Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Design of Complexly Graded Structures inside Three-Dimensional Surface Models by Assigning Volumetric Structures

Design of Complexly Graded Structures inside Three-Dimensional Surface Models by Assigning... Hindawi Journal of Healthcare Engineering Volume 2019, Article ID 6074272, 10 pages https://doi.org/10.1155/2019/6074272 Research Article Design of Complexly Graded Structures inside Three-Dimensional Surface Models by Assigning Volumetric Structures Ronny Bru ¨nler , Robert Hausmann, Maximilian von Mu ¨nchow, Dilbar Aibibu, and Chokri Cherif Institute of Textile Machinery and High Performance Material Technology (ITM), Technische Universita¨t Dresden, Hohe Str. 6, 01069 Dresden, Germany Correspondence should be addressed to Ronny Bru¨nler; ronny.bruenler@tu-dresden.de Received 11 December 2018; Accepted 8 January 2019; Published 4 February 2019 Guest Editor: Saverio Maietta Copyright ©2019 Ronny Bru¨nler et al. *is is anopen accessarticle distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. An innovative approach for designing complex structures from STL-datasets based on novel software for assigning volumetric data to surface models is reported. *e software allows realizing unique complex structures using additive manufacturing technologies. Geometric data as obtained from imaging methods, computer-aided design, or reverse engineering that exist only in the form of surface data are converted into volumetric elements (voxels). Arbitrary machine data can be assigned to each voxel and thereby enable implementing different materials, material morphologies, colors, porosities, etc. within given geometries. *e software features an easy-to-use graphical user interface and allows simple implementation of machine data libraries. To highlight the potential of the modular designed software, an extrusion-based process as well as a two-tier additive manufacturing approach for short fibers and binder process are combined to generate three-dimensional components with complex grading on the material and structural level from STL files. With these methods, different materials cannot be applied 1. Introduction within a part, in particular not within one layer. Additive manufacturing technologies are based on the In the case of the open-space or layer construction layered construction of material into a finished component. methods, it is possible to use various materials [23, 24]. *e production is carried out by solidification of However, the limitations for parts with complex material and structural composition lie in the properties of the file (i) Liquids or gels (curing, drying, crosslinking, etc.) formats [25]. *e most common file format in generative [1–9] production methods is the STL format (standard tessellation (ii) Powders or granules (gluing, sintering, etc.) [10–16] language/standard triangulation language) [26–28]. In niche or applications, the AMF format (additive manufacturing (iii) Pasty and belt-shaped or strand-shaped materials format) and the OBJ format (object format) are used as well (direct deposition without curing processes or so- [29, 30]. lidification, etc.) [17–20] A sphere is used to illustrate the file formats. *e STL file *e limitations for realizing parts with a complex ma- format provides only surface information and uses triangles, terial composition are either found in the process charac- often referred to as vertices, to represent the surfaces as shown in Figure 1(a). In the AMF or OBJ format, the surface in- teristics or in the data formats used. Technologies using powder beds or liquid baths are limited formation can also be supplemented with properties (mate- to a particular material or a particular material composition rial, textures, or metadata), emphasized by a red color scheme which is constant throughout the entire component [21, 22]. in the triangles on the top of the sphere in Figure 1(b). 2 Journal of Healthcare Engineering (a) (b) Figure 1: (a) Visualization of the surface representation of a sphere by means of triangles in the STL format [31]. (b) *e same sphere in the AMF or OBJ format that allow assigning descriptive metadata to individual triangles, shown here as red color scheme in the upper part of the sphere. STL files as well as AMF or OBJ files represent geometric file can be observed in comparison with the anatomy of a bodies exclusively by means of surface information. Re- vertebra (Figure 3(c)). *e STL file features a hollow body and shows almost no structures in the area of the spongy gardless of whether the files are stored as surface or solid bodies, they contain no volume data. Basically, the objects bone in the center of vertebral body and infinitesimally thin walls instead of the dense cortical bone around the spongy are hollow inside and have “outer walls” that are in- finitesimally thin. *e only difference in solid bodies is the bone structure. *ere is no software yet available to fill representation of a filled body. Figure 2(a) shows a graphic certain regions with different materials or structural varia- representation of the previously discussed sphere cut in half tions within STL, AMF, or OBJ files of parts with complex using the software Blender (Blender Foundation). As in all geometry. other software for editing or displaying STL files or similar formats (Netfabb, Cura, Slic3r, and Repetier, amongst 2. Materials and Methods others) as well as CAD software (FreeCAD, SolidWorks, AutoCAD, and CATIA, amongst others), only the surface of 2.1.SoftwareforAccessingtheInnerStructureofSurface-Based the structure can be addressed as it is simply impossible to Bodies. In addition to surface-based file formats, bodies can select or click on other structures than the surface triangles. also be represented by volumetric elements (voxels). Graphic *is points up the decisive limitation of all surface-based file representations of spheres in voxel format with assigned formats: within surface-approximated geometries, for ex- metadata are shown in Figure 4. *is approach is mostly ample, from computer tomography recordings (CT) and known from video games such as Minecraft (Mojang/ magnetic resonance tomography representations (MRT) or Microsoft Studios, 2009) and Blade Runner (Virgin In- 3D scans, property assignments cannot be implemented. teractive, 1997) or simulations [34] to represent terrain fea- In CAD programs, however, objects that contain volume tures and is also widely used in medical imaging formats such information can be designed and stored in the AMF or OBJ as DICOM [35, 36]. However, the voxel formats are not used format, but again have to be regarded as individual surfaces for designing implants for regenerative medicine, prosthetic approximated by triangles. It is thus possible to realize components, or 3D printing applications in reverse engi- structures with grading on the material or structural level, neering as slicer software is usually developed for STL files. supposing that they are designed from scratch as the two- Hence, while it is possible to 3D print or display complex color sphere in Figure 2(b). geometries with different surfaces, it is not possible to realize However, assigning properties within previously defined grading on the material or structural level inside the or given bodies as obtained from CT scans, MRI scans, or structures. radiographs for the determination of defect geometries in To access the inner structure of surface-based bodies, regenerative medicine or 3D scans from reverse engineering novel software for segmenting the structures into voxels and cannot be carried out in these programs because of the manipulating them is developed. *e software is capable of surface-based representation and the related restrictions. processing STL files in ASCII format and is developed in C# Figure 3(a) shows a CT scan of the lower spine and an STL within the development environment Visual Studio Com- file derived from that scan containing geometry information munity (Microsoft Corp.). Figure 5 shows the graphical user of a lumbar vertebra exhibiting a complex geometry. Sub- interface (GUI) that was created within the Windows stantial differences in the structural composition of the STL Presentation Foundation (WPF) framework. Journal of Healthcare Engineering 3 (a) (b) Figure 2: (a) Visualization of the structure of the sphere in surface-based formats that solely allow selecting and thus altering the surface triangles. (b) Image of a sphere with two colors designed in a CAD software as a solid body. Hollow body Thin walls (a) (b) (c) Figure 3: (a) CTscan of the lower spine showing the L3 (top), L4 (middle), and L5 (bottom) lumbar vertebrae [32]. (b) View of the STL file of an L5 lumbar vertebra from [32] exhibiting infinitesimally thin walls and a hollow body. (c) Anatomy of the lumbar vertebrae [33]. (a) (b) Figure 4: (a) Visualization of a sphere in voxel format. (b) *e same sphere with a red color scheme in the upper part of the sphere. Descriptive metadata can be assigned to individual voxels even throughout the inner parts of the body. After importing a file, it is automatically converted into and grid size can be adjusted arbitrarily and separately. In the AMF format and can be viewed, rotated, and zoomed in default mode, the grid size matches the slicing thickness. and out on the “AMF” tab (Figure 5(a)). *e “Slice” tab as *us, the grid subdivides the body into cubic voxels. visualized in Figure 5(b) is used to slice the body and to Generating the slice data for a graphical representation additionally implement a rectangular grid. Slicing thickness of the contours of the body requires the consideration of 4 Journal of Healthcare Engineering (a) (b) Figure 5: Graphical user interface of the developed software: (a) “AMF” tab with imported sphere; (b) “Slice” tab showing the midlayer of the subdivided sphere, the input field for adjusting slicing thickness, and the scroll bar for scrolling through the single slices. several special cases related to the surface representation by layer in the slice tab. *e software layout also allows pro- means of triangles. During the generation of a single slice, cessing more complex bodies, such as the “test your 3D necessarily some triangles are cut by the section plane as evident printer! v2” file from *ingiverse (MakerBot Industries, in Figure 6(a). In total, there are 10 cases describing the situ- LLC) as illustrated in Figure 7(b) [38]. ational relations between section plane and triangles [37]. While simple cases such as triangles located completely above or below 3. Results and Discussion the section plane are easy to process, five specific cases have to be considered more closely (Figure 6(b)): 3.1. Assignment of Multiple Properties and g-Code to Sub- volumes within Given Complex Geometries. *e imported (1) All three edges are located on the section plane body is divided into voxels by the rectangular grid imple- (2) Exactly two edges are located on the section plane mented in the “Slice” tab. *us, all voxels are accessible by (3) One edge is located above, one below, and one on the scrolling through the single layers. *ese features enable section plane assigning arbitrary properties to every single voxel within any given geometry. (4) One edge is located above/below the section plane *e properties assigned to each voxel are used to gen- and two edges are located on the other side erate machine-readable code. As g-code is the most common (5) Exactly one edge is located on the section plane numerical control programming language, it is used for *ese considered cases have a significant effect on the further processing. Specific commands are filed by means of calculation of the intersection points. *e coordinates of a user-friendly editable .txt file. It can be adjusted depending the edges of all triangles describing the edited body are on the manufacturing technology used. *e standard file for stored in a separate class (triangle class) within the soft- extrusion-based additive manufacturing processes is shown ware. *ey are processed during slice data generation and in Figure 8(a). It contains the material id (matid), the serve for intersection point calculation. After using a number of layers necessary to fill one millimeter (fillings/ certain coordinate for calculation, it may be erased from the mm), the path distance (hatch), the path arrangement to-be-processed data. However, depending on the case and (pattern), the digital output command being used to control on the ratio between slicing thickness and triangle size, the extrusion nozzle (tool), the speed of the tool (speed), and certain coordinates have to be used for the calculation of the zero position of the tool used (X-, Y-, and Z-co- the next layer and must not be erased. To ensure a correct ordinates). *e strand thickness can be calculated from the calculation, novel triangles are generated according to “fillings/mm” column as these values are used for the travel Figure 7(a). *e triangle (defined by a, b, and c in of the z-position to lay the strands directly on top of each Figure 7(a)) he is divided into three smaller triangles using other. *us, according to Figure 8(a), 8 fillings/mm are used the intersection points (P1 and P2) with the current section for a strand thickness of 125µm and 4 fillings/mm are used plane. In the case described here, the coordinates a and b for 250µm strands, respectively. can be erased from the triangle class, and only the triangle *e absolute positions of the voxels and the assigned defined by P1, P2 and C is used for calculating the in- fillings, hatches, tools, patterns, speeds, and positions are tersection points in the next layer. used to generate g-code commands. *e black paths in After calculation of all intersection points, a polygon Figure 8(b) directly show the course of the generated path in course is generated automatically and displayed for each g-code. Travels in z direction are handed over from the Journal of Healthcare Engineering 5 Case I Case II Case III Case IV Case V (a) (b) Figure 6: (a) Slicing of a sphere with resulting cuts of the surface triangles. (b) Specific cases for situational relations between the section plane and surface triangles during slicing [37]. i +2 P2 P1 i +3+ k i +2+ k A B Index: i i +1 (a) (b) Figure 7: (a) Method for calculating the polygon course for a graphic representation of single slices. (b) Polygon course of the base layer of a complex 3D structure [38]. Matid Filling (mm) Hatch Pattern Tool Speed X Y Z 1 8 125 zz 1 200 50 50 0 2 8 250 zz 1 200 50 50 0 3 4 250 zz 2 500 150 50 0 4 4 500 zz 2 500 150 50 0 (a) (b) Figure 8: (a) Standard file for extrusion-based printing with two different colored materials and different deposition patterns. *e voxel size in this figure is 1mm and is filled in z direction with 8 red material strands (tool 1; red material; strand thickness 125µm; traversing speed 200mm/min; zero position of the nozzle with respect to tool installation in our machine) and 4 blue strands (tool 2; blue material; strand thickness 250µm; traversing speed 500mm/min; zero position of tool #2), leading to (b) different voxel filling patterns with adjusted path distances (hatch) for the same material used. 6 Journal of Healthcare Engineering information provided by the “fillings/mm” column and the slice thickness. *e file shown in Figure 8(a) leads to different material deposition. *e patterns with “matid 1” and “matid 2” are extruded from the same nozzle (tool 1) and with the same speed but in different path distances. For “matid 1,” the strand thickness matches the path distance and thus fills the voxel area in top view creating a dense structure (Figure 8(b), top left). “Matid 2” deposits the 125µm strands in a path distance of 250µm and thus exhibit filling levels of 50%. In “matid 3” and “matid 4,” another tool is used to deposit larger strands in different patterns. Figure 9: Manufacturing of material grading (left), porosity Figure 9 shows three different structures manufactured grading (center), and material and porosity grading (right) by from the same cuboid STL file (20mm ×20mm ×4mm). means of assigning different nozzles (tool column) and strand Two extruding systems with nozzles 0.4mm in diameter spacings (hatch column). (Nordson EFD) filled with different colored clay were used to manufacture different structures within the STL file. triangles of the processed structure and the hardware used. *e left structure was manufactured with both nozzles *e process of loading and processing the vertebra structure using the same path distance leading to a laydown of one that is defined by 34464 surface triangles into voxels took material (light-blue colored clay) in the inner zone and the 71.6seconds on a dual-core 2.4GHz, 4GB RAM, 256MB other material (orange colored clay) in the outer part while graphics memory system and 14.5seconds on an eight-core both regions featuring a dense path spacing. *e structure in 3.4GHz, 32GB RAM, 1GB graphics memory system, making the middle was manufactured using solely one nozzle fol- the software applicable on a wide range of computing systems. lowing dense path spacing in the outer part and a 0.8mm Contiguous areas are processed with continuous paths by path spacing in the inner region leading to 50% porosity. *e using the zig-zag algorithm according to Figure 10(b). structure on the right was manufactured with both nozzles To show the potential of the novel developed software, using different path distances generating a structure fea- the surface model of a human lumbar vertebra, extracted turing both different materials and different path spacing from a CT scan and provided as STL file by MarioDiniz on and thus porosity. *e different material grading, porosity *ingiverse [32] as shown in Figure 3, is subdivided in voxels grading, or combination of both within the lattice structures and filled with different materials and patterns. *e software were realized within a STL file that usually solely defines the GUI in Figure 11(a) shows polylines based on the calcula- outer geometry by making use of two extrusion nozzles. tions presented, showing walls, remains from scanning parts of the trabecular bones, and artefacts within the STL file. *ese lines function as guidance for assigning the materials 3.2. Additive Manufacturing of Complexly Structured Lattice and structures from the editable configuration file to the Structures from Surface Models. *e software is designed to part. For manufacturing the part, two extrusion-based assign the information from the editable configuration file to nozzles with a diameter of 0.4mm (Nordson EFD) as in the voxels generated by slicing and gridding of the imported Figure 9 are used. *e example part features an orange surface based file. An easy-to-use graphical user interface pattern with narrow strand spacing of 400µm for the areas of was developed to display the voxels and the boundary the cortical bone (compact bone) and a light-blue pattern curve(s) of the imported body as well as the basic properties with a strand spacing of 800µm, leading to a porosity of of the material deposition as defined in the editable .txt file. about 50% in the areas of the cancellous bone (spongy bone) Figure 10(a) shows the GUI of the software during of the vertebral body. Figure 11(b) shows the printed body assigning materials from the standard file to one layer of the featuring a material grading (different colors) as well as a previously discussed sphere. *e standard file is displayed in porosity grading (strand spacing). a reduced form to provide a large area for assigning the materials and patterns. *e voxel size and overall structure size are not limited 3.3.CombinedAdditiveManufacturingofComplexFiber-Based by the software. However, the voxel size should be di- and Strand-Based Structures for Biomedical Applications. mensioned according to the tools used. In the case of *e applications shown so far relate exclusively to the extrusion-based processes, the strand width and the strand production of extrusion-based structures. *e adaptation of spacing should be considered. An appropriate voxel size for the configuration file is used directly to automatically create the extrusion nozzles used for manufacturing the structures a g-code with corresponding strand spacing. *e software, from Figure 9 may lead to different relative disparities in which also serves as a postprocessor, also allows the use of geometries with different sizes. *e comparison between completely different additive manufacturing processes. Figures 10(a) and 11(a) shows larger relative deviations in *e Net Shape Nonwoven Method (NSN) is a unique the sphere with 9mm in diameter and good conformity for technology for the additive production of short fiber-based the life-size vertebra while using the same voxel size. *e structures for regenerative medicine [40]. Similar to processing time is dependent on the number of surface powder bed printing, this technology is a two-tier process. Journal of Healthcare Engineering 7 Starting point 1 23 4 5 6 78 9 10 n =11 1 23 4 5 6 78 9 10 n =11 1 23 4 5 6 78 9 10 n =11 1 1 1 2 2 2 3 3 3 4 4 4 5 5 5 6 6 6 7 7 7 8 8 8 9 9 9 10 10 10 11 11 11 m =12 m =12 m =12 12 3 1 2 3 4 5 6 7 8 9 10 n =11 1 2 3 4 5 6 7 8 9 10 n =11 1 2 3 4 5 6 7 8 9 10 n =11 1 1 1 2 2 2 3 3 3 4 4 4 5 5 5 6 6 6 7 7 7 8 8 8 9 9 9 10 10 10 11 11 11 m =12 m =12 m =12 45 6 (a) (b) FIGURE 10: (a) “Slice” tab showing assigned structures from the browsed standard file within the midlayer of a 9mm sphere. (b) Zig-zag algorithm to create continuous paths in complex slice geometries according to [39]. (a) (b) Figure 11: (a) GUI of the software with a large area (left) for assigning the voxel parameters from the standard file (right) with polylines representing the section plane and a field for scrolling through the layers (upper right). (b) Printed body featuring different colors as well as different strand spacing to demonstrate the software’s ability to generate different g-codes for different nozzles based on the part geometry and the user’s assignments. First, a thin fiber bed is applied, and subsequently a binder sizes and porosities for regenerative medicine can be is applied to selectively bonding the fibers together. Using generated [41]. For the second process step, a piezo- controlled adhesive nozzle is actuated. In order to the customizable configuration file, the developed modular software also allows the production of NSN structures. By achieve precise contours and geometries, path distances of appropriate selection of the tools for fiber application units about 200µm are used. (actuated by speed-controlled stepper motors) and the path Due to the flexible controlling of extrusion nozzles, fiber distances, either a full surface fiber application or a local application units, and adhesive nozzles, completely different fiber application can be realized. Clearly, the width of the additive manufacturing processes can be combined to create fiber track depends on the fiber length used and usually novel structures. Figure 12 shows different structures on the ranges from 0.5mm to 2mm. With these fiber lengths, basis of simple STL files into which different materials have which can be estimated simulation-based, suitable pore been inscribed. 8 Journal of Healthcare Engineering (a) (b) (c) Figure 12: Unique structures combining extrusion-based additive manufacturing processes with a fiber-based additive manufacturing approach. (a) Sandwich structure, (b) core-shell-shell structure, and (c) core-shell structure, demonstrating uses in different applications. *e newly developed approach allows assigning volu- (iv) Transfer information to other units (e.g., bus sys- metric structures in three-dimensional surface models. STL tems, direct digital controls, robot controls, dis- plays, and user feedback) files obtained from CT or MRI scans in medicine, 3D scans from reverse engineering, CAD software, or any other sources *us, any processing technology or application may be solely contain information of the surface of the bodies. With implemented by customizing the spreadsheet file or existing software, assigning properties within the bodies is not implementing other (machine-readable) codes according to possible as only triangles on the surface area can be selected. the users’ needs. *e hosted editable configuration file allows controlling Furthermore, the voxel-based approach allows assigning different extrusion nozzles, fiber application units, and a any information (e.g., materials, material morphologies, piezo-driven adhesive nozzle (via tool column). It features colors, porosities, and metadata) to the imported files. *e arbitrary nozzle diameters, leading to different strand information may be stored and used for other applications thickness or different fiber layer heights (via filling/mm or further processing. column). All materials are deposited in a z-pattern (via pattern column) starting in the x direction in the first layer of a voxel and subsequently changing its direction into the y 4. Conclusions direction which ensure a crossing of the strands and thus *e presented method allows slicing and gridding bodies structural stability especially for strand deposition. *e from STL or AMF files into volumetric elements (voxels) of standard file also allows setting the speed (in mm/min via the arbitrary size. *e underlying software allows assigning dif- speed column) and adjusting the tool position in the ma- ferent tools and features such as path distances, strand chine (in absolute XYZ coordinates (mm) in the respective thicknesses, or traveling speeds to each of these voxels. Ad- columns). *e “matid” column shows a colored pattern with joining voxels with equal properties are combined to sub- dark and light colors representing the materials as well as the volumes and may either be manufactured into structures with porosity to simplify the material/pattern choice during material grading, porosity grading as well as combinations of assigning the properties from the .txt file to the body. both within the different regions or be stored in AMF format After assigning the parameters to the voxels, a g-code can and be used in compatible software or printing technologies. be generated automatically by clicking the “Create machine Strand-based lattice structures with grading on the ma- path” button on the lower right (see Figure 11(a)). Machine terial and structural level can be designed and manufactured paths of adjacent voxels featuring the same “matid” are within a multinozzle additive manufacturing approach and handed over to g-code as a continuous line and are processed can furthermore be combined with a two-tier process for according to the zig-zag pattern shown in Figure 10(b). fiber-based additive manufacturing well suited for applica- *e possible use of the software goes beyond the usage as tions in regenerative medicine. *e combination of both a postprocessor for extrusion-based or fiber-based additive approaches enables, e.g., designing press-fit applications with manufacturing as it basically allows assigning any control flexible transition areas on the basis of geometry data from commands to every single voxel: complex defects. Furthermore, large defects affecting different (i) Positioning and travel commands for axes (e.g., XY tissue types or tissue morphologies, e.g., osteochondral defects tables and XYZ tables) involving bone and cartilage, may be addressed. (ii) On/off and/or speed commands for motors (e.g., material feed/deposition/compaction/removal Data Availability and applying of substrates) (iii) Setting/resetting of digital or analog outputs *e software used to support the findings of this study is (e.g., connected extruders, nozzles, heaters, coolers, described extensively within the article. *e described fans, and lasers) findings can be used to replicate the findings of the study. Journal of Healthcare Engineering 9 Mechanical Behavior of Biomedical Materials, vol. 57, Furthermore, parts of the software used to support the pp. 190–200, 2016. findings of this study are available from the corresponding [13] B. Caulfield, P. E. McHugh, and S. Lohfeld, “Dependence of author upon request. mechanical properties of polyamide components on build parameters in the SLS process,” Journal of Materials Pro- Conflicts of Interest cessing Technology, vol. 182, no. 1–3, pp. 477–488, 2007. [14] J. D. Williams and C. R. Deckard, “Advances in modeling the *e authors declare that they have no conflicts of interest. effects of selected parameters on the SLS process,” Rapid Prototyping Journal, vol. 4, no. 2, pp. 90–100, 1998. Acknowledgments [15] J. M. Williams, A. Adewunmi, R. M. Schek et al., “Bone tissue engineering using polycaprolactone scaffolds fabricated via We would like to thank “Deutsche Forschungsgemeinschaft” selective laser sintering,” Biomaterials, vol. 26, no. 23, for the financial support of the project DFG CH 174/24-2. pp. 4817–4827, 2005. [16] J. P. Kruth, P. Mercelis, J. Van Vaerenbergh, L. Froyen, and References M. Rombouts, “Binding mechanisms in selective laser sin- tering and selective laser melting,” Rapid Prototyping Journal, [1] S.-H. Park, D.-Y. Yang, and K.-S. Lee, “Two-photon stereo- vol. 11, no. 1, pp. 26–36, 2005. lithography for realizing ultraprecise three-dimensional [17] R. Landers and R. Mulhaupt, ¨ “Desktop manufacturing of nano/microdevices,” Laser & Photonics Review, vol. 3, complex objects, prototypes and biomedical scaffolds by no. 1-2, pp. 1–11, 2009. means of computer-assisted design combined with computer- [2] S. A. Skoog, P. L. Goering, and R. J. Narayan, “Stereo- lithography in tissue engineering,” Journal of Materials Sci- guided 3D plotting of polymers and reactive oligomers,” ence: Materials in Medicine, vol. 25, no. 3, pp. 845–856, 2013. Macromolecular Materials and Engineering, vol. 282, no. 1, [3] A. Cazon, ´ P. Morer, and L. Matey, “Polyjet technology for pp. 17–21, 2000. product prototyping: tensile strength and surface roughness [18] S. A. Park, S. H. Lee, and W. D. Kim, “Fabrication of porous properties,” in Proceedings of the Institution of Mechanical polycaprolactone/hydroxyapatite (PCL/HA) blend scaffolds Engineers, Part B: Journal of Engineering Manufacture, using a 3D plotting system for bone tissue engineering,” vol. 228, no. 12, pp. 1664–1675, 2014. Bioprocess and Biosystems Engineering, vol. 34, no. 4, [4] D. Ibrahim, T. L. Broilo, C. Heitz et al., “Dimensional error of pp. 505–513, 2010. selective laser sintering, three-dimensional printing and [19] A. Lode, F. Krujatz, S. Bruggemeier ¨ et al., “Green bioprinting: polyjet models in the reproduction of mandibular anatomy,” Fabrication of photosynthetic algae-laden hydrogel scaffolds Journal of Cranio-Maxillofacial Surgery, vol. 37, no. 3, for biotechnological and medical applications,” Engineering in pp. 167–173, 2009. Life Sciences, vol. 15, no. 2, pp. 177–183, 2015. [5] D. W. Hutmacher, T. Schantz, I. Zein, K. W. Ng, S. H. Teoh, [20] S. Reitmaier, A. Kovtun, J. Schuelke et al., “Strontium(II) and and K. C. Tan, “Mechanical properties and cell cultural re- mechanical loading additively augment bone formation in sponse of polycaprolactone scaffolds designed and fabricated calcium phosphate scaffolds,” Journal of Orthopaedic Re- via fused deposition modeling,” Journal of biomedical ma- search, vol. 31, p. 37, 2017. terials research, vol. 55, no. 2, pp. 203–216, 2001. [21] F. P. W. Melchels, J. Feijen, and D. W. Grijpma, “A review on [6] S. H. Ahn, M. Montero, D. Odell, S. Roundy, and stereolithography and its applications in biomedical engi- P. K. Wright, “Anisotropic material properties of fused de- neering,” Biomaterials, vol. 31, no. 24, pp. 6121–6130, 2010. position modeling ABS,” Rapid Prototyping Journal, vol. 8, [22] E. O. Olakanmi, R. F. Cochrane, and K. W. Dalgarno, “A no. 4, pp. 248–257, 2002. review on selective laser sintering/melting (SLS/SLM) of al- [7] K. *rimurthulu, P. M. Pandey, and N. Venkata Reddy, uminium alloy powders: processing, microstructure, and “Optimum part deposition orientation in fused deposition properties,” Progress in Materials Science, vol. 74, pp. 401–477, modeling,” International Journal of Machine Tools and Manufacture, vol. 44, no. 6, pp. 585–594, 2004. [23] Y. Luo, A. Lode, F. Sonntag, B. Nies, and M. Gelinsky, “Well- [8] I. Zein, D. W. Hutmacher, K. C. Tan, and S. H. Teoh, “Fused ordered biphasic calcium phosphate-alginate scaffolds fabri- deposition modeling of novel scaffold architectures for tissue cated by multi-channel 3D plotting under mild conditions,” engineering applications,” Biomaterials, vol. 23, no. 4, Journal of Materials Chemistry B, vol. 1, no. 33, p. 4088, 2013. pp. 1169–1185, 2002. [24] H.-W. Kang, S. J. Lee, I. K. Ko, C. Kengla, J. J. Yoo, and [9] R. Anitha, S. Arunachalam, and P. Radhakrishnan, “Critical A. Atala, “A 3D bioprinting system to produce human-scale parameters influencing the quality of prototypes in fused tissue constructs with structural integrity,” Nature Bio- deposition modelling,” Journal of Materials Processing Tech- technology, vol. 34, no. 3, pp. 312–319, 2016. nology, vol. 118, no. 1–3, pp. 385–388, 2001. [25] J. L. Rosenberg, “Drowning in triangle soup,” XRDS: Cross- [10] J. Moon, A. C. Caballero, L. Hozer, Y.-M. Chiang, and roads, =e ACM Magazine for Students, vol. 22, no. 3, M. J. Cima, “Fabrication of functionally graded reaction pp. 58–62, 2016. infiltrated SiC-Si composite by three-dimensional printing [26] C. C. Kai, G. G. K. Jacob, and T. Mei, “Interface between CAD (3DP) process,” Materials Science and Engineering: A, vol. 298, and Rapid Prototyping systems. Part 1: a study of existing no. 1-2, pp. 110–119, 2001. interfaces,” International Journal of Advanced Manufacturing [11] K. Lu and W. T. Reynolds, “3DP process for fine mesh structure printing,” Powder Technology, vol. 187, no. 1, Technology, vol. 13, no. 8, pp. 566–570, 1997. [27] P. K. Venuvinod and W. Ma, Rapid Prototyping, Springer, pp. 11–18, 2008. [12] Q. Wei, Y. Wang, X. Li et al., “Study the bonding mechanism Boston, MA, USA, 2004. [28] P. Mohan Pandey, N. Venkata Reddy, S. G. Dhande, and of binders on hydroxyapatite surface and mechanical prop- erties for 3DP fabrication bone scaffolds,” Journal of the S. G. Dhande, “Slicing procedures in layered manufacturing: a 10 Journal of Healthcare Engineering review,” Rapid Prototyping Journal, vol. 9, no. 5, pp. 274–288, [29] ASTM International, Specification for Additive Manufacturing File Format (AMF) Version 1.2, ASTM International, West Conshohocken, PA, USA, 2016. [30] J. D. Murray and W. Vanryper, Encyclopedia of Graphics File Formats, O’Reilly & Associates, Sebastopol, CA, USA, 2nd edition, 1996. [31] J. Hun, M. Chung, M. Park, S. Woo, X. Zhang, and A. Marie, “Generation of realistic particle structures and simulations of internal stress:a numerical/AFM study of LiMn2O4 parti- cles,” Journal of Electrochemical Society, vol.158, no. 4, p. 434, [32] Mariodiniz: *ingiverse: *ing: 1984177: Lumbar vertebra - L3, L4 and L5. [33] Pearson Education, Inc., Anatomy of the Lumbar Vertebrae, Pearson Education, Inc., London, UK, 2010. [34] S. Chmielewski and P. Tompalski, “Estimating outdoor ad- vertising media visibility with voxel-based approach,” Applied Geography, vol. 87, pp. 1–13, 2017. [35] C. E. Kahn, J. A. Carrino, M. J. Flynn, D. J. Peck, and S. C. Horii, “DICOM and radiology: past, present, and future,” Journal of American College of Radiology, vol. 4, no. 9, pp. 652–657, 2007. [36] Medical Imaging and Technology Alliance, DICOM : Digital Imaging and Communications in Medicine, National Electrical Manufacturers Association, Rosslyn, VA, USA, 2019, http:// www.dicomstandard.org/about/. [37] A. C. Brown and D. de Beer, “Development of a stereo- lithography (STL) slicing and G-code generation algorithm for an entry level 3-D printer,” in Proceedings of AFRICON 2013, pp. 1–5, Pointe-Aux-Piments, Mauritius, September [38] ctrlV: *ingiverse: *ing: 1019228:Test your 3D printer! v2. [39] M. Wojcik, I. Pozniak-Koszalka, L. Koszalka, A. Kasprzak, and V. G. Duffy, “Experimentation system for path planning applied to 3D printing,” in Advances in Intelligent Systems and Computing, vol. 481, Springer, Berlin, Germany, 2017. [40] M. Hild, R. Brunler, ¨ M. Jager ¨ et al., “Net Shape Nonwoven: a novel technique for porous three-dimensional nonwoven hybrid scaffolds,” Textile Research Journal, vol. 84, no. 10, pp. 1084–1094, 2014. [41] R. Brunler, ¨ D. Aibibu, M. Woltje, ¨ A.-M. Anthofer, and C. Cherif, “In silico modeling of structural and porosity properties of additive manufactured implants for regenerative medicine,” Materials Science and Engineering: C, vol. 76, pp. 810–817, 2017. International Journal of Advances in Rotating Machinery Multimedia Journal of The Scientific Journal of Engineering World Journal Sensors Hindawi Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 http://www www.hindawi.com .hindawi.com V Volume 2018 olume 2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Journal of Control Science and Engineering Advances in Civil Engineering Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com Journal of Journal of Electrical and Computer Robotics Engineering Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 VLSI Design Advances in OptoElectronics International Journal of Modelling & Aerospace International Journal of Simulation Navigation and in Engineering Engineering Observation Hindawi Hindawi Hindawi Hindawi Volume 2018 Volume 2018 Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com www.hindawi.com www.hindawi.com Volume 2018 International Journal of Active and Passive International Journal of Antennas and Advances in Chemical Engineering Propagation Electronic Components Shock and Vibration Acoustics and Vibration Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Healthcare Engineering Hindawi Publishing Corporation

Design of Complexly Graded Structures inside Three-Dimensional Surface Models by Assigning Volumetric Structures

Loading next page...
 
/lp/hindawi-publishing-corporation/design-of-complexly-graded-structures-inside-three-dimensional-surface-voxPCp0C5O

References (39)

Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2019 Ronny Brünler et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
2040-2295
eISSN
2040-2309
DOI
10.1155/2019/6074272
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Healthcare Engineering Volume 2019, Article ID 6074272, 10 pages https://doi.org/10.1155/2019/6074272 Research Article Design of Complexly Graded Structures inside Three-Dimensional Surface Models by Assigning Volumetric Structures Ronny Bru ¨nler , Robert Hausmann, Maximilian von Mu ¨nchow, Dilbar Aibibu, and Chokri Cherif Institute of Textile Machinery and High Performance Material Technology (ITM), Technische Universita¨t Dresden, Hohe Str. 6, 01069 Dresden, Germany Correspondence should be addressed to Ronny Bru¨nler; ronny.bruenler@tu-dresden.de Received 11 December 2018; Accepted 8 January 2019; Published 4 February 2019 Guest Editor: Saverio Maietta Copyright ©2019 Ronny Bru¨nler et al. *is is anopen accessarticle distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. An innovative approach for designing complex structures from STL-datasets based on novel software for assigning volumetric data to surface models is reported. *e software allows realizing unique complex structures using additive manufacturing technologies. Geometric data as obtained from imaging methods, computer-aided design, or reverse engineering that exist only in the form of surface data are converted into volumetric elements (voxels). Arbitrary machine data can be assigned to each voxel and thereby enable implementing different materials, material morphologies, colors, porosities, etc. within given geometries. *e software features an easy-to-use graphical user interface and allows simple implementation of machine data libraries. To highlight the potential of the modular designed software, an extrusion-based process as well as a two-tier additive manufacturing approach for short fibers and binder process are combined to generate three-dimensional components with complex grading on the material and structural level from STL files. With these methods, different materials cannot be applied 1. Introduction within a part, in particular not within one layer. Additive manufacturing technologies are based on the In the case of the open-space or layer construction layered construction of material into a finished component. methods, it is possible to use various materials [23, 24]. *e production is carried out by solidification of However, the limitations for parts with complex material and structural composition lie in the properties of the file (i) Liquids or gels (curing, drying, crosslinking, etc.) formats [25]. *e most common file format in generative [1–9] production methods is the STL format (standard tessellation (ii) Powders or granules (gluing, sintering, etc.) [10–16] language/standard triangulation language) [26–28]. In niche or applications, the AMF format (additive manufacturing (iii) Pasty and belt-shaped or strand-shaped materials format) and the OBJ format (object format) are used as well (direct deposition without curing processes or so- [29, 30]. lidification, etc.) [17–20] A sphere is used to illustrate the file formats. *e STL file *e limitations for realizing parts with a complex ma- format provides only surface information and uses triangles, terial composition are either found in the process charac- often referred to as vertices, to represent the surfaces as shown in Figure 1(a). In the AMF or OBJ format, the surface in- teristics or in the data formats used. Technologies using powder beds or liquid baths are limited formation can also be supplemented with properties (mate- to a particular material or a particular material composition rial, textures, or metadata), emphasized by a red color scheme which is constant throughout the entire component [21, 22]. in the triangles on the top of the sphere in Figure 1(b). 2 Journal of Healthcare Engineering (a) (b) Figure 1: (a) Visualization of the surface representation of a sphere by means of triangles in the STL format [31]. (b) *e same sphere in the AMF or OBJ format that allow assigning descriptive metadata to individual triangles, shown here as red color scheme in the upper part of the sphere. STL files as well as AMF or OBJ files represent geometric file can be observed in comparison with the anatomy of a bodies exclusively by means of surface information. Re- vertebra (Figure 3(c)). *e STL file features a hollow body and shows almost no structures in the area of the spongy gardless of whether the files are stored as surface or solid bodies, they contain no volume data. Basically, the objects bone in the center of vertebral body and infinitesimally thin walls instead of the dense cortical bone around the spongy are hollow inside and have “outer walls” that are in- finitesimally thin. *e only difference in solid bodies is the bone structure. *ere is no software yet available to fill representation of a filled body. Figure 2(a) shows a graphic certain regions with different materials or structural varia- representation of the previously discussed sphere cut in half tions within STL, AMF, or OBJ files of parts with complex using the software Blender (Blender Foundation). As in all geometry. other software for editing or displaying STL files or similar formats (Netfabb, Cura, Slic3r, and Repetier, amongst 2. Materials and Methods others) as well as CAD software (FreeCAD, SolidWorks, AutoCAD, and CATIA, amongst others), only the surface of 2.1.SoftwareforAccessingtheInnerStructureofSurface-Based the structure can be addressed as it is simply impossible to Bodies. In addition to surface-based file formats, bodies can select or click on other structures than the surface triangles. also be represented by volumetric elements (voxels). Graphic *is points up the decisive limitation of all surface-based file representations of spheres in voxel format with assigned formats: within surface-approximated geometries, for ex- metadata are shown in Figure 4. *is approach is mostly ample, from computer tomography recordings (CT) and known from video games such as Minecraft (Mojang/ magnetic resonance tomography representations (MRT) or Microsoft Studios, 2009) and Blade Runner (Virgin In- 3D scans, property assignments cannot be implemented. teractive, 1997) or simulations [34] to represent terrain fea- In CAD programs, however, objects that contain volume tures and is also widely used in medical imaging formats such information can be designed and stored in the AMF or OBJ as DICOM [35, 36]. However, the voxel formats are not used format, but again have to be regarded as individual surfaces for designing implants for regenerative medicine, prosthetic approximated by triangles. It is thus possible to realize components, or 3D printing applications in reverse engi- structures with grading on the material or structural level, neering as slicer software is usually developed for STL files. supposing that they are designed from scratch as the two- Hence, while it is possible to 3D print or display complex color sphere in Figure 2(b). geometries with different surfaces, it is not possible to realize However, assigning properties within previously defined grading on the material or structural level inside the or given bodies as obtained from CT scans, MRI scans, or structures. radiographs for the determination of defect geometries in To access the inner structure of surface-based bodies, regenerative medicine or 3D scans from reverse engineering novel software for segmenting the structures into voxels and cannot be carried out in these programs because of the manipulating them is developed. *e software is capable of surface-based representation and the related restrictions. processing STL files in ASCII format and is developed in C# Figure 3(a) shows a CT scan of the lower spine and an STL within the development environment Visual Studio Com- file derived from that scan containing geometry information munity (Microsoft Corp.). Figure 5 shows the graphical user of a lumbar vertebra exhibiting a complex geometry. Sub- interface (GUI) that was created within the Windows stantial differences in the structural composition of the STL Presentation Foundation (WPF) framework. Journal of Healthcare Engineering 3 (a) (b) Figure 2: (a) Visualization of the structure of the sphere in surface-based formats that solely allow selecting and thus altering the surface triangles. (b) Image of a sphere with two colors designed in a CAD software as a solid body. Hollow body Thin walls (a) (b) (c) Figure 3: (a) CTscan of the lower spine showing the L3 (top), L4 (middle), and L5 (bottom) lumbar vertebrae [32]. (b) View of the STL file of an L5 lumbar vertebra from [32] exhibiting infinitesimally thin walls and a hollow body. (c) Anatomy of the lumbar vertebrae [33]. (a) (b) Figure 4: (a) Visualization of a sphere in voxel format. (b) *e same sphere with a red color scheme in the upper part of the sphere. Descriptive metadata can be assigned to individual voxels even throughout the inner parts of the body. After importing a file, it is automatically converted into and grid size can be adjusted arbitrarily and separately. In the AMF format and can be viewed, rotated, and zoomed in default mode, the grid size matches the slicing thickness. and out on the “AMF” tab (Figure 5(a)). *e “Slice” tab as *us, the grid subdivides the body into cubic voxels. visualized in Figure 5(b) is used to slice the body and to Generating the slice data for a graphical representation additionally implement a rectangular grid. Slicing thickness of the contours of the body requires the consideration of 4 Journal of Healthcare Engineering (a) (b) Figure 5: Graphical user interface of the developed software: (a) “AMF” tab with imported sphere; (b) “Slice” tab showing the midlayer of the subdivided sphere, the input field for adjusting slicing thickness, and the scroll bar for scrolling through the single slices. several special cases related to the surface representation by layer in the slice tab. *e software layout also allows pro- means of triangles. During the generation of a single slice, cessing more complex bodies, such as the “test your 3D necessarily some triangles are cut by the section plane as evident printer! v2” file from *ingiverse (MakerBot Industries, in Figure 6(a). In total, there are 10 cases describing the situ- LLC) as illustrated in Figure 7(b) [38]. ational relations between section plane and triangles [37]. While simple cases such as triangles located completely above or below 3. Results and Discussion the section plane are easy to process, five specific cases have to be considered more closely (Figure 6(b)): 3.1. Assignment of Multiple Properties and g-Code to Sub- volumes within Given Complex Geometries. *e imported (1) All three edges are located on the section plane body is divided into voxels by the rectangular grid imple- (2) Exactly two edges are located on the section plane mented in the “Slice” tab. *us, all voxels are accessible by (3) One edge is located above, one below, and one on the scrolling through the single layers. *ese features enable section plane assigning arbitrary properties to every single voxel within any given geometry. (4) One edge is located above/below the section plane *e properties assigned to each voxel are used to gen- and two edges are located on the other side erate machine-readable code. As g-code is the most common (5) Exactly one edge is located on the section plane numerical control programming language, it is used for *ese considered cases have a significant effect on the further processing. Specific commands are filed by means of calculation of the intersection points. *e coordinates of a user-friendly editable .txt file. It can be adjusted depending the edges of all triangles describing the edited body are on the manufacturing technology used. *e standard file for stored in a separate class (triangle class) within the soft- extrusion-based additive manufacturing processes is shown ware. *ey are processed during slice data generation and in Figure 8(a). It contains the material id (matid), the serve for intersection point calculation. After using a number of layers necessary to fill one millimeter (fillings/ certain coordinate for calculation, it may be erased from the mm), the path distance (hatch), the path arrangement to-be-processed data. However, depending on the case and (pattern), the digital output command being used to control on the ratio between slicing thickness and triangle size, the extrusion nozzle (tool), the speed of the tool (speed), and certain coordinates have to be used for the calculation of the zero position of the tool used (X-, Y-, and Z-co- the next layer and must not be erased. To ensure a correct ordinates). *e strand thickness can be calculated from the calculation, novel triangles are generated according to “fillings/mm” column as these values are used for the travel Figure 7(a). *e triangle (defined by a, b, and c in of the z-position to lay the strands directly on top of each Figure 7(a)) he is divided into three smaller triangles using other. *us, according to Figure 8(a), 8 fillings/mm are used the intersection points (P1 and P2) with the current section for a strand thickness of 125µm and 4 fillings/mm are used plane. In the case described here, the coordinates a and b for 250µm strands, respectively. can be erased from the triangle class, and only the triangle *e absolute positions of the voxels and the assigned defined by P1, P2 and C is used for calculating the in- fillings, hatches, tools, patterns, speeds, and positions are tersection points in the next layer. used to generate g-code commands. *e black paths in After calculation of all intersection points, a polygon Figure 8(b) directly show the course of the generated path in course is generated automatically and displayed for each g-code. Travels in z direction are handed over from the Journal of Healthcare Engineering 5 Case I Case II Case III Case IV Case V (a) (b) Figure 6: (a) Slicing of a sphere with resulting cuts of the surface triangles. (b) Specific cases for situational relations between the section plane and surface triangles during slicing [37]. i +2 P2 P1 i +3+ k i +2+ k A B Index: i i +1 (a) (b) Figure 7: (a) Method for calculating the polygon course for a graphic representation of single slices. (b) Polygon course of the base layer of a complex 3D structure [38]. Matid Filling (mm) Hatch Pattern Tool Speed X Y Z 1 8 125 zz 1 200 50 50 0 2 8 250 zz 1 200 50 50 0 3 4 250 zz 2 500 150 50 0 4 4 500 zz 2 500 150 50 0 (a) (b) Figure 8: (a) Standard file for extrusion-based printing with two different colored materials and different deposition patterns. *e voxel size in this figure is 1mm and is filled in z direction with 8 red material strands (tool 1; red material; strand thickness 125µm; traversing speed 200mm/min; zero position of the nozzle with respect to tool installation in our machine) and 4 blue strands (tool 2; blue material; strand thickness 250µm; traversing speed 500mm/min; zero position of tool #2), leading to (b) different voxel filling patterns with adjusted path distances (hatch) for the same material used. 6 Journal of Healthcare Engineering information provided by the “fillings/mm” column and the slice thickness. *e file shown in Figure 8(a) leads to different material deposition. *e patterns with “matid 1” and “matid 2” are extruded from the same nozzle (tool 1) and with the same speed but in different path distances. For “matid 1,” the strand thickness matches the path distance and thus fills the voxel area in top view creating a dense structure (Figure 8(b), top left). “Matid 2” deposits the 125µm strands in a path distance of 250µm and thus exhibit filling levels of 50%. In “matid 3” and “matid 4,” another tool is used to deposit larger strands in different patterns. Figure 9: Manufacturing of material grading (left), porosity Figure 9 shows three different structures manufactured grading (center), and material and porosity grading (right) by from the same cuboid STL file (20mm ×20mm ×4mm). means of assigning different nozzles (tool column) and strand Two extruding systems with nozzles 0.4mm in diameter spacings (hatch column). (Nordson EFD) filled with different colored clay were used to manufacture different structures within the STL file. triangles of the processed structure and the hardware used. *e left structure was manufactured with both nozzles *e process of loading and processing the vertebra structure using the same path distance leading to a laydown of one that is defined by 34464 surface triangles into voxels took material (light-blue colored clay) in the inner zone and the 71.6seconds on a dual-core 2.4GHz, 4GB RAM, 256MB other material (orange colored clay) in the outer part while graphics memory system and 14.5seconds on an eight-core both regions featuring a dense path spacing. *e structure in 3.4GHz, 32GB RAM, 1GB graphics memory system, making the middle was manufactured using solely one nozzle fol- the software applicable on a wide range of computing systems. lowing dense path spacing in the outer part and a 0.8mm Contiguous areas are processed with continuous paths by path spacing in the inner region leading to 50% porosity. *e using the zig-zag algorithm according to Figure 10(b). structure on the right was manufactured with both nozzles To show the potential of the novel developed software, using different path distances generating a structure fea- the surface model of a human lumbar vertebra, extracted turing both different materials and different path spacing from a CT scan and provided as STL file by MarioDiniz on and thus porosity. *e different material grading, porosity *ingiverse [32] as shown in Figure 3, is subdivided in voxels grading, or combination of both within the lattice structures and filled with different materials and patterns. *e software were realized within a STL file that usually solely defines the GUI in Figure 11(a) shows polylines based on the calcula- outer geometry by making use of two extrusion nozzles. tions presented, showing walls, remains from scanning parts of the trabecular bones, and artefacts within the STL file. *ese lines function as guidance for assigning the materials 3.2. Additive Manufacturing of Complexly Structured Lattice and structures from the editable configuration file to the Structures from Surface Models. *e software is designed to part. For manufacturing the part, two extrusion-based assign the information from the editable configuration file to nozzles with a diameter of 0.4mm (Nordson EFD) as in the voxels generated by slicing and gridding of the imported Figure 9 are used. *e example part features an orange surface based file. An easy-to-use graphical user interface pattern with narrow strand spacing of 400µm for the areas of was developed to display the voxels and the boundary the cortical bone (compact bone) and a light-blue pattern curve(s) of the imported body as well as the basic properties with a strand spacing of 800µm, leading to a porosity of of the material deposition as defined in the editable .txt file. about 50% in the areas of the cancellous bone (spongy bone) Figure 10(a) shows the GUI of the software during of the vertebral body. Figure 11(b) shows the printed body assigning materials from the standard file to one layer of the featuring a material grading (different colors) as well as a previously discussed sphere. *e standard file is displayed in porosity grading (strand spacing). a reduced form to provide a large area for assigning the materials and patterns. *e voxel size and overall structure size are not limited 3.3.CombinedAdditiveManufacturingofComplexFiber-Based by the software. However, the voxel size should be di- and Strand-Based Structures for Biomedical Applications. mensioned according to the tools used. In the case of *e applications shown so far relate exclusively to the extrusion-based processes, the strand width and the strand production of extrusion-based structures. *e adaptation of spacing should be considered. An appropriate voxel size for the configuration file is used directly to automatically create the extrusion nozzles used for manufacturing the structures a g-code with corresponding strand spacing. *e software, from Figure 9 may lead to different relative disparities in which also serves as a postprocessor, also allows the use of geometries with different sizes. *e comparison between completely different additive manufacturing processes. Figures 10(a) and 11(a) shows larger relative deviations in *e Net Shape Nonwoven Method (NSN) is a unique the sphere with 9mm in diameter and good conformity for technology for the additive production of short fiber-based the life-size vertebra while using the same voxel size. *e structures for regenerative medicine [40]. Similar to processing time is dependent on the number of surface powder bed printing, this technology is a two-tier process. Journal of Healthcare Engineering 7 Starting point 1 23 4 5 6 78 9 10 n =11 1 23 4 5 6 78 9 10 n =11 1 23 4 5 6 78 9 10 n =11 1 1 1 2 2 2 3 3 3 4 4 4 5 5 5 6 6 6 7 7 7 8 8 8 9 9 9 10 10 10 11 11 11 m =12 m =12 m =12 12 3 1 2 3 4 5 6 7 8 9 10 n =11 1 2 3 4 5 6 7 8 9 10 n =11 1 2 3 4 5 6 7 8 9 10 n =11 1 1 1 2 2 2 3 3 3 4 4 4 5 5 5 6 6 6 7 7 7 8 8 8 9 9 9 10 10 10 11 11 11 m =12 m =12 m =12 45 6 (a) (b) FIGURE 10: (a) “Slice” tab showing assigned structures from the browsed standard file within the midlayer of a 9mm sphere. (b) Zig-zag algorithm to create continuous paths in complex slice geometries according to [39]. (a) (b) Figure 11: (a) GUI of the software with a large area (left) for assigning the voxel parameters from the standard file (right) with polylines representing the section plane and a field for scrolling through the layers (upper right). (b) Printed body featuring different colors as well as different strand spacing to demonstrate the software’s ability to generate different g-codes for different nozzles based on the part geometry and the user’s assignments. First, a thin fiber bed is applied, and subsequently a binder sizes and porosities for regenerative medicine can be is applied to selectively bonding the fibers together. Using generated [41]. For the second process step, a piezo- controlled adhesive nozzle is actuated. In order to the customizable configuration file, the developed modular software also allows the production of NSN structures. By achieve precise contours and geometries, path distances of appropriate selection of the tools for fiber application units about 200µm are used. (actuated by speed-controlled stepper motors) and the path Due to the flexible controlling of extrusion nozzles, fiber distances, either a full surface fiber application or a local application units, and adhesive nozzles, completely different fiber application can be realized. Clearly, the width of the additive manufacturing processes can be combined to create fiber track depends on the fiber length used and usually novel structures. Figure 12 shows different structures on the ranges from 0.5mm to 2mm. With these fiber lengths, basis of simple STL files into which different materials have which can be estimated simulation-based, suitable pore been inscribed. 8 Journal of Healthcare Engineering (a) (b) (c) Figure 12: Unique structures combining extrusion-based additive manufacturing processes with a fiber-based additive manufacturing approach. (a) Sandwich structure, (b) core-shell-shell structure, and (c) core-shell structure, demonstrating uses in different applications. *e newly developed approach allows assigning volu- (iv) Transfer information to other units (e.g., bus sys- metric structures in three-dimensional surface models. STL tems, direct digital controls, robot controls, dis- plays, and user feedback) files obtained from CT or MRI scans in medicine, 3D scans from reverse engineering, CAD software, or any other sources *us, any processing technology or application may be solely contain information of the surface of the bodies. With implemented by customizing the spreadsheet file or existing software, assigning properties within the bodies is not implementing other (machine-readable) codes according to possible as only triangles on the surface area can be selected. the users’ needs. *e hosted editable configuration file allows controlling Furthermore, the voxel-based approach allows assigning different extrusion nozzles, fiber application units, and a any information (e.g., materials, material morphologies, piezo-driven adhesive nozzle (via tool column). It features colors, porosities, and metadata) to the imported files. *e arbitrary nozzle diameters, leading to different strand information may be stored and used for other applications thickness or different fiber layer heights (via filling/mm or further processing. column). All materials are deposited in a z-pattern (via pattern column) starting in the x direction in the first layer of a voxel and subsequently changing its direction into the y 4. Conclusions direction which ensure a crossing of the strands and thus *e presented method allows slicing and gridding bodies structural stability especially for strand deposition. *e from STL or AMF files into volumetric elements (voxels) of standard file also allows setting the speed (in mm/min via the arbitrary size. *e underlying software allows assigning dif- speed column) and adjusting the tool position in the ma- ferent tools and features such as path distances, strand chine (in absolute XYZ coordinates (mm) in the respective thicknesses, or traveling speeds to each of these voxels. Ad- columns). *e “matid” column shows a colored pattern with joining voxels with equal properties are combined to sub- dark and light colors representing the materials as well as the volumes and may either be manufactured into structures with porosity to simplify the material/pattern choice during material grading, porosity grading as well as combinations of assigning the properties from the .txt file to the body. both within the different regions or be stored in AMF format After assigning the parameters to the voxels, a g-code can and be used in compatible software or printing technologies. be generated automatically by clicking the “Create machine Strand-based lattice structures with grading on the ma- path” button on the lower right (see Figure 11(a)). Machine terial and structural level can be designed and manufactured paths of adjacent voxels featuring the same “matid” are within a multinozzle additive manufacturing approach and handed over to g-code as a continuous line and are processed can furthermore be combined with a two-tier process for according to the zig-zag pattern shown in Figure 10(b). fiber-based additive manufacturing well suited for applica- *e possible use of the software goes beyond the usage as tions in regenerative medicine. *e combination of both a postprocessor for extrusion-based or fiber-based additive approaches enables, e.g., designing press-fit applications with manufacturing as it basically allows assigning any control flexible transition areas on the basis of geometry data from commands to every single voxel: complex defects. Furthermore, large defects affecting different (i) Positioning and travel commands for axes (e.g., XY tissue types or tissue morphologies, e.g., osteochondral defects tables and XYZ tables) involving bone and cartilage, may be addressed. (ii) On/off and/or speed commands for motors (e.g., material feed/deposition/compaction/removal Data Availability and applying of substrates) (iii) Setting/resetting of digital or analog outputs *e software used to support the findings of this study is (e.g., connected extruders, nozzles, heaters, coolers, described extensively within the article. *e described fans, and lasers) findings can be used to replicate the findings of the study. Journal of Healthcare Engineering 9 Mechanical Behavior of Biomedical Materials, vol. 57, Furthermore, parts of the software used to support the pp. 190–200, 2016. findings of this study are available from the corresponding [13] B. Caulfield, P. E. McHugh, and S. Lohfeld, “Dependence of author upon request. mechanical properties of polyamide components on build parameters in the SLS process,” Journal of Materials Pro- Conflicts of Interest cessing Technology, vol. 182, no. 1–3, pp. 477–488, 2007. [14] J. D. Williams and C. R. Deckard, “Advances in modeling the *e authors declare that they have no conflicts of interest. effects of selected parameters on the SLS process,” Rapid Prototyping Journal, vol. 4, no. 2, pp. 90–100, 1998. Acknowledgments [15] J. M. Williams, A. Adewunmi, R. M. Schek et al., “Bone tissue engineering using polycaprolactone scaffolds fabricated via We would like to thank “Deutsche Forschungsgemeinschaft” selective laser sintering,” Biomaterials, vol. 26, no. 23, for the financial support of the project DFG CH 174/24-2. pp. 4817–4827, 2005. [16] J. P. Kruth, P. Mercelis, J. Van Vaerenbergh, L. Froyen, and References M. Rombouts, “Binding mechanisms in selective laser sin- tering and selective laser melting,” Rapid Prototyping Journal, [1] S.-H. Park, D.-Y. Yang, and K.-S. Lee, “Two-photon stereo- vol. 11, no. 1, pp. 26–36, 2005. lithography for realizing ultraprecise three-dimensional [17] R. Landers and R. Mulhaupt, ¨ “Desktop manufacturing of nano/microdevices,” Laser & Photonics Review, vol. 3, complex objects, prototypes and biomedical scaffolds by no. 1-2, pp. 1–11, 2009. means of computer-assisted design combined with computer- [2] S. A. Skoog, P. L. Goering, and R. J. Narayan, “Stereo- lithography in tissue engineering,” Journal of Materials Sci- guided 3D plotting of polymers and reactive oligomers,” ence: Materials in Medicine, vol. 25, no. 3, pp. 845–856, 2013. Macromolecular Materials and Engineering, vol. 282, no. 1, [3] A. Cazon, ´ P. Morer, and L. Matey, “Polyjet technology for pp. 17–21, 2000. product prototyping: tensile strength and surface roughness [18] S. A. Park, S. H. Lee, and W. D. Kim, “Fabrication of porous properties,” in Proceedings of the Institution of Mechanical polycaprolactone/hydroxyapatite (PCL/HA) blend scaffolds Engineers, Part B: Journal of Engineering Manufacture, using a 3D plotting system for bone tissue engineering,” vol. 228, no. 12, pp. 1664–1675, 2014. Bioprocess and Biosystems Engineering, vol. 34, no. 4, [4] D. Ibrahim, T. L. Broilo, C. Heitz et al., “Dimensional error of pp. 505–513, 2010. selective laser sintering, three-dimensional printing and [19] A. Lode, F. Krujatz, S. Bruggemeier ¨ et al., “Green bioprinting: polyjet models in the reproduction of mandibular anatomy,” Fabrication of photosynthetic algae-laden hydrogel scaffolds Journal of Cranio-Maxillofacial Surgery, vol. 37, no. 3, for biotechnological and medical applications,” Engineering in pp. 167–173, 2009. Life Sciences, vol. 15, no. 2, pp. 177–183, 2015. [5] D. W. Hutmacher, T. Schantz, I. Zein, K. W. Ng, S. H. Teoh, [20] S. Reitmaier, A. Kovtun, J. Schuelke et al., “Strontium(II) and and K. C. Tan, “Mechanical properties and cell cultural re- mechanical loading additively augment bone formation in sponse of polycaprolactone scaffolds designed and fabricated calcium phosphate scaffolds,” Journal of Orthopaedic Re- via fused deposition modeling,” Journal of biomedical ma- search, vol. 31, p. 37, 2017. terials research, vol. 55, no. 2, pp. 203–216, 2001. [21] F. P. W. Melchels, J. Feijen, and D. W. Grijpma, “A review on [6] S. H. Ahn, M. Montero, D. Odell, S. Roundy, and stereolithography and its applications in biomedical engi- P. K. Wright, “Anisotropic material properties of fused de- neering,” Biomaterials, vol. 31, no. 24, pp. 6121–6130, 2010. position modeling ABS,” Rapid Prototyping Journal, vol. 8, [22] E. O. Olakanmi, R. F. Cochrane, and K. W. Dalgarno, “A no. 4, pp. 248–257, 2002. review on selective laser sintering/melting (SLS/SLM) of al- [7] K. *rimurthulu, P. M. Pandey, and N. Venkata Reddy, uminium alloy powders: processing, microstructure, and “Optimum part deposition orientation in fused deposition properties,” Progress in Materials Science, vol. 74, pp. 401–477, modeling,” International Journal of Machine Tools and Manufacture, vol. 44, no. 6, pp. 585–594, 2004. [23] Y. Luo, A. Lode, F. Sonntag, B. Nies, and M. Gelinsky, “Well- [8] I. Zein, D. W. Hutmacher, K. C. Tan, and S. H. Teoh, “Fused ordered biphasic calcium phosphate-alginate scaffolds fabri- deposition modeling of novel scaffold architectures for tissue cated by multi-channel 3D plotting under mild conditions,” engineering applications,” Biomaterials, vol. 23, no. 4, Journal of Materials Chemistry B, vol. 1, no. 33, p. 4088, 2013. pp. 1169–1185, 2002. [24] H.-W. Kang, S. J. Lee, I. K. Ko, C. Kengla, J. J. Yoo, and [9] R. Anitha, S. Arunachalam, and P. Radhakrishnan, “Critical A. Atala, “A 3D bioprinting system to produce human-scale parameters influencing the quality of prototypes in fused tissue constructs with structural integrity,” Nature Bio- deposition modelling,” Journal of Materials Processing Tech- technology, vol. 34, no. 3, pp. 312–319, 2016. nology, vol. 118, no. 1–3, pp. 385–388, 2001. [25] J. L. Rosenberg, “Drowning in triangle soup,” XRDS: Cross- [10] J. Moon, A. C. Caballero, L. Hozer, Y.-M. Chiang, and roads, =e ACM Magazine for Students, vol. 22, no. 3, M. J. Cima, “Fabrication of functionally graded reaction pp. 58–62, 2016. infiltrated SiC-Si composite by three-dimensional printing [26] C. C. Kai, G. G. K. Jacob, and T. Mei, “Interface between CAD (3DP) process,” Materials Science and Engineering: A, vol. 298, and Rapid Prototyping systems. Part 1: a study of existing no. 1-2, pp. 110–119, 2001. interfaces,” International Journal of Advanced Manufacturing [11] K. Lu and W. T. Reynolds, “3DP process for fine mesh structure printing,” Powder Technology, vol. 187, no. 1, Technology, vol. 13, no. 8, pp. 566–570, 1997. [27] P. K. Venuvinod and W. Ma, Rapid Prototyping, Springer, pp. 11–18, 2008. [12] Q. Wei, Y. Wang, X. Li et al., “Study the bonding mechanism Boston, MA, USA, 2004. [28] P. Mohan Pandey, N. Venkata Reddy, S. G. Dhande, and of binders on hydroxyapatite surface and mechanical prop- erties for 3DP fabrication bone scaffolds,” Journal of the S. G. Dhande, “Slicing procedures in layered manufacturing: a 10 Journal of Healthcare Engineering review,” Rapid Prototyping Journal, vol. 9, no. 5, pp. 274–288, [29] ASTM International, Specification for Additive Manufacturing File Format (AMF) Version 1.2, ASTM International, West Conshohocken, PA, USA, 2016. [30] J. D. Murray and W. Vanryper, Encyclopedia of Graphics File Formats, O’Reilly & Associates, Sebastopol, CA, USA, 2nd edition, 1996. [31] J. Hun, M. Chung, M. Park, S. Woo, X. Zhang, and A. Marie, “Generation of realistic particle structures and simulations of internal stress:a numerical/AFM study of LiMn2O4 parti- cles,” Journal of Electrochemical Society, vol.158, no. 4, p. 434, [32] Mariodiniz: *ingiverse: *ing: 1984177: Lumbar vertebra - L3, L4 and L5. [33] Pearson Education, Inc., Anatomy of the Lumbar Vertebrae, Pearson Education, Inc., London, UK, 2010. [34] S. Chmielewski and P. Tompalski, “Estimating outdoor ad- vertising media visibility with voxel-based approach,” Applied Geography, vol. 87, pp. 1–13, 2017. [35] C. E. Kahn, J. A. Carrino, M. J. Flynn, D. J. Peck, and S. C. Horii, “DICOM and radiology: past, present, and future,” Journal of American College of Radiology, vol. 4, no. 9, pp. 652–657, 2007. [36] Medical Imaging and Technology Alliance, DICOM : Digital Imaging and Communications in Medicine, National Electrical Manufacturers Association, Rosslyn, VA, USA, 2019, http:// www.dicomstandard.org/about/. [37] A. C. Brown and D. de Beer, “Development of a stereo- lithography (STL) slicing and G-code generation algorithm for an entry level 3-D printer,” in Proceedings of AFRICON 2013, pp. 1–5, Pointe-Aux-Piments, Mauritius, September [38] ctrlV: *ingiverse: *ing: 1019228:Test your 3D printer! v2. [39] M. Wojcik, I. Pozniak-Koszalka, L. Koszalka, A. Kasprzak, and V. G. Duffy, “Experimentation system for path planning applied to 3D printing,” in Advances in Intelligent Systems and Computing, vol. 481, Springer, Berlin, Germany, 2017. [40] M. Hild, R. Brunler, ¨ M. Jager ¨ et al., “Net Shape Nonwoven: a novel technique for porous three-dimensional nonwoven hybrid scaffolds,” Textile Research Journal, vol. 84, no. 10, pp. 1084–1094, 2014. [41] R. Brunler, ¨ D. Aibibu, M. Woltje, ¨ A.-M. Anthofer, and C. Cherif, “In silico modeling of structural and porosity properties of additive manufactured implants for regenerative medicine,” Materials Science and Engineering: C, vol. 76, pp. 810–817, 2017. International Journal of Advances in Rotating Machinery Multimedia Journal of The Scientific Journal of Engineering World Journal Sensors Hindawi Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 http://www www.hindawi.com .hindawi.com V Volume 2018 olume 2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Journal of Control Science and Engineering Advances in Civil Engineering Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com Journal of Journal of Electrical and Computer Robotics Engineering Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 VLSI Design Advances in OptoElectronics International Journal of Modelling & Aerospace International Journal of Simulation Navigation and in Engineering Engineering Observation Hindawi Hindawi Hindawi Hindawi Volume 2018 Volume 2018 Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com www.hindawi.com www.hindawi.com Volume 2018 International Journal of Active and Passive International Journal of Antennas and Advances in Chemical Engineering Propagation Electronic Components Shock and Vibration Acoustics and Vibration Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal

Journal of Healthcare EngineeringHindawi Publishing Corporation

Published: Feb 4, 2019

There are no references for this article.