Plastics pipe is made from a number of different materials and for many applications they comprise polyethylene and possibly cross-linked polyethylene.

Pipes of large diameter have correspondingly thick walls, although, for some applications even smaller diameter pipes may have relatively thick walls to improve their strength or other characteristics.

Pipes made from plastics materials are generally extruded by an extrusion die. Even immediately after the extrusion head, the skin of the pipe is solid having been cooled by coolant water flow in the die head. However, at this stage it is only the skin which is solid, the wall of the pipe mostly still being liquid.

Consequently, after extrusion the newly formed pipe enters a first cooling tank which comprises a vacuum section serving accurately to size the external diameter of the pipe, as well as to commence the cooling process.

Such a vacuum section might typically be about 6 metres long. As well as spraying cooling water onto the outside of the formed pipe, a cooling medium is typically injecte internally of the pipe also. Such medium might be a gas such as nitrogen having a relatively high thermal conductivity but is seldom liquid due to the difficulty in handling the flow at the end of the extrusion process when the formed pipe is being cut to length.

In any event, after leaving the vacuum section, the formed pipe typically enters two further cooling sections, perhaps each of 6 metres length, until the pipe finally emerges as solid throughout its wall thickness.

It should be borne in mind that it is because plastics is such a poor conductor of heat that it takes such a length of cooling sections before it has completely solidified.

On the other hand, two other factors are also very important, that is to say, the wall thickness and the speed of extrusion. Clearly, from a commercial point of view the speed wants to be as high as possible and typical speeds are about 10 metres per hour for large diameter pipes. The higher the speed, the longer the cooling sections will need to be. However, length is not a parameter that can be adjusted at will. A pipe in the process of formation is driven from both ends. At the die head end it is driven by the extruder although clearly the pipe cannot have any significant compression after exiting the die. Consequently the largest proportion of the driver comes from a hauler at the end of the line and which pulls the pipe along.

Clearly, therefore, the speed of drive must accommodate the need to maintain pipe integrity along its entire length. The hauler generally comprises a series of driven belts pressed against the surface of the pipe, the belts being radially disposed around the circumference of the pipe. Each belt is parallel the long axis of the pipe.

Pipe manufacturers have on the whole devised arrangements which meet the conflicting requirements, but one issue

which continues to cause problems in trying to accommodate is die slump.

When it is extruded the pipe is essentially two annular solid skins, with a liquid central annulus. As the pipe progresses through the cooling sections, the skins become increasingly thick and the central annulus thin.

However, while the central annulus is liquid, gravity has the chance to operate and some of the central annulus can flow downwards thinning the top wall and bulging the bottom wall. In fact, Pitman et al (Intern. Polymer Processing IX (1994) 2, pl30-140"Cooling and Wall Thickness Uniformity in Plastic Pipe Manufacture") found in a 315 mm SDR 11 pipe about 15% variation between the top and bottom thickness of the final pipe. (SDR is Size- to-Diameter ratio, indicating a wall thickness in this case of 315/11 mm (ie about 28.6 mm) Two essential methods of avoiding this problem have been suggested. The first is to distort the original profile of the pipe during extrusion so that the die slump in fact serves to take the pipe true by the end of the extrusion line. The second is to rotate the pipe during extrusion. However, this latter course is impractical because the extrusion die is so large and bulky that to contemplate rotation of it is out of the question.

Nevertheless, it is an object of the present invention to provide a method and apparatus to address the problem of die slump and at least mitigate its effects.

In accordance with the present invention there is provided a method of extruding plastics pipe normally

subject to die slump, the method comprising the step of twisting the pipe after extrusion.

This is equivalent to rotation of the pipe but obviously does not involve rotation of the die head. Also, since it imposes a twist on the pipe it is not something that will evidently work. However, it has been found that a rate of rotation of about 0.007 radians per second will on the one hand effectively counter die slump while at the same time only imposing a twist of about one turn in every three metres of pipe. One advantage of this method is that the twisting will not be uniform along the length of the pipe line (it will be uniform in terms of the number of turns per unit length of finished pipe) because the resistance to twist is higher the more solid (ie the cooler) the pipe becomes. So the rotation gradient will be at its highest at the very points where there is most liquid present and where the vast majority of die slump occurs.

In a different aspect the present invention provides apparatus for extruding plastics pipes normally subject to die slump, the apparatus comprising a fixed extruder die, a vacuum sizing cooling tank, one or more further cooling tanks and a hauler, the hauler being adapted to rotate as well as axially translate the pipe.

Said hauler preferably rotates the pipe sufficiently fast to reduce die slump to less than 10% variation in wall thickness of the finished pipe.

Said hauler preferably rotates the pipe at between 0.004 and 0.009 radians per second, preferably between 0.006

and 0.007 radians per second. Said hauler preferably translates the pipe at between 7 and 15 metres per hour, preferably between 9 and 12 metres per hour.

The pipe preferably has a wall thickness in excess of 20 mm, and preferably an SDR in excess of 7. More especially it may have a wall thickness between 25 and 35 mm and an SDR of between 10 and 20.

The hauler may itself be mounted rotatably in a frame fixed with respect to the die, or instead may be arranged to import said rotation and translation through an integrated drive mechanism.

In one case, said integrated drive mechanism comprises drive devices pressing the outside surface of the pipe but inclined with respect to the axis of the pipe. Said inclination may be between 20° and 30°, preferably about 25°.

Said drive devices can comprise belts but, in order to accommodate the curvature of the pipe, may instead comprise wheels. Each wheel might be driven by its own servo motor so that their speeds may be accurately co- ordinated.

A belt may indeed be wrapped around the pipe.

The invention is further described hereinafter, by way of example, with reference to the accompanying drawings, in which: Figure 1 is a plan view of a pipe extrusion line;

Figure 2a is a plan view of a hauler in an apparatus in accordance with the present invention, Fig 2b being a vector diagram of forces; and Figure 3 is a perspective view of a different embodiment of a hauler for use in accordance with the invention.

In Figure 1 an extruder die 10 extrudes under pressure a pipe 12 which, on exiting the die 10, immediately enters a first tank 14 which includes a number of spray nozzles 16 spraying cold water onto the outside of the pipe 12.

Since the pipe at this point consists of a thin solid skin with liquid central annulus, the tank 14 is under vacuum so that nitrogen injected to the interior of the pipe through the die core (not shown) under atmospheric pressure, the pipe 12 is maintained round by virtue of the pressure differential across its wall. On exiting the tank 14, the pipe enters two further tanks 18,20 which further cool the pipe 12. They likewise have spray nozzles 22 and 24. The pipe 12 is being drawn through the tanks 14,18,20 by a hauler 26 which comprises a number of belts 28 disposed parallel the axis 30 of the pipe 12. The belts 28 are disposed around rollers 32, at least one of which is driven to draw the belt in contact with the pipe 12 to the right in Figure 1. There are maybe five or six of the belts 28 distributed evenly around the circumference of the pipe 12 so that a firm grip may be had on the pipe 12 and it can successfully be pulled through the tanks 14,18,20. Finally, a saw 36 cuts the cooled and solidified pipe 12 into appropriate lengths. The speed of draw of the pipe 12 from the tanks 14, 18,20 depends on a number of factors including the

rate of extrusion from the die 10, the rate of cooling of the molten plastics in the material of the pipe, which is dependent on the thickness of the pipe wall. The tensile strength of the pipe in the tanks 14,18 and 20. However, these aspects have no relevance to the present invention.

On the other hand, the present invention is concerned with die slump and accordingly proposes that the hauler 26 should be modified as shown at 26'in Figure 2a where the belts 28,32 have been replaced by wheels 40, the radial plane of which, 42, is disposed at an angle a to the longitudinal axis 30 of the pipe 12.

Figure 2b shows a vector diagram of velocities where Vx is the desired axial translation of the pipe 12 along the axis 30. V. is a desired rotational velocity further <BR> <BR> <BR> <BR> described below, and VR is the resultant velocity in the direction of angle a and determines the velocity of rotation of the wheels 42, as well as the angle a. That is to say a = tan~1 (Vz/Vx) A rotational velocity Vx imparts a twist in the pipe 12, but at the same time prevents die slump by continuously changing the position of the pipe with respect to gravity, so that die slump does not have a chance to establish. It is found that a rotation of about 0.0015 ms~l is adequate, which, with a desired translational <BR> <BR> <BR> <BR> <BR> velocity V. of about 0.0033 ms-1 results in an angle a of about 26°. On a line of about 25 metres long this results in about 7.23 rotations of the pipe, ie about one twist every 3.5 metres. This is acceptable, and in any event takes place where the pipe is more liquid and therefore

without damage to solidified material.

The wheels 40 each have their own servo motor 44 which is electronically controlled and co-ordinated with the other motors to ensure consistent drive to the pipe 12.

Referring to Figure 3, an alternative belt drive hauler 26"is shown in which a belt 28'is guided by at least two pulleys 32', and driven by at least one of them. The belt is wrapped around the pipe 12 and disposed at an angle thereto such that, drive of the belt imparts both an axial translational and rotational movement to the pipe 12. Alternatively, such a drive might be used in parallel with the traditional drive 26.