Fused Deposition Modeling and Mechanical Properties of Porous Titanium Scaffolds

Abstract

To address issues such as thermal stress concentration in metal bone implants produced via high-energy beam direct additive manufacturing, a method was proposed to fabricate porous titanium scaffolds. This approach combined Fused Deposition Modeling (FDM) with a debinding–sintering process. Ti/ABS composite filaments with titanium volume fractions of 35%, 40%, and 45% were successfully developed via a single-screw extrusion process. Their feasibility in the FDM process was subsequently verified. The effects of different processing parameters on the forming quality and dimensional accuracy of the green bodies were investigated. After debinding and sintering the composite scaffolds prepared with optimized parameters, structurally intact porous titanium scaffolds were obtained. Microscopic characterization shows that the scaffold surface consists primarily of titanium, and the pore structure remains intact. Furthermore, compression tests were performed on three types of porous titanium scaffolds with different porosities. The results indicate that the combination of ABS/titanium alloy composite filaments, FDM technology, and debinding–sintering post-processing enables the high-quality and efficient production of porous titanium scaffolds. The elastic modulus of the resulting scaffolds ranges from 1.2 to 1.6 GPa, and the compressive strength is between 25.7 and 68.3 MPa. The elastic modulus matches that of human cancellous bone. Meanwhile, the compressive strength is significantly higher than that of cancellous bone and falls between the values for cancellous and cortical bone. These mechanical properties meet the requirements for human bone, providing a new approach for the manufacture of orthopedic implants.

1. Introduction

Currently, bone defects caused by fractures, spinal deformities, and falls often require surgical repair and rely on orthopedic implants for fixation or support. Titanium is widely used in orthopedic implants [1,2]. Titanium is a lightweight, high-strength metal with a relatively low density (4.51 g/cm³), excellent biocompatibility, a high specific strength, and an elastic modulus close to that of natural bone, which can effectively reduce the risk of immune rejection. To further enhance the performance of orthopedic implants, porous structural designs have been shown to play a key role in the osseointegration process. Porous titanium scaffolds not only promote bone tissue healing and provide better mechanical support but also improve the interfacial bonding strength between bone and implant [3,4,5]. In addition, porous structures help significantly reduce implant weight and lessen the risk of bone damage caused by stress concentration. Therefore, fabricating porous titanium scaffolds with rational structures has prominent application value and broad development prospects in orthopedics [6,7,8,9,10].

Current methods for preparing porous titanium scaffolds include traditional manufacturing routes such as powder metallurgy [11] and slurry foaming [12], as well as additive manufacturing. Among the conventional processes, powder metallurgy has been widely applied due to its mature technology, cost-effectiveness, and the ability to flexibly adjust the overall porosity by varying the space-holder content. Additionally, laser-induced titanium foaming using biocompatible foaming agents has been successfully developed to fabricate porous structures with rough surfaces, which effectively promotes bone cell adherence on implants [13]. However, these traditional methods still possess notable limitations, such as difficulties in the customized production of complex shapes, complicated processing steps, and poor controllability over the geometric shape and 3D interconnectivity of the macroscopic pore structures.

Additive manufacturing technologies—such as Laser Engineered Net Shaping (LENS) [14], Laser Powder Bed Fusion (LPBF) [15], Electron Beam Powder Bed Fusion (EPBF) [16], and Direct Ink Writing (DIW) [17]—offer new routes for producing porous titanium scaffolds. However, high-energy-beam techniques like LENS, LPBF, and EPBF inherently involve rapid heating and cooling, which tend to induce significant thermal stress gradients, leading to residual stress accumulation, part warping, or even cracking, thereby limiting their application to some extent in manufacturing porous titanium scaffolds [18]. Although Direct Ink Writing (DIW) offers advantages such as high material flexibility and room-temperature processing, its printed wet green bodies are highly susceptible to structural collapse and deformation during the subsequent drying process. To overcome these drawbacks, the indirect metal additive manufacturing process, which utilizes Fused Deposition Modeling (FDM) or Metal Fused Filament Fabrication (MFFF), has emerged as a highly promising alternative. FDM effectively avoids structural collapse because its extruded thermoplastic filaments solidify instantly upon cooling, providing the green bodies with high initial mechanical strength and ensuring the stability of complex porous architectures.

This technique features a smaller temperature change range, more stable temperature control, and does not rely on high-energy heat sources; it also places lower demands on the particle size and quality of metal powders. Through the layer-by-layer deposition of composite filaments loaded with a high volume fraction of metal powder, followed by subsequent debinding and sintering processes, densified metal components are ultimately obtained [19]. Compared with traditional processing, FDM can achieve faster fabrication speeds and produce complex shapes that are difficult to make using conventional methods. By optimizing printing parameters, the surface quality and dimensional accuracy of printed parts can be improved to meet specific medical requirements, which gives FDM broad application prospects in the customized production of biomedical materials [20,21].

Despite these advantages, recently explored extrusion-based metal additive manufacturing technologies have mainly been used for fabricating metal parts like 316L stainless steel and copper. Furthermore, most existing literature focuses primarily on the basic printability and densification processes of solid metal components [22,23]. To extend FDM to the fabrication of high-performance metal porous structures, key issues such as the preparation of composite filaments and the control of structural fidelity during debinding and sintering must be overcome. Currently, there is a limited number of studies on the fabrication of high-porosity pure titanium scaffolds using this technology. In particular, systematic investigations into the morphological retention, shrinkage behavior, and mechanical responses of complex porous structures during the debinding and high-temperature sintering processes are still lacking.

To bridge these research gaps, this study focuses on the FDM forming of high-titanium-content composites, aiming to leverage the layer-by-layer stacking advantage to achieve high-precision fabrication of complex porous scaffold geometries. The novelty of this study mainly consists of the manufacturing feasibility, the investigated material behaviors, and the evaluation of mechanical and biological quantities. Specifically, this study systematically explores the feasibility of fabricating porous pure titanium scaffolds using a combination of FDM and debinding–sintering, with a focus on the morphological fidelity of complex porous structures during the high-temperature densification process. Meanwhile, it analyzes the potential positive effects of titanium and its process-induced compounds (such as titanium oxide and titanium carbide) on the biocompatibility of the scaffolds. Furthermore, a comprehensive quantitative evaluation of the porosity, ultimate compressive strength, and elastic modulus is conducted to demonstrate that the scaffolds prepared using this process can achieve an excellent balance between load-bearing capacity and a low elastic modulus. This study provides a flexible and feasible innovative technical route for manufacturing porous titanium bone scaffolds with outstanding mechanical matching and osteointegration potential.

2. Experimental Section

2.1. Preparation Process of Porous Titanium Scaffolds

Figure 1 illustrates the process flow for fabricating porous titanium scaffolds via indirect additive manufacturing. To ensure uniform dispersion of titanium particles and prevent powder agglomeration, a solution blending approach was first employed. Titanium powder and polymer binders were mixed in a solvent under continuous stirring. The resulting slurry was cast into sheets and vacuum-dried to completely remove the solvent. These solid sheets were then crushed into granules to serve as flowable feedstock for extrusion.

Next, the granules were processed through a single-screw extruder to produce continuous composite filaments with a consistent diameter, ensuring stable material extrusion during 3D printing. These filaments were then utilized in a Fused Deposition Modeling (FDM) process to build the porous green scaffolds layer-by-layer. Finally, the green bodies underwent a controlled debinding and sintering thermal treatment to eliminate the organic matrix and fuse the remaining titanium particles, yielding structurally intact porous metallic scaffolds. Specific equipment and processing parameters for each stage are detailed in subsequent sections.

2.2. Preparation and Characterization of Ti/ABS Filament

Spherical titanium powder with an average particle size of 20–75 μm (Shanghai Naiou Nanotechnology Co., Ltd., Shanghai, China), ABS powder with an average particle size of 50 μm (Shanghai Fengtai Plastics Co., Ltd., Shanghai, China), and MBS and maleic-anhydride-grafted ABS powders, both with average particle sizes of 50 μm (Wanlihong Plastics, Dongguan, China), were used to prepare Ti/ABS granules. N,N-Dimethylformamide (DMF) was mixed with the ABS, MBS, and maleic-anhydride-grafted ABS powders and heated in a 70 °C water bath with continuous stirring until the polymers were completely dissolved in the organic solvent. A predetermined proportion of Ti powder was then added to the ABS solution; after magnetic stirring the ABS/Ti slurry for 30 min, it was poured onto a tin-foil tray and cast into sheets approximately 1 mm thick. The sheets were placed in a vacuum drying oven, heated under reduced pressure to 160 °C and held for 1 h, then removed to cool for 10 min. This cycle was repeated four times.

Subsequently, the dried composite sheets were cut into uniform granules using a paper cutter. The granules were finally extruded into Ti/ABS composite filaments using a single-screw extruder (Wellzoom Desktop Extruder Line II, Shenzhen Misida Technology Co., Ltd., Shenzhen, China). The extruder features an extrusion zone and a melting zone, with the temperatures set to 190 °C and 200 °C, respectively, and the screw speed set to 20 rpm. The diameter of the extruded filament was strictly controlled within the range of 1.75 ± 0.05 mm to ensure stable extrusion during the FDM printing process.

To evaluate the processability and properties of the filaments, several characterization techniques were employed. The Melt Flow Rate (MFR) was measured using a melt flow rate tester (RL-Z1B1, Shanghai Sierda Scientific Instruments, Shanghai, China). The MFR test was conducted at a temperature of 200 °C under a 5 kg load. The cross-sectional microstructure and elemental distribution of the filament were examined via scanning electron microscopy (SEM, SU-8010, Hitachi, Tokyo, Japan). Furthermore, differential scanning calorimetry (DSC) analysis was performed on filaments with three different solid volume fractions using a differential scanning calorimeter (DSC204F1, Netzsch, Selb, Germany) to assess their thermal behavior.

2.3. Study on Printing and Forming Process of Porous Titanium Scaffold Green Bodies

Porous Ti/ABS green bodies were fabricated using an FDM 3D printer (A1, Bambu Lab, Shenzhen, China) with the 45 vol% Ti/ABS composite filament. A 3D model of a porous scaffold measuring 50 mm in length, 50 mm in width, and 10 mm in height was designed and sliced using SolidWorks software (Version 2018, Dassault Systèmes, Vélizy-Villacoublay, France). The scaffold features a prismatic array structure with a strut diameter of 1000 μm and a pore diameter of 800 μm. Studies have shown that this pore size range is more suitable for bone tissue ingrowth [24,25]. During printing, the struts of the scaffold were alternately deposited at 0° and 90° between adjacent layers, creating an interconnected rectangular pore network. To ensure the pores remained completely open, no shells or top/bottom solid layers were added during printing.

To optimize the printing process parameters, a solid rectangular test specimen measuring 15 mm (X-direction) × 20 mm (Y-direction) × 5 mm (Z-direction) was specifically designed. The test specimen was printed with a 100% rectilinear infill pattern and a raster angle of ±45°. Initially, single-factor experiments were conducted to systematically investigate the effects of nozzle temperature (220 °C–260 °C), layer thickness (0.1 mm–0.3 mm), and printing speed (5 mm/s–25 mm/s) on the forming quality. After printing, a surface roughness tester (MarSurf M 300 C, Mahr GmbH, Göttingen, Germany) was used to measure the R_a of the top horizontal surface of three specimens (n = 3) for each condition. The measurement probe traveled along the longitudinal direction (Y-axis) of the specimen, which corresponds to a 45° angle with respect to the printed linear pattern, thereby assessing the spreading and overlapping smoothness of the printing paths. This top-surface roughness measurement specifically excluded the side-staircase effect caused by Z-axis layer-by-layer stacking; the interlayer bonding quality and the evolution of the side-staircase effect were independently evaluated qualitatively using SEM.

Using dimensional accuracy as the primary evaluation metric, a three-factor, three-level L₉(3⁴) orthogonal array was designed to identify the optimal printing process window, with the factor levels detailed in

Table 1. Orthogonal test factor levels table.

Level	A: Layer Height (mm)	B: Nozzle Temperature (°C)	C: Printing Speed (mm/s)
1	0.2	220	10
2	0.25	230	15
3	0.3	240	20

. Layer thickness, nozzle temperature, and printing speed were selected as the three core independent variables. The fourth column of the orthogonal array was assigned as a blank column to estimate the experimental error for subsequent significance testing via analysis of variance (ANOVA). To minimize systematic bias, the sequence of the 9 experimental runs was completely randomized. The actual dimensions of the printed specimens (n = 3 for each condition) along the X, Y, and Z axes were measured using a digital caliper. Both the absolute error (the difference between the actual and designed dimensions) and the relative error (indicating the degree of deviation) were calculated. Ultimately, the range and ANOVA evaluations were conducted based primarily on the absolute errors [26].

2.4. Study on the Debinding and Sintering Process of Porous Ti Scaffolds

To determine and optimize the debinding and sintering process, thermogravimetric analysis was performed on the green bodies using a simultaneous thermal analyzer (TGA/DSC3+, Mettler-Toledo, Zurich, Switzerland). This characterization aimed to precisely identify the thermal decomposition characteristics of the organic binder, providing a factual basis for optimizing the heating program. As shown in Figure 2, the polymer components in the green body undergo significant decomposition between 250 and 550 °C. Following these thermal analysis results, a detailed stepwise heating debinding and sintering profile was established, as illustrated in Figure 3. The process was conducted in a tube furnace (OTF-1200X-S, Hefei Kejing Material Technology Co, Ltd., Hefei, China) under a high-purity argon atmosphere. The specific thermal procedure is as follows: initially, the specimens were heated to 250 °C at a rate of 5 °C/min and held for 60 min. Subsequently, during the critical thermal debinding stage, the heating rate was reduced to 1 °C/min to reach 550 °C, followed by a 60 min isothermal hold to ensure the complete removal of organic components without inducing structural cracks or deformation. After debinding, the heating rate was increased to 5 °C/min to reach the final sintering temperature of 1100 °C, where the specimens were maintained for 120 min. This stepwise strategy, by precisely controlling the thermal stages, provides sufficient driving force for atomic diffusion and the formation of sintering necks between titanium particles while maintaining structural stability [20].

2.5. Characterization of the Porous Ti Scaffolds

The surface morphology of the porous titanium scaffolds before and after sintering was examined using scanning electron microscopy (SEM, SU-8010, Hitachi, Tokyo, Japan). Specifically, cross-sectional observations were conducted on the sintered scaffolds to evaluate the formation of sintering necks between particles. Phase composition was analyzed using X-ray diffraction (XRD, D8 ADVANCE, Bruker, Billerica, MA, USA) with Cu-Kα radiation, operating at a scanning range of 20° to 80° at a scan rate of 6°/min.

To evaluate dimensional changes, a digital caliper was used to measure the X, Y, and Z dimensions of the square porous titanium scaffolds (50 × 50 × 10 mm³) before and after debinding and sintering, from which the linear shrinkage was calculated. Furthermore, the Archimedes method (water displacement) was employed to measure the actual total porosity of the cylindrical scaffolds specifically prepared for mechanical testing (with designed porosities of 40%, 50%, and 60%) at both the green body stage and after debinding and sintering.

Quasi-static compression tests were performed on the sintered cylindrical scaffolds (average sintered diameter of 5.4 mm and height of 10.2 mm) using a biomechanical testing machine (PLD-5, Xi’an Lichuang, Xi’an, China). The testing procedure strictly complied with the GB/T 31930-2015 standard [27]. Three replicate specimens were tested for each porosity condition to evaluate the average mechanical properties. During the test, the compressive load was applied along the Z-axis (the printing stacking direction) at a constant loading rate of 0.5 mm/min until specimen failure. The ultimate compressive strength and elastic modulus were calculated from the recorded representative stress–strain curves.

3. Results and Discussion

3.1. Study on the Quality of Ti/ABS Filaments

Approximately 5 mg samples were placed in aluminum crucibles for DSC testing. Figure 4 shows the DSC curves of three Ti/ABS composite filaments with different solid contents (by volume); all exhibit exothermic peaks at about 40 °C and 120 °C. With increasing Ti content, the positions of the exothermic peaks and the areas of the crystallization peaks show no significant changes, indicating that Ti has little effect on the thermal properties of the composite. Therefore, filaments with different solid contents can be processed using the same nozzle temperature.

Figure 5 presents the cross-sectional SEM image and corresponding elemental maps of a Ti/ABS composite filament with a Ti volume fraction of 45%. The SEM image shows spherical Ti particles uniformly distributed in the ABS polymer matrix, with close particle–particle contact and no obvious pores or agglomeration; the depressions observed at the cross-section are brittle fracture marks produced during SEM sample preparation. The elemental maps further confirm that Ti is highly dispersed throughout the cross-section, and its distribution corresponds to the particle morphology seen in the SEM image. At the same time, the distributions of C, O and N are complementary to the Ti-rich regions, forming a continuous polymer phase network, which indicates tight integration between the organic and inorganic phases.

The MFR of the 45 vol% Ti/ABS composite filament was 1.2 g/10 min, whereas the MFR of the used ABS (PA757) was 1.8 g/10 min. The high density of the metal powder increases the overall density and melt viscosity of the composite system, thereby reducing flowability. Grafted maleic anhydride forms chemical coupling between the ABS matrix and the metal powder surface, enhancing intermolecular interactions and effectively reducing the melt viscosity, thus improving the flowability of the composite filament [28]. The MBS toughening agent, under stress, forms cavities via its soft core and promotes plastic deformation of the material, thereby absorbing impact energy and improving the toughness of the composite filament [29].

3.2. Study on the Printing Process of Ti/ABS Composite Filaments

3.2.1. Effect of Process Parameters on the Surface Quality of the Green Bodies

Figure 6 shows the relationship between nozzle temperature and surface roughness. Experimental data indicate that nozzle temperature regulates surface roughness nonlinearly: when the temperature is set to 230 °C, the surface roughness is minimal (5.19 ± 0.54 μm), at which the material is in the best molten state and interlayer bonding is tight. When the temperature rises to 260 °C, the roughness increases to 12.89 ± 1.71 μm, due to excessively low melt viscosity leading to insufficient interlayer wetting and thus deteriorated surface flatness.

Figure 7 shows the relationship between printing speed and surface roughness. Experimental results indicate that the influence of printing speed on surface roughness is less pronounced than that of nozzle temperature. At a speed of 10 mm/s, the surface quality is optimal, with a roughness of 4.98 ± 0.81 μm, where the molten material is extruded and deposited uniformly. When the speed increases to 25 mm/s, extrusion stability declines, causing larger interlayer positioning deviations and an increased surface roughness of 7.11 ± 0.44 μm. Although the roughness differences within the 5–25 mm/s range are small, speed control remains a key parameter for optimizing surface morphology.

Figure 8 shows the side-surface microstructure of green parts printed with different layer thicknesses. Significant differences exist within the material at different layer thicknesses. At a layer thickness of 0.1 mm (Figure 8a), the interlayer interfaces are not obvious, but the excessively small layer thickness causes extrusion deformation between layers, compromising the forming accuracy of the final print. When the layer thickness is increased to 0.2 mm (Figure 8b), the interlayer bonding is the densest and the forming quality is optimal. However, further increasing the layer thickness to 0.3 mm (Figure 8c) results in clear delamination of the layered structure.

A systematic study of key process parameters—nozzle temperature, printing speed, and layer thickness—shows that each has a significant effect on surface roughness. Reference [30] indicates that a relatively low nozzle temperature favors improved surface quality for ABS; the minimum surface roughness observed at 230 °C in this study is consistent with that finding. As temperature increases, the flow behavior of the ABS polymer changes, producing a nonlinear trend of initial improvement followed by deterioration in surface quality. Regarding printing speed, reference [31] reports a nonlinear influence on surface quality; in this study the best surface quality was achieved at 10 mm/s, while higher speeds reduced extrusion stability and deposition accuracy, increasing surface roughness. Layer thickness has the most pronounced effect on side-surface morphology: this study found the best surface quality at 0.2 mm layer thickness, whereas reducing the layer thickness causes slight material overflow and increasing it amplifies the staircase effect, reducing interlayer bonding quality.

3.2.2. Effect of Process Parameters on the Forming Accuracy of the Green Bodies

The dimensional deviations obtained from the L₉(3⁴) orthogonal experiments are presented in

Table 2. L₉(3⁴) Orthogonal table and test results.

Option	Level			Absolute Error (mm)			Relative Error (%)
	A	B	C	X	Y	Z	X	Y	Z
1	1	1	1	0.06	0.14	0.12	0.4	0.7	2.4
2	1	2	2	0.17	0.13	0.14	1.13	0.65	2.8
3	1	3	3	0.18	0.17	0.17	1.2	0.85	3.4
4	2	1	2	0.11	0.09	0.22	0.73	0.45	4.4
5	2	2	3	0.23	0.27	0.1	1.53	1.35	2
6	2	3	1	0.19	0.15	0.14	1.27	0.75	2.8
7	3	1	3	0.15	0.1	0.31	1	0.5	6.2
8	3	2	1	0.18	0.23	0.35	1.2	1.15	7
9	3	3	2	0.14	0.21	0.33	0.93	1.05	6.6

. To comprehensively evaluate the forming accuracy, both the absolute errors and the relative errors (calculated based on the design dimensions of X = 15 mm, Y = 20 mm, and Z = 5 mm) are included.

(1)

Range analysis: A range analysis was performed on the experimental results to determine the relative importance of each factor on the errors and thus obtain the optimal combination of factor levels. The final analysis is shown in

Table 3. Extreme difference analysis.

	Parameter	A	B	C
X direction	k1	0.137	0.106	0.143
	k2	0.177	0.193	0.140
	k3	0.156	0.170	0.186
	Range R	0.040	0.087	0.046
	Sorting	3	1	2
Y direction	k1	0.156	0.110	0.173
	k2	0.170	0.210	0.143
	k3	0.180	0.186	0.190
	Range R	0.100	0.100	0.046
	Sorting	3	2	1
Z direction	k1	0.144	0.216	0.203
	k2	0.153	0.196	0.230
	k3	0.330	0.213	0.194
	Range R	0.187	0.020	0.036
	Sorting	1	3	2

. Here k denotes the sum of the errors corresponding to a given level of a factor divided by the number of observations at that level (i.e., the average error for that level); the range R is the difference between the maximum and minimum k values. In the X and Y directions, the order of influence (from largest to smallest) is B (nozzle temperature) > C (printing speed) > A (layer thickness). The nozzle temperature has the largest range, indicating it plays a decisive role in the dimensional accuracy on the XY plane; a nozzle temperature of 220 °C helps achieve smaller planar dimensional errors. For the Z direction, the order of influence is A (layer thickness) > C (printing speed) > B (nozzle temperature). The range for layer thickness is much larger than for the other factors, showing its effect is significantly greater than that of printing speed and nozzle temperature and making it the most critical parameter for controlling Z-direction accuracy. A layer thickness of 0.2 mm produced the smallest Z-direction dimensional deviation, indicating that a moderate layer thickness promotes uniform material deposition and stable interlayer bonding.

(2)

Analysis of variance (ANOVA): To further verify the statistical significance of each factor’s influence, ANOVA was performed on the experimental data at the significance level corresponding to an F critical value = 9. The results are shown in

Table 4. Analysis of Variance.

		A	B	C	Error
X direction	Sum of squares (SS)	0.0024	0.0121	0.0041	0.0011
	Degrees of Freedom (DF)	2	2	2	2
	Mean Square (MS)	0.0012	0.0061	0.0021	0.0005
	F	2.24	11.26	3.83
	Significance	Not Significant	Significant	Not Significant
Y direction	Sum of squares (SS)	0.0008	0.0166	0.0036	0.0096
	Degrees of Freedom (DF)	2	2	2	2
	Mean Square (MS)	0.0004	0.00830	0.00180	0.0048
	F	0.09	1.83	0.40
	Significance	Not Significant	Not Significant	Not Significant
Z direction	Sum of squares (SS)	0.0662	0.0007	0.0022	0.0067
	Degrees of Freedom (DF)	2	2	2	2
	Mean Square (MS)	0.0331	0.0004	0.0011	0.0033
	F	9.88	0.10	0.33
	Significance	Significant	Not Significant	Not Significant

. In the X direction, the F value for nozzle temperature (B) exceeds the critical value, indicating a statistically significant effect, whereas the effects of layer thickness (A) and printing speed (C) are not significant. In the Y direction, the F values for all factors are below the critical value and do not reach statistical significance. In the Z direction, the F value for layer thickness (A) is well above the critical value, indicating a highly significant effect.

Nozzle temperature is the primary factor controlling XY-plane accuracy, while layer thickness is the most critical parameter determining Z-direction stacking accuracy. This conclusion is consistent with the FDM process mechanism: nozzle temperature directly affects the material’s melt state and the bond strength of the extruded filament, thereby determining the accuracy of in-plane contours, whereas layer thickness is directly related to Z-axis displacement and is the main cause of interlayer stacking error. To optimize overall dimensional accuracy, the optimal parameters for different directions must be balanced. Since A1 (0.2 mm) is the optimal or near-optimal choice in all three directions and is crucial for Z-direction accuracy, layer thickness should be prioritized at 0.2 mm. For nozzle temperature and printing speed, the optimal combination for the XY plane is B1C2, and for the Z direction is B2C3; however, considering the overwhelming influence of nozzle temperature on planar accuracy and the balanced performance of C2 (15 mm/s) across multiple directions, A1B1C2 (0.2 mm, 220 °C, 15 mm/s) is adopted as the overall optimal process parameter set for subsequent fabrication of porous scaffold green bodies.

3.3. Macro- and Micro-Morphology and Elemental Analysis of Porous Titanium Scaffolds

The macroscopic morphologies of the porous titanium scaffold green body and the subsequently sintered scaffold are shown in Figure 9a and Figure 9b, respectively. The macroscopic geometric features remained essentially unchanged before and after sintering; the sintered scaffold exhibited a pronounced metallic luster, continuous strut diameters, distinct pores, and no structural collapse.

Due to the complete removal of the organic binder during debinding and the densification of metal particles during sintering, the scaffold experienced significant volumetric shrinkage, which exhibited a certain degree of anisotropy. Measurements of the initial 50 × 50 × 10 mm³ square scaffolds revealed that the linear shrinkages in the X, Y, and Z directions were 30.43 ± 0.78%, 30.07 ± 0.75%, and 24.13 ± 0.44%, respectively. The shrinkages in the XY plane were highly consistent and significantly higher than that in the Z direction. This directional difference is primarily attributed to the orientation effect of the polymer chains in the XY plane caused by extrusion during the FDM process, which leads to greater planar retraction upon heating during debinding. In contrast, shrinkage in the Z direction is relatively lower due to the hindrance of the stacked layer interfaces and interlayer frictional constraints. The quantification of this shrinkage characteristic provides crucial data support for scaling and dimensionally compensating the CAD models of porous scaffolds in future applications. Furthermore, the excellent retention of the macroscopic morphology demonstrates that the adopted debinding and sintering process effectively achieves the metallurgical bonding of titanium powder while maintaining the structural stability of the complex porous architecture.

Figure 10a and Figure 11a respectively present the surface and side microstructures of the porous titanium scaffold green bodies; Figure 10b and Figure 11b show the surface and side microstructures of the scaffolds after sintering. The scaffold struts are fully formed and continuous, with no obvious stringing, filament breakage, or warping, and the geometric structure closely matches the design model. After sintering, titanium powder particles diffused and fused under high temperature to form distinct sintering necks [32]. At the same time, the layered structure produced via the FDM process remained stable after sintering, with dense interlayer bonding and no cracking or collapse, which confirms the compatibility and appropriateness of the printing parameters and sintering process and ensures adequate support during forming. In addition, the scaffolds exhibited uniform dimensional shrinkage overall, an inevitable result of binder removal and metal particle densification that conforms to fundamental sintering kinetics and further validates the scientificity and effectiveness of the debinding–sintering parameter settings. This combination of macroscopic pore connectivity and microscopic surface roughness is expected to provide an ideal spatial environment for cell migration and nutrient transport.

To further examine the bonding state between particles, Figure 12 shows the SEM images of the fractured cross-section of the sintered scaffold. At different magnifications, the interconnected internal struts are clearly visible, and the fracture surface exhibits a characteristic granular morphology. High-magnification observation reveals the formation of distinct sintering necks between adjacent spherical titanium particles, where the particle interfaces have fused. This indicates that effective material migration and metallurgical bonding were achieved during the sintering process at 1100 °C, thereby ensuring the structural integrity and stability of the porous scaffold.

The EDS maps in Figure 13 show a uniform distribution of Ti on the scaffold surface. Because the ABS powder contains C and O with mass fractions of 84.71% and 3.4%, respectively, and the other additives are polymeric (mainly composed of C and O), the porous titanium scaffold green bodies contained C and O up to 53.62% by mass (Figure 13a). During debinding, these polymers were removed, and the Ti mass fraction on the surface of the sintered porous titanium scaffold increased to 90.11% (Figure 13b), while C and O were 5.3% and 4.59%, respectively. Compared with the pre-sintered state, the C mass fraction decreased markedly, though a small amount of residual carbon remained. X-ray diffraction (XRD) analysis of the sintered porous titanium scaffolds (Figure 14) shows that the material is primarily composed of Ti and that a TiC phase is present; this finding corroborates the detection of C by EDS and suggests that TiC may have formed from a reaction between titanium and residual carbon from the debinding process.

The porous titanium scaffolds prepared in this study exhibit intact macro- and micro-morphologies, confirming Bankapalli et al.’s [20] conclusion that high-fidelity conversion from polymer composite green bodies to metallic structures can be achieved by precise control of the debinding–sintering process. Microscopically, the internal interlayer bonding is dense with no visible defects, and the surface shows characteristic sintering-neck connections. This contrasts with the smooth molten surfaces of porous titanium produced via EPBF reported by Liu et al. [16], reflecting the differences between indirect forming and direct forming metal additive manufacturing technologies. Further elemental analysis indicates that the scaffold surface is dominated by titanium; XRD phase analysis confirms Ti as the main phase and detects TiC, formed from reactions between residual carbon during binder decomposition and titanium. TiC has been reported to exhibit good biocompatibility with osteogenic cells [33], which is favorable for the scaffold’s potential applications.

3.4. Mechanical Properties of the Porous Titanium Scaffolds

To accurately evaluate the physical characteristics of the porous scaffolds, the actual porosity of the sintered cylindrical specimens was first measured using the Archimedes method. The results indicate that for scaffolds with designed porosities of 40%, 50%, and 60%, the actual total porosities after debinding and sintering reached 47.15%, 56.78%, and 65.94%, respectively. The measured porosities are higher than the designed values, which is primarily attributed to the complete volatilization of the polymer binder during the debinding process and the retention of microscopic interstitial pores between the powder particles due to incomplete densification during the sintering stage.

The stress-strain curves, along with the corresponding compressive strength and elastic modulus, are shown in Figure 15 and Figure 16. In the initial elastic region, the three curves nearly coincide, indicating similar elastic responses at small strains. As strain increases, all three scaffolds pass through a linear elastic stage, reach a peak stress, and then the stress rapidly declines. The scaffold with a designed porosity of 60% reached a peak stress of 26.8 MPa at 3.2% strain; the designed 50-porosity scaffold reached 55.4 MPa at 4.6% strain; and the designed 40-porosity scaffold reached 72.7 MPa at 5.8% strain. This trend indicates that decreasing porosity significantly increases both the load-bearing capacity and the deformation sustained before failure. The average compressive strengths for the 40%, 50%, and 60% designed porosity scaffolds are 68.3 MPa, 52.1 MPa, and 25.7 MPa, respectively, and the average elastic moduli are 1.65 GPa, 1.56 GPa, and 1.21 GPa.

To evaluate the clinical applicability of these porous titanium scaffolds for orthopedic implants, their mechanical properties were systematically compared with literature data on natural human bones. An ideal bone implant must possess mechanical properties that match those of the host bone to avoid the “stress shielding” effect caused by an excessively high modulus. According to reported literature, the ultimate compressive strength of dense human cortical bone typically ranges from 100 to 230 MPa, with an elastic modulus of 10 to 30 GPa; in contrast, human cancellous bone generally exhibits a compressive strength of 2 to 20 MPa and an elastic modulus of 0.1 to 2.0 GPa [34].

The experimental results demonstrate that the high-porosity titanium scaffolds fabricated via the FDM process successfully and significantly reduced the inherently high elastic modulus of dense pure titanium (approximately 110 GPa). The elastic moduli of the three types of scaffolds with actual porosities (47.15%, 56.78%, and 65.94%) ranged from 1.21 to 1.65 GPa, all of which fall well within the range of natural cancellous bone. Furthermore, their compressive strengths (25.7–68.3 MPa) were slightly higher than those of cancellous bone, positioning them between the strength ranges of cancellous and cortical bone. This indicates that the porous titanium scaffolds prepared using this process achieve an excellent balance between load-bearing capacity and a low elastic modulus, demonstrating promising potential for bone defect repair applications.

4. Conclusions

This paper proposes a method for fabricating porous metallic titanium scaffolds by combining fused deposition modeling (FDM) with a debinding–sintering process. The main conclusions are as follows.

(1)

A 45 vol% Ti/ABS composite filament prepared via solution blending has a uniform diameter and meets FDM processing requirements. SEM of the filament cross-section shows an even distribution of titanium powder in the polymer matrix without agglomeration. Although the composite filament’s melt flow rate (MFR = 1.2 g/10 min) is lower than that of pure ABS, the maleic anhydride grafting agent enhances intermolecular interactions in ABS and effectively compensates for the flowability loss due to the high titanium content. DSC curves further confirm that the addition of titanium powder does not significantly alter the material’s thermal properties, supporting process stability.

(2)

Nozzle temperature, printing speed, and layer thickness jointly determine the surface quality and forming accuracy of the green bodies. Optimal printing was achieved at a nozzle temperature of 220 °C, a printing speed of 15 mm/s, and a layer thickness of 0.2 mm, where the material melted uniformly, extrusion was stable, interlayer bonding was dense, and surface roughness was minimized. These optimized parameters provide key process assurance for subsequent fabrication of structurally intact porous scaffold green bodies.

(3)

The debinding–sintering post-treatment successfully converted Ti/ABS composite green bodies into porous scaffolds composed primarily of metallic titanium. A stepwise heating sintering strategy prevented rapid polymer decomposition and structural collapse, preserving the scaffolds’ macroscopic porosity and strut diameter continuity, while forming distinct sintering necks between titanium particles at the microscopic level. EDS analysis shows a titanium mass fraction of 90.11% on the sintered scaffold surface, and XRD phase analysis further confirms titanium as the main phase with a small amount of TiC formed from reactions between residual carbon and titanium. This process enables a reliable conversion from polymer composite green bodies to compositionally controlled, structurally intact pure metal porous scaffolds.

(4)

Compression tests on scaffolds with designed porosities of 40%, 50%, and 60% (corresponding to actual measured porosities of 47.15%, 56.78%, and 65.94%) show that mechanical properties are closely related to porosity: both compressive strength and elastic modulus decrease as porosity increases. The average compressive strengths are 68.3 MPa, 52.1 MPa, and 25.7 MPa, and the average elastic moduli are 1.6 GPa, 1.5 GPa, and 1.2 GPa, respectively. The scaffolds’ elastic moduli (1.2–1.6 GPa) match that of human cancellous bone (0.1–2 GPa), while their compressive strengths (25.7–68.3 MPa) are significantly higher than typical cancellous bone (5–10 MPa) and lie between cancellous and high-strength cortical bone. This “high strength with moderate modulus” combination enables the scaffolds to provide sufficient initial mechanical support while the matched modulus helps alleviate stress shielding, creating a favorable mechanical environment for bone integration.