3D printing of PLA and PMMA multilayered model polymers: an innovative approach for a better-controlled pellet multi-extrusion process

Mohamed Yousfi. Université de Lyon, INSA Lyon, CNRS, UMR 5223, Ingénierie des Matériaux Polymères, F-69621 Villeurbanne, France. Corresponding author: mohamed.yousfi@insa-lyon.fr Ahmed Belhadj. Université de Lyon, INSA Lyon, CNRS, UMR 5223, Ingénierie des Matériaux Polymères, F-69621 Villeurbanne, France. Khalid Lamnawar. Université de Lyon, INSA Lyon, CNRS, UMR 5223, Ingénierie des Matériaux Polymères, F-69621 Villeurbanne, France. Corresponding author: khalid.lamnawar@insa-lyon.fr Abderrahim Maazouz. Université de Lyon, INSA Lyon, CNRS, UMR 5223, Ingénierie des Matériaux Polymères, F-69621 Villeurbanne, France. Hassan II Academy of Science and Technology, Rabat, Morocco. Corresponding author: abderrahim.maazouz@insa-lyon.fr

1 Intr 1 Introduction oduction 3D printing or additive manufacturing (AM) promises to powerfully influence not only production systems, but also life in general. Recently, great initiatives have been devoted to the development of AM machines for multimaterial 3D printing. This multicomponent strategy enhances the printing quality and ultimate performance of objects by grading the composition or type of polymers within the printed layers; this is not easily achievable by conventional manufacturing methods such as multimaterial injection molding [1].
Fused deposition modeling (FDM) requires the preparation of a filament spool. Fused granular fabrication (FGF), in which polymers in the form of pellets are directly printed in 3D, is an increasingly popular AM technology because it saves time and lowers the industrial cost price by removing the filament spool extrusion step.
In the FDM process, each printing layer is in a semi-molten state when deposited. The motion of the nozzle or the ESAFORM 2021. MS12 (Polymer), 10.25518/esaform21.1024 1024/1 hot building platform presses it slightly into the previous layer, which in turn is heated once again. This combination of pressure and heat is reminiscent of plastics welding. When the two layers are in contact, an interface is created, on either side of which the macromolecular chains can interpenetrate or interdiffuse [2][3][4]. The free ends of the chains unite and become entangled in a process called cicatrization (or "healing") [5]. Based on the framework of tube theory [6], Yang and Pitchumani [7] reported that the reptation time is a significant parameter required for full healing of polymer welds and they defined a degree of healing D Dh h(t) (t) in isothermal condition according to: Here, τw(T) is the average chain reptation time and t is the melt process time. The reptation time, which is assimilated to the weld time, is commonly accepted to present an exponential dependence on molar mass (Mw), as τw~Mw 3 in entangled polymers. But in the 3D printing process, the thermal history of the interface is non-isothermal and the time during which chains can interdiffuse is short (less than few seconds).
It should be noted that the temperature of the interface between the two successively deposited layers must exceed the glass transition temperature Tg in order for the polymer chains to diffuse and binding to occur [5,8].
Seppala et al. [14] studied the strength of the weld formed between the layers produced by AM (FFF). They measured the mechanical strength of an FFF weld directly through a torsional test called a "trouser tear" fracture experiment.
Their protocol has been tested with ABS and more recently with biodegradable polymers such as PBS and PBSA [15].
Elsewhere, McIlroy and Olmsted [13] used the entanglements number, Ze, of a melt of molecular mass Mw as an indicator parameter to probe the quality of polymer printing. The Ze number is defined as the ratio between Mw and Me. The latter is the molecular mass between entanglements. According to the authors, the Ze ratio should be in the 20-30 range to provide good interfacial adhesion.
In the present study, PLA and PMMA monocomponents as well as PMMA/PLA multicomponents were prepared by the FGF printing process (Fig. 1). The healing degree of the mono-material based systems was investigated by rheological modeling.
Next, the effect of printing chamber temperature was specifically discussed. The boundary layer interfaces were characterized by SEM and by flexural and short-beam three-point bending experiments.   The Polymer Additive Manufacturing (PAM) Series P 3D printing tool which is a fused granular fabrication (FGF) printer ( Fig. 3), was supplied by Pollen (France) [16]. The PAM machine has four extruders, allowing the manufacture of up to four different materials on a single integrated part.

3D printing w 3D printing wor orkflo kflow w
The 3D printing workflow is divided into three steps as shown in Fig. 4. First, a 3D part design was created by Catia V5 software (USA). The geometry of the specimens was designed to conform to the flexural (ISO 14125), thermomechanical (ISO 6721) and short-beam shear tests (ASTM D3410). After the object was virtually designed, a stereolithography (STL) file was exported into the 3D printer slicing software (Ultimaker Cura, USA). The 3D part was sliced into numerous sections corresponding to the layers to be printed. A G-code file was generated for the 3D printer to use it to print out each test specimen. For multimaterial profiles, two .stl files were created, one for each material.
The .stl file of the first material underwent a slicing operation with a given total thickness h = n x e (number of layers x thickness of the layer), with a linear repetition on Catia V5 with a gap of "2 x e" between each layer. The second .stl 3D printing of PLA and PMMA multilayered model polymers: an innovative approach for a b...

1024/4
file (second material) was the same as the first one but with an offset from the plane of the bed (xy plane on Catia V5) of a distance "e". These two .stl files were imported separately into the Cura software. After the nozzle for printing each material was selected, the two models were merged. This second step is the most important, because it enables the definition of the printing parameters and guarantees good-quality printing (Table 1). Next, the file was exported into the Pollen AM software (Honeyprint), which starts or stops the printing process. Beforehand, the polymer cartridge feeders were filled with the corresponding polymer granules. The main processing parameters defined for printing the PLA, PMMA and PMMA/PLA gradient parts are reported in Table 1.
T To control the temperature of the printed layers, a special blower system with a controlled flow rate of hot air was implemented. The printing chamber temperature (PCT) was measured at different locations in the manufacturing enclosure using thermocouples. The PCT was set at a fixed temperature of 35°C or 55°C.
To control the temperature history profile in the interface between printed layers, the variation over time of the temperature T(t) of a point in the path of the deposition of molten polymer on the interface between an existing layer and a newly deposited road of filament was measured using a type-K thermocouple (Fig. 5). The type-K thermocouple was embedded by slightly pressing it into an existing printed surface layer, and T(t) was recorded using a data acquisition card system. The signal output from the thermocouple was collected at a 50 Hz acquisition frequency, and time-dependent interface temperature profiles were exported for analysis. All measurements were performed using same GF printing parameters.

Eff Effect of infill density on por ect of infill density on porosity osity
First, the apparent density of 3D printing specimens was measured. The percentage of porosity is then deduced using the formula: Fig. 7 illustrates the relationship between porosity percentage and infill density (ID) for the 3D-printed samples. A drastic decrease in porosity is observed as ID increases from 30% to 80%. An infill density of 80% was therefore chosen as the optimum printing parameter value for PMMA/PLA multimaterials. from 185°C to 75°C after 2 minutes, while for PMMA and PMMA/PLA it drops from 235°C to 120°C. During the printing stages, the temperature of the layers always remained higher than the Tg of the two polymers. where |η*| is the complex viscosity modulus, R is the universal gas constant, T is the absolute temperature, and Ea is the activation energy of melt flow. Fig. 10(b) clearly shows that the shear viscosity varies inversely with the processing temperature. The activation energies are 80 kJ/mol and 140 kJ/mol for PLA and PMMA, respectively.     about 80%. This is caused by a decrease in porosity for higher infill densities, as proved earlier via pycnometry.
However, the infill density affects the failure strength to a lesser degree for PMMA than for PLA, because of the greater healing activation energy of PMMA. The σ of 3Dprinted PMMA/PLA parts is close to that of PLA. This indicates that the boundary between PLA and PMMA layers in the multimaterial part has been strengthened compared to the 3D-printed PMMA monocomponent.
The interlaminar shear stress (ILSS) values of different 3D-printed and molded parts are given in Table 3. Compared to PLA and PMMA/PLA specimens, 3D-printed PMMA exhibits lower ILSS values, indicating lower interfacial adhesion between the printed PMMA layers. The SEM images of the fracture facets made after the short-beam flexural tests show a less cohesive fracture for PMMA than for PLA (Fig. 14), in agreement with the rheological modeling discussed above.
In fact, the healing activation energy of the PMMA is double that of PMMA. More surprisingly, the ILSS of PMMA/PLA axisymmetric multilayers is between those of PLA and PMMA. This proves that the boundary interface between PLA and PMMA printing surfaces is more cohesive than the boundary interfaces in the PMMA.  The PCT t The PCT temper emperatur ature is 35°C. e is 35°C.
T Table 3. IL able 3. ILSS data of 3D-print SS data of 3D-printed and injection-molded parts. ed and injection-molded parts.
3D printing of PLA and PMMA multilayered model polymers: an innovative approach for a b...

Conclusions Conclusions
Multimaterial PMMA/PLA parts prepared by an FGF AM process were successfully developed using an original gradientconcept. The PMMA and PLA monocomponent materials were first printed in 3D to determine the optimal printing conditions. Molecular, rheological, thermal, morphological and thermomechanical measurements showed that the healing activation energy is an important factor for controlling interfacial interdiffusion between printed layers.
The flexural mechanical properties depended on the infill density but much less on the printing chamber temperature.
Interlaminar shear stress bending experiments confirmed that the macro-mechanical properties of the printed materials obtained depended on the degree of cohesion inter-layers. Interfacial cohesion is higher for PLA, which forms soft chains (low Tg), compared to the stiffer PMMA. Even more interesting, PMMA/PLA axisymmetric multilayers exhibit interfacial properties situated between those of the PLA and PMMA monolayers. The rheological modeling of the chain interdiffusion involved in the case of interfaces/interphases composed of axisymmetric PMMA and PLA chains during FGF 3D printing will be the subject of another paper in the near future. Finally, the proposed workflow in the present study can be used for manufacturing other 3D-printed multimaterials with different composition ratios and different gradient functional properties in view of numerous industrial applications.