"Improvements in or relating to the manufacture of extrusion dies"

The invention relates to extrusion dies used for producing elongate profiles in

metal (such as aluminium) plastics etc. In an extrusion process it is necessary for all

parts of the material being extruded to pass through the die at substantially the same

velocity, since if this is not the case the extruded profile is likely to be deformed.

As is well known, in an extrusion die the velocity of the extrusion material

through the die, at any particular region ofthe die cavity, depends on the width ofthe

die cavity in that region, its position relative to the centre ofthe die, and the bearing

length ofthe die cavity (i.e. its length in the extrusion direction) in that region.

Since the width and position of each region of the die cavity are essentially

determined for any particular profile to be extruded, it is normally necessary to control

the velocity by adjusting the bearing length ofthe die cavity in different regions thereof

so that the velocity of extrusion material is as uniform as possible through the whole area

of the die cavity. Thus, a narrow part ofthe die cavity will require a shorter bearing

length than a wider part ofthe cavity in order to achieve the same velocity.

This required variation in bearing length (known as the bearing contour) is

normally achieved by forming in the back face ofthe die, i.e. the face furthest from the

billet of material to be extruded through the die, an exit cavity which corresponds to the

general shape ofthe die cavity plus an all-round clearance. The depth ofthe exit cavity

is then varied so as to adjust the effective bearing length ofthe die cavity itself.

Various methods of this kind for manufacturing an extrusion die are described,

for example, in British Patent Specifications Nos. 2143445 and 2184371.

There are numerous well known methods and techniques for providing the

required correlation between bearing length and die cavity shape and position in order

to achieve uniform flow. For example, the required bearing lengths may be achieved by

trial-and-error methods based on the knowledge of an experienced die designer or,

increasingly, computer programs are available to calculate required bearing lengths from

the shape and position ofthe die cavity.

However, the extrusion dies resulting from such prior art methods may suffer

from certain disadvantages. For example, the surface of the extruded profile may be

longitudinally marked by a part ofthe die cavity where there are two adjoining regions

of significantly different bearing lengths, as may frequently occur. Furthermore, since

the die cavity itself has to be worked on and adjusted to control the flow of extrusion

material, it may not be possible to form the die from a material which cannot be readily

worked, or to provide it with a surface finish, such as nitriding, which might otherwise

be desirable to give a better finish to the profile. It would therefore be desirable to

achieve substantially uniform flow through a die cavity which has a substantially uniform,

fixed bearing length so as to avoid marking of the profile due to changes in bearing

lengths and to allow the die to be formed from a material, and have a surface finish, to

give the best possible strength and wear resistance as well as to provide the finest

possible finish on the extruded profile.

One method of achieving such an effect is described in European Patent

Specification No. 0569315. In the method described in that specification, there is

provided on the front, or entry, side ofthe die cavity an enlarged entry cavity the sides

of which converge as they extend towards the cavity in the extrusion direction so as to

provide an "entry angle". This "entry angle" is calculated in reciprocal ratio with the

width of each region ofthe die cavity. Selection of different entry angles to different

regions ofthe die cavity thus controls the velocity of extrusion material towards the die

cavity in such manner that, at the entry to the die cavity, the velocity ofthe extrusion

material at each region is such as to result in a substantially uniform velocity through the

whole area ofthe die cavity. Accordingly, the die cavity itself may be of substantially

constant bearing length. In a preferred embodiment the entry angle is provided by

forming the entry cavity with a series of steps extending inwardly towards the die cavity.

The steps are of constant depth and the entry angle is adjusted by varying the width of

the steps.

While such arrangement has met with some success, it may suffer from certain

disadvantages. For example, where the die cavity is formed with sections which are

closely spaced from one another there may be insufficient room on the entry side of each

section to provide separate and individual entry angles for each region, since the adjacent

stepped entry cavities would overlap. Consequently, in practice such closely adjacent

sections ofthe die cavity have to communicate with a single stepped entry cavity. This

means that there is no individual control over flow through these adjacent regions ofthe

die cavity and this may result in non-uniform flow through the regions if they are of

different widths. Furthermore, the adjustment ofthe flow rate by adjustment ofthe entry

angle does not make use of the long established and well known techniques for

controlling velocity by adjusting bearing length, with the result that die designers must

learn entirely new, and unfamiliar, techniques and parameters in order to put the system

into operation.

Also, although the "entry angle" may be calculated for each region ofthe die

cavity, it is in practice also necessary to make minor adjustments in order to correct

variations in velocity which may show up in initial testing of the die. Such minor

adjustments may be effected by adjusting the bearing length of the die cavity in a

particular region, but this loses the advantage of having a die cavity of substantially

constant bearing length. However, it may be difficult to make accurate minor

adjustments to the entry angle which is the only other means for varying the velocity

through a region of the die. This is presumably why the stepped arrangement is

preferred since it may be easier to adjust the width of a series of steps than it is to

accurately adjust the angle of a continuous inclined surface. However, the provision of

the steps may provide considerable resistance to the flow of material into the die cavity

with the result that the overall velocity of the extrusion material through the die is

reduced. This is undesirable since the productivity of an'extrusion installation depends

on the speed with which extrusions are produced. Also, the stepped arrangement may

cause the generation of excessive heat.

It is also known to provide a lead-in plate on the front side ofthe die, provided

with apertures which communicate with the die cavities. However, such lead-in plates

are generally of constant thickness and the velocity of extrusion material passing through

the apertures in the lead-in plate may only be adjusted by adjusting the width of such

apertures. This is not sufficiently precise to provide accurate velocity control, and

conventional correction ofthe die cavity itself is also required. For continuous extrusion

it is also common practice to provide a weld plate on the front side ofthe die. In this

case the trailing end of each metal billet is sheared off at the front surface ofthe weld

plate and is engaged by the leading surface of a new billet which becomes welded to the

end ofthe previous billet as the junction between the two billets passes through the weld

plate. However, again, the weld plate is not used to control the flow of metal precisely,

and correction ofthe die cavity itself is still required.

The present invention sets out to provide improved forms of extrusion die, and

methods of manufacture of such dies, which may overcome many or all ofthe above-

mentioned disadvantages of the prior art systems and in a preferred embodiment,

provides a fully controUed system where no correction ofthe die cavity itself is required.

According to the invention there is provided an extrusion die comprising a die

cavity having a shape corresponding to the cross-sectional shape of the required

extrusion, and a preform chamber in communication with the die cavity, the preform

chamber being of generally similar shape to the die cavity but of greater cross-sectional

area, so that regions ofthe preform chamber communicate with corresponding regions

respectively of the die cavity, each region of the preform chamber having a bearing

length which is related to the dimensions and position of said region so that, in use,

extrusion material passing through each region ofthe preform chamber is constrained

to move at a velocity such that the material passes through all regions ofthe die cavity

at a substantially uniform velocity.

Since the velocity of the extrusion material is fiilly controlled in the preform

chamber, i.e. before it reaches the die cavity, the die cavity itself may be of constant

bearing length in all regions thereof, with the advantages referred to above. The velocity

of metal through the preform chamber is adjusted by adjusting the width and bearing

length ofthe preform chamber This enables the wealth of experience and/or computer

programs already used in the designing of conventional die cavities to be employed,

resulting in accurate control of the velocity. Furthermore, since no "entry angle" is

required, the side walls ofthe preform chamber may be parallel or substantially parallel,

so that the maximum width ofthe preform chamber may be significantly less than the

maximum width ofthe entry cavity in the prior art "entry angle" arrangement referred

to above, with the result that there is room to provide a separate region ofthe preform

chamber for each region of the die cavity. If two regions of the die cavity are

particularly closely spaced, the enlarged preform chamber communicating with each

region may be made correspondingly narrow, the velocity being controlled by reducing

the bearing length ofthe preform chamber. Alternatively, if the shape ofthe die cavity

permits this, the regions of the preform chamber may be offset relative to their

corresponding regions ofthe die cavity so that they do not interfere with one another,

while remaining in communication with their corresponding regions ofthe die cavity.

To provide precise control ofthe flow through the preform chamber, the side

walls ofthe chamber are preferably exactly parallel.

By appropriate selection of the width of the different regions of the preform

chamber, the number of regions ofthe preform chamber requiring a different bearing

length may be reduced. This allows the number of variable parameters for controlling

the flow of metal through the die aperture to be reduced thus simplifying correction of

the die and rendering such correction more repeatable and reliable.

As mentioned above, variations in velocity can cause the extruded profile to be

deformed and varying the bearing length within the die cavity itself can lead to surface

marking ofthe profile. The present invention may therefore achieve the production of

high quality profiles. Equally importantly however, the invention enables the

manufacturing process itself to be controlled and improved. For example, an extrusion

die will normally incorporate a number of similar die cavities spaced apart over the face

of the die, so as to produce several extruded profiles simultaneously. As they are

extruded, the profiles are drawn by a single puller device. Accordingly, it is necessary

for the profiles from all of the die cavities to be extruded at the same speed since

otherwise the puller device may stretch and thus deform any ofthe profiles which are

being extruded at a slightly slower speed than the rest. Since the present invention

allows the speeds of extrusion to be controlled very accurately it becomes possible to

unify the speeds of extrusion from the various die cavities in the die. The invention also

allows the overall velocity of extrusion to be increased, as will be described, thus

allowing the productivity ofthe die to be increased in a reliable and controlled manner.

Since the velocity through each region of the die cavity is controlled in the

preform chamber before the die cavity is reached, the die cavity will produce an extruded

profile which is of exactly the same shape as the die cavity and it is not necessary, as has

hitherto been the case, to build deformations into the die cavity in order to correct the

profile ofthe extrusion emerging from it. For example, with conventional methods it is

frequently necessary, for some shapes of profile, to incline the walls of the bearing

portion ofthe die cavity in one direction or another in order to compensate for some

deficiency in the shape of the profile which becomes apparent in testing. Also, for

example, where two portions of a profile are required to be at a specified angle to one

another, it may be necessary for the corresponding portions ofthe die cavity to be at a

slightly different angle in order to achieve the required angle in the extruded profile.

Some of these adjustments in the shape ofthe die aperture may be very slight and may

be lost or diminished if the die is not carefully and properly maintained over a prolonged

period of use. Thus, cleaning and polishing ofthe die aperture can, over time, remove

slight correctional variations in the shape ofthe die aperture so that although the die

produces the correct profile when new, it changes with use to begin to produce a slightly

deformed profile. This problem does not arise with the present invention where the

control ofthe metal flow is effected before the metal reaches the die aperture. This sort

of deliberate deformation ofthe die cavity can be avoided with the present invention

where the extrusion material is fully controlled in the preform chamber before it reaches

the die cavity and may be so controlled that the extruded profile produced by the die

cavity is exactly in accordance with the shape ofthe die cavity itself.

The alterations and corrections which a conventional die corrector may make to

a die, in order to achieve the desired profile, may be slight and subtle, being based on the

die corrector's long experience and often being intuitive. Such corrections may therefore

be difficult or impossible to record and to repeat reliably over a succession of similar

dies. By contrast, in the present invention the desired profile is achieved by adjusting

a few clearly-defined parameters of the preform chamber. These parameters may be

measured and recorded, for example in a computer program, and repeated continually,

by precise machine methods, in a succession of dies to give entirely consistent results.

Conventional die correction may require much hand work, which is inherently difficult

to repeat precisely. The present invention may allow all shaping ofthe preform chamber

and die cavity to be carried out by machine, so as to be inherently repeatable.

As mentioned above, the die cavity may be of substantially constant bearing

length in all regions thereof. In particular, the invention allows all regions ofthe die

cavity to be of substantially zero bearing length.

It is known to provide extrusion dies of zero bearing length, and for example

such dies are described in European Patent Specification No. 0186340. However, as

acknowledged in that specification, the design of a conventional zero bearing length die

is such that modification ofthe profile ofthe aperture to hasten or slow the passage of

metal is not possible. Accordingly, zero bearing length dies have hitherto been regarded

as mainly suitable for extruding the minority of sections whose configuration does not

require adjustment or correction. If a conventional zero bearing length die does not

produce an extrusion of the required profile, there is no way in which the die can be

corrected. However, since the present invention allows control ofthe velocity ofthe

metal upstream ofthe die, it allows the use of zero bearing length dies for virtually all

types of section. Thus, the present invention allows the advantages of zero bearing

length dies to be combined with reliable correction and control.

A die cavity of substantially zero bearing length may be formed by providing in

the die plate a die aperture which is negatively tapered throughout its length, i.e. the

walls of the die aperture diverge as they extend from the front surface to the back

surface ofthe die plate. As mentioned in EP 0186340 a negative taper angle of at least

0.8° is preferred so that any friction stress between the walls ofthe die and metal flowing

through it is negligible. It is believed that a negative taper angle of about 1.5° is more

reliable.

It will be appreciated that it is in practice impossible to provide a die cavity

which is literally of zero bearing length, since there will normally be a small radius at the

junction between the negatively tapered die cavity and the front surface ofthe die plate.

EP 0186340 relates to arrangements where this radius of curvature is not greater than

0.2mm. However, for the purposes of this specification the die cavity is regarded as

having zero bearing length where the die cavity increases in width as it extends away

from the front face ofthe die plate, regardless ofthe radius of curvature at the upstream

end ofthe die cavity.

In any ofthe arrangements according to the invention the region of the preform

chamber which is of minimum bearing length may also be of substantially zero bearing

length, increasing to a maximum the overall velocity of extrusion.

At least some of said regions ofthe preform chamber may each have a width

which is the same predetermined percentage greater than the width ofthe respective

corresponding region ofthe die cavity. Alternatively or additionally, at least some of

said regions ofthe preform chamber may each have a width which is greater than the

width ofthe respective corresponding region ofthe die cavity by the same predetermined

amount.

The width of said regions ofthe preform chamber are preferably substantially

symmetrically disposed in relation to the width ofthe corresponding region ofthe die

cavity. However, as previously mentioned, the width of one or more of said regions of

the preform chamber may be offset in relation to the width ofthe corresponding region

ofthe die cavity.

Preferably the bearing length of each region ofthe preform chamber is provided

by a bearing part thereof which is immediately adjacent the corresponding region ofthe

die cavity.

Each region ofthe preform chamber may include a part which is upstream ofthe

bearing part which provides the bearing length, and which increases in width as it

extends away from said bearing part.

The die cavity and preform chamber are preferably formed in separate

components which are clamped together with the preform chamber in communication

with the die cavity. Alternatively the die cavity and preform chamber may be integrally

formed in a single component. However, an advantage of forming the preform chamber

and die cavity in separate components is that it may allow the preform chamber

component to be re-used with a new die cavity component should the original die cavity

component wear out.

The invention also includes within its scope a method of manufacturing an

extrusion die comprising forming the die with a die cavity having a shape corresponding

to the cross-sectional shape of the required extrusion, and a preform chamber in

communication with the die cavity, the preform chamber being of generally similar shape

to the die cavity but of greater cross-sectional area, so that regions of the preform

chamber communicate with corresponding regions respectively of the die cavity, and

adjusting the bearing lengths of different regions ofthe preform chamber in relation to

the dimensions and position of those regions so that, in use, extrusion material passing

through each region ofthe preform chamber is constrained to move at a velocity such

that the material passes through all regions ofthe die cavity at a substantially uniform

velocity.

The following is a more detailed description of embodiments ofthe invention,

by way of example, reference being made to the accompanying drawings in which:

Figure 1 is a diagrammatic front face view of an extrusion die formed with two

simple cavities,

Figure 2 is a diagrammatic section on the Line 2-2 of Figure 1,

Figure 3 is a diagrammatic section on the Line 3-3 of Figure 1,

Figure 4 is a front face view of an extrusion die showing two die cavities of

slightly more complex form than Figure 1,

Figure 5 is a section on the Line 5-5 of Figure 1,

Figure 6 is a diagrammatic front face view of part of a further form of die cavity,

Figure 7 is a diagrammatic section on the line 7-7 of Figure 6,

Figure 8 is a diagrammatic section through a die having a die cavity of zero

bearing length,

Figure 9 is a diagrammatic section through another form of die,

Figure 10 is a diagrammatic section through a further form of die,

Figure 11 is a similar view of a modified version ofthe cavity of Figure 10, and

Figure 12 is a diagrammatic section through a die cavity incoφorating cooling _.

Figure 1 shows the front face 10 of an extrusion die 1 1 formed with two cavities

12 and 13 of generally flattened Z-shape.

In a conventional prior art construction each die cavity 12 or 13 would

communicate with an enlarged divergent exit cavity formed in the back face ofthe die

plate. The bearing length of different regions ofthe die cavity, i.e. its dimension in the

direction of extrusion, would be adjusted by adjusting the depth of this exit cavity. By

this means the bearing length of each part of the die cavity would be adjusted in a

manner to result in a substantially uniform velocity ofthe extrusion material through all

parts ofthe die cavity.

By contrast, in accordance with the present invention, the front face ofthe die

is formed with a preform chamber through which the extrusion material is forced before

it reaches the die cavity 12 or 13, thus enabling the velocity ofthe extrusion material to

be adjusted before it reaches the die cavity itself.

Referring to Figure 2 it will be seen that the die 11 comprises a back plate 14 in

which the die cavity 12 itself is formed. All parts ofthe die cavity 12 have a constant

bearing length 15 which may, for example, be 2mm. An exit cavity 16 leads from the

die cavity 12, the walls ofthe cavity diverging as they extend to the back face 17 ofthe

die plate 14.

Clamped rigidly to the back plate 14 is a front plate 18 which is formed with a

preform chamber 19. The preform chamber is generally similar in shape to the die cavity

12 but the width of all regions ofthe preform chamber is greater than the width ofthe

corresponding regions ofthe die cavity 12. As may be seen from Figure 1, in the case

of the upper die cavity 12 the preform chamber 19 has a width which is increased by

50% all around the die cavity 12 so that the overall width of each region ofthe preform

chamber 19 is twice the overall width of the corresponding region of the die cavity.

Such aπangement will be referred to as a "50% growth" aπangement.

In accordance with the present invention the bearing length 20 (see Figure 2) of

each region ofthe preform chamber 19 is calculated in accordance with the width ofthe

preform chamber in that region, and in accordance with its distance from the centreline

21 ofthe die, to give a required velocity of extrusion material as it enters the die cavity

itself. The velocity at entry to each region ofthe die cavity is selected such that the rate

of subsequent flow through all regions ofthe die cavity is substantially uniform. The

bearing length 20 ofthe preform chamber is controlled by milling into the front face 10

of the front plate 18 an entry cavity 22 of appropriate depth to give the required

resultant bearing length 20 to the preform chamber 19.

The entry cavity 22 comprises a flat nanow shoulder 22a, to define the inlet end

ofthe preform chamber 19 exactly, and surfaces 22b inclined at approximately 45° away

from the chamber 19. Such inclination is necessary to ensure that these surfaces do not

act as a bearing on the extrusion metal so as to alter the bearing effect ofthe preform

chamber 19.

The use of a preform chamber 19 where the side walls ofthe preform chamber

are parallel enables the velocity to be controlled, by adjusting the bearing length 20

using well established means of calculating the required bearing length to achieve the

required velocity. Also, since adjustments to the die to adjust the velocity do not require

any alteration to the die cavity 12 itself, as is the case in most prior an methods _, the die

cavity 12 may be formed in any material to give the required strength and wear

resistance without taking into account any necessity of being able to adjust the bearing

length ofthe die cavity after it has been initially formed. Also, since the bearing cavity

itself remains unchanged, it may be coated with an appropriate finish _, such as by

nitriding, so as to give the best possible surface finish to the extruded profile _.

Also, since the die cavity 12 itself is of constant bearing length, this also

inherently results in a finer finish on the extruded profile, in contrast to the prior art

aπangements where the extrusion is likely to be marked where it passes through a region

ofthe die cavity where two different bearing lengths are adjacent one another.

The extent of increase in width, or "growth", ofthe preform chamber in relation

to the die cavity may be of any required value, depending on the size and shape of the

die cavity itself and its position in relation to the centreline of the die. By way of

example, Figure 1 also shows a die cavity 13 where the preform chamber 23 exhibits

200% growth, i.e. the increased width ofthe preform chamber on each side ofthe die

cavity is twice the width ofthe die cavity 13 itself. Again, an entry cavity 24 is milled

into the front face 10 ofthe front plate 18 ofthe die, the depth ofthe entry cavity 24

being selected to give a required bearing length to the preform chamber 23 and hence

a required velocity ofthe extrusion material as it reaches the die cavity 13 itself.

In the case, such as those shown in Figure 1 , where the percentage "growth" of

the preform chamber is constant for all regions ofthe die cavity, the velocity of extrusion

material through the preform chamber is controlled solely by adjusting the bearing length

of the preform chamber leading to each region. However, in some cases, with more

complex profiles, it may be advantageous also to vary the percentage growth of the

preform chamber in different regions of the die cavity, and Figures 4 and 5 show an

example of this.

Referring to Figures 4 and 5, the extrusion die 25 again comprises a front plate

26 and a back plate 27. The back plate 27 is formed with two identical die cavities, an

upper cavity 28 and a lower cavity 29. Each die cavity has a uniform bearing length of,

for example 2mm- in all regions thereof and leads to an exit cavity 30 which diverges

outwardly to the back face 31 ofthe die.

The front plate 26 is formed with preform chambers 27 and 33 which

communicate with the die cavities 28 and 29 respectively and entry cavities 32 and 34

are milled in the front plate 26 to communicate with the die preform chambers

respectively.

As best seen in Figure 4, the two die cavities 28 and 29 are ofthe same shape,

the upper cavity 28 comprising a central region 28a of generally flattened Z-shape, an

end region 28b of greater width than the central region 28a, and an opposite end region

28c of smaller width than the central region. For example, the central region may have

a width of 2mm the end region 28b a width of 4mm, and the end region 28c a width of

lmm.

As in the previous aπangement the preform chamber 27 is of generally similar

shape to the die cavity 28, and has 50% growth, i.e. the width ofthe preform chamber,

on each side ofthe die cavity, is increased by 50% ofthe width ofthe die cavity.

Also as in the previous aπangement, the bearing lengths ofthe different regions

of the preform chamber 27 are adjusted in relation to the width and position of the

regions ofthe preform chamber, and hence ofthe regions ofthe die cavity with which

they communicate. Thus, the enlarged region 27b ofthe preform chamber will require

a significantly greater bearing length than the region 27a, as may be seen from Figure 5,

in order to reduce the velocity to what is appropriate for the larger area ofthe region of

the die cavity, whereas the smaller region 27c of the preform chamber will require a

smaller bearing length than the region 27a.

In some cases finer control of the velocity of the extrusion material may be

achieved by also varying the percentage growth of different regions of the preform

chamber, in addition to varying their bearing lengths, and such an arrangement is shown

in the case ofthe lower die cavity 29 in Figure 4. In this case the central region 33a of

the preform chamber 33 still has 50% growth, but the enlarged end region 33b ofthe

preform chamber has only 25% growth. The opposite end region 33c ofthe preform

chamber, communicating with the reduced end region 29c ofthe die cavity, has 200%

growth.

Looked at another way, the regions 33a and 33b ofthe preform chamber may

be regarded as having a width which is greater than the width of the respective

coπesponding regions 29a and 29b ofthe die cavity by the same predetermined amount,

even though the region 29b ofthe die cavity is wider than the region 29a.

The effect ofthe proportionally reduced growth ofthe preform chamber region

33b is to decrease the velocity of the extrusion material through that region of the

preform chamber compared with the velocity through the region 33a, so that a shorter

bearing length is required in region 33b to achieve the required velocity through the

region 29b of the die cavity. Similarly the increase in width ofthe region 33c ofthe

preform chamber serves to increase the velocity ofthe extrusion material in a manner

appropriate for such a nanow region ofthe die cavity. This overcomes the possible

problem that, with a uniform percentage growth, it may not be possible, by adjustment

ofthe bearing length alone, to achieve sufficient velocity ofthe extrusion material in the

preform chamber 33c to ensure that the material passes at the required velocity through

the region 29c ofthe die cavity.

In all ofthe above arrangements according to the invention the provision of a

preform chamber coπesponding in shape to the die cavity thus provides great flexibility

in control over the velocity of the extrusion material through the die to enable the

optimum extrusion conditions to be obtained.

It will be appreciated that the simple shapes of die cavity shown are merely by

way of example and the invention is applicable to any profile shape. For example, the

invention is applicable to extrusion dies for extruding hollow shapes. In this case each

preform chamber will be formed partly in the male portion ofthe die and partly in the

female portion so as to provide a preform chamber communicating with the whole ofthe

die cavity.

In the arrangements of Figures 1-5 each region of the preform chamber is

substantially symmetrical with respect to the corresponding region ofthe die cavity, that

is to say the preform chamber region overlaps the die cavity region by a similar amount

on each side. However, this is not essential and in some configurations of die cavity

certain regions of the cavity may be so close together that symmetrically disposed

regions ofthe preform chamber would overlap. In such circumstances the regions ofthe

preform chamber may be offset with respect to the coπesponding regions of the die

cavity so that they do not overlap and may therefore have separate effects on their

respective regions ofthe die cavity. Such an aπangement is shown in Figures 6 and 7.

As best seen in Figure 6, the die cavity 35 is formed at one end to provide two

spaced parallel limbs 36. The limbs 36 of the die cavity may be so close that if the

coπesponding regions 37 of the preform chamber were symmetrically disposed with

respect to the regions 36 ofthe die cavity, they would overlap, thus interfering with the

coπect controlling effect ofthe preform chamber. Accordingly, in this case the regions

37 ofthe preform chamber are offset with respect to their coπesponding regions 36 of

the die cavity, so as to form two separate and distinct regions. Each region 37 ofthe

preform chamber therefore can be adjusted to control accurately the flow of metal to its

coπesponding region of the die cavity. The offsetting of the regions of the preform

chamber has no significant adverse effect on the operation ofthe invention. Provided

that the preform chambers result in the extrusion metal reaching the die cavity at uniform

velocity, it does not matter where the preform chambers are located in relation to the die

cavity.

Since the velocity of the extrusion material through a region of the die is

increased by reducing the bearing length in that region, the overall velocity of the

material through the die may be increased by reducing all bearing lengths. In the

majority of conventional extrusion dies it is necessary to retain significant bearing lengths

in all regions ofthe die cavity itself, since differential variation in such bearing lengths

is the only way of controlling velocity through the different regions ofthe die cavity.

The present invention, however, allows the use of a die cavity of uniform bearing length.

Accordingly, the present invention may be used with a die cavity of so-called zero

bearing length, as previously discussed, and one such aπangement is shown in section

in Figure 8.

In this arrangement the die plate 38 is formed with a die cavity 39 having an inlet

aperture 40 in the shape ofthe required extrusion. The walls 41 ofthe die cavity are

negatively tapered, for example at 1.5°, i.e. they diverge slightly as they extend away

from the aperture 40. The die plate is cut away at the downstream end ofthe die cavity

39, in conventional manner, as indicated at 42.

Since the walls 41 are negatively tapered they do not apply any significant

frictional restraint to metal passing through the aperture 40 and the metal is shaped

solely by the comers 43 around the aperture 40 so that the bearing length ofthe die

cavity is essentially zero. It will be appreciated, however, that the comers 43 require to

be smooth so as to provide a good surface finish on the extruded profile. These comers

will therefore be slightly radiused so that, in practice, there will be a bearing length

which is so small as to be negligible, rather than an actual zero bearing length.

As in all embodiments ofthe present invention, the velocity of extrusion material

through the aperture 40 is controlled by the bearing length ofthe different regions ofthe

enlarged preform chamber on the upstream side of the die cavity. As previously

described, the regions ofthe preform chamber upstream ofthe control bearing length

44a are tapered outwardly, as indicated at 45 in Figure 8, so that there is insignificant

risk of such parts of the preform chamber plate 44 having any bearing effect on the

extrusion material passing through it.

Another way of increasing the overall velocity of material through the die is to

reduce as far as possible the bearing lengths of the different regions of the preform

chamber.

In all the aπangements previously described, the bearing length portion of each

preform chamber region is preferably as close as possible to the die cavity. However,

the invention does not exclude arrangements where the bearing lengths ofthe preform

chamber regions are spaced upstream from the corresponding regions ofthe die cavity.

Figure 9 shows an aπangement where the preform chamber region 50 has a zero bearing

length aperture 51 spaced upstream of a zero bearing length die cavity 52. This

arrangement minimises the overall bearing length of the die and thus provides for

maximum velocity of extrusion material through the die.

In order to retain control of velocity through all regions of the die, only the

region ofthe preform chamber requiring minimum bearing length will be of zero bearing

length. However, this will enable the bearing lengths ofthe other regions to be reduced

by a coπesponding amount, as will be described with reference to Figures 10 and 11.

Figure 10 shows an arrangement in accordance with the present invention

where regions 46, 47 and 48 ofthe preform chamber are of different bearing lengths,

region 46 being of the shortest bearing length. However, the same effect may be

achieved by reducing the bearing length of all regions of the preform chamber by an

amount equal to the bearing length ofthe smallest region 46. As shown in Figure 11,

this may be effected by reducing the bearing length ofthe preform chamber 46 to zero

by applying a negative taper to the sides ofthe chamber as indicated at 46a. The bearing

lengths of the other preform chambers are reduced by a coπesponding amount by

negatively tapering a similar length portion thereof, as indicated at 47a and 48a. Since

the bearing lengths of the three regions of the preform chamber have the same

relationship, the velocity of the extrusion material as it reaches the die plate 49 is

uniform. However, the overall velocity ofthe material is increased as a result ofthe

reduction in effective bearing length of all regions 46, 47 and 48 ofthe preform chamber.

In the aπangements described above the die comprises a separate die plate and

preform chamber plate, the two plates being clamped together face-to-face. However,

in some circumstances it may be desirable and possible to combine the two plates into

a single integral plate formed with the appropriate apertures. However, the two-plate

arrangement will usually be prefeπed since it facilitates correction ofthe bearing lengths

in the preform chamber plate and also allows the preform chamber plate to be re-used

if the die plate wears out first, which is likely to be the case.

Figure 12 shows another situation where a two-plate arrangement is to be

preferred.

In some circumstances it may be desirable to cool the die and the extrusion

material as it passes through the die cavity to reduce the risk of local melting. Cooling

ofthe extrusion material is usually done by injecting a cooled inert gas, usually nitrogen,

into the downstream region ofthe die plate, but cooling ofthe die itself may be difficult.

Two-plate aπangements according to the present invention enable such cooling to be

effected in a simple and convenient way, as illustrated diagrammatically in Figure 12.

In this case a main channel 53 is formed in the die plate 54 closely adjacent the die cavity

55 and passages 56 extend laterally from the channel 53 to open into the downstream

portion of the die cavity. The preform chamber plate 57 then closes the channel 53.

Cooled nitrogen is then pumped under pressure into the channel 53, thereby cooling the

die itself, and is fed therefrom along the passages 56 to cool the extrusion material

passing through the die cavity..