Magnet arrays

Abstract

Method and device for self-regulated flux transfer from a source of magnetic energy into one or more ferromagnetic work pieces, wherein a plurality of magnets, each having at least one N-S pole pair defining a magnetization axis, are disposed in a medium having a first relative permeability, the magnets being arranged in an array in which gaps of predetermined distance are maintained between neighboring magnets in the array and in which the magnetization axes of the magnets are oriented such that immediately neighboring magnets face one another with opposite polarities, such arrangement representing a magnetic tank circuit in which internal flux paths through the medium exist between neighboring magnets and magnetic flux access portals are defined between oppositely polarized pole pieces of such neighboring magnets, and wherein at least one working circuit is created which has a reluctance that is lower than that of the magnetic tank circuit by bringing one or more of the magnetic flux access portals into close vicinity to or contact with a surface of a ferromagnetic body having a second relative permeability that is higher than the first relative permeability, whereby a limit of effective flux transfer from the magnetic tank circuit into the working circuit will be reached when the work piece approaches magnetic saturation and the reluctance of the work circuit substantially equals the reluctance of the tank circuit.

Claims

1. Magnetic device for effecting magnetic flux transfer into a ferromagnetic work piece, comprising: (a) a plurality of magnets, each having at least one N-S pole pair defining a magnetization axis and passive pole extension pieces arranged for extending the magnetic poles of each said N-S pole pair; and (b) a carrier supporting the magnets in an array configuration in a medium having a first relative magnetic permeability which is substantially lower than that of ferromagnetic material; wherein the array configuration is one in which (i) gaps of predetermined distance are maintained between neighboring magnets in the array, (ii) the magnetization axes of neighboring magnets are oriented such that immediately neighboring magnets in the array interact magnetically with one another via magnetic fields extending across said gaps between opposite poles of the respective N-S pole pairs, (iii) the magnets in the array configuration form part of an array-internal magnetic circuit in which array-internal flux paths extend through the medium between opposite poles of the N-S pole pairs of neighboring magnets, and (iv) oppositely magnetized ones of the passive pole extension pieces of the magnets and of neighboring magnets provide magnetic flux access portals through which magnetic flux can be transferred from the array-internal magnetic circuit into a ferromagnetic work piece; whereby a magnetic working circuit can be formed by bringing a ferromagnetic work piece in contact with the magnetic flux access portals and which has an initial reluctance that is lower than that of the array-internal magnetic circuit and in which a limit of effective flux transfer from the array-internal magnetic circuit into the working circuit will be reached when the ferromagnetic work piece reaches magnetic saturation and the reluctance of the working circuit substantially equals the reluctance of the array-internal magnetic circuit, wherein the magnets are dipole permanent magnets having one N-S magnetization axis, wherein the permanent magnets are arranged in one or more concentric, closed circle or oval array(s), and wherein the N-S magnetization axis of each of the permanent magnets extends coaxially with a radius extending from a center of the circle or oval array(s) to the respective permanent magnet.

2. The magnetic device of claim 1, wherein the N-S magnetization axes of the dipole permanent magnets are arranged to extend within a common plane.

3. The magnetic device of claim 1, wherein the permanent magnets are on/off-switchable dipole permanent magnets, the magnets being individually or jointly switchable between an ‘on’ state in which magnetic flux is transferred via the magnetic flux access portals into the working circuit, and an ‘off’ state in which magnetic flux is shunted within the permanent magnets and the associated pole extension pieces.

4. The magnetic device of claim 3, wherein the on/off switchable dipole permanent magnets comprise a first permanent magnet dipole which is held stationary between the associated two passive pole extension pieces such that the passive pole extension pieces are respectively magnetized with opposite polarities, and a second permanent magnet dipole which is held movable relative to the first permanent magnet dipole and the passive pole extension pieces whereby the N-S pole pair of the second permanent magnet dipole can be brought selectively into magnetic alignment with the N-S pole pair of the first permanent magnet dipole to provide the ‘on’ state and into magnetic counter-alignment to provide the ‘off’ state in which a closed magnetic flux circuit is defined between the first and second permanent magnet dipoles and the two passive pole extension pieces.

5. The magnetic device of claim 1, wherein the magnets in the array are arranged such that (i) magnetic flux passing into the work piece through the passive pole extension pieces at each of the magnets flows in a first direction through the work piece and (ii) magnetic flux passing into the work piece through the passive pole extension pieces of neighboring ones of the magnets flows through the work piece in a second direction opposite to the first direction, such resulting in a non-uniform flux flow direction within the work piece.

6. The magnetic device of claim 1, wherein the pole extension pieces are devised to deliver magnetic flux from the array-internal magnetic circuit into the work piece in a direction perpendicular to that of the N-S magnetization axes of the magnets.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a perspective view of an experimental jig incorporating an array of individual, switchable permanent magnet units, being used as a ‘proof of concept’ model embodying a number of aspects of the present invention;

(2) FIG. 2 is a perspective photographic view of a working model of a magnetic lifter device made in accordance with a number of aspects of the present invention;

(3) FIGS. 3a and 3b are perspective schematic illustrations of a single diametrically polarized permanent magnet and a switchable permanent magnet unit as may be employed in the devices of FIGS. 1 and 2;

(4) FIG. 4 is a schematic and highly simplified (side) view of a single, switchable permanent magnet unit illustrating some principles underlying an aspect of the present invention;

(5) FIG. 5 shows a perspective schematic view of the single switchable permanent magnet unit of FIG. 3, illustrating flux exchange areas when the unit is in an activated state and in contact with a ferromagnetic sheet material work piece;

(6) FIG. 6 is a schematic illustration of two linear magnet array configurations in accordance with one aspect of the present invention;

(7) FIG. 7a is a schematic and highly simplified (side) view of a linear array of multiple, switchable permanent magnet units illustrating some of the aspects of the present invention, whereas FIG. 7b represents a perspective schematic view of a three magnet linear array;

(8) FIGS. 8a to 8c are schematic plan bottom views of 3 different circular array magnetic device configurations as contemplated in the present invention, the array of FIG. 8a being embodied physically in the lifter device of FIG. 2;

(9) FIGS. 9a to 9c represent schematic 2-D (or plan view) illustrations of the magnetic field lines that would be detectable in the circular array configurations illustrated in FIG. 8a to 8c, respectively;

(10) FIG. 10 is a schematic plan view of a magnetic field line model of a discontinuous magnet torus, intended to illustrate a further aspect of the present invention related to magnetic flux splitting and self-regulating field intensity; and

(11) FIGS. 11a and b are schematic side views of two switchable permanent magnet units as per FIG. 3b, arranged into a linear array, but which can be incorporated into the magnet array configurations of FIG. 8a and FIG. 10.

DETAILED DESCRIPTION OF EMBODIMENTS

(12) FIG. 1 illustrates a test-rig-style switchable permanent magnet coupling device 10 incorporating one of the basic concepts underlying the present invention. Embodiments of such magnetic devices may be incorporated into more complex (or simple) apparatus and devices to releasably magnetically couple such device or apparatus to a ferromagnetic body, eg a magnetic lifter as illustrated in FIG. 2 adapted for lifting individual, thin, ferromagnetic sheet metal materials from a stack of such sheets.

(13) Such device 10 includes a housing or carrier part 12 of substantially non-ferromagnetic material, in this case having a circular plate-like shape, in which are secured against movement five individual, permanent magnet coupling units 14, as will be described below. The units 14 are mounted in apertures that extend through part 12, and may be permanently secured, eg glued, or otherwise secured to allow exchange of individual units. The units 14 are received at part 12 so that at least the non-visible bottom axial end faces of units 14 are either flush with the circular engagement surface of part 12 or protrude slightly therefrom. In FIG. 1, the magnets are flush with the upper face of the carrier part 12 and accessible to allow switching of each unit 14 between active and inactive magnetisation positions. The units 14 are disposed in a circular array configuration about a central axis of device 10.

(14) As will become clearer from the subsequent description of an individual unit 14 illustrated in FIG. 3b, each unit 14 includes a pair of stacked cylindrical permanent magnets 20 and two pole pieces 16 and 18 that surround the periphery of the magnets to substantially envelope same, wherein the lower (not illustrated) axial end faces of the pole pieces 16, 18, which are made of a soft iron material with high permeability, are either flush with or extend a small amount beyond the corresponding lower axial end face of the lower one of the cylindrical magnets 20.

(15) One of the cylindrical magnets 20 of a unit 14 is shown in FIG. 3a. The magnet is diametrically magnetised across its entire axial length. By that is meant that the notional division between the North Pole (N) 22 and the South Pole (S) 21 of the magnet is provided by a vertical plane 24 that passes along a diameter 28 of the upper face 28 and the lower face 29 of magnet 20. The magnet 20 is still essentially a dipole having a magnetisation axis MA which is perpendicular to the vertical plane 24, wherein however the magnetic field strength along the circumference of the cylinder varies about in sinusoidal manner, wherein a minimum value exists at the N-S interface plane 24, and a maximum exists at about 90 degrees relation along the circumference. Cylindrical (or disc-shaped) magnet 20 is preferably a rare-earth type magnet, for example a neodymium-iron-boron magnet, noting that currently available rare earth magnets will achieve a flux density maximum of around 1.4 Tesla, which is substantially below the saturation densities of good passive ferromagnetic materials that can be used for the pole pieces 16, 18. The present invention also contemplates the use of other active permanent magnetic materials.

(16) Turning next to FIG. 3b, there is shown in disassembled state a switchable, permanent magnet unit 14 which but for the presence of a unit activation and deactivation mechanism 30 is in essence similar to the units 14 shown in FIG. 1.

(17) Unit 14 includes two cylindrical magnets 20a, 20b of the type described above, of similar height dimensions and N-S poles make-up. An example is a 10 mm diameter×8 mm bight cylindrical magnet. The lower magnet 20b is held in surface engaging contact between the two pole pieces 16 and 18, which are identical in shape and cross-section and have a magnet-facing internal surface 32 that is correspondingly curved to match the magnet's external peripheral surface, whereas the upper magnet 20a needs to maintain as minimum as possible gap towards the peripherally facing surfaces 32 of pole pieces 16 and 18 thereby to enable friction free (or minimised) rotation thereof within the pole pieces 18 and 18 and relative to the lower magnet 20b which is itself held immovable. Magnets 20a and 20b are simply stacked above one another along stacking axis A, which defines a longitudinal axis of unit 14, and such that upper magnet 20a may be rotated relative to lower magnet 20b using the actuating mechanism 30.

(18) Further details as to the make-up, possible different configurations of the components of such magnet unit 14 and the principles of operation thereof are described in U.S. Pat. Nos. 6,707,360 and 7,012,495 to which reference should be made for further details.

(19) For present purposes, it is sufficient to note that upper and lower magnets 20a, 20b are received in face to face Juxtaposition within pole piece housing 16, 18, whereby rotation of the upper magnet 20a about axis of rotation A causes time-sequenced passage of the north pole region of upper magnet 20a over the pole regions N and S of lower magnet 20b. When in a position where the north pole of upper magnet 20a substantially aligns and coincides with the south pole of lower magnet, and consequently the south pole of upper magnet 20a substantially overlies the north pole of lower magnet 20b, the first and second magnets act as an internal, active magnetic shunt and as a result the external magnetic field strength from the unit would be ideally zero, assuming equal active magnetic mass in both magnets 20a and 20b and total flux carrying capacity of the pole pieces 18, 18 being higher than flux output of the combined magnets. Rotating the upper magnet 20a 180 degrees about axis of rotation A changes the alignment of the pole pairs of the magnets 20a and 20b, wherein the respective north and south poles of the upper magnet 20a substantially overlie respective north and south poles of lower magnet 20b. In this alignment, the external magnetic field from unit 14 device is quite strong and the device exerts a magnetic force onto a ferromagnetic work piece at the contact surfaces 34 of the unit 14 (provided by the bottom axial end faces of pole pieces 16, 18) thereby firmly securing the unit 14 to the work piece and creating an external magnetic flux path.

(20) The passive pole pieces 16, 18 are important in assisting this magnetic coupling functionality, and are made from a ferromagnetic material with low magnetic reluctance, eg purified iron, soft iron or mild steel. The cross-sectional area of the unit housing wall, which is provided by the pole pieces, is, in the illustrated embodiment, non-uniform, in order to achieve an increase in external magnetic field strength of the pole-piece-‘loaded’ permanent magnets; the external contour of the pole places, ie the wall thickness of the pole pieces 16, 18, is such as to reflect or be a function of the variation of the magnetic field strength around the perimeter of the permanently magnetised cylinders 20a, 20b.

(21) Essentially, the design of the pole pieces follows the variation of the field strength H around the perimeter of the permanent magnet cylinders 20a, 20b, application of the inverse square law of magnetic fields in devising the external shape achieving good results, but use of specific materials for the pole pieces and magnets, and intended application of the overall coupling device 10, require variation of and influence the optimal shape of the pole pieces 16, 18. For further details, refer to the aforementioned US patents.

(22) The external shape of the pole pieces 16, 18 assembled about the cylindrical magnets 20a, 20b alms to maximise the external field strength and assist in holding the unit 14 in place on a work piece in cases of an incomplete ‘external’ magnetic circuit. It is preferred that the pole pieces 16, 18 are of the shortest possible length along axis A. The poles form part of the magnetic circuit (along with the magnets) of each unit 14. The poles have an inherent magnetic resistance (“reluctance”) which results in loss of magnetic energy, even where high permeability materials are employed. In minimising the length of the poles, and overall height (or length) of the coupling units 14, loss of magnetic energy is minimised and hence external field strength maximised. The joint areas 36 that provide the interface between the facing pole pieces is provided with a very high reluctance, but thin layer, thereby maintaining magnetic separation of the pole pieces 16, 18, ie preventing short circuiting.

(23) Finally, the surface area of the axial end feces, see reference numerals 35 and 34, are preferably chosen to provide flux compression functionality. That is, the total cross-sectional (or foot-print) area of pole pieces 16, 18 will be chosen to be smaller than the cross section area of the magnets 20a, 20b, derived from the diameter of the cylinders times the total height. This allows to increase the flux density output of the unit 14 as compared to the maximum flux density which the active material can deliver. For example, since good ferromagnetic materials can reach saturation levels of 2 Tesla and above, it is possible to increase flux density in the poles to this level by reducing the total pole foot print area. Flux compression is not a fixed but a design parameter which is derived from magnetic flux density of the active source material multiplied by its cross section towards the pole pieces, flux saturation levels of the passive ferromagnetic (pole) materiel, and loss factors due to non-linearity of the B-H Curve of the pole piece material.

(24) Turning next to FIGS. 4 and 5, there is illustrated an individual magnetic switching unit 14 in highly schematized fashion, placed in contact on a thin, sheet-like work piece 40, wherein the unit 14 is schematically illustrated in an activated state in which the north and south poles 21 and 22 (FIG. 3a) of the upper and lower magnets 20a and 20b (FIG. 3b) coincide, and an external magnetic field is present; the lighter gray shaded portion of the unit 14 serves to denote the active south pole S that the magnets impose on one of the pole piece 16, and the darker gray shaded portion denotes the north pole polarity N switched onto the other pole piece 18.

(25) The pole piece footprint areas on the work piece 40 are identified at 42 and 43 in FIG. 5, ie in this illustration, the lower axial end surfaces of the pole pieces identified at 34 in FIG. 3b, serve to provide the work piece engagement area of the unit 14. The magnetic flux ‘exiting’ the north pole piece 18 at its contact surface 42 will ‘flow’ through a magnetic flux path across the thickness t of the work piece 60 and ‘enter’ the contact area 43 of the other, south pole piece 18, which is otherwise closed into a magnetic flux loop extending through the vertical interface area between the north and south pole regions of the diametrically polarized cylindrical magnets (20) pole-aligned within the unit 14.

(26) A primary effective flux exchange area 44 within the work piece 40 is that section of the total flux exchange area where flux density saturation is present. Since the magnetic field of the unit 14 is not confined to its footprint area, the total flux exchange area is extended by secondary effective flux exchange areas 46, located traversely to both sides of the central area 44 where the flux density declines with distance from unit 14. These secondary effective flux exchange areas 46 are maintained by flux leakage, which results from the (flux) saturation of the work piece, and the sizes of the flux exchange areas 44, 46 depend on the degree by which the work piece can absorb the flux. High flux absorption results in lower flux leakage and the secondary effective flux exchange areas shrink.

(27) If the thickness t of the work piece and the related total effective flux exchange area (62 and 64) in the work piece is smaller than the footprint area 42 or 43 of an individual pole piece 16 or 18, and/or the flux saturation (properties) of the work piece material are such that saturation occurs at a lower flux density than that of the pole pieces, the flux exchange is restricted and the flux density at the pole contact area drops. The result is a sharp decline of ‘pulling force’ exerted by the unit 14 onto the attached work piece 40, in accordance with the interrelationship between flux density and pulling force: magnetic pulling forces vary with the square of flux density but only linearly with pole area.

(28) As noted, if the work piece 40 cannot carry the whole flux of a unit 14, flux saturation occurs in the work piece 40 and the magnetic field generated by the superimposed individual magnetic fields of the two magnets 20 within the unit 14 extends beyond (in the thickness direction) the work piece 40, as is schematically illustrated at 48 in FIG. 4. Therefore, in attaching to a single sheet material work piece 40, the available magnetic energy which the unit 14 is able to provide in its fully activated state, is only partially utilized. It will be noted that the schematically illustrated magnetic field 48 extends through the thickness of the sheet material and is able to interact with other ferromagnetic work pieces 41 located beneath sheet material 40. Depending on the thickness of additional work piece sheet material 41, which may be a stack of sheets with total thickness t2, and the distance thereof from the saturated work piece sheet 40, the unit 14 will be able to magnetically lift additional sheets 41 up to a combined thickness where the combined flux exchange area of the stacked sheets 40, 41 is about equal to that of the pole piece contact areas 42 or 43 as described above.

(29) The extent by which the magnetic field will go beyond the immediately adjoining work piece 40 will of course depend on the active magnetic material mass present in an individual magnetic coupling unit 14.

(30) In accordance with one aspect of the present invention, instead of using a single or a number of relatively distantly spaced apart units 14, which are rated to provide a specified lifting or coupling force, the necessary active magnetic mass required to provide the necessary coupling force (apart from any force and/or flux transfer magnifying influences which pole piece shaping may contribute, see above), is subdivided into a number of smaller switchable magnet units 14, compare for example the schematic illustrations in FIGS. 7a and 7b. As per FIGS. 1 and 2, the units 14 will be secured and arranged in a larger housing (not shown) of a non-ferromagnetic material. Importantly, the units 14 will be deployed in specific types of array configurations as will be discussed below, compare also the illustrations of FIGS. 8a to 8c and 10, which allow interaction of the individual units 14 to achieve an improved performance.

(31) It will be appropriate to define a further geometric parameter that is necessary to describe not only the overall arrangement of individual units 14 in any given array, but also the relative location of the north and south poles of activated individual units 14. Referring to this end to FIG. 5, there is illustrated a so called polarization (or polar) axis PA of an individual unit 14, which axis is characterized by extending perpendicular to the (vertical) interface plane defined when the individual interface planes 24 (see FIGS. 3a and 3b) of the individual diametrically polarized cylindrical magnets 20a and 20b of the unit 14 are coterminous in that common plane, i.e. when the unit 14 is either in the fully activated or fully deactivated state where the magnetization axes MA of the individual magnets 20a and 20b are parallel aligned. In FIG. 5, the coupling unit is illustrated in its fully activated state. In essence, therefore, the polarization axis PA defines a north to south pole orientation axis in the fully activated state of the unit 14, and may be visualized as being the N-S magnetization axis of a simple dipole bar magnet, compare e.g. FIG. 6, and such simplified (activated) magnet analogy will be used in the further description.

(32) Turning then to FIGS. 7a and 7b, there are schematically illustrated a number of individual coupling units 14 disposed in a linear array wherein the units 14 are held spaced apart from one another by equal gaps (g), the polar axes PA of the individual units 14 being arranged in series and coaxially with one another such that north and south poles of the activated units 14 are arranged in alternating sequence. FIG. 6 illustrates in highly schematised manner the serial alternating array configuration embodied in FIGS. 7a and 7b (represented by simple N-S bar magnets 14′), as well as another serial array configuration in which the polarization axes PA of the units 14′ extend perpendicular to the axis AA of the array. It will be noted there that adjoining (or neighboring) magnets 14 also face one another across the gaps with alternating N-S polarities.

(33) Referring back to FIGS. 7a and 7b, it will be seen that, within the work piece 40, apart from the individual effective flux exchange areas (44 and 46 in FIG. 6) present at each coupling unit 14, there are additional effective flux exchange areas (here termed tertiary flux exchange areas 50) between each pair of units 14 that are formed as consequence of the relative close spatial distance of the individual units 14 in the array line and which exist due to the interaction of the magnetic fields of respectively neighbouring unit pairs. In the illustration of FIG. 7a, the alternating polar arrangement of five units 14 add four effective tertiary flux exchange areas 50, which also assist in confinement of the magnetic field of each individual unit 14. One effect which the tertiary flux exchange areas 50 have is an increase of flux density at the pole contact areas 42, 43 of each unit 14 if that flux density is restricted by high reluctance of the work piece 60 on which the array of units 14 act. Higher pulling forces and improved magnetic efficiency are achieved in this way, as compared to the use of a single unit 14 having the same overall active magnetic mass as the sum of the individual units 14.

(34) The spacing (or linear gap g) between the individual units 14 gives control over the total field magnitude. Short distances g between adjacent units 14 will emphasise the flux exchange between the separate units 14, with a decrease in total field intensity and overall penetration depth. Wider spacing g between units 14 will give more weight to the flux exchange between the magnetic poles of individual units 14, with an overall increase of field strength and deeper flux penetration into work pieces.

(35) FIGS. 8a to 8c show a schematic plan (bottom or top) view of a circular array arrangement of individual units 14, as compared to the linear arrays of FIG. 8. The circular array configuration of FIG. 8a is embodied in the test rig illustrated in FIG. 1 and in the magnetic lifter device 100 shown in FIG. 2, in the lifter device 100 of FIG. 2, six individual units 14 are secured in fixed but removable manner in an outer cylindrical housing part 120 that has a circular face plate 135 against which a work piece (not shown) may be abutted. An actuator module 130 which houses a not illustrated mechanical arm linkage arrangement is bolted to the rear of housing part 120 and provides a means by way of which the equally not illustrated actuating devices (eg as illustrated at 30 in FIG. 3b) of the individual units 14 can be operated to faintly activate and deactivate the individual units 14 as was described above.

(36) It will be noted that the circular array configurations of FIGS. 8a and 8b essentially represent the closing of the free ends of the linear serial arrays with alternating polarities illustrated in FIG. 6, and thereby provide self-contained array configurations where all units 14 have a neighboring unit 14, which allow interaction between unit pairs. For that reason also, circular array configurations are preferred as there is a more homogeneous force field distribution as compared to an open-ended linear, rectangular or other column-row array.

(37) In the array illustrated in FIG. 8a, six units 14 are placed with the respective magnet stacking axis A of each unit 14 extending perpendicular to an imaginary circle of radius r and the drawing plane, with the polar axis PA of each unit 14 extending substantially tangentially at said imaginary circle line that joins the stacking axes A (Le. essentially perpendicular to said radius r) and with the activated north poles of a respective unit 14 facing the activated south pole of a neighboring unit 14 and vice versa. In this array configuration, there are twelve effective flux exchange areas, consisting of six primary and secondary flux exchange areas 44/46 at the individual units 14 and six tertiary flux exchange areas 50 between neighboring units 14.

(38) In the array of FIG. 8a, there are also magnetic field interactions between the north and south poles of non adjacent units 14, however these are in practice marginal and so weak that they do not contribute to the effective overall flux exchange areas 44/46 and 50.

(39) As can be noted in comparing FIGS. 8a, 8b and 8c, circular array configurations of individual units 14 can create different effective flux exchange areas, depending on the relative orientation of the polar axis PA of each unit 14 in the global array and relative to neighbouring units 14. A so called alternating star array configuration is illustrated in FIG. 8b, wherein the same array radius r is present as in the circular array of FIG. 8a. However, in this array configuration, the individual units 14 are disposed with their polar axis PA in a radial arrangement (hub and spoke), substantially coaxial with the respective radii to each unit, with the units 14 having either the active north or south pole facing inwards and the other pole facing outwardly. At the same time, neighbouring units 14 are arranged with alternating poles facing radially inwards and radially outwards whereby active north and south poles of neighbouring units are adjacent.

(40) FIG. 8b illustrates schematically also the effective flux exchange areas that are present in this array configuration, wherein radially inward located tertiary exchange zones 52 are effective flux exchange areas between neighbouring units 14 exhibiting a relative strong exchange as compared to the radially outwardly located tertiary exchange zones 54, due to the increased distance of the radially outward located active poles of neighbouring units as compared to the inward located poles. Equally, due to the relative proximity of opposite polarity active poles of units 14 arranged on diametrically opposite sides of the overall array, there are three effective tertiary flux exchange zones 56 extending between radially facing units 14, the flux exchange zones 56 arranged in an intersecting, star like pattern.

(41) If an increase of flux penetration depth is required, the array of FIG. 8b may be varied in to the array configuration shown in FIG. 8c, wherein whilst the same arrangement of units 14 is present, the activated poles (polarities) of the individual units 14 are disposed such that ail units 14 have the same polarity at an inner radial end of the array, ie the units 14 are again arranged radially with the same pole of each unit 10 facing radially inwards with the other pole facing radially outwardly. In this array formation, the north and south poles of the individual activated units 14 are ‘paralleled’ along the circle defined by radius r and merge effectively into two annular, larger pole units, thereby defining a ring band shaped concentric effective flux exchange zone 58 formed from the individual unit effective flux exchange zone 44, 46. The magnetic field intensities are, however, not homogenous distributed along the exchange band, but reach maxima at the respective poles of the individual units 14. In effect, such array configuration does not have any tertiary flux exchange areas between neighboring units 14, and provides a flux exchange pattern that is comparable (in principle) with that of a common magnet cup design with a radially inner and an radially outer annular magnet pole.

(42) FIGS. 9a to 9c represent idealised 2-D magnetic field fine patterns as would be present at the interface of the arrays of FIGS. 8a to 8c, respectively, when in contact with a very thin ferromagnetic sheet metal or Magpaper, generated using computer assisted modelling. It should be noted that the patterns are visualisation aids only, and represent an idealised model.

(43) The field pattern illustrated in FIG. 9a is a shallow penetrating, relative confined H-field, wherein the arrangement of magnets with opposing polarities in such circular arrangement provides an effective self-regulating H-field, as is explained in greater detail below. In contrast, the field pattern illustrated in FIG. 9b, whilst also shallow penetrating, provides a relatively wider spreading H-field. Finally, the field pattern of 9c clearly illustrates a lack of magnetic interaction between neighboring magnets beyond the resultant compression of field lines of adjacent magnets in the array, whereby the magnetic energy is enlarged and achieving a H-field with deeper penetration perpendicular to the plane of drawing.

(44) As will be apparent from the above description, the number and choice of the sizes of individual magnet units 14, and the spacing layout, can be determined depend on the intended area of use of a magnetic device incorporating the magnet array, eg coupling devices, lifters, etc, but in particular the properties of the ferromagnetic body in contact with which the array is to be brought. For example, the magnetic lifter test-jig illustrated in FIG. 1, employing an array of 5 switchable magnets Version M1008 by Magswitch, with a spacing of 1 mm between them, can exert a pulling force of 145N on a 0.8 mm iron sheet. The pull on a second sheet in direct contact underneath is hardly noticeable in this case.

(45) The following table illustrates some of the basic advantages of subdividing a given mass of magnetically active material into discrete sub-masses and placing the so subdivided masses into a specific array configuration, as per the invention. The table summarises results of a lifting experiment conducted with 6 types of magnetic lifters, the first three in the table being magnetic lifters incorporating an array of six switchable magnets of the type Magswitch M1008 (ie as illustrated in FIGS. 2 and 3, the cylindrical magnets having a dimension of 10 mm diameter and 8 mm height), whereas the subsequent three members in the table employ one larger, switchable magnet of the type M2020, M3020 and M5020 (ie 20 mm diameter×20 mm bight magnets, 30 mm×20 mm and 50 mm×20 mm, respectively). In the table below, ‘Alt Star Array’ designates an array configuration as per FIG. 8b, ‘Joint Star Array’ designates an array configuration as per FIG. 8c and ‘Circular Array’ designates an array configuration as per FIG. 8a.

(46) TABLE-US-00001 Active Pull on 1 Pull on 1 mm magnetic Peak mm sheet fully partial activated material Pull activated to match saturation Volume mm.sup.3 in N in N levels in N 1008 × 6 3768 420 260 Self regulating Alt. Star Array 1008 × 6 3768 450 200 130 Joint Star Array 1008 × 6 3768 220 200 Self regulating Circular Array 2020 8283 450 180 80 3020 14137 750 270 110 5020 39270 1500 320 100

(47) A number of observations are worthwhile. It will be noted that the maximum lifting capacity (peak pull in N) of a single M5020 magnet is only about 3.57 times that of the Alt. Star Array configuration, despite having a total active magnetic material mass of more than 10 times that of the array. The same array, when in engagement with a ferromagnetic sheet having a thickness of 1 mm will have a pull in N which is only 60 N lower than that of the single 5020 magnet, and 60 N higher than a single 2020 magnet which has about double the active material mass contained in the Alt. Star Array lifter. It will also be noted that when a single magnet unit 3020 is switched into a magnetisation state to match the magnetic saturation level capable of being carried by the 1 mm thick metal sheet, so as to practically confine the flux path into the sheet metal work piece and avoid the magnetic field to extend beyond it) that the pulling force is about 1/7 of the peak pull force and less than ½ the value as compared to its fully activated state (in which the magnetic field would extend beyond the thickness of the sheet metal). That is, with single magnets, lowering the magnetising force to avoid H-field extension beyond the work piece boundary, if magnetic flux is ‘bottlenecked, results in a drop of pole flux density, and consequentially a reduction in available pulling force. The array configuration provides for enlargement of the ‘bottlenecked’ flux area, due to the presence of the additional flux paths between neighboring array members, thus loading to an increase in overall pole flux density which results in higher pulling forces.

(48) Of particular interest is, however, that both the Alt. Star Array and the Circular Array configurations exhibit what might be termed a self-regulating H-Field, allowing the pulling force to remain higher than in any of the other lifters listed in the table.

(49) This phenomenon will be explained with reference to FIGS. 10 and 11. In FIG. 10, an idealised 2-D model magnet torus 80 is illustrated, wherein an otherwise closed 6-pole magnet torus is opened at 6 discrete locations 82a to f, thereby defining 6 dipole magnets 84a to 84f which in effect provide an arrangement similar to the circular dipole array configuration of FIG. 8a when activated (but for the slightly curved polarisation axes PA′ of the dipoles 84a to f, given that they are not linear dipoles.

(50) The idealized H-field pattern of a ‘closed circuit’ circular magnet array 80 with alternating polarities N-S in which neighboring magnets 84a to 84f are ‘short-circuited’ (either by bringing the peripherally facing magnet faces into abutment or by inserting a passive ferromagnetic pole piece into each gap so as to bridge each N-S pole pair of adjacent magnets) would be self-contained within the closed circuit and not available for use in nor accessible by an external working circuit. Opening of the torus at one or more locations (e.g. the six gaps 82a to f identified in FIG. 10) provides a number of portals, each of which allow ‘access’ to the magnetic energy stored in the active magnetic material of the (torus) array.

(51) It will be noted in the opened torus 80 that at each gap 82 between neighboring magnets 84a to f, a flux exchange zone exists between opposite N- and S-poles of adjacent magnets 84a to f, thereby providing a flux path through the medium present in the gap volumes, and the overall array arrangement will provide a first (closed) magnetic circuit consisting of the magnets 84a to f and gaps 82a to f. When a ferromagnetic object is brought into magnetic interaction with one or more of the portals defined across 82a to f, magnetic flux available in the ‘tank’ circuit provided by the array is able to divert or ‘split’ at the portals and transferred into the object. A second (closed) circuit consisting of the ferromagnetic object, passive pole extension pieces (not shown) at the N- and S-poles of the adjacent magnets 84a to f against which the object is brought in contact and the two or more magnets 84a to f which the ferromagnetic object bridges can thus be formed, which has a magnetic reluctance that is lower than that of the first circuit, i.e. the array circuit.

(52) The proportion of flux splitting into the second circuit will depend on the reluctance of both circuits. Put another way, if both the first and second magnetic circuit exposed to the same magnetomotive force have the same permeability, an equal flux sharing takes place. Increase of circuit reluctance in one of the circuits will result in a shift of flux from that circuit into the other and vice versa. This basic principle is embodied in the above described Circular and Alternating Star array configurations of FIGS. 8a and b.

(53) The flux-splitting functionality aspect of the present invention may be best exemplified with reference to FIGS. 11a and b, which are schematic side views of two switchable permanent magnet units 240, 242 of the type illustrated in FIG. 3b, and which are arranged in a linear array as illustrated in FIGS. 5 and 6, in fixed positions next to one another with a small air gap 241 between the facing opposite N and S polarities (eg pole pieces 246, 248) of the units 240, 242. It will be appreciated that such idealised two-magnet arrays are also present in the circular arrays of FIGS. 8a and b, as well as the opened torus of FIG. 10.

(54) In FIGS. 11a and b, line 244 simply serves to denote an idealised reluctance free bridge to achieve a closed (short) circuit between the S- and N-poles which do not face one another across the air gap 241 that is maintained between the other N- and S-poles of the units 240 and 242, so that only one portal exists in such arrangement.

(55) Turning first to FIG. 11a, in the absence of a work piece (eg sheet metal piece 250 in FIG. 11b), a flux exchange path between the two magnets 240, 242 exists across the air gap 241 (the circuit being otherwise closed as indicated at 244). The magnitude of flux at a given magnetising force depends here mainly on the width and cross section of the air gap between the magnets 242, 240. Since the permeability of air is linear with flux density, the whole flux transfer behaviour in this part of the path is linear. Reluctance of the air gap magnetic circuit is thus dependent on the flux transfer area geometry and the permeability of the material in the gap, which might be a substance other than air but which should have ideally a very low relative permeability (that of air being about 1), but in any event considerably lower than the relative permeability of the work piece.

(56) As seen in FIG. 11b, when a ferromagnetic work piece 250 with a higher permeability than that of air is brought into magnetic interaction with opposite poles of adjacent magnets 240, 242, an additional flux path between the opposite poles of magnets 240, 242 is created, which has a reluctance that is lower than that across the air gap 241. The amount of flux that will ‘pass’ through this path (or circuit) is governed mainly by the permeability of the work piece material (if the work piece has a small thickness). Flux is ‘drawn’ from the first (air gap) magnetic circuit and diverted into the second (work piece) magnetic circuit. The permeability of the work piece will be initially very high, ie several thousand times higher than air), until flux saturation is reached in the work piece. The permeability of the second circuit will gradually decrease (as the flux density increases), as per the relevant non-linear B-H magnetisation curve applicable for the work piece material, until saturation is reached. The reluctance in the second circuit will then be equal or higher than that of the air gap circuit, and no further magnetic energy will be ‘withdrawn’ from the air gap circuit.

(57) As FIGS. 11a and b illustrate, a flux that may have an initially higher value across the air gap, eg 0.48 Tesla, in the unloaded ‘tank’ circuit, will be split when the work piece bridges opposite poles N and S of adjacent magnets 240, 242, and a lower flux will remain in the air gap 241, eg 0.11 Tesla, once saturation of the diversion circuit across the work piece is finalised

(58) Effectively, magnet array configurations which are devised with the above criteria in mind will provide a magnetic device exhibiting a self-regulated magnetic field strength when brought into magnetic interaction with a ferromagnetic work piece, the non-linear permeability of the work piece serving the purpose of regulating and stabilizing the available magnetizing force (magnetic field strength H) at the access portals within the first magnetic circuit. It should be added here that the overall level of magnetic energy that can be withdrawn from the array through the portals is inverse proportional to the distance between adjacent magnets.

(59) Whilst the above described magnet array configurations utilise switchable permanent magnet units 14, 140, 240 as described also in the above mentioned patents, it will be understood that other dipole magnet units may be employed. The N-S magnetization axis may also not necessarily be straight linear, but could be in particular in the case of circular array formations slightly curved.

(60) The specific geometry of the pole pieces that interact with the active magnetic material in the (switchable) magnet units may also be adapted and varied as required to achieve a desired flux transfer pattern from the active magnetic material into a work piece.

(61) Equally, the material and shape of the housing in which the array of magnets will be held is to be chosen to suit the specific application, as is the precise layout of the array configuration, within the confines noted above.

(62) It will equally be appreciated that FIGS. 9a to c, 10 and 11 illustrate idealised and simplified 2-D models of flux paths, magnetic field geometries and similar, which are based on 3-D artefacts, and which are influenced by numerous other effects and boundary conditions that open and closed (or loaded) magnetic circuits are subject to, eg imperfect magnetic paths, magnetic field leakage, etc. Also, computer modelling introduces some simplifications and inaccuracies in creating the drawings, so that these are to be seen as illustrative only of general principles.

(63) Although the present invention has been principally described with reference to concepts that may find particular application in magnetic lifter and coupling devices. It will be appreciated that magnet arrays can readily be applied to other devices where a magnetisable (ferromagnetic) work piece is to be secured at such device either for holding same, or moving same securely attached to the device, and vice versa.