169 American Academic Scientific Research Journal for Engineering, Technology, and Sciences ISSN (Print) 2313-4410, ISSN (Online) 2313-4402 http://asrjetsjournal.org/ Dimensional Accuracy and Final Density Measurement of One-, Three-, and Eight-Unit Fixed Dental Frameworks Based on Co-Cr, Manufactured by Using Conventional, Additive and Subtractive Technologies Andrea Sinčák Konečná a* , Viktória Rajťúková b , Alena Findrik Balogová c , Ľuboš Chromý d , Gabriela Ižaríková e , Radovan Hudák f a Stomatological clinic, Department of Conservative and Prosthetic dentistry, University Hospital of L.Pasteur, Tr.SNP č.1, 04011 Košice, Slovakia b,c,d,f Department of Biomedical Engineering and Measurement, Faculty of Mechanical Engineering, Technical University of Košice, 040 01 Košice, Slovakia e Department of Applied Mathematics and Informatics, Faculty of Mechanical Engineering, Technical University of Košice, 040 01 Košice, Slovakia a Email: konor24@gmail.com, b Email: viktoria.rajtukova@tuke.sk c Email: alena.findrik.balogova@tuke.sk, d Email: lubos.chromy@tuke.sk e Email: radovan.hudak@tuke.sk, f Email: gabriela.izarikova@tuke.sk Abstract For the aim of this study, 90 fixed metal frameworks based on Co-Cr were made by using conventional lost vax technique, additive 3D printing done by Selective Laser melting (SLM) technique and subtractive manufacturing using CNC milling. Constructions were produced by the Ceramill Motion 2 milling machine (Amann Girrbach) and Mlab Cusing R 3D printer (GE Additive). Samples were made as one-, three- and eight- unit frameworks based on existing clinical cases. Initial stl. model was built up on a scanned plaster model of three clinical cases by ZirconZahn Modellier software. Evaluation of dimensional accuracy was made by comparison of initial stl. model with a scan of manufactured framework and analyzed by measurement software. Density measurements were made by helium (He) based gas pycnometry. Gained data was statistically analyzed by using T-test and F-test for technologies comparison and non-parametric form of ANOVA: Kruskal Wallis was applied in density evaluation of all samples. Keywords: 3D printing; additive technologies; fixed dental prosthetics; 3D printed dental prosthetic; additive technologies in dentistry; subtraction techniques; CNC milling in dentistry; pycnometry; Co-Cr alloy density. ------------------------------------------------------------------------ * Corresponding author. http://asrjetsjournal.org/ American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 170 1. Introduction Development of new technologies affects all of the spheres of human activities; it also impacts medicine, stomatology not excluded. Innovative technologies facilitate, for example, more accurate processing of input data (processing of imprints and models) using the scanning systems, as well as more efficient manufacturing technologies that improve production planning and automatization. A great advantage of digital technologies is the visualisation of the planned product so the patient can see the planned treatment even before starting the process of manufacturing. Despite the fact that new materials are being developed, chrome-cobalt materials still rank among the key materials used in the production of dental replacements. Cobalt- and chrome-based alloys (Co-Cr) are largely used for metal bases for fixed dental prostheses due to their excellent mechanical properties, such as high resistance to corrosion and wear resistance and also good properties of biocompatibility. In the past, CoCr materials were used for manufacturing metal frameworks by traditional method of casting (lost wax method). This process requires several manual steps to be carried out in the manufacturing process and no computer guided systems are used. Co-Cr based materials can be also used by subtractive (CNC milling) and additive (3D printing) technologies. During casting, a big amount of waste material is created. Novel method of CNC milling requires a metal block. Although this method exhibits high accuracy, its disadvantage is a significant amount of waste or unused material. The latest, albeit the least used technology, is the manufacturing using the Selective Laser Melting (SLM) from metal powder. Layers of metal powder are melted into a 3D model using a computer-controlled laser. Advantages of the SLM, when compared to conventional methods, include the possibility of manufacturing customised complex models and parts with a dense structure and predefined surface roughness. This facilitates the fabrication of robust as well as very fine elements, while the quantities of waste materials are negligible as the powder is recyclable [1,2]. One of the most important features of fabricated fixed dental frameworks is marginal fit of the crowns made. Another important property are dimensional accuracy and final mechanical properties of processed metal used in manufacturing which is highly dependent on a production method. The final metal properties after they have gone through the manufacturing process, marginal gaps, and dimensional accuracy of framework (which is related with internal fit of the crowns) are the mostly observed properties when evaluation of accuracy of manufacturing method is done, as they belong to the most important aspects that determine the expected lifespan and function of dental replacements. By creating as accurate construction as possible, microfractures, cement dissolving, bacteria infiltration, accumulation of plaque, formation of caries and gingivitis [3,4]. Also, the inner space between the replacement and the prepped tooth affects the retention and occlusion. If the internal gap is too wide or narrow, the work may become displaced, or it may lead to malocclusion or imperfect insertion of the replacement [5]. Therefore, marginal, and internal gaps are the critical factors that determine the success of fabrication of dental replacements [6,7]. The purpose of this submitted study was to apply conventional casting, 3D printing SLM method and CNC milling technique in order to manufacture samples of fixed metal frameworks of three clinical types (single American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 171 crown, three-unit bridge and eight-unit bridge) made of the CoCr material. After manufacturing all 90 samples, they were backwards scanned and compared throw GOM INSPECT software (GOM, Germany). In addition to dimensional accuracy a testing of density by using gas pycnometry (AccuPyc II 1345, MICorp.), was performed to estimate a qualitative aspect of materials used in manufacturing. The goal of this study was to evaluate all gained data and identify the most precise method of manufacturing a fixed metal framework with the most suitable mechanical and dimensional properties in the range of used technologies and materials. 1.1. Designing and modelling of fixed dental frameworks After selecting three clinical cases (prosthetic treatment by single-crown, three-unit bridge and eight-unit framework), with convergency of preparation about 3–10°, cervico-occlusal reduction 3-4 mm and chamfer design preparation, plaster models were scanned by ARTI S 900 intraoral scanner and transformed into ZIRCONZAHN Modellier software (Figure 1). ased on the obtained data, the models of frameworks were subsequently designed. A created stl. model of each clinical situation vas than used for sample producing by STL technologies and CNC milling, but not for classical method of casting. By this method, frameworks were modelled by hand using wax done by dental technician. Totally ninety samples were made, thirty for each technology and from those ten for each clinical case. Figure 1: Dental frameworks modeled in ZirconZahn modellier software a) Single crown b) 3-unit bridge c) 8-unit bridge Afterwards all manufactured frameworks were backwards scanned and saved in ZirconZahn Archive and further transferred as stl. input files into GOMInspect software (Zeiss) and data sets for dimensional accuracy comparison were made. 1.2. Technologies and materials used As first was created a group of samples made by casting using lost wax technique. All of these samples were made by dental technician, without creating wax preforms of frameworks. Material selected for this technique was Co-Cr based, commonly used Heraenium Pw (Kulzer). The initial phase of reproduction of the situation was carried out in the classic way by pouring the impression by stone plaster and creating split models with American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 172 guide pins and repositioning rings and wax preforms of final frameworks included pouring channels and metal stack. Poured construction was afterwards mechanically processed and prepared for back scanning. Another group of samples was made by Selective Laser Melting (SLM) technology using Mlab Cusing R 3D printer (GE Additive, USA) and selected material was Co-Cr based powder Star Bond Easy Powder 30 (Scheftner). Process started by recording an initial stl. model to CAMBridge software of 3D printer. After positioning the object on construction plate so the occlusal parts are facing the construction plate, designing a support structure, designation of objects and their positioning on a construction plate transformed to cls. format and by sending the file to Mlab Cusing R printer and launch the manufacturing process. Afterwards the samples were mechanically processed and annealed, so the manufacturing process will be identical with the real procedure done in manufacturing a fixed dental prosthetics. Last method used for sample creation was CNC milling technology done by Ceramill Motion 2 (Amann Girbach) by using pressed powder plate Ceramill Sintron (Amann Girbach) based on Co-Cr as the most alike to previously used materials. After transporting initial .stl file to Ceramill Match program, selecting a material and its shape, positioning the objects in plane so there is no overlap of any construction part and designing the length, thickness, and position of attachments. After all initial necessary steps, the .stl file is uploaded to the CNC milling machine software Ceramill Match and the process is initiated. After milling, the process of annealing and mechanical processing is done. 1.3. Method of dimensional accuracy evaluation In evaluation of dimensional accuracy, three groups were defined. In the first group was compared initial stl. model with all 3D printed samples, in the second group was compared initial stl. model with all CNC milled samples. Data gained from these two groups was used for evaluation of accuracy of each method itself. The third group was generated from comparison of referential 3D sample for each clinical case (sample was selected on a criterion of smallest deviation from initial stl. model) with all CNC milled samples. By this comparison was gained data used for this two-method comparison. Samples made by casting, was not incorporated in the final evaluation, as the deviation from initial stl. models (which was not used to generate these samples) or 3D printed, and milled samples was by the software non-evaluable, or the value of deviation differed in repeated comparison for the same sample. The comparison itself was done by GOMInspect software (Figure 2). In each pair of compared samples, one was designated as nominal CAD Body and the compared sample as Mash Body. In groups where stl. model was compared with 3D printed or CNC milled samples, stl. was determined as CAD body. In the group where 3D sample was compared by all CNC milled samples, 3D sample was estimated as initial CAD body. For back- scanning Arti S900 extraoral scanner from ZirconZahn was used, calibrated after every set of samples. Reducing a blunder scan, mattifying spray (IP Sken Spray) was applied over the whole surface of frameworks, so the reflection of metal will be minimalized. After the process of scanning is completed, a scan needs to be cut, saved in stl. format and used for any CAD/ CAM device. American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 173 Figure 2: Visualization of deviation between stl. and CNC milled single crown sample made in GOMInspect software 1.4. Method of density measurement Furthermore, an evaluation of final density of produced dental frameworks was made by using Gas Pycnometry device AccuPyc II 1345 (MICorp.), using helium (He) as the detecting medium. There were five samples chosen from each manufacturing method, and for each sample ten separate measuring was made and open porosity was estimated. Measured amount of density of each sample for each material was compared with optimal density given by manufacturer. Measurement was done by ASTM B923 a USP 699 norms. 2. Results After comparison of all selected samples by software and gaining the complete data sheets of pycnometry analysis, statistical evaluation and analysis was performed. For a better orientation through analyzed data, samples were distributed and labeled (Table 1). Table 1: Samples distribution Casting CNC milling 3D printing Nominal stl. model Single crown 10 (A1-10) 10 (B1-10) 10 (C1-10) X1 3-unit bridge 10 (A11-20) 10 (B11-20) 10 (C11-20) X2 8-unit bridge 10 (A21-30) 10 (B21-30) 10 (C21-30) X3 2.1. Comparison of stl. files with 3D printed and CNC milled samples The comparison was made in several groups. In the initial comparison, nominal samples (marked X1, X2, X3) American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 174 and samples produced by 3D printing (marked C1-C30) and CNC milling (marked B1-B30) were compared. The resulting values are shown in Table 2, 3 and 4. The nominal sample was a stl. model that was created as a production model in the "ZirconZahn Modellier" software for each situation, and samples were produced based on it. Table 2: Comparison of stl. file with 3D printed and CNC milled samples for single crown stl. - 3D print Deviation [µm] stl. - CNC milling Deviation [µm] X1-C1 30,3 X1 -B1 34,7 X1-C2 31,3 X1-B2 32 X1-C3 26,8 X1-B3 31,2 X1-C4 29,9 X1-B4 30,6 X1-C5 32 X1-B5 34,7 X1-C6 27,7 X1-B6 35 X1-C7 36,2 X1-B7 38,6 X1-C8 24,7 X1-B8 41,5 X1-C9 28 X1-B9 40,6 X1-C10 34,3 X1-B10 39,1 Average 30,12 Average 35,8 Variance 0,0122 Variance 0,02 Decisive deviation 3,492 Decisive deviation 4,472 Table 3: Comparison of stl. file with 3D printed and CNC milled samples for 3-unit framework stl. - 3D print Deviation [µm] stl. - CNC milling Deviation [µm] X2-C11 45,5 X2-B11 58,2 X2-C12 47 X2-B12 68,2 X2-C13 46,3 X2-B13 65,8 X2-C14 47,5 X2-B14 51,4 X2-C15 46,4 X2-B15 55,7 X2-C16 40,4 X2-B16 56,4 X2-C17 37,7 X2-B17 52 X2-C18 45,1 X2-B18 63,9 X2-C19 46,4 X2-B19 49,4 X2-C20 39,8 X2-B20 69,9 Average 44,21 Average 35,8 Variance 0,0124 Variance 0,02 Decisive deviation 3,521 Decisive deviation 4,472 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 175 Table 4: Comparison of stl. file with 3D printed and CNC milled samples for 8-unit framework stl. - 3D print Deviation [µm] stl. - CNC milling Deviation [µm] X3-C21 46,6 X3 -B21 86,8 X3-C22 48,4 X3-B22 81,4 X3-C23 35,9 X3-B23 81 X3-C24 45,9 X3-B24 67,1 X3-C25 42,6 X3-B25 137 X3-C26 36,6 X3-B26 83,9 X3-C27 37,8 X3-B27 70,6 X3-C28 47,9 X3-B28 86,5 X3-C29 49,2 X3-B29 88,9 X3-C30 47,6 X3-B30 143,9 Average 43,85 Average 92,71 Variance 0,0273 Variance 0,684 Decisive deviation 5,224 Decisive deviation 26,15 When comparing samples produced by 3D printing and milling with the initial stl. model used as their template, it turned out that the size of work does not have such a significant effect on the increasing tendency of deviation in 3D printed frameworks as expected (Figure 3). In the compared group of stl. and 3D printed three-unit bridge and stl. and 3D printed eight-unit bridge, the average deviation is greater for three-unit works (44.21µm) than for eight-unit works (43.85µm). It was also manifested in the compared group of stl. and milled single crowns, where the average deviation was 35.8 µm, and the group of stl. and milled three-unit bridges with the same average deviation of 35.8 µm. The group with the smallest decisive deviation is the compared group of 3D printed single crowns against the .stl file with a standard deviation of 3.4µm, and the group with the largest standard deviation is the compared group of stl. against the milled eight-unit works with a standard deviation of 26.1µm. When comparing single crown samples produced by 3D printing and milling against the stl. model, the average deviation was 30.12 µm and decisive deviation was 3.492 µm for crowns produced by 3D printing. For milled single crowns, the average deviation was 35.8 µm with a standard deviation of 4.472 µm. When comparing three-unit restorations in the stl. group against 3D printed and milled ones, the resulting value of the average deviation for 3D printed works was 44.21 µm with a decisive deviation of 3.521 µm, and for milling three-unit bridges it was an average deviation of 35.8 µm with a decisive deviation of 4.472 µm. In the compared group of stl. with 3D printing and milling of eight-unit frameworks, the average deviation was 43.85 µm with a decisive deviation of 5.224 µm for 3D printing works, for milling eight-unit works the average deviation is 92.71 µm with a decisive deviation of 26.15 µm. Standard deviation data is more clearly shown in the histogram below. American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 176 Figure 3: Average deviation considering the size and method of production 2.2. Comparison of reference 3D sample and CNC milled samples In this evaluation, one sample from each clinical case produced by 3D printing was selected that had the smallest deviation from the original .stl file and was compared in the GOM Inspect software against all samples of the given group produced by the subtraction method - CNC milling (Table 5, 6, 7). From the samples of single crowns, sample C8 was selected, which showed the smallest deviation from the original stl. file - 24.7 µm and was determined in this comparison as the reference - "CAD Body". For the three-unit bridge, it was sample C17, whose deviation from the stl. file was 37.7 µm, and for the comparison of the eight-unit bridge, sample C23 was selected witch has a deviation of 35.9 µm. Table 5: Difference between 3D printed sample and CNC milled samples of single crown Referential sample of 3D printed – CNC milled samples Deviation [µm] C8-B1 54,2 C8-B2 47,3 C8-B3 52,2 C8-B4 42,1 C8-B5 55,8 C8-B6 38,8 C8-B7 58,6 C8-B8 55 C8-B9 45,9 C8- B10 62,2 Average deviation 51,21 Variance 0,056 Decisive deviation 7,483 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 177 Table 6: Difference between 3D printed sample and CNC milled samples of 3-unit bridge Referential sample of 3D printed – CNC milled sample Deviation [µm] C17-B11 87 C17-B12 66,1 C17-B13 69,3 C17-B14 79,9 C17-B15 56,2 C17-B16 63,9 C17-B17 90,3 C17-B18 64,5 C17-B19 83,7 C17-B20 63,6 Average deviation 72,45 Variance 0,138 Decisive deviation 11,74 Table 7: Difference between 3D printed sample and CNC milled samples of 8-unit bridge Referential sample of 3D printed – CNC milled samples Deviation [µm] C23-B21 74,2 C23-B22 65,8 C23-B23 70,3 C23-B24 55,8 C23-B25 123,3 C23-B26 68,2 C23-B27 55,6 C23-B28 73,3 C23-B29 74,6 C23-B30 132,8 Average deviation 79,39 Variance 0,709 Decisive deviation 26,63 When comparing the selected sample produced by 3D printing and all samples produced by CNC milling technology, the value of the average deviation had an increasing tendency due to the size of the work. A more significant difference appeared only when comparing the standard deviations of individual groups of samples, American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 178 where the difference between three-unit and eight-unit works shows a greater value of the standard deviation than when comparing the decisive deviation of single crown and three-unit works. When comparing the reference 3D sample and milled samples of one-piece work, the average deviation was 51.21µm with a decisive deviation of 7.483µm. When comparing three-unit works, the value of the average deviation was 72.45 µm with a decisive deviation of 11.74 µm. In the compared group of eight-unit works produced by 3D printing and milling, the average deviation was 79.39 µm with decisive deviation of 26.63 µm. For better presentation, the dependence of the deviation of the 3D printed sample to the milled samples with respect to the size of work is shown in the graph below (Figure 4). Figure 4: Graphic representation of the deviation of 3D and milled samples with aspect to the size of work in [µm] The research hypothesis assumes that 3D printing technology is more accurate than milling technology. The hypothesis was verified using F-test and T-test statistical tests. Initial stl. files were compared to works produced by 3D printing and CNC milling technology. When comparing the deviations of single crowns, the resulting T-test value is p=0.003143 (<0.005). When comparing the deviations in the set of three-unit bridges, p=0.00007 (<0.005) and for eight-unit works, p=0.00017 (<0.005). This means that at the significance level of 0.05 we reject the zero hypothesis about the equality of mean values. Thus, a statistically significant difference was confirmed between the technologies. Based on the available production technologies, materials and devices used to produce samples for the purposes of this study, it can be concluded that structures made by 3D printing are more accurate than works made by milling technology. 2.3. Evaluation of sample density By pycnometry testing sample, we obtained real density values of the individual samples. These values were American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 179 compared with the theoretical density that each material has given by the manufacturer. For cast samples from the material "Heraenium PW" the density is 8.9 g/cm³, for the material "Starbond Easy Powder 30" used for 3D printing it is 8.5 g/cm³ and for the milling material "Ceramill Sintron" it is given density 7.9g/cm³. The resulting densities of the tested samples are clearly presented in Table 8. Table 8: Overlook of density values of tested samples Casting Theoretical density Sample density Open porosity Sample 1 8,9 g/cm³ 6,4473 g/cm³ 28% Sample 2 8,9 g/cm³ 6,2305 g/cm³ 30% Sample 3 8,9 g/cm³ 6,3602 g/cm³ 28,6% Sample 4 8,9 g/cm³ 6,8898 g/cm³ 22,59% Sample 5 8,9 g/cm³ 6,6645 g/cm³ 25,12% Average 8,9 g/cm³ 6,51846 g/cm³ 26,862% CNC milling Theoretical density Sample density Open porosity Sample 1 7,9g/cm³ 7,6321 g/cm³ 3,4% Sample 2 7,9g/cm³ 7,7613 g/cm³ 1,6% Sample 3 7,9 g/cm³ 7,607 g/cm³ 3,71% Sample 4 7,9 g/cm³ 7,6969 g/cm³ 2,58% Sample 5 7,9 g/cm³ 7,7209 g/cm³ 2,27% Average 7,9 g/cm³ 7,6836 g/cm³ 2,712% 3D printing Theoretical density Sample density Open porosity Sample 1 8,5 g/cm³ 8,2345 g/cm³ 3,13% Sample 2 8,5 g/cm³ 8,1835 g/cm³ 3,37% Sample 3 8,5 g/cm³ 8,1837 g/cm³ 3,37% Sample 4 8,5 g/cm³ 8,2014 g/cm³ 3,22% Sample 5 8,5 g/cm³ 8,1669 g/cm³ 3,92% Average 8,5 g/cm³ 8,1940 g/cm³ 3,402% Differences between sample densities were verified using statistical tests. In order to use the parametric test (ANOVA), the assumption that the random samples come from a population with a normal distribution must be met. This condition was verified by the Shapiro-Wilks test, at a significance level of 0.05 it was confirmed for all three technologies. The second condition, the homogeneity condition, was verified by the Hartley test, but it is not fulfilled at the 0.05 significance level. A non-parametric form of ANOVA was used to compare the 3 technologies: Kruskal Wallis test. The result of the test is p=0.0374, which means that at the significance level of 0.05 we reject the null hypothesis about the equality of mean values. Thus, a statistically significant difference was confirmed between the technologies. Subsequently, post-hoc. the analysis revealed pairs between which there is a statistically significant difference. A statistically significant difference in sample density was not confirmed when comparing milling and 3D American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 180 printing technology, p=0.72674 (<0.005). Samples produced by classic technology show the highest value of porosity, i.e., have the lowest density. This fact can significantly affect the lifetime of prosthetic work from a mechanical point of view (Figure 5). Figure 5: Graphic representation of density of dental restorations made by different technologies The histogram shows the percentage distribution of the samples with respect to their measured density. It is evident that when using technological procedures using digital technologies, the density of manufactured dental replacements is almost identical to the density indicated in the technical parameters of the materials used in production. This also indicates the achievement of the properties of the prosthetic work, which are determined and optimal for the given material, such as e.g., surface hardness - resistance to abrasion, compressive and tensile strength, release of material molecules in the DÚ environment, etc. In the case of samples produced by the conventional casting method, the density of the samples reached significantly lower values, with an average deviation of 26.862%, which is significantly more than in the technological procedures of 3D printing and milling, where the values of the average deviation reached 2.712% for milled samples and 3.402% for samples produced by 3D printing. The assumed reason for such a significant difference between conventional method and the modern ones can lay in the very process of melting and cooling the alloy in the working process. The human factor can also have a significant impact on the quality of alloy, i.e., non-compliance or alteration of the work procedure. 3. Discussion There are countless methods to determine the accuracy of prosthetic work. The basic factor in comparison study American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 181 is processing the data under the same conditions and the creation of a sufficiently large set of data obtained under the same conditions that can be used for evaluation purposes. For the purposes of this work, we used initially and in a back scan for all samples the same scanner and this created a sufficiently extensive database of scans that could be relevant to compare. Work procedures were strictly followed. In the production of metal structures by casting method, it is not possible to achieve complete accuracy and completely identical repetition in the creation of the same structure (each modeled individually by hand), therefore the samples produced by this method were used only as a reference for mutual comparison and were not further included in the evaluation of dimensional accuracy done by software. Despite the fact that with the additive and subtractive method the production process is digitalized and fully automatic, and the samples were made on the basis of one common stl. model, several factors could influence the data obtained. The first could be the deviation in accuracy of the scanner used in initial scanning and also back scanning of the manufactured structures. Our Scanner: Arti S900 from ZirconZahn has a 10 µm deviation given by manufacturer. Another factor may be the way the software transforms the scanned data into a 3D model, but also the accuracy of the machines used to produce samples from CAD data. The accuracy specified by the manufacturer for the Mlab Cusing R device is 20 µm, for the Ceramill Motion 2 device used for the subtraction method, it is approximately 20-60 µm, but it also depends on the diameter of the manufactured object or the thickness of its walls and the diameter of the selected machining tool. In the production of samples, a new tool was used for each process in order to exclude the occurrence of inaccuracy caused by tool wear and the inability of the instrument to compensate for this phenomenon. In order to exclude differences in the produced samples during milling, the same shape and length of the tool were always used. Initial data was gained by scanning a plaster working model of intraoral situation. In order to exclude the inaccuracy of the obtained data, plaster with a low reflection was used, and a layer of matting spray intended for dental scanners was applied to all scanned metal objects [8-10]. When determining the density of the material, samples made by all three technologies were evaluated, i.e., also samples made by casting, where the individual shape of the samples was not a factor for which they could not be included in testing and compared with other technologies. Each measurement was automatically repeated 10 timesfor each sample to confirm the correctness of the measurement. Currently, any source is not providing an exact number of data determining the size of the marginal gap, which would be considered optimal for specific technology. For constructions made with additive SLM technology, more data are given in the literature, from 40- 120 µm, the value of 40 µm is most often found for single- member work [11-18]. Also, the value of the size of the marginal gap for restorations made by the subtraction method varies in the literature, we can meet values from 50 µm up to 200 µm in larger works, the most often given value as 60 µm is acceptable for a marginal gap with a one- to three-unit restoration [11][13][15-18]. All technologies are moving forward through development, and we can assume that the accuracy of replacements can even now approach the lower limits stated in the literature, thus enabling the production of long-lasting dental replacements. American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 182 4. Conclusion Based on the obtained data and their analysis, it can be concluded that fixed dental restorations produced by SLM 3D printing technology are more accurate compared to the same restorations produced by CNC milling. It also follows from the obtained data that structures produced by 3D printing and CNC milling have a final density very close to the optimal density of the alloy provided by the manufacturer. It is therefore possible to state that by using both of these technologies in the production process, we can achieve equally mechanically satisfying dental structures. From the obtained data for samples produced by casting, it can be concluded that the use of this technology in the production of fixed dental restorations results in less durable and less morphologically satisfactory fixed dental structures. All these conclusions apply only to fixed frameworks made by materials and device used in this study. By summarizing the available literary sources and analyzing the data obtained from this work, the use of modern technologies in the production of dental fixed replacements appears to be a much more effective method compared to the conventional casting method. Additive technology achieved the best results, followed by subtractive technology, and the conventional casting method finished in last place. And this confirms the established research hypothesis of this study, which assumed that 3D printing technology is more accurate than milling technology. The hypothesis was also confirmed statistically. Acknowledgements This research was supported by project VEGA 1/0387/22 - Development and testing of systems for controlled stimulation of cell growth in a bioreactor environment using computer vision and KEGA 050TUKE-4/2022 - Additive manufacturing in medicine - creation of multimedia material and tools to support teaching in biomedical engineering. This work was supported by the Slovak Research and Development Agency under the contract No. APVV-17-0278. This publication is the result of the project implementation Open scientific community for modern interdisciplinary research in medicine (Acronym: OPENMED), ITMS2014+: 313011V455 supported by the Operational Programme Integrated Infrastructure funded by the European Regional Development Fund. The achieved results were created within the investigation of the project no. 2018/14432: 1-26C0, which is supported by the Ministry of Education, Science, Research and Sport of the Slovak Republic within the provided incentives for research and development from the state budget in accordance with Act No. 185/2009 Coll. on incentives for research and development. References [1]. K. Torabi, E. Farjood, and S. Hamedani. “Rapid Prototyping Technologies and their Applications in Prosthodontics, a Review of Literature.” J. Dent., 6:1–9, 2015. [2]. K. Quante, K. Ludwig, and M. Kern. “Marginal and internal fit of metal-ceramic crowns fabricated with a new laser melting technology.” Dent. Mater, 24:1311–1315, 2008. [3]. M. S. Jacobs and A. S. Windeler. “An investigation of dental luting cement solubility as a function of American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) (2022) Volume 89, No 1, pp169-183 183 the marginal gap.” J. Prosthet. Dent., 65:436–442, 1997. [4]. O. Schaefer, H. Kuepper, G. A. Thompson, G. Cachovan, A. F. Hefti and A. Guentsch. “Effect of CNC-milling on the marginal and internal fit of dental ceramics: A pilot study.” Dent. Mater., 29:851– 858, 2013. [5]. J. T. Colpani, M. Borba, and Á. Della Bona. “Evaluation of marginal and internal fit of ceramic crown copings.” Dent. Mater., 29:174–180, 2013. [6]. J. A. Sorensen. “A standardized method for determination of crown margin fidelity.” J. Prosthet. Dent., 64:18–24, 1990. [7]. C. Grenade, A. Mainjot and A. Vanheusden. “Fit of single tooth zirconia copings: comparison between various manufacturing processes.” J. Prosthet. Dent., 105:249–255, 2011. [8]. N. A. Dolgov, T. Dikova, D. Dzhendov, D. Pavlova and M. Simovb. “Mechanical properties of dental Co-Cr alloys fabricated via Casting and SLM.” Mat. Sci. Non-Equilibrium Phase Transform, 2:3–7, 2016. [9]. “Concept laser, dental machines. Internet: www.concept-laser.de/en/industry/dental/machines.html, Jan. 01, 2021. [10]. A. Gribach. “Ceramill Motion 2: Milling machines. Internet: www.amanngirrbach.com/en/products/production-cam/ceramill-motion-2, Sep. 09, 2021. [11]. G. Bosch, A. Ender, A. Mehl. “A 3-dimensional accuracy analysis of chairside CAD/CAM milling processes.” The J. of Prost. Dent., 112(6): 1425-1431, 2014. [12]. E. J. Bae, I. D. Jeong, W. C. Kim, et al. “A comparative study of additive and subtractive manufacturing for dental restoration.” J. Prost. Dent., 118:187-193, 2017. [13]. A. Örtorp, et al. “The fit of cobalt–chromium three-unit fixed dental prostheses fabricated with four different techniques: A comparative in vitro study.” Dent. Mat., 27(4): 356-363, 2011. [14]. H. Nesse, D. M. Ulstein, M. M. Vaage, M. Øilo. “Internal and marginal fit of cobalt-chromium fixed dental prostheses fabricated with 3 different techniques.” J. Prost. Dent., 114(5):686-92, 2015. [15]. R. Mejia, S. M. Tobon. “Marginal fit of metal ceramic restorations subjected to a standardized post soldering technique.” J. of Prost. Dent., 83:535-9, 2000. [16]. P. Limkangwalmongkol, G. J. Chiche, M. B. Blatz. “Precision of fit of two margin designs for metal – Ceramic crowns.” J. of Prost., 16:233-7, 2007. [17]. D. Petteno, G. Schierano, F. Bassi, et al. “Comparison of marginal fit of three different metal - ceramic systems: an in vitro study.” Inter. J. of Prost., 5:405-8, 2000. [18]. Y. Ucar, T. Akova, M. S. Akyli, et al. “Internal fit evaluation of crowns prepared using a new dental crown fabrication technique: laser-sintered Co-Cr crowns.” J. of Prost. Dent., 102: 253-9, 2009. http://www.concept-laser.de/en/industry/dental/machines.html http://www.amanngirrbach.com/en/products/production-cam/ceramill-motion-2