Edited by: Qiyin Fang, McMaster University, Canada
Reviewed by: Dalia Mahmoud, McMaster University, Canada; David Rosen, Georgia Institute of Technology, United States
This article was submitted to Medical Physics and Imaging, a section of the journal Frontiers in Physics
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Additive manufacturing machines, based on the multimaterial jetting technology, are widely used for three-dimensional (3D) printing of sophisticated medical models, which are aimed to be used for preoperative planning and surgical training. Gaining knowledge of process-related influences on mechanical and dimensional properties of 3D-printed parts makes up an essential basis for the design and manufacture of medical models. There are few studies on characterization of multimaterial parts, and those are limited to tests that are not based on standardized methods. Within the scope of this work, mechanical and dimensional investigations were performed on multimaterial parts that were printed using an Objet500–Connex3 3D printer (Stratasys Ltd., Minnesota, Eden Prairie, MN, USA). Among test methods listed in DIN EN ISO 17296-3, tensile tests were chosen for mechanical characterization. In the tensile tests, combinations of four different materials (Tango+, VeroClear, VeroPureWhite, MED610) were tested in three build orientations (XY, YX, ZX). To investigate the orientation-dependent printing accuracy, the tensile specimens were further checked for their dimensional accuracy. Statistically significant variations in the mechanical properties were found between different orientation levels. In general, specimens printed in XY orientation show higher tensile strength than YX- and ZX-oriented specimens. The tensile moduli determined are in the range from 0.2 to 2,500 MPa and compare well with the tensile moduli found in soft biological tissues. Dimensional deviations were found highest for the length of ZX-oriented tensile specimens. For this orientation level, it could be observed that multimaterial specimens, which contain higher percentage of the soft material Tango+, are characterized by higher shrinkage. For tensile specimens printed from the pure photopolymer Tango+, a shrinkage of 4.6% in length was determined. In summary, it was found that with multimaterial jetting technology, the increased shrinkage and lower mechanical strength in the ZX direction must be considered in the design process.
Applications of additive manufacturing (AM) technologies have quickly expanded from the field of traditional engineering to the field of medicine. The flexibility provided by AM enables surgeons to determine the most appropriate implant for each patient and to adapt and optimize the device design before surgery. This improves the performance of implants, lowers the risk of surgical complications, and reduces the duration of the surgery by eliminating the need for implant adjustments during intervention [
Moreover, AM allows the production of complex patient-specific medical models that are based on medical imaging data. AM medical models help doctors to better understand details of the patients' anatomy and the topographical relationship of anatomic structures and thus enhance knowledge and allow surgical training for specific treatments. In cardiovascular research, material jetting is one of the most widely applied AM technologies. This technology enables the production of colored and flexible structures. Medical models printed from flexible materials are especially suitable for planning implantation procedures of medical devices and allow practicing of surgical stitches and cuts. In addition, flexible materials provide possibilities to mimic the compliance of vessels. Because of the differences in mechanical properties of human tissue and polymeric plastics, implementation of cardiovascular models remains difficult, wherein the compliance can be varied over the wall thickness of the vessels [
Available material property specifications, provided by manufacturers, are specified in broad ranges, and influencing factors of the material jetting process (e.g., the anisotropy due to layer-wise build procedure) are not completely understood [
Material jetting is one of the most commonly used AM technologies for the production of medical models. Devices based on this technology are provided by two companies leading in the market, namely, Stratasys Ltd. and 3D Systems. The technology is trademarked under the names PolyJet™ (Stratasys Ltd.) and MultiJet™ (3D Systems), whereas, the correct terminology stated in EN ISO/ASTM 52900-15 is material jetting [
The printing block of the used system (Objet 500-Connex3) is equipped with eight printheads, wherein, always two of them are dedicated for one material, including the support material. The printheads themselves are equipped with numerous fine nozzles (96 per printhead) that allow achieving a high printing resolution of 600 dpi × 300 dpi in the X-Y plane. The material is applied during a double back-and-forth movement of the printing block along the X-axis and is instantly cured with UV light from halogen lamps. To ensure an even surface finish for the following layers, each applied layer is smoothed by means of a roller, located in the printing block. This further prevents collision of the printheads with excess material and thus avoids clogging of the fine nozzles. The layer thickness, which is achieved by lowering the build platform (Z-axis), depends on the chosen printing mode. In the so-called “Digital Material Mode” and in “High Speed Mode,” the layer thickness is 32 μm, whereas, “High Quality Mode” provides a layer thickness of 16 μm. Furthermore, the machine operator is free to choose between a “matte” or “glossy” finish of the topmost model surfaces. Usually, the printed model is completely embedded in support material, including the topmost surfaces to ensure evenness (matte surface finish). If this topmost support material layer is omitted, a glossy model surface is achieved [
In this work, the evaluation of single materials and combinations is reported. In particular, the aim of this work was to perform a standardized mechanical and dimensional characterization of multimaterial AM parts, which shall form the basis for manufacturing of medical models with known mechanical and dimensional properties that can be used for surgical training and preoperative planning.
For characterization of the mechanical properties, standard tensile tests for plastic materials were selected from the test methods listed in DIN EN ISO 17296-3 (compare [
List of studied materials and classification in soft and rigid ones.
Tango+ | RGD8625 |
FLX9950 | VeroClear |
FLX9970 | VeroPureWhite |
FLX9995 | MED610 |
Effects on the mechanical properties were examined for three different build orientations (XY, YX, ZX). A schematic illustration of the resulting anisotropy caused by the unique build procedure is given in
Studied build orientations of the tensile test specimens for investigating the mechanical and dimensional properties, depending on the orientation. Subsequent layers are indicated by an alternating orange-and-white shading, whereas, intersections within the layers, resulting from material deposition of adjacent nozzles, are indicated as cylinders (not in scale).
All type 1A tensile specimens were printed on an Objet 500-Connex3 3D printer in Digital Material Mode with matte surface finish. This printing mode was selected, as it is the only printing mode that allows mixing of materials. In contrast to glossy surface finish, matte surface finish provides a very similar surface quality for all side surfaces of the printed parts. In addition, matte surface finish allows printing of sharp edges, as the parts are completely covered in support material, which prevents the deposited uncured layers from running [
Schematic illustrations (generated in Objet Studio software) of the printing trays. Various colors of the tensile specimens indicate different materials. ZX-oriented specimens are enclosed in an additional support construction (in white color).
Support removal was performed in a custom-made water jet cabin. To remove fine residues of support material, specimens were additionally soaked in a freshly prepared 3% NaOH solution (sodium hydroxide, pearl, 97%; Thermo Fisher Kandel GmbH, Kandel, Germany) for a duration of 15 min. Specimens were then rinsed under running water. To avoid long-term water contact, specimens were immediately dried after support removal by means of a cloth. Support removal of the specimens made from the medical-grade material MED610 was carried out following the guidance “Bio-Compatibility Requirements” [
After support removal, specimens were conditioned for 88 h at a standard atmosphere class 2 (23 ± 2°C and 50 ± 10% r.h.) [
All tensile tests were performed on a uniaxial universal test machine (Beta 10-2,5; Messphysik Materials Testing GmbH, Fürstenfeld, Austria). The used test machine fulfills all requirements listed in DIN EN ISO 527-1/2 and the referred standard ISO 7500-1 [
Tensile test machine including a video extensometer system.
For testing specimens printed from rigid materials, a 5-kN force transducer (TC4/5kN; AEP transducers, Cognento, Italy) was used. The preload value was set to 25 N and is within the limits defined in DIN EN ISO 527-1. As specified for this specimen type, the initial clamping distance is set to 115 mm. The distance of the video extensometer camera was adjusted to obtain a field of view of ~100 mm. Measurements of the tensile modulus are performed at a testing speed of 1 mm/min within the specified strain interval of 0.05–0.25%. The regression method was used to determine the tensile modulus. After a strain value of 0.3% is reached, the testing speed was switched to 20 mm/min, which corresponds to a nominal strain rate of 17.4% min−1. This testing speed was selected among specified values, listed in the standard. In regard to the high viscosity of the materials under investigation, the testing speed was chosen as slow as possible, but still allowing testing of all soft and rigid materials within a reasonable time. Because of the high ductility of the soft materials, different force transducers were required. Determination of the tensile modulus was carried out using a 20-N force transducer (S2/20N; Hottinger Baldwin Messtechnik GmbH, Darmstadt, Germany). After determination of the tensile modulus, a 1-kN load cell (TS C2 1kN, AEP transducers) was used to measure the tensile strength and elongation at break.
Stratasys Ltd. specifies the printing accuracy of the Objet500–Connex3 printer with 85 μm for parts with dimensions smaller than 50 mm and with 200 μm for dimensions >50 mm. These tolerances are provided only for rigid materials, whereas, no information is given about the tolerances for soft materials such as, for Tango+ [
Measuring points located on the type 1A tensile specimen.
DIN EN ISO 527-1 stipulates that dimensions of tensile test specimens, made from rigid plastic materials (
Data evaluation and statistical analysis were done in MATLAB R2017b (The MathWorks Inc., Natick, MA, USA). For statistical testing of orientation-dependent differences in the mechanical properties and in the dimensional accuracy, a one-way analysis of variance (ANOVA) was performed. Adequacy of the measurement data was checked by normal probability plots and by Shapiro–Wilk tests for normality (standard 5% significance level). The decision for this test is based on the highest statistical power compared to other tests for normality (Kolmogorov–Smirnov, Anderson–Darling) [
Mean tensile properties and corresponding standard deviations (SDs) of the tested material types are given in
Mean mechanical data (
Tango+ | XY | 0.359 | 0.059 | 0.472 | 0.017 | 122.50 | 4.27 |
YX | 0.277 | 0.118 | 0.438 | 0.017 | 114.77 | 5.36 | |
ZX | 0.442 | 0.106 | 0.305 | 0.023 | 86.04 | 7.19 | |
FLX9950 | XY | 0.975 | 0.030 | 0.904 | 0.017 | 82.20 | 3.80 |
YX | 0.819 | 0.078 | 0.787 | 0.022 | 84.44 | 2.20 | |
ZX | 0.901 | 0.169 | 0.409 | 0.012 | 67.75 | 2.75 | |
FLX9970 | XY | 2.999 | 0.073 | 1.758 | 0.055 | 62.40 | 1.82 |
YX | 2.134 | 0.082 | 1.480 | 0.049 | 66.55 | 2.03 | |
ZX | 1.873 | 0.333 | 0.573 | 0.046 | 49.03 | 4.92 | |
FLX9995 | XY | 26.55 | 1.77 | 4.955 | 0.084 | 52.20 | 1.32 |
YX | 12.47 | 0.50 | 4.617 | 0.090 | 55.47 | 0.80 | |
ZX | 11.35 | 0.49 | 1.601 | 0.148 | 33.32 | 2.93 | |
RGD8625 | XY | 1,566 | 10 | 39.46 | 0.33 | 12.32 | 3.36 |
YX | 1,537 | 32 | 38.46 | 0.41 | 11.84 | 1.98 | |
ZX | 1,650 | 27 | 17.86 | 0.72 | 1.14 | 0.05 | |
VeroClear | XY | 2,490 | 50 | 63.43 | 0.48 | 8.32 | 2.22 |
YX | 2,396 | 50 | 62.32 | 0.36 | 4.88 | 0.21 | |
ZX | 2,087 | 45 | 18.18 | 1.63 | 0.98 | 0.12 | |
VeroPureWhite | XY | 2,340 | 13 | 58.82 | 0.44 | 8.66 | 2.69 |
YX | 2,264 | 41 | 57.16 | 0.25 | 6.01 | 0.58 | |
ZX | 1,814 | 33 | 18.07 | 1.47 | 1.01 | 0.09 | |
MED610 | XY | 2,475 | 33 | 62.57 | 0.39 | 5.39 | 0.35 |
YX | 2,422 | 29 | 61.32 | 0.56 | 4.18 | 0.21 | |
ZX | 1,955 | 59 | 8.22 | 1.49 | 0.48 | 0.08 |
Boxplots of the mechanical data of all tested materials at different orientation levels (
Soft materials. Sample comparison of the stress-strain curves of the different materials and orientation considered.
Results of the performed one-way ANOVA show statistically significant differences for all materials, except for the tensile modulus of FLX9950. Obtained
Tango+ | XY vs. YX | 0.4956 | 0.1375 | |
XY vs. ZX | 0.4956 | |||
YX vs. ZX | ||||
FLX9950 | XY vs. YX | 0.1147 | 0.6836 | |
XY vs. ZX | 0.8812 | |||
YX vs. ZX | 0.6897 | |||
FLX9970 | XY vs. YX | 0.1290 | ||
XY vs. ZX | ||||
YX vs. ZX | 0.1225 | |||
FLX9995 | XY vs. YX | 0.0811 | ||
XY vs. ZX | ||||
YX vs. ZX | 0.4787 | |||
RGD8625 | XY vs. YX | 0.1825 | 1.0000 | |
XY vs. ZX | ||||
YX vs. ZX | ||||
VeroClear | XY vs. YX | 0.2214 | ||
XY vs. ZX | ||||
YX vs. ZX | ||||
VeroPureWhite | XY vs. YX | |||
XY vs. ZX | ||||
YX vs. ZX | ||||
MED610 | XY vs. YX | 0.1442 | 0.1117 | |
XY vs. ZX | ||||
YX vs. ZX |
All mechanical measurement data and test protocols are available upon request.
Mean dimensional data and corresponding SD, determined on the tensile test specimens, can be taken from
Mean dimensional data (
Tango+ | XY | 3.84 | 0.03 | 9.755 | 0.032 | 174.88 | 0.13 |
YX | 3.83 | 0.02 | 9.894 | 0.040 | 173.71 | 0.18 | |
ZX | 3.82 | 0.01 | 10.102 | 0.050 | 167.02 | 0.25 | |
FLX9950 | XY | 3.91 | 0.03 | 9.850 | 0.025 | 175.46 | 0.09 |
YX | 3.90 | 0.02 | 9.943 | 0.039 | 174.93 | 0.06 | |
ZX | 3.93 | 0.01 | 10.054 | 0.035 | 169.21 | 0.29 | |
FLX9970 | XY | 3.96 | 0.02 | 9.892 | 0.029 | 174.98 | 0.07 |
YX | 3.91 | 0.02 | 9.928 | 0.045 | 175.01 | 0.11 | |
ZX | 3.94 | 0.02 | 10.040 | 0.040 | 171.20 | 0.24 | |
FLX9995 | XY | 4.04 | 0.03 | 10.010 | 0.032 | 174.89 | 0.07 |
YX | 3.95 | 0.02 | 9.902 | 0.030 | 174.92 | 0.05 | |
ZX | 4.05 | 0.01 | 10.072 | 0.051 | 173.49 | 0.11 | |
RGD8625 | XY | 4.019 | 0.029 | 9.884 | 0.017 | 175.10 | 0.04 |
YX | 4.012 | 0.032 | 9.904 | 0.033 | 175.03 | 0.03 | |
ZX | 4.127 | 0.023 | 10.147 | 0.026 | 175.11 | 0.02 | |
VeroClear | XY | 4.019 | 0.034 | 9.921 | 0.034 | 174.99 | 0.01 |
YX | 4.064 | 0.041 | 9.858 | 0.021 | 175.08 | 0.02 | |
ZX | 4.162 | 0.020 | 10.102 | 0.023 | 175.22 | 0.05 | |
VeroPureWhite | XY | 4.012 | 0.035 | 9.838 | 0.037 | 174.93 | 0.01 |
YX | 4.073 | 0.037 | 9.814 | 0.017 | 174.94 | 0.01 | |
ZX | 3.966 | 0.017 | 9.947 | 0.026 | 175.22 | 0.04 | |
MED610 | XY | 3.991 | 0.038 | 9.998 | 0.035 | 175.05 | 0.02 |
YX | 4.046 | 0.038 | 9.967 | 0.034 | 175.09 | 0.02 | |
ZX | 4.180 | 0.013 | 10.212 | 0.025 | 174.98 | 0.04 |
Tango+ | XY vs. YX | 0.1374 | ||
XY vs. ZX | ||||
YX vs. ZX | 1.0000 | |||
FLX9950 | XY vs. YX | |||
XY vs. ZX | ||||
YX vs. ZX | ||||
FLX9970 | XY vs. YX | 1.0000 | ||
XY vs. ZX | ||||
YX vs. ZX | ||||
FLX9995 | XY vs. YX | 1.0000 | ||
XY vs. ZX | 0.2174 | |||
YX vs. ZX | ||||
RGD8625 | XY vs. YX | 0.9058 | ||
XY vs. ZX | 1.0000 | |||
YX vs. ZX | ||||
VeroClear | XY vs. YX | |||
XY vs. ZX | ||||
YX vs. ZX | ||||
VeroPureWhite | XY vs. YX | 0.7871 | ||
XY vs. ZX | ||||
YX vs. ZX | ||||
MED610 | XY vs. YX | 0.1394 | ||
XY vs. ZX | ||||
YX vs. ZX |
Boxplots of the dimensional data of all tested materials at three different orientation levels (
Rigid materials. Sample comparison of the stress-strain curves of the different materials and orientation considered.
Illustration of the material shrinkage of ZX-oriented tensile specimens, manufactured from different digital materials (combination of Tango+ and VeroClear). The rigid material RGD8625 shows the lowest deviation (+0.11 mm on average) from the nominal length (l3 = 175 mm).
All dimensional measurement data and test protocols are available upon request.
Results of the tensile tests have shown the lowest tensile strength for ZX-oriented specimens, followed by YX and XY orientations. While only small differences in strength were found between XY and YX orientations, significantly lower values were determined on ZX-oriented specimens. For rigid materials, ZX-oriented specimens have shown a tensile strength of only 13–45% compared to the strength of XY-oriented specimens. Lower differences in tensile strength were found for the soft materials, whereby, ZX-oriented parts have shown 32–65% of the tensile strength of XY orientations.
Highest variations in the determined tensile moduli could be observed for the softer materials (as shown by the SDs in
In general, rigid pure materials (VeroClear, VeroPureWhite, MED610) follow the same trends, whereby, the tensile modulus, tensile strength, and the nominal strain at break are the highest for XY- oriented specimens, followed by YX and ZX orientations. Highest effects on the mechanical properties were found for ZX-oriented specimens, whereby, the highest difference in tensile strength between orientations was found for MED610. ZX-oriented specimens of this material provide only 13% of the tensile strength of XY-oriented specimens. This could be attributed to the much longer time of chemical loading during the support removal procedure, which was carried out in accordance with the guidance for biocompatibility requirements, provided by the supplier (Stratasys Ltd.). Because of the much higher surface roughness of ZX-oriented specimens, these specimens show a much larger surface area and thus might facilitate impregnation of chemical compounds such as NaOH and isopropanol into the material. Differences in tensile strength between flat-positioned specimens and ZX-oriented specimens, made from soft materials, decrease for material mixtures containing a higher percentage of Tango+ (
In comparison to the flat-positioned specimens, the build procedure of ZX-oriented specimens takes much longer, leading to a higher UV exposure than for flat-positioned parts. In general, higher UV exposure is assumed to result in higher mechanical strength of the parts. Nevertheless, overcuring of the parts could also lead to a more brittle material behavior as seen in rigid ZX-oriented specimens [
Tensile modulus of soft tissues found in literature, reproduced from McKee et al. [
Tendon | 43–1,660 | ~560 |
Muscle | 480 | 480 |
Skin | 21–39 | ~30 |
Liver and kidney | 1–15 | ~10 |
Cornea | 0.1–11.1 | ~3.0 |
Sclera | 0.6–4.9 | ~2.7 |
Spinal cord and gray matter | 0.4–3.6 | ~2 |
Artery and vein | 0.6–3.5 | ~2 |
Determined mechanical values of the investigated materials can be used for simple computational strength calculations. Consideration of anisotropic effects allows correct dimensioning of parts that are mechanically loaded. This is essential for the design of medical devices such as for cutting guides or patient-specific surgical instruments. Especially keeping the mechanical weakness of Z-oriented structures in mind will help to avoid wrong part dimensioning and failure.
Results of the dimensional measurements carried out on the rigid tensile test specimens have shown higher dimensional deviations than specified by the manufacturer (Stratasys Ltd.). Without regard to the build orientation, observed deviations from the nominal dimensions are in the range between −0.2 and +0.2 mm for rigid materials. The observed greater width and thickness of ZX-oriented specimens could be attributed to wobbling of the parts during the build procedure, resulting in a larger width and thickness of the specimens. Although, ZX- aligned parts were printed within a support structure to avoid such movements during printing, this assumption is strengthened by the fact, that those specimens have shown slightly larger dimensions at their upper ends. In general, it can be stated that XY-oriented specimens show the lowest deviations from the nominal thickness and from the nominal width, followed by YX orientations. The higher accuracy in width of XY-oriented specimens could be attributed to the fact that the deposition process of the model and support material takes place simultaneously for dimensions aligned in parallel to the Y-axis of the printer. Not all specimens fulfill the dimensional tolerances (specimens made from MED610), stipulated in DIN EN ISO 527-2. The tolerance of the height h is specified with 4 ± 0.2 mm, whereas, the width b1 is specified with 10 ± 0.2 mm.
Specimens made from soft materials have shown other trends for their dimensional accuracy, compared to that of specimens made from rigid materials (
As the specimens were printed all together (rigid and soft specimens) within a support structure (
For investigating the isotropic behavior of the material shrinkage, it would be most valuable to compare the shrinkage in length, as well as of the other dimensions (width and thickness) of the XY-, YX-, and ZX-aligned specimens. When comparing the length, it could be observed that shrinkage was mainly an issue in the ZX-oriented soft specimens. A reason for the larger shrinkage in the ZX direction could be due to the higher UV exposure in ZX printing orientation. This is mainly due to the fact that a print layer has a height of 32 μm. This results in a number of 125 layers for the flat-positioned specimens, whereas, more than 5,000 layers were required for the specimens produced in the ZX direction. A longer UV exposure time, as it occurred in the ZX direction, or after a postcuring treatment, typically leads to some shrinkage. We therefore think that this longer UV exposure could be the reason for the greater shrinkage in the ZX orientation; however, further investigation would be necessary to thoroughly understand this phenomenon.
A visual comparison of ZX-oriented specimens is given in
Similar trends of orientation effects could be observed for the width of the materials Tango+, FLX9950, and FLX9970. For those specimens, the width was found to be the smallest for XY orientations, followed by YX and ZX orientations. These materials have shown the lowest deviation in width for ZX- and YX-oriented specimens. Without regard to the orientation level, deviations in width and in the thickness are in the range of −0.2 to +0.2 mm for the soft materials.
In general, it could be observed that specimens made from material mixtures, containing a higher percentage of Tango+, show a smaller thickness. Although statistically significant differences between most orientation levels were found for the thickness, no similar trends could be observed for the different material mixtures. Trends between the different orientation levels could be observed for the width of the specimens, whereby, the width was observed to be the smallest for XY-oriented specimens, followed by YX and ZX orientations. As for the rigid materials, it could be assumed that the width of the ZX-oriented specimens was printed larger due to wobbling during the build procedure.
As for the rigid material types, the differences between the mean widths of XY- and YX-oriented specimens could be attributed to the different material deposition process, whereby, the printing resolution in Y direction is limited by discretized nozzle positions of the printheads. This leads to systematic deviations in dimensions aligned along the Y-axis of the printer (width of XY aligned specimens and length of YX aligned specimens). Although, the specimens were printed on different tray positions, these systematic deviations did not reflect in the obtained data, as normal probability plots have shown that dimensions, which should be affected by discretized nozzle positions (width of XY specimens and length of YX specimens), are normally distributed. This suggests that the used measurement setup is not capable of showing this influencing factor.
Gained knowledge of orientation-dependent dimensional printing accuracy helps to ensure keeping necessary tolerances for manufacturing of anatomical models. This is especially relevant if those models are aimed to be used for correct sizing of implants, such as for selection of prosthetic heart valves or fitting surgical plates on anatomical bone structures. As shown in this article, keeping orientation-dependent accuracies in mind gets even more important if soft materials are used for printing. It seems that shrinkage of Z-oriented structures, printed from soft materials, is linear and allows simple correction by scaling the height of the model by applying a constant scaling factor.
Within this study, materials were only tested for their pure elastic behavior. During testing, it could be clearly observed that the materials show high viscosity. An investigation of the viscoelastic properties seems relevant for the design of biomechanical models, especially for cardiovascular applications. Moreover, there are additional materials that need to be characterized if the mechanical properties of the tested materials do not fit for the desired application field. Besides dimensional characterization of the printer, also geometrical (e.g., roundness, parallelism, etc.,) accuracies need to be determined. A potential source of error for all conducted experiments are inconsistent time intervals between manufacturing of the specimens and the performed test [
AM is still a very dynamic and innovative field, which gives rise to a lot of new possibilities in the medical sector. Current standards for AM part characterization are still in development, and up to now, only a few general standards exist. Besides the possibility of different part orientations in AM technologies, resulting in different mechanical characteristics, multimaterial printers, such as the investigated Objet500–Connex3 printer, pose additional challenges for process understanding and require a lot of effort for full machine-material characterization. Published studies on multimaterial part characterization are rare and show limitations based on non-standard compliant investigations. Within the conducted tensile tests, influences of different build orientations on the mechanical properties were investigated. Eight different materials were tested in three different build orientations. With a few exceptions, it turned out that XY-oriented specimens show the highest tensile modulus, followed by YX and ZX orientations. For the tensile strength, a similar trend was found. In general, the lowest strain at break was found for ZX-oriented specimens. For investigation of the influence of orientation effects, dimensional tests were performed on tensile test specimens. For ZX-oriented specimens made from soft materials, high shrinkage (up to 4.6% of the nominal value) was found in the length of the specimens. In general, it can be said that the width and thickness of flat-oriented specimens are printed with an accuracy of ±0.2 mm for all investigated materials.
The original contributions generated for the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.
MK was involved in literature research, study design, experimental work, data analysis, writing the manuscript, and submission process. MS, EU, and CG were involved in study design and 3D printing. FM was involved in funding acquisition, study design, data analysis, writing, and submission process. All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.