Continuous fiber-reinforced plastics (FRP) have excellent mechanical properties in relation to their low weight. For this reason, they are suitable as substitutes for metal components. For instance, the primary structure of the Boeing Dreamliner already consists of 50 wt% of FRP . The reduction of weight offers a high potential to save fuel and emissions during the life cycle and thus is able to compensate for the higher costs of the materials and processes compared with metals. Typical processes for the production of FRP structures based on thermoset plastics are resin transfer molding (RTM), autoclave, and compression prepreg molding. RTM allows for the production of components that require hardly any secondary finishing operations and has a cycle time of between 5 and 25 min . The limiting factor of the RTM process though is the preform. For ribbings, complex preforms as well as an exact positioning in the tool are required, which makes them costly . The production of components by autoclaves causes long cycle times and is expensive due to the manual positioning of the prepregs and the energy required by the autoclave . Compression prepreg molding for producing components without secondary finishing is not known. Contemporarily, the only process that is known to produce thermoset-based fiber-reinforced components with ribbing is the combination of compression prepreg molding and sheet molding compounds , . Here, the problem is the fact that the design freedom of the sheet molding compound structure is severely limited . A new way to produce FRP parts with a high degree of design freedom, for example, for the ribs, based on thermosets could be in-mold-forming with thermosets. This is possible with new fast-curing prepregs (also called snap-curing prepregs). At the starting point of the process, a compression prepreg is heated to forming temperature and inserted into the thermoset injection molding tool. Upon closure of the tool, the sheet is formed in a three-dimensional component. Before final curing, the formed prepreg can be overmolded with short- or long-fiber-reinforced thermosets in the same production cycle. By using optimized precut prepregs, net-shaped parts could be produced as well. The in-mold-forming (“one-shot” processing) of thermoplastics is already a state-of-the-art technology . The big advantage of fiber composites based on thermosets is their high temperature stability and chemical resistance compared with thermoplastic-based composites whereas featuring a lower price than comparable thermoplasts such as polyetheretherketone (PEEK).
Because of slightly different resin systems of master-forming and forming components, the necessary curing conditions can differ. Under the influence of heat, cross-link thermosets cure irreversibly. These irreversible cross-links are a result of primary valency bonding in the molecular structure. The degree of cure of the resin/hardening system depends on curing time and curing temperature, which influences both the viscosity and the hardening speed . Increasing temperatures accelerate the hardening speed and permit higher degrees of cure in shorter times. Curing also changes the viscosity during processing because of the growing molar mass . For a theoretical, completely cured resin, all bonds between the components of the resin are connected. The theoretical degree of cure of a resin is the number of existing cross-linking points in relation to the highest possible number. However, it must be mentioned that there is no method in practice to determine the quantity of the cross-linking points . Nevertheless, resin curing in an exothermic chemical reaction releases heat. In relation to the release of heat, the enthalpy of reaction increases. Because of that, the enthalpy of reaction can be used as a measure of the degree of cure , . The degree of cure influences part properties such as toughness, strength or glass transition temperature . Lower curing temperatures lead to lower degrees of cure because the movement of the molecular chains as well as the cross-linking is limited. The lower number of cross-linking points then leads to lower glass transition temperatures. Investigations by Ziaee and Palmese  for vinyl-ester resins showed that a higher curing temperature improves the elastic modulus, but reduces the strain at break. The effect of the degree of cure on the mechanical properties of polyester resin was investigated by Selden . An incomplete curing is always related to a lower tensile and bending strength, as well as a lower elastic and shear modulus. Other publications also show that the mechanical properties of dental composite materials can be influenced by the degree of conversion , , , . Kummer showed that epoxy-based heat-curing glues have different mechanical properties depending on the curing conditions . The interrelationship of injection molding parameters and mechanical behavior for phenolic novolac compounds was investigated by Höer . A higher shear speed during injection can affect the bending strength negatively. The backpressure, the temperature of the injection unit, and the nozzle diameter do not influence the mechanical properties within the standard deviation. Changes in the curing time interact with the dwell time of the mold and can change the mechanical properties. The bending strength can be improved by raising the mold temperature; however, at the same time, the tensile strength is decreased.
There are several texts that depict the relationship between the degree of cure and the mechanical properties of resins, but there is no comparable analysis of the quantitative effects of epoxy-based injection molding compounds on these properties. To develop a combination of forming and master-forming of thermoset parts, it is essential to understand the relationship between the degree of cure, mechanical properties, and process parameters such as curing time and temperature. In thermoset-in-mold-forming, both components have to be processed with equal curing times and temperatures. Depending on the materials used, this might require some deviations from the optimal processing conditions and thus compromises that might affect the individual properties of the components. Thus, the goal of this work was to find out if it is possible to match the process windows by variation of the curing time and curing temperature. To determine the process window, the degree of cure and resulting mechanical properties were investigated.
2 Materials and methods
For the thermoset injection molding, a glass fiber-reinforced epoxy molding compound of the type Epoxidur EP 3585 T (Raschig GmbH; Ludwigshafen, Germany) was used. The main reasons that this molding compound was chosen were its high impact strength and temperature resistance. Both properties are very important for applications in aerospace or automotive parts. Following the manufacturer’s specifications, the curing time per millimeter was between 5 and 20 s. For the forming component, a continuous fiber-reinforced and preimpregnated sheet of the type Hexply M77/42%/200T2/CHS-3K (Hexcel Corporation; Stamford, USA) was used. This prepregs’ adhesion should be compatible to the molding compound and is one of the prepregs with the shortest curing time. The prepreg is epoxy based and is reinforced with a balanced twill weave fabric with a weight of 200 g/m2. The nominal resin content is 42±3 wt%. Both materials are commercially available.
2.2 Measuring mixer
Tests for the flow and cure relation were made by a measuring mixer of the type 30 (Brabender GmbH & Co. KG; Duisburg, Germany). All tests were carried out with respect to DIN 53764 for pourable thermoset resins. For each parameter setting, three measures were performed.
2.3 Injection molding
For the thermoset injection molding, a regular thermoset injection molding machine of the type KM 80 CX SP 180 (Kraus Maffei; Munich, Germany) was used to produce tension bars according to DIN EN ISO 10724-1. The temperature of the injection unit was 65°C in the feed section and 85°C in the injector section. The injection pressure was 15 bar with an injection speed of 10 mm/s and the holding pressure was 300 bar. The parameters were chosen according to the manufacturer’s recommendations. Mold temperatures were 150°C, 160°C, 170°C, and 190°C for a constant time of 90 s. Process time (curing time) was investigated for 30, 90, 180, and 600 s and a constant temperature of 160°C. According to the manufacturer, the optimum curing time for the compound is between 30 and 90 s with a curing temperature between 150°C and 190°C . After ejection, the tension bars were cooled in air at room temperature for 10 min and then stored in a freezer to prevent postcuring.
2.4 Prepreg compression molding
To vary the degree of cure of the prepregs, a static press APU 100 HPV (Blue Tiger Systems GmbH; Göttingen, Germany) was used. The executed experiments took place in isobaric process control. For the curing process, a single layer of the prepreg was inserted in the preheated cavity and held under pressure for a defined time. The curing parameters used are the same as those described previously for the injection molding. Subsequently, the prepreg was actively cooled under constant pressure at room temperature to reduce the effect of postcuring. In this procedure, the process time corresponds with the heating time of the prepreg in the press. A constant press force of 40 kN was used, which results in a pressure of approximately 10 bar considering the prepreg size of 200 mm×200 mm. According to the manufacturer, the shortest curing time is 90 s at a temperature of 160°C with a pressure over 5 bar .
2.5 Differential scanning calorimetry
Before conducting the mechanical tests, the degree of curing was determined by differential scanning calorimetry. The calculated degree of cure is the difference between the enthalpy of an uncured granulate (reference) and a particular sample. The samples were heated up linearly, from 0°C to 250°C, with a heating rate of 20 K/min. The investigations were carried out with a differential scanning calorimetry device of the type Q 2000 (TA Instruments, New Castle, USA). The implementation was based on the norm DIN EN ISO 1137 reference range.
2.6 Tension tests
The master-forming component was tested according to DIN ISO 527-4 with a type 1B specimen. All experiments were performed at room temperature (23°C) and at a humidity of 50% (DIN EN ISO 291).
The tension tests of the continuous fiber-reinforced component were carried out according to type 3 of DIN ISO 527-4 for a single layer. From the prepregs cured with different settings, rectangular tension bars were cut at dimensions of 180 mm length and 25 mm width. For the test, end tabs with dimensions of 50 mm height and 25 mm width were used. Additionally, 15° phases on the tabs helped to improve the force transmission , . Glass fiber laminates with fibers oriented at ±45° to the specimen axis were used to construct the end tabs. For fixing the tabs, a hot glue gun with an ethylene vinyl acetate-based glue was used.
The orientation and distribution of fibers in the injection molded tension bars were locally analyzed by an optical microscope Axiophot (Carl Zeiss AG; Oberkochen, Germany). The positions of the chosen section in the tension bar are shown in Figure 1. The section was cut with a water-cooled saw and embedded in cold-curing epoxy resin. Finally, it was sanded down to the middle of the sample and polished. For the purpose of improved clarity, in this work, only the lower half of the section is considered. For the investigation of the prepreg surfaces, images from the middle of the tension bars were taken as well.
3 Results and discussion
In Table 1, the results of the tests in the measure mixture are shown. Compared with 150°C, the required torque decreases the time in half to its minimum for a mold temperature of 190°C. The torque minimum is also approximately 79% lower at 190°C, which indicates a lower viscosity of the melt prior to curing after 18 s. The increase in curing speed is caused by the increasing movement of the molecular chains for higher temperatures .
3.1 Injection molding
Figure 2 (left) shows the degrees of cure achieved for the injection molded specimens depending of the mold temperature (curing temperature) with a constant process time (curing time) of 90 s. At 150°C, the master-forming component is only at a curing degree of 86%. Between 160°C and 190°C, the degree of cure remains constant at nearly 100%.
Figure 2 (right) displays the influence of the mold temperature on the tensile strength and elastic modulus of the injection molded specimens. Over all investigated temperatures, the tensile strength stays at nearly the same level within the standard deviation. Notably, the standard deviation becomes bigger for the higher temperatures of 170°C and 190°C. Considering Figure 2 (left), the degree of cure can be excluded as a reason for the higher standard deviation. The higher standard deviation might be explained by the orientation of the fibers due to the injection process. Göschel et al.  explain that the fiber orientation can be influenced by turbulences in the flow of the melt. With higher temperatures, the lower viscosity (also compare Table 1) of the melt can increase the turbulence of flow and thus the orientation of the fibers becomes more heterogeneous. Because of that, the tensile strength at higher temperatures also becomes more heterogeneous, which results in a bigger standard deviation. This thesis is also supported by the microscopy images shown in Figure 3 (right) as discussed later.
The elastic modulus of the master-forming component is also shown in Figure 2 (right). The elastic modulus is at the same level within the standard deviation for temperatures between 160°C and 190°C. Only for 150°C, the elastic modulus is elevated, which differs from the findings of other researchers. According to Ehrenstein and Bittmann , the elastic modulus should increase with an increasing degree of cure. Probably, the decrease of the elastic modulus from 150°C to 160°C (seen in Figure 2, right) is not linked to the rather small change of the degree of cure (14%; Figure 2, left) but caused by differences in fiber orientation originating in melt flow differences. Concerning the fiber orientation of thermosets, Englich  and Goettler  found that the surface layer thickness increases with an increasing mold temperature and that this results in better mechanical properties. Figure 3 shows that at 190°C, the surface layer is not thicker than at 150°C. However, it is obvious that the fiber orientation in the core for 190°C is more anisotropic. This is in accordance with the observations of Höer  for phenolic novoloac compounds. For increasing mold temperatures or rather lower viscosities, the reinforcing materials can orientate more anisotropically due to the plug-flow of the melt. The results of the measurement mixer in Table 1 confirmed a lower viscosity in the form of a smaller torque for 190°C compared with 150°C. A possible relation between mold temperature and shear rate has to be investigated in future research.
Besides the fibers for reinforcement, thermoset compounds for injection molding contain a vast number of different fillers that could also affect the properties. The composition thereby is confidential and has not been released by the manufacturer. The angular white filler in Figure 3 could be wollastonite or talcum. Both minerals are often used as fillers for thermosets and both do not influence results when used with temperatures under 200°C , . The circular bodies in Figure 3 could be a thermoplastic modificator to improve the impact strength. Mold temperatures above 150°C could affect the thermoplastic modificator and thereby the elastic modulus.
Figure 4 (left) shows the influence of curing time on the degrees of cure of the injection molded specimens for a constant mold temperature of 160°C. With at least 98% degrees of cure, the component is almost completely cured for all investigated times. This corresponds to the specifications of the material’s manufacturer.
Despite a constant degree of cure, the tensile strength and elastic modulus of the master-forming component tend to decrease with increasing process times, as shown in Figure 4 (right). Thus, it might be supposed that the degree of cure cannot be the reason for the change in properties. In addition, the fiber orientation cannot change with different curing times. The reason might be a modifier (compare Figure 3) that thermally degrades with the advancing temperature stress. Overall, the decrease of mechanical properties due to thermal degradation is less than 10%.
3.2 Prepreg compression molding
The degree of cure of the prepreg (shown in Figure 5, left) tends to increase with increasing temperature. Nevertheless, a degree of cure of 100% could not be reached in 90 s curing time using the tested temperatures.
Figure 5 (right) displays the mechanical properties of the prepreg in relation to the mold temperature. The scales differ from the scales of the master-forming component because the prepregs have much better mechanical properties than short-fiber-reinforced materials. The tensile strength increases steadily from 150°C to 170°C. For 190°C, the average of the tensile strength decreases and the standard deviation becomes bigger. The increase of the tensile strength up to 170°C mold temperature is in correlation with the increasing degree of cure shown in Figure 5 (left), which indicates the increasing cross-linking density of the molecules. Despite a higher degree of cure than at 170°C, the tensile strength is lower for 190°C. This could be attributed to degradation reactions in the resin system because the maximum mold temperature is specified by the manufacturer as 160°C. Taking Figure 6 into consideration, the surface of the specimens is bad regardless of the mold temperature. This means that for the temperatures less than 190°C, the degree of cure is too low to obtain a good consolidation in 90 s. However, an adequate consolidation of the prepregs cannot be reached by the predefined time of 90 s.
Within the standard deviation, there is no difference in the elastic modulus between 150°C and 170°C. At 190°C, the elastic modulus decreases strongly. This might be explained by the same reasons that were described previously for tensile strength.
Figure 7 shows the results of the variation in the curing time with a constant mold temperature of 160°C. In contrast to the master-forming component, there are improved mechanical properties directly related to the increasing degree of cure as shown in Figure 7 (left and right). The degrees of cure of the prepregs are generally lower as the degrees of cure of the injection molded specimens. With increasing time, the degree of cure increases to 83% after 90 s. According to the manufacturer’s data sheet, the material should be completely cured after 90 s. For further investigations, the process time was thus increased progressively until the prepreg was almost completely cured. The minimum time required was 600 s, which is also shown in Figure 7.
In addition, Figure 8 shows that the surface of the prepreg is improved with a longer process time. Therefore, the greater effect of the degree of cure is linked here with the consolidation. The low degree of cure is due to the low cross-linking rate between the molecules, resulting in a matrix that was dimensionally unstable. When the press opens, the matrix can flow or deform, which results in holes on the surface. With an increasing degree of cure, the viscosity of the matrix becomes higher and more stable while there is still tool contact. A smooth surface and good consolidation of the prepreg is only possible when the material is nearly completely cured before ejection.
The influence of the mold temperature (curing temperature) was investigated for mold temperatures of 150°C–190°C at constant process times (curing time) of 90 s. The master-forming component was almost completely cured for all settings and no influence of the degree of cure on mechanical properties could be determined. The biggest change of the mechanical properties was traced back to the fiber orientation, which depends on the viscosity during the mold filling process. For a lower viscosity, the fiber orientations in the core became more anisotropic and thus influenced the elastic modulus negatively. In future research, a quantitative investigation should be done to see if the shear rate is dependent on the mold temperature. The prepreg had a maximum degree of cure of 95% in this temperature range. Because of the incomplete curing, it did not reach the aspired mechanical properties and surfaces.
An increase of the process time (30–600 s) for a constant mold temperature of 160°C showed a negative trend of the mechanical properties for the master-forming component. This could be caused by degradations in the resin system. The influence of the degree of cure could be excluded because it was practically completely cured for all settings. Because the pure resin of the master-forming component is not available and some of the fillers are the manufacturers’ secret, a complete separation of the effects and relations among them is not possible. The mechanical properties of the forming component increased in relation to degree of cure and time. The full curing needed 600 s and thus was significantly longer than the time given by the manufacturer. It was presented that a complete curing of the prepreg is required for a satisfactory surface quality.
For the combination of master-forming and forming of thermoset-based materials, the prepreg is the limiting factor for the choice of the mold temperature as well as for the process time. The results showed that, based on the process parameters, a combination of master-forming and forming for thermosets is generally possible. However, it was also shown that a complete curing of the prepreg is required for the combination because incompletely cured prepregs had deficient surfaces and worse mechanical properties. However, for the implementation of the process in a series production, 600 s is too long. In addition, the steady time of 600 s makes the master-forming component cure in the molding unit. Thus, it is essential to continue in the following investigations with a faster curing prepreg. Newer tests showed that another commercial prepreg, which cured in approximately 240 s, is available. To overcome the limitations from commercially available master-forming compounds, we should produce our own master-forming compounds. Figure 9 shows the first combination of both components by thermoset in-mold-forming. Based on the new material, tests on hybrid parts and processes are following soon.
This work was carried out as part of the research project “Duro-IMF” (funding no. 20W1503D) funded by the German Federal Ministry for Economic Affairs and Energy according to a decision of the German Federal Parliament. The “Duro-IMF” process was developed together with Schmidt WFT GmbH, C.K. Siebenwurst GmbH & Co. KG, HBW-Gubesch Thermoforming GmbH, and the German Aerospace Center (DLR).
Hale J. Boeing 787 from the Ground Up, Aero, Boeing, 2006. Google Scholar
Neitzel M, Mitschang P. Handbuch Verbundwerkstoffe: Werkstoffe, Verarbeitung, Anwendung, Carl Hanser Verlag, 2004. Google Scholar
Lerch C, Thielmann A, Seelinger W. Wertschöpfungspotenziale im Leichtbau, Leichtbau BW and Frauenhofer-Institut, 2014. Google Scholar
Malnati P. Prepreg compression molding makes its commercial debut, CompositesWorld, 2015. Google Scholar
Akiyama K. Development of PCM* technology, Society of Plastics Engineers, 11zj-Annual Automotive Composites Conference and Exhibition, ACCE 2011. Google Scholar
Schemme M. Langfaserverstärkte Thermoplaste – Status und Perspektiven. Moulding in Aerospace Industry, Procedia Eng. 2014, 81. Google Scholar
Müller T, Al-Sheyyab A, Schmachtenberg E. In-mould-forming (IMF): hybride Hochleistungsverbunde großserientauglich fertigen, in 10. Internationale AVK-Tagung: Stuttgart, 2007, p 139–148. Google Scholar
Endruweit A. Investigation of the Influence of Local Inhomogeneities in the Textile Permeability on the Resin Flow in Liquid Composites Moulding Processes, Swiss Federal Institute of Technology, Zurich, 2003. Google Scholar
Ehrenstein GW, Bittmann E. Duroplaste, Hanser Verlag: Munich, 1997, p 2–17. Google Scholar
Wacker M, Trawiel P, Ehrenstein GW. Härtung von Reaktionsharzen Time-Temperature-Transition-Diagramm, Sonderdrucke Lehrstuhl für Kunstatofftechnik Friedrich-Alexander Universität Erlangen-Nürnberg, 2003. Google Scholar
Schürmann H. Konstruieren mit Faser-Kunststoff-Verbunden, Springer Verlag: Berlin Heidelberg, 2007. Google Scholar
Selden PH. Glasfaserverstärkte Kunststoffe, Springer Verlag: Berlin Heidelberg, 1967. Google Scholar
Freiberg R, Ferracane J. Am. J. Dent. 1998, 11, 214–218. Google Scholar
N, N, Aushärtung von Epoxidharzen und Glasübergangstemperatur (Tg), EPO-Tek PI Kummer Semiconductor Technology. Available at: http://www.jpkummer.com/sites/default/files/Aush%C3%A4rtung%20von%20Epoxidharzen%20und%20Glas%C3%BCbergangstemperatur%20(Tg).pdf. Google Scholar
Höer M. Einfluss der Material- und Verarbeitungseigenschaften von Phenolharzformmassen auf die Qualität spritzgegossener Bauteile, Technische Universität Chemnitz, 2014. Google Scholar
N, N, Processing recommendations, Raschig, 2016. Google Scholar
Hexcel. HexPly® M77 80°C – 160°C curing epoxy matrix Product Data. 2014. Available at: http://www.hexcel.com/Resources/DataSheets/Prepreg-Data-Sheets/M77_eu.pdf. Google Scholar
Englich S. Strukturbildung bei der Verarbeitung von glasfasergefüllten Penolformaldehydharzformmassen, TU Chemnitz, Münster, 2015. Google Scholar
Göschel U, Müller U, Lutz W, Eyerer P. Optimierte Kunststoff-Bauteile durch komplexe Analyse von Werkstoff, Verarbeitung und Konstruktion auf Basis grundlegenden Strukturbetrachtungen, 18. Stuttgarter Kunststoff-Kolloquium, 2003. Google Scholar
Blumenstock T. Analyse der Eigenspannungen während der Aushärtung von Epoxidharzmassen, Universtät Stuttgart, 2002. Google Scholar
Goettler LA. Mod. Plast. 1970, 1, 319. Google Scholar
Matthes S. Mineralogie, Springer Verlag: Berlin Heidelberg, 1990. Google Scholar
Gysau D. Füllstoffe: Grundlagen und Anwendungen, Vicentz Network, Hannover, 2006. Google Scholar
About the article
Published Online: 2017-03-17
Published in Print: 2018-01-26