There are two equally important approaches to the design and optimisation of extruder screws, for both single and twin screw machines. The first is an analytical approach involving the use of mathematical models and computer analysis. The second is a purely experimental approach involving the use of either production equipment or specialist small scale laboratory machines.
|
|
We have demonstrated that optimisation of the extruder screw can lead to improvements in throughput of more than 20% (British Plastics Industry Process Efficiency Audit Report, November 1996, The Nottingham Trent University/DTI, ISBN 0905488768 ( F.Fassihi and P.Prentice)). A case study - in collaboration with B&H Plastics Ltd. (a film blowing company) of Anchorage Works, New Road, Radford, Nottingham, UK - showed that improving the design of the extruder screw led to an increase in productivity of 22%, whilst still maintaining a suitable melt quality, over the entire range of screw speeds - with no increase in power consumption. (See the figure on the left, in which the motor speed is directly proportional to the screw speed). |
The extruder screw lies at the heart of many processing methods. It is obviously one of the most crucial parts of a single screw extruder, but it is also a major component in injection moulding, film blowing, extrusion blow moulding and many other processes. Most screws, until recently, have been designed by purely empirical means with little, or no, scientific reasoning. Here at PolyTech Consultants we have corrected this obvious shortcoming. We have applied the science of plastics processing to the design of processing equipment, based on the experience of many years in research and development; both in industry and in universities.
The role of the extruder screw in all of these processes is to melt the solid polymer and convey the resulting melt to the profiling die of the extruder or to the runner system of an injection moulding machine.
The different thermal properties of all polymers mean that, ideally, the design of the extruder screw should be optimised for each and every polymer type, but we are aware of the impracticability of such an ideal. What we aim to do is to develop a screw suitable for the range of materials commonly used by a company. This means that the resulting screw will work satisfactorily for all the company's materials, but it will not be ideal for any one of them - it will be a compromise.
It is a common misapprehension that the heating elements of the extruder are responsible for melting the polymer, granules or powder. But, the very low thermal conductivity of all polymers means that it is not possible for the heat to be transferred from the barrel wall to the core of the material. It is the rotation of the screw inside the extruder barrel that smears the molten polymer between the solid bed and the barrel wall, generating shear and mechanical heat, that is responsible for most of the heat transfer, and for the major part of the melting of the polymer. It is this work induced heat that is primarily responsible for converting the solid polymer into melt. The screw rotation is also responsible for mixing the molten material during its passage along the screw.
A successful screw design is one in which the amount of shear heat generated is sufficient to ensure that all the polymer is converted to a suitable molten state before the end of the screw.
At PolyTech Consultants we have computer software, developed over the past 18 years, which simulates the passage of the polymer material through a single screw extruder. The program calculates the melting rate, the melt temperature, the mass flow rate and the power consumption. For critical applications such as film blowing, the program also predicts the presence and degree of surging in the extruder. The mathematical model, on which the program is based, assumes that there is no slip of the polymer melt at the walls of the screw and barrel. However, it is also possible to estimate, experimentally, the degree of slip occurring using the instrumented extruder described below. The ratio of the actual output of the screw divided by the theoretical output, based on the screw geometry, is a dimensionless parameter we call Π Q. This parameter can be used to measure the wear of the screw. The greater the wear, the greater the back-flow over the flight, cutting down the screw efficiency and the lower Π Q. The maximum efficiency of the screw, with no slip and no screw wear gives a Π Q of 0.5 (Pure Drag Flow)
No screw can be designed without first knowing all the thermal and rheological properties of the plastic material to be processed. For the screw design to be successful, the physical properties of the polymer must be known precisely. The thermal conductivity of the polymer can be determined most easily, and most cheaply, using the Lee's Disk method. At PolyTech Consultants we require 1 kg of material to determine all the necessary thermal and rheological properties of the polymer, which are then used as the input data for the program.
A full analysis of the material produces the following data, all of which is necessary for a successful screw design:
|
*********************************** Polymer Sample *********************************** S.G. OF MELT.................................... = SPECIFIC HEAT OF SOLID.................[J/kg/°C] = SPECIFIC HEAT OF MELT..................[J/kg/°C] = LATENT HEAT OF FUSION.....................[J/kg] = MELTING TEMP................................[°C] = POWER LAW INDEX................................. = VISCOSITY AT UNIT SHEAR RATE.............[Ns/m2] = TEMP COEFFICIENT OF VISCOSITY.............[°C-1] = THERMAL CONDUCTIVITY OF SOLID...........[W/m/°C] = THERMAL CONDUCTIVITY OF MELT............[W/m/°C] = |
This data, along with the processing variables, such as:
|
|
is used in the Screw Optimisation Program enabling the prediction of the following screw parameters:
|
BARREL DIAMETER..............................[m] = LENGTH OF FEED SECTION.......................[m] = LENGTH OF COMPRESSION SECTION................[m] = LENGTH OF METERING SECTION...................[m] = DEPTH OF FEED SECTION........................[m] = DEPTH OF METERING SECTION....................[m] = RADIAL CLEARANCE.............................[m] = FLIGHT WIDTH.................................[m] = SCREW PITCH..................................[m] = NUMBER OF SCREW STARTS.......................... = |
| Knowing these screw parameters, it is now possible produce a screw using conventional machining methods or it is possible to program a Computer Numerically Controlled (CNC) machining centre to make the screw, since all the dimensions necessary to describe a screw are known. Other factors, such as the root radii of the channels, are left to the designer and can be added at the programming stage. |
|
 |
Unlike other software packages which claim to compute the performance of extruder screws, the results obtained from our computerised optimisation have been compared, over many years, with those obtained using a fully instrumented Francis Shaw extruder (shown on the left) using wide range of screw profiles (shown above). This extruder has been extensively modified; it has, as can be seen, ten pressure transducers and five thermocouples along its barrel. It has a torque transducer between the gearbox and drive motor. The hollow screws are fitted with up to ten thermocouples, embedded at regular intervals along the helical length of the screw and threaded through them to a rotating slip-ring device. One of the most important features of this particular extruder is a crash cooling facility - after stopping the machine using the emergency stop button, the barrel can be flooded with 400 litres of water per minute. This enables the the melt to be frozen onto the screw before both the screw and frozen scroll are extracted from the barrel (as in the figure below left) using a specially designed 30 tonne hydraulic extraction unit. The frozen scroll can then be removed from the screw and sectioned, allowing the melting rate to be determined experimentally. Other experimentally determined parameters are mass flow rate, melt temperature, power consumption and pressures at any point in the barrel. |
The barrel has been further modified so that sleeves with different groove configurations can be included to investigate the effect of grooved feed pockets (below right) on the output of the extruder.
|
|
|
 |
Having removed the frozen scroll
from the screw it is possible to cut it and polish the surface, as
shown here on the left. In this case the material is a white
polystyrene tumbled with a very small quantity of a carbon black
pigment. It can be seen from the uniform grey colour of the
material in the melt pool, even as early as turn number 6, that mixing
is taking place all the way along the screw - not just at the end of the
screw. This brings into question the need for mixing elements
within the screw. The numbers on the left of the plaque refer to the
distance along the screw, measured in turns or diameters.
It can be seen from the polished plaque on the left that a melt pool is formed at the leading edge of the screw flight and grows in size, whilst the solid bed decreases, as the material progresses along the screw . The barrel film is seen to be very thin and constant. It is obvious that the material in this example is within the compression region of the screw, because the channel depth is constantly changing. The rate at which this melt pool develops, or the rate at which the corresponding solid bed disappears, defines the melting rate of the polymer. It is this melting rate that the software described above attempts to model. The results of this modelling are shown on the following page. |
It is obvious that the melting rate of the of the material in the extruder is of paramount importance. Any screw and barrel combination needs to produce a melt of suitable quality at the correct temperature before the end of the screw. The computer program calculates the melting rate for each proposed screw configuration. It also predicts the melt temperature in the melt pool, the pressure gradients within the barrel and the power consumed during melting.
Therefore, using the optimisation program, a screw geometry can be proposed which will produce a suitable melt at the highest possible output rate. This can be achieved without the need for expensive and time-consuming manufacture of screws which are subsequently found not to work.
If a screw is to be manufactured, then this optimised design is the best place to start when it comes to cutting metal.
