Shiri Jstmm 2024 Vf
Shiri Jstmm 2024 Vf
Shiri Jstmm 2024 Vf
10 Abstract: This study focuses on the analysis of the thermoforming process of plastic tubes, combining both experimental
11 and numerical approaches. An "L"-shaped mold was designed to conduct tests on a PA12 tube. The experimental setup,
12 based on a steam station, allowed precise control of air and water flow through the tube. The thermoforming cycle is
13 divided in three distinct phases: heating, cooling, and drying. The experimental tests are carried out using the L9 Taguchi
14 method, varying three process parameters at three levels: heating temperature, holding time, and cooling hold time.
15 Thickness and angle measurements were taken at the elbow of the tube after demolding, providing accurate information
16 on the resulting deformations. The experimental results are collected and statistically analyzed to determine the most
17 significant process factors. In parallel, a numerical model is developed using LS-DYNA software. The simulation results
18 showed a strong correlation with the experimental data, therefore, validating the relevance and accuracy of the model.
19 Keywords: Thermoforming; Tubular plastics; Experimental analysis; Numerical simulation; Taguchi Method.
20 1 Introduction
21 Thermoforming is an essential manufacturing process for producing plastic parts with complex shapes [1-3]. This
22 process involves heating a sheet of plastic to a suitable temperature to allow for deformation, followed by molding to
23 achieve the desired shape [4, 5]. Although this forming technique has been the subject of numerous studies concerning
24 flat plastic and composite structures, the thermoforming of tubular parts remains a relatively unexplored area in the
25 scientific literature. Research has included the thermomechanical behavior of polypropylene sheets [6], and the
26 thermoforming of transparent polycarbonate films [7]. In the composite field, research has considered the technical
27 challenges, which are associated with the thermoforming of multilayer composites, but these studies have primarily
28 concentrated on flat parts [8]. Despite the growing importance of plastic tubes in various industrial applications, such as
29 the automotive conduits, to cite some, a few studies have addressed thermoforming of tubular parts.
30 Tubular parts present specific challenges due to their geometry and material properties. Optimizing this process
31 requires understanding of the thermal and mechanical phenomena, which are involved in the material transformation
32 process. This study aims to analyze the thermoforming process of tubular plastic parts by integrating experimental and
33 numerical approaches. An "L"-shaped mold is designed to conduct tests with a PA12 tube, a material known for its
34 strength and light weight. The experimental setup uses a steam station machine allowing for control of the heating,
35 cooling, and drying conditions of the tube. The experimental strategy follows a L9 Taguchi design, accounting for three
36 process factors at 3 levels each, i.e., the heating temperature, holding time, and cooling time. Thickness and angle
37 measurements are measured at the elbow of the tube after demolding, providing data on the resulting deformations.
38 In the following sections, we discuss the experimental approach, illustrating the steps and methodologies used to
39 conduct these tests. Moreover, we present the numerical modeling study using the LS-DYNA software. The objective is
40 to simulate the interactions between the tube and the mold.
41
1
42 2 Experimental Study
43 An "L"-shaped mold with an angle of 110° was designed to conduct thermoforming tests, as shown in Figure 1. The
44 mold is composed of two arms; a short one and a three times longer arms.
45
46 Fig. 1 "L" shaped mold designed for thermoforming.
47 The mold is installed on a steam station shown in Figure 2, ensuring the tube's connection and allowing air and water
48 passage inside it. The machine is equipped with sensors to monitor the thermoforming process parameters.
49
50 Fig. 2 Steam station used for thermoforming.
51 A PA12 tube, with a thickness of 1 mm, an outer diameter of 10 mm, and a length of 450 mm, is inserted into the
52 mold. Once the ends are secured, the tube is thermoformed following as in Figure 3.
53
55 The thermoforming process is carried out according to three phases (see Figure 3), each playing an important role in
56 obtaining tubular parts.
57 ● Heating the Tube: In this first phase, the plastic tube is heated to a precise temperature to reach a plastic
58 condition, allowing it to deform. This process must be controlled to ensure a uniform distribution of heat, thus
2
59 avoiding hot spots that could lead to forming defects. Once the plastic condition is met, the tube is shaped and
60 retrained within the mold until it becomes irreversible.
61 ● Cooling: After forming, the second phase involves cooling the tube. This cooling must be carried out in a
62 controlled manner to stabilize the newly acquired shape and prevent internal stresses that could lead to cracks
63 and/or nonfunctional deformations therefore, the time and conditions of cooling are essential to ensure the
64 strength and durability of the formed part.
65 ● Drying: The final phase of the process is drying. This phase is important to remove residual moisture from the
66 tube.
67 Experimental tests were conducted following the L9 Taguchi method, which allows for systematic characterization of
68 the main effects in the thermoforming process. Three factors, namely, heating temperature, holding time and cooling time
69 at three levels each, are monitored. Each runs of the L9 plan is repeated three times to measure the experimental error.
70 The objective is to ascertain their impact on the quality of the thermoformed tube. After demolding, measurements of
71 thickness and angle at the elbow of the tube were taken to assess the resulting deformations. Figure 4 illustrates the
72 specimens obtained after thermoforming for the different experimental configurations. Table 1 presents the experimental
73 layout (L9) as well as the process responses obtained, providing an overview of the results.
74
75 Fig. 4 Specimens Obtained After Thermoforming for Different Experimental Configurations.
77
78 Table 2 provides is the ANOVA for the signal-to-noise (S/N) ratios and the means, allowing the evaluation of the
79 main effects of the experiment and suggesting the optimum setting.
80
3
81 Table 2 ANOVA. of the full model for the SN rations and means.
82
83 Table 3 presents the ANOVA of the reduced model for the signal-to-noise (S/N) ratios and means. a) It analyzes the
84 influence of process factors on the S/N ratios, identifying those that significantly affect the quality of the thermoformed
85 parts, accompanied by an effect plot. b) The second part focuses on the means, highlighting the contributions of the
86 factors to the variability of the measurements, also illustrated by an effect plot. Together, these elements provide an
87 overview of the influences of the factors on the quality of the of thermoformed parts.
88 Table 3 a) ANOVA of the reduced model and effect plot for the SN ratios. b) ANOVA of the reduced model and the
89 effect plot for the means.
90
91
92 (a)
4
93
94
95 (b)
96 Table 2 for the S/N ratios and the means indicates that at 5% level of type I error, only the Heating Time factor is
97 statistically significant with p-values 0.038 and 0.018, respectively. This is further confirmed by the high values of the
98 Adj SS values. Also, because the objective is to diminish the relative variation from the nominals (i.e., 1mm for the
99 thermo-formed tube and 120° for the L-shaped mold), the best setting of the Heating Time factor is level 3 (i.e., 150°C)
100 as suggested in Table 3 a) and b). The plot diagrams in Table 3 a) and b) illustrate this choice since the setting at level 3
101 of the Heating Time parameter will maximize the S/N ratio and minimizes the average means of the relative errors,
102 simultaneously. The high Rsq-adj metric of the reduced model is 95.1% suggesting a good reliability of the regression
103 model.
110
5
111 Fig. 5 Mesh used in numerical modeling.
112 A CONTACT_SURFACE_TO_SURFACE (3) type is assigned to manage the interactions between the tube and the
113 mold at the interface. The tube is fixed at both ends and modeled with a SECTION_SHELL_THERMAL element,
114 representing a thermally thick shell. This shell allows modeling a temperature gradient through its thickness, leading to
115 deformations due to thermal expansion variations. The mold, on the other hand, is considered rigid and fixed, with no
116 heat exchange taken into account with the tube. The tube undergoes a three phases thermal flux applied to its lower
117 surface using the BOUNDARY_FLUX_SET command. The phases correspond to the heating, cooling, and drying
118 cycles. The tube material, PA12 with an elastic-plastic behavior whose mechanical properties are retrieved from [10], is
119 subjected to a thermal expansion due to the thermal load. The simulation was configured as a coupled thermal-structural
120 analysis, using an implicit time integration (static). The thermal load is defined using the
121 BOUNDARY_TEMPERATURE command, allowing the thermal solver to compute the temperature fields, which are
122 then transferred to the structural solver to assess the induced thermal expansion. Initially at 20 °C, the temperature of the
123 steel tube gradually increases according to the thermal cycle. The thermal properties of the material are specified using
124 MAT_THERMAL_ISOTROPIC for the thermal solver, while temperature-dependent properties, such as the thermal
125 expansion coefficient, are defined with MAT_ADD_THERMAL_EXPANSION for the structural solver.
126 The post processing results show that the maximum displacement occurs at the elbow of the tube, as illustrated in
127 Figure 6. Furthermore, the maximum plastic deformation occurs at the ends of the tube, confirming the experimental
128 results indicating that rupture under pressure mainly occurs at these locations, as demonstrated in the corresponding
129 Figure 7.
130
131 Fig. 6 Distribution of displacements: maximum displacements at the elbow of the tube (mm).
132
133 Fig. 7 Plastic deformation distribution in the tube after thermoforming.
134 Although the elbow area also experiences plastic deformation, it is less significant than that observed at the ends. The
135 stress distributions shown in Figure 8 reveal a significant reduction in thickness at the ends, particularly in the area close
136 to the elbow.
6
137
138 Fig. 8 Stress distribution in the tube after thermoforming (MPa).
139 The reduction in thickness obtained numerically in the extreme edge of the tube is in excellent agreement with the
140 experimental measurements, with an experimental value of 6% (see Figure 9).
141
142 Fig. 9 Distribution of thickness variation in the tube.
143 After thermoforming, the contact between the tube and the mold is disabled, and an angle variation of 20.1° due to
144 elastic recovery is measured numerically, a result consistent with the experimental value. This correspondence between
145 the numerical and experimental results confirms the reliability of the model to predict material behavior under thermal
146 flux.
147 4. Conclusions
148 In this study, an experimental study is conducted using the Taguchi L9 design. The objective is to characterize the
149 thermoforming process. Three process factors are investigated; Heating Temperature, Holding Time, and CoolingTtime.
150 Measurements taken after demolding allowed for the evaluation of the deformations of the thermoformed tubes. The
151 statistical analysis revealed that the "Heating Time" factor plays a significant role in the quality of the formed parts.
152 Moreover, the run 7 in the L9 plan gives the smallest relative variation in thickness and angle of the thermoformed tubes.
153 The best setting is obtained when tuning the Heating Temperature (150C), Holding Time (20s), and Cooling Time (27s).
154 The numerical simulation using LS-DYNA software validates the developed model because of correlation with the
155 experimental results. This validation reinforces the relevance of simulation in understanding the process.
156 Looking ahead, further investigation of the interaction effects among the operating parameters is needed. An iterative
157 CCD DoE is a good strategy. In addition, refining the numerical model and applying it in an industrial context will
158 enable process optimization and improve the design of parts and molds, contributing to a higher productivity.
7
159 Acknowledgements
160 The authors thank AKWEL Mateur, Tunisia, for their valuable assistance in conducting the experimental tests for this
161 study.
162 References
163 [1] Klein, P. (2022). Fundamentals of plastics thermoforming. Springer Nature.
164 [2] Throne, J. (2024). Thermoforming. In Applied plastics engineering handbook (pp. 449-480). William Andrew
165 Publishing.
166 [3] Subramanian, M. N. (2024). Thermoforming: Processing and Technology. John Wiley & Sons.
167 [4] Gaspar-Cunha, A., Costa, P., Galuppo, W. D. C., Nóbrega, J. M., Duarte, F., & Costa, L. (2021). Multi-
168 Objective Optimization of Plastics Thermoforming. Mathematics, 9(15), 1760.
169 [5] Engelmann, S. (2023). Advanced Thermoforming: Methods, Machines and Materials, Applications, Automation,
170 Sustainability, and the Circular Economy. John Wiley & Sons.
171 [6] Freudenthaler, P. J., Fischer, J., Liu, Y., & Lang, R. W. (2023). Polypropylene post-consumer recyclate
172 compounds for thermoforming packaging applications. Polymers, 15(2), 345.
173 [7] Beltrão, M., Silva, M., Viana, J. C., Duarte, F. M., Dias, D., Marques, R., ... & Paulo, V. (2024). A study on the
174 influence of thermoforming process on the optical properties of polycarbonate films. International Journal of
175 Material Forming, 17(4), 1-15.
176 [8] Wittmann, L. M., & Drummer, D. (2022). Multilayer sheets for thermoforming non thermoformable
177 polymers. Journal of Plastic Film & Sheeting, 38(2), 225-244.
178 [9] Hallquist, J.O. (2006) LS-DYNA Theory Manual. Livermore Software Technology Corporation (LSTC),
179 Livermore.
180 [10] Amstutz, C., Weisse, B., Haeberlin, A., Burger, J., & Zurbuchen, A. (2022). Inverse Finite Element Approach to
181 Identify the Post-Necking Hardening Behavior of Polyamide 12 under Uniaxial Tension. Polymers, 14(17), 3476.