The additive manufacturing method (AM) is a relatively new method for producing solid structures, including metallic removable and fixed partial dentures. ¹ Although the traditional casting method has been employed to create cobalt–chromium (Co–Cr) frames for many years, it has been superseded by AM technology. ² Metallic frameworks can be produced more simply by selective laser melting (SLM) AM technology with appropriate computer-aided design (CAD). ³ , ⁴ , ⁵ , ⁶ Topology optimization and topology modification are computational methodologies employed to generate an optimal arrangement of materials inside a designated design space, considering certain loading circumstances. The primary objective is to minimize the structure's overall weight while simultaneously satisfying other functional criteria. ⁷ , ⁸ , ⁹ Another benefit of the technology is that it enables the creation of finished products with virtually infinite patterns. CAD geometries and features may be manufactured using metal laser sintering (MLS) devices, including undercuts, spaces, and intricate interior geometries. ⁴ , ¹⁰ , ¹¹ The topology-modified designs can be used to create hollow structures, making it possible to construct lighter things while drastically increasing productivity and reducing the quantity of powder needed during manufacturing. ¹² , ¹³ Previous studies analyzed the mechanical characteristics of MLS-produced Co–Cr specimens with internal changes. They indicated that a substantial quantity of metal material could be preserved by modifying the internal geometry, production time could be decreased, and, importantly, the new internal architecture geometry yet complies with the mechanical characteristics required for the International Organization for Standardization (ISO Type 5). ¹⁴ Creating voids or modifying topology in metals with high melting points is extremely difficult, if not impossible when the conventional fabrication methods are used. Even when special chemical media and techniques are used, this can create uncontrolled shapes and sizes of the voids. ¹⁵ As an alternative, metallic objects with pores might be created by AM with no restrictions on the temperature at which they melt. In addition, the pore ratio and structure in an object constructed with AM with an interior porosity may be modified to accommodate the functionally categorized performance of design requirements. ¹⁶ , ¹⁷ Challis et al. stated that Ti6Al4V geometry optimal lattice arrangement manufactured using the SLM technique had superior rigidity and strength compared to analogous composites. ¹⁷ Takezawa et al. utilized topology optimization to determine the optimal lattice network geometry to reduce the performance loss caused by laser powder-removal cavities in the electron beam melting) technique. ¹⁸ Liu et al. discussed the fatigue properties of a CP-Ti lattice network that contains topology-optimized and rhombic cell structures manufactured by SLM. ¹⁹ Other previous studies had reported the benefits of lattice network internal geometry when compared with randomized voids distributed in the printed objects, where the lattice design can avoid local buckling of the face sheets and decrease transmitted stress under impact stresses, showing that the lattice-based structures constructed from stainless steel and SLM technique demonstrated superior rigidity, stress, and energy absorption capacity compared to unevenly shaped and distributed voids. ²⁰ , ²¹ , ²² All aforementioned studies have emphasized the lattice-built internal structure, and its superiority in the mechanical properties improvement when compared with other voids shape, but they neglected the role of these shapes in crack propagation redirection, consequently increasing the stress needed to induce failure in the 3D-printed objects compared to the compact samples. ²³ , ²⁴ As a result, this study aimed to evaluate the influence of three types of internally custom-created geometric voids on the weight reduction and the mechanical characteristics of the 3D-printed Co–Cr samples. The weight, flexural bending strength, and bending modulus of elasticity of the laser-sintered Co–Cr frameworks fabricated by the direct MLS (DMLS) method was evaluated. The study hypothesis is that incorporating custom-designed voids inside the metallic denture base can reduce the 3D-printed Co–Cr prototype's weight, and enhance its mechanical properties.

In the current experimental study, 50 samples were designed using the AutoCAD 2017 software program (AutoDesk Co., San Francisco, USA) to incorporate geometries of self-reinforced custom-designed voids inside the test samples. The samples were fabricated with Co–Cr metal powder (Adentatec, GmbH, Germany) and manufactured using the DMLS method (DMLS) technology by the laser melting machine (Riton, dual-150, Guangdong, China) with fabrication parameters as listed in Table 1. According to the American Society for Testing and Materials (ASTM standard E9) (2010), each subgroup had five samples to be tested. ²⁵ The fabricated samples were divided into four groups. The first group (Group 1, n = 5) was compact without incorporating any internal design representing the control group and the second group was fabricated basically, with spherical internal voids. The third group was manufactured by incorporating ovoid voids as a basic shape. The fourth group was manufactured with diamond voids as a basic shape of the internal design. Each one of the last three groups (except for the control group) was subdivided into three subgroups (n = 5) according to the reduction percentage of the volume (10%, 15%, and 20%).{Table 1}

The samples were fabricated (1 mm thickness, 3 mm width, and 31 mm length) according to the noncasted metallic dental restoration specification of ISO standardization (22674:2016), ²⁶ where the control group samples were printed as compact samples without an internal modification, and the rest of the samples were fabricated with the exact dimensions externally, and with internally incorporated voids. The samples were weighted to find the amount of reduction per mass of the samples using a digital balance (with 0.000 g precision; PG 503-S MonoBloc inside, Mettler Toledo Ltd, Switzerland). Then, the samples were tested using a universal testing machine (GESTER International Co., LTD, Fujian, China) with a 1.5 mm/min crosshead speed. The data gathered were subjected to the following equation to find the ultimate flexural strength (UFS): ²⁷

Flexural strength (σ) = 3PL/2wt ²

Where P is the applied force value, L is the testing length from one supporting pin to the other, which is 20 mm out of the 25 mm length of the sample, w is the width of the sample, and t is the thickness of the sample.

The calculation of the flexural modulus of elasticity (apparent elastic modulus E _a) was conducted according to the following equation: ²⁶

Apparent elastic modulus E _a= PL ³/4dwt ³

Where P is the applied load, L is the length between the supporting pins = 20 mm, d is the highest deflection value, w is the width of the sample, and t is the thickness of the sample.

Incorporated internal geometries

The geometries integrated within the core of the 3-point bending samples with closed cell configuration. One of the connectors was extended to the end of the samples to achieve the enclosed cell printing efficiently to serve as a channel through which the nonsintered alloy particles can exit after the manufacturing procedure. ²⁸ The first group was devoid of any internal geometries, where the created object was compact and used as a control group. In the second group, the incorporated geometry was spherical, where the spheres created ranged from 200 to 500 μm in diameter according to the percentage of volume reduction, and they were connected with tunnels that had ≈ 50%–75% diameter to the diameter of the sphere; also, the length of the cylindrical tunnels among the spheres was ranging from ≈ 50% to 125% to the diameter of the sphere according to the volume reduction percentage. In the third group, ovoid-shaped voids were created and arranged in transverse orientation with 200–500 μm lesser diameter and 400–1000 μm greater diameter. The tunnels also connected the ovoid voids using the lesser diameter of the oval shape as a reference, with the same connector diameter and length ratios used in the second group. In the fourth group, diamond-shaped voids were designed with symmetrical vertical and horizontal diameters of 200–500 μm, and the cylindrical tunnels connected them with the same diameter and length ratios used in the second group, which varied according to the reduction percentage. For all samples designed, the created internal geometries were located on one level in the core of the specimen. The four sides of the specimen had at least 0.5 mm thickness. In contrast, the upper and lower surfaces had a thickness range of 0.5 ± 0.05 mm according to the reduction percentage. The illustration of the designs of the four groups is shown in Figure 1. Figure 1

AutoCAD designs of the four groups. (a) Compact design of the 3-point bending test sample, (b) Sphere-based design of the 3-point bending test sample, (c) Ovoid-based design of the 3-point bending test sample, (d) Diamond-based design of the 3-point bending test sample.

The current study theorized that one of the mechanical properties of the 3D-printed samples could be enhanced by incorporating internal geometries at a certain percentage of reduction. Subsequently, the weight would be reduced in addition to the time of printing and cost. To achieve this goal, three types of internal geometry were incorporated inside the samples to be tested. The weight, UFS (stress at rupture of the samples), and flexural modulus of elasticity were estimated at different degrees of reduction percentages.

Weight reduction results

The statistical analysis of the means and standard deviation of the weight is shown in Table 2. The results revealed that there was a significant decrease in weight for Groups 2–4 in comparison to the weight of the compact samples in Group 1 at a reduction percentage of 10% per volume, the reduction of mass at this percentage estimated about (20.4%–21.5%). Results of the weight reduction at 15% per volume showed that there was a significantly decreased weight of Groups 2–4 when compared to Group 1 compact samples; the percentage of weight reduction at this level was (26.9%–29.3%). Results of the weight reduction at a reduction of 20% per volume showed that the weight mean of Group 1 was significantly higher than those of Groups 2–4, and the samples were reduced by (31.7%–33.2%) per weight. Furthermore, results showed insignificant differences among the groups' means of weight reduction with the same reduction percentage per volume in Groups 2–4.{Table 2}

Flexural strength results

The mean and standard deviation of the flexural strength are shown in Figure 2. The results revealed that the UFS of Group 2 at a reduction (10%) per volume (1821.989 ± 60.103 MPa) was significantly higher than the compact samples, and the rest of the groups, followed by insignificantly different means of Group 1 UFS mean (1675.456 ± 18.360 MPa), and UFS mean of Group 2 at 15% reduction per volume (1674.364 ± 36.207 MPa). The results showed a significant declination of the UFS mean of Group 2 after a 20% reduction per volume (1465.1200 ± 42.505 MPa). The means of UFS in other groups showed significant differences at different reduction percentages of volume in Groups 3 and 4, where UFS mean Group 3 at 10% reduction (1173.262 ± 10.346 MPa), which was lower than 20% reduction of Group 2, and higher than the UFS means of the same group at 15%, and 20% reduction per volume (1128.926 ± 7.242 MPa and 996.414 ± 13.179 MPa), respectively. In Group 4, the UFS mean at a 10% reduction per volume (1131.952 ± 5.970 MPa) was insignificantly different from Group 3 at a 15% reduction per volume, the mean of UFS at 15% reduction (1043.510 ± 50.215 MPa) was significantly lower than the rest of groups at different percentages of reduction, and only significantly higher than UFS mean of the same group at 20% reduction per volume (928.040 ± 76.678 MPa). Figure 2

Bar graph of the statistical analysis of the ultimate flexural strength mean and standard deviation. Bar names with identical letters indicate insignificant differences. UFS: Ultimate flexural strength.

This investigation analyzed the effect of internal structural design on the mechanical properties of DMLS-fabricated Co–Cr specimens. The results revealed that the internal structural design with sphere-based hollow structures at 10% volumetric reduction reduced the specimens' weight significantly and had no effect on their modulus of elasticity. Simultaneously, there was a significantly increased flexural strength of the same reduction percentage of the same group compared to the compact samples of the control group. Therefore, the hypothesis was accepted for the mass reduction, flexural strength, and elastic modulus parameters.

Weighting test of the 3D-printed samples

Results revealed a significant reduction in weighted samples compared to the compact samples. Furthermore, the weights of the tested samples were significantly different at different reduction percentages. The internally integrated micro-voids can decrease the mass of the additively manufactured objects due to the apparent reduction in density, consequently reducing the cost and period of the printing procedure. ¹³ , ²⁸ , ²⁹ These results are consistent with previous studies that confirmed a significant mass reduction when hollow structures are incorporated inside the printed object relative to the reduction in density of the item. ³⁰ , ³¹

Flexural strength of the 3D-printed samples

Results of the flexural strength revealed that the flexural strength was increased in the Group 2, where a sphere-based hollow structure was created at a 10% reduction per volume. The rigidity of the objects decreased with increased volume reduction and different shapes of the internal geometries. The results presented in this paper demonstrated that designed spherical voids offer a substantial potential for strengthening additively manufactured components and frameworks. An increase in fracture resistance, particularly considering that this improvement resulted from a minor geometrical modification, possibly because of crack arrest and termination, whereas the corresponding conventional geometries fractured catastrophically at relatively modest driving forces, the modified geometries resisted more external stresses. This characteristic of crack arrest may be a crucial component of the designing strategy, permitting secure, measurable deformation of an object before its catastrophic breakdown. ²⁹ , ³⁰ , ³¹ Adding micro-voids to a topological strengthening of the materials can offer another freedom level in object design. The capacity to add designs broadens the design characteristics for targeted strengthening. It enables different approaches for targeted strengthening with designed voids, and this would be useful in predicting the directions of crack growth, creating nonlinear propagation of the fracture. ³² , ³³ The planes of voids could generate directed interfaces to control the crack growth and enhance the strength of the material. ³⁴ The increased values of the flexural strength in Group 2 at a 10% reduction per volume, without compromising the elastic modulus values, and the insignificant difference of the same group at a 15% reduction per volume indicate that the spheres designed were capable of absorbing the stress-energy, and, distribute this energy horizontally, in other words, the crack growth was inhibited by the presence of angle-free internal void, where the stress-induced could not concentrate at any point at the metallic shell surrounding the voids. Because its percolated boundary has a rounder junction, which induces less stress concentration and increases resistance to bending deformation, the spherical structure has noticeably higher stiffness and strength due to the higher resistance to bending deformations compared to the compact samples. This effect of voids could also be increased because the tunnels connecting the voids used in the current study may further redirect the crack growth in a horizontal direction. The crack propagation could also be dissipated and split into more than one division by the presence of a sphere angle-free metallic shell around the spherical micro-voids. ³⁵ , ³⁶ Furthermore, the algorithmic rise of the internal gaseous pressure in a spherical void in solid objects could contribute to higher levels of flexural strength. ³⁷ , ³⁸ The results of the current study were totally in agreement with previous studies' findings. ³⁶ , ³⁹ It is partially consistent with previous studies that reported a relatively superior flexural strength induced by better energy absorption after incorporating internally arranged voids. ²⁰ , ²¹ , ⁴⁰ , ⁴¹ The current study showed a significant difference in flexural strength among groups 2–4 with different geometries. The difference could be the result that these shapes have different absorption capacities of the external stresses induced in the three-point bending test, where spherical shapes are devoid of angles that act as a center for stress concentration in the case of Group 4 and long linear courses that are perpendicular to the external force applied in case of Group 3, where the ovoid shapes could act in the same way of ellipsoidal shapes in other studies. ⁴² These results agree with other studies that reported significant differences among different internal geometries incorporated inside the 3D-printed objects due to different energy absorption levels and distribution. ¹⁵ , ³⁵ , ⁴² The current results were inconsistent with other studies that reported insignificant differences among different geometries or better absorption of geometries designed with certain angular architects. ⁴³ The results also revealed that there was a gradual regression in the values of flexural strength and elastic modulus, starting from Group 2 at a 10% reduction per volume, which had the highest means of flexural strength among other groups and reduction ratio, to the least means the flexural strength of Group 4 at 20% reduction per volume. These results indicate that the regression is the factor of reduced density of the printed objects, where the reduced density corresponded to larger voids per volume, which act as focal spots for crack growth. Furthermore, the density of the topologically modified parts plays an important role in flexural strength where at specific high levels of density (mass) reduction, the hollow shapes can reduce the flexural strength of the printed objects so that the density reduction is a critical value and can counteract the intended purpose of its integration at a certain level. ¹⁵ , ²⁹

Elastic modulus

The findings demonstrated that the average elastic modulus gradually decreased in relation to a gradual decrease in the strength and an increase in the reduction percentage within each subgroup. Specifically, the highest mean elastic modulus was observed in the sphere-based structure with a 10% volume reduction. Conversely, the ellipse-based and diamond-based geometries with a 20% volume reduction exhibited the lowest mean elastic modulus. These results indicate a clear correlation between the gradual decrease in mean elastic modulus and the gradual decrease in flexural strength, particularly when the strain of the measured objects is nearly identical. This occurrence is typically characterized by a high degree of brittleness and significantly reduced flexibility. These properties can be attributed to the lower grain dislocation observed in the tested samples.

Furthermore, the samples were printed in a transverse orientation relative to the force exerted throughout the testing process. This further contributes to the increased deflection values of groups with lower stiffness, where the deflection values are inversely proportional to the modulus of elasticity. ⁴¹ , ⁴³ , ⁴⁴

Printing trueness tests

The radiographs of the 3-point bending samples illustrated that the printing of the proposed hollow structures was proved by the presence of the intended geometries inside the test samples. In the current study, the AutoCAD program was used for designing the samples, and the Materialise Magics program was used as a slicing program. These two programs can properly transfer precise dimensions as digital information to the 3D-printing machine program to produce printed items with high precision and accuracy of the proposed dimensions. ²⁹ , ⁴⁵ , ⁴⁶ , ⁴⁷

The analysis of the fractographs of the printed hollow shapes indicated that all geometries were efficiently created using the DMLS technology. Furthermore, examining the cross-sections of the three geometries revealed that the geometrical structures with round forms in the sphere and ellipse were produced satisfactorily, with no or minimal leftover residues. On the other hand, the diamond-shaped objects showed more significant variations, particularly in the upper and lower regions near the sharp inclinations. This suggests that hollow shapes with sharp angles may have lower print quality compared to shapes with circular bases and tops. These findings show that the parameters employed in the DMLS procedure and the selected relocation position were suitable for producing the hollow components. ⁴⁷ , ⁴⁸ Atomic force microscopy microscopic images, SEM images, and XRD test results showed a dendritic surface production of the printing procedure, which was insignificantly different qualitatively and quantitatively among the tested surfaces of the control samples. The walls of the hollow shapes are probably because of the same printing method (DMLS) and the same powder alloy used to fabricate all samples, and especially there was no heat treatment involved in the current study. ² , ⁷ , ¹¹ , ⁴⁸ , ⁴⁹

One of the study limits is that the real-time printing of each specimen could not be estimated. Therefore, the statistical results were not correlated to the printing time factor; also, the thickness variation of the samples was not considered in the current study. Clinical evaluation of the results gathered in this study should be conducted before applying the current research findings.

Within the limitations of the current experimental work, we can state that the graded volumetric reduction significantly reduced the mass of the printed objects, producing items with less density at all levels (10%, 15%, and 20%) of reduction per volume used in this study. Incorporating a sphere-based hollow structure at a 10% reduction per volume can reduce the weight and increase the flexural strength of the 3D printed objects without compromising the elastic modulus property with evident printing quality and quantity of the hollow structures.

Financial support and sponsorship

The research is self-funded.

Conflicts of interest

The authors of this manuscript declare that they have no conflicts of interest, real or perceived, financial or non-financial in this article.