Process–Property Correlation in Sustainable Printing Extrusion of Bio-Based Filaments

: This study investigated the effect of two critical variables for environmental process sustainability, i.e., extruder temperature and printing rate, on thermomechanical performance and accuracy in overall sample sizes, when printing bio-based materials. In this context, 3D specimens produced from basic polylactide (n-PLA) and wood-ﬁlled PLA polymer (f-PLA) were realized using extrusion-based additive manufacturing technology (MEX) by varying the nozzle temperatures (200 ◦ C, 210 ◦ C, and 220 ◦ C) and speed (from 70 mm/s to 130 mm/s). Dynamic mechanical analysis (DMA) was carried out on the produced specimens, providing information on changes in storage modulus at testing temperature of 30 ◦ C (E (cid:48) 30 ) and glass transition temperature (Tg) for each printing condition. Measurements of sample sizes allowed for printing precision considerations as a function of processing temperature and speed. The results revealed similar trends in E (cid:48) 30 changes in printed specimens at a ﬁxed extruder temperature as a function of printing speed for n-PLA and f-PLA. Infrared spectroscopy was performed on printed samples and unextruded material to attest potential material degradation under various operating conditions. Finally, images of sample surface allowed to verify the homogeneity of the diameter of the extruded material and the layer–layer contact at the interface.


Introduction
The expression Industry 4.0, coined in 2011, refers to an ongoing trend in intelligent automation technology that originated in Germany [1], and is driven by advances in data, analytics, human-machine interaction, and robotics, as well as automation to improve working conditions and plant productivity [2].
Additive manufacturing (AM) technologies are well suited for the fourth industrial revolution (Industry 4.0), embodying digital technologies, advanced materials (bio-based or nano), and more modular and configurable products [3].AM provides three "Affordability Pillars" in comparison with existing techniques [1]: (i) value stream mapping, which results in a significant decrease in waste and lead times by eliminating some phases of the manufacturing process (mold and cast realization); (ii) design for performance, which allows for easy adjustment to geometry and design changes; and (iii) bulk reduction through the use of exceptional materials that diminish part count, requiring faster assembly.
The concept of economic, environmental, and social sustainability is then inextricably linked to Industry 4.0, with an emphasis on reducing lead times, increasing energy efficiency, minimizing industrial scrap waste, and improving workplace safety and comfort [4,5].
Extrusion-based additive manufacturing (MEX) is a cutting-edge 3D printing method for processing thermoplastics [6].It is ideal for making prototypes, personalized goods, and intricate shapes, and has the potential to replace conventional methods in various industrial and practical applications from automotive and aerospace [7] to customized prostheses and medical implants [8] up to flexible textiles [9].This technology is typically separated into two main phases.The first process is virtual and involves building a computer-aided model, converting it into a standard tessellation language (STL file), slicing the 3D object into numerous layers, and configuring the printing parameters [10] (such as layer height, infill density and pattern, printing speed, air gap, build orientation, raster angle, extruder and platform temperatures, nozzle diameter, and raster width [11]).The second stage involves physically creating the 3D object.Driven motors are used to transfer a thermoplastic-based material, in the form of pellets or filaments, into a pre-heated chamber at melting or glass transition temperatures.Molten materials are then ejected from a nozzle and deposited layer-by-layer over a heated substrate using movements that correspond to the geometric pattern of the 3D item [6].
However, all of these features that promote the sustainability of MEX technology may have a detrimental impact on the mechanical performance, aesthetics, and durability of the final 3D-printed goods.In a recent review [12], the advantages and disadvantages of each sustainable solution for promoting the environmentally friendly growth of MEX technology were discussed.The use of biopolymers to replace conventional petroleum-based counterparts increased biocompatibility while also resulting in lower mechanical properties, lower resistance to thermal and/or hydrolytic degradation under stress conditions and moisture in the environment.The addition of natural fibers to polymer resins reduced the overall printability of the systems by increasing viscosity, resulting in duct clogging or voids in specimens.The reduction in emissions and energy consumption has been associated with printing settings that reduced processing time, increased printing speed, and lowered processing temperatures.However, lowering the printing temperature and processing time frequently resulted in poor mechanical product properties and printing accuracy.
In this framework, two factors of relevance in sustainable printing process settings can be identified, which are critical for shorter processing times, reduced energy usage, and lower emissions: nozzle temperature and printing speed.Regarding this, a study was conducted to determine how much it is possible to push printing toward sustainable development by using biopolymers and selecting appropriate printing variables without compromising the performance and precision of the final printed products.The thermomechanical characteristics of two bio-based polymers (unfilled and wood-filled PLA) and their relationship to printing speed and nozzle temperature were investigated through dynamic mechanical analysis (DMA).The sample sizes were measured to gain information on printing accuracy for each printing setting.Images of sample surface and spectroscopic measurements were performed to support the discussion in terms of printing defects, bond consolidation, and material degradation.

Materials
Two commercially available filaments with a nominal diameter of 1.75 mm from Eumakers (Barletta, Italy) were chosen for testing.The first was an unfilled neat resin of polylactic acid (here referred to as "n-PLA"), whereas the second was fir wood-PLA composite (referred to as "f-PLA").According to the manufacturer, filaments were cre-ated using Natureworks' Ingeo biopolymer 4032D.Based on previous thermogravimetric analysis (TGA), the wood content in the composites could be around 20 wt.% [26].

Sample Preparation
According to previous investigation [26], the filaments were dried in an oven at 80 • C for 4 h to prevent hydrolytic deterioration.The printing was carried out using a Zortax 3D printer (cod.M200) (Olsztyn, Poland).The infill density was set at 100%, and the support temperature of 70 • C was established.To ensure proper adherence of the first layer of extruded filament to the platform, the temperature should surpass the material's glass transition temperature (Tg) (~60 • C for both filaments [26]).As attested in [27], heating the support bed slightly above the Tg of material filaments improved sample adhesion to the printer platform.
A 0.4 mm nozzle (the smallest) was employed to achieve the lowest possible layer thickness for each material.Based on preliminary printing attempts, depending on the filament type (n-PLA or f-PLA), different retraction movements (distance and speed) and layer thickness were established.
Harsh operating conditions, such as setting the lowest layer thickness and increasing retraction motions (speed and distance), may improve the printing quality of 3D objects while reducing the amount of material ejected from the nozzle.In the case of biocomposite printing, this could lead to filler accumulation and/or misalignment in the orifice, preventing material and polymer chains from flowing smoothly out the nozzle.To ensure a constant flow from the printing extruder, f-PLA printing required smaller retraction motions and thicker layers than n-PLA printing.Retraction speeds of 27 and 20 mm/s, as well as retraction distances of 2.7 and 1 mm/s, were set for n-PLA and f-PLA, respectively.The layer thicknesses for n-PLA and f-PLA were 0.09 mm and 0.19 mm, respectively.
Different printing speeds ranging from 70 mm/s to 130 mm/s were selected at nozzle temperatures of 200 • C, 210 • C, and 220 • C (see Table 1).To establish the printing speed, a 30% reduction and increase were considered, resulting in 70 and 130 mm/s, respectively, compared to the default printer setting of 0% (100 mm/s).Rectangular 3D specimens, 2 × 5 × 25 mm 3 in size, were developed to perform dynamic mechanical analysis.All the 3D-printed specimens were constructed on top of a preparatory printed structure (also known as "raft") to limit warping and detachment of first layers from the heated plates, and allow for the same interface to all layers.The raft was removed manually following printing.At a given geometry (nominal extruded mass of 5 g), raising the printing speed from 70 to 130 mm/s reduced the process time by around 30-35% (Table 1).
At 200 • C, the printing speed could be no more than 115 mm/s.This was explained by the previous work by Patti et al. [28], on rheological measurements of PLA-based polymers and experimental printing evidence.The authors concluded that a specific combination of printing speed and melt viscosity, which is directly impacted by extruder temperature, played a crucial role in avoiding printing nozzle failure.
At 200 °C, the printing speed could be no more than 115 mm/s.This was explained by the previous work by Patti et al. [28], on rheological measurements of PLA-based polymers and experimental printing evidence.The authors concluded that a specific combination of printing speed and melt viscosity, which is directly impacted by extruder temperature, played a crucial role in avoiding printing nozzle failure.

Dynamic Mechanical Analysis (DMA)
DMA is a commonly used and versatile polymer characterization method that offers information on the material's composition, physical behavior, and viscoelastic properties.The samples are subjected to sinusoidal load (stress or strain) and the corresponding response (strain or stress) is recorded, while the temperature is changed at a constant rate.Under linear conditions, stress (σ) and strain (ε) are described by sinusoidal function of time (t): where  is the frequency,  is the stress amplitude,  is the strain amplitude, and δ is the phase shift between the stress signal and the strain signal (δ = 0° corresponds to a perfect elastic; δ = 90° to a pure viscous; 0° < δ < 90 for a viscoelastic material).A complex modulus (E*) can be defined as follows: E* includes two contributions: an elastic part (storage modulus: E′) and a viscous part (loss modulus: E″) described by the following equations: It is worth mentioning that the complex modulus measured in DMA is not identical to Young's modulus.Young's modulus is measured using uniaxial stress, while complex

Dynamic Mechanical Analysis (DMA)
DMA is a commonly used and versatile polymer characterization method that offers information on the material's composition, physical behavior, and viscoelastic properties.The samples are subjected to sinusoidal load (stress or strain) and the corresponding response (strain or stress) is recorded, while the temperature is changed at a constant rate.Under linear conditions, stress (σ) and strain (ε) are described by sinusoidal function of time (t): where ω is the frequency, σ 0 is the stress amplitude, ε 0 is the strain amplitude, and δ is the phase shift between the stress signal and the strain signal (δ = 0 • corresponds to a perfect elastic; δ = 90 • to a pure viscous; 0 • < δ < 90 for a viscoelastic material).A complex modulus (E*) can be defined as follows: E* includes two contributions: an elastic part (storage modulus: E ) and a viscous part (loss modulus: E ) described by the following equations: It is worth mentioning that the complex modulus measured in DMA is not identical to Young's modulus.Young's modulus is measured using uniaxial stress, while complex modulus is estimated using small mechanical oscillatory solicitations.However, DMA is currently becoming widely used to characterize mechanical properties of heterogeneous materials, analyze parameters for quality control, for research and development, and to determine optimal processing settings [29].
Here, the Tritec 2000 DMA machine from Triton Technology Ltd. in Leicestershire (UK) was employed to perform dynamic mechanical analysis on 3D specimens.A single cantilever holder was used for testing in air.The sample span length was 10 mm, the am-plitude was 0.05 mm, and the frequency was set at 1 Hz for each sample.The temperature was increased from room to 75 • C at a rate of 2 • C/min.Three specimens were measured for each printing condition.

Spectroscopic Measurements
Fourier transform infrared spectroscopy-attenuated total reflectance (FTIR-ATR) was used to determine the presence and absence of specific functional groups, the chemical structure of polymer materials, and the resulting alterations.Spectra were recorded using Perkin Elmer Spectrum 65 FTIR spectrometer (Waltham, MA, USA).Measurements were taken on the surface of virgin filaments and printed samples in a wavenumber range of 4000-650 cm −1 with 16 scans each measurement and a camera resolution of 4 cm −1 .

Visual Analysis of the Sample Surface
The sample surface (bottom and lateral side) was inspected using a USB digital microscope with a magnification of 1600× and manual focal length (640 × 480 resolution).

Accuracy in Sample Sizes
Tables 2 and 3 reported the average sizes of prismatic 3D specimens according to measurements described in Section 2.3.1 at different nozzle temperatures and printing speeds for n-PLA and f-PLA filaments, respectively, and corresponding errors according to Equation ( 6): Error = measured value − design value (6) Sample dimensions had discrepancies of roughly 200-300 microns compared to nominal measurements.At 200 • C, the n-PLA samples had W and L slightly below the nominal values, whereas for f-PLA, the same dimensions exceeded the nominal values.As the temperature rose, the dimensions of n-PLA increased, whereas those of f-PLA decreased.However, these effects converged in the same result: when the temperature rose, the W and L dimensions approached their nominal values.Increasing the speed resulted in an overall reduction in W and L values, with an effect more evident for n-PLA.The thickness showed the greatest variation in dimensional errors, especially in the case of pure polymer.In the case of f-PLA, inaccuracies in thickness were comparable to those in other dimensions.Errors in thickness could be caused by a variety of factors, including the printer's skill, material type, and operating conditions, as well as uneven raft removal.Eliminating the raft resulted in some of the beads sticking to the sample, as well as a potential loss of 1-2 layers.
Thus, focusing only on W and L for each prismatic sample to determine the ideal printing settings for precise sample sizes, n-PLA dimensions appeared to agree with standards at 210-220 • C and 100-115 mm/s, and f-PLA dimensions aligned with standards at 220 • C with a slight influence on speed.

Dynamic Mechanical Analysis (DMA)
Figure 2 shows typical curves of storage (E ) modulus (elastic response) and loss modulus (E ) (viscous response) vs. testing temperature for specimens produced at 210 • C. By varying the nozzle temperatures and printing speed, the thermomechanical properties of 3D specimens made from n-PLA and f-PLA remained qualitatively unchanged.Analogous considerations might be qualitatively derived from the experimental curves, regardless of material and operating conditions.
The storage modulus (E ) corresponds to the stiffness of the system, while the loss modulus (E ) is correlated to the dissipation of energy due to viscous behavior.These characteristics fluctuate dramatically with temperature and frequency, particularly around polymer relaxation events, such as for the glass transition, during which the storage modulus lowers, the loss modulus increases [30].
In both cases (n-PLA and f-PLA), when the testing temperature increased from room temperature to 75 • C, the storage modulus (E ) trended downward.After an initial constant zone, the parameter decreased by three or two orders of magnitude (depending on n-PLA or f-PLA) when the temperature rose over 55 • C.This was linked to the evolution from   3 showed average E′ values at testing temperature of 30 °C (E′30) as a function of printing speed at three different nozzle temperatures for both n-PLA and f-PLA.
In this case, comparable considerations could be drawn for both materials.At a nozzle temperature of 200 °C (green bars), starting from 70 mm/s, increasing the printing speed caused E′30 to initially decline and then grow.Both n-PLA and f-PLA showed comparable values at 200 °C and 70 mm/s or 130 mm/s.At 210 °C (light blue bars), positive and negative E′30 changes occurred alternately.However, under these conditions, Increasing the testing temperature increased the loss modulus (E ) of f-PLA by one order of magnitude and even more for n-PLA.Then, the parameter peaked at roughly 60 • C and dropped by over three orders of magnitude for n-PLA and two orders of magnitude for f-PLA.The loss modulus is usually linked to internal friction and is affected by several factors such as molecular movements, transitions, relaxation processes, morphology, and structural heterogeneity [31].As the temperature approached the glass transition, the material became more deformable; the stiffness of the material was reduced while boosting the mobility of the large polymer segments.This facilitated energy dissipation via friction (E rose).When the glass transition was exceeded, the material was easier to deform, resulting in less energy dissipated by macromolecular motions (E decreased).
The peak of the loss modulus is conventionally associated with the glass transition temperature (Tg).The glass transition remained around 60 • C both for n-PLA and f-PLA (Table 4).A slight drop in glass transition temperature (approximately 2 • C) of n-PLA at higher nozzle temperatures (220 • C) was observed.Changes in viscoelastic properties are greatly affected by segment motions and chain mobility.As a result, any factor that influences macromolecular mobility and changes in molecular structure (degradation, aging, crystallinity, adsorption of small molecules, etc.) has an impact on relaxation processes and, more broadly, on viscoelastic features [30].
The decrement in Tg was attributed to thermal degradation events and chain scission mechanisms, which could not be ruled out in biopolymers subjected to thermomechanical stress during the printing process.In f-PLA, the addition of the filler slowed the polymer chain motions, limiting the impacts of degradation at elevated nozzle temperatures (220 • C) on the Tg, which remained constant.

Spectroscopic Analyses
Polymer heat resistance and deterioration under extreme conditions have been studied using a variety of techniques, including X-ray diffraction, microscopes, thermal analyzers (TGA, DSC, and DMTA), rheological measurements, and spectroscopy (infrared, near-infrared, Raman, ultrasound, and UV-vis) [32].Evaluations of the molar mass, glass transition, and mechanical properties of extrudates were conducted to assess the impact of processing settings (screw speed and processing temperature) on poly(L-lactic acid) (PLLA) degradation during melt extrusion [33].
PLA's thermal degradation involves multiple pathways, including random main scission reaction, intramolecular and intermolecular transesterification, hydrolysis, pyrolytic elimination, and hydrogen transfer.Intermolecular transesterification alters the order of chain portions while intramolecular transesterification produces cyclic oligomers of lactic acid and lactide [34].Hydrolysis of the ester linkage results in a random chain scission and depolymerization that reduces molar mass and can be accompanied by a significant increase of carboxyl (COOH) and/or hydroxyl end (OH) groups [32,35].Pyrolytic elimination produces molecules with conjugated double bonds (C=C) from the carbonyl group (C=O) [33].The hydrogen transfer process involves in the attachment of the C=O double bond, producing a vinyl ester (CH2=CH-) and acid end groups (COOH) [35].In this case, comparable considerations could be drawn for both materials.At a nozzle temperature of 200 • C (green bars), starting from 70 mm/s, increasing the printing speed caused E 30 to initially decline and then grow.Both n-PLA and f-PLA showed comparable values at 200 • C and 70 mm/s or 130 mm/s.At 210 • C (light blue bars), positive and negative E 30 changes occurred alternately.However, under these conditions, the E 30 values at 130 mm/s were higher than those at 70 mm/s.At 220 • C (red bars), E 30 showed an increasing trend as printing increased, albeit with a slight effect.

Spectroscopic Analyses
Polymer heat resistance and deterioration under extreme conditions have been studied using a variety of techniques, including X-ray diffraction, microscopes, thermal analyzers (TGA, DSC, and DMTA), rheological measurements, and spectroscopy (infrared, nearinfrared, Raman, ultrasound, and UV-vis) [32].Evaluations of the molar mass, glass transition, and mechanical properties of extrudates were conducted to assess the impact of processing settings (screw speed and processing temperature) on poly(L-lactic acid) (PLLA) degradation during melt extrusion [33].
PLA's thermal degradation involves multiple pathways, including random main scission reaction, intramolecular and intermolecular transesterification, hydrolysis, pyrolytic elimination, and hydrogen transfer.Intermolecular transesterification alters the order of chain portions while intramolecular transesterification produces cyclic oligomers of lactic acid and lactide [34].Hydrolysis of the ester linkage results in a random chain scission and depolymerization that reduces molar mass and can be accompanied by a significant increase of carboxyl (COOH) and/or hydroxyl end (OH) groups [32,35].Pyrolytic elimination produces molecules with conjugated double bonds (C=C) from the carbonyl group (C=O) [33].The hydrogen transfer process involves in the attachment of the C=O double bond, producing a vinyl ester (CH 2 =CH-) and acid end groups (COOH) [35].
Fourier transform infrared spectroscopy (FTIR) is a well-established method for analyzing the conversion of functional groups in polymer materials, enabling the assessment of chemical and physical transformations such as ageing, interactions, crystallinity, and resin and composite cure [32].Here, attenuated total reflectance spectroscopy (ATR) was used to qualitatively analyze potential material degradation under various printing speeds and temperature settings.In all cases, spectra were normalized to an internal standard for PLA polymer (wavenumber: 1455 cm −1 ), which corresponds to methyl group vibration (CH 3 ) [18].
Figure 4 compares absorbance intensity as a function of wavenumber (cm −1 ) for unextruded filament and printed samples.For n-PLA, at a process temperature of 200 °C (Figure 4a), the sample printed at a higher speed (115 mm/s, green line) exhibited an absorption spectrum that was remarkably similar to the non-extruded material (blue line).Then, compared to non-extruded material, the spectrum of the sample produced at lower speeds (70 mm/s, red line) revealed increased intensities in the range of 3000-3800 cm −1 , corresponding to the stretching vibration of hydroxyl group (υO-H), and in the range of 1600-1650 cm −1 , attributed to bending vibration in the OH group.On the other hand, the signals related to the carbonyl group vibration at 1750 cm −1 (υc=o) and vibration of ester group at 1183 cm −1 (υc-o) were signifi- For n-PLA, at a process temperature of 200 • C (Figure 4a), the sample printed at a higher speed (115 mm/s, green line) exhibited an absorption spectrum that was remarkably similar to the non-extruded material (blue line).Then, compared to non-extruded material, the spectrum of the sample produced at lower speeds (70 mm/s, red line) revealed increased intensities in the range of 3000-3800 cm −1 , corresponding to the stretching vibration of hydroxyl group (υ O-H ), and in the range of 1600-1650 cm −1 , attributed to bending vibration in the OH group.On the other hand, the signals related to the carbonyl group vibration at 1750 cm −1 (υ c=o ) and vibration of ester group at 1183 cm −1 (υ c-o ) were significantly diminished.These findings were interpreted as indicators of ester bond hydrolysis, which resulted in the formation of additional hydroxyl end groups [18].Then, the increase in absorbance intensity at 1635 cm −1 (υ c=c ), potentially caused by the vibration of a carbon double bond, at 2920 and 2855 cm −1 (υ C-H ), attributed to symmetric and asymmetric stretching vibrations of CH 2 [36], could suggest the presence of another degradation mechanism, such as pyrolytic elimination or hydrogen transfer [18].Similar considerations can be made for n-PLA printed at a process temperature of 220 • C (Figure 4b).Both specimens at different printing speeds (70 mm/s and 130 mm/s) showed signs of polymer deterioration, more relevant in the case of lower printing speed (70 mm/s, red line).
The spectra of unextruded material and specimens, made from f-PLA, printed at 200 • C (Figure 4c) and 220 • C (Figure 4d) revealed the same intensity variations (increments/decrements) as those reported for n-PLA at the same wavenumbers.This result was interpreted as the occurrence of degrading events in f-PLA samples similar to those documented earlier for printed n-PLA specimens.At 200 • C (Figure 4c), the intensity changes for a speed of 70 mm/s (red line) were more significant than those obtained at 115 mm/s (green line).Thus, at low processing temperatures, the deteriorating effect of printing on material appeared to be more relevant when the printing motions were slow.On the contrary, the spectra of printed specimens at 220 • C (Figure 4d) were fairly similar, indicating that the speed had a small effect on polymer degradation at higher temperatures.
The bands at 871 and 756 cm −1 were often assigned to PLA's amorphous and crystalline phases [18,24].In all cases, adsorption peaks at 756 cm −1 and 871 cm −1 were substantially equal, indicating no significant changes in crystallinity during the printing process.

Visual Aspects of Sample Surfaces
Figures 5 and 6 show surface pictures (bottom and lateral side) of samples manufactured from n-PLA and f-PLA under various processing settings.The visual inspection of the sample bottom highlighted areas where the first layers were accidentally removed during the raft asportation or areas where raft beads have remained attached.These parts of surfaces allowed for observing differences in the diameter of the extruded material on heated bed support, such as voids, or defects, whereas those of the lateral surface revealed information on layer-by-layer adhesion.
The worst conditions for significant filament diameter change in n-PLA specimens were discovered at 200 • C and 115 mm/s (Figure 5c).At low temperatures (related to decreased viscosity) and high speeds, the filament's integrity was compromised, resulting in diameter irregularities.On the other hand, raising the speed and temperature caused a progressive rise in bond consolidation at the layer-layer interface.Starting from conditions of 200 • C and 70 mm/s (Figure 5b), the lateral surface of the sample presented with a deposition of layers that were rather parallel to one another.Then, by raising the speed, the distinction between layers was weakened and phenomena of interpenetration of the layers became more visible (Figure 5d), with an increasing effect as the extrusion temperature increased (Figure 5f,h).
The same observations made in the case of n-PLA were also collected in the case of f-PLA.At 200 • C and 70 mm/s, defects in the form of voids (Figure 6a) and filament diameter irregularities (Figure 6b) were detected.The same problems become more pronounced when the speed was increased to 200 • C and 115 mm/s (Figure 6c,d).Increased processing temperatures resulted in greater filament diameter homogeneity and enabled the elimination of voids (Figure 6e-h).In terms of layer-to-layer contact at the interphase, stratification was clearly visible in all cases, with no evidence of strong interpenetration between adjacent layers.In f-PLA specimens, the impact of printing speed was less significant than in n-PLA specimens, and the printing temperature appeared to be the primary driver of sample morphology.The same observations made in the case of n-PLA were also collected in the case of PLA.At 200 °C and 70 mm/s, defects in the form of voids (Figure 6a) and filament diamete irregularities (Figure 6b) were detected.The same problems become more pronounce when the speed was increased to 200 °C and 115 mm/s (Figure 6c,d).Increased processin temperatures resulted in greater filament diameter homogeneity and enabled the elim nation of voids (Figure 6e-h).In terms of layer-to-layer contact at the interphase, stratifi cation was clearly visible in all cases, with no evidence of strong interpenetration betwee adjacent layers.In f-PLA specimens, the impact of printing speed was less significant tha in n-PLA specimens, and the printing temperature appeared to be the primary driver o

Discussion
This study examined the effect of two process variables (printing time and nozzle temperature) on the dimensional changes and thermomechanical properties of polylactidebased materials, highlighting the challenges of forecasting the process-property relationship due to multiple interconnected and opposing factors.
The impact of temperature and printing speed on the dimensional accuracy and mechanical properties of 3D-printed items has been the subject of conflicting findings in the literature.
Extrusion temperature is crucial to dimensional accuracy of printed parts since it affects the material viscosity and its ability to extrusion.An extruder temperature that is too high can result in excessive material fluidity and dripping, while an extruder temperature that is too low makes the material difficult to work with and print [37].Speeding up the printing process, improving the printing speed, can result in falls and vibrations that lower the quality of the printed specimens [38].According to Ansari and Kamil [39], at higher extrusion temperatures and lower print speeds, the dimensional variations of 3D items made of PLA were reduced.Buj-Corral et al. [38] reported that low temperature prevented the material from becoming overly fluid, resulting in the lowest dimensional inaccuracy; then, high layer height, high temperature, and low speed produced the highest roughness values.Alafaghani et al. [40] observed a negative dimensional error on the overall length in PLA specimens with an increase in speed (70 mm/s−170 mm/s) and temperature (175−205 • C), while the errors on the width were positive with a minor effect of the speed.They also suggested setting a lower temperature to increase the dimensional accuracy.De Freitas and Pegado [41] showed a decrease in average length and width error values in low-density PLA specimens when nozzle temperature increased from 220 and 250 • C. Alsoufi et al. [42] analyzed the effect of nozzle temperature (from 205 • C to 250 • C, speed of 30 mm/s) in 3D parts made from two different types of PLA.The dimensions throughout the width and length were always lower than those of the CAD model due to warping/shrinking effects.The authors suggested that higher nozzle temperatures resulted in smaller distorted shape mistakes and minimal uncertainty.
In this case, as the speed increased and the temperature reduced, the W and L dimensions of n-PLA decreased in comparison to the CAD model.For f-PLA, the dimensional errors on W and L were all positive and decreased as the temperature increased, with small effects of speed.Aside from the fact that f-PLA was printed under different conditions in layer thickness and retraction movements compared to n-PLA, this result could be due to the different effect of temperature on the rheological behaviour of melted materials (viscosity trend, flow instability, die swell), as well as a different shrinkage/warping tendency when particles were incorporated into the polymer matrix.Reduced nozzle temperature and increased printing speed (shorter cooling time) could result in higher levels of internal stresses, triggering warpage events in PLA specimens.The inclusion of wood particles could have exerted a stabilizing effect on the microstructure of the molten material, minimizing the stresses experienced during cooling.
Using the raft to minimize the warping phenomena did not eliminate warpage distortion; rather, it caused a large error in thickness due to the difficulty in separating the two surfaces (raft and sample) downstream of the printing process.
The nozzle temperature may have an opposite effect on the mechanical properties of a 3D-printed element.On the one hand, too low temperatures can provide inadequate fluidity in the melt material, which not properly attach to the deposited layer once extruded, resulting in reduced mechanical performance.On the other hand, excessively high temperatures can affect thermal stability of materials, resulting in polymer chain breakdown [43].
Similarly to temperature increment, an increase in printing speed might have an opposite effect on the mechanical properties of 3D specimens.Dou et al. [44] reported that the tensile strength of 3D-printed parts made from continuous carbon fiber-reinforced polylactic acid (PLA) composites was decreased by increasing the printing speed from 50 to 400 mm/s.In the work of Rezaeian et al. [45], the testing results revealed that specimens made from acrylonitrile butadiene styrene (ABS) printed at 70 mm/s outperformed alternative lower nozzle speeds (10, 30, and 50 mm/s) in terms of elongation and fracture resistance.Finally, Wang et al. [46] showed that increasing printing speed from 17 to 26 mm/s resulted in non-monotonous variations in tensile characteristics, and surface roughness of printed thermoplastic polyetheretherketone (PEEK).For a 0.4 mm nozzle, tensile strength reached the maximum at 20 mm/s and then decreased, while layer roughness reached its lowest at 23 mm/s and then increased.The authors concluded that printing speeds might affect the contact time between the new layer and the previously deposited layer, which might be insufficient to allow for macromolecules to diffuse and form an effective connection.
The impact of processing temperature and speed on mechanical performance of two printed materials was primarily assessed as a change in storage modulus at testing temperature of 30 • C (E 30 ), measured through DMA.This parameter appeared to vary consistently throughout different operating circumstances, regardless of material (Figure 3).Overall, increasing the nozzle temperature and doubling the speed resulted in higher E 30 values.
By increasing printing speed, at a nozzle temperature of 200 • C, E 30 was initially reduced before improving; at nozzle temperature of 210 • C, positive and negative changes in E 30 alternated; at nozzle temperatures of 220 • C, E 30 values increased, albeit with a small effect.
These variations were attributed to opposing phenomena happening as a function of printing speed and nozzle temperature.
The lowest nozzle temperature impacted the fluidity of the material, resulting in poor sample filling, defects, and irregularities in diameter of deposited layers (Figures 5c and 6c) and/or higher mechanical stress acting on polymers, especially when printing speed increased.However, shortening the printing time (i.e., raising the speed) reduced the sample's residence time in the molten chamber, resulting in potential less polymer chain breakdown (as confirmed in Section 3.3).An increase in printing speed could also result in a reduction in the cooling time of the deposited layer before the next one was deposited on it.This could imply improved conditions for macromolecules diffusion at the interface, as well as increased adhesion at the layer-to-layer contact surface (as confirmed in Section 3.4), particularly evident for neat polymer.
Higher temperatures enhanced material fluidity and sample filling, improved the layer-to layer contact, and reduced the mechanical stress acting on the polymer during printing, but they might also promote thermal breakdown.When the speed was raised, elevated nozzle temperatures avoided flaws and filament diameter anomalies, resulting in a slightly altered sample microstructure, while shorter printing times limited polymer chain degradation events, and promoted layer adhesion.

Conclusions
This study aimed to find a next step in pushing the MEX process toward sustainable development by utilizing bio-based materials (unfilled and wood-filled PLA) and appropriate printing parameters that reduced processing time and temperatures while maintaining the mechanical performance and printing accuracy of developed goods.Final considerations were derived using sample measurements (prismatic geometry) and thermomechanical data (small deformations).
Measurements of the sample sizes as the working conditions changed revealed 210-220 • C and 100-115 mm/s speeds (especially for the n-PLA) as the best for the dimensional accuracy of the 3D parts.
A similar trend in storage modulus at 30 • C (E 30 ) of samples produced at the same extrusion temperatures was confirmed by varying the printing speed for both materials.In this instance, a rise in the extrusion temperature led to an increase in mechanical properties, with a more pronounced effect as the printing rate increased.
Temperature and speed were two process parameters that had a significant impact on the final properties of the 3D product, influencing material degradation phenomena (as attested through ATR), melt fluidity, and welding between deposited neighboring layers (as attested through images of sample surface).Increasing the nozzle temperature decreased melt viscosity, resulting in better filament diameter homogeneity, less fraying, roughness, and defects; then, it worsened thermal breakdown events in polymer chains while improving connectivity at contact layers, particularly in the case of neat polymer.Regardless of material, increasing the speed could have a negative impact at low processing temperatures on the uniformity of the filament deposited on the support, promoting flaws and defects.However, the increased speed lowered the material's residence time in the extruder chamber (i.e., minimizing degradation phenomena), as well as the cooling time of the deposited layer before the next one was deposited on it (i.e., promoting adhesion).
Finally, beneficial printing conditions can be achieved at medium-high printing speeds and elevated temperatures both for n-PLA and f-PLA.
Funding: This research received no external funding.

Figure 1 .
Figure 1.Position of measurements on 3D specimens.Three measurements of width: W1, W2, and W3.Three measurements of the thickness: T1, T2, and T3.Three measurements of the length: L1, L2, and L3.W, T, and L represent the average width, thickness, and length of each prismatic specimen.

Figure 1 .
Figure 1.Position of measurements on 3D specimens.Three measurements of width: W1, W2, and W3.Three measurements of the thickness: T1, T2, and T3.Three measurements of the length: L1, L2, and L3.W, T, and L represent the average width, thickness, and length of each prismatic specimen.

Figure 2 .
Figure 2. Representative curves of storage (E′) and loss (E″) moduli of 3D specimens made from neat PLA (n-PLA) and composite (f-PLA) filaments printed at 210 °C and varied printing speeds.

Figure 2 .
Figure 2. Representative curves of storage (E ) and loss (E ) moduli of 3D specimens made from neat PLA (n-PLA) and composite (f-PLA) filaments printed at 210 • C and varied printing speeds.

Figure 3
Figure 3 showed average E values at testing temperature of 30 • C (E 30 ) as a function of printing speed at three different nozzle temperatures for both n-PLA and f-PLA.

Figure 3 .
Figure 3.Effect of printing speed on changes in E′30 in printed samples at different nozzle temperatures (200 °C, 210 °C, and 220 °C).

Figure 3 .
Figure 3.Effect of printing speed on changes in E 30 in printed samples at different nozzle temperatures (200 • C, 210 • C, and 220 • C).

Figure 4 .
Figure 4. Normalized ATR spectra of unextruded n-PLA and f-PLA and printed samples at 200 °C and 70 mm/s or 115 mm/s: (a) and (c); 220 °C and 70 mm/s or 130 mm/s: (b) and (d).

Figure 4 .
Figure 4. Normalized ATR spectra of unextruded n-PLA and f-PLA and printed samples at 200 • C and 70 mm/s or 115 mm/s: (a) and (c); 220 • C and 70 mm/s or 130 mm/s: (b) and (d).

Table 1 .
Technological parameters in the printing process to prepare specimens.The printer management software computed the printing time.

Table 2
. T, W, and L as a function of printing speed at various nozzle temperatures and errors in mm on T, W and L compared to nominal size for n-PLA.

Table 3 .
T, W, and L as a function of printing speed at various nozzle temperatures and errors in mm on T, W and L compared to nominal size for f-PLA.

Table 4 .
Glass transition temperature (Tg) for n-PLA and f-PLA as function of operating conditions.

Table 4 .
Glass transition temperature (Tg) for n-PLA and f-PLA as function of operating conditions.