EP0074447B1

EP0074447B1 - Apparatus and method for sorting articles

Info

Publication number: EP0074447B1
Application number: EP19810304231
Authority: EP
Inventors: Reginald Harold Clark; John Duncan Macarthur; Michael Sayer; William David Wilder
Original assignee: RESOURCE RECOVERY Ltd
Current assignee: RESOURCE RECOVERY Ltd
Priority date: 1981-09-15
Filing date: 1981-09-15
Publication date: 1987-01-07
Also published as: EP0074447A1; DE3175795D1

Description

This invention relates to an apparatus and method for sorting articles and more particularly for sorting mixed metal pieces dependent on the type of metal.
Methods have previously been proposed whereby articles have been sorted manually as they progressed along a conveyor belt. Once identified, such articles would be manually removed from the conveyor belt and deposited in appropriately identified receptacles. A method and apparatus is known for the separation of uranium bearing rock and this consists of a vibratory feeding mechanism together with a translucent conveyor belt. A light source device is provided to measure the rock size together with a radio-active counter which meaasures the radiation rate from each rock. From the measurements, a product of the rock size and radiation rate is computed electronically and a signal is produced to cause actuation of air jets which separate the rocks into two categories at the end of the conveyor belt. Attempts have been made to utilize this apparatus for sorting other items, such as pieces of scrap metal, into different categories but such attempts were not successful.
Apparatus is known for sorting mixed metals using differential melting techniques. It is believed that this process is relatively inefficient and consumes large amounts of energy.
As it will be appreciated, apparatus for sorting scrap metal would be particularly attractive from a commercial point of view having regard to the amount of scrap metal which is presently located in different scrap metal yards as, for example, an end product of the automobile industry.
U.S. Patent Specification 3,233,720 describes a conveyor apparatus having support members extending transversely of the conveying path and controlled for movement between a supporting position and a non-supporting position to discharge articles from the conveyor.
An object of the present invention is to provide apparatus for sorting objects which is applicable to the sorting of articles such as pieces of scrap metal and in which the above-mentioned disadvantages are obviated or substantially reduced.
According to this aspect, there is provided conveyor apparatus for sorting a plurality of articles fed sequentially thereto, of the type including a conveyor for conveying the articles along a predetermined path in sequential order, scanning means adjacent the conveyor arranged to scan each article sequentially and provide an output signal corresponding to a selected characteristic of the article and control means arranged to direct a selected article in response to said output signal along a flowpath selected from a plurality of flowpaths branching off from said predetermined path, characterized in that:

(a) said conveyor includes a plurality of spaced apart support members extending transversely of said predetermined path and being arranged, on the one hand, for sequential movement along said path, and on the other hand for movement between an article supporting position and an article non-supporting position;
(b) sensor means are provided to determine the number of support members supporting each article, to thereby establish the length of an article;
(c) means are provided to determine the position of each article, relative to a reference position fixed in relation to said conveyor, as a function of the number of support members from the article to the reference position; and
(d) the control means is arranged to select one of the flow paths for a selected article by triggering movement of the members supporting the selected article from the supporting to the non-supporting positions in response to said output signal and the length of the article when said article arrives at the selected receiving flowpath.

Another object of the present invention is to provide a method of sorting objects which is particularly applicable to the sorting of scrap metal and in which the above-mentioned disadvantages are obviated or substantially reduced.
According to this aspect there is provided a method of sorting articles according to a selected characteristic thereof, wherein the articles are fed sequentially along a predetermined path by means of a conveyor and wherein each article is scanned as it moves along the conveyor so as to generate an output signal corresponding to the selected characteristic, which signal is fed to a control means arranged to direct a selected article. in response to said output signal along a flowpath selected from a plurality of flowpaths branching off from said predetermined path characterized by:

(a) conveying the articles to be sorted sequentially by means of a plurality of spaced apart support members extending transversely of said path and being arranged, on the one hand, for sequential movement along said path and, on the other hand, for movement between an article supporting position and an article non-supporting position;
(b) determining the length of each article as a function of the number of support members supporting it;
(c) measuring the position of each article with respect to a reference position as a function of the number of support members from the article to said reference position; and
(d) directing a selected article to a selected one of the flowpaths by triggering through said control means movement of the members supporting the selected article from the supporting to the non-supporting positions in response to said output signal and the length of the article when said article arrives at the selected flowpath.

Embodiments of the present invention will now be described, by way of example, with reference to the accompanying drawings in which:-

Figure 1 is a diagrammatic representation, in plan view, of one embodiment of an apparatus for sorting scrap metal;
Figure 2 is a side view of the apparatus illustrated in Figure 1;
Figure 3 is a plan view on an enlarged scale of part of the apparatus shown in Figure 1 so as to illustrate details thereof;
Figure 4 is a cross-sectional view of part of Figure 3 taken on the line IV-IV;
Figure 5 is a diagrammatic representation to show the use of an X-ray fluorescence unit;
Figure 6 is a block schematic representation of the electronic control circuits for the apparatus illustrated in Figure 1;
Figure 7 is a more detailed block schematic diagram of part of the electronic control circuits;
Figure 8 is a schematic outline of a software program for the apparatus of Figure 1;
Figure 9 is a diagrammatic representation of light-emitting diode sources and associated optical detectors in the light head.
Figure 10 diagrammatically illustrates the solenoid driver stages; and
Figure 11 is a diagrammatic representation of the power circuit of the solenoid stages.

Referring to Figures 1 and 2, there is diagrammatically illustrated an apparatus for sorting scrap metal. The mixed pieces of scrap metal travel along a feed conveyor belt system 2 of any conventional design onto a sorter conveyor system 4 arranged, for convenience, in a circular manner as illustrated in Figure 1. It is important that the pieces of metal should fall onto conveyor 4 piece by piece so that at least some linear separation between pieces is achieved, as scanning and discrimination between pieces is difficult if not impossible if material is fed too quickly so that heaps or piles of pieces occur. The circular or carousel conveyor 4 comprises a plurality of individual members or keys adapted to support the pieces of scrap metal fed thereon in sequential order from the conveyor belt 2. In use, the carousel conveyor 4 in Figure 1 moves in a clockwise direction. Thus, each piece of scrap metal, separated from each succeeding piece is supported by one or more members 6, dependent on its size, and passes, first of all through an overhead detection unit 7 and then through a plurality of vertical light beams emanating downwardly from a light head, unit 8, where the size of the piece may be determined. Information signals as to both the length and width of the piece and its presence on the carousel conveyor are fed to a computer unit as described below.
After passing the light head, unit 8, the respective piece of metal material continues along the carousel conveyor and under an X-ray fluorescence or other scanning unit 10. This unit produces an output signal representative of the elements present in the scrap metal and passes this information to the computer unit which then analyzes all information signals received and produces resultant output control signals. These resultant output control signals are dependent on the type of elements determined to exist in the piece of scrap metal and also on the size of the piece of metal. The computer's output signals are fed to a selected one of a plurality of exit stations, dependent on the type of metal. The exit stations are identified in Figure 1 as stations 12, 14, 16, 18 and 20. Each station is adapted to receive scrap metal of a particular type, for example, iron, brass, zinc or aluminum. At the station where aluminum is deposited a metal detector is provided beneath the members 6. Only if the piece of scrap material is determined to be metallic, is the piece deposited here. Consequently, non-metallic pieces continue along the carousel conveyor to a discard bin 22.
The construction of the carousel conveyor 4 will now be considered in greater detail, particularly having regard to the construction of the individual members or keys 6. Referring to Figure 1, the carousel conveyor consists of a circular wheel, or table, 24 mounted in a horizontal plane and carrying a plurality of metal plates, such as 26, rigidly mounted around the periphery of the table 24. Each metal plate 26 supports a group of nineteen individual members in a manner which will be described in greater detail with reference to Figure 3.
Each individual member 6 consists of a key made of plastic or other material producing X-rays or other identifying signals which do not interfere with the identification process. The key which is about ten inches long and a quarter-inch square cross section for convenience. Each key is supported at its inner end on the respective metal plates 26 in a pivotal manner by means of a metal rod 28. One such rod is shown in Figure 3 in a remote location so as to indicate how it would be inserted through an aperture in the respective member 6 and aligned apertures in finger portions 30 and 32 on either side of the respective member 6. Thus, each key is supported at its inner end so that it can rotate about the respective metal rod 28.
The other end of the key portion 6 is normally supported by a smooth metal plate 34 which extends around the outer periphery of the carousel conveyor 4. Thus, the other end of each member 6 can slide over a smooth metal plate 34 during normal rotation of the carousel.
At the various exit stations 12 through 22, the continuity of the smooth metal plate 34 is interrupted. The interruption is filled by a slidable metal plate 36 (Figure 3) which can be retracted under control of the solenoid device 38 so as to cause the free end of respective keys 6 to drop as the key rotates about its metal rod or pin 28. .Referring particularly to Figure 4, it will be seen that the solenoid device 38 comprises a solenoid coil unit 40 having a movable armature 42. Attached thereto is a rod 44 which supports the slidable metal plate 36 in the manner illustrated. Energization of the coil unit 40 causes the armature 42 to move in the direction A pulling the metal plate 36 with it and allowing the respective key 6 to rotate as described above. However, as soon as current is removed from the coil unit 40, the spring memory is effective to cause plate 36 to return to its original position where it supports the succeeding members 6 as they, in turn, move with the carousel conveyor. A slidable metal plate 36 and associated solenoid device 38 is provided at each of the exit stations 12 through 22. The operation of the respective solenoid devices is controlled by a computer unit, to be described, in dependence on the signals produced by the light head unit 8 and the X-ray fluorescence unit 10.
After passing exit station 22 those keys which are still in the horizontal position are released so that all keys are in a vertical position. The released keys are then restored to the horizontal position between station 22 and feed conveyor 2, by means of a sloping ramp or bar 28 extending from the periphery of table 24 tangentially upwardly towards and merging with plate 34 as shown in Figure 1.
In Figure 5, the X-ray fluorescence system is diagrammatically illustrated in a little greater detail so as to provide a greater understanding of its operation. For convenience, pieces of scrap metal 48 and 50 are shown as moving along a standard conveyor 52. The piece 48 has reached the examination position and fluorescence is produced by an 1251 source 54 irradiating the sample of scrap metal 48. Specific X-rays 56 are produced and are detected with a Si (Li) detector unit 58. As will be understood, the charge produced in the silicon wafer thereof is fed to a preamplifier and then is amplified by a pulse processor within unit 60.
An analog electrical signal is produced and this is digitized by an analog to digital convertor within unit 60. The resultant digital information is then fed to a computer unit 62 through an interface within unit 60.
The computer unit 62 then analyzes the information received in order to produce an output signal on line 68 whereby control of the selected one of the exit stations 12 through 22 can be effected. In this way, the type of metal in a piece of scrap metal can be determined and, at the corresponding respective exit station, the keys 6 can be caused to rotate whereby the piece of metal drops at that exit station into a chute and, for example, a receiving bin for that particular type of metal. At the exit station 20, where aluminum is to be deposited, a metal detector unit 64 is located below the conveyor as illustrated in Figure 1. This unit overrides the control signal to this station if the material is non-metallic so as to prevent the dropping of the members 6. In this way, all the scrap metal of a particular type can be collected in a particular bin for future processing.
As will be appreciated, the number of keys 6 which are caused to drop, i.e. rotate, by retraction of the respective slidable metal plate 36 (Figure 3) is dependent on the size of the piece of scrap metal. This is determined by the light head unit 8 of Figure 1 which comprises two horizontal metal bars, one of which is placed below the position of the keys 6. This metal bar incorporates sixteen infra-red emitting light sources (type TIL 32) and optical lenses to focus the light whilst the other metal bar is placed above the keys 6 and incorporates sixteen solid state infra-red detectors (type TIL 63). The sixteen detectors and the sixteen emitters are recessed in the respective metal bar so that the light beam from a given emitter is received by only the corresponding detector. Fifteen of the light beams are utilized in the detection of objects on the conveyor, e.g. pieces of scrap metal, whilst one of the light beams, the one closest to the perimeter of the wheel, is utilized to provide a pulse to interrupt the computer and to provide a pulse to the logic circuits used for test purposes. The logic circuits are designed to prevent any action being taken merely because successive keys pass through the light beam. The logic circuits are designed to respond to the presence of pieces of scrap metal. It will be apparent that the circuits to provide the interrupt signal and perform the above logic can readily be suitably designed.
In Figure 6, there is diagrammatically illustrated, in block form the various units which are incorporated into the apparatus together with their interconnections. The table 24 is associated with the optical detector unit 8 as well as the X-ray detector unit 10. An output from the X-ray detector unit 10 is fed to a pulse process unit 70 (Kevex Corp. model #4532-P), then through an analogue-to-digital convertor (ADC) unit 74 (Northern Scientific model #TN1313) and an interface 72 to the computer unit 62. Units 70, 72 and 74 are indicated in Figure 5 as the single unit 60. A teletype unit 76 and a display unit 78 are associated with the computer 62 whilst signals pass between the computer 62 and automation module unit 80. The automation module unit 80 is operational to receive signals from the optical detector unit 8 and pass the information on to the central processor unit for analysis. Control signals pass through the automation module unit 80 to control a relay unit 82 whereby the selected one of the exit stations 12 through 22 is provided with information signals to initiate its operation at a time when the respective piece of scrap metal is over the output chute for that particular exit station. In Figure 7, there is diagrammatically illustrated, in block form, part of the electronic stages which are incorporated in the units illustrated in Figure 6. It is believed that the function and operation of the stages ilustrated in Figure 7 will be clear from the labelling thereof and it will be seen that the stages have been grouped into the respective groups, data input circuits 84, output drive circuits 86 and height reject circuits 88. Thus, the illustrated stages may be considered as the electronics for the light head stage 10 of Figure 1 and the driver circuits for the solenoid stages such as illustrated in Figure 4.
In Figure 8, there is drawn a schematic outline of the software program when the light head stage 8 (Figure 1) produces an interrupt operation. The outline is the main decision-making routine in the computer 62 (Figure 6) which is programmed to control the reaction of the sorting table 24 of Figures 1 and 6 and its associated apparatus. The simple program normally running in the computer displays the X-ray spectrum which is accumulating in the computer's memory. When an interrupt occurs as a result of a "peg" pulse (as hereinafter defined), the display program is broken and the sequence of operations illustrated occurs. The operation of the outline shown in Figure 8 will be clear to an expert skilled in the art having regard to the labelling used thereon.
In Figure 9, there is diagrammatically illustrated the light-emitting diode sources and the associated optical detectors in the light head 8 (Figure 1). The use of diode soruces and the optical detectors permits close spacing between the light beams and this allows objects to be located on the keys with a high degree of accuracy. This if of importance in making decisions as to whether two objects are located side-by-side, or deciding whether an object is located in a suitable position so that it will be satisfactorily sorted by the detecting unit 10. The light beams are arranged to be perpendicular to the axis of the conveyor and each light beam is interrupted by the movement of a key under the head. If a beam is interrupted within this space, simple counting of the number of keys which pass under the head whilst such an interruption continues gives the apparatus a measure of the length of the object independently of the speed of the conveyor 4.
As will be appreciated, the movement of the regularly spaced keys through the light beams allows the position of a piece of scrap metal to be determined as it moves with the carousel conveyor. Since each successive pulse which is generated when the beam is broken represents the movement of the conveyor 4 by a distance corresponding to one key spacing, the position of the object on the table can be located by counting pulses from some arbitrary position, the light head. This is completely independent of variations in the speed of the conveyor and it has been demonstrated that no other method of object location need be provided.
With reference to Figure 9, it will be seen that each detector is incorporated in a transistor emitter-follower circuit. The low impedance output is connected via a multi-conductor cable to an integrated circuit amplifier and sixteen separate outputs are selected. These are fed to the digital computer which evaluates which of the beams in the series of sixteen are occulted at the time that an interrupt pulse is generated.
To produce a pulse as each key passes through the light beams, the light beam closest to the perimeter of the conveyor 4 is emitted, detected and then amplified as described above. As the light beam reappears after the passage of a key, the voltage step in the light detector is fed to an astable multivibrator which generates a pulse of a duration approximately equal to one-half that of the time for which the light beam will be on. At the end of this pulse, a second astable multivibrator generates a pulse of relatively short duration which is provided to the computer as an interrupt signal. It is during this pulse, that the computer reads the information about which light beams are occulted.
The display monitor circuitry displays the signals presented to the computer on a set of light-emitting diodes. The outputs are also combined through a sequence of gates to activate a light-emitting diode when an object is detected between the keys. The status of this indicator only changes during the computer-read pulse.
In Figure 10 there is diagrammatically illustrated the arrangement for the solenoid driver stages, whilst in Figure 13 the power circuit for the solenoid stages is shown. Signals generated by the computer are arranged to cause a specific solenoid, like 40 (Figure 4), at a respective control station (Figure 1) to be activated. These signals are passed by way of a connecting cable to a single stage transistor amplifier (Figure 11), whose output is connected to a solenoid driver unit. As will be seen in Figure 11, this comprises a power circuit utilizing an A.C. source, a transformer, a full-wave bridge rectifier circuit and a current limiting resistor. The power supply charges a capacitor which may be connected across the terminals of the solenoid by the incoming pulse applied to the base of a power transistor used in a searching mode. This arrangement provides a strong initial pulse to activate the solenoid and a weaker holding current appropriate to the permitted power dissipation in the solenoid coil.
The solenoid driving circuit is repeated in accordance with the number of solenoids provided. At one of the control stations, an override circuit is provided utilizing a commercial metal detector and a Schmitt trigger circuit to only activate the solenoid if the object is metallic in nature. All other objects are treated as non-metallic and remain on the conveyor unit 4 until a discard outlet is reached.
From the above and with reference to Figure 5 it will be appreciated that the illustrated circuit design has two functions incorporated within it, as set forth below:-

a) The provision to the computer of the information which includes:
- i) An interrupt signal to denote the movement of a key under the light head unit 10. This pulse forms a peg counter for object location on the conveyor 4, and also is utilized to enable the digital computer to alter information sorted in its internal registers.
- ii) A series of voltage levels which are high or low depending on whether any given light beam is interrupted. These levels are transferred to the computer registers only during the above-mentioned interrupt signal.
b) The provision of a test facility which includes an illuminated display of the status of each light beam and an indicator to show whether any light beam is interrupted by an object. This feature is believed to be useful for routine testing and setting up of the detector with respect to the keys. The front panel lamp display is a set of light-emitting diodes which are not illuminated if a beam is broken. If an object is detected by any beam, the light-emitting diode is lit and a voltage appears at a test point on the front panel.

After the analysis has taken place, the digital computer changes the voltage level within a register appropriate to sorting the metal into a particular bin. This level operates a particular solenoid through the respective output drive circuit.
The X-ray fluorescence unit operates to sort non-metallic materials towards the bin allocated for aluminium. The solenoid driver circuit for this bin is fitted with an override circuit whereby unless a commercial metal detector placed immediately in front of the bin is triggered, the material will not be sorted and will continue to a discard exit.
To prevent excessively high pieces of material from damaging the light head unit 8 or the detector unit 10, a horizontal light beam in unit 7 (Figure 1) may be provided at a conveniently set height (approximately three inches in the apparatus illustrated) above the members 6. This is positioned just after the place where the pieces of scrap metal are fed from the feeding conveyor. If the light beam is occulted some twenty keys are dropped at an exit station 9, situated just after this horizontal light beam and similar to stations 12 to 22.
Apparatus according to the present embodiment of this invention has been described above. Consideration will now be given to the operation and use of the apparatus having particular regard to the sorting of. shredded automobile scrap metal. This is usually non-ferrous but it will -be appreciated that this embodiment can equally be applied to ferrous scrap material. Non-magnetic automobile scrap material can usually be classified into the following groups:-

1) Zinc alloys.
2) (a) Copper and brass. (b) Copper wire with some form of insulation.
3) Stainless steel.
4) Aluminum.

As mentioned above, soon after a sample arrives on the table from the feed conveyor belt system 2, it passes through the linear array of infrared light beams 8 which are set perpendicular to its path and which are arranged vertically so that they can pass between the keys on the rotating wheel. If a sample covers part of the opening between two keys, some of the 16 light beams will be occulted. The position of each light beam occulted is passed to the computer. The electronic units necessary to effect this transfer can be readily determined from the above description and will be seen to consist of an amplifier, a comparator, and a pulse-shaping circuit. As mentioned above, the signal from one light beam, on the rim of the table is sometimes called a "peg" pulse and is specially treated whereby it is delayed approximately seven milliseconds before being sent as a relatively short signal to the computer 62 (Figure 5). All the signals pass through the I/O interface within the Tracor Northern 1310 interface section within unit 80 and are then fed to the computer. The peg pulse causes what is called an "interrupt" in the computer which then accepts the information from the light head. The computer determines which of the light beams are occulted in each opening betwen the keys and from this information the computer notes:-

(1) where the sample is radially on the keys in order to decide if the sample will pass under the X-ray fluorescence detector,
(2) if there is more than one sample side-by-side on the table in order to cancel the X-ray analysis and thus prevent missorting,
(3) the number of openings between keys in which at least one light beam is occulted in order to determine the length of the sample.

The X-ray fluorescence system was described above with reference to Figure 5 and it will be understood that when the material is excited by radiation, part of the incident energy is lost by the emission of the X-rays which have energies characteristic of the elements present in the samples. The energy and intensity of such characteristic X-rays serve as a unique signature of a given material.
Radiation from the radioactive source ¹²⁵1 is incident on the sample under investigation which then emits characteristic X-rays. These are then detected by a lithium drifted silicon counter unit 58 (Figure 5). The output identifying signals from this counter consists of a series of voltage pulses of amplitude proportional to X-ray energy. The pulses are amplified and shaped by a standard nuclear electronics stage, and the number of pulses corresponding to a given energy (element) are sorted into a spectrum and displayed using the computer stage 62. Using this spectrum, the minicomputer can make decisions about the type of object presented to the detector and provide command control signals to operate the mechanical sorting equipment.
As will be understood, the computer associates with each object an identification made by the X-ray detector and prepares subsequent components to discharge the respective object at the respective solenoid for the particular type of material. The computer keeps track of the position of the total number of objects (normally in the range thirty to a hundred) as they move around the table by counting the keys as they pass under the optical light head. Besides noting the passage of each key the light head, with the help of the computer, measures the length of the object by noting the number of keys which pass the head whilst one or more of the infrared beams is occulted by the respective object.
As mentioned above, at a number of stations around the outer rim of the sorting table there are provided metal slides which can be withdrawn or inserted by means of a solenoid. Withdrawing the slides allows the keys to rotate about their pinned end to discharge objects off the table at the location of the respective solenoid. The operation of these solenoids is controlled by the computer.
As illustrated in Figure 3, the movable section can be approximately 2,54 cm (one inch) long and is on the end of the plunger of a solenoid. When the respective section is to be withdrawn, i.e. when the first part of a sample to be dropped at this station arrives there, the solenoid is simply energized to withdraw the support. When all the keys supporting the respective sample have dropped through the gap in the supporting surface, the solenoid is released and it springs back. Since the keys are somewhat flexible, no difficulty is experienced in the operation of the table if one of the keys is hit by the returning section of the support surface.
The energizing of the respective solenoid is effected by the above-mentioned computer stage since it monitors where each sample is as it moves around the sorting table.
The computer system which is used in the constructed practical embodiment works on the interrupt basis or in real time. Most of the time, it is simply displaying an X-ray spectrum it has in its memory. Two types of interrupt could occur. One occuring if the ADC has completed digitizing a signal from the X-ray detector and the ADC interface (TN1313) interrupts the central processor in the computer and directly modifies a memory location. This is normally referred to as direct memory access (DMA) and .involves no program steps in the actual transfer if the interface is initialized to operate this way.
The second interrupt occurs when the signal from the peg pulse arrives at the computer. It initiates a sequence of events. Firstly the interrupt indicates to the computer that a key has passed the light head and therefore every sample on the table has moved further along. The computer produces a corresponding adjustment in the entry of its memory for each sample and causes the appropriate action, e.g. firing a solenoid at the appropriate station or starting an analysis at the X-ray fluorescence detector etc., to occur.
If all the light beams are not on, the computer determines which light beams are off and whether more than one group of lights is off. This information together with similar information from the previous gaps between the keys allows the computer to decide of a single sample is on a path going under the X-ray detector and therefore than an analysis should be effected when the sample reaches the detector.
In Figure 8 there is actually shown the schematic outline of a software program when the light head produces an interrupt. This is the main decision-making routine in the computer programmed to control action of the sorting table.
As mentioned above, the computer, is supplied by Tracor Northern, and is used to control all functions involved in the sorting operation. It collects the data from the X-ray fluorescence detector, decides what type of material has passed under the detector, notes the passage of each key under the light head and whether a piece of material is sitting on that key and subsequently activates the appropriate solenoid as the respective object reaches it.
As will be clear, the software (Figure 8) for performing these operations was specially written and consisted of two main parts, the analysis part and the table control part. In the first part, the number of counts in several regions of the X-ray spectrum was determined after the sample object had passed the detector. These regions corresponded to those X-rays which are characteristic of Fe, Ni, Cu, Zn and a background. If the largest number of counts occurs in the Fe or Cu regions, then the sample is said to be iron or brass respectively. If the Ni region had the greatest number of counts, then the Cu/Ni and Zn/Ni ratios determined whether the sample was brass or zinc (many brass and zinc automobile parts are flash coated with nickel to facilitate bonding of the subsequent bright chrome or similar plating). If the Zn region had the greatest number of counts, then the relative amount of Cu present, i.e. Zn/Cu ratio determined whether the sample was zinc or brass.
If the highest number of counts occurred in the background region then the material was aluminum or some non-metallic material. Consequently on the solenoid for aluminum material, a metal detector was provided to check the object for metal content before the solenoid was released.
The second function of the software was to monitor the position of each object as it moved around the sorting table. To do this, information about each sample on the table was stored in a section of the computer's memory. This information consisted of (1) the position of the sample relative to the light head (2) the length of the sample, in order to drop the correct number of keys, and (3) whether the sample had been analysed and, if so, the type of material so that the sample would be deposited at the appropriate solenoid exit station and exit along a respective selected path.
The digital information from the X-ray detector entered the computer through the TN1313 interface unit 72 whilst the control information, i.e. the passage of a key or the status of the solenoids, entered through two input-output units in the TN1310 within unit 90.
A commercial unit was assembled and tested. The sample of scrap used was unwashed and had been shredded into pieces, the average weight of each piece was 44 gms so that a material flow rate of one ton/hr. implied a sorting rate of 20,000/hr. or about 5 per second. Each piece was approximately 5 cm (2 inches) in size and about 60% of the brass and zinc samples were plated.
Using the X-ray analysis it was found that the materials were well characterized by the components zinc (Zn), brass (Zn, Cu), wire (with lead in the insulation), stainless steel (Fe, Cr), aluminium (with no characteristic peaks). In the plated samples, only zinc or brass were found to be plated and the plating invariably contained nickel (Ni) and copper (Cu). Since the technique using 1251 sampled the surface, nickel constituted the major detected element for both plated zinc and plated brass. However, on the basis of the samples examined, the two materials could be distinguised with greater than 90% certainty by measurement of the Ni:Cu ratio and the Cu:Zn ratio. By producing the results graphically, it was found that plated zinc fell amost exclusively above a particular level whilst plated brass had a higher copper content and fell below the respective level, i.e. line drawn on the graph.
The explanation for this resides in the fact that the nickel acts as a barrier for those X-rays, characteristic of copper or zinc as they return to the detector (Figure 5). Furthermore, because the characteristic K X-ray of zinc has an energy greater than the binding energy of the K electrons in nickel while the K X-ray of copper does not, the most abundant X-rays from zinc are very strongly absorbed and the discrimination between plated zinc and brass is effected.
It was found that the peaks for all the elements found in the scrap were distinct and their heights could be compared in a simple manner. No problems were encountered due to dirt, and if the sample of scrap examined was representative of the industrial material then no washing would appear to be required.
Experimentally it was estimated that approximately 1000 counts in the whole spectrum were required in order to make a clear and reliable recognition of the material. This figure and the time for which a given specimen is in front of the detecting head determines the counting rate required for a given speed of operation.
If scrap material is presented as single pieces separated on 10 cm centres, the conveyor system must travel at 0.5 m/s (1.1 mph) for a material throughput of 1 ton/hr. A rough estimate suggests that if the sample is presented to the detector system for 0.1 sec and 1000 counts are required for a decision, then the counting rate is 10,000/s. Standard nuclear electronics can operate effectively up to 50,000/s so that the principal limitation on counting speed is the strength of the exciting radioactive source.
Sources of a few Curie strength are commercially available and it is to be noted that because the radiation is weakly penetrating, it may be easily confined by simple radiation shields whereby radioactive hazards are minimal.
It will be appreciated that the categories of brass and zinc could be futher subdivided into plated and unplated samples with considerble reliability using the apparatus above. Furthermore, the presence of iron samples as distinct from stainless steel could also be detected.
The embodiments of the invention have been described above in regard to a particular application, i.e. the separation of mixtures of shredded metallic pieces. However, it will be appreciated that it can be readily adapted to other uses and for some of these applications X-ray fluorescence may be a suitable method of analysis. The apparatus can obviously be adapted to the separation of alloys of the same class (e.g., the separation of stainless steels, brasses or nickel alloys). Furthermore, other methods of analysis could readily be employed with the sorting table and the following is a partial list of the measurements which can be made to provide the criteria for separation:-

(a) size and shape
(b) mass
(c) radioactivity
(d) surface features
(e) temperature
(f) air resistance
(g) colour
(h) premarking or tagging.

Appropriate combinations of these measurements may also be employed to determine the separation criteria.
The sorting table itself may, also be employed for a variety of other purposes. It is envisaged that it could be modified in the following ways:-

(a) Size: The keys can be made of any desired length, width and shape to accommodate items of appropriate shape and size.
(b) Configuration: The keys can be incorporated into a table of circular design, a linear conveying system or may be stacked.
(c) Materials of Construction: The sorting system can be constructed in a variety of materials to suit the particular operating conditions which might, on occasion, involve the immersion of the system in a special atmosphere or liquid.

It will be appreciated that the computer may readily incorporate _'microprocessors or other microcircuit devices.

(d) Key Design: For special purposes the mechanism for key support, release and spacing may be redesigned.
(e) Light Head: The components incorporated within the light head may readily be changed for use in other applications as may the number of light beams. In the present embodiment of the invention sixteen beams were used to facilitate the transfer of information from the light head to the sixteen bit computer.

It will also be appreciated that the sorting mechanism can readily be employed as a feeding system for particles or manufactured parts.

Claims

1. Conveyor apparatus for sorting a plurality of articles fed sequentially thereto, of the type including a conveyor (4) for conveying the articles (48, 50) along a predetermined path in sequential order, scanning means (10) adjacent the conveyor arranged to scan each article sequentially and provide an output signal corresponding to a selected characteristic of the article and control means (62) arranged to direct a selected article in response to said output signal along a flowpath selected from a plurality of flowpaths (12, 14, 16, 18, 20) branching off from said predetermined path, characterized in that:

(a) said conveyor includes a plurality of spaced apart support members (6) extending transversely of said predetermined path and being arranged, on the one hand, for sequential movement along said path, and on the other hand for movement between an article supporting position and an article non-supporting position;

(b) sensor means (8) are provided to determine the number of support members (6) supporting each article, to thereby establish the length of an article;

(c) means (8, 62) are provided to determine the position of each article, relative to a reference position fixed in relation to said conveyor (4), as a function of the number of support members (6) from the article to the reference position; and

(d) the control means (62) is arranged to select one of the flowpaths for a selected article by triggering movement of the members supporting the selected article from the supporting to the non-supporting positions in response to said output signal and the legnth of the article when said article arrives at the selected receiving flowpath.

2. Apparatus as claimed in Claim 1 characterized in that said support members (6) comprise keys each pivotally mounted at one end, substantially horizontal in said supporting position with the free end thereof arranged to rotate downwardly to said non-supporting position at a selected receiving flowpath.

3. Apparatus as claimed in Claim- 2 characterized in that said keys (6) rotate under gravity at a selected receiving flowpath.

4. Apparatus as claimed in Claim 1 characterized in that said conveyor (4) is a horizontal conveyor and said supporting position is horizontal while said non-supporting position is substantially vertical.

5. Apparatus as claimed in Claim 2 characterized in that the free end of each key (6) is supported on a supporting element (34, 36) as the respective key moves along said predetermined path, at least a portion (36) of said supporting element being selectively retractable at each said receiving flowpath whereby the said free ends of the keys are released when arriving at a selected said receiving flowpath and the selected keys rotate to said non-supporting position.

6. Apparatus as claimed in Claim 1 characterized in that said articles (48, 50) to be sorted comprise scrap metal pieces and said scanning means (10) provides an output signal characteristic of the type of metal contained in said scrap metal pieces.

7. An apparatus as claimed in Claim 6 characterized in that the scanning means comprises an X-ray fluorescence scanning means (10).

8. Apparatus as claimed in Claim 1 characterized in that said sensor means is a light head (8) capable of monitoring the passage of each support member (6) as it travels along the predetermined path, the position of said light head defining the reference position.

9. Apparatus as claimed in Claim 8 characterized in that said position determining means comprises a computer unit (62) receiving signals from said light head (8) each time a support member (6) travels past said reference position, said computer unit counting said signals to determine when a particular article arrives at at least one predetermined location.

10. Apparatus as claimed in Claim 9 characterized in that said computer unit (62) of said position determining means is arranged to determine the number of support members (6) supporting a selected article on said conveyor (4).

11. Apparatus as claimed in Claim 10 characterized in that, in addition to the flowpaths, a plurality of discharge locations (9, 22) is provided, said control means (62) operating, when a selected article arrives at a selected discharge location, to cause said number of support members (6) to be moved to said non-supporting position whereby the selected article exits at said selected discharge location.

12. Apparatus as claimed in Claim 8 characterized in that said light head (8) comprises a plurality of light sources spaced from each other transversely across the conveyor (4) to provide a plurality of parallel light beams and a corresponding plurality of photocell devices on the opposite side of the conveyor each arranged to receive a respective one of said light beams so as to facilitate pattern recognition of said articles.

13. A method of sorting articles according to a selected characteristic thereof, wherein the articles are fed sequentially along a predetermined-path by means of a conveyor (4) and wherein each article is scanned as it moves along the conveyor so as to generate an output signal corresponding to the selected characteristic, which signal is fed to a control means (62) arranged to direct a selected article in response to said output signal along a flowpath selected from a plurality of flowpaths (12, 14, 16, 18, 20) branching off from said predetermined path characterized by:

(a) conveying the articles to be sorted sequentially by means of a plurality of spaced apart support members (6) extending transversely of said path and being arranged, on the one hand, for sequential movement along said path and, on the other hand, for movement between an article supporting position and an article non-supporting position,

(b) determining the length of each article as a function of the number of support members (6) supporting it;

(c) measuring the position of each article with respect to a reference position as a function of the number of support members (6) from the article to said reference position; and

(d) directing a selected article to a selected one of the flowpaths by triggering through said control means (62) movement of the members (6) supporting the selected article from the supporting to the non-supporting positions in response to said output signal and the length of the article when said article arrives at the selected flowpath.

14. A method as claimed in Claim 13, characterized in that said articles are scrap metal pieces.

15. A method as claimed in Claim 14 characterized by scanning each scrap metal piece with radiation from a radioactive source whereby it emits an X-ray pattern characteristic of the metal in said scrap metal piece, and detecting said X-ray pattern to produce an output signal corresponding to said metal.