Table 1.
Devices, materials and printing parameters for 3D printing.
Fig 1.
Test bodies and parameters for 3D printing comparison.
Individual test bodies shown for every analysed parameter and orientation within the printers.
Fig 2.
Processing of CAD files to printed parts (example X-resolution test body).
All test bodies were designed in Solidworks 2017 software (A) and converted into STL-file format (B). The STL-file was then processed into the device-specific G-codes using the respective company owned software for the FFF printers (C), SLA printer (D) and the SLS printer (E). The resulting printed part after post processing (ready to use) shows significant differences in surface quality and accuracy (F-H). The nozzle or laser path is shown in blue.
Fig 3.
Evaluation of printed test bodies using Fiji macros.
Microscopic pictures of the test bodies were taken and the area of interest was cropped (dotted line). The shape was identified and analysed using self-written Fiji macros (red line). (A) For X and Y resolution test bodies (X and Y) the maximum width and for the Z resolution test body (Z) the maximum height was measured (blue line). For the horizontal printed channels (B) as well as the vertical printed channels (C), the feret diameter and the roundness was calculated. (D) The angle of the appropriate test body was measured manually using the angle measuring tool from Fiji (red line). Scale bar equals 1000 μm.
Fig 4.
Accuracy of different 3D printing methods compared to each other.
Test bodies printed with FFF (Raise3D Pro 2 printer), SLA (Form 2 printer) and SLS (Lisa Pro printer) technique were analysed to their deviation of the respective geometry that was designed by CAD before. (A) Accuracy of the base axes X, Y and Z. (B) Diameter and roundness of horizontal and vertical printed channels. (C) Angled overhangs, printed without support structures. (D) Leakage investigated for different wall thicknesses. n = 3.
Table 2.
Evaluation of the printing techniques to each other.
Fig 5.
Overview of the 3D printing guidance.
(1) Download the STL-files of the test bodies attached to this publication. (2) Transfer the STL-files into G-code or similar, using the individual software and printer settings. (3) 3D print the test bodies. (4) Image the printed parts. (5) Download the Fiji macros attached to this publication and analyse the images after cropping the area of interest. (6) Compare the results to each other or to different printers if they match the desired quality or not. After printing, the sterilisation test (7), mechanical test (EN ISO 527 and 178) (8) or biocompatibility tests (ISO 10993–5) (9) can be performed.
Fig 6.
Accuracy of different FFF printers compared to each other.
Test bodies, which were FFF-printed with Raise3D Pro 2 (RP2), Ultimaker 3 (UM3) and Ultimaker S5 printer (UM5) were analysed regarding their deviation of the respective geometry that was designed by CAD before. Data from Raise3D Pro 2 are equal to Fig 4 (FFF). (A) Accuracy of the base axes X, Y and Z. (B) Diameter and roundness of horizontal and vertical printed channels. (C) Angled overhangs, printed without support structures. (D) Leakage investigated for different wall thicknesses. n = 3.
Fig 7.
Effect of autoclaving and plasma sterilisation to printed materials.
Tension (A) and bending (B) test were performed according to EN ISO 527 and 178. FFF printed material showed no change in material properties when treated by plasma sterilisation up to 3 times, but autoclaving significantly reduced tensile and flexural strength of the material. In SLA-printed material, no significant loss of tensile and flexural strength could be measured, but autoclaving tends to be more corruptive to the material over time. Materials printed by SLS show a significant increase of the tensile strength when sterilised for the first time, but no further impact when sterilised up to 3 times. In flexural strength, however, both methods tend to downgrade the material properties but do not show significances. All significances are referring to the control and are indicated by *. n = 3.
Fig 8.
Comparison of 3D printed and milled bioreactor system for the culture of perfused hydrogels.
Overview of the 3D printed and the milled bioreactor. Both bioreactors are closed with a top lid and a bottom lid, sealed via O-rings. The bottom is connected to the bioreactor via a plug connection, whereas the top lid is fixed with screws. In contrast to the 3D printed bioreactor, the milled bioreactor consists of nine parts. The bioreactor body has two opposing inlets to hold two peripheral venous catheters for perfusion. Barb-like structures protrude inwards from the openings to anchor the hydrogel. For the 3D printed bioreactor the barb-like structures are part of the bioreactor body. For the milled bioreactor they are screwed into the openings as well as the connectors for the peripheral venous catheters. Arrows indicate the medium flow.
Fig 9.
Workflow for starting the bioreactor with cell-laden collagen hydrogel.
HdF-containing collagen gel is cast in the assembled bioreactor enclosing the inserted cannula of the peripheral venous catheter. After 24 h, the cannula is removed. Thereby, a channel-like structure is created, which enables perfusion of the hydrogel with medium. HdmECs are seeded into the channel and the hydrogel is cultured under dynamic flow conditions. After 14 days, the hydrogel is removed from the bioreactor for further analysis. The arrow shows the location of the channel in the fixed hydrogel.
Fig 10.
Analysis of the biofabricated hydrogel with channel.
(A) The schematic representation of the collagen hydrogel shows the position of the channels and the sectional plane for the following cross-sections. (B) Qualitative MTT-staining of half of a tissue construct. The blue color indicates viable cells within the hydrogel. (C) The central channel is visible in the HE-stained cross-section of the hydrogel after 14 days of culture in the bioreactor under dynamic flow conditions. The inset shows in detail the immunohistological staining of the channel. It reveals CD31-positive cells lining the channel lumen, indicating colonisation with endothelial cells. (D) Immunofluorescence staining of cross-sections of the channel visualizes the presence of endothelial cells and fibroblasts. Positive staining for CD31 (red) shows endothelial cells at the channel surface and positive staining for vimentin (green) shows fibroblasts within the hydrogel. Cell nuclei are labeled with DAPI (blue). (E) Magnification of D. The asterisk marks the channel lumen. Scale bars 500 μm (C), 100 μm (D) and 50 μm (inset in C, E).