1 Introduction

Additive manufacturing (AM) technologies have gained an increasing industrial interest for the production of both prototypal and commercial components and are becoming a key aspect of the Industry 4.0 transition [1]. AM offers strong environmental impact reduction opportunities by enabling the production of complex and high-quality parts, with several possible materials and minimum waste production [2,3,4,5]. The advantages of AM can also be paired with composite materials, adding reinforcement phases in polymeric, ceramic, or metallic matrixes; this allows the customization of the mechanical and physical properties of the materials to suit specific applications [6,7,8].

Among the available technologies, one of the most used in industrial applications is vat photopolymerization (VP) in which 3D parts are built layer by layer through the cross-linking of a photocurable resin under the action of a light source (e.g., laser or LED lights) [7, 9]. Its large use in different sectors, including mechanical, biomedical, and electronics [10], has prompted studies for tailored optimization of the properties of the resins used as raw materials. Several analyses can be found in the scientific literature concerning the use of filler particles and nanoparticles in resins (e.g., graphene, graphene oxide, nanofibers) to improve their mechanical [11] and electrical [12] properties.

For example, Saadi et al. investigated the mechanical and piezoresistive behavior of VP resin with low contents of carbon nanotubes (CNT), proving that 3D printing can be a valuable solution for multifunctional system production [13]. Graphene nanocomposites with filler alignment were produced using VP technology by Markandan et al. They promoted part anisotropy and investigated the filler concentration that maximize the part properties [14]. High-strength nanocomposites reinforced with graphene oxide were also investigated by Manapat et al. [15]. Fu et al. investigated the production of dielectrically functional gradient materials by adding ceramic functional fillers into a UV-curable resin matrix [16]. Antibacterial properties, along with improved mechanical response, were observed by Vidakis et al. as copper nanoparticles were added in raw resin [17]. Typically, small percentages of reinforcement can strongly improve the properties of raw resins for VP applications, resulting in high-performance composite materials [18, 19].

From an environmental perspective, raw resins and reinforcement materials can be an issue due to high unitary impact values (for example, in terms of equivalent CO2 emissions) [20, 21]. Efforts have been made to improve the raw material sustainability, mainly by considering bio-based alternatives to the fossil-based resins [22, 23] and natural fillers [24]. In addition, considering other 3D printing technologies such as fused deposition modeling (FDM), literature analyses concerning recycled matrixes and fillers can be found [25, 26]. However, literature lacks of studies focused on the use of recovered materials as filler for VP resins with the goal of reducing the carbon footprint of printed components. Moreover, no sustainability assessment of improved 3D printing with such systems can be found.

In this context, the present study investigates the use of recovered polyamide 12 (PA12) powder as a filler for 3D printing photocurable resins. This thermoplastic material is characterized by good mechanical properties (high ductility and toughness), flame resistance, and low weight [27, 28]. For these reasons, it is widely used for several industrial applications. Hence, in order to improve a sustainable development of 3D printing, efficient ways to reuse this plastic powder are required to minimize the quantity of powder waste generated.

Different powder percentages were considered (up to 10%) to assess the feasibility of the powder reuse and to investigate the effects of the filler on the mechanical performances and the environmental sustainability of 3D printed parts. In order to do so, specimens were produced by means of a LCD vat photopolymerization process and tensile tests were carried out to evaluate their mechanical properties. The life cycle assessment methodology was employed to evaluate the environmental sustainability of the proposed recovery process. Scanning electron microscopy (SEM) and computed tomography (CT) image analyses were also carried out. The ultimate goal is to develop new recovered materials with improved mechanical properties and low carbon footprint, contributing to make industrial 3D printing processes more sustainable.

2 Materials and method

2.1 Materials and specimen preparation

The material used in the present investigation was the ELEGOO translucent standard commercial LCD photocurable resin, used in vat photopolymerization AM technology; according to the material datasheet, it is mainly composed by epoxy acrylate resin and monomers and about 3–5% in weight of photoinitiators. The translucent resin allows to enhance the contrast with the dark filler powder. In order to evaluate the effect of the plastic filler on the mechanical properties of the resin, the Sinterit polyamide 12 (PA12) industrial powder for selective laser sintering (SLS) was used. The material datasheet reports a tensile modulus of 1840 MPa and a tensile strength of 52.3 MPa. The powder was recovered from the LISA PRO SLS machine after a total of five printing jobs were carried out. The powder has a refreshing ratio equal to 30%. No fresh powder was added after the fifth printing job to have only recovered powder used as a filler. Nine different material configurations were prepared and used to fabricate the samples with different weight content filler powder. Starting from the neat resin, increasing filler content was added to the resin up to 10% in weight (0.5%, 1%, 1.5%, 2%, 2.5%, 5%, 7.5%, 10%). For each configuration, raw resin and PA12 powder were weighted to achieve the target mix weights and were mixed together by means of a mechanical stirrer [29]. After that, the obtained filled resin was poured into the SLA machine tank and let sit for 1 h before the printing process to allow trapped gasses to escape. The ELEGOO Saturn 2 LCD 3D printer is based on the masked stereolithography (MSLA) technology, in which LED lights are masked by a selectively transparent LCD screen to precisely cure resin layers.

Irrespective of the filler weight content, tensile specimens were obtained with the following process parameters: bottom exposure equal to 60 s for the first 5 layers, normal exposure of 8 s, and layer height equal to 0.05 mm.

The simple geometry of the specimens did not require the use of support material. After the SLA process, the specimens were washed in isopropyl alcohol for 5 min to completely remove uncured resin excess and post-cured via UV lights for 30 min to ensure complete curing of the parts. The dedicated washing and curing machines (ELEGOO Mercury X Bundle) were used.

2.2 Tensile test

Dog bone specimens were manufactured to assess the tensile properties of both pure and powder filled resin in accordance with the ASTM D638-22 standard for plastic (type IV specimens). The samples were characterized by a constant nominal thickness, gauge length, and width equal to 3.6, 33, and 6 mm, respectively. At least five specimens for each configuration were tested using the MTS 810® universal servo-hydraulic testing machine. Tensile tests were carried out at a constant crosshead speed of 5 mm/min. During tests, the load and strain along the loading direction were acquired using a load cell of 50 kN and an extensometer; they were then plotted as tensile stress–tensile strain curves, by which Young modulus, ultimate tensile strength (UTS), and strain to failure were derived as a function of the filler powder weight content.

2.3 SEM and CT

Fracture surfaces of the tensile specimens were observed by means of the SEM-FEG Zeiss Supra 40 electron scanning microscope with an electron high tension (EHT) of 5.00 kV. Fracture mechanism and filler behavior within the resin were investigated at different magnifications (ranging from 100× up to 4000×). SEM images were also acquired for the recovered PA 12 powder. Fracture surfaces and recovered powder were made conductive by means of a gold metallization process to allow SEM image acquisition.

Computed tomography (CT) scans were also performed by means of the Zeiss METROTOM 1500 to identify any defects (e.g., voids, filler particle aggregations) throughout the entire specimen thickness, not limiting to the outer surfaces. Specimens were placed on metal supports to be held in vertical position between the X-ray source and X-ray detector of the machine chamber. A voltage of 130 kV, a current of 80 μA, and a magnification of 20× were set for the tests.

2.4 Life cycle assessment analysis

The environmental sustainability of the production of stereolithography 3D printed parts with recovered SLS powder as a filler was evaluated by means of the of life cycle assessment (LCA) methodology. The LCA analysis was carried out following the four iterative phases defined by the ISO 14040 and 14044 standards: goal and scope definition, life cycle inventory (LCI), life cycle impact assessment (LCIA), and results and discussion.

2.4.1 Goal and scope definition

The present LCA analysis aims at quantifying and comparing the environmental impacts associated with the production of 3D printed parts with different PA powder contents in order to evaluate the effect of the presence of recovered thermoplastic powder as filler in photocurable resin on the environmental impact.

The functional unit (FU) was defined as the production of a tensile specimen with dimensions defined according to the ASTM D638-22 standard (see Section 2.2) by means of stereolithography 3D printing technology. The FU was chosen to be as generic as possible to guarantee validity and replicability of the study.

System boundaries and scenario description

The analysis can be defined as “from cradle to gate” since all impacts from the extraction of raw materials to the post-processing of the printed parts are included within the system boundaries. More in detail, the following impact items were considered: raw material extraction (photocurable resin and polyamide powder), raw material transport, transport packaging, consumable materials, part production, part washing, and curing (see Section 2.1 for details about the manufacturing procedures). The same scenarios of the mechanical characterization analysis were considered to assess the environmental impacts of parts produced using raw resin and resin with different filler percentages (from 0.5 to 10%).

Machine production (i.e., 3D printer, washing, and curing machines) were considered out of the system boundaries since their impacts are the same in the different scenarios and would not have influence in the comparative analysis. Moreover, machine production would most likely result in negligible impacts due to their long service life compared to the analyzed production time [30]. Packaging material was considered to be sent to landfill after the end of their life.

2.4.2 Life cycle inventory

The life cycle inventory phase was conducted considering both primary (measured) and secondary data (taken from literature research and from the Ecoinvent 3.1 commercial database).

For every scenario, the weight of each specimen was directly measured, and the weight of its constituents (raw resin and PA powder) was estimated considering its nominal composition (e.g., the percentage in weight of matrix and filler for the resin mix). The resin production was modeled considering its composition (97% monomers and oligomers and 3% photoinitiators) and Ecoinvent datasets related to their production were used according to Mele et al. [31]. The impacts related to the production of PA12 powder for SLS were retrieved from previous literature studies [32, 33]. Since allocation procedures and modeling of recycled materials could significantly influence LCA results, two different approaches were employed: the first one is the cut-off approach in which the recovered material brings no environmental burden from its previous life cycle phases and only the recovery process is taken into account; this is suitable for the present case study since the SLS powder undergoes significant reductions in properties during the printing procedures and the damaged material is frequently disposed of in landfill facilities after several printing jobs [34]. The second one is the “as virgin approach” that considers that the recovered powder can still be used for SLS processing with adequate refreshing ratio and can substitute virgin powder; therefore, its impacts can be considered the same as those of the virgin material.

The environmental impact of the isopropanol used for the washing phase was obtained from the Ecoinvent commercial database; it was allocated to the functional unit considering the quantity needed to fill the washing tank (5 l), the maximum number of washing cycles before its substitution (200), and the number of parts that can be washed simultaneously [31].

The weight of the packaging material (cardboard box, plastic wrap, and HDPE bottle) for both the resin and the powder was measured. The impacts of the resin and powder packaging were allocated to the functional unit considering its weight and the capacities of the bottles; it is worth to notice that the FU weight slightly varies between the scenarios due to different filler percentages and overall different material density (e.g., it reduces from 6.14 to 5.85 g between neat resin and 10% powder filled resin). Transport distances were estimated considering the geographical position of the material suppliers and the part production facility. Both road and air transport were considered.

Energy consumptions of the 3D printing, washing, and curing machines were directly measured by means of a power meter. Their impacts are the same in all the scenarios since the machine parameters and working time were kept unchanged during the study. Impacts related to road and air transport, packaging and consumable raw materials, and low-voltage electric energy were modeled by means of the Ecoinvent commercial database. Table 1 summarizes the main inventory data considered in this LCA analysis.

Table 1 Relevant LCI data related to the FU production based on direct measurements

The LCIA phase was performed using the dedicated software SimaPro; in line with several LCA literature studies, cumulative energy demand and global warming potential were selected as representative impact categories to provide an overview of the possible impacts of the considered scenarios [35].

3 Results and discussion

3.1 Tensile test results

In order to evaluate the effect of recovered powder filler contents on the mechanical properties of the light-curing resin used in SLA processes, tensile tests were performed at room temperature on 3D printed samples characterized by different powder contents in percentage. Fig. 1A shows the printed specimens, obtained using different filler percentages, before the tensile tests; it can be observed the parts became darker as the powder percentage increases. No voids or macroscopic defects appear on the specimen surfaces. The results of tensile tests were reported in Fig. 1B as typical nominal stress (s) vs. nominal strain (e) curves at different filler percentages. Irrespective of the recovered powder contents investigated, PA12 filled resin is characterized by an initial region in which stress linearly rises with strain; then stress further increases with strain with lower slope until fracture occurs. As far as the tensile stress vs. tensile curves of reinforced specimens is concerned, Fig. 1B shows that, for a given strain value, the specimens obtained using recovered powder filler are characterized by stress values higher than that obtained by virgin resin. Such result demonstrates that the PA powder acts as a reinforcement for the resin; consequently, the tensile strength of the 3D printed components, obtained by vat photopolymerization with recovered plastic filler, can be significantly improved with respect to the raw resin. Such improvement tends to decrease with increasing the value of filler percentage, as shown by the decreasing stress values achieved, for a given strain value, as the recovered powder content increases. Furthermore, it can be observed that the reinforcement in the resin caused by the PA powder filler reduces the strain to failure as compared to the value exhibited by the virgin resin. As a matter of fact, the ductility exhibited by 3D printed specimens using reinforced matrix is always lower than the base scenario for every filler percentages, with a decrease in strain to failure ranging from 23% (for 0.5% filler content) to 42% (for 10% filler content).

Fig. 1
figure 1

A Printed specimens in photocurable resin with different powder filler percentages; B Typical nominal tress–tensile strain curves obtained by tensile tests on 3D printed specimens with different filler contents

These phenomena can be attributed to the behavior of the powder filler within the resin; while, on one hand, the filler powder promotes resin reinforcement effects and increases the material strength, on the other hand, it can induce stress concentration, leading to crack formation and premature failure. These aspects are better discussed in Section 3.2 considering the scanning electron microscopy results.

The ultimate tensile strength and Young modulus values, obtained by the nominal stress vs. nominal strain curve of Fig. 1B, were plotted as a function of the recovered material percentage in weight (Fig. 2). It can be observed that the highest mechanical properties of the light-curing resin used in vat photopolymerization processes can be obtained using the lowest value of filler percentage investigated (0.5%). In this condition, the ultimate tensile strength and Young modulus values increased by 62% and 107% with respect to the unreinforced resin.

Fig. 2
figure 2

Effect of the recovered powder content on the A ultimate tensile strength and B Young modulus values obtained by tensile tests on the printed specimens in photocurable resin

As the recovered powder percentage increases, both the tensile strength and the elastic modulus decrease; for the former, the neat resin values are reached for a filler percentage of about 7.5% and became 18% lower than the base scenario for a filler percentage of 10%. The elastic modulus never falls below the virgin resin value even though it almost reaches it for the highest reinforcement percentages. Similar behavior has already been observed in literature studies concerning particle reinforced resins where, due to filler agglomeration or other crack propagation mechanisms, the mechanical properties decreased after reaching a peak value [18].

In order to define the relationship between strength and filler weight content, the average stresses (s) at pre-established strain values, with 0.05 mm/mm increments, were plotted against the filler percentage (wt%). Since s-wt% curves showed an exponential trend in a bilinear plot, data were analyzed in a log-log scale with a linear interpolation. To this purpose, Fig. 3 shows log (s) vs. log (wt%) curves at the different strain values investigated.

Fig. 3
figure 3

Stress values as a function of the filler percentages in a logarithmic scale graph. Same colored points (and dotted lines) refer to the same strain value

It is noted that the interpolation lines referred to each strain value considered are almost parallel (constant slope) with y-intercept increasing with strain. The nominal stress vs. filler content curves can be mathematically described by the following equation:

$$\mathit{\log}(s)= Alog\left( wt\%\right)+K$$
(1)

with s, the tensile stress; A, the constant curve gradient; wt% the filler weight percentage; and K, the stress-filler content curve y-intercept.

Moreover, by plotting the y-intercept as a function of strain, a linear relation was found (Fig. 4):

$$K= Be+C$$
(2)
Fig. 4
figure 4

K values as function of the strain

with B, the gradient of the K-strain curve; e the strain; and C the K-strain rate curve y-intercept.

Finally, by substituting Eq. 2 to K in Eq. 1, the model allowing to predict the stress as a function of filler content and applied strain can be obtained:

$$\mathit{\log}(s)= Alog\left({w}_t\right)+ Be+C$$
(3)

The model allows to calculate the nominal stress at every strain level for every defined filler weight content. In the present case, the following values were obtained equal to A = −0.22, B = 0.59, and C = 0.64.

3.2 SEM and CT results

SEM images of the recovered polyamide powder are shown in Fig. 5. Spherical and slightly elongated shaped particles, with diameter ranging from 40 to 80 μm, can be observed. According to [36], surface imperfections and cracks are present. Furthermore, no noticeable powder agglomerates were observed.

Fig. 5
figure 5

SEM images of the polyamide recovered powder at two different magnifications: A 500× and B 1.5k×

Figure 6 shows the tensile fracture surfaces of neat resin and filled resin with different PA powder percentages (0.5%, 2.5%, and 10%).

Fig. 6
figure 6

SEM images of the fracture surface with a 100× magnification of specimens with different filler percentages: neat resin (A), 0.5% (B), 2.5% (C), 10% (D)

According to literature [37], the fracture surface of the neat resin appears smooth (Fig. 6A). As far as the 0.5% filled resin specimen is concerned (Fig. 6B), phenomena related to reinforcing mechanisms in particles filled resins can be identified, confirming the beneficial effects of the recovered powder on material strength. Tails behind the powder particles (circled in the figure) are markers of crack pinning [18]: the crack propagation along the plane perpendicular to the load direction is obstructed by the rigid reinforcement, leading to a bowing in the crack front and the tails (bifurcation of the crack). Debonding at the interface between matrix and reinforcement can also be detected; the rigid fillers cause stress concentration, leading to void and microcrack formation on their interface with matrix and consequently determining debonding of the powder [38]. An example of void caused by a powder particle debonding is highlighted in the figure by the black square. A homogeneous dispersion of the powder was achieved.

Similar considerations can be made for the 2.5% filled resin specimen (Fig. 6C). The increased powder percentage leads to the formation of small agglomerates of 2–3 spherical particles even though overall homogeneous material dispersion is observed. The agglomeration phenomenon is remarkable for the 10% filler resin powder percentage, where particle cluster is easily detected across the whole fracture surface (Figs. 6D and 7A). A preferential direction of crack propagation can still be identified by observing the tails and the different height fracture surfaces; moreover, microcracks connecting powder particles can be seen in different directions throughout the whole image. The rigid reinforcement determines localized stress, triggering the formation and propagation of cracks. This can lead to early failures and a reduction in ductility of the specimens. Such result is in accordance with tensile test outcomes that show a progressive decrease in strain to failure with increasing reinforcement phase percentage.

Fig. 7
figure 7

Fracture surface of the 10% filler specimen with 500× magnification (A) and zoom in at 3k× on a damaged powder particle (B)

Fig. 7 shows a detail of a polyamide damaged particle embedded in resin matrix (Fig. 7B) and several voids in its proximity caused by debonding phenomena (Fig. 7A); according to literature [39], particles with higher aspect ratio show higher interfacial strength. Hence, even if the use of damaged powder with rough surfaces is not recommended in SLS processes and could lead to low-quality parts [34], it could contribute to improve resin reinforcement effect if used as filler.

Figure 8 shows computed tomography images of specimens obtained with neat resin and 10% filled one. Irrespective of the presence of filler powder, few printing defects were observed.

Fig. 8
figure 8

Computed tomography images of specimens in A neat resin and B resin with 10% filler

While no defects of appreciable size are observed in pure resin specimens (Fig. 8A), few voids with diameters of 0.05–0.06 mm are detected in filled resin specimens (Fig. 8B). This indicates that the powder-resin stirring did not cause relevant void formation phenomena; nevertheless, further void reduction could be achieved by means of pre-processing operations on the printing material (e.g., vacuum chamber degassing) [40].

3.3 Life cycle impact assessment

The results of LCIA are reported in terms of global warming potential (Fig. 9A) and cumulative energy demand (Fig. 9B); both figures present two total impact columns referred to “as virgin” and “cut-off” approaches to model the recovered powder. The main difference between the different approaches is represented by the PA powder impacts that are equal to zero in the “cut-off” approach and equal to those of the virgin powder production in the “as virgin approach.” The two impact categories show similar trends.

Fig. 9
figure 9

Life cycle impact assessment results in terms of global warming potential and cumulative energy demand

As far as the impacts of the powder production are considered (“as virgin”), the total impacts slightly decrease as the filler content increases; in particular, the impact values decrease of about 4.5% in GWP and 3.1% for CED from neat resin to 10% filled resin. The most relevant items are the photocurable resin production (3.85 kg CO2 eq per kg of resin) and its transport (12.75 kg CO2 eq per kg of resin), accounting for about 17–30% and 56–63% of the total impacts, respectively, depending on the scenario and the impact category. Resin transport impacts are mainly determined by the long haul aircraft transport while PA12 powder transport only relies on road transport, leading to negligible emissions. Hence, even considering high environmental loads associated with the PA12 powder production (15.39 kg CO2 eq and 345 MJ per kg), its use leads to environmental benefits by reducing the quantity of MSLA resin and lowering the impacts related to its production and transport. Printing, washing, and curing phases are characterized by low energy consumption and their contribution accounts for about 15% of GWP (mostly due to the isopropanol) and less than 1% of CED.

Indeed, in industrial applications, damaged powder not suitable for SLS printing and destined to landfill disposal would be used as a recovered filler with much greater carbon footprint reduction. This scenario is well represented by the cut-off allocation with no impacts related to the filler material production and transport; in this case, the environmental impacts decrease by 12% for GWP and 14% for CED with a 10% filler content with respect to the neat resin base scenario (from 0.125 to 1.11 kg CO2 eq and from 1.92 to 1.65 MJ). These environmental benefits, combined with quick and inexpensive preparation of the recovered filled resin and with the improvement in mechanical performances of the filled material, could lead to more sustainable 3D printed products. Similar outcomes are expected as plastic or other materials waste are recycled and pulverized by means of low-impact processes (e.g., milling and mechanical grinding) and used as fillers [41].

The environmental impact results shown in Fig. 9 were obtained by using a FU defined by only geometrical features, not considering the effect of filler content on the resin mechanical properties. However, in applications in which strength and/or stiffness are a design constraint, such effect is a crucial issue that can also affect environmental impacts. For this reason, a different functional unit can be defined to take into account the filled resin stiffness in the environmental assessment. The target stiffness value was defined as that exhibited by neat resin tensile specimens 3D printed according to the ASTM D638-22 standard.

Since the specimens were produced using raw materials with different Young moduli due to the presence of different filler weight content, they can exhibit the same displacement as subjected to a defined tensile load if they have different cross section areas. In particular, the ratio between the cross section area of each specimen and the neat resin reference must be equal to the reciprocal of the ratio of their elastic moduli [42].

The updated FU is defined as a tensile specimen with length defined by the ASTM D638-22 standard and cross section area defined so that the displacement along the load direction, irrespective of the applied load within the elastic region, is equal to that of the one of the neat resin reference scenario.

The updated FU allows to obtain filled resin specimens with lower weight as compared with the neat resin one due to their higher Young moduli. By considering the new specimen dimensions and weight, the LCI data were updated. Fig. 10 shows the results of the analysis in terms of GWP and CED. It is worth noticing that, due to linear relationship between elastic moduli, cross section area, and weight, the choice of a different reference scenario for the definition of the new FU would have not changed the results trend.

Fig. 10
figure 10

LCIA results in terms of GWP (A) and CED (B) for the different scenarios in the case of parts with the same tensile stiffness

The 0.5% filler specimen showed the highest stiffness and therefore needs the lowest amount of material (2.8 g vs. 6.1 g of the neat resin base scenario) to ensure the defined requirement. This implies a reduction of the raw resin quantity higher than 50%, with a marked decrease in resin production and transport impacts; the overall reduction in carbon footprint was equal to 48% (0.65 kg CO2 eq vs. 1.25 kg CO2 eq). At the same time, the filler powder impacts are almost negligible due to very low powder content; moreover, as compared to other filler materials used in low percentages, PA12 powder is characterized by low environmental impacts and provides high stiffening effects [21]. Due to lower part thickness, the printing phase is faster than that of the neat resin and requires lower energy consumption. Environmental impacts increase with filler percentage due to the decrease in elastic modulus of filled resins.

However, since the elastic modulus of the filled resins is always higher than the neat resin one, impacts are always lower than the neat resin scenario value of at least 15%.

4 Conclusions and further developments

In this paper, the reuse of PA12 SLS powder as a filler for vat photopolymerization was widely investigated. To this purpose, 3D printed specimens in neat resin and filled resin with different PA12 powder contents (from 0.5 to 10%) were subjected to tensile tests and computed tomography. Furthermore, fracture surfaces were investigated by means of scanning electron microscopy. Filler content effects on mechanical properties of 3D printed specimens were investigated and failure mechanisms were identified. Finally, sustainability of the innovative filled resin was assessed with the LCA methodology. The main outcomes are reported as follows:

  • Strength and stiffness of neat resin can be significantly improved by filling the matrix resin with PA12 powders. The lowest filler content investigated (0.5%) leads to the highest ultimate tensile strength and Young modulus, with an increment of 62% and 107%, respectively, as compared to the neat resin.

  • The strain to failure decreases as the filler content increases, with the highest reduction with respect to the neat resin, equal to 42%, obtained with the 10% filled resin specimens.

  • A predictive mathematical model was developed to estimate the mechanical strength of 3D printed components as a function of both filler content and applied strain.

  • SEM and CT images show a homogeneous dispersion of the filler particles within matrix. Small powder clusters and small size voids can be detected for the highest filler content parts (10%). Particle reinforcement mechanisms, such as crack pinning and tailing, were observed.

  • Life cycle assessment analysis shows a reduction in impacts associated with the recovered powder up to 14% for the production of the specimens as the functional unit is defined considering only geometric criteria. This reduction can become as high as 48% if the increment of Young modulus obtained by means of the reinforcement and the specimen stiffness are considered to define the functional unit.

Research activities are in progress to investigate the use of different recycled materials and higher filler contents. This would also allow to confirm and extend the validity of the predictive mathematical model developed in the present work. Cost analyses are also being carried out to compare the traditional printing materials with the innovative recovered materials and to prove the industrial relevance of the proposed model.