# A numerical investigation of the use of novel melt processed PET-hemp fiber composites for thermoforming applications.

INTRODUCTIONThe thermoforming process is one of the most important plastic and composite material forming techniques, with extensive uses in various industrial and commercial applications. It consists in heating a material sheet which is either extruded or cut until its softened state and subsequently deforming it into the mold shape by an applied pressure, a vacuum, a moving plug or a combination of the three. The constant innovation of the thermoforming market results in the processing of more complex geometries and an expansion to a wider list of potential materials, which selection is based on their cost, their working properties, and their processability. In this regard, sufficient work has not been done to elucidate the viscoelastic behavior of thermoplastic matrices reinforced with vegetal fibers especially towards the thermoforming process. Moreover, high temperature melting thermoplastic reinforced with vegetal fibers have only recently regained researchers' attention and been developed [1, 2], making them a good target for thermoforming studies.

In fact, as described below, current observations show that low and high temperature melting thermoplastics exhibit different kind of behavior with respect to the reinforcement by vegetal fibers and with respect to secondary forming processes like thermoforming.

During the past decades, a significant amount of work has been done for the development of low temperature melting thermoplastic matrices such as polypropylene (PP), high density polyethylene (HDPE) and Nylon reinforced with natural fibers such as wood flour, sawdust and bast fibers like hemp. The low processing temperature of these matrices is rather advantageous for the thermal stability of vegetal fiber and a justification of the high volume of their applications with wood-related reinforcements as high as 70% by weight [1]. Moreover, vegetal reinforcements have been reported to be more advantageous due to their low density, their attractive low cost and their nonabrasive properties. They are additionally abundant, recyclable and they have a limited toxicity [3, 4]. These composite materials have recently been found in different fields of application in the example of construction materials, the aerospace industry, as well as furniture and car parts.

There are only a few challenges in processing low temperature melting thermoplastics with vegetal fiber including the chemical incompatibility between hydrophilic fibers and hydrophobic matrices which results in disappointing mechanical properties [5-7] and the limited processable volume of bast fibers due to their high surface to volume ratio. Many methods of tackling the chemical incompatibility challenge have been reported in the literature based on improving the bonding quality between the fiber cellulose molecules and the matrix either through a chemical modification of the components or the addition of a coupling agent to connect its components [4, 7].

Contrary to the composites with low temperature melting thermoplastics, it is more challenging but advantageous to reinforce high temperature melting thermoplastics with vegetal fibers. The temperature difference between the onset of thermal degradation of vegetal fibers ([T.sub.d] = 190[degrees]C) [8] and the melting point of high temperature melting thermoplastic matrices ([T.sub.m] > 200[degrees]C) is the main challenge to processing such composite material without thermal degradation of vegetal fibers. However, polar matrices like PET offer the possibility of hydrogen bonding between the carbonyl groups of the matrix and the hydroxyl groups of cellulose. In fact, our previous works have shown an improvement of the mechanical and structural properties of PET-hemp fiber composites in the absence of coupling agents [1].

Apart from the challenges discussed in the previous sections, many other parameters influence the final properties of a composite material made of plastics and vegetal fibers. They include the reinforcing load and its geometric parameters (L/D ratio, granulometric distribution), the proportion of the coupling agent as well as the processing conditions [3-7, 9], the blend quality, the thermal degradation of vegetal fibers and the blend viscosity. In fact, an effective mixing process is crucial to achieve a homogeneous blend resulting in optimal composite properties. Additionally, various authors have shown that the presence of vegetal fibers such as wood chips significantly increase the viscosity of the mixture, thereby limiting its ability to be formed with standard equipment such as the twin screw extruder [8, 9] thus the necessity to investigate their behavior towards the thermoforming process.

This work investigates the elaboration and the numerical thermoformability of the novel melt processed polyethylene terephthalate (PET)-hemp fiber composites. PET-hemp fiber composites were successfully processed by combining an alkaline fiber treatment to the melting point depression of the matrix, followed by compounding in a Torque-based Rheometer and injection molding.

The dynamic explicit problem of forming a thin viscoelastic composite structure subjected to an air flow loading was used during the numerical analysis [10] while the virtual external work which is involved in the finite elements' formulation was expressed in the form of a volume integral [11]. The pressure was then derived from the Redlich-Kwong's real gas equation of state [12]. The process modeling was done by an application of the finite element method based on the Lagrangian formulation and assuming both the membrane theory and the incompressibility of the composite material. The membrane structure was discretized by plane finite elements [13] while the Christensen model and the characterization technique were considered for their ability to describe free inflations [14], The influence of the Lodge constitutive model on the thickness and the stress distribution in the free inflation of five PET-hemp fiber composite formulations were investigated.

EXPERIMENTAL

The PET-hemp fiber composite samples were processed by combining an alkaline treatment of the fibers with the melting point depression of the PET matrix, followed successively by compounding and injection molding. Furthermore, the three groups of composite properties described below were necessary for the intended thermoformability investigation.

Materials

PET grade AA-48 (Eastman, Montreal, QC, Canada), polycaprolactone (PCL, Sigma-Aldrich, Oakville, ON, Canada) and hemp fibers of composite grade from Lanaupole (Berthierville, QC, Canada) were used in this work. These fibers had an average length and diameter of respectively 50 mm and 20-25 pm. The preliminary work investigated the effect of some additives such as clay grade Cloisite 30B (Southern Clay Products, Gonzales, TX), pyromelitic dianhydride (PMDA) and glycidyl methacrylate (GMA) from Sigma Aldrich (Oakville, ON, Canada) were respectively used as fire retardant, chain extender and fiber's coating agent [15-17]. Moreover, triethylamine, hydroquinone and sodium hydroxide were used in different stages of the process as chemical reagents.

Composite Elaboration

The virgin hemp fibers were treated for an hour with 5N NaOH solution in order to increase their thermal stability based on an alternative to the procedure of K. Specht et al. [18]. The melting point depression of PET was achieved by blending with 5% (w/w) polycaprolactone in an internal batch mixer (Haake Rheomix, Polylab OS system, USA), based on a previously published method by Papageorgiou et al. [19]. The compounding process of alkaline treated fibers and modified PET was done in an internal batch mixer (Haake Rheomix, Polylab OS system), with the melting chamber heated at 250[degrees]C. Furthermore, injection molding of the mechanical and rheological samples was done at 250[degrees]C with a Haake Minijet, with a mold heated at 50[degrees]C.

All the hygroscopic substances such as PET, the modified PET and the compounded PET-hemp formulations were pre-dried prior to compounding and injection molding to avoid sample degradation as reported by authors like La Mantia and Morreale [20] and Awaja et al. [15]. In fact, PET being synthesized by esterification reaction between terephthalic acid and ethylene glycol with water as a by product, would decompose when heated in the presence of water.

Composites Characterization

The mechanical properties of the investigated composites were evaluated using tensile tests with an Instron model 4206 at a cross head speed of 5 mm [min.sup.-1] based on ASTM D638-08. The data were analyzed by inbuilt software to determine the influence of both the additives used and the variation of hemp fibers concentration.

The structural properties of the investigated composites were analyzed with a Scanning Electron Microscope Philips (model XL 30, USA) to determine the effects of the additives and the reinforcing concentrations. The samples were cryo-fractured and gold coated prior to the analysis.

The loss and storage moduli of the investigated PET reinforced with 0, 1,5, 10, and 15% (w/w) hemp fibers were determined by small amplitude oscillatory shear test [21] and the results applied to the finite element ThermoForm[R] code developed by the senior researcher for both an identification of their thermo-rheological behavior and the determination of their relaxation properties.

MECHANICAL AND STRUCTURAL IMPACTS

The mechanical and structural properties of the investigated PET-hemp fiber composites indicated the same trend earlier observed by other authors about the behavior of low temperature melting thermoplastic matrices reinforced with vegetal fibers [22] and the need for a trade-off between the required properties, the additives and the targeted applications is necessary. In fact, our previous works indicated an increase of both the elastic modulus and the maximum strength following an increase of the hemp concentration from 1 to 20% (w/w) and the same behavior in the presence of different additives. In the contrary, the strain at break dropped consistently from above 40 for virgin PET to around 5%. Moreover, an improved interface was observed from the micrographs of all the PET-hemp fiber formulations without additives, indicating a good fiber-matrix bonding [1]. The micrographs of all the investigated formulations confirmed the thermal stability of the treated hemp fibers in the suggested processing method and the thermal stability of those formulations were studied for further processing [1, 2]. However, above 10% (w/w) of hemp fibers concentration, more fiber-fiber contact could be observed indicating potential weak structural points.

Modeling and Simulation

The behavior of the formulated PET-hemp fiber composites towards the thermoforming process was numerically investigated using an application based on the thermoforming of a circular membrane with 15 cm diameter and initial thickness h0 of 1.47 mm. The processing load being expressed in terms of nonlinear air flow rate as shown in Fig. 2.

PRELIMINARY CONSIDERATIONS

In this work, apart from the membrane theory and the incompressibility of the composite materials, PET-hemp fiber composite materials were considered above their glass transition temperature. Therefore, the investigated materials were assumed to satisfy a viscoelastic isotropic constitutive equation.

CONSTITUTIVE EQUATION

The Christensen constitutive model [13] was applied to describe the behavior of PET-hemp fiber composite materials, considered isotropic. Such viscoelastic integral model which is more appropriate for the description of polymeric behavior in both their semi-solid or molten state can be used for the description of the thermoforming of semi-solid materials. The Christensen model describes the true matrix stress [[sigma](t)] at time t as a function of the Lagrangian strain history [E(t)] as shown in Eq. 1; where [F] is the deformations gradient, p is the isostatic pressure, [g.sub.0] is the hyper elastic modulus and [g.sub.1] is the material relaxation function described by Eq. 2.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

The relaxation function of the material is a relaxation spectrum with moduli [c.sub.k] and relaxation time [tau].sub.k]. The dependency of these models on temperature is accounted for by using the WLF function [16].

CHARACTERIZATION TECHNIQUES

Different characterization techniques were used to identify the material properties of PET-[0, 1, 5, 10, and 15% (w/w)] hemp fiber composites needed for the simulation process. The linear properties were obtained from small amplitude oscillatory shear tests [21] known for the determination of the storage and loss moduli as a function of the frequency as shown in Fig. 1. Moreover, their relaxation spectra were determined by optimizing the dynamic data and minimizing the objective function F defined by Eq. 3, where N is the number of data points, ([G'.sub.i.exp.], [G".sub.i.exp.]) are available from the dynamic experiments and ([G'.sub.i,fit], [G'.sub.i,fit.]) are the best fit values based on Eq. 4.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], (4)

Taking advantage of the fact that low-frequency behavior is dominated by the long relaxation times, just as high-frequency response is controlled by the short relaxation times, the linear viscoelastic behavior of the composite material is described over a wide range of time values with just a few constants.

Figure 1 shows that the Christensen model adequately describes the behavior of the investigated PET-hemp fiber composites while fitting the experimental data and the theoretical model. The associated relaxation strengths [g.sub.k] are given in Table 1. This behavior slightly differs from the observations made by Erchiqui et al. [23] about the formulations of HDPE reinforced with 0, 20, 30, 40, 50, and 60% (w/w) wood flours. In fact, an application of the same fitting method in the latter case resulted in the description of the composite behavior by the Lodge model. Furthermore, the variations of different parameters related to the numerical thermoforming of PET-hemp fiber composites are further analyzed in the following sections.

FINITE ELEMENTS DISCRETIZATION

The thermoforming of PET-hemp fiber composite membranes was achieved by applying the space and time discretization to the explicit dynamic finite element method [24], The principle of virtual work was thus applied on the undeformed configuration for both the inertial effects and internal work. The time discretization was required to handle the inertial forces while avoiding their associated instability. It was achieved by the use of a conditionally stable centered finite difference technique [25], The thermoforming problem was then reduced to the discrete system of Eq. 41 [11]. [F.sub.ext], [F.sub.gray] and [F.sub.int] are respectively the external, body and internal global nodal force vectors experienced by the composite membranes, and M is the mass matrix.

A diagonalization of Eq. 41 yields Eq. 42 in which the M matrix is transformed into a diagonal form and each degree of freedom independently controlled. [M.sup.d.sub.ii] are the diagonal components of the matrix [M.sup.d].

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (6)

The Courant-Friedrichs-Lewy criterion of Eq. 7 was applied to the thermoforming of PET-hemp fiber composites as the common convergence criterion applied to explicit dynamic finite element method for nonlinear problems. The parameters C and / denote, respectively the wave speed in the medium and the element size; thus the ratio l/c is the time needed for a wave propagation across an element of size l and [epsilon] is a proportionality constant related to the applied integration scheme.

[DELTA]t [less than or equal to] [DELTA][t.sub.crt] = [epsilon]. l/c (7)

The initial conditions given in Eq. 8 stipulate that the displacement and velocity vectors are assumed to be zero at the beginning of the thermoforming process.

[u.sub.i]([t.sub.0]) = 0

[[??].sub.i]([t.sub.0]) = 0 (8)

PLANE-STRESS ASSUMPTION

An accurate evaluation of the internal forces is necessary for the numerical thermoforming of PET-hemp fiber composite membranes. However, such process requires a computation of the stress deformation relationship for each element under a plane-stress and an incompressibility assumption. Consequently, the components of the Cauchy stress tensor satisfy the conditions of Eq. 9.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (9)

GAS EQUATION OF STATE AND PRESSURE LOADING

The forming forces were expressed in terms of the air flow on the composite membranes during the simulation process. Additional assumptions needed for the calculation of the dynamic pressure process are listed below.

The gas temperature ([T.sub.gas]) is constant;

The pressure ([P.sub.0]) between the composite membranes and the mold is constant;

The initial volume ([V.sub.0]) enclosing the composite membranes at the initial time ([t.sub.0]) and containing (nn) moles of gas is evaluated by formulating the external virtual work and pressure in terms of a closed volume and the Redlich-Kwong's equation of state [11, 12] as shown in Eqs. 10 and 11; n(t), P(t), and F(t) are respectively the moles of air, the internal pressure and the volume occupied by the composite membranes at time t. The parameters a and b are determined from the critical pressure ([P.sub.c]) and temperature ([T.sub.c]) of the gas used in the blowing process as shown in Eq. 12.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (10)

and

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (11)

a = 0, 42748 x [[bar.R].sup.2] [T.sup.2,5.sub.c]/[P.sub.c]/ b = 0, 08664 x [[bar.R][T.sub.c]/[P.sub.c] (12)

Equation 77 is also the time variations of the pressure inside the composite membrane following the variations of their internal volume. Such expression of the load in term of gas How formulated based on the thermodynamic state equation is an advantageous way to deal with the load-deformation while avoiding the instabilities associated with the classical pressure loading. In the contrary, the use of a constant pressure as a loading force instead of the gas flow velocity in the applied quasi-static and dynamic finite element formulations yields a divergence of pressure values computation beyond a given critical point [25, 26],

ENERGY AND POWER

The energy required for the thermoforming process is equivalent to the mechanical work done by the external forces on the composite membranes during the forming stage. Its compact form is shown in Eq. 13 where {[F.sub.ext]} is the global nodal external force vector and [{[u.sup.n]}.sup.T] is the associated global nodal displacement vector. The power associated with the global nodal external force vector is the energy per unit time transferred to the composite membrane during the forming stage as shown in Eq. 14.

[W.sup.ext] = [{[u.sup.n]}.sup.T] x {[F.sub.ext]} (13)

[p.sup.ext] = [W.sup.ext]/[t.sub.forming] (14)

The energy and power associated with the forming phase of the material significantly impact its part manufacturing cost and its industrial feasibility. They were evaluated for the potential thermoforming of the investigated PET-hemp fiber composite formulations.

RESULTS AND DISCUSSION

Many thermoforming parameters were analyzed after the ThermoForm[R] code was used to implement the dynamic finite element method outlined in the previous sections [10]. They include the stress and deformations of the composite membranes during the process as well as the associated energy. The targeted applications derive from circular PET-hemp fiber composite membranes of 15 cm diameter and an initial thickness h0 of 1.47 mm.

An expression of the processing load was done in terms of nonlinear air flow rate. Its variations shown in Fig. 2 indicate a maximum value of 0.250 moles per seconds at 0.2 s. The mold geometry and the composite membranes were discretized using triangular membrane elements. The discretized mold consisted of 629 elements and 338 nodes; while the discretized composite membranes consisted of 2784 elements, 1422 nodes and fixed sides. The method was repeated for PET-hemp fiber composite formulations containing 0, 1,5, 10, and 15% (w/w) reinforcements.

The deformations of PET-hemp fiber composite membranes at selected stages of their forming process and their contact with the mold are shown in Fig. 3. The red nodes indicate the contact points between the membrane and the mold. A progression of the contact points from the clamping section to the base of the mold is observed at respectively 0.0, 0.90, 0.134, 0.149, 0.156, and 0.164 s. Moreover, 0.164 s represents the end of the process, while most of the deformation happened during the last two tracking times.

The behavior of the investigated materials throughout the forming process is further described by the variations of the principal extension and the von Mises stress in the parts molded with representative composites formulations reinforced with 1 and 5% (w/w) hemp fibers as illustrated in Figs. 4 and 5.

The principal extension is found to decrease from the clamping point to the base of the mold. A similar behavior was previously observed by Szvegda [27] who associated it to additional stress from the clamping forces. Moreover, it varies from 2.59E-4 to 7.97E-4 mm [mm.sup.-1] and from 2.81E-4 to 7.21E-4 mm [mm.sup.-1] in the presence of 1 and 5% (w/w) hemp fiber respectively. Contrary to the principal extension, the von Mises stress increases from the clamping point to the base of the mold, varying from 8.94E-2 to 4.32E-1 (MPa) and from 1.13E-4 to 3E-1 (MPa) in the presence of 1 and 5% (w/w) hemp fiber, respectively.

Figures 6 and 7 give the time variations of respectively the internal pressure and the volume of thermoformed parts with the five investigated PET-hemp fiber composite formulations. The time variations of the volume were nearly identical for all the composite formulations. In the contrary, the variations of the internal pressure consisted of an increase until a short plateau is reached, followed by a sudden increase at the end of the process. The material formulations can be further partitioned into two groups with identical internal pressure based on their reinforcement concentration. The first group is reinforced with 0 and 5% (w/w) hemp fibers, while the second is reinforced with 1, 10, and 15% (w/w) hemp fibers. An identical partition is shown by the variations of the internal pressure with the volume.

The observed behavior can be linked to different parameters such as the effect of 5% (w/w) PCL, an impact of the important volume to weight ratio of the fibers as well as their general structure. In the contrary of man-made fibers, natural fiber properties vary with different conditions and consequently hardly follow a specific trend as earlier reported in the literature [28]. The relative good performance of PET-1% (w/w) hemp fiber composites with respect to other formulations is probably caused by a combination of the high number of carbonyl groups deriving from both PET and PCL molecules, and the limited possibilities of weakened structure from fiber-fiber contacts. In fact, such high number of carbonyl groups contributes to increasing the sites of interactions with 1% (w/w) hemp fibers which is too low a concentration to favor fiber-fiber contact. In the contrary, higher reinforcement concentrations like 10 and 15% (w/w) increase the possibilities of fiber-fiber contacts which yield weak points of the composite structure; moreover, they increase the possibilities for some thermo-degradation of the fiber due to the challenges associated to processing their large volumes with comparatively lower matrix volumes especially with laboratory scale equipment. Consequently, the formulations reinforced with 10 and 15% (w/w) hemp fibers showed identical behavior. This is an indication that the best formulations based on these parameters derive from processing with lower fiber reinforcements.

The maximal pressure derived from the data are all higher than 18 kPa and they are reached after a period that varies with the fiber concentration. It appears that a high fiber concentration yields challenging processes due to the hardening of the composite membrane. The formulations reinforced with 1, 10, and 15% (w/w) hemp fibers showed the maximum forming pressures higher than 25 kPa indicating more difficult and expensive forming processes than the formulations reinforced with 0 and 5% hemp fibers. This further implies that lower reinforcement concentrations are more interesting for the thermoforming process while PET-5% (w/w) hemp fiber is optimal.

Some important parameters for the simulation of the blow molding and thermoforming processes as well as parts design are the thickness prediction and the stress estimates on the half planes of symmetry. Those parameters are discussed in the following sections. In fact, the residual stress which occurs during the forming process and the shape stability predictions are strongly related to the estimated stress. In the same manner, the localized thinning effects of the deformed membranes are generally accompanied by an increase in the true stresses of the material.

The distributions of the final von Mises stress (creq.) on the XZ half plane of symmetry of PET-hemp fiber composite membranes also known as the median line of the structure is given in Fig. 8. Such distribution deriving from the Christensen's constitutive model is similar for all the composite formulations. Moreover, it shows a maximum of about 0.411 MPa in the presence of 10% (w/w) hemp fibers and a minimum of about 0.292 MPa for virgin PET. A trend similar to the observations made in the previous sections about the time variations of the internal pressure is also observed except at the center of the hemisphere.

Moreover, the critical values of the von Mises stresses in the final shape of the thermoformed part with different composite formulations are presented in Table 2. A similar variation was observed for all the formulations with symmetry at the center of the mold.

All these observations imply that the composite formulations can also be grouped in two groups reinforced with 0 and 5% (w/w) hemp fibers on one hand, and 1, 10, and 15% (w/w) hemp fibers on the other hand. They confirm the fact that higher reinforcement concentrations are less interesting for the thermoforming process as they require powerful air flow during the forming process.

The extension distribution in the final shapes of the thermoformed part for each PET-hemp fiber composite formulation is given in Fig. 9. It shows similar variations for all the formulations, with a maximum of about 0.72 close to the clamping points, and a minimum of about 0.43 at 1.5 cm of the membrane center. The observed similarities in thickness distributions is in agreement with previous reports by DeLorenzi and Nied [29] relating the behavior to the incompressibility of isotropic material. Indeed, these authors have reported that thickness distribution is less dependent on the material behavior than it is on the mold geometry. Finally, the von Mises stress distribution and the localized thinning effect indicate that large deformations induced by inflation is most likely to cause material failure at about 2.0 cm of the membrane center as shown in Fig. 9.

Furthermore, Fig. 10 illustrates the main extensions ([[lambda].sub.1], [[lambda].sub.2], and [[lambda].sub.3]), on the XZ half plane of symmetry in the membrane for the composite formulations reinforced with respectively 0 and 10% hemp fibers. The variations observed in the two cases are similar with close values.

Figure 11 gives the time variation of the work required for molding the needed parts from the investigated PET-hemp fiber composite formulations. It varies with the fiber concentration, however the same two groups of formulation earlier mentioned can be observed based on this parameter. Moreover, these observations fully agree with previous ones made with other parameters and the previous explanations are applicable. This implies that the energy used for thermoforming the parts with the investigated PET-hemp fiber composites are related to their fiber concentration. In comparison with the energy required for thermoforming virgin PET, 1,380 times more energy is required for the composite reinforced with 10% (w/w) hemp fibers. In the same manner, the ratios of energy required with other formulations are 1,394 for 1% (w/w) fibers, 0,944 for 5%(w/w) fibers and 1,367 for 15% (w/w) fibers.

The power variations derived from the data follow a trend which is similar to the variations of previous parameters like the internal pressure and the work required for the thermoforming process. The power required is respectively 44.95, 60.82, 43.21, 59.65, and 59.65 Watt for the composites reinforced with 0, 1, 5, 10, and 15% (w/w) hemp fibers. Similarly, to previous observations, an analysis of investigated composite formulations based on the required power shows that they can be partitioned into the previously mentioned two groups which are (0, 5% (w/w)) on hand and (1, 10, 15% (w/w)) on the other. The observations have a significant impact on the choice of PET-hemp fiber composite formulations which are suitable for the thermoforming process based on their performance, the quality of the final product and the cost of the process.

In a cost-benefit perspective, the thermoforming cycle which is a pertinent industrial parameter also shows non-monotonic variations with the fibers concentration. In fact, the parts forming cycle with the composite containing 1 and 5% (w/w) fibers are respectively 0.170 and 0.162 s. Consequently, a consideration of both the forming energy and the process cycle indicate that the profitability of the thermoforming process depends on the fiber concentration.

Such simple investigation of the suitability of newly formulated PET-hemp fiber composites for the thermoforming process shows the advantage of applying the dynamic finite element method based on both a total Lagrangian approach and the air flow loading derived from thermodynamic theories. An analysis of different molding parameters suggests that more work must be done in order to optimize both the composite formulation and the thermoforming process. This includes for example taking into consideration the energy dissipated by frictional contact between the mold and the composite membranes, and using data from industrial scale equipment.

Our future work will focus on refining the methodology for processing such high temperature melting thermoplastics reinforced with vegetal fibers while using numerical tools to facilitate their applications through the thermoforming process. Furthermore, we intend to validate the suitability of the formulations reinforced with low fiber concentrations (1 and 5% (w/w)) for the thermoforming process based on industrial tools and structural instrumentation techniques.

CONCLUSIONS

The mechanical and structural properties of novel melt processed PET-hemp fiber composites were investigated, followed by a numerical investigation of their suitability for the thermoforming process. The variations of the process parameters such as the air flow and the associated energy with the fiber concentration, as well as the variations of the material parameters such as the thickness, the stress, internal pressure and stretch with the fiber concentration were analyzed for the formulations reinforced with 0, 1,5, 10, and 15% (w/w) hemp fibers.

Our study showed that the constitutive behavior of such composite formulations has little influence on the final thickness distribution in the thermoformed part. Moreover, important parameters such as the process cycle and the energy required for the thermoforming process as well as the stress and stretch distribution do not show monotonic variations with the fiber concentration. However, the investigated composite formulations can be partitioned into two groups with respect to both their behavior and the fiber concentration. The composites of the first group are reinforced with 0 and 5% (w/w) fibers, while those of the second are reinforced with 1, 10, and 15% (w/w) hemp fibers.

Consequently, the formulations with low fiber concentrations such as 0, 1, and 5% (w/w) are the most suitable for the thermoforming process; however, a better valorization of other formulations would either require a combination of woven reinforcements and resin transfer molding, or the use of elastomers to tackle the challenges related to their brittle nature.

REFERENCES

[1.] A.S. Fotso Talla, E. Mfoumou, S. Jeson, D. Page, and F. Erchiqui, Reinf. Plast. Compos., 32, 1526 (2013).

[2.] A.S. Fotso Talla. F. Erchiqui, H. Kaddami, and D. Kocaefe, J. Appl. Polym. Sci., 132, (2015).

[3.] M. Bengtsson, P. Gatenholm, and K. Oksman, Compos. Sci. Technol., 65, 1468 (2005).

[4.] P.J. Herrera-Franco and A. Valadez-Gonzalez, Compos. B, 36, 597 (2005).

[5.] P. Balasuriya, L. Ye, and Y.W. Mai, Compos. A, 32, 619 (2001).

[6.] R.P. Kumar, M.L.G. Arama, and S. Thomas, J. Appl. Polym. Sci., 58, 597 (1995).

[7.] M. Lu, Wood Fiber Reinforced Polypropylene Composites, Louisiana State University (1997).

[8.] D.N. Saheb and J.P. Jog, Adv. Polym. Techno!., 18, 351 (1999).

[9.] A.K. Bledski, M. Letman, A. Viksne, and L. Renee, Compos. A, 36, 789 (2005).

[10.] F. Erchiqui, A. Gakwaya, and M. Rachik, Polym. Eng. Sci., 45, 125 (2005).

[11.] F. Erchiqui, Polym. Eng. Sci., 46, 1554 (2006).

[12.] O. Redlich and J.N.S. Kwong, Chem. Rev., 44. 233 (1949).

[13.] O.C. Zienkiewicz and R.L. Taylor, The Finite Element Method, Vol. 1, McGraw-Hill Book Company, Europe (1994).

[14.] A.S. Lodge, Elastic Liquids, Academic Press, London (1964).

[15.] F. Awaja, F. Daver, and E. Kosior, Polym. Eng. Sci., 44, 1579 (2004).

[16.] M. Pracella, D. Chionna, I. Anguillesi, Z. Kulinski, and E. Piorkowska, Compos. Sci. Technol., 66, 2218 (2006).

[17.] H. Qin, S. Zhang, G. Hu, and M. Yang, Polymer, 46, 8386 (2005).

[18.] A.K. Bledzki, H.P. Fink, and K. Specht, Appl. Polym. Sci., 93, 2150 (2004).

[19.] G. Z. Papageorgiou, D.S. Achillas, and G.P. Karayannidis, Polym. Int., 53, 1367 (2004).

[20.] F.P. La Mantia and M. Morreale, Compos. A, 42, 579 (2011).

[21.] C. Papanastasiou, L.E. Scriven, and C.W. Macosko, J. Rheol. 27, 387 (1983).

[22.] F. Godard, M. Vincent, J.F. Agassant, and B. Vergnes, Etudes du comportement rheologique et des proprietesmecaniques de composites sciures de bois-polyethylene haute densite, Rheologie. 136, 21 (2008).

[23.] F. Erchiqui, F. Godard, A. Koubba, M. Vincent, and H. Kaddami, J. Reinf. Plast. Compos., 28, 1153 (2008).

[24.] F. Erchiqui, A. Gakwaya, M. Rachik, A. Gakwaya, and M. Rachik, Polym. Eng. Sci., 45, 125 (2005).

[25.] K. Subbaraj and M.A. Dokainish, Comput. Struct., 32, 1387 (1989).

[26.] E. Verron, G. Marckmann, and B. Peseux, Int. J. Numer. Methods Eng., 50, 1233 (2001).

[27.] D. Szegda, Experimental Investigation and Computational Modelling of the Thermoforming Process of Thermoplastic Starch, Brunei University (2009).

[28.] B. Madsen, Properties of Plant Fibre Yarn Polymer Composites: An Experimental Study, Technical University of Denmark (2004).

[29.] H.G. DeLorenzi and H.F. Nied, "Finite Element Simulation of Thermoforming and Blow Molding," in Progress in Polymer Processing, A.I. Isayev, Ed., Carl Hanser Verlag, Munich, Germany, 171 (1991).

F. Erchiqui, (1) A.S. Fotso Talla, (1) H. Kaddami (1,2)

(1) Departement Des Sciences Appliquees, Universite Du Quebec En Abitibi-Temiscamingue, Rouyn-Noranda, Quebec, Canada

(2) Laboratoire De Chimie Bioorganique Et Macromoleculaire, Universite Kadi Ayyad, Faculte Des Sciences Et Techniques, Gueliz Marrakech, Maroc

Correspondence to: F. Erchiqui; e-mail; fouad.erchiqui@uqat.ca

DOI 10.1002/pen.24332

Published online in Wiley Online Library (wileyonlinelibrary.com).

TABLE 1. Relaxation spectrum of PET-hemp fiber composites compounded with a mixing chamber heated at 250[degrees]C. Hemp fiber load [% (w/w)] 0 1 5 Co (MPa) 0.5 0.5 0.5 [[tau].sub.k] [g.sub.k] (MPa) [g.sub.k] (MPa) [g.sub.k] (MPa) (sec) 0.010 1.091 2.095 0.700 0.050 1.312 1.448 1.340 0.500 0.250 -0.740 0.083 1.000 -1.348 0.465 1.199 Hemp fiber load [% (w/w)] 10 15 Co (MPa) 0.5 0.5 [[tau].sub.k] [g.sub.k] (MPa) [g.sub.k] (MPa) (sec) 0.010 1.636 1.977 0.050 1.307 1.314 0.500 -0.490 0.741 1.000 -0.555 -0.400 TABLE 2. The critical values of the von Mises stress for PET-hemp fiber composite parts. Hemp fiber load [% (w/w)] 0 1 5 10 15 [[segma].sub.eq] (MPa) 0.292 0.410 0.294 0.412 0.394

----------

Please note: Some tables or figures were omitted from this article.

Printer friendly Cite/link Email Feedback | |

Author: | Erchiqui, F.; Talla, A.S. Fotso; Kaddami, H. |
---|---|

Publication: | Polymer Engineering and Science |

Article Type: | Report |

Date: | Sep 1, 2016 |

Words: | 5883 |

Previous Article: | The influence of molecular weight distribution of industrial polystyrene on its melt extensional and ultimate properties. |

Next Article: | Carbon nanotube-based thermoplastic polyurethane-poly (methyl methacrylate) nanocomposites for pressure sensing applications. |

Topics: |