Micro-robotic deposition guidelines by a design of experiments approach to maximize fabrication reliability for the bone scaffold application
Introduction
Micro-robotic deposition (μRD) is a solid freeform fabrication process in which a colloidal suspension, or ink, is extruded through a micron-sized nozzle in a defined trajectory to form a three-dimensional structure (Fig. 1). The term micro-robotic deposition is chosen since the extrusion nozzle is usually positioned by a robotic device with resulting part feature sizes between 1 and 1000 μm. The extruded ink forms semisolid rods that can span distances of up to 2 mm [1], which permits the fabrication of porous structures without lost molds. μRD technology has been applied to composite [2], [3], microfluidic [4], photonic bandgap [5], and tissue engineering [6], [7], [8] structures and is referred to in the literature as robocasting [6], [8], [9], robotic deposition [1], [4], direct-write assembly [4], [7], or slurry micro-extrusion [10]. To date, the majority of the μRD research has focused on developing new materials appropriate for deposition [1], [7], [11], [12], [13] as well as decreasing the feature size [5]. In order to make μRD a reliable and viable manufacturing process the number and severity of fabrication defects must be reduced. However, deposition reliability has received relatively little research attention. The work presented here fills this gap in the literature and will help facilitate the transition of μRD technology from the research bench to a manufacturing environment by developing general guidelines that maximize process reliability.
Although μRD is capable of fabricating structures with submicron resolution [5], many applications require structures that are macro-sized in total dimension with micron-sized features. Of primary interest here are artificial bone scaffolds that are large enough to fill anatomical defects, yet have a controlled porous microstructure appropriate for bone cell proliferation [14]. While the material and structure for this study are chosen specifically for use as artificial bone scaffolds, the procedures and results are relevant to the μRD of colloids in general, including micro-sized structures [9], piezoelectric actuators [2], and dental implants [10]. Commonly, the quality of the finished part is compromised by a variety of defects derived from either a momentary loss in material flow, a clogged nozzle, or a material deformation. Each of these types of defects reduces part quality and must be minimized in order to make μRD technology a reliable manufacturing process. To that end, a design of experiments (DoE) is devised that analyzes which combination of manufacturing treatment levels yields the highest μRD process quality.
There are a number of treatment options that may potentially affect the quality of finished μRD parts. For this study, only treatments that directly affect the rate of manufacture and those that are appropriate for micro-sized features are considered. While deterministic of structure quality, colloidal ink fabrication treatments such as colloidal solids loading, dispersant concentration, and pH, have been researched previously [1], [11], [13] and therefore are kept constant here in order to isolate the effects of the selected manufacturing treatments. A 2 × 2 × 3 full factorial DoE examines the effects of two calcination time (CT), two nozzle size (NS), and three deposition speed (DS) treatment levels on structure quality. The test structure has a lattice architecture, consisting of alternating layers of orthogonal rods, which is a common architecture found in bone scaffolding [15], [16] and has been used as a benchmark in previous μRD research [1], [12].
The specific treatment levels chosen satisfy at least one of two criteria: (1) they must have been proven to work well, either by experience [17] or the published literature [1], [7], to specifically focus on optimizing the μRD process or (2) they must have the potential to decrease fabrication cost. The following arguments justify the chosen treatment levels. Ceramic powder calcination is a time-consuming and expensive manufacturing step that smoothes the particle morphology. However, a smoother particle morphology improves particle consolidation and green body density [18] and the added costs of extending CT may significantly improve the finished product enough to justify the cost. Particle morphology at four CT treatment levels (1/2, 2, 10, and 20 h) is analyzed and of these four, two (1/2 and 10 h) are selected for the DoE to study their effect on deposition defects. The two NS treatment levels (250 and 410 μm internal diameter) chosen reflect the upper and lower limits of feature sizes typically found in bone scaffold structures [7], [15]. The DS affects the part fabrication rate and therefore manufacturing cost. The three DS treatment levels include speeds that have been commonly used in μRD bone scaffold fabrication (5 and 10 mm s−1) [19] as well as one higher speed (15 mm s−1) that, if it could be used successfully, would reduce fabrication time and cost.
To analyze the effects of the treatment variables, it is important to have quantitative metrics for evaluation. Here we define and quantify structural quality using the five weighted cost functions, or dependent variables, that are described in detail in Section 2.3. A multivariate analysis of variance (ANOVA) and three complementary statistical tests are used to evaluate which treatments significantly affect the dependent variables. The DoE presented provides statistical correlations between μRD process inputs (i.e. treatment variables) and outputs (i.e. dependent variables). From these correlations, inferences on the mechanisms governing deposition are made.
The following outlines the content of the paper. Section 2 provides details on the materials and instruments used, the DoE setup, deposition procedure, defect quantification method, and statistical analysis. Section 3 presents material processing, lattice deposition, and statistical results. Section 4 discusses the experimental results, analyzing the correlations between manufacturing treatments and part quality and postulating possible mechanisms that govern the process. Section 5 includes a short experiment summary and concludes with resulting insight gained.
Section snippets
Powder processing and characterization
Hydroxyapatite (HA) powder (Riedel-de Haen lot 50270) was used as the solid phase material in the colloidal ink. To smoothen the surface morphology, HA powder was calcined in batches (Electric furnace, Paragon Industries TNFQ11A) at 1100 °C for 1/2, 2, 10, or 20 h and furnace cooled to room temperature. Calcined powder was subsequently ball milled in ethanol with cylindrical alumina media (9 mm × 14 mm) for 13 h. The morphology of the as-received, calcined, and calcined and ball milled powders were
Powder and ink characterization
As-received HA powder has a rough surface morphology (Fig. 2a) and consequently a relatively high SSA of 67.49 m2 g−1 (Fig. 3a). The SEM images (Fig. 2) show the evolution of the surface morphology over the range of CTs tested. Surface morphology improves markedly between the as-received and 1/2 h CT powders (Fig. 2b). The particle surface continues to smooth as CT increases, until a 10 h CT, after which there is little noticeable improvement. Measured morphological data are shown in Fig. 3. Here
Discussion
The primary intention of this research is to develop general deposition guidelines that will aid the transition of μRD technology from the lab bench to a mass manufacturing environment. The general guidelines are as follows: to achieve maximum deposition reliability within the defined ranges, powder CT should be extended to sufficiently smoothen particle morphology, the largest NS allowable by the application should be selected, and DS should be sufficiently high. Although these guidelines are
Summary and conclusions
Research efforts have expanded the number of materials appropriate for μRD and have decreased feature sizes to submicron scales [36]. However, there has been less focus on assessing the process reliability for larger structures fabricated by μRD despite the many important applications. This research utilized a DoE approach to determine which manufacturing treatments maximize μRD process reliability for the fabrication of HA artificial bone scaffolds. Although there are many manufacturing
Acknowledgements
The authors would like to acknowledge Dr. Adam Martinsek for statistics consultation, Ranjeet Rao for his help with rheometry, Danchin Chen for his work with XRD, and Amanda Hilldore for her work with Micro-CT. Additionally, we would like to acknowledge the Beckman Institute, Center for Cement Composite Materials, Colloidal Assembly Group, and the Frederick Seitz Materials Research Laboratory Center for Microanalysis of Materials for the use of their characterization facilities. This work was
References (36)
- et al.
Concentrated hydroxyapatite inks for direct-write assembly of 3-D periodic scaffolds
Biomaterials
(2005) - et al.
Sintering and robocasting of β-tricalcium phosphate scaffolds for orthopaedic applications
Acta Biomater
(2006) - et al.
Control of the viscous behavior of highly concentrated mullite suspensions for robocasting
J Mater Process Technol
(2003) - et al.
Rheological and extrusion behavior of dental porcelain slurries for rapid prototyping applications
Mater Sci Eng A
(2005) - et al.
3D fiber-deposited scaffolds for tissue engineering: influence of pores geometry and architecture on dynamic mechanical properties
Biomaterials
(2006) Rheology of monodisperse lattices
Adv Colloid Interf Sci
(1972)- et al.
Colloidal inks for directed assembly of 3-D periodic structures
Langmuir
(2002) - et al.
Piezoelectric properties of 3-X periodic Pb(ZrxTi1 − x)O3-polymer composites
J Appl Phys
(2002) - et al.
Robocasting and cofiring of functionally graded Si3N4–W materials
Ceram Eng Sci Proc
(2001) - et al.
Chaotic mixing in three-dimensional microvascular networks fabricated by direct-write assembly
Nat Mater
(2003)
Direct writing of three-dimensional webs
Nature
Customization of load-bearing hydroxyapatite lattice scaffolds
Int J Appl Ceramic Technol
Robocasting provides moldless fabrication from slurry deposition
Ceram Ind
Solid freeform fabrication of permanent dental restorations via slurry micro-extrusion
J Am Ceram Soc
Nanoparticle inks for directed assembly of three-dimensional periodic structures
Adv Mater
Bioceramic bone graft substitutes: influence of porosity and chemistry
Int J Appl Ceram Technol
Hydroxyapatite implants with designed internal architecture
J Mater Sci: Mater Med
Reliability guidelines and flowrate modulation for a micro robotic deposition system
Cited by (33)
Medical applications of polymer/functionalized nanoparticle composite systems, renewable polymers, and polymer-metal oxide composites
2022, Renewable Polymers and Polymer-Metal Oxide Composites: Synthesis, Properties, and ApplicationsProcesses and materials used for direct writing technologies: A review
2021, Results in EngineeringFactors governing the dimensional accuracy and fracture modes under compression of regular and shifted orthogonal scaffolds
2020, Journal of the European Ceramic SocietyMineralization in micropores of calcium phosphate scaffolds
2019, Acta BiomaterialiaCitation Excerpt :Those traditional techniques offer little control over the pore size or shape, either at the macro- or at the microscale. The recent development of fabrication techniques using additive manufacturing, or 3D printing, and the use of sacrificial porogens allow better control of the size, fraction and interconnectivity of both macropores (Fig. 1b) and micropores (Fig. 1d) [36,37,96]. More control of macro- and micropore characteristics allows for better design and exploitation of scaffold porosity at the macro- and microscale.
Accelerated hardening of nanotextured 3D-plotted self-setting calcium phosphate inks
2018, Acta BiomaterialiaMicrostructural control of modular peptide release from microporous biphasic calcium phosphate
2017, Materials Science and Engineering CCitation Excerpt :To make the “ink” for deposition, HA powder is combined with additives in order to tune the rheological properties of the ink for printing. The mass of HA powder in a batch determines the specific amount of additives, which include deionized (DI) water, an anti-foaming reagent, gelling reagents, and surfactants [16,17]. The viscosity of the ink is adjusted by adding the base NH4OH until the ink is viscous enough to hold its shape when extruded.