Statement of problem
Selective laser melting (SLM) Ti-6Al-4V has been used for removable partial dentures, but the impact of SLM Ti-6Al-4V build orientation is not evident.
The purpose of this in vitro study was to investigate the microstructure and properties of SLM Ti-6Al-4V clasps with different build orientations compared with cast Ti-6Al-4V clasps.
Material and methods
Forty-eight clasps were made from Ti-6Al-4V alloys—by SLM with 3 different build orientations (SLM0, SLM45, and SLM90) and cast (CAST) as a control. The microstructure was investigated by using a metallographic microscope and a confocal laser scanning microscope. The fit and surface roughness of the clasps were measured, and the physical properties were evaluated. In addition, the von Mises stresses in the clasps were calculated by finite element analysis. All specimens were then subjected to insertion and removal tests in artificial saliva to model 5 years of clinical use. After these tests, 3-point bend tests were used to analyze the fracture surface of the clasp arms, which were observed by using a scanning electron microscope. All data were statistically analyzed (α=.05).
The microstructure of the Ti-6Al-4V specimens was a comixture of α+β phases. In addition, growth directions of β grains were approximately parallel to the build orientation, with acicular α grains present between β grains. SLM0 and SLM45 had significantly higher roughness than SLM90. Even though the fit was inferior to that of SLM90, SLM0 and SLM45 still performed better than cast specimens ( P <.05). The finite element analysis showed that the maximum von Mises stress was located on the middle part of the retainer arms and that the values of the 0.50-mm undercut clasps were much lower than the elastic limit. In addition, the decrease of retentive force in SLM90 clasps was less than that of the CAST group ( P <.05). CAST clasps showed brittle fracture, whereas all SLM clasps showed ductile fracture.
The microstructure of SLM Ti-6Al-4V affected the properties of clasps by changing the anisotropy of specimens. Among the tested groups, SLM90 clasps had the best fit, the lowest surface roughness, and the best fatigue resistance. Furthermore, SLM Ti-6Al-4V clasps could be engaged into 0.50-mm undercuts.
SLM Ti-6Al-4V clasps fabricated at 90 degrees to the long axis showed the best properties. These results may guide technicians in fabricating SLM Ti-6Al-4V RPDs.
Removable partial dentures (RPDs) are still widely used to treat partial edentulism because they are widely indicated and their cost is lower than that of fixed or implant-supported prostheses. Because RPD clasps experience repeated flexure, fatigue has been reported to be one of the main complications of RPDs. A 25-year retrospective study concluded that clasp fracture was the first damage observed in RPDs.
Selective laser melting (SLM) Ti-6Al-4V has been used in RPD frameworks in recent years because its mechanical properties are better than those of traditional cast cobalt-chromium (Co-Cr) alloys. In addition, SLM frameworks avoid casting problems such as deformation. SLM is an advanced manufacturing technique that can satisfy clinical requirements, and appliances made with SLM have been reported to have improved biocompatibility, yield strength, tensile strength, bend strength, and corrosion resistance. However, anisotropy has been reported in components made by SLM.
RPD clasps fabricated by SLM have been prepared mainly from Co-Cr alloys. Kajima et al reported that the fatigue strength of SLM Co-Cr clasps was affected by different build orientations, but the authors are unaware of reports on the fatigue strength of Ti-6Al-4V alloy clasps prepared by SLM. In addition, previous studies on the fatigue resistance of prostheses have mainly been performed under dry conditions, unlike the oral environment. The objective of this study was to investigate the microstructure and properties of RPD clasps fabricated in Ti-6Al-4V by SLM with different build orientations. The null hypothesis was that build orientations would have no effect on the microstructure and properties of SLM Ti-6Al-4V clasps.
Material and methods
A mesioocclusal rest (2.5-mm long, 2.5-mm wide, and 1.5-mm deep) and a mesial guide plane were prepared on an acrylic resin mandibular first molar tooth. The lost wax process was used to fabricate 4 metal abutment teeth by using Co-Cr alloys (Wirobond C; Bego) from the resin tooth. All abutment teeth were welded to Co-Cr plates (15×15×2 mm) and polished by the same technician to maintain uniformity.
Refractory dies were duplicated from 1 of the metal abutment teeth. The 0.25-mm and 0.50-mm undercut locations were marked on the distobuccal surfaces of dies using a dental surveyor. One-third of the retentive clasp arm was placed in the undercuts (length of the clasp arm=12 mm, thickness of the clasp arm=1.1 mm, width of clasp shoulder=2.5 mm, width of clasp tip=1.5 mm). Wax clasps with 5-mm-diameter sprue rods were manufactured according to the marks. The sprue rods were later used as fixed links to attach the clasps to the universal testing machine. Twelve wax clasps (n=6 for each undercut) were fabricated in Ti-6Al-4V ingots (Ti-6Al-4V; Neodent). The remaining 3 metal teeth were scanned by using a 3D optical scanner (D810; 3Shape) to build the digital models. Clasps with the same geometric shape as the cast clasps were designed according to the models. The data were imported into a 3D-printing software program (Magics; Materialise) to make the longitudinal axes of the clasp arms at 0 degrees, 45 degrees, and 90 degrees to the build direction ( Fig. 1 ). The clasp specimens were divided into the following groups (n=6): CAST, cast clasps used Ti-6Al-4V ingots; SLM0, SLM clasps built at an angle of 0 degrees; SLM45, SLM clasps built at an angle of 45 degrees; and SLM90, SLM clasps built at an angle of 90 degrees.
Titanium alloy powder (Ti-6Al-4V; EOS) with an average 38-μm particle size was used to prepare the specimens with an SLM machine ( Fig. 2 ). The process parameters of SLM clasps are listed in Table 1 . All clasps were heated to an annealing temperature of 800°C for 2 hours, followed by cooling in air. To ensure uniformity, only electropolishing procedures were performed on the intaglio surfaces of the clasps. The chemical composition of the Ti-6Al-4V ingot and power is listed in Table 2 .
|Laser power (W)||200|
|Layer thickness (μm)||30|
|Point distance (μm)||45|
|Exposure time (μs)||200|
|Scan speed (mm/s)||225|
|Laser spot size (μm)||70|
|Raw material||Titanium (Ti)||Aluminum (Al)||Vanadium (V)|
|Ti-6Al-4V powder (mean ±SD)||89.0 ±2.0||6.4 ±0.4||4.0 ±0.2|
Clasps were selected by using a random sampling method. All clasps from each group were given a number from 1 to 6. Random selection was performed automatically by using a software program (Multi Random Data Generator, v1.01; SoonWare) to select 4 clasps for subsequent experiment. The selected clasps were cut along the cross-section of the fixed link to create microstructure specimens (5 mm in diameter and 2 mm in height). The specimens were polished by using a polishing machine (Tegramin-30; Struers) up to 2000 grit, followed by etching for 10 seconds in a solution of 50 mL H 2 O, 25 mL HNO 3 , and 5 mL HF. Finally, the specimens were ultrasonically cleaned for 5 minutes in anhydrous alcohol (ethyl alcohol; Jinan Yunxiang Chemicals Co, Ltd). The microstructure was investigated by using a metallographic microscope (Axio Imager M2m; Zeiss) and a confocal laser scanning microscope (LSM 700; Zeiss). The specimens for the microstructure evaluation are listed as follows: MC, the specimen cut from the fixed link of the cast clasp; ML0, the specimen cut from the fixed link of the SLM90 clasp (parallel to the build orientation); ML45, the specimen cut from the fixed link of the SLM45 clasp (inclined to the build orientation); ML90, the specimen cut from the fixed link of the SLM0 clasp (vertical to the build orientation).
The SLM and cast clasps were detected by using an X-ray apparatus (exposure conditions: tube voltage 70 kV, tube current 200 mA, exposure time 63 ms, exposure distance of approximately 20 cm). Pores, cracks, and other defects in the clasps on the X-ray plate were observed with the naked eye. Subsequently, the intaglio surfaces of the clasps were placed under the confocal laser scanning microscope to obtain 3D images of the clasps ( Fig. 3 ). The surface roughness (Ra) data were obtained from the 3D images.
Each clasp was placed on its associated abutment. The gaps between the clasp arms and the lingual side of the teeth were then measured by using a stereomicroscope (M205A; Leica) under 30× magnification. The gaps of the tip, middle, and initial part of the clasp arm were measured. The average of the readings represented the fit of the clasps.
Digital models of 2 undercut SLM clasps were transformed into entity models by using Geomagic software (3D SYSTEMS). Finite element analysis was carried out by using the ANSYS 17.0 software (ANSYS). The characteristics for element analysis are listed in Table 3 . The nodes and elements of the 0.25-mm undercut model were 37 922 and 23 568, and the nodes and elements of the 0.50-mm undercut model were 35 599 and 22 228 ( Fig. 4 ). The von Mises stress values and distributions of the clasps were investigated as the clasps were moved perpendicularly on and off the abutment teeth.
|Specimen||ρ (kg/m 3 )||E (GPa)||UTS (MPa)||0.2% YS (MPa)||Poisson Ratio|