Journal of Mechanical Science and Technology 31 (1) (2017) 133~139 www.springerlink.com/content/1738-494x(Print)/1976-3824(Online) DOI 10.1007/s12206-016-1212-4 Influence of building orientation on the flexural strength of laminated object manufacturing specimens† D Olivier1, J A Travieso-Rodriguez2, S Borros1, G Reyes1 and R Jerez-Mesa2,* Materials Engineering Research Group, Institut Quimic de Sarria, Universitat Ramon Llull, Barcelona, Spain Mechanical Engineering Department, Universitat Politècnica de Catalunya, Barcelona, Spain (Manuscript Received November 18, 2015; Revised July 5, 2016; Accepted September 8, 2016) Abstract This study aims to define the best building orientation for components produced via the Laminated object manufacturing (LOM) technique to enhance their flexural performance Results of previous research show that components produced via LOM are capable of withstanding higher deflections than components produced through other layer manufacturing techniques However, the relation between the building orientation and flexural strength of components has not yet been assessed Four types of specimens have been manufactured using different building orientations for each type The specimens have been tested in a machine with four loading points to evaluate their failure mode and identify the best building orientation toward flexural loading The best building orientation in terms of maximum load before failure is 45° Furthermore, a repetitive failure pattern is found for each tested condition Building orientation is confirmed to be a relevant parameter in LOM manufacturing by influencing the mechanical properties of components Keywords: Flexural strength; Laminated object manufacturing; Additive manufacturing; Rapid prototyping Introduction Additive manufacturing (AM) can be defined as a group of techniques that aims to obtain final components and prototypes within a short period from a 3D file generated using CAD software [1] These processes make performing activities that support the design process of prototypes possible by validating their definitive design or selecting their best geometric configuration In recent years, these techniques have expanded in the area of serial production of final components, mainly because of the development of new machines that can decrease costs and manufacturing time [2] AM techniques are also frequently selected because of the high geometric complexity achieved by their manufactured components and the confidentiality they provide when dealing with the development of special components These techniques have also been enhanced by developing new materials, such as powder metal [3] and glass [4]; improving the accuracy of their manufacturing devices; and further simplifying the required postprocesses, coupled with the extended possibilities provided by currently available CAD software [5] Given the aforementioned reasons, AM has transformed into a commodity system; it is easy to perform using devices derived from industrial * Corresponding author Tel.: +34 934 137 431 E-mail address: ramon.jerez@upc.edu † Recommended by Associate Editor Vikas Tomar © KSME & Springer 2017 AM systems These desktop systems are frequently referred to as 3D printers [6] Laminated object manufacturing (LOM) belongs to the group of the aforementioned AM techniques [7] It allows the acquisition of desired shapes by attaching successive sheets of a rolled material Whenever a layer is deposited, it is cut with a tool driven via a simple numerical control, which defines the shape of the final component and allows the removal of undesired materials by the end of the manufacturing process [8] LOM is used to manufacture paper models, which are materials of high interest given that their prototypes are cheap, readily available, and easy to cut [9] However, as the technique evolves and the availability of building materials is improved, interest has turned toward making functional parts out of metals [10], ceramics [11-13], and composites [14, 15] Examples of final products for industrial applications created via LOM are rapid tooling [16] and pattern making in the foundry industry [17] The LOM system used in this study is based on the superposition of PVC film layers, which are securely glued to each other using a commercial adhesive When a building is completed, a compact layered block is formed A secondary postprocessing operation is required to peel off the layered support material, thereby showing the final component Numerous authors have focused their research on optimizing the postprocessing phase to increase its productivity For example, 134 D Olivier et al / Journal of Mechanical Science and Technology 31 (1) (2017) 133~139 Kechagias and Iakovakis (2009) realized this objective by applying neural networks [18] The post-processing stage is typically time-consuming and is frequently cited as the main disadvantage of the LOM technique; thus, other authors have focused on optimizing the overall manufacturing design strategy to reduce total manufacturing time [19, 20] Nevertheless, past research on LOM systems has shown the potential of this technique Expanding knowledge on LOM and characterizing the properties of components manufactured using this method are of high industrial interest; they allow designers and engineers to determine the actual potential of this technology to produce final components [21] LOM has been positively cited for its high dimensional accuracy Park et al (2000) [22] concluded that LOM components demonstrated high accuracy; they identified optimal manufacturing parameters to reduce the highest dimensional error, which occurred along the vertical axis These results were later confirmed and specified for a wider range of test factors by Kechagias (2007) [23] by applying a design of experiments approach LOM has also been studied in terms of surface roughness Campbell et al (2002) [24] concluded that the final surface roughness of LOM components was easy to predict, and thus, could be considered a remarkable advantage compared with other AM techniques Kechagias et al (2007) [25] determined that system temperature and layer thickness were the most influential parameters for final surface roughness; they also established an arithmetic model for predicting final surface roughness value depending on the manufacturing parameters Ahn et al (2012) [26] designed a finite element model and solved new mathematical expressions to predict the surface roughness of LOM components The numerical computation values were confirmed through comparison with real specimens With regard to mechanical strength, Chryssolouris et al (2003) [27] found that layer thickness was the most influential parameter for the tensile strength of LOM specimens Similar conclusions were also obtained by Paul and Voorakarnam (2001) [28], who confirmed that the mechanical properties of LOM components were highly dependent on the type of raw material used not only because of its composition but also because of the state in which it was introduced into the system Liao and Chin (2006) [29] added that thermal history was an influential factor for the strength and economy of the process Es-Said et al (2000) also studied building orientation [30] They found that 0° was the optimal building direction to maximize impact and tensile strength Bonding strength is also a relevant factor that should be defined, and thus, certain constructive patterns that maximize the bonding surface between layers are desired to maximize the resistance of the final component, as highlighted by Liao and Chiu (2001) [31] The mechanical strength of the layered components results from two main mechanisms The first mechanism is related to the strength of the raw laminated material, i.e., along the direction of alignment of the polymer molecules The second mecha- Table Configuration of the tested LOM specimens Specimen Laminate deposition direction Loading force direction with regard to layers #1 Parallel to the x-axis Perpendicular #2 Parallel to the y-axis Perpendicular #3 45° to the x-axis Perpendicular #4 Parallel to the x-axis Parallel Fig LOM specimens produced under different conditions nism is defined by the bonding strength between layers [32] Studies on the flexural strength of LOM specimens remain limited Thus, based on a previous basic research presented by some of the authors of this paper [33], the current work selects building orientation as the objective factor of study to determine its influence on flexural strength Given a fixed specimen geometry, building orientation defines the direction of the application of testing strength with regard to the orientation of the polymer molecules and the manner in which layers are deposited and glued together to generate the final volume Therefore, this study evaluates the aforementioned influence and focuses on the failure pattern exhibited by LOM specimens when subjected to a four-point bending test Materials and methods A Solido SD300 LOM 3D printer was used to manufacture 80×10×4 mm3 bars according to the ASTM D6272 standard The x- and y-axes precision of the system is ±0.1 mm and 0.168 mm at the z-axis The raw material was a laminated SolVC-105 PVC with a thickness of 0.12 mm SolGL-101 was the selected adhesive agent, and SolAG-154 was the complementary anti-adhesive Four types of specimen were manufactured, and each one demonstrated a different building orientation, as shown in Table For each type, four specimens were considered for the sake of repeatability, as presented in Fig The flexural properties of the specimens were evaluated using a universal testing machine MTS100 An MTT flexural testing device with four loading points was utilized, and the load and support spans were set according to ASTM D6272 standard (Fig 2) The calculation and test inputs specified in the standard are shown in Tables and D Olivier et al / Journal of Mechanical Science and Technology 31 (1) (2017) 133~139 135 Table Test parameters according to ASTM D6272 standard Parameter Value Unit Loading span 21.3 mm Length of yield segment % Length of slope segment % Support span 64 mm Yield offset 0.002 mm/mm Point at strain 3.5 % Table Required calculation inputs Parameter Value Unit Final strain point 3.5 % Failure sensitivity 90 % Acquisition data rate 10 Hz Initial speed 1.9 mm/min Fig Failure crack propagation through the thickness: (a) Frontal view of specimen #4; (b) detail of specimen failure; (c) transverse view of zip-shaped crack propagation Fig Experimental setup of the tests with four bending points according to ASTM D6272 standard Different tested specimens were subjected to visual inspection, and the characteristic failure features were manually measured to analyze the failure mechanism of the material Furthermore, a stereomicroscope (Leica M60) was used to observe the exhibited crack formation in the specimen layers Result discussion The first approach in the flexural study was performed through a three-point flexural test according to the ASTM D790 standard However, significant failure patterns within the deformation range specified at the same standard were not evident in the specimens Consequently, the four-point bending test was selected to develop the research A thorough inspection of the tested specimens showed a repetitive deformation pattern with a relevant component of elastic deformation Fig 3(a) shows one of the tested specimens The cracks that opened at the outer layer were propagated along the z-direction in all cases, as shown by the crack detail in Fig 3(b), and throughout the entire thickness in some cases Crack formation was observed on both sides of the specimen The manner in which this crack penetrated into the inner layers could not be observed because the components were opaque Although crack propagation appeared superficial, a closer inspection of the tested parts exhibited a zipper-shaped deeper failure mechanism along the transverse section of the components (Fig 3(c)) Conversely, crack propagation demonstrated varying behavior when the results of different testing conditions were compared Some of the results displayed single cracks, whereas others presented multiple crack paths Some cases had parallel paths This first visual inspection procedure showed that each set of conditions would result in different failure mechanisms In multiple-crack failure specimens, the distance between the cracks was measured, and considerable repeatability was observed The different macroscopic and microscopic bonding mechanisms of the material affected the global macroscopic behavior of the layered part Bonding was performed through a chemical adhesive deposited among different homogeneous layers stacked along the z-direction at a macroscopic level Internally, the orientation of the extruded PVC sheet molecules favored the anisotropic behavior of the mechanical properties of a component, which increased strength bonding 136 D Olivier et al / Journal of Mechanical Science and Technology 31 (1) (2017) 133~139 Table Average values of crack propagation patterns found on specimen #1 Left edge (mm) Mean SD Table Average values of crack propagation patterns observed on specimen #2 Right edge (mm) Left edge (mm) d1 d2 d3 d1 d2 d3 9.44 6.54 36.68 9.57 6.63 36.49 0.08 0.37 0.26 0.06 0.20 0.22 Mean SD Right edge (mm) d1 d2 d3 d1 d2 d3 20.69 20.14 38.38 20.44 19.98 38.40 0.25 0.43 0.17 0.59 0.67 0.72 Fig Schematic representation of failure points on specimen #1 Fig Schematic representation of failure points on specimen #2 along the direction of orientation and reduced strength at the transversal direction Different macroscopic responses are expected to be obtained depending on the combination of the two bonding phenomena, as discussed in the following section to the center, were separated at approximately 20 mm from each side of the component The average experimental results are presented in Table In this case, polymer chains were at right angles with the longest length of the specimen and the deposition direction, which was caused by the selected building direction Conversely, the loading device was parallel to the polymer chain orientation; that is, the specimen was loaded along the weakest direction of the microscopic polymer orientation 3.1 Specimen #1 For the first tested condition, flexural behavior and failure pattern are illustrated in Fig The average distance between different propagated cracks, measured for a set of five samples of specimen #1, is shown in Table along with its standard deviation Three failure points were observed One of these points was located at the center of the specimen This central crack was visually the deepest The other two failure points were coincident with the support points and were symmetric In this case, the orientation of the extruded raw material was parallel to the building direction; that is, the longest dimension of the specimen was parallel to the building direction, and its layers were stacked along its thickness Thus, when the specimen was loaded, the polymer chains and bonding adhesive were at right angles with the loading force 3.2 Specimen #2 Fig presents the configuration of the resulting failure mode for specimen #2 Three failure points were observed One of these points was slightly biased from the geometric center Two additional failure points, which were equidistant 3.3 Specimen #3 The failure mode of specimen #3, which shows two crack points, is presented in Fig These points were not symmetrical in the top view of the specimen representation, unlike the cracks that developed on specimens #1 and #2 Considering the location of the cracks from both edges, d3 represents the mismatch between the cracks observed on both sides of the specimens and measures the asymmetry level The average values are presented in Table Although the values obtained from the location of the failure points (d1 and d2) had a considerably higher dispersion in this test, the obtained gap value (d3) was highly consistent In this case, the polymer chains, i.e., the strongest direction of the PVC sheet, were oriented at an angle of 45° with the longest specimen direction, as a result of the extrusion direction of the stock material The failure mode of these specimens depicted the manner in which stresses are concentrated, following the direction in which the load was applied The 137 D Olivier et al / Journal of Mechanical Science and Technology 31 (1) (2017) 133~139 Table Average values of crack propagation patterns found on specimen #3 Left edge (mm) Mean SD Table Average values of crack propagation patterns observed on specimen #4 Right edge (mm) d1 d2 d3 d1′ d2′ 18.82 47.29 4.72 23.54 51.68 3.74 4.64 0.60 3.97 4.22 Up (mm) Down (mm) d1 d1′ 39.90 40.05 0.39 0.72 Mean SD Table Mean characteristic values of the stress–strain curves for each type of specimen, including deviations Specimen Maximum load (N) UTS (MPa) Young’s modulus (MPa) Mean SD Mean SD Mean SD #1 50,70 1,42 20,30 0,57 1004,37 29,13 #2 59,17 0,94 23,68 0,38 1177,77 17,26 #3 51,25 1,63 20,50 0,65 1047,33 40,20 #4 38,98 0,54 15,63 0,20 1053,80 27,59 Fig Schematic representation of failure points on specimen #3 Fig Flexural curve according to ASTM D6272 standard set of experiments because the loading force was perpendicular to the stacked layers of the PVC bound with polymeric adhesive 3.5 Flexural curve Fig Schematic representation of failure points on specimen #4 crack still propagated along a direction at right angles with the polymer chains 3.4 Specimen #4 The results of the fourth set of experiments are presented in Fig This case showed a single failure point located approximately at the center of the specimen The failure point went through the entire thickness, thereby exhibiting the mark of plastic deformation The numerical values measured for the different specimens are provided in Table This testing condition was particularly different from those used in the other After analyzing the four testing conditions, the influences of the macroscopic and microscopic contributions on the bonding of LOM components and their flexural strength are shown in Fig Despite the heterogeneous nature of this layered material, the four curves in that figure indicated the typical behavior of plastic components subjected to pure flexural load Pure elastic behavior could be initially observed, followed by an inflexion region that led to a new plastic–elastic deformation region At a deformation point of 3.5 %, the behavior of the four tested conditions was clearly differentiated, and the maximum bending stress was obtained This behavior could be compared to that observed by Sun (1996) [32] for laminated components, which specified the theoretical failure criterion for the manufactured components by following similar techniques, as is the case of LOM Future lines of investigation should focus on the development of new failure criterion 138 D Olivier et al / Journal of Mechanical Science and Technology 31 (1) (2017) 133~139 theories to explain the behavior of LOM specimens, such as those included in this study The general mechanical strength of the layered parts results from two main mechanisms The first mechanism is related to the strength of the raw laminated material, i.e., along the direction of alignment of polymer molecules The second mechanism is defined by the bonding strength between layers A higher resistance was observed in specimen #3, a lower but significantly similar strength for specimens #1 and #2, and a considerably lower strength for specimen #4 This behavior is consistent with the failure mechanisms described previously for each set of specimens and is a consequence of the overall contribution of different bonding mechanisms that are internally present In specimen #3, the influence of the adhesive forces along the z-direction is reinforced by the polymer orientation direction that is orthogonal to the loading force Therefore, the higher resistance exhibited in this case results from both coupled mechanisms In specimen #1, the load is applied perpendicularly to the polymer chain direction, which leads to the addition of macroscopic and microscopic contributions, as well as good resistance properties The failure mechanism located at the lower supports is attributed to tensile stress applied to the lower side of the specimen In specimen #2, two failure points appear beside the central point In this case, the weakest direction of the oriented film is located parallel to the load application This condition explains the defect formation that is located close to the loading supports, in which the loaded specimen is mainly subjected to compressive stress In specimen #4, plastic deformation occurs with lower loads Moreover, a lower strength is registered because of the heterogeneous contributions present in this configuration Specimens manufactured at an angle of 45° with the global reference axis exhibit better performance against flexural stress This finding asserts that parts with future similar working regimes should be manufactured to maintain this orientation Conclusions The contribution of the original manufacturing process of the stock polymer and the characteristics of the LOM additive process have demonstrated a combined effect on the macroscopic response expressed in terms of flexural properties The analysis of the failure mechanism under flexural testing for additive manufactured LOM specimens shows the following - Each group of tested conditions exhibits a repetitive failure pattern Thus, the manufacturing orientation of LOM specimens can be inferred to influence their flexural behavior - The stronger direction of LOM components matches the orientation of the polymer chains because of the sheet extrusion process of the stock - LOM components produced with a building orientation of 45° present the highest flexural strength References [1] H Bikas, P Stavropoulos and G Chryssolouris, Additive manufacturing methods and modelling approaches: a critical review, International Journal of Advanced Manufacturing Technology, 83 (1-4) (2016) 389-405 [2] D Günther, B Heymel, J F Günther and I Ederer, Continuous 3D-printing for additive manufacturing, Rapid Prototyping Journal, 20 (4) (2014) 320-327 [3] Y Zhang, X He, S Du and J Zhang, Al2O3 ceramics preparation by LOM (Laminated Object Manufacturing), International Journal of Advanced Manufacturing Technology, 17 (7) (2001) 531-534 [4] G Marchelli, R Prabhakar, D Storti and M Ganter, The guide to glass 3D printing: developments, methods, diagnostics and results, Rapid Prototyping Journal, 17 (3) (2001) 187-194 [5] J J Beaman, C Atwood, T L Bergman, D Bourell, S Hollister and D Rosen, Additive/subtractive manufacturing research and development in Europe, WTEC Panel Report (2004), http://wtec.org/additive/report/additive-report.pdf [6] B Berman, 3-D printing: The new industrial revolution, Business Horizons, 55 (2) (2012) 155-162 [7] Y Chiu, Y S Liao and C C Ho, Automatic fabrication for bridged laminated object manufacturing (LOM) process, Journal of Materials Processing Technology, 140 (2003) 179-184 [8] A Das, G Madras, N Dasgupta and A M Umarji, Binder removal studies in ceramic tick shapes made by laminated object manufacturing, Journal of the European Ceramic Society, 23 (2003) 1013-1017 [9] G Yu, Y Ding, D Li and Y Tang, A low cost cutter-based paper lamination rapid prototyping system, International Journal of Machine Tools and Manufacture, 43 (11) (2003) 1079-1086 [10] Y Shuping, F Liu, J Zhang and S Xiong, Study of the key technologies of LOM for functional metal parts, Journal of Materials Processing Technology, 150 (2004) 175-181 [11] H Windsheimer, N Travitzky, A Hofenauer and P Greil, Laminated Object Manufacturing of Preceramic-Paper-Derived SiC Composites, Advanced Materials, 19 (24) (2007) 45154519 [12] L Weisensel, N Travitzky, H Sieber and P Greil, Laminated object manufacturing (LOM) of SiSiC composites, Advanced Engineering Materials, (11) (2004) 899-903 [13] H Zhong, X Yao, Y Zhu, J Zhang, D Jiang, J Chen, Z Chen, X Liu and Z Huang, Preparation of SiC Ceramics by Laminated Object Manufacturing and Pressureless Sintering, Journal of Ceramic Science and Technology, (2) (2014) 133-140 [14] C Gomes, N Travitzky, P Greil, W Acchar, H Birol, A P Novaes de Oliveira and D Hotza, Laminated object manufacturing of LZSA glass-ceramics, Rapid Prototyping Journal, 17 (6) (2011) 424-428 [15] D Klosterman, R Chartoff, G Graves, N Osborne and B D Olivier et al / Journal of Mechanical Science and Technology 31 (1) (2017) 133~139 Priore, Interfacial characteristics of composites fabricated by laminated object manufacturing, Composites Part A: Applied Science and Manufacturing, 29 (9) (1998) 1165-1174 [16] W Wang, J G Conley and H W Stoll, Rapid tooling for sand casting using laminated object manufacturing process, Rapid Prototyping Journal, (3) (1999) 134-141 [17] B Mueller and D Kochan, Laminated object manufacturing for rapid tooling and patternmaking in foundry industry, Computers in Industry, 39 (1) (1999) 47-53 [18] J Kechagias and V Iakovakis, A neural network solution for LOM process performance, International Journal of Advanced Manufacturing Technology, 43 (11-12) (2009) 12141222 [19] A P King Wah and A Joneja, Geometric techniques for efficient waste removal in LOM, Journal of Manufacturing Systems, 22 (3) (2003) 248-263 [20] J Kechagias, S Maropoulos and S Karagiannis, Process build-time estimator algorithm for laminated object manufacturing, Rapid Prototyping Journal, 10 (55) (2004) 297304 [21] B G Mekonnen, G Bright and A Walker, A Study on State of the Art Technology of Laminated Object Manufacturing (LOM), CAD/CAM, Robotics and Factories of the Future, Springer India (207-216) (2016) [22] J Park, M K Tari and H T Hahn, Characterization of the laminated object manufacturing (LOM) process, Rapid Prototyping Journal, (1) (2000) 36-50 [23] J Kechagias, S Maropoulos and S Karagiannis, Investigation of LOM process quality using design of experiments approach, Rapid Prototyping Journal, 13 (4) (2007) 316-323 [24] R I Campbell, M Martorelli and H S Lee, Surface roughness visualisation for rapid prototyping models, Computer-Aided Design, 34 (10) (2002) 717-725 [25] J Kechagias, An experimental investigation of the surface roughness of parts produced by LOM process, Rapid Prototyping Journal, 13 (1) (2007) 17-22 [26] D Ahn, J H Kweon, J Choi and S Lee, Quantification of surface roughness of parts processed by laminated object manufacturing, Journal of Materials Processing Technology, 212 (2012) 339-346 [27] G Chryssolouris, J Kechagias, P Moustakas and E Koutras, An experimental investigation of the tensile strength of parts produced by laminated object manufacturing (LOM) process, CIRP Journal of Manufacturing Systems, 32 (5) (2003) 319-322 139 [28] K Paul and V Voorakarnam, Effect of layer thickness and orientation angle on surface roughness in LOM, Journal of Manufacturing Processes, (2) (2001) 94-101 [29] Y S Liao, H C Li and Y Y Chiu, Study of laminated object manufacturing with separately applied heating and pressing, International Journal of Advanced Manufacturing Technology, 27 (7-8) (2006) 703-707 [30] O S Es Said, J Foyos, R Noorani, M Mandelson, R Marloth and B A Pregger, Effect of layer orientation on mechanical properties of rapid prototyped samples, Materials and Manufacturing Processes, 15 (1) (2000) 107-122 [31] Y S Liao and Y Y Chiu, Adaptive crosshatch approach for the laminated object manufacturing (LOM) process, International Journal of Production Research, 39 (15) (2001) 3479-3490 [32] C T Sun, Comparative evaluation of failure analysis methods for composite laminates, Technical report for the US Department of Transportation No DOT/FAA/AR-95/109 [33] D Olivier, J A Travieso-Rodriguez, S Borrós and G Reyes-Pozo, Flexural properties and failure mechanism assessment for additive manufactured LOM bars on different building orientations, Society of Plastics EngineersEUROTEC 2011 Conference Proceedings (2011) 1-5 Djamila Olivier holds a master’s degree in Materials Engineering She is currently a Ph.D candidate and an expert in 3D Printing Materials Her field of expertise focuses on the Development and Validation of new Polymeric Materials, and the Applications of 3D Printing and Additive Manufacturing Systems Ramón Jerez Mesa is an Industrial Engineer and a Ph.D candidate at the Mechanical Engineering Department of the Polytechnic University of Catalonia His research activity focuses on Ultrasonic Finishing Processes, the Characterization of Rapid Manufacturing, and 3D Printing via Laminated Object Manufacturing and Fused Deposition Modeling ... manufacturing orientation of LOM specimens can be inferred to influence their flexural behavior - The stronger direction of LOM components matches the orientation of the polymer chains because of the sheet... analyzing the four testing conditions, the influences of the macroscopic and microscopic contributions on the bonding of LOM components and their flexural strength are shown in Fig Despite the heterogeneous... is related to the strength of the raw laminated material, i.e., along the direction of alignment of the polymer molecules The second mecha- Table Configuration of the tested LOM specimens Specimen