FR2644250A3 - Defect-identification device for identifying defects on moving material portions - Google Patents

Abstract

The defect-identification device comprises: a light source located on one side of the material portion, an optical system containing a rotary wheel with mirrors, this being located on the same side or on the other side of the material portion, a photoelectric converter which is placed on the path of the light rays reflected by the mirrors of the wheel and a processing circuit connected to the output of the photoelectric converter. It includes a Fresnel lens 3, a beam splitter 6, a diaphragm 8A, 8B and an optical filter 7A, 7B in each path of the split beam. Electronic processing circuits 100 allow the various types of defect to be distinguished by processing the various signals coming from the photoelectric converters 9A, 9B.

Description

DEFECT IDENTIFICATION DEVICE FOR IDENTIFYING
DEFECTS ON PARTS OF MOVING MATERIAL
The invention relates to a defect identification apparatus for identifying faults on moving parts of material such as, for example, ribbon objects or bottles, comprising: a light source situated on one side of the object , an optical installation containing a rotating mirror wheel located on the same side or the other side of the material part, a photoelectric converter which is placed on the path of the light rays reflected by the wheel mirrors and a circuit which is connected to the output of the photoelectric converter.

 Such installations are used, for example, during quality control during the manufacture of textiles. Devices of this type can be used, for example, in textiles manufactured in the form of strips, in which longitudinal defects due for example to the breaking of a warp wire can be identified, also transverse defects due, for example to breaking. weft thread.

 The problem of identifying defects in strips of moving material is not only important for the manufacture of textiles, but also for example during the manufacture of metal strips, sheets (plastic or metal), plates (plastic or metal), tablecloths and woven glass (in particular those used in the electronics industry), paper and tufted objects (carpets), In the usual quality control devices in the textile industry, work is usually carried out in through light, that is to say that a device is used in which a light source is located on one side of a strip of material, while on the other side, there is an optical installation followed by an evaluation (default) installation. On the contrary, in the quality control of metal strips, we work in reflected light. The present invention relates in particular to an installation working in through or transmitted light.

 In the area of textile manufacturing, there has been an increasing use of circular knitting machines recently. These automatic knitting machines are equipped with several thousand needles which make a knitting having the shape of a tube. The tubular material produced by the machine moves in front of the machine in the longitudinal direction while being rotated around the longitudinal axis. Given the very high cost of this equipment, it is imperative that downtime be as short as possible and, to do this, it is not only essential to recognize faults as quickly as possible, but also to differentiate them. each other, because not all faults require stopping and repairing the machine. On the other hand, it should be avoided that before having recognized a fault, for example due to a broken knitting needle, a large amount of defective, unusable knitted material is produced.

 Defects occurring with circular knitting machines are typically so-called longitudinal defects, i.e. those caused by defects in the circular direction with respect to the material (and which have an annular shape or spiral). There are also transverse defects, that is to say defects which are caused by a defective knitting needle, and in addition point defects, which are either buttons (extra thickness) or small holes.

 Document DE-A-2 925 734 describes an installation for identifying faults which are of different types, using the technique which has just been described below. In order to be able to distinguish faults of different types, it is proposed in this document to equip a rotating mirror wheel with two types of alternately active reflecting surfaces. Each type of reflecting surface produces a spot of well-defined shape in the plane of exploration (that is to say the plane on which the defects of the strip material are examined).

 In addition to the fact that it is very expensive to have to prepare mirrors which are specially adapted for each examination of the defect, it is also necessary to design a processing circuit which is specific to each type of mirror.

 From document US-A-3 797 943, an installation is known for identifying the defects of moving flat materials. It uses a flat diaphragm provided with three parallel diaphragm slots of different lengths, the images of which are projected onto different photodiodes. In this way, the strip of material is swept over its entire width, using the three diaphragm slots.

 The above devices allow the identification of defects in strip materials moving only in their longitudinal direction and non-homogeneously in the direction of their width. Indeed, the sensitivity of the fault identification process changes depending on whether one is in the middle or at the edge of the strip of material. The identification in the center is more precise than on the edge.

 The guidance of the material during its passage must be excellent, since there is no depth of field for the examination. The slightest flutter of the material to be examined immediately causes a measurement error. The lack of depth of field comes from the fact that one does not use parabolic mirrors, nor Fresnel lenses.

 Recognition of transverse faults is not possible. In fact, transverse faults only cause a difference in level around a reference point in the width of the overall signal used for identification.

 Too little or too much illumination of the entire width of the material (transverse defect) causes a variation in the DC voltage over the entire scanning width. Such variations in DC voltage are then eliminated using a differentiator circuit (FIGS. 5 (a), 5 (b) of document US-A-3,797,943).

The object of the invention is to propose an apparatus of the above type using only very simple means and which, however, in its multiple aspects, can be used for the identification of faults of all kinds on different type parts. of materials
To this end, according to the invention, the identification device comprises a beam splitter which divides the light beam coming from the mirror wheel into at least two partial beams and on each partial beam path is arranged a diaphragm and / or an optical filter which differs respectively from the diaphragm or from the optical filter situated on the other beam path by its characteristics of aperture shape or filtering respectively.

 To implement the device according to the invention, it is possible to work with a common and simple mirror wheel.

The distinction between the types of defects is made by separation of the optical signals into two separate beams of rays.

A slit diaphragm can be placed in the transverse direction along the path of a partial beam. On the other path of the partial beam; one can have a slot acting as a diaphragm in the longitudinal direction. The image of the two slit diaphragms is in the plane of the strip of moving material, and it is in the form of a cross spot light. Thanks to these provisions, the device according to the invention can be used especially for circular knitting machines. Besides, there is no reason to limit the invention to such a field of application. It is also applicable in the case of looms.

 Apart from the relatively cheap polygonal mirror wheel, the other elements of the apparatus according to the invention are constituents which are easily found on the market and which are inexpensive. Thus, for example, the beam splitter can for example be a current mirror with partial transmission, which is known in various fields of application of optics.

 In order to judiciously take into account the different types of faults, partial beams are associated with each path, an amplification / filtering circuit. The operating parameters of this amplification / filtering circuit are chosen so that the processing of the signals is carried out for each of them under the best conditions. The amplification / filtering circuit can for example include a bandpass filter in order to filter different domains of parasitic frequencies. It is then possible to provide one or two low-pass filters which allow only the frequencies of the spectrum domain to pass, for which a real identification of the faults can be made.

 In particular, it is provided according to the invention that the output signals from the amplification / filtering assembly are transmitted to a multiplier. The signal from this multiplication device is characteristic of long transverse faults. Such transverse defects are due to bad knitting needles in the case of circular knitting machines, and to a faulty operation of the weft yarn dispensing device in the case of fabrics.

 By well-designed differentiation and subtraction operations of the two signals from the two separate beams of rays, it is possible to achieve that the parasitic signals cancel each other out and that only the useful signals add up.

 A great advantage of the present invention lies in the fact that it is possible to use as a light source a fluorescent tube common on the market, which is operated in alternating current with a frequency of about 50 kHz.

All known fault identification devices nowadays work with a laser light source. However, when a laser light source is used, in fact a monochromatic light is used which is not well suited for materials in color or on which there are drawings. On the other hand, a fluorescent tube as used according to the present invention makes it possible to have a wide useful spectral width, and it can be obtained at a very low price.

 It has been shown that when an alternating voltage source which has a sufficiently high frequency is used, the disturbing alternative components which usually disturb the emission do not manifest themselves. In other words, for a sufficiently high working frequency, the light tube behaves like a light source supplied with direct current.

 The identification of transverse defects (needle scratch, breakage of the weft thread, split weft) is based on the comparison of two superficial explorations or scans (longitudinal and transverse diaphragms), which are carried out in total synchronism in space and in time and which, however, require a strictly parallel passage of the transverse defects through the field of exploration. This parallelism of the transverse defects is not obtained in practice for strip widths which go up to 2000 mm. Tensions within the strip material in fact cause a crescent-shaped deformation of the transverse defects.

A preferred embodiment of the invention, to solve this problem, consists in combining, over the entire width of the material, individual fault identification systems according to the invention, each working on a small exploration area. The width of exploration of such a theme system preferably has a value of 150 mm × 15 mm, so that over a width of 150 mm, it is possible to identify a transverse defect having a gradient of 15 mm from across the exploration area
The invention can also be adapted in sorting installations, in order for example to sort different objects according to their color. To this end, provision is made to identify the different colors of the objects using different color filters placed in the two partial beam paths to obtain, for example, a criterion for the removal of objects of a certain color. In particular for recycling bottles, the fault identification device according to the invention is economically well suited. By scanning the individual bottles in the longitudinal direction, it is possible to obtain a logic signal which is, for example, characteristic of bottles of green, brown or white color. Using these signals, bottles of a specific color can be collected.

 US-A-4,076,979 describes an automatic installation for automatic discrimination between brown bottles and green bottles. In this device, the light sent onto the surface to be explored is divided by a semi-transparent mirror. One part reaches a first photocell via a transparent green filter, and the other part of the light beam reaches a second photocell via a transparent brown filter. However, the exploration of surfaces is done, in this case, statically and punctually. Therefore, an exact alignment according to the size and shape of the bottles must be carried out.

This means that the bottles must be roughly sorted by hand beforehand.

 Optical exploration according to the invention is carried out, however, preferably over an overall height of 150 mm and over a width of 15 mm. It follows that no preliminary sorting is necessary and that the efficiency of the process does not depend on the size or the shape of the bottles.

 In addition, an identification can be carried out on the basis of the trichromatic principle, by providing a downstream processing circuit.

 While in the installation according to the American patent, the identification is distorted or made impossible by the presence of punctual objects such as labels, inscriptions applied hot or lines, these parasitic influences are on the contrary completely eliminated thanks to the exploration mode and to the associated processing circuit implemented in the fault identification apparatus according to the present invention.

Other objects, advantages and characteristics will appear on reading the description of an embodiment of the invention, given without limitation, and with reference to the appended drawing where:
- Figure 1 shows a partial schematic view of a device for identifying faults in a moving strip of material;
FIG. 2 is a detailed synoptic view of the amplification / filtering and signal processing circuits shown diagrammatically in FIG. 1,
- Figure 3 shows a variant of the processing circuit shown in Figure 2;
- Figure 4 is a schematic sketch explaining the various defects that occur in a strip material; and
- Figure 5 shows by way of example a fault video signal emitted by the device.

 We will first refer to Figure 4, which shows a section of a strip of tubular material. It is. woollens produced using a circular knitting machine. In products made with this kind of technique, the tube or sleeve produced moves in a spiral as shown by the arrow. The defects have different shapes depending on their origin: they can be defects from the breakage of one or more knitting needles, yarn defects, etc. I1 exist transverse defects Q, which are transverse with respect to the direction of movement of the strip of material 1, defects L, running parallel to the direction of movement, extra thicknesses (buttons) D, and also holes H which are present as clear areas.

 The tube-shaped strip material 1 is shown in Figure I in the upper left corner. It moves in the longitudinal direction along the arrow PL, and also in a circumferential direction corresponding to the arrow PR.

As a result, a given point located on the strip material 1 moves on a spiral path.

 Inside the strip of material, leaving a circular knitting machine which is not illustrated here, is a light source, constituted by a fluorescent tube 2 of the trade and which is supplied with high frequency alternating current at a frequency of 50 kHz.

The light (white) coming from the luminescent tube 2 falls on the interior surface of the material 1. It crosses it and ends in a Fresnel lens 3 placed outside the strip of material, and which focuses the light in a good place determined on a facet F of a polygonal mirror wheel 4 which rotates in the direction indicated by the arrow
Ps 6
While the facet F of the mirror wheel 4 touched by the focused light beam rotates by a value corresponding to a facet length, a line in the plane of exploration on the material 1 is scanned in correspondence with the angular displacement. The light reflected by the facet F of the mirror wheel 4 takes the direction indicated by the arrow PL towards the beam collecting lens 5.

This light corresponds at all times to a luminous point crossing in a very precise location the material 1 along the scanning line.

 After passing through the collecting lens 5, the light ends at a beam splitter 6. The latter separates the beam into two more or less equal parts; one part follows the beam path A, while the other is deflected on the beam path B.

The partial light beam path A comprises an optical filter, for example a green color filter 7A, and a slit diaphragm 8A. Light from the slit diaphragm
A falls on a photodiode 9A serving as a photoelectric converter which transforms light into an electrical signal.

The output signal from the photodiode 9A leads to an amplification / -filtration circuit 10A.

 The path of the partial light beam B is analogous to the path A and includes a color filter 7B, which for example is also a green color filter. The path also includes a slot diaphragm 8B and a photodiode 9B at the output of which is connected an amplification / filtering circuit 10B. The slots of the diaphragms 8A and 8B are oriented in such a way that, in the scanning plane of the strip material 1, their image defines an exploration spot in the shape of a cross 11. The output signals of the amplification circuits / filtering 10A and lOB are brought to an electronic signal processing circuit 100 which, on the basis of the signals, makes it possible to obtain fault identification signals.

 FIG. 2 represents the constructive arrangement of the amplification / filtering circuits lOB. Circuit 10A has the same structure as circuit 10B. The amplification / filtering circuit 10B comprises a pre-amplifier 12, a band-pass filter 13, an amplifier 14, a low-pass filter 15, a peak signal detector 16 connected to the output of the low-pass filter 15 At the output of this peak signal detector 16 is connected an automatic gain control circuit (AVR) 17, the output of which is connected to the amplifier 12.

After the peak signal detector 16, a low-pass filter 18 is also provided.

 The output signal from the low-pass filter 15 is delivered, via a conductor 19, as a video signal representative of the transverse faults, to the output 25 of the processing circuit 100. The output signal from the low-pass filter 18, which has a characteristic operating curve different from that of the low-pass filter 15, is sent as a first input signal to a multiplier 23 which is part of the processing installation 100.

 The output signals of the amplification / filtering circuits 10A are sent, via a conductor 22, as video signals representative of the longitudinal errors, to the output 28 of the processing circuit, or via a conductor 21 as the second input signal to the multiplier 23. The two input signals, after their multiplication by the multiplier 23, are then sent to the output 24 where they are representative of signals corresponding to transverse faults.

 The signal present on the conductor 22 is inverted using the inverter 29, it undergoes a subtraction after passage through an addition and differentiation circuit 26. On the basis of the video signal for the longitudinal errors Vl and of the signal video for Vci transverse errors, a global signal is obtained.

 With the aid of the above-mentioned inversion and addition processes, which in fact result in a subtraction, the signal obtained at output 27 is for the most part devoid of any disturbing component at high frequency. The signal obtained at output 27 comprises, as seen in FIG. 2, front and rear ridges, which correspond to the start and the end of an exploration period, and which correspond to a facet F of the wheel. with mirrors 4. The signal in the form of pulses and which is between these two peaks corresponds to a fault.

 If we consider the scanning spot 11 in FIG. 1 and the different types of possible faults, we can see that, depending on the specific types of faults, the fault signal obtained occurs either during a plurality of scanning periods succeeding each other, ie during a very small number of successive scanning periods. In the case of a transverse fault, a relatively long signal occurs in the first partial light beam path and the associated circuit. For the other path, on the other hand, the signal obtained is relatively short. Indeed, for the latter, the slit-shaped diaphragms are oriented transversely to the transverse defect.

 In the case of longitudinal faults, the pulse signals obtained are therefore very short. If a correct choice is made of the time constants relating to the low-pass filters 15 and 18 and to the amplification / filtering circuits 10A and lOB, we arrive at having sufficient discrimination between transverse and longitudinal faults and short faults and long in the material strip 1.

 FIG. 3 represents a variant 100 ′ of the processing circuit. The multiplier 23 corresponds to the multiplier 23 shown in Figure 2. I1 is used to signal transverse errors. Likewise, the signal on the conductor 22 is inverted by an inverter 102.

 In addition to the differentiation circuit 101, which corresponds to the differentiation circuit 26 of FIG. 2, in the circuit 100 'according to FIG. 3, a second differentiation circuit 103 is provided which provides the second derivative dv2 / dt2 of the sum signal. (V1 + VQ) An integrator 107 is interconnected between the differentiator circuits 101 and 103. The first differentiator circuit only eliminates the steep sides of the fault signal, without however affecting the parasitic signals. This first differentiation, however, distorts the polarity of the fault signal. The integrator 107 dampens the high frequency components in the output signal of the differentiator circuit 101 which performs the first derivation.

The second differentiator circuit 103 which follows the integrator 107 cancels the polarity errors sorted by the first branch circuit.

 From the output signal of the differentiator circuit 103 an analyzer 104 provided with 5 processors CPU1 to CPU5 highlights the transverse, longitudinal and punctual faults.

A computer 105 is connected to the analyzer 104 in order to evaluate the faults, and possibly to activate a display device or an alarm.

 All the optical, mechanical and electrical constructive parts of the fault identification apparatus according to the invention are commercially available parts at an advantageous price.

 The optical filters 7A and 7B shown in FIG. 1 can be chosen as a function of the color of the strip of material 1 so as to obtain in the channels corresponding to the path of the partial beams color signals A or B.

 Referring to Figure 5, we will further explain below the operation of the optical scanning installation according to the invention.

FIG. 5 represents video signals of faults emitted by the scanning device in the transverse fault channel and in the longitudinal fault channel, where the various marks have the following meaning:
T = inspection time
t = dead time (corresponds to the change of mirror
reflecting the polygonal mirror wheel)
A = signal amplitude level
T + t = operating time of a mirror surface
of the polygonal mirror wheel.

 During a change of reflective mirror surface, the receiver (photodiode) is illuminated by a light intensity twice that corresponding to the inspection time. Optical scanning causes the amplitude "A" in the faulty video signal at the change point, regardless of the diaphragm. A peak value detector 16 (FIG. 2) located at the end of the chain provides access to the DC voltage value of this amplitude "A".

 This value serves as an input signal to the following two circuits: on the one hand the AVR-17 circuit (figure 2) for automatic amplitude adjustment, and on the other hand the signal passes through the low-pass filter 18 ( Figure 2), followed by the multiplier 23 (Figure 2) with differential input. The lowpass filter 18 lets through only the peak variations which are caused by several changes in the surface of the reflecting mirror.

 In the event of a transverse defect, the peak variation of the amplitude "A" of the photodiode with transverse diaphragm is notably greater than that of the photodiode corresponding to the diaphragm in length. The two low-pass filters 18 of one and the other circuits 10A and 10B (FIG. 2) send their respective signals to the differential inputs of the multiplier 23. In the latter, thanks to the differential inputs, the identical parts of the two signals 1 on conductors 20 and 21 are eliminated, and the effective variation in amplitude of the two peak values is only multiplied by a constant factor. The signal produced at the output of the multiplier device is thus undoubtedly a signal indicating a transverse fault.

 Of course, the present invention is not limited to the embodiment described and shown, but it is susceptible of numerous variants accessible to those skilled in the art without departing from the spirit of the invention. .

Claims (18)

Claims  1. Defect identification device for identifying faults on moving parts of material comprising: a light source situated on one side of the object, an optical installation containing a rotating mirror wheel situated on the same side or of the other side of the material part, a photoelectric converter which is placed on the path of the light rays reflected by the mirrors of the wheel and a processing circuit which is connected to the output of the photoelectric converter, characterized in that it comprises a beam splitter (6) which divides the light beam coming from the mirror wheel (4) into at least two partial beams and in that on each partial beam path (A, B) is arranged a diaphragm (8A, 8B) and / or an optical filter (7A, 7B) which differs respectively from the diaphragm or the optical filter located on the other beam path by its characteristics of aperture shape or respectively of f iltrage.  2. Device according to claim 1, characterized in that the beam splitter is a partially transparent mirror (6).  3. Device according to claim 1 or 2, characterized in that it is used for monitoring material defects in the form of strips.  4. Device according to one of claims l to 3, characterized in that the diaphragms (8A, 8B) on the two paths (A, B) of the partial beams are slit diaphragms whose images intersect substantially perpendicularly on a strip of material (1) and there define an optical scanning spot.  5. Device according to one of claims 1 to 4, characterized in that on each partial beam path is interposed an amplification / filtering circuit (10A, 10B).  6. Device according to claim 5, characterized in that a multiplier circuit (23) is coupled to the output of the amplification / filtering circuit.  7. Device according to claim 5 or 6, characterized in that at the output of the amplification / filtering circuit (10A, 10B), is coupled a differentiation subtraction circuit which performs subtraction and differentiation operations.  8. Device according to one of claims 1 to 7, characterized in that the light source is a fluorescent tube (2).  9. Device according to claim 8, characterized in that the fluorescent tube is connected to an alternating current source whose frequency is at least 50 kHz.  10. Device according to one of claims 3 to 9, characterized in that the strip of material is a tubular strip produced by a circular knitting machine, and in that the light source (2) is placed inside the tubular strip, the optical installation containing the mirror wheel (4) being outside, or vice versa.  He. Device according to one of claims 1 to 10, characterized in that the mirror wheel is polygonal.  12. Device according to one of claims I to 11, characterized in that a Fresnel lens (3) is disposed between the strip of material and the mirror wheel (4).  13. Device according to one of claims 1 to 12, characterized in that the processing circuit (100, 100 ') comprises a circuit for detecting longitudinal faults, a circuit for detecting transverse faults and a circuit for detecting faults punctual.  14. Device according to one of claims 1 to 9, for sorting, according to color, parts of material, for example for recycling used bottles, characterized in that, on each partial beam path, is arranged a optical filter (7A, 7B) whose characteristic of chromatic transparency is different from that of the filter located on the other partial beam path.  15. Device according to one of claims 4 to 14, characterized in that each of the amplification / filtering circuits (10A, 10B) comprises a peak signal detector (16) to which is connected, downstream, a filter low pass (18).  16. Device according to claim 15, characterized in that an adjustable amplifier (12) is interposed upstream of the peak signal detector (16) and in that an automatic gain control circuit (17) is connected between the output of the peak signal detector (16) and an input for controlling the adjustable amplifier (12).  17. Device according to one of claims 5 to 16, characterized in that the multiplier circuit (23) has a differentiation input.  18. Device according to one of claims 4 to 17, characterized in that the outputs of the two amplification / filtering circuits (10A, 10B) are connected directly or via an inverter (29) to a differentiator (26, 101 ) performing an addition.  19. Device according to claim 18, characterized in that the differentiator (101) is coupled downstream to an integrator (107) itself followed by another differentiator (103).