Prof. Dr. Christian Hopmann, Daniel Schneider
Integral or sandwich components are used in multiple applications. For example, steering wheels and headrests in vehicles or highly flexible shoe soles with excellent damping properties are produced of flexible polyurethane (PU) integral foams [1, 2]. Rigid integral foams are used as casings for medical equipment or as foot prostheses in orthopaedic applications [1, 3]. Hereby, advantages are their good mechanical properties at a low weight and the high abrasion resistance. Other applications include armrests for aircrafts, technical enclosures or collection boxes in the solar thermal energy sector [2, 3, 4, 5].
For sandwich components the top-layer and the core material can also consist of different materials. Thus, a composite of metallic or thermoplastic liners with an insulating rigid foam core are used  in insulation elements for the automotive industry, for the cladding of industrial buildings or manufacturing for the cooling equipment. In supporting structural sandwich components a hard foam core is supplemented with two fibre-reinforced top-layers. Here, mainly thermoplastics, such as polyamide, or thermosetting epoxy resins are used as matrix materials [3, 6]. However, the use of polyurethanes as a matrix material is also possible.
In summary, both component classes (integral and sandwich panels) share different advantages, such as:
• Good mechanical properties at low weight
• Adjustable stiffness
• High abrasion resistance
• Good chemical resistance
• Low thermal conductivity
However, disadvantages can still be found in the production processes of sandwich or integral components. High demands regarding the surface quality of the components lead to the use of high quality polyurethane system. This results in high raw material costs and consequently high part costs of the integral components, because the entire part is made of one material. On the other side, sandwich components are produced in multi-step processes leading to high waste-rates and thus high manufacturing costs. These disadvantages can be overcome by a single-step process in which the top-layer material and the core material are injected successively into the cavity. In such a process different PU systems can be used, so that the core material can be a low-cost system and the top-layer material is fulfilling the high requested quality standards. Furthermore, the use of semi-finished textiles as reinforcing top-layers to improve the mechanical properties is possible.
Objective and approach
The objective of the IGF research project “Sandwich-RIM (Reaction Injection Moulding)”, conducted at the Institute of Plastics Processing at RWTH Aachen in industry and crafts (IKV), is to develop a process for the one-step production of sandwich components with a foamed core and compact top-layers made of polyurethane. In a following test series the integration of a semi-finished textile (e.g. as ± 45°glass-fibre textiles) in the top-layers will be analysed. The impregnation of the reinforcing fibres could be achieved by the foaming pressure of the core material.
To produce the components in the Sandwich-RIM process, first the top-layer and afterwards the core material are injected in a cavity (Figure 1). The injection is done vertically, so that the top-layer material leads to a partial filling of the mould. With the injection of the core material, the top-layer material is distributed uniformly alongside the cavity walls until the cavity is completely filled. Thereafter, the mould is moved into a horizontal position and the cavity volume is increased by using a punch to allow the foaming of the core material. The manufacturing process takes place in a mould that is adapted to the process requirements. These requirements include:
• Injection possibility of two different PU systems using a high pressure metering-unit
• Homogeneous distribution of the dosing volume over the width of the mould
• “Breathing” of the cavity
• Fixing possibilities for semi-finished textiles
• Production of different component thicknesses
Furthermore, the requirements of different sandwich components are decisive for the choice and the characteristics of the used polyurethane systems. Is a component with high desired tactile qualities; an elastic polyurethane top-layer material is selected. For particularly high demands regarding the mechanical properties, thermosetting top-layers and rigid foams as a core material are used.
Separated analysis of the process steps
In the fundamental studies, two aspects of the newly developed process chain were considered separately in the beginning. In a series of experiments analysis for the process control of the consecutive injection of different material systems were carried out in order to obtain knowledge of the flow conditions in the cavity. In a deviating experimental setup, the integration of textile semi-finished fibre products in the top-layers of the sandwich component and their impregnation by the foaming pressure of the core material was examined.
For the fundamental analysis of the flow conditions in a sequential injection of two different PU systems, as required by the developed process, an elastic material from Bayer MaterialScience (Covestro) was used as the top-layer material. The raw components are Baytech PU30BV08 (polyol) and Desmodur PU0309 (isocyanate). The core material is a flexible water-blown foam-system from Bayer MaterialScience with the raw components Bayfill UP.PU53FF40 (polyol) and Desmodur UP.PU58F07 (isocyanate). Particular attention was given to the pre-reaction time of the top-layer material and the thereby influenced flow characteristics of the two components, so that the sandwich-parts were produced in a conventional RIM tool that is not providing a controlled breathing of the cavity. The pre-reaction time (tpre) is defined by the period between the end of the injection of the top-layer material and the start of the injection of the core material. In Figure 2 sandwich-parts are presented in the order of increasing pre-reaction time. It turns out that, for a short pre-reaction time (tpre = 2 s) the viscosity of the top-layer material is too low, so that it flows back partially and resulting in an inhomogeneous layer thickness distribution. Thus, areas where there is only top-layer material at the sandwich-part edges occur. Selecting a long pre-reaction time (tpre = 18 s), the top-layer material is no longer able to form a uniform 2-component flow due to already high viscosity. In this case, it also results in areas where no core material is present. An optimally selected pre-reaction (tpre = 6-12 s) leads to a uniform distribution of the core material and a good forming of the desired sandwich structure throughout the whole part. Thus, the evidence can be provided that the developed process is capable of producing homogeneous formed sandwich structures. Driven by the results of these experiments, there is the need for a determination of the viscosity curves of both PU systems to obtain a deeper understanding of the processes during the injection. Also, this would allow the determination of process-relevant parameters in advance, so an elaborate determination of these parameters when changing material combination can be omitted.
In a second series of experiments, the analysis of impregnating a semi-finished textile in the top-layers was done. Due to the changed experimental conditions (test mould with a vertical flash face and rectangular cavity and no option for mounting a mixing head), PU-systems were used which allow for a manual injection. The top-layer material was a PU-system for RIM applications from Rühl Puromer GmbH with the raw components Purorim 185IT (polyol) and Puronate was 900 (isocyanate). The core material was a rigid foam system, also from Rühl Puromer GmbH, with the raw components Purotherm 448LF (polyol) and Puronate 900 (isocyanate). A semi-finished glass-fibre textile was exemplary selected with a fibre orientation of ± 45°. Figure 3 shows microscopy images by a cross section of a produced sandwich-part. In the microscopic analysis, both a clear separation between the core and top-layer and the good impregnation quality of the semi-finished fibre textile (only a few small air bubbles in the top-layers and high penetration of the glass fibre filaments) can be shown. However, this quality could not be obtained over the entire part. Firstly, the required pre-reaction time and a constant injection speed could not be realised due to the manual injection. Thus, it is expected that further studies with the adapted test mould and a high-pressure metering of the PU systems leads to a significantly more homogeneous sandwich-part quality.
The mould concept for the Sandwich-RIM-process
The mould concept is designed to meet the needs identified at the project start and the results obtained from the fundamental studies. Figure 4 shows an overview of the developed mould concept. To realise the injection of different PU-systems two gates are provided each with a mounting option for a high-pressure injection head. In both gates the injection is made in a circular runner. Both runners lead to a coathanger geometry which distributes the injected volume homogeneously over the cavity width. The injection level is always in the middle of the cavity height, defined by the fixed upper half of the mould. So, a symmetric sandwich structure can be achieved. From studies of previous research projects , an ideal range of the viscosity ratio (kη) of two different materials for a uniform layer formation is known. A closer look at the influence of the viscosity conditions takes place in the next chapter.
By means of a wedge mechanism the cavity volume space can be increased after reaching the horizontal position (“breathing” of the cavity) to provide the foaming of the core material. Adapted to the reaction kinematics of the selected foam systems different opening speeds can be realised. Furthermore, a direct influence on the foam density and thus the component thickness is possible. For a constant injection volume and at a constant top-layer thickness, parts with different core densities can be produced.
To investigate the impregnation of semi-finished textiles in the top-layer by the foaming pressure, magnets are placed both in the bottom and in the upper cavity wall, which serve to fix the semi-finished textiles. Before the process starts, semi-finished textiles are placed in the open mould and fixed to the cavity walls by means of thin metal plates in the area of the bar magnets. Thus, the textiles remain in the top-layers at the end of the manufacturing process. In the absence of fixation instead, the semi-finished textiles can slip during the injection process or during the breathing of the cavity and contribute no longer to the desired increase in stiffness of the top-layers. The metallic plates for fixation remain in the sandwich-part.
Additionally, in order to achieve various, especially higher, part thicknesses (> 20 mm), while maintaining an injection of the core material in the middle of the cavity height, mould inserts for the upper mould half are manufactured with different cavity heights, as shown in Figure 5. This ensures that when the process starts, the injection of the foam material always takes place in the middle of cavity height and the skin material is distributed homogeneously on both sides of the cavity walls. On the side of the stamp the initial cavity height can be adjusted by a wedge mechanism to the desired component thickness. Thus, sandwich-parts with thicknesses of 4-80 mm can be produced.
Further, various sensors are integrated near and in the cavity surface of the upper mould half. A temperature sensor used for measuring the cavity-near mould temperature. Two pressure sensors (near the gate and far from the gate) are used to check the pressure ratios at two locations. Thus, it can be examined by means of these two measuring points, if an uneven distribution of both material systems after injection (see Fig. 2) leads to a pressure variation in the foaming process. If so, a process integrated quality assurance can already be assessed. Furthermore, can be checked, which influence the foaming pressure has on the impregnation of the semi-finished textiles.
Development of a reaction-viscometer for material characterisations
As already explained, the viscosity ratio kη of the two material systems in the injection process is critical to the forming of the sandwich structure. Here, values of the viscosity ratio between 0.5 < kη < 5 be sought  in order to achieve the formation of a 2-component flow. To determine the viscosity ratio in the injection of the two different PU systems, knowledge of the rheological properties of the two materials is necessary. The rheological investigation of compact reactive systems is carried out for example by means of a parallel-plate rheometer. The analysis of foaming systems, however, is not possible with the current methods for compact systems. The foaming system flows down from the measuring volume in the course of measurement, making any further inspection impossible. Therefore, within the framework of this research project, a reaction viscometer was developed (Figure 6). This is based on an electric mixer, driven by a stepping motor. The torque of the mixer is the measure unit to determine the viscosity of the sample material. Further characteristics of the measuring cell are the possibility of injecting the sample material by means of conventional high-pressure and low-pressure metering-unit, but also a manual addition of the reactants is possible (for compact and foaming reactive polymer materials). The adjustable control of the stepping motor enables a temporal influencing of the mixing head, so that using the manual injection a high speed can be pre-set for the mixing at the start, which is then reduced during the course of the measurement. Thus, the foam formation is affected as little as possible. Furthermore, a temperature control and pressurisation of the cell is possible, in order to consider the process conditions.
Figure 7 shows a viscosity curve measured by the reaction viscometer in comparison with the result of an analysis by means of a parallel-plate rheometer. With the same experimental conditions (compact polyurethane system; component temperatures and measurement cell temperature) there are significant divergences. Due to the non-adiabatic measurement process and the energy introduced by the mixing head, the viscosity increase proceeds much faster using the reaction viscometer. For these reasons, the measured quantity is called apparent viscosity (ηsch) and serves mainly as a qualitative benchmark for different PU systems (compact and foaming) to provide a first indication of the appropriate selection of the process parameters (pre-reaction time, mould and component temperatures).
Thus, the developed reaction viscometer offers the possibility of a rapid quantitative analysis of the rheological properties of compact and foaming PU systems under processing conditions. Based on the results from the rheological analysis, conclusions about the viscosity ratio during the injection of the foaming PU system can be drawn. Thus, the results obtain support in the process parameter determination and analysis of the effects observed during the component formation.
In summary, the first fundamental investigations show the high potential of the developed process. It could be shown both the appropriate process control of the formation of a 2-component flow can be realised, which is decisive for the homogenous thickness of the single sandwich-layer, as well as an impregnation of the semi-finished textiles through the top-layer material is possible. Additionally, suitable process parameters can be identified fast and statements about the occurring component defects are taken by the newly developed reaction viscometer.
In the current systematic studies, the impact made on the component forming process will be analysed using the adapted test mould. After these series of experiments, the transfer to a different material combination (e.g. a rigid foam core with rigid top-layer) will be done in order to use the technology in other fields of application, as well as the transfer to sandwich components with higher component thickness. In this case, both the influence of the layer thickness of the core and the top-layer will be considered. At the end of the systematic studies a detailed feasibility study for the impregnation of different semi-finished textiles in the top-layer will be done.
The research project IGF 18046 N of the Forschungsvereinigung Kunststoffverarbeitung was sponsored as part of the “Industrielle Gemeinschaftsforschung und -entwicklung (IGF)” by the German Bundesministerium für Wirtschaft und Technologie (BMWi) due to an enactment of the German Bundestag through the AiF. We would like to extend our thanks to all organizations mentioned.
 Gansen, P.; Richter, H.-J.; Hagel, G.: Emissionsarmer, Kohlendioxid-getriebener PUR-Integralschaum für Automobilanwendungen. FAPU 20 (2003) 5, S. 34-39
 Leppkes, R.: Polyurethane – Werkstoff mit vielen Gesichtern. Landsberg/Lech: Verlag Moderne Industrie, 2003
 Liman, U.: Polyurethane (PUR) – Maßgeschneiderte Alleskönner für zahlreiche innovative Anwendungen. Kunststoffe 10/2014, S. 108-117
 Becker, K.; Gansen, P.; Hagel, G.: Anwendungen von Polyurethanen in der Orthopädie. FAPU 48 (2008) 3, S. 34-3
 Ruthmann, H.: PUR-Anwendungen im Bereich der Solarthermie. PUR 2009. Nürnberg, 2009
 Horn, K.; Schneiders, F.; Pophusen, D.: Polycarbonat (PC) – Wachstumschancen mit innovativer Beleuchtung und leichten, hochwertigen Autos. Kunststoffe 10/2014, S. 78–84
 Hopmann, Ch.: Reaktives Mehrkomponenten-Spritzgießen von flächigen Bauteilen mit einer Polyurethan-Außenhaut. Institut für Kunststoffverarbeitung, RWTH Aachen, Abschlussbericht zum IGF-Forschungsvorhaben 16396N, 2012
Institute for Plastics Processing (IKV) in Industrie und Handwerk der RWTH-Aachen, Pontstraße 49, 52062 Aachen, Germany
Prof. Dr.-Ing. Christian Hopmann
Director of the Institute of Plastics Processing
Daniel Schneider, M.Sc.
Leader Working Group Polyurethane Technology / Foaming
Phone: 0241/80 93673