The Construction of Biologically Relevant Fiber-Reinforced Hydrogel Geometries Using Air-Assisted Dual-Polarity Electrospinning

Fiber-reinforced hydrogels are a class of soft composites, where the hydrogel matrix is reinforced with soft polymeric fibers [1–9]. The presence of the electrospun reinforcing fibers results in both improved mechanical properties (when compared to the pure hydrogel) as well as morphological heterogeneity replicating the extra cellular matrix found in soft tissues. As described by Shafiee and Atala [10], the fabrication of biological analogs can be classified based on their level of complexity with the simplest biologically relevant geometry being that of flat, laminated constructs (Fig. 1(a)). This is followed in increasing complexity by cylindrical constructs ((Fig. 1(b)) and hollow constructs (Fig. 1(c)). Current manufacturing processes associated with fiber-reinforced hydrogels are primarily limited to flat geometries and some examples of basic cylindrical tubes. Thus, there is a need to develop versatile manufacturing processes to realize all three types of constructs relevant to diverse biomedical applications.

Electrospinning has been used to incorporate micro/nanoscale polymer fibers (10¹–10³ nm in diameter) directly onto a layer of hydrogel to realize flat, laminated geometries [11,12]. However, the electrospinning process is limited to thicknesses below 1 mm due to deterioration of the electric field as fiber and hydrogel layers are sequentially deposited. Methods for producing cylindrical fiber-reinforced hydrogel geometries rely on a three-step process: viz., (i) electrospinning a randomly oriented mat of polymer fibers, (ii) rolling the mat onto a cylindrical mandrel, and (iii) infusing hydrogel [13–16]. However, this approach is unable to account for more complex tubular geometries, such as bifurcations and mesoscale texturing on the surface of the fiber-reinforced hydrogel geometry. Hollow, nontubular geometries have not been realized using fiber-reinforced hydrogels due to two limitations of the electrospinning process. First, electrospinning requires the use of a grounded metallic collection substrate to collect the fibers, which poses a challenge when dealing with a three-dimensional (3D), hollow geometry. Second, the realization of free-standing, hollow hydrogel constructs like the ones depicted in Fig. 1(c) requires the use of a relatively thick fiber layer (>1 mm thick), to support the overall weight of the construct.

Given the current processing limitations of electrospinning when applied to fiber-reinforced hydrogels, this paper presents the development of a novel air-assisted, dual-polarity electrospinning printhead to manufacture fiber-reinforced hydrogel geometries across all three classes of biologically relevant geometries, identified in Fig. 1. This novel printhead eliminates the substrate shape/electrical-conductivity dependency of the electrospinning process. This capability when combined effectively with different fiber collector configurations resulted in the realization of (i) tubular geometries with bifurcations and mesoscale texturing, (ii) hollow, nontubular geometries, and (iii) flat, laminated geometries with varying amounts of microscale fibrous reinforcement. In addition, all three classes of prototype geometries were mechanically characterized to have properties that were in line with those observed in living soft tissues.

The remainder of this paper is organized as follows: Sec. 2 presents the design and characterization of the air-assisted, dual-polarity electrospinning printhead. Section 3 describes the process and prototype geometries that were produced in this work. Section 4 details the mechanical characterization performed on all three geometric classes, which include uniaxial tensile load testing, pressurized rupture testing, and unconstrained compressive load testing. Finally, Sec. 5 presents specific conclusions that can be drawn from this work.

Far-field electrospinning is a common process used to realize the polymer fiber reinforcement necessary to create fiber-reinforced hydrogels. Electrospinning utilizes high-strength electric fields to produce polymer fibers that have diameters ranging from 10¹ to 10³ nm [17–19]. However, the use of electrospinning has been limited to two-dimensional mats, and it has been a challenge to expand this process to 3D geometries. This limitation is due to the required generation of an electric field between the deposition needle and collection substrate. This dependency on the electric field requires that the collection substrate be (i) electrically conductive (i.e., metallic) and (ii) simple in geometric complexity (e.g., flat, cylindrical) to avoid disrupting the directionality and distribution of the vectorial field. Fiber mats that are produced on these electrically conductive substrates have a maximum thickness of the order of 10² μm. Inherently, this limits the feasibility of using this process toward the creation of biologically relevant 3D freeform geometries, as shown in Fig. 1.

A system-level view of the air-assisted, dual-polarity electrospinning printhead addressing these limitations is shown in Fig. 2(a). This dual-polarity process charges two blunt-ended fiber deposition needles to opposite polarities (i.e., positive and negative charges). It has been reported that fibers that are deposited through the electrospinning process carry a residual charge, based on their deposition polarity [20]. This residual charge (both positive and negative) remains on the fiber, until they are grounded, and as such results in an attractive force between the fibers generated within this printhead. The attractive force causes the fibers to collect in the center of the printhead housing, where a continuous, low-velocity stream of air is then used to propel the fibers toward the fiber collection zone.

Given the absence of a grounded collection substrate, any fiber collection template (regardless of its electrical conductivity or geometry) can be placed within the fiber collection zone (Fig. 2(a)). When motion is applied to this template, a conformal coating of electrospun fibers is produced on its surface. This capability of the printhead is demonstrated ahead in Sec. 3 using 3D thermoplastics, hydrogels, and collapsible elastomeric constructs as collection substrates.

In order to deposit polymer fibers, the air-assisted, dual-polarity electrospinning printhead relies on the convergence of two fields, viz., (i) the electrostatic field to generate fibers between the deposition needles and (ii) the air velocity field to propel the fibers into the fiber collection zone and then onto a collection template. Finite element simulations were performed to quantify the combinatorial impact of these two fields on the overall fiber deposition process using (i) the Flow Simulation package included in solidworks™ 3D Modeling software and (ii) quickfield™, electrostatic field simulation software. Details associated with the design simulation parameters can be found in the Supplementary Materials on the ASME Digital Collection, with specific boundary condition schematics seen in Figs. S1 and S2 available in the Supplementary Materials. These simulation results are shown in Figs. 2(b) and 2(c), respectively. Within the volume of the printhead, there are two regions identified in Figs. 2(b) and 2(c), viz., (i) Region-1 (the annular region closest to the two fiber deposition needles) and (ii) Region-2 (the circular region near the axisymmetric centerline of the cylindrical printhead).

The air velocity field within the printhead has two requirements to aid effective fiber deposition. First, the air flow within Region-1 must be near-zero, to avoid any detrimental effects on the Taylor cone formed by the electrostatic forces acting on the polymer solution at the tip of the fiber deposition needles. Second, the air flow within Region-2 must possess a sufficient velocity component along the Y-axis (i.e., V_Y), to displace and deposit fibers into the fiber collection zone. Figure 2(b) displays the simulation results for the air velocity field observed within the printhead volume. As seen in Fig. 2(b), an inlet diffuser was used as a design feature to shape the air flow-field. The diffuser was designed to increase the cross-sectional area of the inlet to reduce the magnitude of the air velocity entering the printhead. Within Region-1, the velocity of the air flow remains near-zero, especially close to location of the fiber deposition needles. In Region-2, the air velocity profile expands radially outward from the centerline of the printhead after it passes through the diffuser. In addition, the velocity profile maintains a velocity of ∼0.5 m/s down the Y-axis (Fig. 2(b)). This velocity magnitude is sufficient to displace the fibers into the fiber collection zone.

The electrostatic forces are responsible for the generation and ejection of the polymer fibers. Therefore, the purpose of the electric field is to (i) overcome the surface tension forces of the polymer solution to eject a fiber and (ii) provide a residual charge on the fibers so that they are attracted to one another, from each needle. The simulated magnitude of the electric field (E) within the printhead volume is shown in Fig. 2(c). The magnitude of the electric field is critical in this process, as it will influence the attraction between the fibers ejected from each needle. The deposition needles were charged to an electrical potential of +8 kV and −8 kV in this simulation. Within Region-1, near the fiber deposition needles, the strength of the electric field approaches 300 V/mm, which is the magnitude required for electrospinning processes [21,22]. As one moves radially inwards away from the Region-1, and toward the centerline in Region-2, the electric field decreases to ∼150 V/mm. The magnitude of the electric field also continues to decrease along the Y-axis, toward the fiber collection zone [Fig. 2(c)].

Figure 2(d) displays the spatial variation seen in the electric field magnitude, E, and air velocity, V_Y, across regions 1 and 2. The radially varying values in Fig. 2(d) are plotted for the height corresponding to the tip of the fiber deposition needles. As seen in Fig. 2(d), the electric field generates the primary force in Region-1. As a result, the ejected fibers experience a radial pull toward the centerline of printhead. Then, as the ejected fibers enter into Region-2, the force from the airflow overcomes the net forces from the electric field and residual charges contained on the fibers. The ensuing effect is a downward (+Y direction) displacement of the fibers toward the fiber collection zone. In addition, as the fibers continue to be displaced downward, the magnitude of the electric field acting on the deposited fiber continues to decrease, while the air flow velocity remains relatively constant. This further promotes a downward displacement of the deposited fibers toward the fiber collection zone.

In order to realize the air-assisted, dual-polarity electrospinning printhead, the following components were integrated. Two high voltage power supplies, one positive polarity, and one negative polarity (ES-Series, Gamma High Voltage, Ormond Beach, FL) were used to charge the stainless steel, blunt-ended fiber deposition needles (25-gauge by 12 mm long), in order to generate an electric field. The polymer solution was independently supplied to each needle using two syringe pumps (Razel^™, St. Albans, VT, Pt. No. R-99). A push-to-connect fitting was used to supply compressed air to the center of the cylindrical printhead, which had an overall diameter of 125 mm. The velocity of the compressed air passing through the printhead was controlled using a metering valve (Nordson, East Providence, RI). The cylindrical printhead housing was produced using a stereolithography rapid prototyping technique (FormLabs, Somerville, MA).

The electrospinning solution was similar to the one reported in previous studies [6,11,22,23]. The polymer solution consisted of polycaprolactone (PCL) (10 wt.%, Mn = 80,000, Sigma Aldrich, St. Louis, MO), pluronic acid, F-127 (1 wt.%, Sigma Aldrich), and reagent-grade acetone (balance, VWR International, Radnor, PA). Pluronic Acid F-127 improved the hydrophilicity of PCL, which promoted the infusion of the subsequent hydrogel solutions. The polymer solution was mixed at 60 °C for 12 h. The kinematic viscosity, surface tension, and electrical conductivity values for this solution are around 11.15 ± 0.05 (mm²/s), 20.2 ± 1.0 (mN/m), and 0.302 ± 0.014 (μS/cm), respectively [23].

The electrospinning parameters used for this solution with the air-assisted, dual-polarity, electrospinning printhead are shown in Table 1. Scanning electron microscopy (SEM) of the ensuing fiber preforms shows that the individual fibers average around 0.5 ± 0.2 μm (n = 300) in diameter. The use of the dual polarity electrospinning head results in lateral adhesion between some individual fibers to yield larger effective diameters averaging around 1.3 ± 0.3 μm (n = 96). A representative SEM image can be seen in Fig. S3 in the Supplementary Materials on the ASME Digital Collection.

The hydrogel solution used to create the tubular and hollow geometries (Secs. 3.1 and 3.2, ahead) was sodium alginate (3.0 wt.%, Sigma Aldrich) and de-ionized (DI) water (balance). For the flat, laminated fiber-reinforced hydrogel geometries (Sec. 3.3), the hydrogel solution was similar to one reported by an extrusion-based bioprinting process [24]. The solution consisted of sodium alginate (2.5 wt.%, Sigma Aldrich), gelatin (8.0 wt.%, Sigma-Aldrich), and DI water (balance). This gelatin/alginate hydrogel solution was favorable for extrusion-based deposition because it allowed for partial solidification in-between each hydrogel layer deposition step, through a reduction in temperature caused by the presence of gelatin, followed by a final ionic cross-linking of the alginate. Commercially available food coloring was added to the solutions for visualization purposes. The ionic cross-linking solution for all geometries consisted of calcium chloride (5 wt.%, Sigma Aldrich) and DI water (balance).

Figure 3(a) outlines the manufacturing process used to create tubular geometries. This process follows the same principle as the fiber film infusion process seen in the manufacturing of fiber-reinforced hydrogel composites [25]. However, the two-dimensional fiber film has been replaced with a 3D fiber preform. Polymer fibers are deposited onto the thermoplastic template rotating at a low rotational speed (20 rpm). Details associated with the fabrication and assembly of the template are discussed in the Supplementary Materials on the ASME Digital Collection (Part 3). First, the template is assembled into the shape of the desired fiber-reinforced hydrogel geometry (Step 1). Next, the fiber is deposited onto the template using the air-assisted, dual-polarity, electrospinning printhead (Step 2). Then, the 3D fiber preform and template are submerged sequentially in the hydrogel and cross-linking solution for 15–30 min each, depending on the size of the prototype (steps 3 and 4). Finally, after cross-linking, the template is disassembled and removed from the tubular fiber-reinforced hydrogel (Step 5).

There are two benefits of using thermoplastic templates to create the tubular geometries. First, thermoplastics allow for easier removal of the fiber perform due to lower surface energy and surface roughness than traditional metallic collection substrates. Second, the thermoplastic material allows for more complex 3D printed geometries to be produced, as shown in Fig. 3(b). These freeform geometries, such as bifurcations and tubes with mesoscale texturing, can be realized by designing the template out of multiple pieces that can be fastened together during the fiber deposition process (Step 2), and then disassembled and removed following the cross-linking of the hydrogel ((Fig. 3(b)). In addition, mesoscale texturing can be applied on the surface of the tubular geometries. Such templates can be printed using rapid prototyping processes to have customized textures and geometries that mimic those found in the human anatomy.

Figures 4(a)–4(d) show four tubular prototype geometries, produced using the aforementioned process (Fig. 3(a)). A straight annular tube is shown in Fig. 4(a). The dimensions of the tube are 13 mm (inside diameter) × 60 mm (tall). This geometry was used for tensile testing and pressurized flow testing, as described ahead in Sec. 4.1. The fiber weight fractions in these prototypes were estimated to be 5.3 ± 0.7 wt.%, based on fiber preform weight measurements taken before and after hydrogel infusion/cross-linking. A 45 deg bifurcation and 90 deg bifurcation are shown in Figs. 4(b) and 4(c), respectively. These tubes have an inside diameter of 25 mm, which is representative of the diameter of large vasculature (e.g., aorta) seen in the human body. Finally, an annular tube geometry with mesoscale texturing is shown in Fig. 4(d). The tube is 30 mm in diameter and 115 mm long and mimics the geometry seen in structures such as the intestinal tract. The textured ridges present along the axial length of the part geometry have a depth of ∼2 mm and have a periodic spacing of 10 mm. In order to provide enough fiber reinforcement to support the mesoscale ridges, the total fiber deposition time was ∼120 min for the prototype in Fig. 4(d).

The process to create hollow, nontubular geometries is like the one presented for tubular geometries (Fig. 3(a)) and is shown in detail in Fig. S4 (see the Supplementary Materials on the ASME Digital Collection). However, here the thermoplastic template is replaced with an inflatable elastomeric material that serves as a pneumatically morphable template to collect the deposited fibers. Details associated with the pneumatically morphable template are discussed in the Supplementary Materials (Part 3). To ensure conformal coating of the hollow nontubular geometries the pneumatically morphable template is rotated at 20 rpm to collect the fibers from the air-assisted, dual-polarity electrospinning printhead. The preform is then infused with hydrogel solution and ionically crosslinked. After this, the pneumatically morphable template is collapsed and extracted out of the hydrogel structure.

Two types of hollow geometries can be produced using this process, viz., (i) single entrance geometries and (ii) dual-entrance geometries, which mimic the general shape of hollow organs, such as the stomach or bladder. In order to achieve these two types of hollow geometries, the pneumatically morphable templates are shown in Fig. 3(c). Single-entrance geometries utilize one spherically inflated elastomeric template (see Fig. S4 available in the Supplementary Materials), while dual-entrance geometries utilize a second spherically inflated elastomeric template. The second template is rotated in the same direction and speed using an additional rotary stage. Three prototype geometries of hollow fiber-reinforced hydrogels were produced using the aforementioned process and shown in Fig. 5. The single-entrance sphere (Fig. 5(a)) was ∼100 mm in diameter and had a 25 mm diameter entrance. This geometry was used for mechanical characterization by pressurized rupture testing, as described ahead in Sec. 4.2. A dual-entrance, hollow sphere is shown in Fig. 5(b), and a dual-entrance, bi-spherical geometry is shown in Fig. 5(c). The diameter of the spheres shown in both Figs. 5(b) and 5(c) is ∼100 mm.

Unlike the first two processes that involved continuous fiber deposition on a rotating template to create a 3D fiber preform, the process required to create flat, laminated geometries involves a layer-by-layer approach with alternating steps of hydrogel and fiber deposition. The layer-by-layer approach is commonly used in the processing of fiber-reinforced hydrogels, but is limited in height and resolution by the far-field electrospinning process [25]. Figure 6 outlines the manufacturing process for constructing flat, laminated fiber-reinforced geometries. The manufacturing process involves printing a layer of the thermo-responsive gelatin/alginate hydrogel via syringe deposition (Step 1). After deposition and semisolidification of the hydrogel layer (∼2–3 min), the fiber reinforcement layer is deposited using the printhead (Step 2). Since the deposition area of the printhead is larger than the cross-sectional area of the part, a trimming step is at times needed for the fibers at the periphery of the layer (Step 3). This alternating sequence of hydrogel deposition, fiber deposition, and fiber trimming is repeated until the desired part height is reached (Step 4). Since there is no grounded collection substrate required for the fiber deposition process, large build heights (10 s of mm in height) can be achieved. After all of the hydrogel and fiber layers were printed, the part was placed in an ionic cross-linking solution for 12 h, given the volume and thickness of the hydrogel printed during the process.

Two prototype geometries of flat, laminated fiber-reinforced hydrogels were produced. The two geometries are a circular pillar and an intervertebral disk geometry (shown in Figs. 7(a) and 7(b), respectively). The circular pillar was 40 mm in diameter by 10 mm tall (30 layers). This geometry will be used ahead in Sec. 4.3, for mechanical characterization of the compressive strength of the flat fiber-reinforced hydrogel. The intervertebral disk geometry was 55 mm × 35 mm × 30 mm tall (90 layers). Figure 7(c) displays a binary optical image of the cross section of the laminated fiber reinforced hydrogel. The fiber reinforcement can be seen deposited at equal distances along the build direction of the part. The denser regions of white pixels indicate an increase in concentration of fiber reinforcement within the layer, which is indicative of the random deposition pattern from the fiber deposition printhead. The cross section images indicate the syringe-deposited hydrogel layer to be 300–350 μm thick. In addition, there are no voids present between the individual layers, which is indicative of good adhesion between the layers of fiber reinforcement and the hydrogel matrix.

Table 2 presents the processing times for the different classes of prototype geometries that were discussed in Secs. 3.2–3.4. The flat, laminated geometries possess the longest processing time given (i) the layer-by-layer process and (ii) the long cross-linking time required for the thick hydrogel sample (10 mm).

The tubular and hollow, nontubular geometries are processed using a similar approach (Figs. 3 and S4). Given these approaches, the longest processing step is the fiber deposition process using the air-assisted, dual-polarity electrospinning printhead. The fiber deposition time is directly proportional to the surface area of the geometry being produced. This is illustrated by the fact that the prismatic cylinder geometry (Fig. 4(a)) required 0.5 h of electrospinning, while the larger dual-entrance, bi-spherical geometry (Fig. 5(c)) required ∼8 h. In addition, the 45 deg and 90 deg bifurcation geometries required repositioning during the fiber deposition step so that the rotational motion could be applied concentric to each of the three cylindrical segments present in these geometries. Both geometries required three equal duration, repositioning steps during fiber deposition.

Given that the fiber deposition process is the main bottleneck, the following two options can be explored to reduce the time required to create a 3D fiber preform. First, the current printhead used only two fiber deposition needles. The inclusion of additional needles would increase the throughput of fibers being deposited from this printhead. Second, collection templates can be assembled in parallel to create multiple 3D fiber preforms at once. This would split the overall processing time required to create these geometries across multiple prototype geometries.

The mechanical strength of the fiber-reinforced hydrogels was characterized using (i) tensile load tests, (ii) pressurized flow tests, (iii) pressurized rupture tests, or (iv) compressive load testing, based on the class of the geometric construct (Sec. 3). Sections 4.1 and 4.2 explain the methods and results for these mechanical characterization tests in detail.

In order to characterize the tensile strength of the tubular fiber-reinforced hydrogels that were produced (Fig. 4(a)), tensile tests were carried out in air using a motion platform and piezo-electric load cell (Kistler Pt. No. 9256C1, Switzerland). The tests were performed with a crosshead velocity of 1.0 mm/min and a force sampling frequency of 100 Hz. The wall thickness of the fiber preforms used for the tensile tests was measured using a digital micrometer (Mitutoyo Digital APB-2D, Resolution = 1 μm) to be 400 ± 50 μm. Information on the measurement protocol can be found in the Supplementary Materials on the ASME Digital Collection (Part 6). A total of three tensile tests were performed for the fiber-reinforced hydrogel and dry PCL fiber preform. A summary of the resulting tensile properties from this work is shown in Table 3.

Pure alginate hydrogel possesses low tensile properties of 10–65 kPa for the elastic modulus and 10–40 kPa for the ultimate tensile strength [26]. The addition of the electrospun polymer fiber-reinforcement resulted in an approximately 20- and 18-fold increase in the elastic modulus, and ultimate tensile stress, respectively, as compared to literature-reported values of pure alginate. As expected, the material properties of the fiber-reinforced hydrogel are higher than those of the pure hydrogel matrix, but lower than those for the fiber reinforcement phase.

Figure 8 benchmarks the ultimate tensile strength and tensile elastic modulus of the fiber-reinforced hydrogels produced in this work, to the tensile properties of vasculature tissues reported in literature [27–31]. As seen the ultimate tensile strength and elastic modulus reported in this work, fall within the range of reported values. Further tuning of these mechanical properties could be performed with different hydrogel/fiber material combinations.

Pressurized flow characterization was performed to test the fluid perfusion capabilities of the tubular fiber-reinforced hydrogel geometries. Details associated with the pressurized flow characterization are presented in the Supplementary Materials (Part 6). For this test, water flow pressure was increased, using a pneumatic controller (Performus V EFD, Nordson), until the rupture of the tubular structure. The wall thickness of the tubular geometries was varied based on the electrospinning time to create the fiber preforms (Fig. 3(a), Step 2). Each wall thickness category was tested three times. The burst pressure of the tubes as a function of the wall thickness is shown in Fig. 9. As seen in Fig. 9, the burst pressure of the tubes increases based on the wall thickness of the fiber-reinforced hydrogel. Tubes with a wall thickness greater than 0.2 mm can withstand pressures larger than those seen in the human blood flow, without rupturing.

In order to characterize the mechanical properties of the hollow, nontubular fiber-reinforced hydrogel geometries, pressurized airflow-based rupture tests were carried out on the single-entrance, hollow spheres (Fig. 5(a)). The wall thickness of the fiber preforms was varied based on the electrospinning time (Fig. 3(a), Step 2). A total of three fiber preforms were first manufactured for each thickness variation and then destructively tested by increasing the gauge pressure of the compressed air introduced into the hollow geometry. An optical camera was used to record the inflation and rupture of the hollow fiber-reinforced hydrogel geometries. The experimental setup and images from a pressurized rupture test are shown in Figs. 10(a) and 10(b), respectively. The failure point, as indicated in image (iv) and (v) in Fig. 10(b), occurred near the apex of the hollow, nontubular geometry for all samples tested.

where the variables are identified in Fig. 10(a). It should be noted that while Eq. (1) is derived for a spherical geometry, the geometries seen in this work were elliptical in nature (Fig. 10(b)). To account for this elliptical shape, a nominal radius, R_NOM, was calculated based on the average of the major and minor radii measured in the samples during this rupture test. The values for the rupture stress, wall thickness, and prerupture nominal radius are shown in Table 4.

As seen in Table 4, the wall thickness and rupture stress of the fiber-reinforced hydrogel can be tuned based on the electrospinning time, with longer electrospinning times creating thicker and stronger hollow geometries. The increase in the rupture stress is due to the larger amount of fiber reinforcement present within the composite. A four-fold increase in electrospinning time (15–60 min) is seen to result in a 61.4% increase in the associated rupture stress.

Unconstrained, compressive load tests were carried out on the circular pillar geometry (depicted in Fig. 7(a)) using a linear motion platen in conjunction with a piezo-electric load cell (Kistler Instrument Corp., Pt. No. 9256C1, Switzerland). The compression tests were performed with a crosshead velocity of 0.1 mm/min, and a force sampling frequency of 100 Hz. A total of three compression tests were carried out on the fiber-reinforced hydrogel sample to obtain an average compressive modulus value. The compressive load tests were performed to strain values that remained within the elastic region of deformation to avoid nonlinear plastic deformation effects and to stay below the rated load on the motion stage (20 N).

Characteristic compressive load–displacement curves were obtained for the neat hydrogel samples (i.e., with no fiber reinforcement), and for both the low and high-fiber density fiber-reinforced hydrogels. The fiber density was varied based on the fiber deposition time during the layer-by-layer manufacturing process (Step 2, Fig. 6), as seen in optical images shown in Fig. S5 (see the Supplementary Materials on the ASME Digital Collection). The low-fiber density and high-fiber density samples had fiber deposition times of 10 s and 30 s, respectively. The fiber number density, ς, of the samples was determined from five optical microscopy images (Zeta 20, Zeta Instruments, Milpitas, CA). The fiber number density for the low and high fiber density samples was determined to be ∼590 fibers/mm², and ∼2900 fibers/mm², respectively.

Table 5 presents the compressive modulus (E_C), of each of the flat, laminated fiber-reinforced hydrogel samples. The addition of the low fiber density and high fiber density reinforcement, respectively, resulted in a 37.4% and 109% increase in the compressive modulus as compared to the neat hydrogel. This indicates that the fiber deposition time can be tuned to deposit the desired amount of fibrous reinforcement per layer to result in a targeted compressive modulus value for a given application.

The measured values for the compressive modulus fall within the 50–250 kPa range that has been reported for the compressive modulus for the knee meniscus [33]. Given the anisotropic behavior seen for load-bearing soft tissues, such as the knee meniscus, further tuning of the mechanical properties of interest can be easily performed by varying the fiber reinforcement material and fiber loading, using the air-assisted, dual-polarity electrospinning printhead.

An air-assisted, dual-polarity electrospinning printhead was designed, implemented, and demonstrated in this work. Through the convergence of the electrostatic forces with a low velocity air flow, the need for a grounded metallic collection substrate has been circumvented. Configurations of deformable templates were deployed to collect the deposited electrospun polymer fibers. Using this printhead, three classes of fiber-reinforced hydrogel geometries were produced as use-cases to showcase the versatility of this process. Mechanical characterization was carried out on these fiber-reinforced hydrogels and showed comparable properties to soft tissues seen in the human anatomy. The flexibility of this fiber deposition process when used in conjunction with different deformable template configurations, lends itself toward a variety of biofabrication application areas, including cell seeding for in vitro tissue models.

• The U.S. National Science Foundation, Division of Civil, Mechanical and Manufacturing Innovation (Award Nos. CMMI CAREER 13-51275 and 14-62648; Funder ID: 10.13039/100000147).

• Rensselaer Polytechnic Institute (Funder ID: 10.13039/100007092).

• U.S. National Science Foundation Graduate Research Fellowship Program, Division of Graduate Education (Grant No. DGE 12-47271; Funder ID: 10.13039/100000082).