devices cannot discriminate between closed cavity and open cavity. Therefore, such monitoring systems bring to reject all the products that have a predetermined type of defect, without determining the characteristics and the position of the same. Moreover, in certain cases a closed cavity defect can be embedded in the thickness of the object in the middle of the thickness and then does not involve any risk of break or contamination of the product. Yet, it is possible that the closed cavity defect is located near the inner or external surface and can turn into an open cavity by the mechanical or thermal stress to which the object is subjected.

In any case, an indiscriminate disposal requires an increase of the costs and of the time for production. If the position and the kind of defect are known, the tube can be used in applications compatible with the defects shown.

Summary of the invention

It is therefore a feature of the present invention to provide a method for measuring the quality of a transparent tubular object, in particular a transparent tube, capable of determining the three-dimensional position of impurities or other defects, such as air bubbles, in the object.

It is another feature of the present invention to provide such a method for checking the three-dimensional geometry, the position and the kind of impurities or defects present in the object.

It is another feature of the present invention to provide such a method for determining a level of quality of the object according to the dimension, shape, and position of the impurities or defects in it detected.

It is another feature of the present invention to provide an apparatus for measuring the quality of a transparent tubular object that achieves the same objects.

It is also a feature of the present invention to provide such an apparatus that is structurally easy and cheap to make.

It is also a feature of the present invention to provide such an apparatus that can be applied and arranged in a desired production line.

These and other objects are achieved by a method for measuring the quality of a tubular body, said tubular body being movable along a conveying direction Z, said tubular body being transparent with respect to a predetermined radiation, said method comprising the steps of:

- measuring and/or detecting the geometry of said tubular body, in particular a tube having a predetermined inner and outer diameter;

- arranging in a first position next to said conveying direction a source of said radiation;

- causing said tubular body to cross at least one first beam of said radiation and a second beam of said radiation coming from said source so that said first and second radiation beams are incident to each other according to a predetermined angle at said object;

- arranging in a second position next to said feeding direction, and opposite to said first position with respect to said tubular body, a radiation sensor;

- measuring said first radiation beam and said second radiation beam, by said sensor, after that said first and second radiation beams have crossed said tubular body, said measuring step determining an alteration which occurs when said tubular body is crossed by said first and second radiation beams, said measuring step providing a first and a second alteration signal (x, y) in case said alteration exceeds a predetermined threshold,

- analysing said first and second alteration signals (x, y) in order to determine in a first cross section (Z-i) of said object :

- a position (r, Θ) of said alteration in said first cross section;

- a first coordinate of axial position (z^ of said alteration for said first cross section;

- repeating according to a predetermined sampling frequency said measuring and analysing step along said conveying direction in order to determine, in a second, third, i^th, n^th cross section (Z₂, Z₃, ... , Z,, Z_n) of said tubular body, corresponding first and second alteration signals (x, y)₂.._n, as well as n positions (r, θ)₂.._η of said alteration, which are associated to respective coordinates of axial position (z₂, Z3, z- z_n);

determining a parameter of quality of said tubular body according to the radial position, angular and axial elongation of said alteration

This way, the method, through the first and second radiation beams, allows positioning the defect shape in the tube wall scanning it from two different observation directions with the first and second radiation beams, which are incident according to a known angle a. By determining the three- dimensional position of the shape and angle a, a triangulation is obtained for determining a depth position of the shape in the tube wall. Such method is applicable both when the tube slides continuously, and when the tube is in a fixed scanning position.

The sampling frequency can be high, i.e. between 10 and 100 microseconds, in order to achieve a high resolution of the defects in the tube, i.e. an air bubble in the glass tube wall that extends for a predetermined length of the tube.

This way, starting from two points of observation, the centre of the glass tube is detected. The first point of observation sees an the alteration signal with respect to the tube centre. The second point of observation sees another the alteration signal with respect to the tube centre. The interval of the alteration signal changes responsive to the position of the cause of the alteration signal in the wall of the tube, i.e. between a possible interval corresponding to a glass defect located at the outer surface of the tube, to an interval corresponding to a glass defect located at the inner surface of the tube.

In particular, the limit values of the interval can be calculated with either a simulative model or experimental tests, either on-line or off-line, for reproducing the geometric features of the hollow tube under inspection and the behaviour of the beam responsive to the features of the tube. Such simulation allows determining for each interval of the first altered signal all the relative possible intervals of the second signal responsive to the position of the impurity that has altered the signal, as well as clearly responsive to the outer and inner diameter of the tube.

The intervals are calculated instant-by-instant in accordance to the sampling frequency of the inspection device. The acquisition of all the intervals makes it possible to determine two contours that show the shape of the impurities from the two different points of observation. However, even if taken singularly the two contours are not a three-dimensional representation of the shape of the impurity that has caused the two alterations of the signal from the two points of observation, they allow to calculate a space in which the impurity is certainly present, and then to determine approximately their three- dimensional shape.

Advantageously, said sensor comprises a linear array of optical sensors, and said first and second alteration signals (x, y) consist of a first and a second interval of said pixels (Dx, Dy) of said linear array of optical sensors that receive an altered radiation

In particular, the step of analysing said first and second alteration signals (x, y) provides the control of said first and second intervals of said pixels (D_x, D_y) of said linear array of optical sensors that receive an altered radiation, in order to determine in each cross section of said object a position interval (r-_\, r₂; θ-ι , θ₂) of said alteration in said cross section.

This way, from said alterations of the signals and relative contours it is possible to determine the distance of said first and second contours (D_x, D_y) with respect at the centre of said tube. Thus, the distance of the first and of the second contour from the tube centre observed from the two measurement points allow to determine an three-dimensional shape information of the contour and of its depth. In particular, it can be ascertained if an air bubble in a glass tube is completely in the glass, or if emerges on the surface.

According to at least one feature as above defined, it is possible to determine an area where the shape contour of said alteration is contained with respect to said first and second signal.

Advantageously, a step is provided of determining along said conveying direction an approximated three-dimensional shape of said alteration.

In particular, the step of approximating a shape of a defect is effected by a step of comparing said first and second intervals of said pixels (Dx, Dy) for each cross section with known geometric parameters of defects or impurities.

The step of analysing the approximated three-dimensional shape of said defect can be advantageously associated with the step of selection of the quality of the tubular body. This way, the three-dimensional position coordinates and other size parameters of what has been detected can then be used to determine an index of quality of the glass tube, which is used for deciding whether accepting or rejecting the glass tube in production line.

In particular, said step of correlating comprises a step of comparing said first and second shape contour with known geometric parameters of defects or impurities in order to reduce the error of said approximated three-dimensional shape. For example, the linear air bubbles can have very elongated thin and circular three-dimensional shape. According to the requirements of the application, the three-dimensional shape can be approximated in an extended way, in order to include certainly the real three-dimensional shape. The approximated three-dimensional shape of the defect allows to position the shape three-dimensionally in the tube wall glass.

In particular, said tubular body is a tube and said step of detecting and/or acquiring the geometry of said tube comprises the step of detecting the edges of said tube and the step of detecting the centre of said tube, in particular said step of detecting the edges of said tube allows determining the outer diameter and the inner diameter of said tube.

A step can be provided of rejection of the forces to which said object is subjected along said feeding direction, said step of rejection comprising a step of repeating said step of detecting the centre of said tube instant-by-instant, in order to eliminate position artifacts of said tube. This way, the system is unaffected by vibrations since the points of observation look at the totality of the inspected objects (for example a glass tube cross section). Such arrangement makes it possible to calculate instant-by-instant the centre of the tube starting from the sides of the tube and to align to it in all the measurements.

Preferably, a step is provided of correction of said first and second alteration signals responsive to refractions that occur when the radiation crosses said tubular body. In particular, in case of a tube, the model and the experiments must consider that the light radiations cross both two curved surfaces separated by an air zone.

In particular, said step of determining a parameter of quality is fed back to a production step of said transparent object in order to correlate the features of said alteration for adjusting the process for forming said object.

In a first exemplary embodiment, said step of causing a first radiation beam and a second radiation beam to pass through said object comprises a step of arranging a first and a second source of radiation rotationally spaced from each other.

In a second exemplary embodiment, said step of causing a first radiation beam and a second radiation beam to pass through the hollow tube comprises a step of arranging a single source of radiation to emit an emitted radiation beam and a step of selectively directing said emitted radiation beam according to a first direction, in order to provide said first radiation beam, and according to a second direction, in order to provide said second radiation beam.

In particular, said step of directing said emitted radiation beam comprises a step of rotating a source according to said angle. The source and the sensor can in this case rotate about the tube at a high speed, for example at a predetermined sampling frequency (for example, between 10 and 100 microseconds per turn).

In the above described exemplary embodiments, said step of measuring comprises a step of prearranging a first sensor and a second sensor spaced by said angle.

In a third exemplary embodiment, said step of causing a first radiation beam and a second radiation beam to pass through the hollow tube, comprises a step of transmission of said emitted radiation beam and a step of rotating said object according to said angle set between a first and a second position spaced by said angle, wherein when said object is in said first position said emitted radiation beam is called first radiation beam, and when said object is in said second position said emitted radiation beam is called second radiation beam. In the third exemplary embodiment it is possible to determine with a single sensor the position of the defect through a single point of observation causing the object to rotate by a degrees.

Alternatively, said step of measuring comprises a step of prearranging a single sensor movable between said first and second position spaced by said angle.

In the above described particular cases, it is possible to adjust angle a, in order to maximize the positioning precision of the defect. For example, for inspecting defects along all the hollow tube it is necessary convey the latter through the points of observation, or to move the points of observation, i.e. to look at the totality of the tube from the points of observation.

According to another aspect of the invention, an apparatus for measuring the quality of a tubular body, said tubular body being movable along a feeding direction, said tubular body being transparent with respect to a predetermined radiation, comprising:

- a detection means and/or detecting the geometry of said tubular body, in particular a tube having a predetermined inner and outer diameter;

- a source of said radiation arranged in a first position next to said feeding direction, said source arranged to cause at least one first beam of said radiation and a second beam of said radiation to pass through said tubular body so that said first and second radiation beams are incident to each other according to a predetermined angle at said object;

- a radiation sensor arranged in a second position next to said feeding direction, and opposite to said first position with respect to said tubular body, said sensor arranged to measure said first radiation beam and said second radiation beam after that said first and second radiation beams have crossed said tubular body, said sensor arranged to provide a first and a second alteration signal (x, y) that occurs when said tubular body is crossed by said first and second radiation beams,

- a means for analysing said first and second alteration signals (x, y) in order to determine in a first cross section (Z-i) of said object :

- a position (r, Θ) of said alteration in said first cross section;

- a first coordinate of axial position (zi) of said alteration for said first cross section;

- a sampling means for causing said sensor and said means for analysing to operate in a repeating way according to a predetermined sampling frequency while said object moves along said feeding direction, said means for analysing arranged to determine, in a second, third, i^th, n^th cross section (Z₂, Z_3j ... , Z,, Z_n) of said tubular body, corresponding first and second alteration signals (x, y)₂.._n, as well as n positions (r, θ)₂.._η of said alteration, which are associated to respective coordinates of axial position (z₂, z₃, ¾, z_n);

- a means for determining a parameter of quality of said tubular body according to the radial position, angular and axial elongation of said alteration obtained from said means for analysing.

In particular, the sensor can comprise two linear arrays of optical sensors opposite to a respective first and second source.

Advantageously, said means for causing a first radiation beam and a second radiation beam to pass through said object comprises a first and a second source of radiation rotationally spaced from each other according to said angle, in particular said first and second source of radiation are a first and a second light source, in particular are lighting devices at high uniformity and highly collimated radiations.

Alternatively, said means for causing a first radiation beam and a second radiation beam to pass through said tube comprises a single source of radiation, in particular a light source, arranged to emit a source light beam and a means for directing selectively said source light beam according to a first direction, in order to provide said first light beam, and according to a second direction, in order to provide said second light beam.

In particular, said means for directing said source light beam comprises a means for rotating said single light source according to said angle.

In a further alternative, said means for causing a first radiation beam and a second radiation beam to pass through said hollow tube comprises a single source of radiation, in particular a light source, arranged to emit a source light beam and a means for rotating said object according to said predetermined angle between a first and a second position spaced by said angle a, wherein when said tube is in said first position said source light beam is called first light beam, and when said object is in said second position said source light beam is called second light beam. The rotation according to angle a is carried out with respect to a point between the light source and the sensor, and said point can also be located ideally in the transparent object.

In particular, said detection means comprises a first sensor and a second sensor spaced by said angle a, in particular said first and second sensors are video cameras, in particular linear video cameras.

Advantageously, said first and second video cameras are located so that the linear sensor inspects orthogonally and radially said object. In particular, in case of a cylindrical glass hollow object, said linear video cameras are located radially to the cylinder, orthogonally with respect to the longitudinal axis of the cylinder. In this case, the simulative model must simulate the movement of the radiation of the light source through the glass tube towards the linear video camera considering the geometric features of the cylinder and angle a formed by the two directions of observation the two video cameras that inspect the same portion of cylinder. The single radiations of the light sources cross two times the walls of the glass tube, and then are subjected to four refractions that change the original direction of the light radiations. For this purpose, a correction means is provided for determining the path of the radiations through the tube.

In particular, a plurality of video cameras can be arranged associated with respective light sources, in order to inspect corresponding portions of the tube. This way, it is possible to choose arbitrarily angle a at the centre of the cylinder formed both by the directions of observation of the video cameras and by the light sources. This can be chosen in order to maximize the resolution or to minimize the number of the video camera.

For optimizing the apparatus, the movement of the object or the video camera can be determined according to the preferences. Also a succession of views can be obtained, with an enhanced resolution.

For example, the configuration with angles at the centre of the cylinder equal to 45° with four video cameras allows observing the most common types of glass cylinders available on the market, and a good resolution. Such configuration provides four video cameras located at a same side of the cylinder or, alternatively, one of the two central video cameras at the opposite side of the cylinder. It should be noted that the method for inspection remains valid changing both the angle at the centre, and the number of axes, as well as the arrangement of the video camera on either of the two sides of the tube. Furthermore, the axes of the points of observation can be located also on different planes, for example along the longitudinal axis of the cylinder.

The simulative model makes it possible to determine for each position, where a first video camera observes the cavity, the interval of the positions where the recess can be seen by a second video camera. The extreme positions of such intervals correspond to points of the impurities located at the outer and inner surfaces of the glass cylinder.

Furthermore, said first and second light sources rotationally spaced from each other are a first and a second laser sources, in particular each of said first and second laser sources is configured to emit a single beam that hits a respective mirror element to obtain a variable incident beam arranged to scan a respective portion of said object, the combination between a first and a second variable incident beam allows scanning completely said object and measuring the position of the defect.

In particular, said detection means associated with said first and second laser sources comprises a first and a second couple of photosensitive diodes that detect the refraction of each of said first and second variable incident beam.

Advantageously, a first and a second array of laser radiations rotationally spaced from each other can be arranged, which are adapted to generate each a plurality of radiations parallel that embraces completely said object and an array of sensors opposite to said array of laser radiations. In particular, a simulation means is provided for simulating a path of refraction of said beams of radiation through said transparent object, in order to determine the actual position of said point in said transparent object.

Brief description of the drawings

The invention will be now shown with the description of an exemplary embodiment thereof, exemplifying but not limitative, with reference to the attached drawings, in which like reference characters designate the same or similar parts, throughout the figures of which:

Fig. 1 shows a diagrammatical view of an apparatus and of a method for measuring the geometric position of a critical point or of a contour in an object, according to the invention, according to a first exemplary embodiment with a video camera technology;

- Fig. 1A shows a diagrammatical view of an exemplary embodiment of the method of Fig. 1 ;

- Fig. 1 B shows a view of the methodology of recognition of the coordinates and of the contour of a defect in a cross section of a tubular hollow body;

- Fig. 2 shows a view of the methodology of recognition of the axial coordinates of a defect in several cross sections of a tubular hollow body;

- Fig. 3 shows a table of recognition of the three-dimensional profile of a defect;

- Fig. 4 shows a diagram of a possible implementation of the method for detecting defects in a hollow tube;

- Fig. 5 shows a perspective view and relative enlarged views 5A, 5B and 5C of a glass tube and of a variety of types of defect that can be identified through the apparatus of Fig. 1 ;

- Fig. 6 shows a diagrammatical view of a second exemplary embodiment that provides a single light source capable of rotating according to a predetermined angle;

Fig. 7 shows a diagrammatical view of a further exemplary embodiment that provides two light sources and a single sensor arranged to rotate between a first position and a second position rotationally spaced from each other;

- Figs. 7A and 7B show a diagrammatical view of a further exemplary embodiment that provides two light sources and a single fixed sensor and a means for rotating the object between a first and a second position rotationally spaced from each other;

Fig. 7C shows a diagrammatical view of an exemplary embodiment that comprises a single light source and a single video camera, both movable between a first and a second position, in order to observe the defect from two different points spaced of a predetermined angle from each other; Fig. 8 shows a diagrammatical view of the apparatus for measuring the geometric position of a defect in an object in a fourth exemplary embodiment that adopts radiation source with laser technology with photosensitive diodes;

- Fig. 9 shows a diagram for controlling data coming from the video camera or photodiodes and for determining the position and the dimensional characteristics of a critical point or of a contour observed in the object; finally, Figs. 10, 1 and 12 show a diagrammatical view of a further exemplary embodiment that adopts more video cameras and respective sources, in order to maximize the resolution of inspection.

Description of preferred exemplary embodiments

With reference to Fig. 1 , a first exemplary embodiment is shown of an apparatus and a method for measuring the quality of a tubular body 1 movable along a conveying direction Z. Tubular body 1 is transparent with respect to a predetermined radiation 10, 20. The quality of the object is responsive to the presence in the walls of an alteration to the movement of the light caused by a defect or impurities 70.

In particular, the apparatus comprises a supply means 100 for causing 1 at least one first radiation beam 10 and a second radiation beam 20 to pass through tubular body, so that the first 10 and beam second beam 20 of radiation form to each other a predetermined angle a.

In particular, according to a first exemplary embodiment, if the object is of glass, the first 10 and second 20 radiation beams are light beams obtained from respective light sources 110 and 120, with directions rotationally spaced from each other according to an angle a.

The apparatus further comprises a detection means 200 for measuring first light beam 10 and second light beam 20 after that they have crossed object 1. Both light beams cross the curved surfaces of the tubular body two times each, being subjected to four refraction effects. Such refraction effects cause a factor of noise that is corrected by means of known mathematical algorithms, avoiding that the radial and angular positioning is affected by noise.

In particular, the detection means 200 is configured to measure a light alteration (x, y) that occurs when the first 10 and second 20 light beams cross the object and to provide a first 150 and a second 160 alteration signal in case this alteration exceeds a predetermined threshold.

More precisely, for alteration a predetermined value of light intensity is intended that is detected by detectors 200. This implies that the light beam has been deviated by a critical point, like a glass defect 50, such that the detector detects a light intensity less than the predetermined value, as described below in detail.

In Fig. 1 , it is supposed that the defect is of small size that can be approximated to a point 50, and that then it determines an alteration signal in a single critical point X and Y, respectively on detection means 210 and 220.

In particular, point 50 has a position X, which is compared with acceptable positions, where the critical point or contour is within the object, and two limit positions, where the defect is located respectively on the inner surface 1a or external surface 1 b of the tube.

In particular, first light source 110 is configured to emit first light beam 10 incident on object 1 and a first detection sensor 210 is arranged along a first plane diametrically opposite to first light source 110. In particular, first light sensor 210 is adapted to generate the first alteration signal due to a faulty exit from object 1 , in order to determine a first reference position X of defect 70. In a same way, second light source 120 is configured to emit second light beam 20 and a second light sensor 220, which is arranged along a second plane diametrically opposite with respect to second light source 120, generates second alteration signal Y.

The apparatus also comprises a program means 300 associated with the means for detecting 200 and arranged to analyse first alteration signal 150 and second alteration signal 160, in order to determine, starting from the known position of detection sensors 210, 220, a geometric position of point 50, through a simulative model shown in Fig. 9.

More precisely, computer 300 receives as input first 150 and second 160 light alteration signals and, by knowing the angle of incidence of beams 10/20 in addition to the geometric position of the two sensors 210, 220, executes a triangulation from which it is possible to determine the geometric position of the defect in object 1 . This way, it is possible to discriminate defects embedded in the thickness of the object, or defects close to its inner surface a or external surface 1 b. More precisely, as shown in Fig. 5, defect 70 can be positioned in the thickness 1 c (Fig.5A) or on the surface of object 1 , in particular on the inner surface 1 a (Fig.5B) or on the outer surface 1 b (Fig.5C). In one of these two extreme cases, the defect can represent a crack capable of triggering a fracture, causing the tube to be extremely vulnerable, subjected to an easy break to a minimum action, with subsequent hazards (for example, break of a vial or ampoule of glass. Alternatively, the defect that emerges on the surface can determine a loss of tightness, for example

Structurally, according to a first exemplary embodiment, as shown in Figs. 1 and 1A, first sensor 210 and second 220 are video cameras, in particular linear video cameras, whereas first 1 10 and second 120 light sources are lighting devices at high uniformity and highly collimated radiations. This way, the critical point or contour is identified by the first video camera 210 as an alteration of the light beam 10 at a point Po used as first reference, second video camera 220 is used, instead, for determining the geometric position of the critical point or contour in the wall and deducing then its three- dimensional position coordinates through the alteration of second light beam 20. More precisely, second video camera 220 can see this alteration and then defect 70 in limit positions that correspond respectively to the position of defect 70 at the inner surface 1 a of object 1 or at the outer surface 1 b of the object.

In figures 1A and 1 B a diagrammatical view is shown of an exemplary embodiment of the method of Fig. 1 , in which it determines a linear contour Dx and Dy of defect 50, if it has a not point-like shape. In Fig. 1 B a view is shown of the methodology of recognition of the coordinates and of the contour of a defect in a cross section of a tubular hollow body.

Fig. 2 shows a view of the methodology of recognition of the axial coordinates of a defect in several cross sections of the tube, which are indicated as Z. By measuring the first radiation beam and the second radiation beam, by sensor 200, after that the first and second radiation beams have crossed tubular body 1 , the measurement step provides the measurement of an alteration which occurs when the first and second radiation beams cross tubular body 1 , supplying a first and a second alteration signal (x, y) in case the alteration exceeds a predetermined threshold. Then, the first and the second alteration signals (x, y) are computed in order to determine, in a first cross section (Zi) of the object a position (r, Θ), the alteration in the first cross section and a first coordinate of axial position (z-i) of the alteration in the first cross section, then the measurement step is repeated according to a predetermined sampling frequency and an analysis step is effected along the conveying direction in order to determine, in a second, third, i^th, n^th cross section (Z₂, Z₃, Z_h Z_n) of the tubular body of the corresponding first and second alteration signals (x, y)₂.._n, as well as n positions (r, θ)₂.._η of the alteration, which are associated to respective coordinates of axial position (z₂, z_3i z_u z_n). Finally, a parameter of quality of the tubular body is determined according to the radial position, angular and axial elongation of the alteration.

An example of the described method can be given also as hereafter reported. After choosing two points of observation A and B (not shown), the centre of the glass tube is detected. The point of observation A sees a signal alteration between X1 and X2 with respect to the tube centre. Then, the other signal alteration is seen from the second point of observation B between Y1 and Y2 with respect to the tube centre. The interval ΔΥ relative to the alteration signal from Y1 to Y2 changes responsive to the position of the cause of the alteration signal which can be present in the wall of the tube: DY can change from a possible interval DYI corresponding to a glass defect located at the outer surface of the cylinder, to an interval AYF corresponding to a glass defect located at the inner surface of the cylinder.

In particular, the values among which Y1 and Y2 can be chosen can be calculated with a simulative model or experimental tests, either on-line or offline, for reproducing the geometric features of the hollow tube to inspect and the behaviour of the beam responsive to the features of the tube. Such simulation allows determining for each interval of altered signal on A all the relative possible DY intervals on B responsive to the position of the impurity that has altered the signal, as well as responsive to the outer and inner diameter of the tube.

The intervals DX and DY are calculated instant-by-instant in accordance to the sampling frequency of the inspection device. The acquisition of all the intervals DXn and DYn makes it possible to determine two contours that show the shape of the impurities from the two different points of observation A and B. However, taken singularly the two contours are not a three-dimensional representation of the shape of the impurity that has caused the two alterations of the signal from the two points of observation.

The first and the second contour allow calculating a space in which the impurity is certainly present, but not its exact three-dimensional shape.

The numerical values of the two limit positions, responsive to the position Po seen from first video camera 210, are computed by computer 300 comprising dedicated program means that compute the signals following the diagram of Fig. 9. In other words, the limit positions defining an interval in which second video camera 220 can get the impurities, responsive to the point Po where first video camera 210 finds impurity 70. The obtained intervals are used for setting the value beyond which impurity 70 is considered located at the inner wall 1 a or at the outer wall 1 b of the cylinder.

The program means compute signals 150/160, in order to include the effects of the refraction of the radiations on the glass and to take into account that the optics on the video camera 210, 220 focuses objects in the focal plane, whereas the more the distance from the focal plane the more the objects are out of focus. The program means must then consider that the light beams cross two curved surfaces and are then deflected four times, by introducing then a factor of error that has to be balanced in the simulative model in order to have measurements not significantly affected by error.

In particular, for calculating the position of a shape of a defect detected in the cylinder a step is provided of detection and/or acquisition of the geometry of object 1. For example, for a tubular body with cylindrical cross section, as shown in Fig. 1 , it is necessary to measure/detect the diameter of the outer circumference d_est and the diameter of the inner circumference d_int, from which the actual thickness s of the wall of the tube is evaluated. This way, chosen the two points of observation, i.e. the directions of first 10 and second 20 light beam, it is possible to determine the centre O of the glass tube. In particular, the first point of observation sees the impurities as a deviation of the signal that can evidence a single critical point or a contour, i.e. a set of critical points connected to each other, at a relative distance from the tube centre. Then, the alteration signal is observed on the same portion of the objects inspected from the second point of B at a distance from the tube centre. The distance changes responsive to the position of the critical point, or of a contour, within the wall of the tube: the distance changes between a first limit value corresponding to the impurities located at the outer surface of the cylinder, to a second limit value, corresponding to the impurities located at the inner surface of the cylinder.

In particular, the limit values can be calculated with a simulative model that reproduces the geometric features of the object to inspect and of the radiation beams and allows determining for each point the relative limit values responsive to the geometry of the object.

Furthermore, in addition to setting the interval at least one safety threshold can be added in such interval for narrowing it, or for adding more or less strict quality criteria. This way, the step of positioning the detected contours can be adapted to a desired rate of quality, with respect to particular positions of the defect in the wall of tube 1. For example, a cylinder that has a closed cavity defect that is located near the inner 1 a or external 1 b surface of the cylinder can be considered as a product to reject. This allows rejecting objects with a closed cavity defect very close to the surface less than a predetermined value. In fact, such a closed cavity can turn into a open cavity by a mechanical stress.

The positioning step of the defect can be carried out considering a succession of said points, which define the overall shape of the defect.

In particular, for setting the video camera 210, 220 and to set up the apparatus for inspection it is enough carry out a simulation for each type of cylinder to inspect, responsive to the outer diameter d_est and to the thickness s of the cylinder, as well as to the type of glass and to the type of lighting.

In a second exemplary embodiment, diagrammatically shown in Fig. 6, the means 100 for supplying a first light beam and a second light beam comprises a single light source 130, arranged to emit a source light beam 30 and a means for directing selectively the source light beam 30 according to a first direction, in order to provide the first light beam 10, and according to a second direction, in order to provide second light beam 20. In particular, the means for directing the source light beam 30 comprises a means for rotating (not shown) the light source 130 according to an angle a.

An alternative solution, diagrammatically shown in Fig. 7, provides the two light sources 1 10, 120 and a single sensor 230 arranged to rotate between a first position C and a second position D spaced by angle a. During the rotation the tube does not move. Such solution can provide, alternatively, as shown in Figs. 4A and 4B, keeping fixed the single sensor 230 and rotating object 1 by angle a. In both cases, it is possible to determine the position of defect 70 through a single point of observation by rotating point of observation 230 by a degrees about object 1 , or causing the object 1 to rotate by angle a with respect to the point of observation 230.

Even alternatively, as shown in Fig. 7C, a single mobile light source 130 is provided, and a single sensor 230 also movable between a first position A and a second position B. During the rotation the tube is still. In this step, several acquisitions with small angles can be made for increasing the precision of the inspection and reducing the margin of approximation. This way, in first position A, the source light beam is first light beam 10, whereas in second position B, the source light beam is second light beam 20;

Fig. 8 shows a further exemplary embodiment of the apparatus made with laser technology. In particular, this exemplary embodiment provides a first laser 140 and a second laser 150 arranged to scan linearly the object within a known time, as shown with reference to laser 140, and a couple of respective photosensitive diodes 240/250 located at opposite sides of the cylinder, in order to capture the laser radiations refracted by the object, in detail cylinder 1. In particular, each laser source 140/150 emits a respective beam source 142/152, which hits a respective mirror element 145/155, in detail a oscillating mirror, in order to generate a movable output beam 146 arranged to scan the cross section of cylinder 1 from wall to wall. In detail, Fig. 8 shows, only relatively to mirror 145, a plurality of radiations that show the variation of incidence of the output laser beam 146 coming from mirror 145. The scanning action of output movable beam comes from mirror 155. More in particular, the oscillation frequency of each moving mirror 145/155 determines the scanning frequency of laser source 140/150. When the beam scans a portion of cylinder, the radiations are in part refracted towards one or both photodiodes 240 according to a known pattern. If the expected pattern on the photosensitive diodes varies, it means that a glass defect 70 has been found.

In particular, when photodiodes 240/250 not receive the light signal, coming from laser 140/150, it means that the beam has been deviated by a defect 70. This way, on the basis of the frequency of oscillating mirror 145/155 it is possible to determine a first direction of the beam that has hit a defect. In a same way, the direction is measured of the second beam coming from the second laser source 150. By knowing such directions and their angle of incidence, as well as the position of photodiodes 240/250, it is possible to determine then the position of defect 70 in object 1.

The photodiode laser scanning technology allows checking the defect like the video camera technology described above. By providing a same algorithm according to the block diagram of Fig. 9, it is therefore possible to trace three- dimensionally the shape of the contour (i.e. defect 70) if it is observed by couple of photodiodes 240/250, in known position, after having hit the object and spaced angularly by known angle a.

Alternatively to the moving mirrors 145/155 a first and a second array of highly collimated laser radiations can be used that generate each a plurality of radiations parallel to each other that embrace from wall to wall object 1 , and that then carry out the same function of a movable light beam at the exit from mirrors 145/155. Even in this case, each array of laser radiations is rotationally spaced from each other by angle a. The sensor is in front of it and has an array of sensors.

Relatively to the video camera detection technology, in a preferred exemplary embodiment, as diagrammatically shown in Figs. 7, 8 and 9, a plurality of video cameras can be arranged associated with respective light sources, in order to inspect corresponding portions 1c, 1d, 1e, 1f of said object 1. This way, it is possible to choose arbitrarily an angle a formed by the two directions of observation by the video cameras. This can be chosen in order to maximize the resolution or to minimize the number of video cameras.

For example, the configuration with angles at the centre of the cylinder equal to 45° allows observing the most common sizes available on the market of glass cylinders 1 with four video cameras A, B, C and D and respective light sources A', B', C and D' obtaining a good resolution. In particular, as shown in Figs. 7 and 8, the portions of cylinder identified by a same symbol are observed by a same couple of video cameras. Such configuration, sees the four video cameras A, B, C and D located at a same side of the cylinder (Fig.10) or, alternatively, sees a video camera B or C positioned at the opposite side of cylinder 1 (Fig.11). It should be noted that the method for inspection remains valid changing both angle a, and the number of axes, as well as the arrangement of the video camera on either of the two sides of cylinder 1. Furthermore, the axes of the points of observation can be located also on different planes, for example along the longitudinal axis of the cylinder. The simulative model makes it possible to determine for each position, where a first video camera observes the cavity, the interval where the same recess can be seen by a second video camera. The extreme limits of such intervals correspond to a glass defect located at the outer and inner surfaces of the glass cylinder.

It is useful to specify that sensors and source of electromagnetic signals (radio, microwaves, infrared, visible, UV, X-ray, gamma radiations) can be used in function also of the type of material. Preferably, the best radiation and relative sensor types are those within the visible wavelength range.

The foregoing description of an embodiment of the method and of the apparatus according to the invention, and of the way of using the apparatus, will so fully reveal the invention according to the conceptual point of view, so that others, by applying current knowledge, will be able to modify and/or adapt for various applications such embodiment without further research and without parting from the invention, and, then it is therefore to be understood that such adaptations and modifications will have to be considered as equivalent to the specific embodiment. The means and the materials to realise the different functions described herein could have a different nature without, for this reason, departing from the field of the invention. It is to be understood that the phraseology or terminology the is employed herein is for the purpose of description and not of limitation.

Claims

- 1 - CLAIMS

1. A method for measuring the quality of a tubular body (1), said tubular body (1) being movable along a feeding direction, said tubular body (1) being transparent with respect to a predetermined radiation, said method comprising the steps of:

- measuring and/or detecting the geometry of said tubular body (1), in particular a tube having a predetermined inner and outer diameter;

- causing at least one first beam of said radiation and a second beam of said radiation coming from said source to cross said tubular body (1) so that said first and second radiation beams are incident to each other according to a predetermined angle at said object;

- arranging in a second position next to said feeding direction, and opposite to said first position with respect to said tubular body (1 ), a radiation sensor;

- measuring said first radiation beam and said second radiation beam, by said sensor, after that said first and second radiation beams have crossed said tubular body (1), said measuring step determining an alteration which occurs when said tubular body is crossed (1) by said first and second radiation beams, said measuring step providing a first and a second alteration signal (x, y) in case said alteration exceeds a predetermined threshold,

- analysing said first and second alteration signals (x, y) in order to determine in a first cross section (Zi) of said object :

- a position (r, Θ) of said alteration in said first cross section;

- repeating according to a predetermined sampling frequency said measuring and analysing step along said conveying direction in order to determine, in a second, third, i^th, n^th cross section (Z₂, Z₃, ... , Z,, Z_n) of said tubular body (1), the corresponding first and second alteration signals (x, y)₂.._n, as well as n positions (r, θ)_2..η of said alteration, which are associated to respective coordinates of axial position (z₂, z_3, ... , z,, - 2 - z_n);

- determining a parameter of quality of said tubular body (1) according to the radial position, angular and axial elongation of said alteration.

2. A method, according to claim 1 , wherein said sensor (210, 220) comprises a linear array of optical sensors, and said first and second alteration signals (x, y) consist of a first and a second interval of said pixels (Dx, Dy) of said linear array of optical sensors that receive an altered radiation

3. A method, according to claim 2, wherein said step of analysing said first and second alteration signals (x, y) provides the control of said first and second intervals of said pixels (D_x, D_y) of said linear array of optical sensors that receive an altered radiation in order to determine in each cross section of said object a position interval (η, r₂; θι, θ₂) of said alteration in said cross section.

4. A method, according to claim 3, wherein a step is provided of determining along said conveying direction an approximated three-dimensional shape of said alteration.

5. A method, according to claim 2, wherein a step is provided of approximating a shape of a defect by a step of comparing said first and second intervals of said pixels (D_x, D_y) for each cross section with known geometric parameters of defects or impurities.

6. A method, according to claim 1 , wherein said tubular body (1) is a tube and said step of detecting and/or acquiring the geometry of said tube comprises a step of detecting the edges of said tube and a step of detecting the centre of said tube, in particular said step of detecting the edges of said tube allows determining the outer diameter and the inner diameter of said tube.

7. A method, according to claim 6, wherein a step is provided of rejection of the forces to which said object is subjected along said feeding direction, said step of rejection comprising a step of repeating said step of detecting the centre of said tube instant-by-instant, in order to eliminate position artifacts of said tube.

8. A method, according to claim 1 , wherein a step is provided of correction of said first and second alteration signals responsive to refractions that occur when the radiation crosses said tubular body (1).

9. A method, according to claim 1 , wherein said step of determining a parameter - 3 - of quality provides a signal of feedback to a production step of said transparent object in order to use the features of said alteration for adjusting the process for forming said object.

10. A method, according to claim 1 , wherein said step of causing a first radiation beam and a second radiation beam to pass through said object comprises a step of arranging a first and a second source of radiation rotationally spaced from each other according to said predetermined known angle a.

11. A method, according to claim 1 , wherein said step of causing a first radiation beam and a second radiation beam to pass through said object comprises a step of arranging a single source of radiation to emit an emitted radiation beam and a step of selectively directing said emitted radiation beam according to a first direction, in order to provide said first radiation beam, and according to a second direction, in order to provide said second radiation beam, in particular said step of directing said emitted radiation beam comprises a step of rotating said single source of radiation according to said angle a.

12. A method, according to claim 1 , wherein said step of causing a first radiation beam and a second radiation beam to pass through said object comprises a step of transmission of said emitted radiation beam and a step of rotating said object according to said angle set between a first and a second position spaced by said angle, wherein when said object is in said first position said emitted radiation beam is called first radiation beam, and when said object is in said second position said emitted radiation beam is called second radiation beam.

13. A method, according to claim 1 , wherein said step of measuring is selected from the group consisting of: a step of:

• prearranging a first sensor and a second sensor spaced by said angle;

• prearranging a single sensor movable between said first and second position spaced by said angle.

14. An apparatus for measuring the quality of a tubular body (1), said tubular body (1) being movable along a feeding direction, said tubular body (1) being transparent with respect to a predetermined radiation, comprising: - 4 -

- a detection means for detecting the geometry of said tubular body (1 ), in particular a tube having a predetermined inner and outer diameter;

- a source of said radiation arranged in a first position next to said feeding direction, said source arranged to cause at least one first beam of said radiation and a second beam of said radiation to pass through said tubular body (1 ) so that said first and second radiation beams are incident to each other according to a predetermined angle at said object (1 );

- a radiation sensor arranged in a second position next to said feeding direction, and opposite to said first position with respect to said tubular body (1 ), said sensor arranged to measure said first radiation beam and said second radiation beam after that said first and second radiation beams have crossed said tubular body (1), said sensor arranged to provide a first and a second alteration signal (x, y) that occurs when said tubular body is crossed (1) by said first and second radiation beams,

- a means for analysing said first and second alteration signals (x, y) in order to determine in a first cross section (Zi) of said object :

- a position (r, Θ) of said alteration in said first cross section;

- a first coordinate of axial position (z-i ) of said alteration for said first cross section;

- a sampling means for causing said sensor and said means for analysing to operate in a repeating way according to a predetermined sampling frequency while said object moves along said feeding direction, said means for analysing arranged to determine, in a second, third, i^th, n^th cross section (Z₂, Z3, ... , Z,, Z_n) of said tubular body (1 ) the corresponding first and second alteration signals (x, y)2.._{n <} as well as n positions (r, θ)2.._η of said alteration, which are associated to respective coordinates of axial position (z₂, z_3, ... , z,, z_n);

- a means for determining a parameter of quality of said tubular body (1 ) according to the radial position, angular and axial elongation of said alteration obtained from said means for analysing.

15. An apparatus, according to claim 14, wherein said sensor comprises two linear arrays of optical sensors opposite to a respective first and second source.

16. An apparatus, according to claim 14, wherein said means for causing a first - 5 - radiation beam and a second radiation beam to pass through said object are selected from the group consisting of:

- a first and a second source of radiation rotationally spaced from each other according to said angle, in particular said first and second source of radiation are a first and a second light source, in particular are lighting devices at high uniformity and highly collimated radiations;

- an single source of radiation arranged to emit a source radiation beam and a means for directing selectively said source radiation beam according to a first direction, in order to provide said first radiation beam, and according to a second direction, in order to provide said second radiation beam, in particular said means for directing said source radiation beam comprises a means for rotating said single source of radiation according to said angle;

- a single source of radiation arranged to emit a radiation beam source and a means for rotating said object according to said predetermined angle between a first and a second position spaced by said angle a, wherein when said object is in said first position said radiation beam source is called first radiation beam, and when said object is in said second position said radiation beam source is called second radiation beam.

17. An apparatus, according to claim 14, wherein said detection means comprises a first sensor and a second sensor spaced from each other by said angle a, in particular said first and second sensors are video cameras, in particular linear video cameras, in particular said first and second video cameras are located orthogonally with respect to said object, in particular said detection means comprises a plurality of video cameras, in particular linear video cameras, rotationally spaced from each other, each video camera being associated with a respective light source, in order to inspect corresponding portions of said object.

18. An apparatus, according to claim 14, wherein said first and second light sources rotationally spaced from each other are a first and a second laser sources, in particular each of said first and second laser sources is configured to emit a single beam that hits a respective mirror element to obtain a variable - 6 - incident beam arranged to scan a respective portion of said object, the combination between a first and a second variable incident beam allows scanning completely said object and measuring the position of the defect, in particular said detection means associated with said first and second laser sources comprises a first and a second couple of photosensitive diodes that detect the refraction of each of said first and second variable incident beams, in particular a first and a second array of laser radiations can be arranged rotationally spaced from each other, to generate each a plurality of radiations parallel that embraces completely said object, said first and second array of laser radiations being associated with respective photodiode sensors.

19. An apparatus, according to claim 14, wherein a simulation means is provided for simulating a path of refraction of said beams of radiation through said transparent object, in order to determine the actual position of said defect in said transparent object.