In-line rheological testing of thermoplastics and a monitored device for an injection moulding machine: Application to raw and recycled polypropylene

Abstract

A methodology for the rheological testing of polymers during the injection moulding process has been developed. This method has been designed to consider the non-conventional features of the plastication phase that result from the injection of recycled thermoplastics. The majority of previous research has been focused on extrusion machines or injection moulds. In this study, a measuring device is attached directly to the nozzle of the plasticising unit, enabling in-line measurements of the pressure drop and temperature that are necessary to calculate the apparent viscosity at different shear rates. Optimisation of the slit geometry also allows the production of parts without disturbing the operating parameters. The results have been corrected to characterise a mineral-filled polypropylene and its recycled form and to obtain the constants of its Carreau-WLF rheological model with SIM-Fit software. The material model constants have been implemented in Cadmould 3D-F simulation software. A comparison of the results with experimental injection moulding is consistent with the expected trends.

Introduction

The use of the plastic injection moulding process continues to increase with many parts manufactured using this method. New challenges in plastics, including reuse, newer lightweight materials, microcellular foaming and nanocomposites, have led to new technological developments in processing. New non-conventional injection technologies have recently emerged to cover a larger market, Berins [1]. These new techniques employ processes, moulds and material features that differ from those used in conventional thermoplastic injection moulding, Hettinga [2], Rothe [3], Rosato et al. [4].

CAE systems for conventional injection moulding simulation are available to aid in mould design and to optimise the process, Gucerin [5], Lee and Castro [6], Tucker [7]. At present, none of the 1-D beams used in 2-D and 21/2-D filling analysis or the newer 3-D filling analysis predict the magnitude of the shear-induced filling and melt imbalances in hot runner systems and multi-cavity moulds, Beaumont [8]. For this reason, pressure predictions diverge from real cases; a lower pressure drop in hot runners than that measured experimentally is predicted because the viscosity of the molten polymer is reduced with increasing melt temperature along the runner, but additional complex flow behaviour is not taken into account. These CAE systems do not consider non-conventional features for their calculations and, therefore, they cannot properly address non-conventional injection moulding (microcellular injection or long glass fibre injection, for example). This is the case for the plastication phase. An approximation for CAE simulations of non-conventional injection can be performed by characterising the material under non-conventional conditions, directly in an injection moulding machine after plasticising the material, and then introducing these material parameters into conventional CAE injection systems.

Thus, there is a need to characterise non-conventional injection moulding from the perspective of describing the rheological behaviour of a polymer affected by specific, different non-conventional material features. Non-conventional material features refer to recycled materials, Ames [9], Hoffman [10], Nichetti and Manas-Zloczower [11], Scriven and Sykes [12]; microcellular foamed materials Ruogu et al. [13], Tomasko et al. [14] or nanocomposites, Krishnamoorti and Yurekli [15], Wu et al. [16]; or compounds, Braun et al. [17].

The rheological characterisation of a material is commonly performed using a capillary rheometer, Abbas [18], Haghtalab et al. [19], Pillo et al. [20], Rao [21], Rosenbaum and Hatzikiriakos [22]; however, this method is not available for non-conventional circumstances, such as previous exposure, Burke and Kazmer [23], Dontula et al. [24], Fann et al. [25]. Therefore, it is necessary to use other characterisation methods, such as real tests conducted in a similar way to the one used in this study, which consists of a rectangular monitored nozzle with pressure sensors.

Previously published studies have developed methodologies and applications to measure the apparent viscosity in extrusion machines, Hay et al. [26], Son [27], Chen et al. [28], or in injection moulds, Clavería et al. [29], Bariani et al. [30]. The work presented here develops a device and methodology to directly measure the apparent viscosity in-line on an injection moulding machine. The application of a similar device for in-line process monitoring was applied in a 3000 ton injection moulding machine for the production of washing machine tubs, Fernández et al. [31].

The use of a slit-die rheometer joined to a piston after an extruder, Hay and Mackay [32], revealed that the apparent viscosity measured using a slit-die rheometer or capillary rheometer differed from the viscosity measured using a parallel plate apparatus. The flow is affected by several factors, as listed below:

• Slip

• Pressure-dependent viscosity

• Viscous heating

• Heat transfer

• Entrance and exit effects

The results obtained indicated that the apparent viscosity measured using a slit-die rheometer is higher than that measured using a parallel plate rheometer and that the difference increases with higher shear rates and slit die height. Adding a piston after an extruder allows the viscosity of the material to be measured in-line, but it prevents extruded parts from being produced continuously. A similar apparatus designed to correct shear rate calculations, Son [27], was used to improve real viscosity calculations by introducing shear-thinning effects in the lateral walls of the slit die for a H/W > 0.1 ratio. This study did not describe any details about the device used and presented results only in the lowest shear rate region (<100 1/s), which is far from the typical values in injection moulding.

A mould with pressure sensors and slit die geometry, Bariani et al. [30], was used to evaluate the influence of plastication on the viscosity of polypropylene in an injection moulding machine without temperature measurements. The comparison of the results with those from a capillary rheometer was not discussed and showed differences in apparent viscosity near 20%. The use of a mould for rheological characterisation does not allow in-line monitoring for the injection moulding process, and is not supportive of cost reduction needs because its use implies investment in a complex and precise mould, with the added costs of assembling, starting-up and disassembling from the machine each time the mould is used between batch productions.

The use of a spiral mould to determine apparent viscosity, Clavería et al. [29], allowed the calculation of apparent viscosity over a short range of shear rates; therefore, the viscosity model has to be used with rough extrapolations. In-line monitoring is impossible with this device. The goal of the methodology described in this work is its application to non-conventional features in the mould, such as textile injection, injection of recycled material without rheological information or asymmetric mould temperatures.

Several approaches for the in-line monitoring of viscosity under high shear rates have been investigated, Knappe [33], Friesenbichler et al. [34], Aho and Syrjälä [35], Gou et al. [36], Zhang and Gilchrist [37]. Of special interest is Friesenbichler et al.'s [34] injection mould because it directly and automatically allows the determination of viscosity curves, and the rheological model in an injection moulding machine under high shear rates.

The methodology and device presented in this study enable rheological characterisation using a slit-die rheometer in an injection moulding machine. This capability is achieved because the apparatus with sensors is directly connected at the end of the barrel replacing the conventional nozzle. The goal of this device is to enable the apparent viscosity to be calculated for a wide range of shear rates (from 50 1/s to 10.000 1/s) using 2 sensors for both temperature and pressure measurements. Pressure measurements are used for rheological calculations based on the pressure drop between the positions of the sensors. Temperature measurements are only taken to provide control information for the flow conditions, and a wire potentiometer attached to the reciprocating screw is used to measure the flow rate. The differences between the results obtained with the capillary rheometer are explained by complex flow behaviour and the factors listed above, Hay and Mackay [32], Laun [38]. Differences between the results show that the calculated apparent viscosity is higher than the real viscosity, as expected; thus, the results are of greater reliability if applied to a robust process window design. This device can remain mounted on the machine, and can be used to determine the apparent viscosity, during injection with different moulds.