WO2003033238A1

WO2003033238A1 - Improved circumferential distribution in a circular extrusion die

Info

Publication number: WO2003033238A1
Application number: PCT/EP2002/012192
Authority: WO
Inventors: Ole-Bendt Rasmussen
Original assignee: Ole-Bendt Rasmussen
Priority date: 2001-10-12
Filing date: 2002-10-14
Publication date: 2003-04-24
Also published as: GB0214424D0; TW548175B

Abstract

Improvement of the method and apparatus in which a generally equal circumferential distribution in an annular extrusion die is achieved by dividing the flow between inlet (1a, 10) and circular exit orifice (21) into a number of partflows through channels (4, 14) which have generally helical form at least through a portion of each channel, with space (15) provided for overflow between these helical channels. The improvement consists in either a) controlling the overflow by an adjustable insert (101) which provides a variable spacing (15) for the overflow between the helical channels (4, 14), or alternatively b) providing a constraint (115, 116, 117) against the flow in a circularly extending zone after the helical part of the channels, but before the exit orifice and in case of coextrusion before the joining of the components (19, 20). This constraint is adjustable by adjustment means (44, 45) from location to location along its circumference.

Description

Improved Circumferential Distribution in a Circular Extrusion die

The invention concerns a process of forming a tubular film by extruding at least one thermoplastic polymer material A by means of a circular extrusion die which has at least one inlet for A and an exit passageway ending in a circular exit orifice. In this process A passes an arrangement of dieparts having generally planar, conical or cylindrical surfaces supplied with grooves shaped to form channels in a construction which helps to equalize the flow over the circumference of the exit orifice. The invention is an improvement on the commonly used type of equalization in which the flow between the inlet or inlets and the exit is divided into a number of partflows of generally helical form at least through a portion of each channel with space provided for overflow between said portions. (The latter are hereinafter collectively referred to as the helical die zone) . The invention is primarily but not entirely related to coextrusion of the thermoplastic polymer material A with one or more further thermoplastic polymer materials . The circumferential distribution by use of helical grooves with space provided for overflow in cylindrical surfaces - originally grooves formed in cylindrical surfaces - was first described about 30 years ago. In this system of distribution the cross-sections of each helical groove and of the space between adjacent grooves which allow overflow, are adapted so that gradually less and less material flows through each groove, and more and more passes over to the neighbour groove, while gradually the depth of the grooves reaches zero.

In addition to the effect of circumferential equalization, the helical grooves with overflow between neighbours also serve the important purpose to shear-out the "dielines" which otherwise will occur where partflows have met each other. It has been claimed that a single helical groove, extending over several revolutions around the circular die can make a perfect circumferential distribution, provided the design of the groove and intervening spaces for overflow is exactly adapted to the rheological properties of the molten polymer material under the prevailing conditions. However, this is theory, and in practice the polymer flow must first in one or another way be divided into several part flows, each of these proceeding into a helical groove with space provided for overflow between different grooves. The higher the number of part flows and thereby the number of grooves, the shorter the helical portion of each groove can be, but in any case the design of the grooves and the spaces for overflow is essentially dependent on the rheological properties of the molten polymer material.

Therefore, if the helical die zone is constructed to provide perfect circumferential distribution for one given, commonly used polymer material, and the die with the same helical die zone is applied for another commonly used polymer, it will not be exceptional if the variations in the distribution the will amount e.g. to about +/- 30% (as the inventor has experienced) .

In order to make the die applicable to rheologically different materials, GB-A-1384979 (Farrell) which relates to coextrusion, has disclosed that the helical die zone may consist of exchangeable inserts. However, in this connection it should be noted that disassembling and reassembling a die - which also necessarily includes cleaning - is a time consuming and expensive operation.

Farrell also discloses that there may be installed an adjustable annular bar after the termination of the spiral grooves and before the polymer material in question joins with another polymer material. By means of push-pull screws this annular bar can set up a higher or lower overall resistance to the flow. However as a person familiar with extrusion will understand, while increased resistance at this location in the die will tend to improve the circumferential equalization, it also results in drawbacks, in particular that the increased back-pressure on the extruder leads to a higher consumption of energy and increased temperature in the melt which leaves the extruder. The higher melt temperature has a negative influence on the behaviour of the bubble hauled off from the die. As a general rule, the resistance to the flow should be minimised as much as conveniently possible overall in the die.

The circumferential equalization before the exit passageway is much less critical in monoextrusion as it is in coextrusion, since there exist efficient methods of eliminating gauge variations in the final tubular film. Thus, controlled by feed-back from automatic inline measurements of gauge variations, either the air cooling of the bubble or the temperature of the exit orifice can be varied and set differently at different locations around the circumference of the bubble or the exit orifice.

There also exists a system in which a lip on the exit orifice of an annular extrusion die can be made flexible and adjustable from location to location along its circumference. In the case of coextruded tubular film, however, there is a need for improvement of the possibilities to reduce the gauge variations in the individual layer, and quickly to adapt the coextrusion die to changes in components. In this connection it should be remembered that some layers in a coextruded film, e.g. barrier layers or tie layers, often consist of very expensive material, and that their thicknesses nowhere must be lower than a certain critical value, thus thickness variations can have a very essential influence on the economy of the production. With existing technology the practical consistency in thickness of the individual layers in coextruded film is much lower for films from annular dies than it is for coextruded films from flat dies, and that is probably the main reason why flat dies usually are preferred for coextrusion in spite of the other advantages which annular extrusion dies offer, e.g. the simple change of film width.

The process according to the invention of forming a tubular film by extruding at least one thermoplastic polymer material A by means of a circular extrusion die having at least one inlet for A and having an exit passageway ending in a circular exit orifice, and in which process A passes an arrangement of dieparts having generally planar or conical or cylindrical surfaces, whereby said surfaces are supplied with grooves shaped to form channels tending to equalize the flow over the circumference of the exit orifice, the flow between the inlet or inlets and the exit being divided in a helical diezone into a number of partflows of generally helical form at least through a portion of each channel with space provided for overflow between said portions, is characterised by one or both of the following features: either a) said overflow is controlled by an adjustable insert which provides a variable spacing for the overflow in at least a part of the helical die zone, and/or b) there is provided a constraint against the flow in a circularly extending constraint zone after the helical die zone and before the exit orifice and, in case of coextrusion, before A joins with any other extruded material, and this constraint is adjustable from location to location along its circumference with means for different adjustments at the different locations . The coextrusion die for carrying out the process of the invention is similarly characterised. It is noted that the adjustable constraint bar mentioned in GB-A-1384949 (Farrell) , which is installed immediately after the helical die zone, is not adjustable from location to location along its circumference with means for different adjustments at the different locations, and therefore exerts a different action.

In this specification the terms helix and helical are not restricted to a curve formed on a cylindrical or conical surface which would, if the surface were developed to a flat surface, be a straight line, but include also curves on such surfaces which would not be rectilinear and also plane curves traced by a point which winds about an axis on the plane and which continuously, regularly or somewhat irregularly, recedes from the axis. The condition that the part flows and channels must be of a generally helical form in the helical die zone therefore does not limit the invention to the regular helical form. Although such a regular form usually is very suitable it is not always needed for proper equalization.

Thus as an example, if there are many part flows, e.g. 16 or more, the "generally helical" portion of each can be relatively short and can be substantially linear and directed at a small angle to the tangent to the circle which meets the inner extent of this short linear portion formed by rotation of that point around the die axis. Another example of an irregular but generally helical form which can be suited for the shaping of the channels, is a staggered form in which a first segment of a generally helical partflow follows a channel which is circular around the die axis, then just before this partflow would meet the adjacent partflow the channel bends to project the first mentioned partflow into a second "orbit" further apart from the inlet to the die, as illustrated in fig. 5b. In this second "orbit" a second segment of the channel continues circularly, later again before the two part flows would meet each other, the channel deviates to third "orbit", and so on. As it shall be explained later such a staggered form can be advantageous in connection with the present invention. In this figure the distribution pattern is shown in a planar arrangement, but a similar pattern can be formed in conical or cylindrical arrangement.

As mentioned above the flow of A is divided into several part flows before entering the helical die zones. The dividing into part flows should preferably take place by the system which in US-A-4403934 (Rasmussen, et al) is referred to as labyrinthine dividing although there may be some dividing carried out by other systems prior to the labyrinthine dividing. Labyrinthine dividing is easiest understood by a reference to figs. 7a, b and e. It means that a main flow branches out to two generally circularly arched equally long and mutually symmetrical first branch-flows, which together occupy essentially 50% of the circumference of the corresponding circle, whereafter each of the first branch-flows branch out to two, in similar way generally circularly arched second branch flows. These in total four second branch flows also occupying together essentially 50% of the circumference of the corresponding circle. The dividing may continue in a similar manner to form 8 or 16 or 32 even 64 part flows. There may be small modifications of the circular arrangement, e.g. the four second branch- flows may form four of the sides in a regular octagon, the eight third branch-flows may form eight of the sides in a regular 16-sided polygon, etc. The labyrinthine dividing has first been described in US-A-2820249 (Colombo) in connection with extrusion coating of cylindrical items. The first description of labyrinthine dividing for extrusion of blown film and in connection with a subsequent equalization by means of helical channels with overflow is found in the above mentioned US-A-4403934 (Rasmussen et al) . At least a part of the channels for the labyrinthine dividing may be formed integrally with the channels for the generally helical flow between the cylindrical planar or conical surfaces of said first dieparts by grooves in at least one surface of a pair of contacting surfaces. This is shown in fig. 7b.

The labyrinthine dividing is here extending in a planar pattern, but can also extend in a conical pattern or as in the above mentioned US patent in a cylindrical pattern.

A first embodiment of the invention, based on the solution that an insert provides an adjustable control of overflow in the helical die zone, is characterised in that the insert is nested in a diepart facing the grooves in another diepart, in such a way that the nesting allows the insert to be moved in the axial direction of the die to adjust the overflow. A series of adjustment devices is provided for the movement. This is illustrated in figs. 3a and 3b. In order to avoid leakage or stagnation of A between the insert and the diepart in which the insert is nested, this embodiment can be carried out in such a way that the insert is hinged at its two ends through a flexible generally annular sheet to the said diepart. This is illustrated in figs. 5a and 6.

The possibilities for adjustment need not comprise the entire helical die zone. On the contrary, in this embodiment of the invention it will normally be an advantage to have unchangeable dimensions at the end of this zone, designed as a compromise between the different planned uses of the die.

A second embodiment of the invention, which also operates with variable overflow in the helical die zone, effects the adjustments by rotational movement of the insert, and is characterised in that the insert is nested in a diepart facing the generally helical grooves in another diepart (hereinafter the first set of grooves) whereby the surface of the insert on the side facing the first set of grooves also is supplied with generally helical grooves (hereinafter the second set of grooves) . The geometrical pattern of the grooves in the first and second sets generally correspond to each other, and the nesting of the insert allows the latter to be rotated around the axis of the die to adjust the position of the helical ridges on the second set of grooves relative to the position of the helical ridges on the first set of grooves, so that said ridges can cover each other to a greater or lesser extent. There are adjustment devices provided for this rotation. This embodiment is illustrated in fig. 4.

In this embodiment it will normally be advantageous to have unchangeable dimensions at the end of the helical die zone, designed as a compromise between the different planned uses of die, i.e. the part of the insert which is supplied with grooves should normally not occupy the entire helical die zone. A third embodiment of the invention, still operating with variable overflow in the helical die zone, is similar to the first embodiment in that the adjustments are effected by axial displacements of the insert . It is limited to a flow pattern in which the partflows are of a generally planar spiral form where the control takes place, flowing outward or inward in relation to the die axis, and the embodiment is characterised in that the insert consists of a flexible generally flat annular sheet facing the grooves, which sheet may comprise thicker areas for stabilisation against the pressure in the flow, and which at its inward and outward boundaries is fixed to a stiff diepart. A circular row of adjustment devices is provided on the side of the sheet which is opposite to the grooves. As shown in these sketches it may be carried out in such a way that the spaces for overflow can become independently adjusted at different stages of the flow. As mentioned above, an alternative or a supplement to the adjustments of overflow in the helical die zone is the provision of a constraint zone after that zone, with means for different adjustments at different circumferential locations. As further mentioned the constraint zone is provided before the exit orifice and, in the case of coextrusion, before A joins with any other extruded material .

A preferred embodiment of the invention following this solution is characterised in that the flow in the constraint zone is generally planar or conical with the genetrix of such cone relatively close to 90° to the axis. The flow is outward or inward in relation to the die axis, and the adjustable constraint is established by means of a flexible generally flat annular sheet facing the grooves, which sheet may comprise thicker areas for stabilization against the pressure in the flow, and which at its inward and outward boundaries is fixed to a stiff diepart. As means for the different adjustments at the different locations a circular array of adjustment devices is provided on the side of the sheet which is opposite to the grooves.

This enables a particularly exact equalization over the circumference with minimum increase in back pressure. The part-equalization in the helical die zone can with advantage be preceded by part-equalization in a labyrinthine dividing system (as described above) . Such labyrinthine system should preferably comprise at least three steps of branching out . For the same efficiency of the equalization the labyrinthine system, which properly designed, causes less back pressure than the helical system, and were it not for the dielines which can be eliminated in a helical die zone but not by labyrinthine dividing, this embodiment of the invention could with advantage be carried out with the total helical zone in the die substituted by an efficient labyrinthine distribution. It is usually important to eliminate the dielines, therefore the helical die zone normally is indispensable, however its length can often be reduced to a minimum.

As a separate aspect of the invention a die has labyrinthine distribution of the extrusion flow by at least three steps of branching out to the circularly extending zone before the exit orifice or joining zone, as the case may be, wherein a constraint against the flow is provided which is as defined in the embodiment of the first aspect comprising the constraint zone.

The invention shall now be described in further detail with reference to the drawings.

Fig. 1 represents prior art relating to annular coextrusion with helical die zones for the circumferential equalization and for elimination of dielines. It shows an axial section through the coextrusion die. The coextrusion is of the sequential type, which probably presently is the most used type of coextrusion when four or more layers are joined. The drawing is derived from WO-A-9800283 (Planeta) .

Fig 2 is a bottom view of one of the "mandrels" of the known art die of fig. 1, also derive from W098-A- 00283 (Planeta) .

Fig. 3a is a magnification of the die section which in fig. 1 is framed and marked "Dl", however modified so that it represents the first embodiment of the present invention, in which an axially adjustable insert is nested in a diepart to provide control of overflow from groove to groove . Fig.3b is analogous to fig. 3a, but representing a die construction in which the helical die zone is conical .

Fig.4 is a magnification of the die section which in fig. 1 is framed and marked "Dl", however modified so that it represents the second embodiment of the present invention, in which an insert which is adjustable by rotation, is nested in a diepart to provide control of overflow from groove to groove.

Fig. 5a, representing the third embodiment of the invention, is a view similar to that of fig. 4a, but showing the die insert as a flexible generally flat sheet with a number of rigid rings, which are independently adjustable.

Fig. 5b represents the section which in fig. 5a is indicated by the line b-b, and shows a modification of the generally helical form of the grooves.

Fig.6 shows a modification of fig 3b in which, in order to avoid leakage or stagnation, the insert is hinged at its two ends through flexible annular sheets to a major diepart. Fig. 7a illustrates a preferred coextrusion die construction, of general principles different from the prior art die of fig. 1, and for each extruded component supplied with a variable constraint immediately after the helical die zone. This constraint is made variable with means for different adjustments at different circumferential locations.

Fig.7b represents the section which in fig. 7a is indicated by c-c and shows the pattern of grooves starting as a labyrinthine system and continuing in a helical system.

Fig. 7c is a magnification of the die section which in fig. 7a is framed and marked "D2", and illustrates in detail the variable constraint.

Fig. 7d is an unfolding of the circular section formed by rotation of the line d-d in fig. 7c. fig. 7a is an unfolding of the circular section formed by rotating each of the lines e-e in fig. 7a around the die axis. It shows the first two steps of the labyrinthine dividing. Fig.7f is a magnification of the die section which in fig. 7a is framed and marked D3 , and illustrates in detail the joining of the three coextruded components. The prior art die shown in figs. 1 and 2 has axis (1) and consists of clamped together discs and "shell" - or "bowl" formed parts. Thus (2a) and (2b) together discs form a shell or bowl, and (3a) to (3i) are discs fitting, into this bowl. Five components are fed into the die for coextrusion, of which the inlets for two (la) are shown. Apart from the inlet channels all channels for the five components and the common flow of two or more of these components are formed by spaces between the disc- or bowl- formed parts, thus the equalization of each component over the circumference is in a first step established by a generally annular channel (lb) which is gradually tapering towards the downstream (see fig. 2) , and in a second step by helical grooves (4a) to (4e) which extend generally along a plane perpendicular to the axis (1) and in fig. 1 are seen almost in cross-section. These grooves are formed in the surface of one of a pair of adjacent discs or between the bowl and the adjacent disc. Alternatively there might be grooves in both surfaces facing each other.

As the drawing shows there is arranged an overflow between grooves which are adjacent when seen in cross- section. Each groove starts relatively deep but gradually becomes shallower to end at zero depth. The proportions between the different dimensions in such a spiral distribution system is oritical for the equalization of the flow over the circumferences and depends critically on the rheological parameters of the extruded melt under the given conditions of temperature and throughput.

This construction of an extrusion die has the advantage that it allows coextrusion of many components, but has the drawback that these components must have relatively similar rheologies, otherwise the thickness of the individual layers may become uneven.

In the known art coextrusion die which is shown in figs. 1 and 2, the flows follow a planar pattern during the passage through the helical die zone, and they proceed from the inner of the die outwardly towards the circumference until they meet the common annular channel after which they proceeds generally axially. However it is also well known, as e.g. disclosed in US-A-4403934, to let the flows follow a conical pattern during the passage through the helical die zone.

Finally it is known that in similar sequential extrusion, no matter whether the flows follow a planar or a conical pattern during the passage through the helical die zone, the inlets can be at the outer circumference while the flows proceed inwardly during the passage through the helical die zone.

The present invention can with advantage be applied in each of these cases.

In fig.3a, which as mentioned above is a detail from fig.l but modified according to the invention, the annular insert (101) , here seen in an axial section, is nested in the diepart (2a) opposite the grooves in the diepart (3a) . The nesting allows axial movement of this insert. The polymer flow urges the insert away from the grooves, but the insert is held in position by the spirally curved tap (102) . The tap is a rotatable element having an outer spiral cam surface, which bears against the insert and which can be rotated through its shaft (103) . Due to its spiral curving, the axial position of the insert (101) , which controls the amount of overflow, can hereby be adjusted.

There is a plurality of such curved taps (102) around the circumference, the number depending on the stiffness of the insert, and their adjustment through shafts (103) is preferably coordinated by transmission means. These means and the means for fixing the position is not shown. In the fitting of the insert (101) into part (2a) there should be a clearance, e.g. 0.1mm, which is sufficient to allow a very small drain of polymer from the main flow between (2a) and (3a) into the space (104) under the insert, and therefrom out through the drain channel (105) . Hereby there is provided lubrication for the adjustments of the mandrel, and furthermore a degradation of polymer is avoided, where the flow crosses the borders between diepart (2b) and insert (101) . Fig. 3b shows how the first embodiment of the invention is carried out when the construction of the diepart is conical. Except for the conical shapes the constructions are similar to those in fig. 3a.

Fig. 4 is generally similar to fig. 3a except that the adjustments of the overflow takes place by rotational instead of axial displacements of the insert (101) . Thus the insert is nested rotatably in diepart (2a) . The insert is supplied with grooves (106) and the pattern of the latter exactly match with that of the grooves in diepart (3a) . The drawing shows the insert in a position in which there is maximum overflow. The spiral pattern is a pattern of regular spirals (not e.g. a spiral-like arrangement as that shown in fig. 5b) so the spacing which provides for the overflow can be continuously varied by turning the insert (101) . This is here done by means of a gear-wheel (107) .

The gear-wheel fits into a gearing at the bottom of the insert and is turnable through the shaft (103) . In the principle one gear-wheel is enough, but in practice there should normally be at least two, e.g. arranged diametrically opposite each other. Means for coordination of these gear-wheels are not shown. At (107a) the surface of the insert (101) and the surface into which it fits are conical to provide sealing, but for reasons of safety there is still arranged a drain (105) . Such rotational adjustment can of course be modified for a conical and even for cylindrical construction of the dieparts (as that in fig. 2b) . In fig. 5a the insert which controls overflow in a part of the helical die zone, is adapted to allow independently adjustable amounts of overflow at different distances from the die axis. For this purpose the beginning of the grooves - see fig. 5b - have the staggered form already mentioned, in which a first segment of a generally helical partflow in "orbit 1" marked (I) follows a channel Which is circular around the die axis, then just before this partflow would meet the adjacent partflow the channel bends to project the first mentioned partflow into "orbit 2" marked (II) . In this "orbit" a second segment of the channel continuous circularly, later again before the two part flows would meet each other, the channel deviates to "orbit 3" marked (III) , and so on. In fig. 5b the "ridges" on the grooves are shown as (108) .

There is arranged independent adjustment of the flow over the circular parts of "ridges" in different "orbits". This is achieved by giving the insert the form of a membrane with flexible annular zones (109) and much thicker, stiff zones (110) . Each of the annular stiff zones corresponds to an orbit of the ridges (108) . At the inner and outer boundary the membrane is also supplied with a thicker zone (111) and (112) , but in this case for fitting it into the diepart (2b) . The outermost and innermost thickenings (112 and 111) are screwed together with (2a) by bolts (113) .

The axial position of each annular thickening (110) is adjusted by means of spirally curved taps, of which of one (102) for the innermost annular thickening is shown. It is turnable through the shaft (103). Similar spirally curved taps for the other annular thickenings are not shown but would appear in other axial sections of the die. For each of the annular thickenings (110) there are several such taps (102) and they are mutually coordinated by mechanical means (not shown) . The required number of such taps depends on the stiffness of the thickenings and the expected pressure in the flow.

A practical technical equivalent to such coordinated, spirally curved taps, is a ring which is turnable around the die axis like the ring (101) in fig. 4, and which is supplied with a number of wedges which press against corresponding wedges at the bottom side of a ring (101) for adjustment of this ring's axial position. In this construction there should be one such adjustment device for each ring (110) .

The drain (105) is not strictly needed in this construction provided parts (111) and (112) fit tightly with diepart (2a) , but is preferable as a safety precaution. The die construction shown in fig. 6, which is closely related to the construction shown in fig. 3b, also aims at a tight fitting between the inset (101) and diepart (2a) by forming the inner and outer parts as hinges, with a thin, flat, annular zones (114) which can be flexed, and at the boundaries of the insert an annular thickened zone, (111) and (112) respectively, which allows the insert to be firmly joined with diepart (2a) . As in fig. 5a the outer boundary (112) of the insert is screwed to diepart (2a) by means of bolts (113) . The adjustment by means of spirally curved taps (102) with shafts (103) is similar to the adjustment means in fig. 3b.

In order to allow axial movement of the insert (101) by such hinging, the flow must be generally planar at the inlet to and at the exit from this part of the helical die zone, as shown.

It is obvious that the construction shown in fig. 6 can be modified for use of generally flat annular shapes instead of conical shapes of the insert and the immediately adjacent regions of parts (2a) and (3a) . However, it can also be modified for generally cylindrical shape of the insert and the immediately adjacent region of part (2a) and (3a) . In this case the insert must also be supplied with helical grooves on the side which faces the flow, and these grooves must correspond to the grooves in the fixed diepart - like in fig- 4, but now in a cylindrical instead of a planar arrangement. Then the insert is adjusted to an axial position in which the ridges on one set of grooves directly faces the ridges on the other set of grooves, the overflow will be at minimum, and if the axial position of the insert is continuously moved away from this position, the overflow will continuously increase.

In the embodiment of the invention which is shown in figs. 7a to e, the means to make exact adjustments of the evenness of each flow over the circumference, consists of a constraint device after the helical zone and before the location of flow-joining. The major part of the coextrusion die, which has axis (1) , is made from two shell or bowl formed parts (5) and (6) and two disc- formed parts (7a) and (7b) , which parts are clamped together. These parts form the ends of a labyrinthine dividing system for each of three polymer materials A, B and C, and form helical grooves with overflow, as well as a common exit passageway and an exit orifice, which here is shown at the periphery. The die also comprises three smaller disc-formed parts (32) , (33) and (34) , which together form the beginning of the labyrinthine dividing channel system.

From the inlets (10) , here holes in part (32) , each of the molten polymer materials A, B and C divide out into the two channel branches (35a) and (35b) , see fig. 7e. The branches are formed by grooves in both parts (32) and (33) , but it could be a groove in one part only. From each end of these branches each component passes through a hole in the disc (33) , and at the other surface of (33) each of the two part-flows divide out into two part-flows (36a) and (36b) , in total four branches, so that each component A, B and C now has become four part- flows. At the end of each of the four branches each component passed through a hole (37) in disc (34) which leads into the dieparts (5) , (7a) or (7b) .

Each hole (37) continues as a bore (38) through shell-part (5) , see fig. 7a. For component B the bores (38) directly form the four inlets to the system of grooves between (5) and (7a) . For component A and C the bores (38) are continued as bores (39) through disc (7a) . For component A the bores (39) directly form the four inlets to the system of grooves between discs (7a) and (7b) . For component C the bores (39) are continued as bores (40) through disc (7b) , and these bores directly form the four inlets to the system of grooves between disc (7b) and bowl (6) . Since the three sections marked c-c are considered identical except for the inlets, fig. 7b does in fact show the continued system of flow of each component B, A and C. The dieparts (5) , (7a) , (7b) and (6) are clamped together by bolts (not shown) .

As shown in fig. 7b each of the four partflows divide out into two, so that each component forms a total of eight partflows (13), and each of these eight partflows proceeds through one of the helical grooves (14) with overflow in the spaces (15) .

Alternatively, not only the four but all eight partflows of each component may be formed by labyrinthine dividing upstream of the dieparts (5) , (7a) and (€) , or it may be advantageous, especially for dies of a large exit orifice diameter, to divide to more than eight partflows, e.g. to 16 or 32 partflows. There is not shown any devices for control of the amount of overflow.

The serrated, but generally circular, broken curve (16) in fig. 7b shows where the overflow begins. The downstream limit of the helical die zone is in fig. 7b indicated by the broken circle (16b) . This limit is also shown in fig. 7c. The broken lines (17) in fig. 7a indicate that the channels which are seen almost in cross-sections are connected outside the section which is represented in these drawings .

Figs . 7c and d show the constraint device between the helical die zone and the zone where the polymers join. This device has a flexible annular part (115) , and upstream and downstream of this it is stiff in zones (116) and (117) . The flexible part (115) can be considered an annular membrane, however it has a semi- flexible annular zone (115a) in its middle, as shall be explained below. The stiff part (117) on the downstream side is fixed to the adjacent die disc (7c) by a circular row of bolts of which one (43) is shown.

The pressure in component A pushes the membrane part of the constraint device towards a circular row of spirally curved taps (44) each on a turnable shaft (45) which is nested in a bore in the die disc (7b) . There are many such shafts with taps, and they extend in a starlike manner through the disc (7b) very close to each other. The distance from axis to axis of these taps can in practice be down to about 15-20mm. By turning the shafts the positions of the membrane and thereby the overflow between the helical grooves can be continuously adjusted. The stiffness of the semi-flexible annular zone (115a) must be adapted to the distance between the taps to give an appropriate variation of the constraint over the circumference. The means for turning the many shafts (45) and fixing their positions, e.g. by the use of spindles and spindle wheels, are not shown. If the taps are very close to each other, the shafts should point towards different axial levels in the hollow space in the middle of the die, where these devices are arranged.

An alternative to this arrangement of taps is disclosed at the end of this description. Having passed the helical die zone of channels and the adjustable constraint device, A, B and C proceed towards the common circular exit channel (18) whereby B and A join at the internal orifice (20) while A and C join at the inlet orifice (19) (see fig. 7a) . The two internal orifices are here shown directly opposite each other, which is an advantage but not a necessity. 5 The joining here, as shown in detail in figure 7f, takes place with a special, advantageous but inessential feature, namely so that the annular edges (118) on dieparts (7a) and (7b) , over which the joining takes place, are almost freely moveable in the axial direction,

10 since they are "hinged" on the main body of the respective diepart (7a or b) through the annular, membrane-like thin portions (119) . In principle the latter should be almost as thin as possible, and in order to avoid them becoming damaged during cleaning, a

15 practical compromise is about 0,3 mm thickness. The material is hardened steel, and the length of the "hinges" can conveniently be about 10 mm.

It is generally considered that the velocity of two polymer streams joining each other should be about equal

2.0 where they meet. This is with some approximation automatically achieved with this "free-floating lip" construction. It is noted that such a floating, annular lip, is known from the inventor's copending patent application, WO-A-0178966. In WO-A-0178966 the lip is

25 disclosed as a solution to problems in a very special kind of coextrusion, but it is also generally applicable to coextrusion with annular dies.

After the internal orifices (19) and (20) , the common flow follows the exit channel (18) which ends in

30 the exit orifice (21) .

Contrary to the normal construction of extrusion dies, the flow of joined B, A and C proceeds radially out of the exit orifice (21) which is located at the periphery of the die .

35 Having left the exit orifice, as shown in figure 7a, the molten tubular B/A/C film is turned over the cooled ring (22) and is hauled off, blown and aircooled by conventional means (not shown) . The ring (22) is directly fixed to the shell-part (6) of the die through a heat insulating material (23) . The ring (22) is hollow, and the cooling takes place by circulation of water or oil, which may be temperature-controlled. This cooling medium is pumped into and out of the ring (22) through pipes, of which one (24) for the inlet is shown. These pipes are preferably passed through the die cavity in the region around the axis of the die . One of the circular lips (25) of the orifice (21) is preferably made flexible as indicated and is made adjustable by means of row of screws close to each other of which one (26) is shown. Such adjustment is well known from the construction of ordinary flat dies, and in fact the die of fig. 7a can be considered a flat die, although the exit orifice (21) is not straight but circular. Screw (26) is shown pressing on a lip of the die (25) . Although the screw could also be a push-pull screw, the pressure in the melt may give a sufficient opening force for it only to push the lip.

This peripherical extrusion, by which the exit orifice is made adjustable like in flat extrusion dies, adds very efficient adjustment means for the gauge of the final multilayer film to the means for adjusting the individual layers. However, it is clear that the exit channel (18) can alternatively be bent towards the axial direction so that the exit (21) can be located at the end of the die as it is normal in extrusion of tubular film. As it appears from figs. 7a and b there is preferably provided a relatively large continuous hollow space extending from the die axis (1) to the innermost cylindrical surfaces of the clamped-together dieparts (which surfaces may e.g'. be conical instead of cylindrical) . This space can be very useful e.g. to establish an efficient internal cooling of the extruded tubular film, and for the adjustment devices at the end of the shafts (45) . In order not to make the drawings 7a-e too difficult to read, they are simplified on several points. Thus heating elements are not shown, and some of the drawings do not show any drainage system, which is normally indispensable when channels for the extrusion are formed between clamped-together dieparts.

Without a suitable drainage, unavoidable leakages may build up to high pressures between the dieparts. Since such drainage is well known in the art it is not further described here.

In figs. 7a to e the extrusion of the three components is shown as taking place outwardly. This is an advantage in particular when a large diameter of the exit orifice (21) is wanted. However, in a similar arrangement the extrusion can take place inwardly to achieve an exit orifice of small diameter. Fig 7a can also illustrate this, if the die axis (1) is moved from the right to the left side of the drawing, and if the film when leaving the cooling ring (22) is shown moving towards the right instead of towards the left. Fig. 7b, however, should in this case be changed to show the helical die zone inward of the labyrinthine system.

A graph showing the throughput of each component of coextrusion in dependence of the circumferential position, e.g. indicated in degrees, will normally show a number of relative maxima and relative minima, which for one tour around the circumference each will equal the number of spirals. If the die is evenly constructed and evenly heated/cooled, and if the components are fed into the die with a homogenous temperature, the curve segments between two adjacent maxima will all be very similar. In that case the construction of the constraint device after the helical grooves can be simplified. Thus, if the distance between two neighbour grooves is about 5 cm or less, measured along the circle at the end of the grooves, it may be sufficient to have adjustments screws or the like (e.g. push-pull screws) at the positions where the maxima and the minima are located. All screws for the maxima may be supplied with gearing and be synchronised by means of a gear-ring, which has its axis coinciding with the die axis (1) and all screws for the minima may be similarly synchronised under use of another, similar gear-ring. The adjustment of these two gear-rings is similar to what is explained in connection with fig. 4.

Claims

1. A process of forming a tubular film by extruding at least one thermoplastic polymer material A by means of a circular extrusion die having at least one inlet for A and having an exit passageway ending in a circular exit orifice, and in which process A passes an arrangement of dieparts having generally planar or conical or cylindrical surfaces, whereby said surfaces are supplied with grooves shaped to form channels tending to equalize the flow over the circumference of the exit orifice, the flow between the inlet or inlets and the exit being divided in a helical die zone into a number of partflows of generally helical form at least through a portion of each channel with space provided for overflow between said portions characterised in that in order to improve said equalization either: a) said overflow is controlled by an adjustable insert which provides a variable spacing for the overflow in at least a part of the helical die zone, or b) there is provided a constraint against the flow in a circularly extending constraint zone after the helical die zone and before the exit orifice and in case of coextrusion before A joins with any other extruded material, said constraint being adjustable from location to location along its circumference with means for different adjustments at the different locations .

2. A process according to claim 1 in which the flow control is established within the said helical die zone, characterised in that said insert is nested in a diepart facing the grooves in another diepart, said nesting allowing the insert to be moved in the axial direction of the die to adjust the overflow, and a series of adjustment devices being provided for such movement.

3. A process according to claim 2, characterised in that in order to avoid leakage or stagnation of A between the insert and the diepart in which the insert is nested, the insert is hinged at its two ends through a flexible generally annular sheet to the said diepart.

4. A process according .to claim 1 in which the flow control is established within the said helical die zone, characterised in that said insert is nested in a diepart facing the generally helical grooves in another diepart, which are the first set of grooves, and the surface of the insert on the side facing the first set of grooves is also supplied with generally helical grooves, which are the second set of grooves, the geometrical pattern of the grooves in the first and second set generally corresponding to each other, and the nesting of the insert allows the insert to be rotated around the axis of the die to adjust the position of the helical ridges on the second set of grooves relative to the position of the helical ridges on the first set of grooves, so that said ridges can cover each other to a greater or lesser extent and in that adjustment devices are provided for such rotation.

5. A process according to claim 1 in which the flow control is established within said helical die zone and the partflows are of a generally planar spiral form where the control takes place and flow outwards or inward in relation to the die axis, characterised in that the insert consists of a flexible generally flat annular sheet facing the grooves which sheet may comprise thicker areas for stabilization against the pressure in the flow and which at its inward and outward boundaries is fixed to a stiff diepart, and in that a circular row of adjustment devices is provided on the side of the sheet which is opposite to the grooves.

6. A process according to claim 1, in which the flow control takes place in the said constraint zone after the helical die zone, characterised in that the flow in the constraint zone is generally planar or conical with the genetrix of a cone being relatively close to 90° to the axis, the flow being outward or inward in relation to the die axis, and in that the adjustable constraint, which is variable from location to location along its circumference, is established by means of a flexible generally flat annular sheet facing the grooves, which sheet may comprise thicker areas for stabilization against the pressure in the flow and which at its inward and outward boundaries is fixed to a stiff diepart, and a circular row of adjustment devices, which are provided on the side of the sheet which is opposite to the grooves .

7. A process according to claim 6, characterised in that prior to the passage through the helical die zone the equalization over the circumference is started by labyrinthine dividing with at least three steps of branching out .

8. A circular extrusion die having at least one inlet for A (la, 10) and having an exit passageway (18) ending in a circular exit orifice (21) , comprising an arrangement of dieparts (5, 6, 7a, 7b) having generally planar or conical or cylindrical surfaces, which surfaces are supplied with grooves (4, 14) shaped to form channels tending to equalize the flow of A over the circumference of the exit orifice, the flow between the inlet or inlets (la, 10) and the exit (21) being divided in a helical die zone into a number of partflows of generally helical form at least through a portion of each channel with space (15) provided for overflow between said portions characterised by comprising either: a) an adjustable insert. (101) which provides a variable spacing -(15) for said overflow in at least part of the helical die zone; or b) a constraint zone comprising a circularly extending constraint (115, 116, 117) after the helical die zone, upstream of the exit orifice (21) , or in the case of a coextrusion die, upstream of the joining (19, 20) of the flow of A with the flow of the coextruded material, said constraint being adjustable (44, 45) from location to location along its circumference with means for different adjustments at the different locations.

9. Die according to claim 8 comprising said insert, characterised in that the insert (101) is nested in a diepart (2a) facing the grooves in another diepart, said nesting allowing the insert to be moved in the axial direction of the die to adjust the overflow, and a series of adjustment devices (102) being provided for such movement .

10. Die according to claim 9 in which the insert (110) is hinged at each of its two ends through a flexible generally annular sheet (109) to the said diepart .

11. Die according to claim 8 comprising said insert, characterised in that the insert is nested in a diepart (2a) facing the generally helical grooves in another diepart (3a) , which are the first set of grooves, and the surface of the insert on the side facing the first set of grooves is also supplied with generally helical grooves (106) , which are the second set of grooves, the geometrical pattern of the grooves in the first and second set generally corresponding to each other, and the nesting of the insert allows the insert (101) to be rotated around the axis (1) of the die to adjust the position of the helical ridges on the second set of grooves relative to the position of the helical ridges on the first set of grooves, so that said ridges can cover each other to a greater or lesser extent and in that an adjustment device (107) are provided for such rotation.

12. Die according to claim 8 comprising said insert, in which the grooves forming the part flows of A are of a generally planar spiral form where the control takes place and are shaped so that A flows outwards or inward in relation to the die axis, characterised in that the insert (101) consists of a flexible generally flat annular sheet (109) facing the grooves which sheet may comprise thicker areas (110) for stabilization against the pressure in the flow and which at its inward (111) and outward (112) boundaries is fixed to a stiff diepart (2a) , and in that a circular row of adjustment devices (102) is provided on the side of the sheet which is opposite to the grooves (108) .

13. Die according to claim 8 comprising said constraint characterised in that the grooves form channels for a generally planar or conical flow of A in the constraint zone, the genetrix of a cone being relatively close to 90° to the axis (1) , the flow being outward or inward in relation to the die axis (1) , and in that the adjustable constraint (115, 116, 117) , which is variable from location to location along its circumference, is established by means of a flexible generally flat annular sheet (115) facing the grooves, which sheet may comprise thicker areas (115a) for stabilization against the pressure in the flow and which at its inward (117) and outward (116) boundaries is fixed to a stiff diepart (7b) , and a circular row of adjustment devices (44), which are provided on the side of the sheet (115) which is opposite to the grooves (14) .

14. Die according to claim 13 characterised in that, upstream of the helical die zone, equalization of the flow of A over the circumference of the die is started by labyrinthine dividing of the flow through channels formed by grooves (35a, 35b, 36a, 36b) in the dieparts (32, 33, 34), the labyrinthine dividing comprising at least three steps of branching.