GB2148498A - Method and apparatus for measuring the fibre length distribution of textile fibres - Google Patents

Abstract

A sample 26 of fibres is disposed on a plane transparent glass sheet 27 illuminated substantially uniformly by at least one set comprising a tubular lamp 28, cylindrical reflector 29 and cylindrical condenser lens 30 arranged with their axes parallel to an x direction corresponding to the length direction of the fibres in the sample 26. An objective 33 projects a real image at an image plane 35 where an array of photodetectors or the mosaic of a television camera tube is located. The image plane 35 is scanned electronically in at least one direction and the signals from the array or mosaic are digitally processed to eliminate non-uniformities due to characteristics of the apparatus and to produce a representation of the fibre length distribution. Other embodiments may employ an array of capacitive sensing elements to sense the sample directly. <IMAGE>

Description

SPECIFICATION Method and apparatus for measuring the fibre length distribution of textile fibres The present invention relates to a method and apparatus for measuring the fibre length distribution of a sample of textile fibres.

An accurate description of the fibre length distribution by one or several diagrams and by a set of parameters is necessary for many of the more important decisions and operations in a textile mill such as selection of the appropriate raw materials, blending of these raw materials, adjustment of the machine settings, prediction of performance factors, for example waste of material and frequency of "ends down" in spinning, prediction of yarn properties, etc......

Furthermore, the price of wool increases with mean fibre length in a very significant way.

For all these reasons, several attempts have been made to develop methods and apparatus for measuring the fibre length distribution. A first generation of instruments was purely mechanical.

Some of them were designed for the measurement of the length of individual fibres and were very slow in operation. Other instruments in this group used the "comb sorter" principle: fibres disposed in a field of combs were sorted into classes of predetermined length intervals. The fibre length distribution was finally obtained by weighing the fibres in each class.

Progress in industrial electronics led to the appearance of a second generation of instruments, using capacitive or optical sensors to measure the local mass or density (by transparency) of a sample of fibres, and evaluating the fibre length distribution by analysis of the corresponding signal. These instruments use well defined samples of fibres, constituted by a relatively thin layer of parallel fibres representative of the total population of fibres in the lot. Two types of samples are preferred. The first one, called a "numerical sample", contains parallel fibres all starting from the same origin line, perpendicular to the direction of the fibres.The second one, called a "sample in cross-section" or "length biased sample", is obtained by gripping a sliver or other assembly of parallel fibres in a transverse clamp, and removing, by combing, all the fibres which are not gripped in this clamp. The present invention also preferably employs one of these two types of samples, and the "numerical sample", which has several advantages leading to a more accurate determination of the parameters of the length distribution is especially preferred.

The sample being rather thin can be considered as lying in a plane.

In order to obtain a measure of the full fibre length distribution, it is necessary, in electonic instruments using either optical or capacitive sensors, to obtain a signal corresponding to the local optical density or local mass along the whole length of the sample. In the known instruments, this is done by using a sensor arranged to be transverse to the direction of the fibres, and having a dimension in the direction of the fibres which is very small compared with the mean fibre length. As a result, the sensor can be considered to approximate to a transverse line across the plane of the parallel fibres.A signal for the whole length of the sample is then obtained in the existing instruments by causing, by the use of mechanical means, a relative displacement between the sample and to the transverse sensor, amounting to a mechanical scanning movement of the sensor along the whole length of the sample. In the known optical instrument most used for cotton and short fibres, this is done by keeping the sample of fibres fixed in the clamp, and moving an optical sensor along the length of the fibres. In the known instrument used for wool and long fibres, the sample is held between two plastic sheets, and moves at constant speed through a transverse capacitive sensor. The mechanical scanning takes in both known instruments a relatively long time of 20 to 60 seconds, depending on the maximum length of the fibres.

The measurement of the fibre length distribution in non-homogeneous raw materials, in all the components of a blend (farm lots, for instance) and at several stages in processing of the fibres, will require in the future a very large number of determinations of the fibre length averaged over many samples. It is thus necessary to speed up and automate the measurement as much as possible.

The present invention in a preferred form provides a new sensor arrangement characterised by one or both of the following principles: 1) the sensor covers the whole area of the sample or of an optical image of the sample instead of having a narrow dimension in the direction of the length of the fibres 2) the sensor arrangement contains a relatively large number of sensing elements, which are electronically scanned in a very short time, for example, of the order of a millisecond.

The sensing elements of the sensor arrangement may be electro-optical, or capacitive. For both types of sensor arrangement, the thin sample is contained in a plane, and hereinafter the direction parallel to the fibres will be called the x direction, while the direction perpendicular to the x direction and in the plane of the sample will be called the y direction.

Where an electro-optical sensor arrangement is used, the whole area encompassing the sample is uniformly illuminated by means of suitable sources of light, reflectors, condensers and, possibly, diffusing surfaces. The illuminated object plane containing the sample is then imaged by an appropriate objective lens. The method of illumination can be either brightfield or darkfield depending on the relative positions of the light sources, the object plane and the objective lens. An electro-optical sensor arrangement is placed in the image plane in order to analyse the image of the sample to allow the determination of the length distribution.In the brightfield mode the image plane is uniformly bright in the absence of a sample, while the presence of a fibre at a given (x, y) position in the object plane gives rise to a decrease in the light intensity at the corresponding position in the image plane. On the other hand in the darkfield mode the presence of a fibre gives rise to an increase in the light intensity which is in principle zero without fibres. in both cases the light intensity variation due to the presence of the sample has to be measured at every point in the image plane.

With the capacitive arrangement, the sample of fibres is inserted between the two plates of the sensor arrangement, which covers the whole area of the sample itself.

Some embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings, in which: Figure 1 is a view of one form of apparatus according to the invention, employing brightfield illumination.

Figure 2 is an optical ray diagram for the apparatus of Fig. 1; Figure 3 shows the geometry of the sample together with the coordinate frame; Figure 4 shows one possible arrangement of a detector relative to the image of the sample; Figure 5 shows another possible arrangement of a detector relative to the image of the sample; Figure 5a shows an alternative method of illumination of the sample; Figure 6 is a view of another form of apparatus according to the invention, employing darkfield illumination; Figure 7 is an optical ray diagram for the apparatus of Fig.6; Figure 8 shows one possible arrangement of a light trap used with the apparatus of Fig. 6; Figure 9 is a view of a further form of apparatus according to the invention, capable of operating with either brightfield or darkfield illumination;; Figure 10 is an optical ray diagram for the apparatus of Fig. 9; Figure ii illustrates schematically a measuring capacitor arrangement, having an array of electrodes, of an embodiment of the invention; Figure 12 is a circuit diagram of circuits for coupling a measuring capacitor arrangement to a computer; and Figure 13 is a circuit diagram of alternative circuitry for coupling a measuring capacitor arrangement to a computer.

In the apparatus shown in Figs. 1 and 2, a source of light is provided which consists of a tubular lamp 1 with an effectively straight filament disposed in the x direction. A cylindrical reflector 2 is placed below the lamp 1 whose filament coincides with the cylinder axis of the refelctor 2. An elongate condenser 3 is placed above the lamp and may consist of a cylindrical lens or may have a more elaborate profile in order to get a uniform illumination in the y direction beyond the condenser. The condenser 3 is so positioned that the emerging beam 4 of the light is concentrated on a sheet of transparent glass 5 or any other transparent material placed in the x-y plane at some distance from the condenser, this plane lying perpendicular to the optical axis.A thin sample 6 consisting of fibres parallel to the x direction and each having one end aligned along a line parallel to the y direction and defining the y axis is placed on the transparent sheet 5. The lighting device, comprising the lamp 1, the reflector 2 and the condenser 3, is arranged to provide uniform brightfield illumination of the sample 6. The area of the uniformly illuminated zone containing the sample will be called the illumination area. By way of example, when measuring fibres like wool which can be as long as 25cm, the length of the illumination area should be 25cm or more. A corresponding width of at least 5 to 1 Ocm is required in order to allow the sample to contain a statistically representative number of fibres without being too dense.Along the optical axis which extends in the z direction, and centered with respect to the illumination area, at a distance d1 above it, is placed the objective lens 7. In a preferred embodiment, the image of the filament of the lamp 1 through the condenser 3 is formed at the z coordinate where the objective lens 7 lies. The sample is imaged by the objective lens 7 in a plane 8 called the image plane, at a distance d2 above the objective lens 7.

A photodetector (not shown) is placed with its sensitive surface in the image plane 8.

Different types of photo detectors and different arrangements of photo detectors can be used, leading to different types of subsequent analysis of the detected signals.

When there are no fibres in the illumination area, the image plane 8 is uniformly illuminated.

The presence of fibres at a given position in the illumination area leads to a decrease of the light intensity at the corresponding positon in the image plane 8.

Let the quality of fibre at any position (x, y) in the object plane be described by the function s(x, y). The x axis is defined by an arbitrarily selected line at one side of the sample in the object plane. The fibres lie parallel to the x axis, as shown in Fig. 3, with their ends aligned at the origin of the x axis, so that the x coordinate corresponds to the length coordinate. Integrating s(x, y) along the y axis gives the function:

where the limits of the integral correspond to the full width of the samples. F(x) is tho cumulative length distribution.The length distribution (frequency distribution) f(x) can bo calculated therefrom:

or equivalently

Considering now the image plane, the same set of coordinates can be used since the sample and its image are homothetic.

What is to be measured by the photo detector is the illuminance E(x, y) at any point (x, y) of the image plane.

In a simple approach, satisfactory for the description of the operation principle of the instrument, the illuminance is given by: E(x, y) = E0 (1 - s(x, y)) where E0 is the uniform illuminance in the absence of sample. For low values of s(x, y), namely for a thin sample, this formula corresponds to Lambert's law of absorption. Integrating the measured illuminance along the y axis we get:

hence:

w being the width of the sample image or more precisely the width of the image of the illumination area. In one embodiment of the invention, the photo detector consists of a tube from a standard video camera. In that case, the photo detector is arranged so that the whole frame covers the image 9 of the illuminated area, with the video lines parallel to the y axis, namely perpendicular to the fibres.Preferably the scanning of the sensitive area of the camera tube is effected sequentially, i.e. without interlacing. Each video line thus corresponds to a particular x value. By integrating the video signal along each line and memorizing the results individually for each line, we obtain in a discrete form:

= 1 m k being the proportionality constant between the detected electrical signal and the correspond ing illuminance, and m being the number of lines covering the length of the image of the illumination area. The distribution F(x) is then calculated directly from G(x,). By way of example, using 500 lines from a standard TV camera tube to measure samples up to 25cm long will give a length distribution with classes of 0.5mm.

In another embodiment of the invention, the detector consists of a matrix array of semiconductor detector elements (photodiodes or CCD (charge-coupled device)) comprising n rows along y and m columns along x. The detector elements are scanned element by element, row after row, giving for each element an electric signal directly proportional to the corresponding illuminance. With this detector, the signals from each element in a row are summed and memorized, giving:

from where the distribution F(x,) is immediately obtained. By way of example, m and n can be as great as 576 and 384 respectively; for textile fibres generally, m = 244 and n = 1 90 is suitable.

In another embodiment of the invention, the detector consists of linear array of m adjacent semiconductor detector elements (photodiodes or CCD) which are scanned successively at an adjustable frequency, giving successive electrical signals porportional to the corresponding illuminance. In one form of the apparatus, the detector 10 is arranged parallel to x and thus sees one line of the sample parallel to the fibres as shown in Fig. 4. Each detector element 11 corresponds to a particular x value. As the image of the sample and the detector are moved relative to each other in the y direction, the detector successively sees all the fibres in the sample. The mutual displacement of the image and the detector can be done in different ways.

In one embodiment of the invention, the displacement is obtained by moving the sample itself along the y direction. In another embodiment of the invention, th displacement is done by moving the detector along the y direction. In practice, the latter arrangement is preferred to the former.

A typical detector comprising m = 256 elements is 6.25mm long. When measuring samples up to 250mm long, the linear sizes of the image must then be 40 times smaller than those of the sample. The required mechanical displacement is thus 40 times smaller in the second arrangement than in the first one and can therefore be much faster. In still another embodiment of the invention, the mutual displacement is obtained by placing a mirror or prism in the path of the image-forming rays and rotating it about the x axis.In any of these arrangements for producing mutual displacement of the image and the detector, the relative displacement during one scan period of the detector has to be equal to the width of the sensitive surface 11 of the detector 10, which is typically 25cm. In this way the whole image will be scanned without overlapping of the successive scans of the detector. Let n be the number of scans of the detector necessary to cover the whole image. At the jth scan one measures successively: kE(xj, yj) where i = 1 ...., m At each scan, the measured signal is accumulated with the previous one, individually for each of the m elements. Finally:

where i= 1,...., m which yields again the cumulative length distribution F(xi).

In another form of the apparatus, the same type of detector is used, but arranged perpendicular to the fibres, parallel to the y axis as shown in Fig. 5. The detector 1 2 thus sees one line of the sample correspodning to a particular x value. A mutual displacement of the image and the detector is performed by one of the methods according to the previously described form of the apparatus, with the same considerations concerning the speed of the i displacement, the scan period of the detector and its width.With this arrangement of the detector, the analysis proceeds as follows: during one scan period, corresponding to a particular value xi, the sum of the values measured for each element of the detector is computed and memorized, this being

Doing the same for each xj value yields the distribution: kG(xj) where i= 1,...., n and hence F(xj), where n is the number of scans of the detector necessary to cover the whole image.

In both forms of apparatus using a linear array of photodetectors, i.e. with the array aligned either along the x axis or the y axis, another form of optical arrangement can be used advantageously for the sake of simplicity. It is observed that when a linear array is used, the whole area of the sample need not be simultaneously illuminated. Only one line at a time on the sample, corresponding to the line seen by the detector array, need be illuminated. Consequently the whole system of illumination as shown in Figs. 1 and 2 can be replaced by a simpler one as shown in Fig. 5a. A narrow beam of light 36, preferably from a laser head 37, passes through a cylindrical lens 38 whence it forms a thin diverging beam of light 39. The orientation of this beam is such that the beam illuminates a line 40 on the sample 41 in the direction y in the example of Fig. 5a.At each point (x, y) on that particular line, the illuminance E(x, y) beyond the sample is related to the local sample density S(x, y) as before. The values of E(x, y) corresponding to the illuminated line can then be directly measured by a linear detector array 42 placed beyond the sample and covering the whole width of the sample. As this width can be larger than currently available linear detector arrays, means are preferably used to concentrate the light rays passing through the sample onto a detector array smaller than the width of the sample. In other words, the thin beam of light is made convergent rather than divergent beyond the sample. The means used for this purpose can simply be a converging lens, which is preferably a Fresnel lens which is light and thin, placed just below or just above the sample.

Another means consist of a row of light guides, preferably optical fibres, picking up the light passing through the sample and conducting this light to a row of detectors, each light guide corresponding to an individual detector, and the detectors being electronically scanned in the same order as the ends of the light guides occur at the sample.

As the values of E(x, y) have to be determined over the whole area of the sample, relative displacement of the sample with respect to the optical system, along the x axis, is effected by suitable means.

In another form of apparatus, the same system of illumination and detection is used with the thin diverging beam of light illuminating a line on the sample but parallel to the x axis (i.e. to the lengths of the fibres). In that case, the orientation of the linear detector array and the direction of the relative displacement are accordingly adapted.

In any form of the apparatus working in the brightfield mode as described above, except for the examples in which a thin diverging beam of light is used to illuminate a line on the sample, a sheet of translucent material, e.g. opal glass, can be added to the system between the condenser 3 and the sample 6. The presence of this diffusing material helps to provide very uniform illumination in the plane of the sample. However its effect is also to reduce drastically the illuminance of the image 8. In addition it also reduces the contrast between the fibres and the illuminated background, that is, the visibility of the fibres. This visibility depends among others on the distance between the diffusing plate and the sample: the smaller the distance, the lower the visibility.Advantage can be taken of this property by adjusting the position of the diffusing sheet so as to match the range of light levels in the image with the dynamic range of the detector sensitivity. In some embodments of the invention, the sheet of glass 5 supporting the sample can be the diffusing sheet itself. When not using a diffusing sheet in any form of the appartus described above, a neutral density filter or a colour filter can be placed between the condenser and the sample to adjust the illuminance to the sensitivity of the detector. An infrared-absorbing filter can also be used to avoid excessive heating of the sample.

The apparatus shown in Figs. 6 and 7 corresponds to the darkfield configuration. In this form of the apparatus a thin sample of fibres 1 3 is placed on a sheet of transparent glass 14, the fibres all lying parallel to an x axis with one end aligned along a y axis, perpendicular to the x axis.Sources of light, there being preferably two or more, are placed above the sample in such a way as to produce a resultant uniform illumination of the sample 1 3. In a preferred embodiment of the invention the sources should consist of a tubular lamp 1 5 with an effectively straight filament, a cylindrical reflector 1 6 placed behind the lamp 15, with the axis of the cylinder coinciding with the filament, and a condenser 1 7, for example cylindrical lens, placed in front of the lamp, the axis of these elements all being parallel to the x axis.The sources should be arranged in such a way that the emerging light beams 1 8 just encompass the sample, with each beam axis 1 9 inclined with respect to the z axis so that no specular reflection on the glass sheet 14 occurs in the z direction of neighbouring directions. An objective lens 20 is placed a distance d1 above the sample along the z axis and centered on the illumination area where the sample 1 3 lies. This objective lens 20 collects the backscattered light 21, comprising reflection and refraction from the fibres, to image the sample in the image plane 22 located at a distance d2 above the objective lens 20. In the image plane, the sample thus appears bright on a dark background.In order to obtain a background as dark as possible, the direct light 1 8 from the sources passing through the transparent glass sheet 14 has to be prevented from returning upwards. This can be done by placing a light trap 23 below the transparent glass sheet. Fig. 8 shows the vertical section of an example of light trap consisting of a box with black-painted absorbing walls 24 arranged so that light rays 25 entering the box can come back upwards only after a great number of reflections at each of which they are strongly attenuated.

Using the same formalism as for the brightfield configuration described above, the illuminance at position (x, y) in the image plane is now expressed by:

where : Eó is the constant illuminance in the object plane, a is a constant taking into account the ratio of the illuminance in the image plane to the corresponding illuminance in the object plane, s(x, y) is a function proportional to the quantity of fibre at the position in the object plane corresponding to the position (x, y) in the image plane (the correspondence between these positions is homothetic due to the imaging properties of the lens).

measuring E(x, y) everywhere in the image plane and integrating along y yields:

which gives the cumulative length distribution F(x) Compared with the brightfield configuration, it is seen that the wanted distribution F(x), and hence f(x), can be obtained in both case from G(x) with a different relation between F(x) and G(x) in either case. Thus, all the methods of detection used in the brightfield mode to obtain G(x) (or a discrete form of this function G(xi)) are applicable to the darkfield mode.

Therefore, according to the invention, the form of apparatus corresponding to Figs. 6 and 7 as described above can be combined with any type of detector and its corresponding arrangement(s) as described above for the form of apparatus operating in the brightfield mode.

The corresponding methods of analysis of the measured signals have simply to be accordingly modified, due to the differences between equations (1) and (2).

In the darkfield configuration, no translucent material can be placed behind the sample contrary to the brightfield configuration. According to the invention, as alternative form of apparatus, shown in Figs. 9 and 10 can operate either in brightfield or. darkfield mode by a slight modification. In this configuration, a thin sample of fibres 26 lies on a transparent sheet of glass 27 with the fibres all parallel to an x axis, and one end of the fibres aligned on a y axis, perpendicular to x. Sources of light, there being preferably two or more, are placed below the sheet of glass 27 in such a way as to produce a resultant uniform illumination of the sample.In a preferred embodiment of the invention the sources consist of a tubular lamp 28 with an effectively straight filament, a cylindrical reflector 29 placed behind the lamp, with the axis of the cylinder coinciding with the filament, and a condenser 30, for example a cylindrical lens, placed in front of the lamp 28, the axis of these elements all being parallel to the x axis. The sources are arranged in such a way that the emerging light beams 31 just encompass the sample 26, with the beam axis 32 inclined with respect to the z axis so that no direct ray 31 passing through the sheet of glass 27 propagates in the z direction or neighbouring directions.

An objective lens 33 is placed a distance d1 above the sample along the z axis and centered on the illumination area where the sample lies. This objective lens collects the forward scattered light 34, comprising reflection and refraction from the fibres, to image the sample in the image plane 35 located at a distance d2 above the objective lens. This system thus operates in the dark field mode: the sample appears bright on a dark background. However, by simply placing a diffusing sheet, e.g. opal glass, below the sample 26, between the light sources and the sheet of glass 27, the system operates in brightfield mode. In some embodiments of the invention, the sheet of glass 27 can be the diffusing sheet itself.

In both configurations of the present form of the apparatus, namely with or without a diffusing sheet, corresponding respectively to brightfield or darkfield mode, any of the detectors and their respective arrangements as described hereinbefore with reference to Figs. 1, 2 6 and 7 can be used together with the corresponding methods of analysis. In one form of apparatus according to the invention two or more detectors can be arranged to allow the simultaneous detection of the image of the sample in brightfield and darkfield modes. For instance, with reference to the apparatus shown in Figs. 1 and 2, the objective lens 7 images the sample in brightfield mode, in the image plane 8.However, simultaneously, an additional objective lens looking at the sample 6 with its optical axis inclined with respect to the z axis, both axes being in the same y-z plane, would image the sample in darkfield.

The optical embodiments described hereinbefore have the advantage that they allow measurement of the fibre length distribution of blends contain antistatic or conductive fibres. However, a capacitive measuring system delivers a better defined signal, proportional to the mass of the fibres contained between the electrodes, and is much less sensitive to several factors such as the thickness of the sample, evenness of the spreading of the sample, and the diameter and colour of the fibres. In the majority of cases, where no conductive fibres are used, a capacitive sensor is therefore preferred. However, in order to obtain a universal instrument, a better solution is to use both types of sensor on the same sample and to compare the two signals by an algorithm detecting anomalies.For example, the presence of antistatic fibres will result in the presence of peaks in the histogram delivered by the capacitive sensor. Automatic selection of the best signal or a combination of the two signals to produce an improved signal is thus possible.

A capacitive sensor designed to provide an nearly instantaneous measurement of the complete fibres length distribution is shown in Fig. 11. The sensor consists of two plates, preferably rectangular, situated in two planes parallel to the plane of the sample, one being above and the other below the sample during a measurement. Each of these plates covers the entire area of the sample, so that the longest fibre is smaller than the length of the plate. One of these plates (represented in Fig. 11) is divided into a relatively large number of rectangular strips. Each of these strips has a dimension in the x direction of the fibres which is much smaller than the mean length of the fibres, for example, 5mm for wool fibres or similar long fibre material. The other dimension is chosen to cover the whole width of the sample.Each of these rectangular strips is made of electrically conductive material and is insulated from the other strips. For instance, the plate can be made in the form of a printed circuit, where an epoxy substate is used as insulation material and the strips are formed as thin copper layers. The second plate, on the opposite side, can be a single large rectangular metal plate (or a rectangular copper layer on a printed circuit board) having approximately the same external dimensions as the first plate divided in strips.

The second single plate is fed preferably with high frequency current by a generator. The strips of the first plate are successively connected to high frequency amplifiers through an analog multiplexer switch. At a given time, the strip number i acts as a measuring electrode; the two neighbouring strips numbered (i - 1) and (i + 1) act as guard electrodes. The successive strips from i = 1 to i = 50 for instance, are scanned in this way. In a preferred electronic circuit arrangement, an additional signal compensation electrode, having approximately the same area as one of the rectangular strips, but not necessarily the same linear dimensions, is deposited to one side of the first plate close to the strips. The purpose of this compensation electrode is to correct for temperature and humidity variations of the atmosphere.

In one form of apparatus according to the invention, an analog switch array is arranged so that the measuring capacitor elements are connected one at a time to a bridge as depicted in Fig. 1 2. A high frequency voltage generator 43 drives the bridge. At the output of the transformer 44 a signal is obtained whose amplitude is related to the difference between the capacities of the compensation capacitor 49 and the measuring capacitor element which is connected to the circuit at that time. This signal is amplified by an amplifier 45 and fed to the input of an amplitude detector 46. The detection is carried out preferably by means of either a rectifier followed by a low-pass filter, or by mixing the transformer output signal with a signal of same frequency and fixed amplitude, and low-pass filtering the mixed signals. The detector output is connected to an analog-to-digital converter 47 to obtain digital signals which are processed by the computer system 48.

In another form of apparatus according to the invention an analog switch array is arranged as illustrated in Fig. 1 3 so that any one of the measuring capacitor elements or the compensation capacitor 50 can be connected to either a fixed voltage generator 51 or a current integrator 52.

In the first step of operation, all the switches, except switch 56, are closed so that each capacitor is charged to a fixed voltage, determined by the generator 51. In the second step, all the capacitors are discharged one by one successively through the integrator system 52 by closing the switch 56 and then successively closing the respective switches of the measuring elements. Since the output of the integrator system is proportional to

i being the discharge current and

the capacity of each capacitor is obtained from

where

is a charge variation in a capacitor of capacity C and

the corresponding voltage variation.

An improvement in the electric field configuration is obtained if the second plate is also formed of strips, switched synchronously with the strips of the first plate, but at the expense of a more elaborate circuitry.

The increase in the capacity of each of the individual measuring capacitor elements, due to the presence of the fibres is proportional to the mass of the fibres contained in this individual capacitor element, formed by strip number i for example. After the high frequency amplifiers connected to bridge circuits formed by each strip and the compensating electrode, a demodulator delivers a signal proportional to the mass of the portion of fibres contained in each individual capacitor element.

The electronic scanning from i = 1 to i = 50 provides a signal corresponding to a cumulative fibre length distribution biased by cross-section.

In both optical and capacitive sensors, the principle of construction involving a large number of sensitive elements used in succession by electronic scanning to build up the complete fibre length distribution, creates a problem due to the differences in electrical level and sensitivity between these elements. These differences are relatively significant for elements of photodiode or CCD linear arrays and matrices, but also appear between signals produced by individual capacitor strips. A procedure is therefore provided to compensate for these difference, involving a periodical automatic calibration of each element and the correction of the signals produced by each individual element, by correction factors obtained during the calibration procedure and stored in the RAM of a computer which controls and reads the sensor.

At a period of one day for example, or every time the instrument is switched on, a calibration procedure is automatically started before the first measurement, without sample in the sensor.

For the optical sensors, at a fixed level of illumination corresponding to the one used during the measurements, the "zero level" signals of each photodetecting element, corresponding to the absence of fibres, is recorded and stored in memory in digital form. The illumination level is then reduced by say 50% and the signal produced by each element is again recorded and stored in memory. The comparison of the signals obtained at the two illumination levels defines the sensitivity or gain of each element.

During the measurement of a sample, the signal from each element will be corrected on the basis of the two sets of data stored in memory, using a linear transformation. Basically, this transformation amounts to a substraction of the "zero level" or "white level" and to a multiplication by a gain factor for each element.

For the capacitive sensor, the "zero level" is also recorded and stored for each capacitor strip connected in the bridge, in the absence of the sample. The "sensitivity factor" is obtained by carrying out a second calibration run with a uniform dielectric sheet inside the plates of the capacitor, giving a reference signal which is stored in memory for all elements.

For a optical sensor the number of elements in a matrix is relatively high. To reduce the memory size, the calibration signal corresponding to the elements of a line transverse to the direction x of the fibres can be average before storage in memory.

In those cases where the sample doesn't move relative to the sensor during the measurement, the total measurement time being very short, of the order of one millisecond, the sample is allowed to move during the measurement period at a speed not exceeding 10 to 30 m./min when the classes of the length distribution are at least 1 mm long. It is for instance possible to place the sensors directly on the mechanical preparation machine where the samples are formed, and to make the measurements in a zone where the samples are transferred.

By using a computer to control the sensor and process the signals generated thereby in apparatus embodying the present invention it is possible to provide very fast measurement of the fibre length distribution of a sample of textile fibres. The computer can be a microprocessor based unit with a control program dedicated to the measurement to be carried out. Thus the complete apparatus can be very compact.

Claims (21)

1. A method of measuring fibre length distribution, the method comprising the steps of: disposing a sample of fibres on a plane with the fibres lying substantially parallel to one another, scanning an area of the said plane containing the whole sample so as to produce electrical signals representative of the presence or absence of fibre at a sequence of points or elemental areas forming the said area, and processing the said signals so as to produce therefrom a representation of the fibre length distribution of the fibres in the sample, the said scannng being, in at least one direction in the said plane, carried out by electrical switching of a sequence of scanning elements.

2. A method according to claim 1, wherein the scanning is effected electro-optically, an image of the said area being projected onto a plane at which electro-optical sensing means are provided.

3. A method according to claim 2, wherein the electro-optical sensing means has an area of sensing elements and the scanning is carried out by electrical switching of successive sensing elements, the area of sensing elements covering the said image.

4. A method according to claim 1, wherein the scanning is such as to produce electrical signals representative of quantity of fibre when fibre is present.

5. A method according to claim 4, wherein the scanning is carried out by electrical switching of capacitive sensing elements, and the said signals are representative of quantity of fibre or absence of fibre at a sequence of elemental areas each of which is defined by a respective one of the capacitive sensing elements.

6. Apparatus for use in measuring fibre length distribution, the apparatus comprising: means for supporting a sample of fibres on a plane with the fibres lying substantially parallel to one another, means for scanning an area of the said plane containing the whole sample so as to produce electrical signals representative of the presence of absence of fibre at a sequence of points or elemental areas forming the said area, and means for processing the said signals so as to produce therefrom a representation of the fibre length distribution of the fibres in the sample, the said scanning means including electrical switching means adapted to enable scanning to be carried out, at least in one direction in the said plane, by electrical switching of a sequence of scanning elements.

7. Apparatus according to claim 6, wherein the scanning means includes optical means for projecting an image of the said area onto a plane at which the scanning means includes a plurality of electro-optical sensing elements.

8. Apparatus according to claim 7, wherein the scanning means includes an area of the said electro-optical sensing elements and the switching means is arranged to activate each of the sensing elements in succession, the said area of sensing elements covering the said image.

9. Apparatus according to claim 6, wherein the scanning means is such as to produce electrical signals representative of quantity of fibre when fibre is present.

10. Apparatus according to claim 9, wherein the scanning means includes a plurality of capacitive sensing elements, each capacitive sensing element defining a respective elemental area of the said area of the plane.

11. Apparatus according to claim 10, wherein the capacitive sensing elements are defined by parallel strips of conductive material lying in a plane spaced from the means for supporting the sample, the strips being arranged alongside one another and extending transversely of the direction of the lengths of fibres in a sample.

1 2. Apparatus according to claim 7, wherein the optical means includes at least one source of light arranged to illuminate the said area of the said plane, the supporting means comprising a sheet of transparent or translucent material.

1 3. Apparatus according to claim 12, wherein the supporting means comprises a sheet of transparent material and the source or sources of light are arranged to provide darkfield illumination of the sample.

14. Apparatus according to claim 12, wherein the optical means includes an optical objective arranged to provide said image for brightfield illumination of the sample.

1 5. Apparatus according to claim 6, wherein the means for processing the said signals includes means for applying corrections to the said signals to compensate for non-uniformity in signals obtained from the scanning means in the absence of a sample.

1 6. Apparatus according to claim 15, wherein the scanning means includes optical means for projecting an image of the said area onto a plane at which the scanning means include a plurality of electro-optical sensing elements, and the said means for applying corrections is adapted to apply a correction for non-uniformity of illumination of the said area and a correction for non-uniformity in the responses of the electro-optical sensing elements.

1 7. Apparatus according to claim 10, wherein a compensating capacitive element is coupled to the processing means which is adapted to utilize signals from the compensating capacitive element to correct for environmental variables affecting samples.

1 8. Apparatus according to any one of claims 6 to 16, wherein the processing means comprises digital computing means.

19. A method according to claim 1, wherein the processing of the said signals includes applying corrections for non-uniformity in corresponding signals obtained by scanning in the absence of a sample.

20. A method of measuring fibre length distribution, substantially as described herein before with reference to the accompanying drawings.

21. Apparatus for measuring fibre length distribution, substantially as described hereinbefore with reference to Figs. 1 and 2, or Figs. 4 and 5, or Figs. 6, 7 and 8, or Figs. 9 and 10, or Figs. 11 and 12, or Figs. 11 and 1 3 of the accompanying drawings.