Additive manufacturing and mechanical characterization of graded porosity scaffolds designed based on triply periodic minimal surface architectures

Highlights

• TPMS structures were used to design porosity graded scaffolds.

• Two graded scaffold structures were 3D printed and subjected to compression.

• Stretching dominated structure showed higher specific mechanical properties.

• Graded scaffolds showed brittle failure at lower strains than uniform scaffolds.

• Deformation mechanisms highly affected mechanical properties of scaffolds.

Abstract

Since the advent of additive manufacturing techniques, triply periodic minimal surfaces have emerged as a novel tool for designing porous scaffolds. Whereas scaffolds are expected to provide multifunctional performance, spatially changing pore patterns have been a promising approach to integrate mechanical characteristics of different architectures into a unique scaffold. Smooth morphological variations are also frequently seen in nature particularly in bone and cartilage structures and can be inspiring for designing of artificial tissues. In this study, we carried out experimental and numerical procedures to uncover the mechanical properties and deformation mechanisms of linearly graded porosity scaffolds for two different mathematically defined pore structures. Among TPMS-based scaffolds, P and D surfaces were subjected to gradient modeling to explore the mechanical responses for stretching and bending dominated deformations, respectively. Moreover, the results were compared to their corresponding uniform porosity structures. Mechanical properties were found to be by far greater for the stretching dominated structure (P-Surface). For bending dominated architecture (D-Surface), although there was no global fracture for uniform structures, graded structure showed a brittle fracture at 0.08 strain. A layer by layer deformation mechanism for stretching dominated structure was observed. For bending dominated scaffolds, deformation was accompanied by development of 45° shearing bands. Finite element simulations were also performed and the results showed a good agreement with the experimental observations.

Introduction

Functionally graded scaffold design has attracted huge attention during the last decades thanks to the dawn of additive manufacturing techniques thereby development of architecture-controlled porous biomaterials. A variety of human tissues including cartilage, bone, heart valves, nerves, muscles, liver etc. are being revived through tissue engineering scaffolds which are employed as a temporary replacement for the sake of tissue grafting and repair. Designing porous biomaterials, which are served as a three-dimensional extra cellular matrix (ECM) for initial cell attachment and subsequent tissue/organ formation is accompanied by conflicting requirements from both structural and biological standpoints (Sobral et al., 2011). Ideally, a porous scaffold is expected to best mimic the host tissue from mechanical and biological standpoints (Lin et al., 2004, Yan et al., 2015). Biomechanical properties are expected to match those tissues at the site of implantation (Bahraminasab et al., 2014, Kadkhodapour et al., 2015, Rajagopalan and Robb, 2006) and since TPMS structures are previously shown to provide high values of permeability (Kapfer et al., 2011), high surface to volume ratio, and topological complexity for cell ingrowth, structural modulation of TPMS-based scaffolds can provide promising designs in which many requirements of tissue engineering scaffolds are satisfied.

Although increasing porosity favors cell ingrowth and biological activities, it diminishes mechanical strength. On the other hand, high Young׳s modulus of the scaffold biomaterial with respect to the host tissue leads to the stress shielding effects and should be prevented by tailoring porosity and pore architecture. Given the complexity of the design requirements, the current trend for porous architecture design is to locally manipulate the internal pore architecture so that the above-mentioned demands can be satisfied by simultaneous function of different regions (Khoda et al., 2010). In this way, a compromise is made between mechanical strength and permeability (which is associated to the potential for cell ingrowth). For instance, where scaffold is designed to fill in the cylindrical voids, a radial gradient in volume fraction leads to a scaffold structure in which peripheral areas contribute in biological penetration while the central areas play role in load bearing. Also, during linkage replacement, a smooth longitudinal variation could be desired in porous design procedure to mimic the incipient tissue.

Based on the aforementioned pre-requisites, one major goal in scaffolds production is to maintain control over macro- (e.g., spatial form, mechanical strength, density, porosity) and micro-structural (e.g., pore size, pore distribution, pore interconnectivity) properties (Yoo, 2012, Yoo, 2011). Development of additive manufacture (AM) techniques such as solid freeform fabrication (SFF) and rapid prototyping (RP) has significantly improved control over the pore network architecture of tissue engineering scaffolds (Amirkhani et al., 2012, Parthasarathy et al., 2011, Yoo, 2012, Yoo, 2013). While fabrication methods such as SFF or RP have established remarkable advances in the biomaterials and tissue engineering, there have been many challenging issues in the design of scaffolds to meet multiple biophysical and biological requirements due to the limitations of currently available commercial CAD systems in modeling massive porous architectures. Such challenges are basically originated from two major sources: first, the mathematical definition of micro-architectures along with the macro-structure requires a huge memory on current CAD systems. Secondly, even if such mathematical definition can be generated using CAD systems, it would be almost impossible to transfer such detailed micro-structure information to SFF or RP machines using the current standard file formats such as STL due to massive resulted files (Yang et al., 2014; Yoo, 2012, Yoo, 2012b). In another study, (Jande et al., 2014) the mechanical and physical properties of a uniform and porous polyamide structure were characterized, which were then utilized for producing a composite structure of porous polyamide infiltered with epoxy using selective laser sintering 3D printing method. The effect of different processing parameters of selective laser sintering for fabrication of porous structures with a wide range of porosities was also investigated to find a controllable manner of porous and composite structures fabrication in terms of porosity. Furthermore, (Erdal et al., 2010) it was proposed that mechanical properties of functionally graded porous structures can be controlled by regulating the energy density level applied during selective laser sintering using mechanical testing techniques, due to the effect of energy density on porosity of the manufactured models.

Triply periodic minimal surfaces have emerged as a promising tool for designing regular porous architecture of scaffolds (Kapfer et al., 2011, Yoo, 2012, Yoo, 2014). Apart from the regular architectures which have proven to provide enhanced cell penetration thanks to increasing permeability (Melchels et al., 2010, Olivares et al., 2009), there are many advantages of TPMS-based designs since their open cellular nature along with their complexity which makes them suitable for cell differentiation. TPMS architectures provide smooth joints leading to less stress concentrations which are seen in typical lattice network structure. Whereas TPMS structures are mathematically defined, gradient structures in which either porosity or pore architecture varies throughout the scaffold structure, can be readily generated by defining weight functions (for changing pore architecture) or spatial depended porosity function (for locally controlling porosity). The great interest of many scientists to TPMS is also apparently due to their existence in both the nature and human body (Yoo, 2012, Yoo, 2012b).

Sun et al. (2015) introduced a real function based method to form complex heterogeneous porous architectures and indicated its versatility for designing hybrid, multi-scale, gradient, and stochastic structures and a combination of them as well. They evaluated the capability of their design through additive manufacturing of models in combination with micro-CT images of bone. Almeida and Bartolo (2014) also assessed two different types of TPMS topologies, namely Schwarz and Schoen and addressed their mechanical behavior using finite elements method for such biomimetic scaffolds with high surface to volume ratio and high porosity. Moreover, they carried out a parametric study on thickness and surface radius effect on compressive mechanical properties. In another study (Yang and Zhou, 2014), an effective approach for producing cellular structures was presented by combining operation of multiple substructures based on given substructures and boundaries. They reported the high potential for gradient porous scaffold design and fabrication using additive manufacturing techniques. The studies by Yang et al. (2014) suggested two CAD methods in which different TPMS structures are combined with the given transition boundaries. Their approaches were based on using Sigmon Function (SF) and Gaussian radial basis function for more general cases which demonstrated to be applied effectively for designing cellular structures fabricated by AM methods.

Apart from the extensive amount of data available in the literature on mechanical behavior of stochastic graded materials (Beletskii et al., 2003), the studies on mechanical properties of gradient structures based on regular architectures are scarce (Ajdari et al., 2011, Khanoki and Pasini, 2012). Gradient patterns hold promise to design scaffolds for those tissues in which the structure and material spatially change. For instance, in order to mimic the behavior of host tissue in osteochondral defects, the scaffold at bottom layers is expected to mechanically match cancellous bone while in next layers it should be close to calcified cartilage and eventually articular cartilage. Hence, a gradation of porosity can help to tailor the scaffold for cartilage tissue engineering (Li et al., 2015). With this in mind, we show how mechanical properties for regular bending and stretching dominated porous architectures can be controlled by considering a longitudinal and linear gradient pattern in porosity. To this, we utilized triply periodic minimal surface based architectures (P and D surfaces) and introduced a linear increase in relative density from 30 to 60% in the CAD modeling, and utilized additive manufacturing techniques to manufacture the scaffolds. Mechanical properties were discussed along with the failure mechanisms in order to put the results together with those obtained in our previous studies (Kadkhodapour et al., 2014) on uniform TPMS-based topologies studied in the present work.