Analysis of the Welding Strength in Hybrid Polypropylene Composites as a Function of the Forming and Overmolding Parameters.
Hybrid thermoplastic composites (HTPCs) are increasingly used to satisfy the lightweight requirements of the automotive industry. Reducing vehicle weight to decrease energy consumption and C[O.sub.2] emission is one of the most important goals required by the regional regulation of transport [1-4], End of life considerations promote thermoplastic over thermoset composites, because they are suitable to be recycled  and they could be easily combined with other materials by means of different joining mechanisms . HTPC are widespread in automotive industry also due to their higher manufacturability. These characteristics make such materials good candidates for metal replacement.
In the last decade, several technologies that incorporate the injection molding process were proposed to manufacture hybrid parts. In fact, it is suitable for high volume production and enables the implementation of in-mold assembly or joining [7, 8]. Very recently, the in-mold forming technology--also known with the commercial names of FiberForm, Organomelt, and Spriform--was proposed for the manufacturing of HTPC. In this process, an outer thermoformed shell, reinforced with continuous fibers, is internally stiffened with an overmolded ribs system [9-12]. This technology allows directly coupling of thermoforming and injection molding. The adhesion between parts is ensured by interdiffusion of the macromolecules across the interface. The welding is based on the reptation theory and it is promoted by high interface temperature during processing. For semi-crystalline polymers, the degree of adhesion is related to the residence time of the interface above the crystallization temperature [13-18], and good adhesion is favored by the compatibility of the matrices.
As in the overmolding phase, the interdiffusion window is short and strongly nonisothermal, the maximum adhesion strength is usually lower than the theoretical maximum, that is, the matrix strength [19, 20]. Moreover, for good-quality welding, the weakest zone becomes the interface between the matrix and the glass-woven reinforcement inside the laminate, which is likely to yield a delamination failure [21-23].
This work analyses how the main injection molding process parameters are related to welding strength in a hybrid T-joint specimen made of a polypropylene-based HTPC.
The effects of melt temperature, holding pressure, and mold temperature (controlled by a variothermal system) on welding strength were investigated in relation to the characteristics of the interface, such as residual stress and topography. The resistance of the welding was determined through tensile tests and the ultimate tensile strength (UTS) was assumed as the response variable of the experimental plan. Injection molding process simulations were coupled with structural analyses to investigate the residual stress induced by the process.
The conditions of the interface were analyzed varying the following strategies for the pre-heating of the laminate: (a) without preheating; (b) selective heating and cooling of the welding area, (c) selective heating without cooling, and (d) preheating of the whole laminate.
MATERIALS AND METHODS
The hybrid T-joint specimen includes a 2-mm-thick composite laminate base and an overmolded stem, which creates an interface area of 4 mm X 20 mm. The geometry is represented in Fig. 1.
The base is a rectangular cutout from a laminate made of polypropylene reinforced with 50 wt% glass fibers with balanced woven fabrics (Bond Laminates, Tepex Dynalite 104RG600(4)). The stem is made of polypropylene reinforced with 30 wt% long glass fibers (Celanese, Celstran PP-GF30-0304).
A two-cavity mold was designed to manufacture the T-joint samples. The welding occurs at the end of each cavity were the laminate bases are inserted in two transversal slots machined in the mold (Fig. 2). These are perpendicular to the parting plane to facilitate the positioning of the laminates after preheating. During the demolding phase, ejector pins push both the stems and the laminates avoiding shear loading of the welding area.
Three 8-mm-diameter heating/cooling channels were placed close to the interface area, providing better heat exchange efficiency. A variothermal control unit (Wittmann Battenfeld, TEMPRO plus D 180 Vario) was used for mold thermal control. All the specimens were manufactured on a 1,000 kN electrical injection molding machine (Engel, E-Motion 440/100).
The preheating system used in the experiments is mainly composed of an IR lamp (200 mm X 200 mm, by Krelus AG) installed at the top of a tower frame and directed to a lower grid tray, where the specimen bases were placed. A combination of power supply, residence time, and distance of the IR source from the specimen bases was set to reach a target temperature before the overmolding process. During the experiments, the temperature at the interface was checked using an IR pyrometer (Seitron, PorTerBir). This pyrometer was previously calibrated by comparing its measurements with a type-K thermocouple that was placed inside a test laminate and connected to a thermometer (Hasco, Z251/2).
Overmolding of the T-Joint Specimens
The injection molding experiments were set in accordance to a two-level full-factorial design, varying the following three factors as reported in Table 1: melt temperature ([T.sub.melt]), holding pressure ([P.sub.hold]). and mold temperature ([T.sub.mold]).
The mold temperature was controlled by setting a constant temperature of 80[degrees]C at the lower level and by using the variothermal system at the upper level. The variothermal cycle was set to heat the entire mold, including the transversal bar, up to 130[degrees]C before placing the specimen bases. The mold was then cooled down to 80[degrees]C after the packing phase, to allow a correct ejection of the specimens. Figure 3 shows the temperature variation at the interface between the overinjected stem and the laminate base during the variothermal cycle, as calculated by the numerical simulation described in the following section. In the first round of overmolding experiments, the specimen bases were used at room temperature (30[degrees]C), without preheating. Each molding experiment was replicated 7 times. All the other process parameters were kept constant, as reported in Table 2.
Numerical Simulation of the Overmolding Process
Each run of the experimental plan described in the previous section (Table 1) was numerically simulated using a thermo/ fluid dynamic code (Autodesk Moldflow Insight). A 3D cooling analysis was carried out considering both the variothermal cycle and the part-insert cooling after having placed the insert in the mold, after the preheating phase. The average volumetric shrinkage, that is, the average value of volumetric shrinkage over the half-gap thickness for 3D models, was calculated for each experimental run. The volumetric shrinkage is the percentage increase in local density from the end of the packing phase to when the part has cooled to the ambient reference temperature (default value of 25[degrees]C). Its calculations begin once the cavity is filled, based on the difference between the current state and the reference state, as described by the modified 2-domain Tait pressure-volume-temperature (pvT) model :
VS = [[rho].sub.av](t)/[rho]([T.sub.amb], [p.sub.atm]) (1)
where VS is the volumetric shrinkage, [[rho].sub.av](t) is the average density over the half-gap thickness, [rho]([T.sub.amb], [p.sub.atm]) is the density at ambient pressure and temperature.
As the mass of an element changes, shrinkage continues to vary according to the pvT-state of the element. Once the mass stabilizes, the current pvT-state of the element is fixed in the shrinkage calculation as the reference state. The mass of an element stops varying when the cavity pressure has decayed to zero and the volumetric shrinkage becomes a constant.
The residual stress distribution was calculated by transferring the volumetric shrinkage to a structural simulation (Autodesk Simulation Mechanical), through direct transfer of the nodal and elemental information. The contact between stem and laminate was set as bonded to guarantee continuity between the interfacial nodes. The process parameters effect on the residual stress was evaluated considering the average value of the Von Mises equivalent stress at the stem-side interface.
The handling of the specimen bases from the preheating system to the mold took 20 s. However, the temperature of the interface was measured after 30 s using an IR pyrometer (Seitron, PorTerBir), as it took 10 s to close the mold and fill the cavity after positioning the specimen bases. The procedure was repeated for each run of the experimental plan.
During the uniform preheating, the specimen bases were heated up to 220[degrees]C. After 30 s (i.e., at the end of the cavity filling), their surface temperature reached the value of 160[degrees]C, with variothermal control, and 150[degrees]C without variothermal control, respectively.
The effect of selective preheating before the injection phase was evaluated by using the steel screen shown in Figs. 4 and 5. The full factorial design described in Table 1 was conducted also to overmold the laminates that were selectively preheated. Each factorial run was repeated 7 times.
Using the screen, it was not possible to reach the same initial temperature of the uniform heating (i.e., 220[degrees]C). The laminate cutouts were uniformly heated up to 170[degrees]C. A further increase of the selective preheating time would extend the heated surface, due to conduction from the screen, damaging the laminates. The handling time was minimized to provide the overmolding process with cutouts that had surface temperature as close as possible to the values reached with the uniform heating. The laminates were at 142[degrees]C and 135[degrees]C at the beginning of the filling phase, respectively, with and without variothermal control.
In summary, a comparison was performed considering the laminates in four different conditions before the overmolding phase:
a. without preheating,
b. after selective heating and cooling of the welding area,
c. after selective heating but without cooling (135[degrees]C before the filling phase), and
d. preheating the whole laminate (150[degrees]C before the filling phase).
The injection molding parameters (i.e., melt temperature, holding pressure, and mold temperature) that maximized the welding resistance in the first investigation were used and kept constant for manufacturing the specimens. The resistance of the welding was determined through tensile tests and the UTS was assumed as response.
Moreover, the surface topography of the laminates was characterized for all the four preheating conditions, to evaluate the effect of the interface topography (laminate side) on the welding resistance. The surface topography of the interface was characterized by stopping the overmolding cycle before the injection phase and removing the laminate from the cavity.
Surface Characterization of the Interface
Additional injection molding tests were conducted, stopping the cycle before the injection phase and ejecting the specimen bases to allow the surface characterization of the interface on the laminate side, using a 3D optical profiler (Sensofar, Plu Neox, in confocal mode with 20 X magnitude lens). With this procedure, the effects of preheating and mold clamping on the surface topography of the specimen bases were analyzed. The point-cloud output provided by the profiler software was used for the reconstruction of the surface and the measurement of the area using a solid modeling CAD program (Dassault Systemes, SolidWorks).
Moreover, a computed tomography system (Nikon Metrology, X-Tek MCT225) was used to characterize the interface of 4 overmolded T-joint specimens, considering the effect of the different preheating conditions. The purpose of the investigation was the reconstruction of the interface borders between the laminate and the overmolded reinforced polymer, focusing on the orientation of the fibers and the local macroscopic mixing of the two matrices.
Tensile Testing of T-Joint
The tensile tests of the T-specimens were performed on a universal tensile testing machine MTS 322 with a load cell of 5 kN. The basis was not clamped but it was secured using a steel plate with a central rectangular hole having sides 1 mm longer than the sides of the interface area (Fig. 6). The lateral surfaces of the stem were free to slide without interference with the steel plate. This device allowed the stem clamping and prevented any kind of bulk loading of the sample during the closing of the clamps. Each test was displacement controlled with a rate of 2 mm/min and a frequency acquisition of 20 Hz.
RESULTS AND DISCUSSION
Effect of Process Parameters
The analysis of variance (ANOVA) results (Table 3) of the experimental plan conducted on the tensile tested specimens indicate that both the melt temperature ([T.sub.melt]) and the holding pressure ([P.sub.hold]) significantly affect the bonding resistance (Fig. 7). The main effect of the mold temperature ([T.sub.mold]) is not significant. However, an interpretation of the results based only on the main effects is not possible, as melt and mold temperature present a significant interaction (Fig. 8). When [T.sub.mold] is low, the different levels of [T.sub.melt] influence the UTS but when [T.sub.mold] is high, different levels of [T.sub.melt] have no significant influence on UTS. In other words, the levels of [T.sub.m]oid eliminate the contributions from [T.sub.melt]. Conversely, [T.sub.mold] has a significant effect only when [T.sub.melt] is high. As [P.sub.hold] is not involved in any interaction, its influence on UTS can be interpreted analyzing its main effect ,
To understand the complex dependence of welding strength on melt and mold temperature it is necessary to consider the residual stress at the interface induced by the shrinkage of the injected polymer.
The influence of the process parameters on the Von Mises equivalent stress on the stem-side interface was calculated by transferring the thermal results from a thermal/fluid dynamic simulation of the process to a structural simulation of the welding. With reference to the interaction plot of the previous section (Fig. 8), it was supposed that the negative effect of mold temperature on the UTS is related to a higher residual stress induced by a larger shrinkage of the overmolded polymer at the interface. This hypothesis is supported by the simulation results, which show a strong dependence of residual stress on the mold temperature (Fig. 9).
Effect of Selective Heating
The graph in Fig. 10 shows the ultimate tensile strength of the specimens that were selectively and uniformly heated. Considering the UTS of the best welding from the first DOE as reference value for welding strength (11.65 MPa), which is the last bar on the right, the selective heating of the interface is responsible of a reduction of 24% (8.86 MPa). The laminate uniformly heated and provided at 135[degrees]C leads a lower reduction of 7.3% (10.74 MPa).
Although the reduction in temperature leads to a modest reduction of the UTS, the localized heating results in a significant reduction of the mechanical strength at the interface. With a selective heating, only the top layers were softened while the rest of the base remains in the solid state.
The analysis of variance (ANOVA) results (Table 4) of the screening plan on the tensile tested samples show that the melting temperature and mold temperature significantly affect the welding resistance when the laminate is selectively heated. The main effects plot in Fig. 11 shows the same influence of the process parameter on the welding strength that was noticed with the laminate uniformly heated.
The main difference from the previous results on the full-thermoplastic hybrid specimens is the more significant negative effect of the mold temperature. Nevertheless, the negative effect of the mold temperature is consistent with the results of the process simulation that showed how a higher mold temperature causes a significant increase of the residual stress at the interface.
The melt temperature always contributes in increasing of the welding strength. Higher temperatures at the interface promote the wetting of the laminate surface and the macromolecules interdiffusion. In addition, the holding pressure positively contributes to bring the material at the interface in close contact but it provides a stronger contribute when the laminate is preheated. When the laminate is uniformly heated at a higher temperature, the overinjected material can penetrate the laminate and fill the void of matrix due to swelling of the first laminate. As a consequence of this phenomenon, the two polymers are provided with a rougher and extended interface where the macromolecules can interdiffuse. This observation is supported by the micro-CT scans shown in Fig. 12. In this case, the laminates were heated over the melting point and the solid matrix around the heated area could not block the flowing of the overinjected material.
Effect of the Surface Topography
The four surface topographies are shown in Fig. 13. The comparison of the welding performance of the four different specimens shows that a higher preheating temperature leads a progressive increase of the welding strength (Fig. 14).
The four preheating conditions of the laminate are shown in Fig. 15 with the respective surface topography characterization and the surface reconstruction starting from the point cloud, which was performed using a CAD (Dassault Systemes, SolidWorks).
The value of the area that were measured and the relative percentage increase, which is obtained by the comparison to the first condition (a), are summarized in Table 5 and are compared to the percentage increase of the welding strength.
The area of the interface was respectively 72.92 [mm.sup.2] for the as-provided laminate, 73.22 [mm.sup.2] for the laminate heated by using the mask and cooled down, 84.1 [mm.sup.2] for the laminate heated by using the mask and clamped in the cavity (T = 135[degrees]C) and 94.09 [mm.sup.2] for the laminate uniformly heated and clamped in the cavity (T = 150[degrees]C).
The investigation aimed at understanding how the process affects the surface topography on the laminate side before the filling phase and how this effect influences the welding strength. In the first and the second conditions, the laminates are positioned in the cavity at room temperature. The increase of the area in the second condition is due exclusively to the swelling of the matrix outside the woven glass fabric, induced by the heating cycle, and it provides an increase of the welding resistance of 26%. Therefore, the interdiffusion of the macromolecules takes place over an extended area. The same effects are present in the third and in the fourth conditions but in these cases, the increased area and the swelling of the woven fibers promote a mechanical interlocking. Moreover, when the laminate is heated at the highest temperature, the overinjected material can penetrate the laminate and replace the matrix. This interpretation is supported by the tomography results shown in Fig. 16.
The CT scans highlight the problem related to the structural integrity of a hybrid composite part during the closing of the mold. Even if the morphological modification promotes an increase of the welding resistance, it could act as a local defect. Being the matrix weaker and the layers subject to external forces in the direction parallel to the plane on which they are arranged, the layer are free to move one another and the fabric is able to deform. If this occurs, the mechanical properties of the laminate undergo a sharp deterioration because the stresses are no longer parallel to the fibers in the areas in which the laminate have been deformed. Therefore, the stresses are no longer supported exclusively by the reinforcement but they distribute within the matrix.
The error bars in the histograms of Fig. 14 highlight the variability of the UTS results. The fracture area was investigated to understand the variability of the results and to highlight the factors that affect the behavior of the welding during the test. Two different typology of fracture area are shown in Fig. 17. The interface in Fig. 17a results deprived of the matrix surface revealing a complete delamination of the first layer. Only a small portion of the weft yam results cut and peeled in the lower left-side corner. This portion is close to the warp-yams that stiff the system. On the right side of the picture, a portion on the warp-yam completely overlays the weft-yam, avoiding its breakage and peeling.
The contrary occurs in the interface presented in Fig. 17b. On the left side, the warp-yam was completely broken and peeled by the stem. The causes can be related both to the edge of the mold cavity, where the laminate is pushed during the mold closing, or directly due to the shear stress induced by the traction of the stem. Moreover, it can be noticed that on the right side, the weft-yam was not broken and peeled because of the vicinity with the warp-yams.
The overlapping of the stem with the laminate, and in particular with the woven fibers, highly affects the behavior of the interface during the tensile test and it is crucial for determining the welding resistance. Since the thickness of the stem is equal to the width of the yam, the interface can overlays a random portion of the laminate that can include areas where the yam surfaces and some other that are occupied by the matrix, which are located at the edge of the zones where the weft cross the warp. Figure 18 shows an example of 6 different overlapping cases. The high variability of ultimate tensile test can be explained by a gradual distribution of the mechanical strength between the fibers and the matrix.
With reference to Fig. 19, two typical areas can be distinguished. The zones marked in green represent the portion rich of matrix, where the injected stem can create a stronger welding. The red circles highlight the area where the weft-yam surfaces in correspondence of the shorter side of the interface. In this area, where the stress is higher and concentrated, the matrix in the welding is scarce of and the delamination is promoted.
Fiber bridging depends on how the overinjected material can create a physical interlocking with the fibers of the yam. Indeed it is promoted especially when the laminate is preheated up to a temperature above the crystallization point (both selective and uniform heating). In this case, the laminate can swell inside the cavity and the small undercuts, which develop inside the yams and between the fibers, are filled by the overinjected material. In addition, the interdiffusion between matrices is promoted because it takes place over a larger surface and in a volume that is poor of fiber.
This work investigated the effects of the in-mold-forming process on the welding strength between the laminate and the overmolded part of a full-thermoplastic hybrid composite. The effects of the main process parameters on welding strength were analyzed in accordance to the design of experiments approach and the surface topography at the interface was investigated and discussed. The results of the experimental plan showed that the melt temperature and the holding pressure contribute to increase the welding strength, because they promote the intimate contact between the two polymers and the interdiffusion of the macromolecules. The mold temperature induces a decrease of the welding strength due to higher residual stress at the interface, induced by the larger shrinkage of the overmolded polymer. The surface characterization of the interface topography (laminate side) allowed understanding that preheating the laminate induces a squeezing of the matrix during mold clamping that contributes to increase the interface area. In addition, the swelling of the molten matrix inside the laminate allows the injected material to wet and penetrate the fibers of the woven at the interface. Even if the swelling of the laminate inside the cavity contributes in increasing of the area of the interface, it could constitute a local defect that could affect the overall performance of the hybrid part. For this reason, a selective preheating of the laminate, which is less invasive than the uniform heating, could represent the best tradeoff between welding strength performance and robustness.
The authors gratefully thank Prof. Simone Carmignato and Dr Filippo Zanini for the p-CT scans of the welding.
[1.] Regulation (EC) No 443/2009 of the European Parliament and of the Council (2009).
[2.] U.S. Environmental Protection Agency Office of Transportation and Air Quality. EPA and NHTSA Set Standards to Reduce Greenhouse Gases and Improve Fuel Economy for Model Years 2017-2025 Cars and Light Trucks, Report EPA-420-F-12-051 (2012).
[3.] Ministry of Environmental Protection of the People's Republic of China. Limits and Measurement Methods for Emissions from Light-Duty Vehicles, (China 5) GB 18352.5-2013 (2013).
[4.] Japanese Ministry of Economy, Trade and Industry. Act on the Rational Use of Energy 1979 (rev. 2013).
[5.] Directive 2000/53/EC on End-of-Life Vehicles of the European Parliament and of the Council.
[6.] F. Awaya, M. Gilbert, G. Kelly, B. Fox, and J. Pigram, Adhes. Polym. Prog. Polym. Sci., 34, 948 (2012).
[7.] M. Grujicic, V. Sellappan, M.A. Omar, N. Seyr, A. Obieglo, M. Erdmann, and J. Holzleitner, J. Mater. Process. Technol., 197, 363 (2008).
[8.] J. Kragl, Plast. Technol., 57. 34 (2012).
[9.] A. C. Long, Composite Forming Technologies. Woodhead Publishing Limited, Cambridge, UK, 256 (2007).
[10.] T. Muller, T. Osswald, and I. Kuehnert, Annual Technical Conference --ANTEC, Conference Proceedings, 2, 1233 (2009).
[11.] J. Kragl, ACCE 2012: Unleashing the Power of Design, All (2012).
[12.] G. P. Holzinger, M. Wurtele, S. Schierl, and Kruppa, Proceedings of the Second International MERGE Technologies Conference, IMTC 2015 Lightweight Structures, 119 (2015).
[13.] R. P. Wool, Final Report Proposal No. 22966-MS, U.S. Army, 217, 222 (1990).
[14.] L.J. Bastien and J.W. Gillespie, Polym. Eng. Sci., 31, 1720 (1991).
[15.] C.J.G. Plummer, P.E. Bourban, J.E. Zanetto, G. Smith, and C.J. Manson, J. Appl. Polym. Sci., 87. 1267 (2003).
[16.] F. Yang and R. Pitchumani, Polym. Compos., 24, 263 (2003).
[17.] E. Lafranche, T. Renault, and P. Krawczak, Key. Eng. Mat., 611-612, 821 (2014).
[18.] V. Kagan, J. Mazza, I. Palley, SAE World Congress, Detroit, MI (2001).
[19.] M. Fiorotto and G. Lucchetta, AIP Conf. Proc., 1353, 797 (2011).
[20.] J. Aurrekoetxea, G. Castillo, F. Cortes, M.A. Sarrionandia, and I. Urrutibeascoa, J. Appl. Polym. Sci., 102, 261 (2006).
[21.] M. Fiorotto and G. Lucchetta, Int. J. Mater. Form., 6, 179 (2013).
[22.] R. Giusti and G. Lucchetta, Key. Eng. Mat., 611-612, 915 (2014).
[23.] E. Moritzer, C. Budde, T. Troster, and S. Pohler, Euro Hybrid Mater. Struct., 188 (2014).
[24.] P. S. Cook, Z. Zuo, and F. S. Costa, Annu. Tech. Conf. ANTEC, Conf. Proc., 2, 1199 (2016).
[25.] J. C. Reinard, Communication Research Statistics, Sage Publications, CA, 219 (2006).
Ruggero Giusti (iD), Giovanni Lucchetta
Department of Industrial Engineering, University of Padua, via Venezia 1, Padova 35131, Italy
Correspondence to: G. Lucchetta; e-mail: giovanni.Iucchetta@unipd.it
Caption: FIG. 1. Design of the T-joint specimen. All dimensions are in millimeters.
Caption: FIG. 2. Injection mold for the hybrid T-joint specimens. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 3. Mold temperature variation at the interface during the variothermal injection molding cycle. Colored line on top indicates the variothermal phases; the bottom one indicates the injection molding phases. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 4. Lateral view of the screen assembly.
Caption: FIG. 5. Screen for the selective heating of the interface. [Color figure can be viewed at wileyonlinelibrary.coml
Caption: FIG. 6. Clamping system for the base of the specimen that was used in the tensile test. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 7. Main effects plot for the process parameters on the UTS of the hybrid T-joint specimen. [Color figure can be viewed at wileyonlinelibrary. com]
Caption: FIG. 8. Interaction plot for mold temperature and melt temperature on the UTS of the hybrid T-joint specimen. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 9. Effects of the process parameters on the average Von Mises equivalent stress at the interface of the hybrid T-joint specimen, as calculated by the numerical simulations. [Color figure can be viewed at wileyonlinelibrary. com]
Caption: FIG. 10. Mean values of UTS of the specimens overmolded with a selective and a uniform heating of the laminate. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 11. Main effects plot for the process parameters on the UTS of the hybrid T-joint specimen with the selective heating of the laminate. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 12. Micro-CT scans at the interface that shows the penetration of the injected material inside the laminate. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 13. Interface surface topography on the laminate side before the filling phase: (a) as provided, without preheating, (b) after selective heating and cooling of the welding area, (c) after selective heating but without cooling (135[degrees]C before the filling phase), and (d) after uniform preheating of the whole laminate (150[degrees]C before the filling phase). [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 14. Mean values of the UTS of the specimens manufactured with the laminate provided with different preheating conditions. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 15. Reconstruction of the surface starting from the point clouds. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 16. Micro-CT scans of the interface. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 17. Delamination of the matrix and presence of fiber bridging.
Caption: FIG. 18. Examples of overlapping of the interface on the woven fabric. [Color figure can be viewed at wileyonlinelibrary.com]
Caption: FIG. 19. Morphological details of the surface of the fracture. [Color figure can be viewed at wileyonlinelibrary.com]
TABLE 1. Experimental design of the overmolding process parameters. Investigated parameters (DoE) Levels Low High Melt temperature ([degrees]C) 220 260 Holding pressure (bar) 100 200 Mold temperature ([degrees]C) 80 130/80 (Vario) TABLE 2. Constant injection molding process parameters. Injection molding parameters (constant) Injection flow ([cm.sup.3]/s) 63 Back pressure (bar) 30 Holding time (s) 35 Cooling time (total S) 65 Contact time before closure (S) 10 TABLE 3. Analysis of variance results for the process parameters on the UTS of the hybrid T-joint specimen. Factor Sum of square [T.sub.melt] 14.060 [T.sub.mold] 3.912 [P.sub.hold] 16.966 [T.sub.melt] x [T.sub.mold] 10.090 [T.sub.melt] x [T.sub.holdt] 1.213 [T.sub.mold] x [P.sub.hold] 2.392 [T.sub.melt] x [T.sub.mold] x [P.sub.hold] 2.643 Error 116.61 Factor Mean square [T.sub.melt] 14.060 [T.sub.mold] 3.912 [P.sub.hold] 16.966 [T.sub.melt] x [T.sub.mold] 10.090 [T.sub.melt] x [T.sub.holdt] 1.213 [T.sub.mold] x [P.sub.hold] 2.392 [T.sub.melt] x [T.sub.mold] x [P.sub.hold] 2.643 Error 2.429 Factor F value [T.sub.melt] 5.79 [T.sub.mold] 1.61 [P.sub.hold] 6.98 [T.sub.melt] x [T.sub.mold] 4.15 [T.sub.melt] x [T.sub.holdt] 0.50 [T.sub.mold] x [P.sub.hold] 0.98 [T.sub.melt] x [T.sub.mold] x [P.sub.hold] 1.09 Error Factor P value [T.sub.melt] 0.020 [T.sub.mold] 0.211 [P.sub.hold] 0.011 [T.sub.melt] x [T.sub.mold] 0.047 [T.sub.melt] x [T.sub.holdt] 0.483 [T.sub.mold] x [P.sub.hold] 0.326 0.302 [T.sub.melt] x [T.sub.mold] x [P.sub.hold] Error UTS = 10,10,046 + 0.501; [T.sub.melt], 0.264; [T.sub.mold] +, 0.550; [P.sub.hold], 0.424; [T.sub.melt] x [T.sub.mold], 0.147; [T.sub.melt] x [P.sub.hold], +0.207; [T.sub.mold] x [P.sub.hold], -0.217; [T.sub.melt] x [T.sub.mold] x [P.sub.hold]. TABLE 4. ANOVA for the influence of the process parameters on the UTS of the hybrid T-joint specimen with the selective heating of the laminate. Factor Sum of square Mean square F value [T.sub.melt] 14.105 14.105 7.88 [T.sub.mold] 32.923 32.923 18.10 [P.sub.hold] 0.921 0.921 0.51 [T.sub.melt] x [T.sub.mold] 5.273 5.273 2.95 [T.sub.melt] x [P.sub.hold] 2.110 2.110 1.18 [T.sub.mold] x [P.sub.hold] 0.580 0.580 0.32 [T.sub.melt] x [P.sub.mold] 0.056 0.056 0.03 x [P.sub.hold] Error 85.87 1.789 Factor P value [T.sub.melt] 0.007 [T.sub.mold] 0.000 [P.sub.hold] 0.477 [T.sub.melt] x [T.sub.mold] 0.92 [T.sub.melt] x [P.sub.hold] 0.283 [T.sub.mold] x [P.sub.hold] 0.572 [T.sub.melt] x [P.sub.mold] 0.861 x [P.sub.hold] Error UTS = -24.5+ 0.138; [T.sub.melt], +0.174; rmo,d, +0.089; [T.sub.hold], -0.00080; [T.sub.melt] x [T.sub.mold], -0.000326; [T.sub.melt] x [P.sub.hold], +0.00038; [T.sub.mold] x [P.sub.hold], -0.000001; [T.sub.melt] x [T.sub.mold] x [P.sub.hold]. TABLE 5. Comparison of the interfacial area and of the welding performance. Condition Area ([mm.sup.2]) Increase (%) Strength (MPa) 1 72.9 Ref. 6.20 2 73.2 0.4 7.80 3 84.1 15.4 8.80 4 94.1 29.1 11.65 Condition Increase (%) 1 Ref. 2 26 3 42 4 89
|Printer friendly Cite/link Email Feedback|
|Author:||Giusti, Ruggero; Lucchetta, Giovanni|
|Publication:||Polymer Engineering and Science|
|Date:||Apr 1, 2018|
|Previous Article:||Dynamic Local Temperature Control in Micro-Injection Molding: Effects on Poly(lactic acid) Morphology.|
|Next Article:||An Evaluation of Three Different Techniques for Melt Impregnation of Glass Fiber Bundles With Polyamide 12.|