WO2009141600A1

WO2009141600A1 - Pile levelling system

Info

Publication number: WO2009141600A1
Application number: PCT/GB2009/001249
Authority: WO
Inventors: Adrian Richard Marshall
Original assignee: Crafty Tech Limited; Food Processing Faraday Partnership Limited
Priority date: 2008-05-20
Filing date: 2009-05-19
Publication date: 2009-11-26
Also published as: GB0809115D0

Abstract

There is described a conveyor assembly comprising at least one drive roller (1), a conveyor belt (5) and a profiled support member suitable for spreading out a point feed of particulate product over a wider area (1). The support member comprises first imput end comprising an upwardly facing concave shape. Adjacent to the second output end of the support member, the support includes a portion with convex cross section. There is also described a method and apparatus related thereto.

Description

Pile Levelling System

Field of the invention

The present invention relates to an assembly for spreading out a point feed of particulate product over a wider area and to a method and apparatus related thereto.

Background to the invention

Particulate material when dropped, for example from a filling nozzle, will naturally tend to adopt a "piled" repost, which may for example, be frusto conical in shape. This is illustrated in Figure 2 herein. The particular size and shape of the pile may depend, inter alia, on which depends on a variety of factors such as particle shape, size, stickiness etc.

In manufacturing and/or packing industries, such as food manufacturing/packaging, a "pile" of material is undesirable since this may lead to unsatisfactory filling, etc. In packaging industries in particular, this is usually avoided by scraping the material pile level so that it presents a substantially horizontal upper surface. Conventional levelling and re-distribution systems use scrapers, ploughs, contra-rotating rollers or spiral brushes to relocate high spots to lower regions.

However, all of these techniques suffer from the disadvantage that they invariably apply localised compressive forces to the product and/or will exert substantial shear forces on the product at or near the 'cut ofP height. These compressive forces tend to compact which may alter the flow properties and ease of packaging. Further the shear forces may damage fragile products. There is therefore a need for a pile levelling system which achieves the desired product levelling but without using undesirable compressive or shear forces and thereby minimises the risk of damage to the product.

A conveyor whose belt follows the desired sequence of flexing over its length is described, together with details of methodology for creating appropriately shaped humps that make such a belt practical by ensuring that belt stretching is minimised and belt tracking can be achieved.

Summary of the invention

The present invention therefore provides a low stress method of spreading out a particulate product over a wider area by use of a levelling conveyor which achieves the desired product levelling substantially without using compressive or forces and so with minimum risk of damage to the product and an assembly related thereto.

Therefore, according to a first aspect of the invention we provide a conveyor assembly comprising at least one drive roller, a conveyor belt and a profiled support member.

The profiled support member may be a dead plate or base plate as is understood in the art but which is profiled to provide one or more concave and convex regions. It is the use of the profiled support member according to the invention that achieves the 'unfolding' of the material pile by an appropriate sequence of flexations of the base of the conveyor belt on which the pile is initially deposited. The sequence of flexations produces a fϊuidising effect on the material which allows the material to settle under its own weight

Thus, it is desirable that the profile of the conveyor support may vary along the length of the support. The support will generally comprise a first "input" end which is position at the end of the conveyor where the material is fed, and generally piled, onto the belt, and a second output end where the material leaves the conveyor to the next stage of the manufacturing or packing process. However, it should be understood that the preferred loading point, e.g. for the deposit of material, will be in the deepest part of the concave region rather than at the conventional start of the belt, since feeding the material at the infeed roller, then the forming of the initial V by the hump, may lead to compression of the product which is clearly undesirable. Thus, the profiled support creates one or more humps or troughs, i.e. convex or concave regions, in the conveyor belt and the presence of the humps or troughs in the belt encourage the product to settle under its own weight. It will be understood here by the skilled person that the input end of the hump is different to the input end of the conveyor, since the input end of the conveyor will be substantially flat to enable it to go around the end roller of the conveyor assembly. Thus, it is the support shape alters between the infeed and the outfeed, and is essentially of a fixed geometry, substantially set when designing the "hump" for a given application which is further described herein. It will be understood that a conveyor belt will naturally tend to take a straight line and therefore the role of the hump or trough is to ensure that there is a curved element to the path of the belt. Therefore, the longitudinal cross section of the support may at its input end comprise an upwardly facing concave shape e.g. in the form of a valley, i.e. may be substantially V-shaped whilst at the output end the support may be substantially flat or planar. Intermediate between the input end and the output end the cross-sectional profile of the support may vary. Therefore, from the input end to the output end the support preferentially defines an area of decreasing concavity and it is the travel of the belt over area of decreasing concavity that causes the opening up of the product and reduction of the pile as hereinbefore defined. The travel of the conveyer over the concave depression causes the pile of the material to collapse around itself and therefore form a substantially planar upper surface, although the outer edges of the surface may still retain a slight incline. The peripheral incline may also be removed by including a portion of the support which is convex in cross section. Preferably this convex portion of the support will be adjacent to the output end of the conveyor.

The hump or trough should not extend the complete length of the belt, e.g. to the end rollers, since in order, for example, to prevent bouncing as the belt goes from the hump to the substantially flat profile, which is necessary to enable the belt to travel around the end rollers, one or more conventional dead plates or base plates can be used. Thus, combinations of travel paths which may be used include, but shall not be limited to:

roller / free belt/ hump (valley to Ridge)/ free belt/ roller; and deepening V dead plate/ hump (valley to Ridge)/ Decreasingly ridged dead plate/ roller The conveyer assembly support of the invention may comprise a static device, for example a pair of hinged plates, but can also be carried out as a continuous process by contriving a conveyor belt whose surface carries out this pattern of flexing with the additional requirement of going round straight rollers at both ends of the conveyor. When the conveyor assembly comprises a pair of hinged plates, the pair of hinged plates can be made into a concave shape, e.g. a valley shape. In use, the particulate produce may then be poured into the concave or valley form into a pile oi peak. The hinged plates may then be opened out bringing the pile to a substantially flat profile.

The rollers used in the conveyor assembly may generally be conventional rollers. However, in one aspect of the invention one or more of the rollers may be a barrel shaped roller. Barrel shaped rollers are desirable since, inter alia, they improve belt tracking. Furthermore, the degree of slope on the barrel of the rollers may be used to further correct the belt tension.

It is desirable that all sections across the belt should be at a similar tension, to minimise belt wear and motor loading, etc. In one aspect of the invention this may be achieved by ensuring that the path length at all sections across the belt's width are substantially the same or similar.

The assembly may also include a flat 'arrowhead' platform that defines a series of sections that take the belt from flat to a valley, then to flat to a ridge and back to flat. Such an arrowhead platform maybe integral to the concave/convex platform hereinbefore described. It is generally desirable to minimise the difference in path lengths across the width of the conveyor belt by offsetting the "arrowhead" upstream (which also makes it asymmetric to accommodate the now dissimilar rising and falling belt angles). It will be understood that moving the "arrowhead" upstream makes the rising belt shorter and steeper, with its edges becoming relatively tauter, whilst simultaneously making the falling belt longer with its centre becoming less taut. This is advantageous in that, inter alia, it reduces the variation in the tension across the belt's width.

In the method of the invention the conveyer belt may be lengthened and the upper surface of the support built up to match the angles of the rising and falling sides of belt and so give a smooth belt transition. The methodology comprises defining the valley and ridge angles. Thus, given the target belt width and length the centred support lift height can be calculated. The belt lengths at a number of sections across the belt can then be calculated and graphs may be produced of the belt section path length vs position across belt width like. Moving the arrow upstream and allowing it to go asymmetric will allow the reduction in the variability of path lengths across the belt width.

A curve is desirable to give a smooth transition between the rising and falling sections of belt even for the tightest section of the belt. It is desirable to identify a range of suitable curves that work with an increased length of belt to give smooth transitions and also give appropriate additional path length. The arrow position may be varied to give enough difference in achievable path lengths to take up the slack at some areas across the belt and so give even belt tension in a given case. The curve calculation can be approached numerically, e.g. as a segment of an ellipse and a Bezier curve. This is further defined in the examples herein.

It is important that the belt used in the conveyor assembly of the invention has sufficient shear strength to withstand the stresses created by the profiled support. Usually the weakest point on a conveyor belt is the joint between each end of the belt. Therefore, it is desirable that joints of the ends of the conveyor belt have a shear strength comparable to that of the rest of the belt. To enable the belt to flow over the hump, it must be able to shear locally, i.e. the belt must be weak in shear otherwise the tension in the belt required to cause shearing will be so large that the belt will not slide over the hump. Thus, mesh belts joined using conventional 'adhesive' type joints have a non-shearable section. In order to provide adequate shear, which is necessary for the belt to be utilized with the profiled support of the present invention, we provide a belt joint that maintains tensile strength but has a lower (than usual) shear strength or lower rigidity, i.e. greater flexibility. It will be understood that shear strength is normally not relevant to conveyor belt manufacturers, but in the present case it is important to allow redistribution of the tension incompatibilities across the width of the belt, i.e. caused by the slackness of the belt in the concave region and taughtness in the convex region. Hence, it is desirable to use a low friction mesh belt. Conventional mesh belt joints may be problematical in that, inter alia, they may be very rigid, e.g. in a shear direction, because the ends are just overlapped and then sandwiched between adhesive films on each side which makes the belt solid at the joint. Thus, such a conventional belt will not shear and thus will not travel over the hump smoothly. Thus, it is desirable for the joint to be discontinuous and to possess similar shear characteristics to the surrounding belt. In a preferred aspect of the invention the joining method comprises creating short loops on the end of the belt by stitching the last belt mesh back on itself and then cross-stitching between loops. Although a variety of thread materials may be used a preferred thread material comprises man-made fibres e.g. polyesters, polyolefϊns, acrylics, PVC, regenerated cellulosics and their derivatives, polyamides and the like, aramid threads e.g. para and meta-aramid threads such as Kevlar®.

Similarly, the belt may comprise a variety of materials depending upon, inter alia, the shear stress that it will experience, etc. Desirably the belt comprises a low friction material, such low friction materials may include, but shall not be limited to polyolefϊns, such as polyethylene, halogenated polyolefϊns, such as, polytetrafluoroethylene (PTFE), and aramids fibres e.g. para and meta-aramids, such as, Kevlar®. Most preferably the belt material may comprise a fabric coated with a low friction material, such as PTFE₅ e.g. PTFE coated Kevlar®. The belt may comprise a fabric mesh since a mesh will have virtually no intrinsic shear strength whilst retaining robustness as hereinbefore described. The size of the mesh may vary and the selection of the mesh size may be based upon the apertures of the mesh being small enough so that the product does not fall through the mesh, but are big enough so that the shear strength is reduced. Thus, the mesh belt will comprise a series of interconnected strands defining apertures therebetween and the strand size and the aperture size may therefore be varied depending upon, inter alia, the nature of the product, the detail and flexibility of the strands, etc. In order to maintain belt tracking the assembly desirably include one or more tracking guides. Although a variety of such guides may be used, one form of guide is a guide running in slots or grooves along the hump of the assembly. The positioning of these grooves in the hump preferably allows for the narrowing of the belt due to, for example, local shearing of the belt, which allows the hump to produce an even belt tension. Side guides may also be provided on or adjacent the belt so that, when in use, the product will spread up against them. Such guides can optionally be fixed. Preferably the belt can be fitted with corrugated sidewalls which can, for example, be heat-sealed onto the tracking guides hereinbefore described.

The profiled support member as hereinbefore described is also novel per se. Therefore, according to a yet further aspect of the invention we provide a profiled conveyor support in which the support comprises at least one concave region.

Desirably the profiled support member also includes a convex region. Furthermore, the profiled support member may comprise a single unit or may optionally comprise one or more modular units. One advantage of a modular unit is that the profile of the unit may be varied according, inter αliα, to the nature of the material being transported on the conveyer. In addition, the support may comprise a flat 'arrowhead' platform that defines a series of sections that take the belt from fiat to a valley, then to flat to a ridge and back to flat.

The conveyer and profiled support according the invention also provide novel methods of material transfer and a novel method of levelling a pile of particulate material. Therefore, in a yet further aspect of the invention we provide a method of conveying a material which comprises the use of a conveyor assembly as hereinbefore described.

The method of the invention is especially advantageous in conveying particulate foodstuffs and more particularly in levelling piled particulate foodstuffs although it will be understood that the method and assembly of the invention may suitably be used with any conventionally known particulate material.

It will be further understood by a person skilled in the art that the method of the invention may be used in conjunction with a wide variety of particulate foodstuffs. Such particulate material may comprise flakes, powders granules, agglomerates and the like. The method may be applicable to any conventionally known foodstuffs that may be piled during the manufacturing or packaging process. The method is particularly suited to fragile, brittle, sticky, interlocking or compactable products. Thus, the method may be suitably used for any product for which compression is undesirable, for example, freshly grated cheese strands present considerable difficulties from a manufacturers perspective and may be suitably treated with the method of the present invention. In addition the main application of the conveyor assembly of the present invention will be in the spreading out feeds of materials evenly across, for example, oven bands and steamer bands, so as to ensure even heat transfer. For example, this would include drying and baking breakfast cereals and baked crisps, steaming of pasta shapes etc, or evenly loading vegetables into continuous washing tanks. The conveyor of the invention with its progressively changing profiled belt surface is advantageous in that, inter alia, it provides a continuous non-compacting levelling action, which enables piled feeds of fragile items to be dismantled and levelled with minimal stress, and fed onto subsequent processes as a uniform thickness layer, or as a monolayer.

The levelling conveyor takes a heaped feed, deposited onto a belt, and unfolds it by manipulating the profile of the underlying conveyor, facilitating the product's flow under its own weight into a level distribution. This is a minimal stress manipulation with product moving under gravity.

According to a further aspect of the invention we provide a kit suitable for spreading out a particulate product, the kit comprising a conveyor assembly including at least one drive roller, a conveyor belt and a profiled support member.

The invention will now be illustrated by way of example only. The following examples are intended to illustrate the invention and are not to be construed as being limitations thereon. Any abbreviations used are those conventional in the art. The invention will now also be illustrated with reference to the drawings that follow in which Figure 1 is a perspective view of a conveyor assembly according to the invention;

Figure 2 is a schematic representation of a particulate product's 'angle of repose'; Figures 3a to g are each cross-sectional view of the conveyor support at different points along its length; Figure 4 is a perspective view of a conveyer fitted with a flat 'arrowhead' platform; Figures 5a to c are schematic representations of path length variations at different sections across a belt raised by a flat arrowed platform;

Figure 6 is a schematic representation of a typical hump blending the belt profile; Figure 7 is a schematic representation of the elements contributing to the path length of the conveyor belt;

Figures 8a to c illustrate how a segment of an ellipse can be used to give a smooth transition between the rising and falling areas of the belt;

Figures 9a to c illustrate how Bezier curves can be used to give a smooth transition between the rising and falling areas of the belt; Figure 10 is an illustration of an analysis of a 'natural' hump profile; and

Figure 11 is a perspective view of the mounting of side walls on the conveyor assembly (not shown).

Figure descriptions Figure 1 shows an overview of a levelling conveyor with its highly deformed belt 5 changing from flat around the infeed roller 1 to a valley 11 through to a ridge 12 then back to flat outfeed 2 as it rises over a carefully profiled hump 3.

Figure 2 illustrates a product's 'angle of repose', the angle 13 that a pile of particulates 14 will naturally tend to adopt, which depends on a variety of factors such as particle shape, size, stickiness etc.

Figure 3 illustrates a number of stages in manipulation of the support 22 under a pile 21 that can, in sequence, create a more level distribution: Figure 3a shows the underlying surface at rest Figure 3b shows the surface bent into a valley, and the product piled into that valley

Figure 3c shows the valley partially flattened out, which opens up the centre of the pile producing a hole for the peak to collapse into Figure 3d shows the material from the peak collapsed into the voids created

Figure 3e shows the surface flat with a substantially level distribution apart from at the edges

Figure 3f shows the surface folded back on itself allowing the product to spread to the edges according to the product's angle of repose Figure 3g shows the surface unfolded back flat, with an even distribution of material ready for use on subsequent processes.

Figure 4 shows a simplified conveyor bed profile that has a flat 'arrowhead' platform 26 that defines a series of sections that take the belt from flat to a valley, then to flat to a ridge and back to flat.

Figure 5 shows examples of path length variation at different sections across a belt raised by a flat arrowed platform

Figure 5a shows a centrally positioned platform 26, which gives a 7% variation 27 in belt path length across the belt's width,

Figure 5b shows the platform 28 moved upstream somewhat, and made asymmetric to suit the dissimilar rise and fall angles, which gives a 2% variation 29 in belt path length across the belt's width Figure 5c shows the platform 30 moved further upstream, and made asymmetric to suit the dissimilar rise and fall angles, which gives a 0.6% variation 30 in belt path length across the belt's width

Figure 6 shows a typical hump 3 that smoothly blends the belt profile from flat through valley 11 and ridge 12 to flat, but whose exact shape and position has been carefully calculated to give consistent belt tension across the belt width.

Figure 7 identifies the elements contributing to the path length at a given section across the belt's width, which include the hump arc length 19

Figure 8 illustrates how a segment of an ellipse can be used to give a smooth transition between the rising and falling areas of the belt, and how using different sorts of ellipses gives different path lengths in the three cases shown.

Figure 9 illustrates how Bezier curves can be used to give a smooth transition between the rising and falling areas of the belt, and how using different 'sharpness' Bezier curves gives different path lengths in the three cases shown.

Figure 10 shows analysis of a 'natural' hump profile, where the valley and ridge profiles have been imposed on a sample of belt, but the belt has been left free to find its own transitional form, which has then carefully mapped, and the corresponding hump has been manufactured to echo this 'natural' form Figure 11 shows the mounting of side walls 43 and tracking guides 7 with the mesh belt 5 trapped between them.

Description Referring to figure 3 there are a number of key stages in achieving the levelling action by flexing the support base:

Valley to flat

Depositing a pile 21 into a support which is already folded to give a 'valley' section 11 (preferably an angle related to the product's natural angle of repose) and then opening out this valley to a flat surface, can be thought of as creating a void 23 at its centre into which the peak 24 will collapse under its own weight as shown in figures

3b and 3c. In fact the opening up action will tend to generate a fluidisation effect through the majority of the body of the product, allowing particles to settle relatively freely under gravity.

The resultant product distribution will tend to have ramped edges (unless the initial fill in the valley reached the sidewalls 43)

Flat to ridge

These ramped edges 48 can preferably be removed by a second stage, which consists of flexing the support backwards so that it forms a ridge 12 (preferably at an angle related to the product material's natural angle of repose), at which point any ramped edges 48 will be filled in by 'scree-like' flow 25 of the product material slipping down until it reaches the support sidewalls 43 as shown in figures 3f and 3g. At this point the product material will be at substantially the same depth across the conveyor width.

Continuous version: changing belt profile

The actions described can be achieved in a static device, for example a pair of hinged plates, but can also be carried out as a continuous process by contriving a. conveyor belt whose surface carries out this pattern of flexing with the additional requirement of going round straight rollers at both ends of the conveyor.

The belt surface therefore should preferably go through the full sequence shown in figures 3a-g, flat, valley, (flat), ridge, back to flat.

Arrow shaped platform Referring to figure 4, since the concave regions of belts cannot be held under tension, a valley must be achieved by raising the infeed edges 44 up above a taut centre path 45, and a ridge created by keeping the centre 47 up above taut exit edges 46, and this can be visualised schematically by raising an arrow shaped platform 26 above a conventional flat conveyor bed as shown in figure 4, so creating the required valley and ridge.

In principle just an arrow shaped plate will be sufficient but the use of a flat arrow not only produces sharp edges for the belt and so locally stresses it (particularly at the nook of the upstream V) but the difference in path lengths at sections across the belt are large, as illustrated in figure 5a, so even with an elastic belt, areas of the belt would be under great tension while others would be slack or nearly loose.

Getting belt to deform in this way goes against conventional industry practice, and a number of special techniques have had to be used to make the levelling conveyor of the present invention practical.

Achieving uniform belt Tension

Ideally all sections across the belt should be at a similar tension, to minimise belt wear and motor loading resulting from the high drag produced by taut lengths of belt being pulled over the convex hump, and to achieve this, preferably the path length at all sections across the belt's width should be made the same.

Arrow platform offset The first step is to minimise the difference in path lengths across the width of the belt by offsetting the arrow head upstream (which also makes it asymmetric to accommodate the now dissimilar rising and falling belt angles).

Moving the arrow head upstream makes the rising belt shorter and steeper, with its edges becoming relatively tauter, whilst simultaneously making the falling belt longer with its centre becoming less taut.

These effects combine to reduce the variation across the belt width as shown in the examples in figure 5 The case illustrated in Figure 5a with a central arrowhead 26 would require a variation in the belt path length between the belt edges and the centre (27) of 7% of the conveyor length.

Putting an arrowhead upstream 28 as shown in figure 5b reduces this to 2.4% (29).

However this technique cannot equalise all the path lengths in this way; even if the arrowhead is moved back (upstream) significantly (30 as shown for a particular geometry in figure 5c), where the edges and the centre have the same path lengths, there is still about a 0.6% variation (31) remaining part way across each side of the belt.

Hump profile

Referring to figure 7, in order to correct the remaining differences, the whole belt can preferably be lengthened somewhat, and then the upper surface of the arrow can be built up in a hump 3, to firstly match the angles of the rising 16 and falling 18 belt meeting the hump at the correct angles 15 and 17 (and so give a smooth belt transition) but also by making the different sections of the hump 3 stand higher, the remaining difference between slack and taut sections can be accommodated. An example of such a hump is shown in figure 6

The methodology for designing a hump is first to define the valley and ridge angles (preferably they should be related to the angle of repose of your product). Given the target belt width, and length, the centred arrow lift height can then be calculated. The belt lengths at a number of sections across the belt can then be calculated and graphs produced of belt section path length vs position across belt width like those in figure 5. Moving the arrow upstream and allowing it to go asymmetric will allow the reduction in the variability of path lengths across the belt width to be observed.

A hump, to accommodate the remaining differences across the belt width, now has to be designed.

Even for the tightest section of the belt, a curve is needed to provide a smooth transition between the rising and falling sections of belt, and the slackest section of the belt will need a high hump to 'lose' significantly more belt length than this.

The task then is to identify a range of suitable curves that work with an increased length belt to give smooth transitions and also give appropriate additional path length.

However there are additional constraints since there is a maximum path length (if the path length is too high the hump's peak will become too sharp and very shallow humps will give a insufficiently smooth entry and exit curves), so the arrow position needs to be determined to provide enough difference in achievable hump path lengths to take up the slack at some areas across the belt and so give even belt tension in a given case. Curve calculation

For a given pair of entry angle 15 and exit angle 17 there are many smooth curves 19 that could be used, but calculating the path length for any other than a few special cases of curve is difficult and making hump design also difficult.

Two cases can be approached mathematically (or at least numerically), namely a segment of an ellipse and a Bezier curve.

Curves from an ellipse Referring to figure 8, for a given set of entry 15 and exit angles 17 and flat path length 20, there will be a family of elliptic segments that pass through the end points, and match the gradients.

Members of the family will have different sizes, aspect ratio, orientation and centre coordinates, and each one will have a different path length, and it is possible (though not trivial) to back calculate from a desired path length to a unique ellipse segment.

Three independent parameters are required to define a rotated ellipse (preferably the major axis, minor axis and angle of rotation), and to define a straight line takes four independent parameters (preferably the x and y coordinates of both its endpoints). Requiring the line endpoints to lie on a rotated ellipse provides two non-linear equations relating the seven parameters required to define a segment of a rotated ellipse. The segment-of-the-ellipse hump shape that we wish to use to accommodate belt slack in a particular case is uniquely determined by providing five additional independent geometrical constraints, namely a specified length and gradient for the segment chord, a specified length for the segment arc, and a specified angle between the tangent to the ellipse and the horizontal at each endpoint

Expressed mathematically, these five further constraints thus complete a set of seven non-linear simultaneous equations relating the seven unknown parameters.

By algebraic manipulation, this set of equations can be simplified to an irreducible set of three equations in three unknowns, preferably the x and y coordinates of one of the endpoints, and the angle of rotation of the ellipse.

These may be solved by numerical methods, and the other four unknowns are then determined by back-substitution, allowing the parameters that define an appropriate elliptic hump segment with the required additional path length to be calculated.

Figure 8 gives three examples from a family of ellipses for a particular configuration, showing widely differing path lengths.

In figure 8a the segment (32) from a large fat tilted ellipse gives a path length of 4% longer than the flat path length, the segment (33) from an intermediate ellipse in figure 8b gives 2.6% longer path, and the segment (34) from small narrow nearly horizontal ellipse gives only a 0.6% increase. Different ellipse segments should be calculated for a number of sections across the belt, with the differences between their path lengths contrived to take up the remaining slack.

Manufacturing a hump from these elliptic sections can be done by laminating thick ellipses cut using an ellipse router guide rig or using rapid prototyping techniques.

Bezier curves

Bezier curves are commonly used in graphics packages, and are relatively easy to compute, but it's not easy to predict the curve length for a particular case in advance, (although having once defined points on a Bezier curve, the length can be simply approximated by summing the length of the series of straight lines joining points calculated to draw the curve).

Referring to figure 9, for a two point Bezier curve, the shape is defined by the angle and length of 'handles' 38 &39 controlling the direction and 'inertia' of the line at each end.

In the case of the present invention the handle direction is simply the direction of the adjoining belt, but sensible values for the handles' lengths range from 0 to 100% of the lengths of the sides of an unsmoothed triangular hump 49.

Changing both end's handle position together from 0 to 100% gives a family of different 'sharpness' Bezier curves, whose length can be then calculated retrospectively. An iterative technique can be used to find the a Bezier curve with the desired path length, for example, try a particular sharpness, calculate the approximated resultant length, and increase the sharpness a bit if the length is not enough, or reduce the sharpness if the length is too long.

The 95% sharpness Bezier in figure 9a gives a path length (35) that is 6.7% longer than the flat path length (20), the 50% sharpness Bezier in Figure 9b gives a curved path (36) that is 2.4% longer path and the 33% sharpness Bezier in figure 9c gives a curved path (37) which is only a 0.9% increase.

The manufacture of a Bezier based hump can be done, for example, by laminating

Bezier curves generated in CAD and printed out to form full sized templates, or using rapid prototyping techniques.

Natural hump

Referring to figure 10, an alternative approach is to force a sample of belt to the required valley 40 and ridge 41 forms, but allow it to adopt its own transition route

42.

If this 'natural transition form' that the belt adopts is mapped, an appropriate 'natural hump' can then be fabricated which will support the belt in this form when it's loaded with product.

The contours of such a 'natural hump' are shown in figure 10 Belt requirements

The use of correctly profiled humps will ensure that the tension across the belt is even, but only if the belt is able to shear locally to allow different sections to creep ahead of or fall behind their neighbouring slices at different points along the belt.

An elastic or stretchy solid belt might be able to do this but will experience a lot of drag over the hump, so preferably a mesh belt should be used, because it will have virtually no intrinsic shear strength whilst being otherwise robust, and so will permit the belt shear required.

Belt joints

Most conventional mesh belt joining techniques such as overlapping, interweaving or bonding, tend to locally stiffen the belt, reducing its ability to shear, which reduce the belt's ability to deform and so obstruct the creation of the valley and the ridge.

Preferably the joint's shear strength should be comparable to that of the rest of the belt, and a jointing method involving creating short loops on the end of the belt by stitching the last mesh back on itself and then cross-stitching between loops with Kevlar® thread has produced joints with shear properties that are similar to the rest of the fabric.

The belt needs to be low friction e.g. PTFE coated Kevlar®, so that it doesn't generate too much resistance as it's dragged over the hump under tension, and the mesh size needs to be compatible with the product's dimensions. Belt tracking

In order to maintain belt tracking (particularly at the valley end) the belt should preferably have tracking guides or grooves running in slots along the hump. The positioning of these grooves in the hump will need to allow for the narrowing of the belt due to the local shearing of the belt which allows the hump to produce even belt tension.

Belt edging On the ridge stage particularly, side guides are required up against which the product will spread, and these can be fixed, or preferably the belt can be fitted with corrugated sidewalls 43 which for example can be heat-sealed through onto the tracking guides 7 mounted on the other side of the mesh as shown in figure 11

Example of operation

Referring to figure I₃ product is poured continuously 9 into the rising valley 11, and is unfolded and spread to the belt sidewalls 6 to give a uniform depth of product 10 at the belt discharge.

If the feed flow rates and belt speed are appropriate a monolayer can be created using this process.

Other embodiments

Some difference in belt tension can be removed by using profiled end rollers (barrel shaped to suit the pattern of the belt tending to be tighter at the edges due to creation of the valley and ridge), and this can be used with in combination with humps to help achieve even tension.

Another way of fine tuning the tension is to have the pointed end of the arrow on an adjustable height mechanism so by raising it slightly, you preferentially tension the centre of the belt and by lowering it (and taking up the slack on a conventional tensioning roller), the edges can be preferentially tightened.

And a mirroring effect can be created by raising and lowering the tail end of the arrowhead.

Figure Labels

Claims

1. A conveyor assembly comprising at least one drive roller, a conveyor belt and a profiled support member.

2. A conveyor assembly according to claim 1 wherein the profiled support member is adapted to provide a sequence of flexations of the base of the conveyor belt.

3. A conveyor assembly according to claim 1 wherein the profile of the conveyor support varies along the length of the support.

4. A conveyor assembly according to claim 1 wherein the support comprises a first "input" end and a second output end and where the input end comprises an upwardly facing concave shape.

5. A conveyor assembly according to claim 4 wherein the concave input end comprises a substantially V-shaped valley.

6. A conveyor assembly according to claim 4 wherein the output end of the belt is substantially flat or planar.

7. A conveyor assembly according to claim 4 wherein intermediate between the input end and the output end the cross-sectional profile of the support defines an increasingly shallow concave valley.

8. A conveyor assembly according to claim 4 wherein adjacent the output end the support includes a portion comprising a convex cross section.

9. A conveyor assembly according to claim 1 wherein the joints of the ends of the conveyor belt have a shear strength comparable to that of the rest of the belt.

10. A conveyor assembly according to claim 9 wherein the joining method comprises creating short loops on the end of the belt by stitching the last belt mesh back on itself and then cross-stitching between loops.

11. A conveyor assembly according to claim 10 wherein the stitches comprise Kevlar® threads.

12. A conveyor assembly according to claim 1 wherein the conveyor belt comprises a low friction material.

13. A conveyor assembly according to claim 12 wherein the low friction material is a PTFE coated Kevlar® mesh.

14. A profiled conveyor support in which the support comprises at least one concave region for use in conjunction with a conveyor assembly according to claim 1.

15. A profiled conveyor support member according to claim 14 wherein the support member includes a convex region.

16. A profiled conveyor support member according to claim 14 wherein the support member comprises a single unit.

17. A profiled conveyor support member according to claim 14 wherein the support member comprises one or more modular units.

18. A profiled conveyor support member according to claim 14 wherein the support comprises a flat 'arrowhead' platform that defines a series of sections.

19. A method of conveying a material which comprises the use of a conveyor assembly comprising at least one drive roller, a conveyor belt and a profiled support member.

20. A method according to claim 19 which comprises conveying of particulate foodstuffs.

21. A method according to claim 20 which comprises conveying and levelling piled particulate foodstuffs.

22. A method according to claim 19 which comprises calculating curves from an ellipse.

23. A method according to claim 19 which comprises calculating curves from a Bezier curve.

24. A kit suitable for spreading out a particulate product, the kit comprising a conveyor assembly including at least one drive roller, a conveyor belt and a profiled support member.

25. A conveyor assembly, conveyor support. Method or kit substantially as hereinbefore described with reference to the accompanying drawings.