Impact of printing orientation on inter and intra-layer bonds in 3D printed thermoplastic elastomers: a study using acoustic emission and tensile tests

Polymer 283 (2023) 126241 Available online 1 August 2023 0032-3861/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Impact of printing orientation on inter and intra-layer bonds in 3D printed thermoplastic elastomers: A study using acoustic emission and tensile tests B` arbara Adrover-Monserrat , Silvia García-Vilana , David S´ anchez-Molina , Jordi Llum` a , Ram´ on Jerez-Mesa , Eva Martinez-Gonzalez, J. Antonio Travieso-Rodriguez * Universitat Polit` ecnica de Catalunya, Escola d’Enginyeria de Barcelona Est, Av. d’Eduard Maristany, 10-16, 08019, Barcelona, Spain ARTICLE INFO Keywords: Additive manufacturing Material extrusion Thermoplastic elastomer Acoustic emission Inter and intra-layer bonding Mechanical characterization ABSTRACT The use of the Material Extrusion technique with Thermoplastic Elastomers is currently growing because of the large number of benefits of this family of materials. They are processable materials with high flexibility, which makes them very useful, for example, in biomedical applications that require flexible objects with complex geometries. This study aims to characterize a specific polymer, namely polyether-block-amide-based polymer (PEBA), by analyzing its anisotropic behavior in printed samples and investigating the mechanical properties based on different printing orientations. Three orientations (X, Y, and Z) were used to relate the printing orientation to the type of bonds formed in the samples: intra-layer bonds, inter-layer bonds, and the deposited filament. Tensile tests following ASTM D638 were conducted to measure sample rigidity, while Acoustic Emission, an advanced Non-Destructive Technique, was employed to examine the trend of the failure process. The presence of voids was also observed to assess printing quality, which is influenced by the printing orientation and alters the quality of bonds. The results revealed that samples printed horizontally exhibited higher Young’s Modulus values and fewer voids in the inner parts. Vertically printed samples displayed inferior mechanical properties and a greater number of voids. Consequently, the intra-layer yielded better bond formation and minimized voids. Acoustic Emission analysis corroborated these findings by demonstrating distinct energy distribution patterns among the different printing orientations. Hits were concentrated at maximum stresses, primarily observed in the vertically printed samples, which experienced macroscopic failure. Furthermore, this particular specimen exhibited a vertical asymptote near the maximum stress level. The analysis of the energy of Acoustic Emission hits demonstrated a reasonably good fit with the Gutenberg-Richter (GBR) law based on the printing direction. 1. Introduction It is well known that engineering materials can have anisotropy, which can be caused by intrinsic properties of the material (e.g., the thermal history during crystallization in metallic materials or during polymerization) or can be induced by the manufacturing process to which it is subjected. The case of 3D printing can be identified with the latter because of the way the material is deposited, namely layer by layer. The occurrence of orthotropic behavior induced by 3D printing is occasionally not fully taken into consideration. Due to this reason, it is necessary to study the behavior of 3D-printed specimens under different manufacturing conditions, as the mechanical performance of a printed specimen will be determined by the weakest segment or union between filaments. To analyze this anisotropy, advanced detection techniques must be used to recognize failures. An analysis through an advanced detection method, known as Acoustic Emission (AE), is proposed in this paper. Material Extrusion (MEX) is one of the most valued 3D printing techniques due to its good performance with elaborated geometries and its low prices. MEX consists of the fabrication of parts, layer upon layer, from a filament made of polymeric material that follows a programmed trajectory mastered by a CNC routine. In addition, it is possible to achieve other properties among thermoplastics by adding fibers and particles or by processing a copolymer. MEX allows the use of flexible * Corresponding author. E-mail addresses: [email protected] (B. Adrover-Monserrat), [email protected] (S. García-Vilana), [email protected] (D. S´ anchez-Molina), [email protected] (J. Llum` a), [email protected] (R. Jerez-Mesa), [email protected] (E. Martinez-Gonzalez), antonio. [email protected] (J.A. Travieso-Rodriguez). Contents lists available at ScienceDirect Polymer journal homepage: www.elsevier.com/locate/polymer https://doi.org/10.1016/j.polymer.2023.126241 Received 18 January 2023; Received in revised form 31 July 2023; Accepted 31 July 2023 Polymer 283 (2023) 126241 2 materials, such as Thermoplastic Elastomers (TPE), which are copolymeric materials with elastomeric properties that might provide the parts with softness and flexibility. The use of materials with elastomeric properties in 3D printing is increasingly interesting as they present good mechanical behavior along with novel properties that can be used in industries such as soft robotics or medical devices [1–4]. Hence, the mechanical characterization of TPEs processed through 3D printing is essential. Therefore, it is essential to undertake a scientific investigation of TPEs’ mechanical properties post 3D printing. Although previous studies [5–11], have explored the mechanical characteristics of TPEs, there remains a need to further delve into this domain. One of the most relevant concepts that define the mechanical behavior of 3D printed samples is the creation of bonds between deposited filaments. This is known as inter and intra-layer bonding, which can be analyzed with procedures such as cross-sectional image analysis or mechanical studies such as tensile tests. The stronger the connections, the better mechanical response a tested part will have. However, the resistance of these welded unions can be weaker due to the presence of voids for rigid materials [12–17] and for flexible materials [5,6,12,18] because of the geometrical shape of the filaments, printing defects, or the inadequate selection of the manufacturing parameters. In this sense, the orientation of the printed specimens might be one of the most important parameters, as it affects the creation of unions between filaments [19–22]. In fact, the presented works in past reviewed literature demonstrate that the mechanical properties of 3D-printed samples are due to change depending on the selected manufacturing parameters. The mechanical properties of 3D-printed samples can be determined through different techniques. Non-destructive techniques (NDT) such as AE can be used to evaluate the failure of specimens while testing them. AE is a passive technique that detects the transient elastic waves generated by the rapid release of energy from some sources within a material. Some fracture mechanisms and changes in the microstructure of the material can be detected by means of AE [23–25]. This technique was used to measure the progress of damage in different materials [25–28], and even predict their eventual failure stress based on the estimated damage [29]. In addition, other research showed that AE allowed the determination of the nature and the type of micro-failure by means of the characteristics of the AE detected [30]. The previous results show that AE could be a useful and interesting technique to study the anisotropy of the material processed through MEX if it can detect how its mechanical behavior changes depending on its microstructure orientation caused by the manufacturing routine selected to generate it. In fact, AE has previously been introduced in MEX 3D-printed samples [31,32], even some have applied AE in rigid materials to distinguish between different printing configurations [33,34]. However, most of the references do not apply AE during the mechanical testing of the samples, instead, the authors usually use AE to evaluate the process of 3D printing in different situations such as for the detection of filament breaking during the printing [35], to diagnose machine faults like extruder blockage or running out of material [36], for the detection and identification of possible failures during the printing of the first layer [37,38], or the recognition of common MEX distortions as warping [39]. Therefore, the challenge is to analyze the mechanical behavior of the MEX 3D printed samples through AE. Deepening in the MEX technique, the study of bonds between layers and AE is processed through techniques normally focused on rigid materials. However, the behavior of rigid materials is expected to differ significantly from the flexible ones such as PEBA’s (polyether-blockamide-based polymer). Indeed, PEBA is a flexible TPE based on polyamide capable of transmitting vibrations [40] which has been rarely studied. The present work aims to study the anisotropic behavior in TPE specimens obtained through MEX. To do this, PEBA specimens were 3Dprinted in three orientations, to study the effect of the exposure of the extruded filaments, the intra-layer bonds, and the inter-layer bonds to load. The study of variations in the mechanical behavior of specimens was done by conducting tensile tests and incorporating the AE technique during these tests. It is usually assumed that the best printing orientation for maximizing tensile strength involves depositing filaments aligned with the tensile force. However, this study aims to investigate the validity of this assumption and determine whether it holds true for a flexible material processed via the MEX technique. As evidenced by the results of this study, differences in the stiffness and damage mechanisms depending on the printing direction are observed. Moreover, the possibility of implementing the AE technique in TPEs processed through MEX is demonstrated. To our knowledge, this is the first study analyzing the three printing directions of TPE and incorporating the AE technique to evaluate the damage evolution. 2. Materials and methods A thermoplastic elastomer (TPE) polyether-block-amide-based polymer, with hardness 90 Shore A (PEBA 90A), supplied by Fillamentum (Hulín, Czech Republic) was used to evaluate the inter and intra-layer behavior. PEBA is a copolymeric material with flexible performance that can be processed through MEX. It is characterized by its good transmission of vibrations and good mechanical performance. Also because of its potential to be used for the manufacturing of orthosis or models to be applied in the biomedical industry. This work consists of three main sections. The first part involves conducting tensile tests in accordance with ASTM D638 standards to determine the rigidity of the samples. In the second section, which is done during the tensile tests of the first part, AE was used to monitor and analyze the failure process’s progression of the samples when tested by placing a sensor on the samples while loaded (Fig. 1). The third part of the study is visual support, where the presence of voids to assess the printing quality was observed, which is influenced by the printing orientation and affects the overall quality of bonds. 2.1. 3D Printing Ultimaker Cura 4.3 (Utrecht, The Netherlands) was used to slice the 3D models. An Ender-Pro-3 (Shenzhen Creality 3D Technology Co., Ltd., Shenzhen, China) was used to manufacture specimens. Because this paper focuses on the analysis of the anisotropy of parts, only the printing orientation was variable, keeping all the other printing parameters constant for all samples. The study involved a printing strategy to establish unions either between filaments or between layers (creating two types of bonds) with the goal of analyzing the mechanical behavior of these bonds, encompassing factors like rigidity and strength. Moreover, a comparison was made by evaluating the force required to test a Fig. 1. Position of acoustic emission sensors in the tensile samples. B. Adrover-Monserrat et al. Polymer 283 (2023) 126241 3 block of filaments against the bonds. Thus, the primary objective was to ascertain if the unions created were stronger or weaker than individual filaments. Therefore, three different printing orientations were selected, resulting in three configurations. The first configuration was obtained by placing the model oriented longitudinally along the X-axis (configuration X) with filaments oriented with the tensile force. The second one was aligned longitudinally along the Y-axis (configuration Y) exposing the intra-layer bonds. The third configuration was placed along the Zaxis (configuration Z) exposing inter-layer bonds, as the printer’s extruder moves along axes X and Z, covering the Y-axis by movement of the heated platform. Fig. 2 provides visual support for a better understanding of the printing arrangement. The printing velocity (1300 mm/min), printing temperature (245 ◦C), layer height (0.2 mm), and nozzle diameter (0.4 mm) were decided paying attention to previous work done in the research group. The 3D Printing process was challenging as all filaments that build the parts had to be deposited in the same orientation. Additionally, the extrusion of TPEs in general is not as simple as a thermoplastic material. The manufacturing process of the specimens used was a long iterative process. Eventually, it was realized that some critical steps should be followed for successful printing. For example, to avoid stuck problems, all the retractions were disabled. The feed rate was constant for all the samples. An optimized rate of value 1 was set to maintain the dimensional accuracy of the samples. With a bed temperature of 60 ◦C and a good calibration of the platform, there were no warping problems with the specimens printed horizontally as they had all their surfaces in contact with the heated bed. 2.2. Tensile tests To obtain the mechanical properties of the specimens, tensile tests were performed with a Universal Testing Machine, Zwick AllaroundLine table-top 5 kN (ZwickRoell S.L., Ulm, Germany), equipped with a load cell of 50 N. The standard ASTM D638 [41] was followed for defining the testing parameters and the design of the specimens. Specifically, the shape and dimensions of the printed parts were as determined in the standard type I. Moreover, five specimens were printed in each orientation to ensure repeatability. A testing speed of 50 mm/min was set. The specimens were tested up to 550% deformation, as no further deformation is typically needed for elastomeric parts applications [42–44]. 2.3. Acoustic emission As previously introduced, the AE technique is a non-destructive testing technique used to detect elastic waves, which are generated by the release of elastic energy during the material’s response to an external stimulus [29]. This technique is based on the use of sensors placed in contact with the specimen during the mechanical test (Fig. 1), which detect the AE events generated during the increase of damage. AE signals emitted during the tensile test were detected and recorded with a Vallen System GmbH (Bayern, Germany). Two AE passive piezoelectric sensors (VS150, Vallen System Gmbh) were firmly attached to the specimen close to the extremes. Their frequency response is characterized by a peak at 150 kHz, where they exhibit resonance. Two 34 dB pre-amplifiers (AEP4, Vallen System GmbH) and a fourchannel system (AMSY-5, Vallen System GmbH) were also used. The AE system provides frequency filtering in the pre-amplifiers and in the AMSY-5 board. AE measurements were filtered using a low threshold of 34 dB. Eventually, signals with zero duration and/or zero rise time were removed. Each specimen’s final AE dataset consisted of the total number of hits or AE events, their energy, and the associated stress achieved when each AE signal was detected. The AE signals were analyzed using the Gutenberg-Richter law (GBR). GBR law relates the number of seismic events in a region during a pre-established time interval [45]. This law was previously used to analyze the structural integrity of materials such as cement [46] or coal [47]. In uniaxial compression studies, the value of its exponent ε has been shown to be related to the degree of damage that occurs in the material [48] or to the homogeneity of the sample [47]. Moreover, in the literature, it was observed that the value of the exponent ε of the GBR law is indicative of the type of material used (for example, in animal bone, this exponent takes a value between 1.3 and 1.7, being higher with the age of the animal). An increase in the exponent ε supposes an increase in the probability of longer emissions. That is, given an AE that increases to maximum energy and then decreases, the emission will have a longer duration [49]. The GBR distribution for elastic wave energies released in an elastic process is defined in Eq. (1), pGBR(E) = ε −1 Emin (E Emin)− ε ∀E≥Emin (1) with E being the energy of the event or hit, Emin being the lower limit of the energy or threshold, and ε being the Gutenberg–Richter exponent. However, this well-known law has a direct relationship with the exponential distribution, thus, it can be rewritten using a change of random variable, being η =ln (E/Emin )whose inverse is E=Emine η . Consequently, the distribution can be expressed as shown in Eq. (2), obtaining the exponential distribution where λ= ε −1. pexp( η ) = [ ε −1 Emin e− εη ]Emine η =λe−λ η (2) In this way, the energies above the threshold Emin unequivocally follow a GBR distribution when ln (E/Emin )follows an exponential distribution. Eventually, the GRB exponent can be obtained from parameter λ. The same concept about the energy of elastic waves produced in earthquakes seems to be applicable to elastic waves produced by AE [48]. Fig. 2. Printing orientation of the tensile specimens. B. Adrover-Monserrat et al. Polymer 283 (2023) 126241 4 2.4. Statistical analysis Mean and standard deviation values were computed for the mechanical properties and AE parameters obtained. The statistical influence of the printing orientation in the mechanical properties and the acoustic emission parameters was assessed using the Kruskal-Wallis test for comparing the three directions, and Wilcoxon test for the comparison of two directions. A p-value <0.05 was considered to be statistically significant. 3. Results The results obtained regarding the printability of the samples, the mechanical behavior of PEBA, and the anisotropic AE in (PEBA), are presented in this section. 3.1. Printability Cross-sectional images of the different specimens were conducted and analyzed to investigate voids’ formation. It was found that depending on the manufacturing orientation of the specimens, their porosity changes. In order to assess the compactness and porosity of the specimens in relation to their manufacturing orientation, which correlates with the obtained mechanical and acoustic emission findings, cross-section images were captured (see Fig. 3). The analysis of the images revealed that Y-oriented samples exhibited a reduced number of voids, whereas Z-oriented samples displayed a higher void content, making this one the configuration with the highest porosity. Notice that the parallel lines in the images are a consequence of the cutting process, they are not part of the sample nor the printing process. The Z configuration was the one that shows more imperfections since it was manufactured vertically. This configuration was affected by the vibrations of the heated platform, which decreased the stability to create necks between layers and filaments. On the contrary, the Y configuration had the highest density since the bonds could be created more easily mainly for one reason: the time between the deposition of one filament to the next one was shorter than in the case of the X configuration (transverse paths are shorter than the longitudinal ones), therefore, the temperature of the bonded filaments was higher, thus, better unions could be created. 3.2. Tensile tests From tensile tests, the strain-stress curves for all the specimens were obtained (Fig. 4). ISO 527-1 standard [50] was used to measure the mechanical properties, which are reported in Table 1. The tensile properties such as the Young’s modulus (YM) and yield strength (Rp 0.2 ) showed to be different depending on the printing orientation (p =0.005). Specimens oriented along the Z direction showed the lowest YM as expected, with a resulting rigidity of 77.23 ± 7.02 MPa. However, the Y-oriented specimens reported the highest YM (102.39 ±8.20 MPa) whereas the YM obtained from X-oriented specimens was 86.69 ±6.42 MPa. Differences among paired groups showed to be statistically significant (X vs. Y p =0.016, Y vs. Z p =0.008) excepting the comparison between YM of orientations X and Z, which was nearly significant (p =0.056). Regarding the yield strength, Rp 0.2 was different between the three printing orientations (p =0.008) being 1.19 ±0.19 MPa for the X-oriented specimens; 0.99 ±0.09 MPa for those Y-oriented; and 2.09 ±0.60 MPa for Z-oriented specimens. The comparison between paired groups showed significant differences between X and Z orientations (p =0.032) and Y and Z orientations (p = 0.008), but not for the X and Y groups (p =0.095). 3.3. Acoustic emission The study of the AE data obtained during the uniaxial traction tests was carried out. After the preliminary analysis of the data, signals not associated with the breaking of the material caused by friction or other phenomena that cause noise were ruled out. To that end, a filter Emin = 4.5eu (or equivalently ln(Emin) = 1.5) was established. This value ensured an adequate number of events, as well as properly adjusting the signal energies to an exponential distribution. The right graph of Fig. 4 shows the relative number of hits (N/N total ) versus relative strain (S x /S max ) for two samples in each direction. It can be observed that the number of hits in the X direction is, on average, lower than the number observed in the Y and Z directions. In addition, the Z direction is the only configuration with a clear vertical asymptote defined by the number of hits at very high-stress values (concerning the maximum stress observed) unlike samples X and Y. In this way, most of the hits in the Z specimens are accumulated at high stress values near the catastrophic failure. This accumulation is not observed in the other two directions, where a more uniform distribution is displayed throughout the data trend. The energies associated with the detected events were analyzed using exponential distributions as indicated in Eq. (2). The value of the parameter λ of the distribution for each sample was obtained (Fig. 5). Eventually, the value of the Gutenberg Richter exponent can be determined as ε =λ+1. All the samples passed the KolmogorovSmirnov and Chi-square tests when adjusting the exponential distribution. The average results of the samples are shown in Table 2. No statistical differences were observed between the mean number of hits (Nprom) detected for each printing orientation (p =0.516). On the other hand, it can be observed that the λ parameter is statistically different for the three printing directions, being the highest value for the X direction and the lowest for the Z direction (p =0.026). Consequently, being both parameters λ and ε directly related, this same variation is observed in the parameter ε , whose value was verified for each sample. Regarding the comparison between paired groups, significant differences were only observed between the orientations X and Z (p =0.008), but not between X and Y (p =0.310) nor Y and Z (p =0.151). 4. Discussion The present study focused on analyzing the influence of the anisotropy of 3D-printed polyether-block-amide-based polymer (PEBA 90A) specimens by using the acoustic emission technique and mechanical tensile tests. This is the first study analyzing the mechanical properties and introducing the AE technique in 3D-printed PEBA specimens and Fig. 3. Cross-section images of the presence of the different specimens’ orientations: (a) X-oriented; (b) Y-oriented; (c) Z-oriented. Samples were cut perpendicularly to the direction of the deposited filaments. B. Adrover-Monserrat et al. Polymer 283 (2023) 126241 5 Fig. 4. Representation of two samples for each printing direction: (left) strain-stress tensile curve; (right) the relative number of hits versus relative strain. Table 1 Mean value and standard deviation for Young’s modulus (YM) and yield strength (Rp 0.2 ) obtained for the different printing orientations. Printing Direction YM p-value Rp 0.2 p-value X 86.69 ±6.42 0.005* 1.19 ±0.19 0.008* Y 102.39 ±8.20 0.99 ±0.09 Z 77.23 ±7.02 2.09 ±0.60 Fig. 5. Exponential fittings of the energy for one sample of each direction and comparison of the distributions in the XYZ directions with the mean lambda parameter obtained. Table 2 Average values and deviation of the number of hits N and the parameters λ and ε for the three directions X Y Z. Printing Direction N prom λ ε X 128 ±49 0.669 ±0.046 1.669 ±0.046 Y 230 ±186 0.581 ±0.145 1.581 ±0.145 Z 186 ±106 0.456 ±0.038 1.456 ±0.038 B. Adrover-Monserrat et al. Polymer 283 (2023) 126241 6 with different printing orientations. Therefore, the structure of the specimens obtained with this technique is formed by filaments of polymeric material arranged in adjacent rows and layers, which gives the specimen orthogonal symmetry [51]. In addition, this microstructure contains microdefects due to the small spaces between fibers, which, depending on their distribution in the microstructure, causes the breakage of these materials to be irregular, and with certain randomness, as has been previously observed in other viscoelastic materials [29,52,53]. Moreover, for constant printing parameters (e.g. speed or temperature), the orientation of the specimen taken during extrusion will affect its properties, among other things, due to the main orientation of the fibers and the order of deposition. The latter will influence the temperature maintained by the fibers before the deposition of the next layer, which is expected in this study to affect the strength of the inter-fiber bonds. To study this effect, in this research, PEBA 90A specimens were obtained by 3D printing in three different fabrication directions (according to the three orthogonal directions) in order to test purely one type of bonds or filaments. This was done to study the differences in the mechanical behavior of the samples. The anisotropy of the mechanical behavior was studied by tensile tests and the use of AE during the whole test to monitor the damage generated in the specimen. From these tests, the mechanical properties, the acoustic events (or hits) associated with the evolution of the damage, and their energy were measured. 4.1. Tensile tests The effect of printing orientation on the mechanical behavior of samples was analyzed from tensile tests, where Young Modulus (YM) and yield strength (Rp 0.2 ) were determined for the specimens oriented in X, Y, and Z directions. In this way, it was ensured that either only one type of bond (inter on intra-layer) or the filaments were tested. This type of test has been used since it allows the use of constant cross-section specimens (obtaining properties that only depend on the material), the determination of the deformation from the measured elongation, and the implementation of the AE technique by placing sensors in contact with the sample. As previously mentioned, a uniform stress field is achieved along the specimen and thus, the differences measured in the results for the three orientations are related to the inter and intra-layer characteristics of each orientation. Moreover, cross-sectional images of the different orientations were captured in order to relate the internal structure with the results obtained (Fig. 2). The results showed that Y-oriented specimens had the highest YM, followed by the X-oriented specimens. The Z-oriented ones showed the lowest YM. These results seem to be related to the microstructural images from the cross-images analysis, where the Z-oriented specimen showed a higher number and size of voids in its microstructure, followed by the X-oriented samples. In this sense, this observation agrees with the YM values reported, as it can be assumed that the presence of voids makes the samples less rigid, reporting lower values of YM. Regarding the Rp 0.2 values, as expected, their increment had an inverse behavior of YM, as in this property, voids play a less significant role. In elastomeric materials, the YM is the slope of the very initial part of the curve. This fact explains that when the YM is increased, the intersecting point of Rp 0.2 is notably decreased. Hence the shape of the curve is quite similar, but not the slope of the parallel line. The comparison between Y and Z-oriented specimens is intriguing due to their perpendicular filament deposition concerning the applied force, despite their markedly distinct mechanical behavior. The critical factor lies in the nature of the bonds exposed to tensile forces. During the layer-by-layer deposition, the temperature of the specific layer is higher, leading to more robust intra-layer bonds compared to inter-layer bonds. Regarding Y-oriented specimens, it is observed that the bonds exposed to the tensile force are intra-layer, exhibiting superior mechanical strength when compared to Z-oriented specimens, where the inter-layer bonds are subjected to the tensile force. In other words, Y-oriented specimens demonstrate stronger tensile bonds due to the thermal effects occurring within a single layer. No previous literature was found studying the influence of three manufacturing orientations obtained through 3D-printing in PEBA. In fact, the studies that previously investigated the effect of printing directions, have done it with rigid thermoplastics such as ABS, PLA, or HDPE. The authors that work with these materials tend to find that the direction that performs better in mechanical testing is the one that prints the filaments oriented along the load direction (X-oriented) [21,54,55], whereas in the present study, the Y-oriented samples (where the intra-layer bonds are exposed to the load) showed higher YM values. However, Schirmeister [22] concluded that no massive influence of the printing direction was detected as they achieved really good bonding between the extruded filaments in all directions. This founding is related to the results presented in this present work, where it is demonstrated that the printing orientation is essential for 3D printing: good results will depend on the printing quality and on the bonds created. As for the literature working with TPEs, few papers were found studying the inter and intra-layer bonds by testing them mechanically. Le´ on [12] concluded the same results as the ones obtained in this work. The X-oriented configuration has a higher YM than the Z-oriented configuration, where an improvement in rigidity is observed. Nevertheless, Le´ on [12] did not consider the Y-oriented direction. In another study, Bachtiar [5] printed layers by alternating the deposition angles. The intra-layer bonds could not be studied individually as the joints were compound by both, intra and inter-unions. Thermoplastic elastomers can achieve better bonds between layers and filaments than rigid thermoplastics [12], therefore, as the results of this paper indicate, it is possible to get good results when printing perpendicularly to the applied force (Y-oriented). To achieve it, it is important to define the optimal parameters for the printing process. The continuity of the material, avoiding voids, is essential for good mechanical behavior. The mechanical testing results suggest that, with optimal printing parameters, the unions created in Y-oriented are stiffer than the filaments themselves. 4.2. Acoustic emission During the tensile tests, AE sensors were placed in contact with the specimen to detect the acoustic events propagating through the specimen, which are elastic waves produced as a result of breakage or irreversible events. Thus, for each AE, the stress of the specimen was measured (the stress-strain state is of uniform plane stress, in the region of interest). Thus, the objective was to detect the hits associated with the damage process in the specimens, which was expected to be related to different mechanisms due to the different structural configurations of the orientations achieved during the 3D-printing manufacturing. Eventually, through an analysis of AE, the printing orientation (and therefore the type of bonds) would be characterized. The results for the AE spectrum showed remarkable differences in the hits-strain behavior between the three distinctive printing directions. Essentially, the number of hits observed in the X direction is, on average, less than that detected in the Y and Z directions (however, this difference was not statistically significant, due to the deviation between samples). It should be noted that samples in the X and Y directions did not break, unlike the Z direction. Nevertheless, the behavior of samples Y and Z is similar to each other. These differences exist given that in the X direction, the PEBA filaments are located along the stress direction, and it can be conjectured that the hits detected in the samples manufactured in the X direction would be mostly related to filament breaking. However, when one of the other two directions (Y or Z) is tested, the filaments are perpendicular to the external stress, where it is suspected that the AE is related to the intraand inter-layer bonds breakage respectively. This fact shows that the direction of the filaments regarding the applied stress notoriously alters the failure behavior of the samples. In this manner, the detected hits in the samples produced in the X direction are primarily B. Adrover-Monserrat et al. Polymer 283 (2023) 126241 7 associated with filament breaking. However, the hits captured in the Y and Z directions would be associated with the separation between layers (debonding). When comparing samples Y and Z, it was observed that in the former, hits were distributed throughout the tension, while in the latter, these hits are mostly concentrated in values very close to the maximum tension. Hence, the Z stress-strain graph contains a vertical asymptote at stress values near the maximum stress. Previous studies showed how in different materials, hits occur for stresses close to the maximum strength of the material, generating an asymptote [26,29]. This vertical asymptote is interpreted as an avalanche effect, in which many micro-failures chain together, creating a macroscopic break [56]. These facts explain the reason why the asymptote of AE is only observed in the Z direction, where macroscopic failure was reached. Note that in the Y and X directions no asymptote is observed. This is due to not having reached the maximum stress of the sample, aside from possible other microstructural differences (namely, the manufacturing conditions). The energies of the hits observed according to the printing direction were also analyzed and fit reasonably well with the well-known Gutenberg-Ricther (GBR) law. The GBR law has been widely used in the study of earthquakes, as well as in the analysis of the evolution of damage in cement and its relationship with the microstructure [46,47, 57]. From the GBR law, the value of the exponent ε of GBR (Eq. (1)) was determined, which in this study was related to the parameter of an exponential distribution. This relation exists since when the energies follow a GBR law, the logarithms of the energies follow an exponential law, thus being both approaches equivalent. The higher values of the ε exponent (shown in Table 2) were obtained for the X direction (1.669 ±0.046), followed by the Y-oriented specimens (1.581 ±0.145) and the Z-oriented ones (1.456 ±0.038). The differences between the mean values are statistically significant, so they showed to vary according to the request direction, being higher in the X direction and lower in Z. This indicates that the damage mechanisms of the three orientations are different among the three orientations. Indeed, in the specimens made with filaments, the AE signals would involve micro failures of groups of filaments or the links between the components of the microstructure. It is known that the λ parameter of the exponential distribution is related to the inverse of the expected value of the said distribution (E[ln (E/Emin )] = 1/λ). In this way, it can be concluded that the energetic mean observed in the events in the three directions is statistically different. The most plausible explanation for the difference observed in the GBR parameter (or equivalently in the λ parameter) is the nature of the most frequent micro faults produced in the samples. On the one hand, and as mentioned above, the X-oriented direction specimens have the filaments aligned in the direction of the force. On the other hand, even though in the samples requested according to the Y and Z directions, the filaments are perpendicular to the direction of request, the Y and Z directions are printed differently. In the Y-oriented direction samples, the filaments stacked in the direction of the applied stress are deposited one after the other (until completing the entire plane in that direction). While in the Z-oriented direction samples, the printed layer is the cross-section of the stress. In this way, it is logical to think that the bond between the Y-filaments in the tensile direction is stronger as the temperature when filaments are deposited is higher. Thus, the joints that resist tensile in the Y specimens would be more rigid and with higher strength than in the Z direction. This fact would explain three consequences. First, greater resistance was observed in the Y samples (which do not reach maximum rupture) compared to the Z. Second, the concentration of the hits at maximum stresses was higher in the Z direction. Third, differences in the energy distribution of the events were observed. From the emissions detected with AE, an approximation of the damage recorded by this technique can be calculated as dk=Nk/λk, with N being the total number of hits and 1/λ being its average intensity. The following damages were calculated: dX=193.8±79 <dY= 366.6±323.5<dZ=416.5±242.6, being precisely the Z direction, the only one that has reached complete breaking. Unfortunately, the energy analysis by itself does not allow direct discernment between types of events (breaking of filaments or union between filaments). Some events may involve the breaking of groups of filaments or several unions, which would require a statistical microscopic detailed analysis of the micro faults, which is outside the objectives of this study. Regarding the obtained values, it was observed in the literature that the parameter ε approaches 1 when the level of total damage increases [48]. Compared to the results of this work it would imply that the samples in Z, whose value of ε is lower, would exhibit a greater degree of damage, which is again consistent with the fact that it is the only sample that breaks. Among other values in the literature, it was observed that coal exhibits a parameter of ε =1.4–1.7, being higher in the most heterogeneous samples. Note that these values are similar to those obtained in this study, which is not surprising, since previous research shows that different systems can be classified under the same universality classes, given that they show almost identical exponents [58]. Based on the results, it would be interesting to conduct further investigations to obtain samples with specific characteristics that reach rupture in three dimensions. That would allow the research team to analyze the evolution of the acoustic emission signals. 5. Conclusions This work analyzes the influence of the printing orientation on the mechanical and damage behavior of 3D-printed PEBA specimens. With this objective, the thermoplastic elastomeric material PEBA was processed through MEX in three different printing orientations, and uniaxially tested under the ASTM D638 standard. This could be the first study analyzing three different orientations of 3D-printed TPE, which involved different types of bonds (intra and inter-layer bonds). Moreover, the Acoustic Emission technique was implemented during the tensile test, detecting the elastic waves spreading through the material due to the irreversible processes that occur in the specimens during the load increase. The present study shows that the implementation of the AE technique in thermoplastic elastomeric materials processed through MEX was possible. Moreover, the results suggest that the mechanisms responsible for the irreversible processes in the material with the load increase are different among the three directions analyzed, where a higher Gutenberg-Richter exponent was observed for the specimens, whose filaments were oriented in the load direction, followed by those with the intra-layer bonds subjected to the load. Thus, this indicates that the transverse printing orientations Y and Z were notably different, and moreover, the Young’s modulus values obtained from the tensile tests showed that the Y-oriented specimens were stiffer than those whose filaments were oriented in the load direction. Z-oriented samples were the weakest due to the poor unions created. This direction presented the concentration of the hits at maximum stresses as macroscopic failure was reached. It cannot be assumed that the behavior of PEBA is the same as the behavior of a thermoplastic such as PLA as the created bonds can be higher for this TPE material. However, the printing orientation is a key parameter to consider as its effect is statistically significant. The detection method presented in this paper can effectively analyze the damage of the 3D printed sample exposed to tensile loading, distinguishing the printing directions selected. Funding The first author gratefully acknowledges the Universitat Polit` ecnica de Catalunya for the financial supportof her predoctoral grant FPUUPC, with the collaboration of Banco de Santander. The second author acknowledges the Ministerio de Universidades of Spain for her Margarita Salas Grant. B. Adrover-Monserrat et al. Polymer 283 (2023) 126241 8 CRediT authorship contribution statement B` arbara Adrover-Monserrat: Data curation, Formal analysis, Investigation, Software, Funding acquisition, Writing – original draft. Silvia García-Vilana: Data curation, Formal analysis, Investigation, Software, Funding acquisition, Writing – original draft. David S´ anchezMolina: Conceptualization, Formal analysis, Methodology, Validation, Writing – review & editing. Jordi Llum` a: Conceptualization, Formal analysis, Methodology, Validation, Software, Visualization, Writing – review & editing. Ram´ on Jerez-Mesa: Formal analysis, Methodology, Visualization, Writing – review & editing. Eva Martinez-Gonzalez: Formal analysis, Methodology, Validation, Software, Visualization. J. Antonio Travieso-Rodriguez: Conceptualization, Formal analysis, Methodology, Validation, Software, Resources, Project administration, Supervision, Writing – review & editing. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper named “Influence of the printing orientation on the inter and intra-layer bonds via Acoustic Emission of a 3D printed Thermoplastic Elastomer”. Data availability Data will be made available on request. Acknowledgments The authors would like to thank Fillamentum Manufacturing Czech s.r.o. for their support. References [1] X. Zhou, P.S. Lee, Three-dimensional printing of tactile sensors for soft robotics, MRS Bull. 46 (2021) 330–336, https://doi.org/10.1557/s43577-021-00079-3. [2] C.P. Paul, K. Dileep, A.N. Jinoop, A.C. Paul, B.K. S, in: H.K. Dave, J.P. Davim (Eds.), Fused filament fabrication for external medical devices, Mater. Forming, Mach. Tribol., Springer, 2021, pp. 299–322, https://doi.org/10.1007/978-3-030-680244_16. [3] R.N.M. Delda, R.B. Basuel, R.P. Hacla, D.W.C. Martinez, J.-J. Cabibihan, J.R. C. Dizon, 3D printing polymeric materials for robots with embedded systems, Technologies 9 (2021) 82, https://doi.org/10.3390/technologies9040082. [4] G. Stano, G. Percoco, Additive manufacturing aimed to soft robots fabrication: a review, Extrem. Mech. Lett. 42 (2021), 101079, https://doi.org/10.1016/j. eml.2020.101079. [5] E.O. Bachtiar, O. Erol, M. Millrod, R. Tao, D.H. Gracias, L.H. Romer, S.H. Kang, 3D printing and characterization of a soft and biostable elastomer with high flexibility and strength for biomedical applications, J. Mech. Behav. Biomed. Mater. 104 (2020), 103649, https://doi.org/10.1016/j.jmbbm.2020.103649. [6] X. Lin, P. Coates, M. Hebda, R. Wang, Y. Lu, L. Zhang, Experimental analysis of the tensile property of FFF-printed elastomers, Polym. Test. 90 (2020), 106687, https://doi.org/10.1016/j.polymertesting.2020.106687. [7] M. Robinson, S. Soe, R. Johnston, R. Adams, B. Hanna, R. Burek, G. McShane, R. Celeghini, M. Alves, P. Theobald, Mechanical characterisation of additively manufactured elastomeric structures for variable strain rate applications, Addit. Manuf. 27 (2019) 398–407, https://doi.org/10.1016/j.addma.2019.03.022. [8] X. Lin, J. Gao, J. Wang, R. Wang, M. Gong, L.L. Zhang, Y. Lu, D. Wang, L.L. Zhang, Desktop printing of 3D thermoplastic polyurethane parts with enhanced mechanical performance using filaments with varying stiffness, Addit. Manuf. 47 (2021), 102267, https://doi.org/10.1016/j.addma.2021.102267. [9] V. Kanyanta, A. Ivankovic, Mechanical characterisation of polyurethane elastomer for biomedical applications, J. Mech. Behav. Biomed. Mater. 3 (2010) 51–62, https://doi.org/10.1016/j.jmbbm.2009.03.005. [10] B. Arifvianto, T.N. Iman, B.T. Prayoga, R. Dharmastiti, U.A. Salim, Tensile properties of the FFF-processed thermoplastic polyurethane (TPU) elastomer, Int. J. Adv. Manuf. Technol. 117 (2021) 1709–1719, https://doi.org/10.1007/s00170021-07712-0. [11] R.R. Ford, A.K. Pal, S.C.E. Brandon, M. Misra, K.A. Mohanty, The influence of fused filament fabrication printing parameters on the mechanical properties of a thermoplastic elastomer, Rapid Prototyp. J. 28 (2022) 1906–1919, https://doi. org/10.1108/RPJ-10-2021-0274. [12] A.S. de Le´ on, A. Domínguez-Calvo, S.I. Molina, Materials with enhanced adhesive properties based on acrylonitrile-butadiene-styrene (ABS)/thermoplastic polyurethane (TPU) blends for fused filament fabrication (FFF), Mater. Des. 182 (2019), 108044, https://doi.org/10.1016/j.matdes.2019.108044. [13] J. Ghorbani, P. Koirala, Y.L. Shen, M. Tehrani, Eliminating voids and reducing mechanical anisotropy in fused filament fabrication parts by adjusting the filament extrusion rate, J. Manuf. Process. 80 (2022) 651–658, https://doi.org/10.1016/j. jmapro.2022.06.026. [14] D. Yang, H. Zhang, J. Wu, E.D. McCarthy, Fibre flow and void formation in 3D printing of short-fibre reinforced thermoplastic composites: an experimental benchmark exercise, Addit. Manuf. 37 (2021), 101686, https://doi.org/10.1016/j. addma.2020.101686. [15] B. Yin, Q. He, L. Ye, Effects of deposition speed and extrusion temperature on fusion between filaments in single-layer polymer films printed with FFF, Adv. Ind. Eng. Polym. Res. 4 (2021) 270–276, https://doi.org/10.1016/j.aiepr.2021.07.002. [16] 3D printing in Formula 1, 3D print, Ind, https://3dprintingindustry.com/news/in terview-alpines-pat-warner-offers-a-unique-insight-into-the-secretive-world-of-3dprinting-in-formula-1-209139/, 2023 (accessed February 9, 2023). [17] Y. Tao, F. Kong, Z. Li, J. Zhang, X. Zhao, Q. Yin, D. Xing, P. Li, A review on voids of 3D printed parts by fused filament fabrication, J. Mater. Res. Technol. 15 (2021) 4860–4879, https://doi.org/10.1016/j.jmrt.2021.10.108. [18] C.Y. Liaw, J.W. Tolbert, L.W. Chow, M. Guvendiren, Interlayer bonding strength of 3D printed PEEK specimens, Soft Matter 17 (2021) 4775–4789, https://doi.org/ 10.1039/d1sm00417d. [19] K. Rodze´ n, E. Harkin-Jones, M. Wegrzyn, P.K. Sharma, A. Zhigunov, Improvement of the layer-layer adhesion in FFF 3D printed PEEK/carbon fibre composites, Compos. Part A Appl. Sci. Manuf. 149 (2021), 106532, https://doi.org/10.1016/j. compositesa.2021.106532. [20] T. Yao, H. Ouyang, S. Dai, Z. Deng, K. Zhang, Effects of manufacturing microstructure on vibration of FFF 3D printing plates: material characterisation, numerical analysis and experimental study, Compos. Struct. 268 (2021), https:// doi.org/10.1016/j.compstruct.2021.113970. [21] T. Yao, K. Zhang, Z. Deng, J. Ye, A novel generalized stress invariant-based strength model for inter-layer failure of FFF 3D printing PLA material, Mater. Des. 193 (2020), https://doi.org/10.1016/j.matdes.2020.108799. [22] C.G. Schirmeister, T. Hees, E.H. Licht, R. Mülhaupt, 3D printing of high density polyethylene by fused filament fabrication, Addit. Manuf. 28 (2019) 152–159, https://doi.org/10.1016/j.addma.2019.05.003. [23] S. Hanagud, R.G. Clinton, J.P. Lopez, Acoustic emission in bone substance, in: Proc. Biomech. Symp. Am. Soc. Mech. Eng., ASME, New York, 1973. [24] R.-P. Franke, P. Dorner, H.J. Schwalbe, B. Ziegler, Acoustic emission measurement system for the orthopedic diagnostics of the human femur and knee joint, J. Acoust. 22 (2004) 236–242. https://www.ndt.net/article/jae/papers/22-236.pdf%0Ahtt p://212.8.206.21/article/jae/papers/22-236.pdf. [25] D.G. Aggelis, M. Strantza, O. Louis, F. Boulpaep, D. Polyzos, D. van Hemelrijck, Fracture of human femur tissue monitored by acoustic emission sensors, Sensors (Switzerland) 15 (2015) 5803–5819, https://doi.org/10.3390/s150305803. [26] D. S´ anchez-Molina, E. Martínez-Gonz´ alez, J. Vel´ azquez-Ameijide, J. Llum` a, M.C. R. Soria, C. Arregui-Dalmases, A stochastic model for soft tissue failure using acoustic emission data, J. Mech. Behav. Biomed. Mater. 51 (2015) 328–336, https://doi.org/10.1016/j.jmbbm.2015.07.002. [27] A. Carpinteri, G. Lacidogna, F. Accornero, A.C. Mpalaskas, T.E. Matikas, D. G. Aggelis, Influence of damage in the acoustic emission parameters, Cem. Concr. Compos. 44 (2013) 9–16, https://doi.org/10.1016/j.cemconcomp.2013.08.001. [28] G. Shen, Z. Wu, J. Zhang, Advances in Acoustic Emission, Springer, Switzerland, 2015, https://doi.org/10.1007/978-3-319-29052-2. [29] S. García-Vilana, D. S´ anchez-Molina, J. Llum` a, I. Fern´ andez-Osete, J. Vel´ aquezAmeijide, E. Martínez-Gonz´ alez, A predictive model for fracture in human ribs based on in vitro acoustic emission data, Med. Phys. 48 (2021) 5540–5548, https://doi.org/10.1002/mp.15082. [30] E. Martinez-Gonzalez, I. Picas, J. Romeu, D. Casellas, Filtering of acoustic emission signals for the accurate identification of fracture mechanisms in bending tests, Mater. Trans. 54 (2013) 1087–1094, https://doi.org/10.2320/matertrans. M2013089. [31] M.S. Hossain, H. Taheri, In situ process monitoring for additive manufacturing through acoustic techniques, J. Mater. Eng. Perform. 29 (2020) 6249–6262, https://doi.org/10.1007/s11665-020-05125-w. [32] Y. AbouelNour, N. Gupta, In-situ monitoring of sub-surface and internal defects in additive manufacturing: a review, Mater. Des. 222 (2022), https://doi.org/ 10.1016/j.matdes.2022.111063. [33] Z. bo Pan, W. Zhou, K. Zhang, L. hua Ma, J. Liu, Flexural damage and failure behavior of 3D printed continuous fiber composites by complementary nondestructive testing technology, Polym. Compos. 43 (2022) 2864–2877, https:// doi.org/10.1002/pc.26582. [34] F. Li, Z. Yu, Z. Yang, Failure characterization of PLA parts fabricated by fused deposition modeling using acoustic emission, Rapid Prototyp. J. 26 (2020) 1177–1182, https://doi.org/10.1108/RPJ-09-2019-0247. [35] Z. Yang, L. Jin, Y. Yan, Y. Mei, Filament breakage monitoring in fused deposition modeling using acoustic emission technique, Sensors (Switzerland) 18 (2018) 1–16, https://doi.org/10.3390/s18030749. [36] Y. Lu, Y. Wang, Machine fault diagnosis of fused filament fabrication process with physics-constrained dictionary learning, in: Procedia Manuf., Elsevier B.V., 2021, pp. 726–734, https://doi.org/10.1016/j.promfg.2021.06.071. [37] H. Wu, Z. Yu, Y. Wang, Experimental study of the process failure diagnosis in additive manufacturing based on acoustic emission, Meas. J. Int. Meas. Confed. 136 (2019) 445–453, https://doi.org/10.1016/j.measurement.2018.12.067. B. Adrover-Monserrat et al. Polymer 283 (2023) 126241 9 [38] P. Bhavsar, B. Sharma, W. Moscoso-Kingsley, V. Madhavan, Detecting First Layer Bond Quality during FDM 3D Printing Using a Discrete Wavelet Energy Approach, Elsevier B.V., 2020, pp. 718–724, https://doi.org/10.1016/j.promfg.2020.05.104. [39] F. Li, Z. Yu, Z. Yang, X. Shen, Real-time distortion monitoring during fused deposition modeling via acoustic emission, Struct. Health Monit. 19 (2020) 412–423, https://doi.org/10.1177/1475921719849700. [40] B. Adrover-Monserrat, J. Llum` a, R. Jerez-Mesa, J.A. Travieso-Rodriguez, Study of the influence of the manufacturing parameters on tensile properties of thermoplastic elastomers, Polymers (Basel) 14 (2022) 576, https://doi.org/ 10.3390/polym14030576. [41] ASTM D638-14, Standard Test Method for Tensile Properties of Plastics (n.d.), htt ps://doi.org/10.1520/D0638-14, 2022. [42] W. Geng, T.J. Cuthbert, C. Menon, Conductive thermoplastic elastomer composite capacitive strain sensors and their application in a wearable device for quantitative joint angle prediction, ACS Appl. Polym. Mater. 3 (2021) 122–129, https://doi. org/10.1021/acsapm.0c00708. [43] A. Leber, B. Cholst, J. Sandt, N. Vogel, M. Kolle, Stretchable thermoplastic elastomer optical fibers for sensing of extreme deformations, Adv. Funct. Mater. 29 (2019), https://doi.org/10.1002/adfm.201802629. [44] D. Plant, B. Blackman, P. Leevers, Mechanical and rheological testing to develop thermoplastic elastomer-polyborodimethylsiloxane blends for personal impact protection, Polym. Test. 86 (2020), https://doi.org/10.1016/j. polymertesting.2020.106477. [45] A. Clauset, C.R. Shalizi, M.E.J. Newman, Power-law distributions in empirical data, SIAM Rev. 51 (2009) 661–703, https://doi.org/10.1137/070710111. [46] D.L. Carnì, C. Scuro, F. Lamonaca, R.S. Olivito, D. Grimaldi, Damage analysis of concrete structures by means of acoustic emissions technique, Compos. B Eng. 115 (2017) 79–86, https://doi.org/10.1016/j.compositesb.2016.10.031. [47] Y. Xu, A.G. Borrego, A. Planes, X. Ding, E. Vives, Criticality in failure under compression: acoustic emission study of coal and charcoal with different microstructures, Phys. Rev. E. 99 (2019), https://doi.org/10.1103/ PhysRevE.99.033001. [48] J.D. Riera, I. Iturrioz, The Gutenberg-Richter and similar laws and their relation with numerical and experimental laboratory results, in: 15th World Conf. Earthq. Eng. Lisbon Port, 2012. [49] J. Bar´ o, P. Shyu, S. Pang, I.M. Jasiuk, E. Vives, E.K.H. Salje, A. Planes, Avalanche criticality during compression of porcine cortical bone of different ages, Phys. Rev. E. 93 (2016) 1–8, https://doi.org/10.1103/PhysRevE.93.053001. [50] ISO 527-1:2020, Plastics - Determination of Tensile Properties - Part 1: General Principles (ISO 527-1:2019) (n.d.), 2020. [51] Q.-S. Zheng, Theory of representations for tensor Functions—a unified invariant approach to constitutive equations, Appl. Mech. Rev. 47 (1994) 545–587, https:// doi.org/10.1115/1.3111066. [52] D. S´ anchez-Molina, S. García-Vilana, J. Vel´ azquez-Ameijide, C. Arregui-Dalmases, Probabilistic assessment for clavicle fracture under compression loading: ratedependent behavior, Biomed. Eng. Appl. Basis Commun. 32 (2020), 2050040, https://doi.org/10.4015/S1016237220500404. [53] J. Vel´ azquez-Ameijide, S. García-Vilana, D. S´ anchez-Molina, J. Llum` a, E. MartínezGonz´ alez, M.C. Rebollo-Soria, C. Arregui-Dalmases, Prediction of mechanical properties of human rib cortical bone using fractal dimension, Comput. Methods Biomech. Biomed. Eng. 24 (2021) 506–516, https://doi.org/10.1080/ 10255842.2020.1836623. [54] K. Fayazbakhsh, M. Movahedi, J. Kalman, The impact of defects on tensile properties of 3D printed parts manufactured by fused filament fabrication, Mater. Today Commun. 18 (2019) 140–148, https://doi.org/10.1016/j. mtcomm.2018.12.003. [55] J.J. Laureto, J.M. Pearce, Anisotropic mechanical property variance between ASTM D638-14 type i and type iv fused filament fabricated specimens, Polym. Test. 68 (2018) 294–301, https://doi.org/10.1016/j.polymertesting.2018.04.029. [56] A. Johansen, D. Sornette, Critical ruptures, Eur. Phys. J. B 18 (2000) 163–181, https://doi.org/10.1007/s100510070089. [57] J. Perez-Oregon, F. Angulo-Brown, N.V. Sarlis, Nowcasting avalanches as earthquakes and the predictability of strong avalanches in the olami-federchristensen model, Entropy 22 (2020) 1–21, https://doi.org/10.3390/e22111228. [58] V. Navas-Portella, ´ A. Gonz´ alez, I. Serra, E. Vives, ´ A. Corral, Universality of powerlaw exponents by means of maximum-likelihood estimation, Phys. Rev. E. 100 (2019) 1–14, https://doi.org/10.1103/PhysRevE.100.062106. B. Adrover-Monserrat et al.