TRANSMITTED LIGHT

* * * Background of the invention

Field of the Invention and Prior Art

The present invention refers to devices and methods devoted to detection of the weft threads in textile jet looms by means of optical sensors based on transmitted light.

In particular, the invention refers to devices of the known type indicated in the pre-characterizing portion of claim 1, and to methods of the known type indicated in the pre-characterizing portion of claim 10. Devices and methods of this type are for instance known from EP 0 137 380, EP 0 290 706, EP 0 212 727, US 5 063 973, US 4 565 224 and WO 98 24957. The weaving technique

The production of a defect-free fabric depends on a number of factors related to the availability of proper yarns and efficient machines, and to the rate of control in the various phases of the weaving process. As regards the weft thread contribution to the fabric formation, stable weaving conditions are required in order to obtain a constant weft spacing that is a constant number of weft threads per unit length of fabric.

This target depends first of all on defect-free and correctly inserted wefts, so that a special attention must be paid to the weft insertion into the shed, and to the following beat-up of the reed.

Each weft insertion requires a thread length greater than the fabric width (or height) due to the wavy shape assumed in the final interwoven structure of the fabric by both the weft and warp yarns.

It is then necessary to control the length of the inserted weft thread and its integrity along the full width (or height) of the fabric being produced, at each reed beat-up, and during the whole weaving process.

To this aim, it is conventional to secure two weft sensors along the reed, one at a short distance from the other, at that point of the reed corresponding to the required width, and hence not necessarily at the end of the reed. The first sensor signals the insufficient length of a weft thread (short weft)

and the second sensor signals a possible thread breaking in an upstream point of the reed channel (long weft).

Only once the two sensors have confirmed even weft insertion, proper devices cut the weft threads at the fabric ends, and form the selvage to block the side warp threads while tightening the inserted weft ends.

(According to another known technique, the above mentioned pair of sensors are applied to a strechhing device...) ...)The above mentioned pair of sensors are sometimes applied also to a stretching device, the "reed extensor", which provides a final tension on the weft yarn at the beat-up, and it is located in prosecution of the reed.

For this reason the reed extensor, whose possible advantages are not considered here, requires the availability of a numbers of reeds to allow for the production of fabrics with different widths. This aspect is not considered here, since the present invention relates to a sensor method and apparatus applicable at any point of the reed, so allowing the production of any possible fabric width from a single reed.

The design of weft thread sensors must take into account some conditions and requirements from weaving in general, and from weaving with jet looms in particular. Under the action of the main jet, the nozzles, and the reed reciprocating movement, the yarn proceeds along the reed channel with a wave-like movement, so that each time it may pass at any point of the channel cross-section, practically in a random manner, although it passes more frequently at certain zones of said section. In general, it is possible to define a sensitive zone, inside the reed channel geometrical section, in which the yarn can be expected with a specified rate of probability.

The weft contraction due to the beating-up of the reed after the shed inversion requires the presence of a contrasting device to limit the strong warp yarn tension against the dents (or "lamellae") at both ends of the part of the reed engaged by the fabric. This device, "the temple", is located as closest as possible to the weaving line, and is supported by a structure fixed to the loom: for this reason it is not possible to position parts of the sensors in front of the reed channel.

The oscillations of the batten, to which the reed is secured just by its lower base, originate longitudinal shocks to the reed, which reacts with quite complex

vibrations of its lamellae. Amplitudes and frequencies of said vibrations depend on loom speed, lamellae density, lateral air jets, type and number of warp yarns. This effect has no influence, in general, on the warp density distribution, but must be taken into great account in the design and use of weft sensors, in particular of not invasive sensors.

The fabric production frequently requires the series insertion, into the shed, and according to a program, of weft yarns very different as for material, physical and chemical treatments, colour, size. Weaving with a plurality of wefts leads to ask for weft thread sensors sensitive to the size, and not to the optically varying characteristics of the yarn. Electro-optical sensors

Electro-optical techniques are used in general in the size measurement of a moving yarn, and in particular in the detection of the presence of a weft thread when weaving with air-jet looms. In an electro-optical sensor a beam, or beams, of light (electromagnetic radiation from ultraviolet to infrared) has a definite geometrical path, or paths, between one or more light generators (LED, laser) and one or more light detectors (pin, phototransistors), in said path being present, or not, passive optical elements. Both the source and the detector may be (or may be not) integrated by separate optical systems, i.e. a transmitting optics and a receiving optics, having a common axis of symmetry, which is the optical axis of the system.

The amount of energy emitted by the . source per time unit (sec), and contained within the radiation beam, is the "radiation flux" of the beam, and it is measured by radiometric units (watt = joule/sec). The amount of flux onto the detector, whose responsivity (ampere/watt) is assumed to be constant with time, is converted into an electric signal.

The radiation flux is constant in air, that is, it has the same value through any section of the beam, but it is modified when the beam crosses optical media different from air, or it meets reflecting or opaque bodies, or diffusing particles in the air.

The optical path from source to detector may then be thought as a sequence of different parts: if they all maintain their characteristics constant in time, the final radiation flux and its corresponding electrical signal remain constant.

The optical properties varying with time (variations of dimension, shape, position, refractive index, absorption, reflectivity, colour, ..) of one or more parts

of the optical path, or simply the intermittent presence of objects in one or more of said parts, produce a time variation in the final flux, and then a corresponding variation in the electrical signal.

In particular, the insertion of a textile thread in a radiation beam reduces the beam flux according to a function of the flux distribution and of the transverse dimension of the thread.

Going into details, it must be considered that a textile thread is a very complex object with respect to its constituting material, the form and structure of its elementary fibres, the form and structure received by the spinning process, the absorbed dyeing substances.

The interaction of a thread with a radiation beam is consequently a complex event, whose description may be essentially reduced, for the purposes of the present invention, to two types of processes: reflection and absorption.

A distinction must be made between "diffuse reflection" (Rayleigh) from elementary fibres, and "specular reflection" (Snellius) from relatively plane and extended, but variously oriented, elements of the thread.

The absorption, that is, the extinction of part of the incident flux, with a thread thermal level increase, is in general wavelength dependent and selective, as it depends from nature and morphology of the fibres and also from their minute applied pigment particles. For this reason, the re-emitted light spectral distribution results different from that incident on the thread, so giving the thread its colour, specific with the incident light. The selective absorption then determines the spectral components and the relative level of both the specular and, in particular, the diffuse reflection. Given a spectral composition of the incident wave, said processes,- absorption, diffusion, reflection -, arise with different weights depending on the thread, and, together, determine the value Φ _f of the total intercepted flux :

Φf = Φfd + Φfr + Φfa

being Φ _f( j that part of the intercepted flux which is found as diffused , Φ _fr as reflected , and Φ _fa as absorbed flux.

On the other hand, as said before, for a given flux distribution, the amount of the total intercepted flux Φ _f depends from d , the tranverse dimension of the thread :

Φ _f = Φ (d)

It follows, from the above notes, that, in the generality of the applications, two types of electro- optical sensors are mainly possible, namely based on transmitted light or on reflected light, or on combinations of the two. In the weft thread control on air jet looms both types are used in different forms, with advantages and disadvantages not evenly distributed. However the present state-of-the-art devices show frequently, in their optical design, an attention not proportionally paid to the complexity of the problem, and give to electronics, downward the sensor, decision tasks on the basis of misleading information, with a quite modest final reliability.

Reflected light sensors In reflected light sensors one or more light sources radiate the reed channel sensitive zone, to obtain a signal from the light reflected by the thread, which is then the true source of information-carrying radiation to the detector. For this reason in a properly designed reflected light sensor, light from the illuminator must not reach the detector directly nor after reflections from not pertaining-to- thread elements, as, for instance, from dent edges of the channel sensitive zone.

The above features must be pursued taking into due account that the signal from light reflected by the thread mainly depends on variable elements not related to the thread dimension.

In particular, the reflected flux Φg. from the thread is a fraction, small and strongly variable, of the total intercepted flux Φ _f .

In fact, the division of the intercepted flux among reflection, diffusion and absorption depends, as shown, on the relationship between incident wavelength and thread morphology.

A further wavelength-dependence comes in the electric signal from the reflected flux, as the spectral responsivity of the detector varies by one order of magnitude from visible to infrared.

Furthermore, a strong angular dependence, may be expected in the detector response due to the possible different positions of the thread in the sensitive zone. Moreover , the reflected flux is distributed, in a varying and not uniform way, in all directions with respect to the incident light, so that a receiving optical

system, even if complex, can only catch a minimal fraction of it.

Because of the foregoing reasons, the use of the flux reflected by a thread to detect presence of the thread, and, even more, to measure its size, is an hard task which can't be correctly accomplished by means of rough methods. It is for these reasons that present reflective sensors are frequently not able to detect low titre yarns, or coloured yarns with a small reflecting power.

A typical failure may hence be expected when weaving with a plurality of wefts .

For instance, a not properly designed reflected sensor which necessarily leads to modest results is described in WO9927172Al .

The reflective sensors work with a good efficiency degree in a number of applications, but not, at present, in the detection of textile threads.

Transmitted light sensors

In transmitted light sensors, a light beam permanently crosses the sensitive volume assigned to the sensor in the weft channel, its optical axis being definite and fixed with respect to the reed. Then, as in all types of sensors, the design starts from the definition of the sensitive zone.

Let-Φo be the final beam flux value, in absence of yarn, at the end of the optical chain of possible passive components from the source to the detector, including the reed, which always conditions the sensor design.

Assuming constant in time the relative position and the characteristics of said components, also the value V ₀ of the electric signal in which the flux is converted remains constant in time.

This value V ₀ of the electric signal in absence of yarn allows to know the sensor efficiency, and the rate of optical surfaces contamination from textile particles and chemical substances released by threads weaving.

In a correct transmitted-light sensor design, the light diffused and reflected from a yarn present in the sensitive volume does not reach the detector.

Given this essential condition, the presence of a yarn in the sensitive volume reduces the flux Φ _o to the value Φ _s of the unchanged flux (that is, "transmitted"), so that:

Φ _o - Φ _s = Φ _f .

The difference between the two flux values measures the value Φ _f of the

subtracted flux, and it does not depend on how this subtracted flux might be distributed among diffusion, reflection and absorption, but only on the projected dimension from the thread inside the beam, and hence on the thread titre.

For the above reasons, the signal from a transmitted light sensor does not depend on yarn colour, nor on the used wavelength, and it is possible to make it independent on the yarn position in the sensitive zone. In that case, the longitudinal sections projected by two yarns which differ only on their colour are equal, as are their subtracted fluxes, and then their corresponding signals.

The transmitted light method, in line of principle, offers every element for an efficient detection in all weaving situations, without the intrinsic limits of the reflected-light method, but for a correct application the transrnitted-light method requires a more complex and expensive technology.

Furthermore, the transmitted light sensors produced or suggested thus far mainly have an optical element in common, whose shape and dimension enables it to be inserted between two adjacent reed dents, in order to generate an optical field in the weft channel which is constant in level and flux distribution.

In accordance with the above description, the weft thread, while crossing the optical field, reduces the flux, and operates the sensor.

The above mentioned optical element is designed, manufactured and protected in various ways by different producers. It guides the light produced and detected through the sensitive zone by electro-optical components which are located in the sensor body, on a same side of the reed.

The transmitted-light signal detection, in said known solutions, is therefore obtained in a way which is invasive of the space between reed dents. Devices of this kind are, for instance, known from EP 0 137 380, EP 0 290

706, EP 0 212 727, US 5 063 973, US 4 565 224 and WO 98 24957.

In many cases and for a plurality of reasons, the insertion between reed lamellae of parts of the sensor may have negative consequences in a subsequent use of the reed where a wider fabric is produced. In fact, the couple of lamellae which have temporarily received these parts of the sensor in the previous position of the sensor, may result worn out or injured in some measure, so that the subsequently produced fabric may show unacceptable discontinuities at that point.

The above facts may more probably happen with high fineness reeds, if the thickness of the inserted optics largely exceeds the space allowed by the elasticity

of two adjacent dents.

The transmitted-light sensors which have parts to be inserted between reed dents may, on their turn, be worn out and injured for a number of reasons: limits and defects of the sensor design or of its manufacturing process, the transit itself of the weft yarn, an incorrect sensor mounting to the reed, rough and destroying cleaning and maintenance operations of the exposed optical parts. Reflection / Transmission Comparison

Ih the following a comparison is made between the reflected and transmitted light methods with reference to the most important parameters of the use of sensors to detect weft yarns.

From a general point of view, and for said reasons, the features of the reflected-light method allow neither a safe yarn detection in all possible different situations, nor, less than ever, its measurement.

On the contrary, the transmitted-light method has all the necessary characteristics for a perfect information, not only about yarn presence in all possible cases, but about its measurement too.

For what concerns applicability, the transmitted-light method is more technology-demanding, as it requires more complex technical solutions, which appear simpler and less expensive with the refiected-light method. However, the transmission method may lead to good results in yarn detection, while from the reflection method only poor results can be expected.

Furthermore, the transmitted-light sensors can be positioned at every point of backside reed, while the fabric carrier with the temple support can limit continuous positioning of reflected-light sensors in frontside reed. As indicated above, transmission devices are always invasive, whether for the insertion of optical parts (with or without mechanical protection), or simply because of the insertion of mechanical parts having connecting or expanding functions. Compact and not invasive sensors can be devised from the reflection method. For the above reasons, with invasive transmission devices the risk exists of injuring the reed, with possible deformations and wear of the lamellae due to mechanical and chemical facts, and of injuring the sensor too, with possible damage to optical parts due to incorrect positioning and/or wear from thread and lamellae. Furthermore, damages to the fabric are possible, with imperfections corresponding to the reed points where the sensor is previously located.

No risk of damage to reed or fabric exists with reflection sensors. The transmission method is superior in particular for what concerns "reliability", since the sensitive zone may be optically covered more efficiently in all situations, even if with possible errors with lower titles, while, with reflection sensors, said optical covering is not homogeneous or absent with lower titles or dark colours.

Sensitivity is acceptable in transmission sensors in general but less with lower titres which may ask for a reduction of the sensitive zone, while, in reflection sensors, said sensitivity is modest, always dependent on colour and ' thread position, till to be scarcely adequate with lower titres.

Both systems, in the known solutions, are not adequate for what concerns optical surface sensor contamination, which may soon rise up to unacceptable levels. Cleaning operation may require time and special care. The transmission system is advantageous in that it gives a continuous measure of contamination, not possible with reflection systems . Object of the invention

The purpose of the present invention is to define a method to detect weft yarns in textile jet looms, and to provide, for the application of this method, a sensor characterized by being not invasive of the space between the lamellae or dents of the reed, with the possibility of continuous adjustment of the position of the sensor at any point along the reed, with high reliability and based on transmitted light, and finally with high sensitivity. Summary of the Invention

In order to achieve this object, subject of the invention is a device whose characteristics are indicated in the appended claim 1, and a method whose characteristics are indicated in the appended claim 10. Further advantageous features of the invention are indicated in the dependent claims.

In the case of the invention, the reed weft channel of a textile jet loom is crossed by an extended and collimated light beam, which illuminates two or more reed lamellae and is generated and received outside the reed, without inserting any element, neither optical components nor elements of any other type, between said lamellae.

The thread, when present in the reed channel, intercepts and attenuates the light beam, so originating a signal which, for a given textile material, depends only on the titre of that part of thread, and not on its position in the channel.

This principle can be accomplished by several distinct embodiments of electro-optical transmission sensors, which are not invasive and which can be continuously positioned along the full length of the reed, while ensuring high reliability and sensitivity. More in detail, in the device according to the invention, a collimated light beam having a substantially rectangular cross-section and predetermined dimensions, comes out from the external surface of an optical transmitting system contained in the sensor body, and is directed so that it goes through the reed, at the weft channel. The optical axis of this beam is parallel to the plane of the dents or lamellae, and forms an angle relative to a direction orthogonal to the reed, whose amplitude is determined as a function of the loom geometry, and may be typically (but not exclusively) of about 60°.

In these conditions, the front edges and, to some extent, the internal surface of the reed dents, intercept part of the beam, by absorbing and reflecting it in various directions: this part of the beam is not used to detect the weft yarn.

What remains of the beam moves forward through the spaces between the dents, goes through the full free space of the reed channel, proceeds again through the spaces between the dents, and finally emerges from the reed being still partially collimated.

The beam presents now an intensity distribution, which has, in the longitudinal sense of the reed, a minimum value, near to zero, right at each illuminated dent, and a maximum value between the dents. Moreover , on the full beam front, and superimposed to said longitudinally periodic distribution, there is a further intensity variation which depends not only on the incident beam intensity distribution, but, in the, real cases ( that is with real reeds, optics and sources), depends particularly on the different path lengths between dents of the different parts of the beam.

In the limits of the approximation allowed by this sensor application, it is possible to say that a short segment of a reed made of collimated light emerges from the loom reed.

This part of the beam contains the information relative to the presence of the weft yarn, and it is used for its detection.

In any sensor, the value of the electric signal due to a yarn in the reed channel is, normally, a function of its position in the channel. Said function

depends on two dimensions, that is, on the coordinates of the channel points, in any reference system in the dents plane.

In certain sensor types, usually in reflection sensors, the signal depends also on the axial reed coordinate. This fact shows the limits of the methods adopted till now, and how critic are data obtained from the corresponding real sensors.

In transmission sensors, the electric signal value represents the beam attenuation due to the yarn presence in the weft channel, by means of the ratio of the subtracted light to the total light in absence of yarn.

If the transmission sensor uses a collimated light beam, the electric signal is a function of one dimension only, that is, of the "h" coordinate, transverse to the beam in the dents plane. In fact, the same value of radiation flux is found, by the definition itself of collimated beam, in all points of the weft channel belonging to the same straight line parallel to the beam optical axis.

If, moreover, the collimated beam is uniform, the radiation flux has the same value in all channel points, hence the yarn attenuation does not vary even with h, and the signal is independent on its position in the channel.

In this way, for different yarns made of the same textile material, cotton for instance, that is, with same specific weight, same spinning and twisting, the sensor signal depends only on the yarn titre, and it does not appreciably change with its position.

Even if the main purpose of the sensor according to the invention is the detection of yarn presence and not its diameter measurement, a constant level signal, for a given yarn, is the condition for a certain detection, and for the best planning of the command and control electronics of the full weaving operation. Moreover, by a sensor as described above, the loom operator may check and control the sensor performance itself, when he knows the nominal titre of the weft yarn.

Advantages of the invention

Due to the above mentioned characteristics, the invention provides a number of important advantages from many points of view. It has all features necessary to a correct and full information not only about the yarn presence in all possible cases, but also for its measurement.

Although technologically demanding, it obtains better results than by any known device for what concerns yarn detection and its measurement. It is possible to position the sensor at any point along the reed. The system

according to the invention is not invasive, it doesn't damage the reed, it is not injured by weaving. The fabric is not impaired by previous positioning. A homogeneous and complete covering of the sensitive zone gives high reliability, even with lower titres. In the preferred embodiment, a double control of the yarn in the optical channel quadruples the sensitivity versus the yarn titre, the more so as lower is the titre.

Contamination is at its minimum level, due to total continuity between the relatively large and planar outer optical surfaces and the sensor body. Cleaning is easy, immediate, doesn't require sensor removing, nor skilled personnel. The contamination control is possible, easy, and continuous in that the contamination level is known from the signal in absence of yarn.

Brief description of the drawings

Further features and advantages of some preferred embodiments of the invention will become apparent from the following description, with reference to the accopanying drawings, which are provided only by way of non lnniting example, and in which :

- figure 1 is a schematic side view of a simplified embodiment of the invention with single optical path, that is, with light emitter and light detector located on opposite sides of the reed, so that the light beam crosses the reed only once in the yarn detecting zone,

- figure 2 is a perspective view of a working solution corresponding to the simplified embodiment of figure 1, in which the optical beam is illustrated as a solid part for sake of clarity, - figure 3 is a perspective view of the device of figure 2, from the back-side of the reed,

- figure 4 is a further perspective view schematically illustrating the optical beam once it has crossed the reed, said beam being represented as a solid part which has a reed shape too, - figure 5 is a schematic side view of a second embodiment with a two-way optical path, that is, with optical beam emitter and detector located at the same side of the reed, and a reflecting optics located on the opposite side to address the optical beam back to the reed, once said beam has crossed said reed a first time in the weft detecting zone, - figure 6 is a schematic view in the plane of the outward and backward

optical beam axes in the device of figure 5,

- figure 7 is a perspective view of the device of figure 5,

- figure 8 is a different perspective view of a different embodiment of the same device, - figure 9 is a perspective view corresponding to figure 7, where an upper clamp has been removed for clarity,

- figure 10 is a perspective view, showing solely the device according to the invention, with the emitting and receiving parts of the optical beam, and the reflecting part of said optical beam, where the optical beam has been visualized as a solid part, for clarity,

- figures 11, 12, 13 are perspective views from the back side and the front side of the reed, showing the components of the device of figures 5 - 10,

- figures 14 and 15 respectively show the optical scheme and a perspective view of a further embodiment of the invention, in which the outward and backward optical beams are superimposed in the same optical path, backcrossing the reed at the same point of the weft detecting zone, instead of being directed along two parallel and spaced paths as shown in the embodiment of figure 6.

Detailed description of some preferred embodiments of the invention

In the enclosed drawings, reference number 1 generally designates the loom reed, including a plurality of dents or lamellae 2 which define a weft channel 3. Moreover, A and B respectively designate the components which, in all embodiments of the invention, are respectively located on the two opposite sides of reed 1, and are supported in such a way to be fully external to the reed, in order to avoid all possible mechanical interferences with dents 2 of reed 1.

Reference number 4 designates the slay to which clamp structures 5,6 (figure 5) are fixed, the structures 5,6 being continuously adjustable in their longitudinal position along the reed, to support the components A and B. The fabric support 7 represents a fixed part of the loom that the reed 1 approaches when the shed is closing.

Reference L indicates the optical beam, with optical axis O. In embodiments with a two-way optical path, the optical beam presents a forward beam Ll and a backward beam L2, with their own optical axes Ol , 02 (figure 6). With reference to figure 4, the optical beam, once it has crossed the reed, consequently takes up a

reed shape, with, dents I which in figure 4 are represented as solid parts for clarity.

In the case of the embodiment with a two-way optical path (see for instance figures 5, 6), the device part A includes both the light emitting device (indicated by S in figure 6) and the light receiving device R. Both in the embodiment with a two-way optical path, and in the embodiment of figure 1 with a one-way optical path, the beam emitting system has an optics OT to shape the optical beam output from S into the collimated beam L, or Ll, L2. Similarly, the beam receiving system has an optics OR (figure 6) to focus the collimated light beam into the receiver device R which is of any known type able to give output electrical signals representative of the received optical signal.

In the two-way optical path embodiment (figure 6), part B includes a reflecting optics ORF able to reflect the optical beam twice, through a 180° deviation, from Ol into the opposite direction 02, in order to send back the beam toward the receiving device R. Figure 7 shows the device of a second embodiment of the invention, highlighting how A, B components can be easily and continuously positioned along the longitudinal direction of reed 1, thanks to that no parts are to be introduced through dents 2 of reed 1.

Figure 9 shows the device of figure 7, with upper clamp 6 removed for clarity.

Hence, in all embodiments the sensor body is made of two independent parts A and B, which are located, with respect to the loom operator, front-side and back-side of the reed, by properly using the slay (or, according to figure 8, by using the slay only for the front-side part, and the upper base of the reed for the back-side part).

Parts A and B of the sensor don't insert any element, containing or not optical components, between reed dents, so that the sensor is, as a whole, not invasive.

The front part of the sensor, part A for instance, takes up some part of the free space among slay, reed, fabric and temple support, in the reed forward position of shed closing, that is, of weft beating up. The back part B takes up some part of the free space among slay, reed, and warp throwing system, in the reed backward position of shed opening.

In the above described conditions, there are no obstacles, due to moving or fixed parts of the loom, to position the sensor at any point of the reed.

The two parts of the sensor are aligned, with proper and easy procedure, one with respect to the other, and permanently fixed to the slay, or to the slay and to the upper part of the reed, once the desired position of the sensor along the reed has been found, in order to achieve the fabric required width. The two parts of the sensor contain electronic circuits, electro-optic components (that is, light emitters and receivers), optical components (that is, lenses, mirrors, windows, prisms, diaphragms, splitters), in various ways, depending on solutions, in order to accomplish some important functions.

A primary function is the generation of a collimated light beam between A and B.

A source S, that is, a light emitter, - a LED or a diode laser - and a first optical system, called transmitting optics OT, are provided to said purpose.

Source S converts an electric input (ampere) into a radiation flux (watt), and operates typically in the near-infrared (0.8÷1.0 μm). In details, the collimated beam of light (that is, a beam which is neither divergent, nor convergent, and thus having a cross-section constant both in shape and dimension) is to be generated in A and received in B, or vice versa, and crosses the reed, with a due slope, typically (but not exclusively) from down in an upward direction, at the weft channel level. Once crossed the reed, the collimated beam turns into a reed of light, which contains information about the possible weft presence in the channel. Shape and dimension of the beam cross-section are selected to illuminate the whole weft channel on a width of two or more dents.

If crossing through the reed were not to alter the intensity distribution in the beam wavefront points, that is, -the beam being collimated-, in the points of its normal (i.e. orthogonal) cross-section , the search for constant attenuation ratio would be reduced, in the most simpler and intuitive way, to design a collimated beam, with uniform intensity distribution, and with a normal cross-section of a rectangular shape. The width L of the beam section would determine the illuminated dents number, for a given reed dents density, and the height H would fill out the sensitive zone section in the weft channel, according to the beam slope versus reed. In these conditions, the ratio of the subtracted quantity of light to the total quantity of beam light is given by the ratio of the area of the normal beam cross- section taken up by the yarn, to the full area of said section. On its turn, said ratio,

not depending on the beam width, is given by the ratio of the projected yarn diameter d, to the height H of the rectangular beam cross-section.

Therefore, for instance, in the described simplifying hypothesis, the attenuation introduced by a 0.1 mm diameter yarn on a 5.0 mm height beam is 2%. Said attenuation is independent on yarn position in the weft channel because the beam has been assumed to be collimated, uniform, and of constant width in the normal cross-sections of equal height which can ideally be traced through the weft channel.

In reality, the reed does change the wavefront flux distribution in the two modes previously described. The transformation of a continuous incident beam into a discontinuous reed-of-light structure is inevitable whichever may be the reed point and the beam slope, but has consequences only on the total level signal and does not change, by itself, the ratio of projected yarn width to the beam height. At worst, the collimated beam may enlighten only two reed dents, so that one light dent alone emerges from the reed, the second change being however present in it, namely an intensity variation along the beam height . With said variation, a yarn, moving along the weft channel, projects the same diameter value, but takes away different fractions from title total radiant flux, depending on its position in the weft channel.

In order to balance the optical channel, a compensation is required on radiant flux differences introduced by the reed versus beam height, so that the same attenuation may result at different beam levels, what can be efficiently be done by the parts of the optical passive chain which in the sensor come before and after the reed.

A radiant flux distribution on the receiver, equivalent to that originated by a rectangular, uniform and collimated beam in absence of by-reed-introduced distortions, is to be obtained with the balancing operation. To said equivalent distribution there are related the calculations of diameters, and so of titres, from yarn attenuations.

The necessary technology for the above operation is a possibly known argument of applied optics, which, as such, will not be explained nor claimed in the present invention.

The information about yarn presence may be made available in B, for instance, at the rear-side of the reed, by converting the residual radiation flux

(watt) into an electrical signal (ampere) which, in yarn absence, is the sensor base- signal. This is the task of a second optical system, the receiving optics OR, and of a photodetector R, - a PIN diode, or a phototransistor-, whose responsivity (ampere/watt) is assumed substantially constant in time. The first, simpler, "one optical path" configuration is so made up : it has only one beam transit through the reed, with a part of the sensor (for instance A) generating the beam, and the other part ( for instance B) receiving it. The base configuration of figure 1, applies the claimed principle in its simpler scheme, what does not exclude specific difficulties of design and realization. In the second embodiment (figure 5), the light beam, once crossed the reed a first time, instead of being converted into an electric signal, is to be reflected toward the reed to cross it again. To this purpose a component or, more generally, a passive optical system, the reflecting optics ORF, is to be located in the B part of the sensor. The light beam suffers, for a second time, the described radiant flux losses and intensity distribution variations. A once again modified reed of light emerges from the reed.

In this second passage through the reed as well, the light beam, which had a previous memory of the yarn presence in the weft channel due to the first passage, is again intercepted by the yarn, and acquires for the second time the same information.

The conversion of residual radiation flux in electrical signal, is obtained, in this case, with undoubted practical advantages, in the same sensor part (for instance, part A, front-side reed) where the collimated beam is produced.

The twofold beam passage through the reed, with a double quantitative check of the same yarn in the weft channel, gives up an almost double detection sensitivity, as regards yarn diameter, as it will be seen in the following.

The second embodiment is then provided according to the above mentioned "two optical paths" principle (see, for instance, figures 5 and 6), that is, with two transits of the light beam through the reed. Only one sensor part (for instance A) is dedicated to the beam generation and detection, the other part (for instance B on reed backside) to its reflection.

The above "twofold optical path" principle is preferably (but not exclusively) applied at least in two schemes which differ in the optical paths layout. In the first scheme (figures 5-13) the two paths Ll, L2, - outward and backward - are side by side, parallel and distinct, and therefore the reflecting

optics ORF operates a beam translation as well.

In a second scheme (figures 14-15), with coincident and inverted paths, the two paths overlap each other, in such a way the beam comes back on itself, and the reflecting optics ORF inverts the upper with the lower part of the beam. In both said configurations, i.e. either with side-by-side or overlapped paths, between its emission from source S and its arrival onto detector R the radiation beam meets a chain of components or passive optical subsystems - which are simpler in the first case, more complex in the second one-, among which the reed is included, which "filters" the optical beam two times in both cases. The reed is a rather complex passive optical component, by way of its structure made of parallel dents or lamellae, which operates as a multiple slit and an asymmetric attenuator. The design of all the other components of said chain depends on the reed.

Ih the configuration of figure 1, the full chain includes the radiation beam emitting source S, the beam forming transmitting optics OT, the reed P which channels the weft transit, attenuates the radiation flux and modifies the beam distribution, the receiving optics OR converging the beam into the photodetector

R ₅ which converts the radiation flux into an electric signal.

In the two paths configuration, side-by-side or overlapped, the full chain includes : the source S, the transmitting optics OT, the reed P in forward direction, the reflecting optics ORF which reflects and shifts, or reflects and inverts the ^"" beam, the reednnJTbackward direction, the receiving optics OR, and ^" THe ^~ photodetector R. In the overlapped path configuration, a single external optical part belongs to both the transmitting and receiving optics. Each component of the chain from S to R receives the radiation flux emerging from the previous component, modifies said flux according to its own function, and transmits it to the following component.

As known, the attenuation α of an optical component or system is the ratio of the flux Φ _s it subtracts from the incident flux Φj, to the flux Φi itself : α = Φ _s / Φi .

Similarly, the transmittance τ of a component or of an optical system is the value of the ratio of the radiation flux Φ _e transmitted to the following component, fo the incident flux Φj coming from the previous component : Being : Φ _s = Φj - Φ _e , the relationship between attenuation and

transmittance of a component is : τ = 1 - α .

When α = 0 , τ = 1 .

As the emerging flux from a component is the product of the incident flux on that component by its transmittance :

Φ _e = Φr τ , the emerging flux from a chain of n components is the product of the incident flux on the first component by the product of the transmittances of all n components : Φ _e , _n = Φi,i ^• T ₁ ^• τ ₂ ^■ . . . ^• τ _n

The reflections at the surfaces, the optical materials absorptions, the diaphragms, the uncoupling among input and output beams, all operate in optical systems to reduce their transmittance. The reflections and absorptions associated to the beam transit among dents operate to reduce the reed transmittance.

Also the active optical volume, that is, the space limited by weft channel and crossed by the radiation beam, may be considered as an element of the chain of optical fixed components.

Its transmittance assumes the value 1 in absence of weft yarn, as the emerging flux does not differ from incident flux, and, in presence of yarn, assumes the value X _f which represents the ratio of the radiation flux Φ _e , _f attenuated by yarn presence, to the incident radiation flux Φi _f in which the yarn is inserted :

Φ i,f

Let us assume that:

Φs is the radiation flux level emitted by S, Φ _O T is the flux emerging from OT,

Φ _pl is the flux emerging from the reed in the transit from OT to ORF, ΦORF is the flux emerging from ORF,

Φ _P2 is the flux emerging from the reed in the transit from ORF to OR, ΦOR is the flux emerging from OR, ΦR is the flux received by R.

and further: τoτ , τ _p i , T ₀ RF , τ _p2 , T _O R are me transmittances of OT, Pl, ORF, P2, OR, respectively.

The Φs level of radiation flux emitted by source S is reduced, with the beam passage through the optical passive chain elements, to the Φ _O R level of flux emitted by receiving optics, a fraction η ≤ 1 of which is received by the photodetector R .

In the absence of the weft yarn, for the single path configuration, it is :

ΦR, ₀ = ΪJ - ΦOR ^ ¹ V ^■ Φs ^• τor τpi - τoR ,

and, for double path (side by side or overlapped) configurations :

ΦR,O = V " ΦOR = V " Φs ' τoτ ^■ τ _P1 ^■ XORF ^• τp ₂ ^• T _O R.

If ₅ when the loom is working, and in the absence of weft yarn, the sensor parts, and hence, the whole of optical line elements, including the reed, don't present relative mechanical, electrical, and/or optical perturbations, the wavefront intensity distribution remains constant in time, as the total radiation flux level ΦR on the detector, and so the electrical signal V ₀ in which said flux is converted. V ₀ is the base signal of the sensor.

In the previous hypothesis, that is, with constant transmittances of the various elements of the optical passive chain, but in the presence of weft yarn, the product of transmittances with which, from flux Φs emitted by the source it is obtained the flux ΦR received by photodetector R, is simply completed with the factor T _f for the single path configuration:

$R,f,(i) = V ' ΦOR = V ^• Φs ^• TOT ' ^τ Pi " ^T OR " tf,

or with the factor X _f ² for the double paths configurations :

ΦR,f,(2) = V ^• ΦOR =

= rj ^■ Φs ^• TOT ' ^p ₁ ^• TORF ^■ τp ₂ ^• TOR ^• Tf _>

by assuming, for simplicity, :

τfl ^• Tf ₂ = Tf

In the single-path, the flux attenuation Φ _f due to the thread is : ocf,(i) = 1 - Tf .

Ih double-path, the flux attenuation Φf due to the thread, in two transits through the reed, is :

0Cf,(2) = 1 - Tf = αfe _s (i) ,

that is, said attenuation is equal to the attenuation due to an "equivalent" thread which, in a single path sensor, would give the active volume the transmittance :

Tfe,(l) ⁼ Tf

In the hypothesis of the equivalent distribution due to a uniform, rectangular, collirnated, distortions-free beam, the radiation flux attenuation caused by weft thread, with single-path, is :

H being d the projected yarn diameter, and H the beam normal section height.

The incident "on yarn" flux, (to say shortly "on the sensitive volume") is proportional to H, and the "downstream of yarn" transmitted flux, (to say shortly "downstream of the yam-containing sensitive volume"), is proportional to H-d . Then the transmittance T _f "of the weft yarn", when present in the passive optical components chain, that is, the transmittance. "of the sensitive volume when it contains the yarn", is : Tf = ^H ~ ^d = 1 - Of

H

In the double-path case, the attenuation of the equivalent yarn is:

and then its diameter :

d _fe , ₍ i ₎ = H ^■ α _f e _j(1) = — (2H -d) .

Jl

Therefore, the two-paths sensor configurations, either side-by-side or overlapped, generate, from a d diameter yarn, a signal, for which, in a single-path configuration, a d _fe ,(i) diameter yarn would be necessary. The ratio of the equivalent yarn diameter to the true diameter is : d _m) 2H -d d d H H ' that is, the diameter d _f ^ ₎ of the equivalent yarn, as seen by a single-path sensor, the more is near to twice the true diameter d seen by a two-paths sensor, the small is this d diameter with respect to H.

Going on with the example :

d= 0,1 mm , H = 5,0 mm ,

it is : α _f = 0,02 , → τ _f = 0,98 ,

τ _fe) (i) = τ _f ² = 0, 9604 ,

and then : d _fe = H ^• θ _fe)( i ₎ = 0,198 mm.

The two-paths sensor configuration generates, from a 0.1mm diameter yarn, a signal for that, in a single-path configuration, a 0.198 mm diameter yarn would be required, that is twice in practice, being : — — = 1,98. d

Also for comparatively small H values, and high d values, the equivalent diameter is quite near to twice the effective diameter. For instance :

with H = 3.0 mm, and d = 0.3 mm , d _fe = 0,57 mm, e ^ = 1,9 . d

The yarn titre, being defined as the mass of unit length, varies, for a given textile material, according to its transversal section, and then, assuming circular

said section, to its squared diameter.

Therefore the two-paths configuration allows detecting the presence of a weft thread whose titre is, with a good approximation, one-fourth of the smallest detectable yarn in a single-path configuration. The comparison between the two configurations in which it is possible, typically but not exclusively, to realize the principle of the invention, is to be completed by observing that, on its turn, the single-path configuration is, in spite of the lower sensitivity respect to the two-paths configuration, under any aspect, and then also for what concerns reliability and sensitivity of detection, highly superior to any other way might be devised in the present state-of-the-art of sensors, both of transmitting and in particular of reflecting type.

For what concerns the problems related to the provision of two-paths sensors, it is to be considered that the yarn passing through the first incident beam reduces its flux in the beam section points engaged by the yarn. If ₅ in the second passing, the yarn engages the same beam wavefront points engaged in the previous passing, the flux attenuation is lower, and the higher sensitivity from double passing gets down.

This possibility must be taken into account when designing the reflecting optics, taking however into account that the yarn position, due to its waving, is not rigorously parallel to the longitudinal axis of the reed. Moreover, in the coincident and overlapped paths solutions, the inversion produced by reflecting optics illuminates the yarn from two opposite directions, so completely eliminating the above problem.

How it has been clarified, the previous considerations have been made assuming that the sensor parts, and then in particular the optical line elements, including the reed, don't undergo relative mechanical, electrical and optical perturbations induced by the loom, being so constant in time the distribution and total flux value in the detection wavefront.

Reality complicates matters in a various way, and particularly, as seen, with the presence of reed lamellae oscillations with respect to their rest position, so determining reed transmittance variations, as for both the emitted and the received light beam, being source and detector fixed with each other and with the slay.

The reed transmittance variations produce variations of the flux value received by the detector, and then, of the base signal Vo , that is of the signal from the sensor firmly fixed at a point of the reed, and in absence of yarn and disturbs.

To avoid reading of said variations as of yarn presence signals, different criteria may be adopted depending on the loom control scheme, but at large based on the following considerations.

A threshold coefficient k _s links the maximum disturbance value V _n;inax to the base signal V ₀ : ■

Sensor design measures, so as loom operative criteria, may limit disturbance level, then increasing the k _s value, as characteristic of a given sensor when applied to a given reed, on its turn mounted on a given loom. Similarly, let's define a β _s coefficient, to link V ₀ to the minimum value

V _f i _m i _π of the signal due to the yarn presence:

V ₀ = βs ^• Vfcnm

The foregoing applies to not-invasive sensors in general. About not-invasive and transmitted-light sensors it must be yet observed that V _fjTnm is the minimum value that may be subtracted, that is, the smallest signal fraction subtracted to V ₀ which certainly identifies the yarn presence, and not the existence of a disturbance.

The maximum disturbance level then determines the upper threshold for the possible β values, as it must be : β _s < k _s . hi the sensors of the present invention, which are not-invasive and based on transmitted light, and yet characterized by the use of an equivalent collimated and uniform beam having height H and width L, the coefficient β _s expresses the ratio of height H to the minimum detectable diameter d _f , _e of the equivalent yarn : β _s = f- .

In the simpler model of the origin of disturbances is : k _s = p ^• L

Once limited, as much as possible, the dents oscillations, by means of mechanical parts of the sensor itself, the k _s value may be increased by the L value.

As it must be: β _s < k _s , and H is inferiorly limited by the necessary extent of weft channel sensitive zone, the sensors sensitivity is limited by disturbances level. hi fact, the β _s value gives both the minimum diameter value detectable with

a pre-fixed H value, and the maximum H value to detect a pre-fixed yarn diameter.

Hence, in exceptional cases with high disturbances due to the reed, and/or with extremely thin yarns, the minimum necessary height of the sensitive zone may be divided into two or more sections, separately addressed by the receiving optics to two or more independent parts of the detector.

In conclusion, from the foregoing it is clear that the goal achieved with the present invention is to provide a device and a method for weft yarn detection in a jet loom which is not-invasive towards the reed dents, which further can be continuously and easily positioned along the longitudinal reed direction, and finally which has a high reliability.

This last result is obtained thanks to that the sensor output signal only depends on the yarn geometry and not on yarn optical varying characteristics, like colour, reflectivity, diffusivity. The output signal from the device according to the invention does not depend on the yarn position in the weft channel, does not significantly depend on sensor surfaces contamination due to weaving, and overcomes the problems arising from possible transversal oscillations of reed dents.

The above advantages are so relevant that the device according to the invention results suitable to the yarn measurement also, and then not only to the detection of its presence.

All the objects and advantages described above are obtained thanks to that the device according to the invention operates with transmitted light, with a collimated optical beam, with a cross-section of constant shape and dimension, and with a uniform final flux distribution.

Particularly useful embodiments of the invention are the above described solutions with double optical path, side-by-side or overlapped, which greatly improve the device sensitivity.

Naturally, while the principle of the invention remains the same, the details of construction and the embodiments may widely vary with respect to what has been described above purely by way of example, without departing from the scope of the present invention.