Super Abrasive Machining of Integral Rotary Components Using Grinding Flank Tools

Manufacturing techniques that are applied to turbomachinery components represent a challenge in the aeronautic sector. These components require high resistant super-alloys in order to satisfy the extreme working conditions they have to support during their useful life. Besides, in the particular case of Integrally Bladed Rotors (IBR), usually present complex geometries that need to be roughed and finished by milling and grinding processes, respectively. In order to improve their manufacturing processes, Super Abrasive Machining (SAM) is presented as a solution because it combines the advantages of the use of grinding tools with milling feed rates. However, this innovative technique usually needed high tool rotary speed and pure cutting oils cooling. These issues implied that SAM technique was not feasible in conventional machining centers. In this work, these matters were tackled and the possibility of using SAM in these five-axis centers with emulsion coolants was achieved. To verify this approach, Inconel 718 single blades with non-ruled surfaces were manufactured with Flank-SAM technique and conventional milling process, analyzing cutting forces, surface roughness, and dimension accuracy in both cases. The results show that SAM implies a suitable, controllable, and predictable process to improve the manufacture of aeronautical critical components, such as IBR.


Introduction
Aeronautical industry is on the path to steady growth, is one of the most strategic sectors in the worldwide economy.In fact, TEDAE association (Spatial, aeronautical, and defense technology Spanish association) states that the turnover of this industry is about 7600 million euros, from which 9.5% is aimed at R & D [1].This economic investment stems from the need of facing global competitiveness.This is being reflected in high requirements that are related not only to mechanical properties required through all the life-cycle of these components, but also on the optimization of manufacturing process.This is especially relevant for critical aeronautical turbomachinery components where their reliability directly affects to aeronautical jet engines security.Among them, noteworthy components are axial compressors and turbines in civil aviation.In particular, this work deals with manufacturing processes of Bladed Disks (Blisk) and Integrally Bladed Rotors (IBR).The main difference between them is the junction of the blades to the disk.As it is shown in Figure 1, Blisk components have their blades separated from the disk, assembled through fir-trees.IBR, however, are integrated components, which are manufactured from the same blank, what avoids unions' issues [2].
Metals 2018, 8, 24 2 of 11 components have their blades separated from the disk, assembled through fir-trees.IBR, however, are integrated components, which are manufactured from the same blank, what avoids unions' issues [2].It should be noted that during its entire lifetime, these components are usually subjected to extreme working conditions, what requires high heat resistant alloys, such as titanium or nickel-based alloys.These materials are characterized for combining hardness with low thermal conduction and good ductility [3,4].However, these alloys are known as difficult-to-cut materials, implying premature tool wear and high cutting forces [5][6][7].Additionally, in the particular case of Blisk, difficulties are not only presented in blade manufacturing processes, but also as an extra challenge consists of fir-trees junctions manufacturing.This is due to the fact that these geometries are crucial for the functionality of the parts, and they possess complex geometry, together with very tight dimensional and finishing tolerances [8].On the other hand, fir-trees have been traditionally manufactured by Electro Discharge Machining (EDM).This non-conventional manufacturing technology stands out for facilitating complex geometrical cavities of high hardness materials with dimensional accuracy and excellent finishing surface [9][10][11].Notwithstanding, the main inconvenience of this technology is based on low material removal rates, what means elevated manufacturing time.For this reason, some studies look for innovative technologies to cover this deficiency.Among these technologies, Super Abrasive Machining (SAM) was presented in [12] as a solution to increase machining efficiency during the production of blades and turbine disks.
In particular, SAM is characterized by using grinding technology with machining rates.Therefore, this technology provides the precision of grinding process with similar feeds and costs to the use of single point machining.Moreover, SAM is more versatile than other grinding techniques.For example, if it is compared with creep fatigue grinding (the closest rival), the use of SAM achieves higher speeds, higher material removal rates-up to 1000 mm 3 /s-with lower workpiece loads, and more accurate dimensional tolerances [12,13] is what makes this technology a suitable and efficient alternative to manufacture IBRs nickel-based super alloys [14,15].Its application to blisk fir-trees has been realized [16,17], as it is shown in Figure 2, due to the fact that the use of this technique increases efficiency significantly.It should be noted that during its entire lifetime, these components are usually subjected to extreme working conditions, what requires high heat resistant alloys, such as titanium or nickel-based alloys.These materials are characterized for combining hardness with low thermal conduction and good ductility [3,4].However, these alloys are known as difficult-to-cut materials, implying premature tool wear and high cutting forces [5][6][7].Additionally, in the particular case of Blisk, difficulties are not only presented in blade manufacturing processes, but also as an extra challenge consists of fir-trees junctions manufacturing.This is due to the fact that these geometries are crucial for the functionality of the parts, and they possess complex geometry, together with very tight dimensional and finishing tolerances [8].On the other hand, fir-trees have been traditionally manufactured by Electro Discharge Machining (EDM).This non-conventional manufacturing technology stands out for facilitating complex geometrical cavities of high hardness materials with dimensional accuracy and excellent finishing surface [9][10][11].Notwithstanding, the main inconvenience of this technology is based on low material removal rates, what means elevated manufacturing time.For this reason, some studies look for innovative technologies to cover this deficiency.Among these technologies, Super Abrasive Machining (SAM) was presented in [12] as a solution to increase machining efficiency during the production of blades and turbine disks.
In particular, SAM is characterized by using grinding technology with machining rates.Therefore, this technology provides the precision of grinding process with similar feeds and costs to the use of single point machining.Moreover, SAM is more versatile than other grinding techniques.For example, if it is compared with creep fatigue grinding (the closest rival), the use of SAM achieves higher speeds, higher material removal rates-up to 1000 mm 3 /s-with lower workpiece loads, and more accurate dimensional tolerances [12,13] is what makes this technology a suitable and efficient alternative to manufacture IBRs nickel-based super alloys [14,15].Its application to blisk fir-trees has been realized [16,17], as it is shown in Figure 2, due to the fact that the use of this technique increases efficiency significantly.
Furthermore, SAM was tested with other more complex geometries, such as the blades from IBRs or impellers.The main advantages are higher material removal rates at high speed [18] and a near shape surface.In fact, Rolls-Royce claims, under the correct performance, that the process is capable of stock removal at a rate of 80 cubic millimeters per second per millimeter of wheel width.That is, eight times the achievable rate using plated CBN wheel technology for super abrasive machining of nickel alloys on a conventional grinding machine.The process can also achieve higher removal higher rates for fishing grinding operations than alternative methods [19].Additionally, it is noteworthy that using these kinds of tools, due to process temperature and extreme cutting conditions, cutting fluids Metals 2018, 8, 24 3 of 11 are required, particularly cutting oil [20][21][22].Regarding cutting conditions, in line with [23], they can be optimized with the aim of developing the exact windows parameters.Moreover, another restrictive requirement is having appropriate equipment with high speed spindles (60,000-90,000 rpm) [17], what makes this process unachievable to the most part of the current machining centers whose spindles rotary speed capacity are below 24,000 rpm.Furthermore, SAM was tested with other more complex geometries, such as the blades from IBRs or impellers.The main advantages are higher material removal rates at high speed [18] and a near shape surface.In fact, Rolls-Royce claims, under the correct performance, that the process is capable of stock removal at a rate of 80 cubic millimeters per second per millimeter of wheel width.That is, eight times the achievable rate using plated CBN wheel technology for super abrasive machining of nickel alloys on a conventional grinding machine.The process can also achieve higher removal higher rates for fishing grinding operations than alternative methods [19].Additionally, it is noteworthy that using these kinds of tools, due to process temperature and extreme cutting conditions, cutting fluids are required, particularly cutting oil [20][21][22].Regarding cutting conditions, in line with [23], they can be optimized with the aim of developing the exact windows parameters.Moreover, another restrictive requirement is having appropriate equipment with high speed spindles (60,000-90,000 rpm) [17], what makes this process unachievable to the most part of the current machining centers whose spindles rotary speed capacity are below 24,000 rpm.
In this line, the novelty of this work stems from the idea of using SAM technology instead of conventional milling technology for roughing and finishing IBR blades geometry with non-ruled surfaces, (a) using conventional machining centers with a spindle, which does not reach the high rotary speeds needed, and (b) combined with oil emulsions instead of mineral cutting oils.For this, cutting forces, surface finishing, and dimension accuracy using SAM technology were analyzed and compared with conventional milling process.

Experimental Setup and Tests Performance
Experimental tests were carried out in a five-axis machining center, Ibarmia ZV-25/U600 model (IBARMIA INNOVATEK, S.L.U., Guipuzkoa, Spain).This machine consists of five-axis divided into three linear axes (X, Y, Z), two rotary axes (A, C), and a spindle speed capacity of 18,000 rpm, 18 KW.Regarding machining processes, in order to compare the two different techniques (SAM and conventional milling), roughing and finishing operations with both techniques were carried out.During machining process, cutting forces were recorded with tri-axial force transducer piezoelectric dynamometer, Kistler 9255 and OROS ® OR35 analyzer (OROS, Inovallee, France), with a sampling frequency of 16,384 samples per second.Figure 3 shows the experimental setup used to perform this test.In this line, the novelty of this work stems from the idea of using SAM technology instead of conventional milling technology for roughing and finishing IBR blades geometry with non-ruled surfaces, (a) using conventional machining centers with a spindle, which does not reach the high rotary speeds needed, and (b) combined with oil emulsions instead of mineral cutting oils.For this, cutting forces, surface finishing, and dimension accuracy using SAM technology were analyzed and compared with conventional milling process.

Experimental Setup and Tests Performance
Experimental tests were carried out in a five-axis machining center, Ibarmia ZV-25/U600 model (IBARMIA INNOVATEK, S.L.U., Guipuzkoa, Spain).This machine consists of five-axis divided into three linear axes (X, Y, Z), two rotary axes (A, C), and a spindle speed capacity of 18,000 rpm, 18 KW.Regarding machining processes, in order to compare the two different techniques (SAM and conventional milling), roughing and finishing operations with both techniques were carried out.During machining process, cutting forces were recorded with tri-axial force transducer piezoelectric dynamometer, Kistler 9255 and OROS ® OR35 analyzer (OROS, Inovallee, France), with a sampling frequency of 16,384 samples per second.Figure 3 shows the experimental setup used to perform this test.
Due to aggressive conditions at which IBRs are exposed, such as high temperatures and aggressive chemical environments, they are manufactured using nickel-base superalloys [24].In particular Inconel 718 was chosen.This material is a heat-resistant Ni-Fe alloy, which is hardened by precipitation of secondary phases into the metal matrix [25], achieving in this case an average value of hardness 42 HRC, with strain hardening affected areas on several surface points due to the initial saw cuts.The chemical composition, mechanical, and physical properties of tested materials are shown in Table 1.Due to aggressive conditions at which IBRs are exposed, such as high temperatures and aggressive chemical environments, they are manufactured using nickel-base superalloys [24].In particular Inconel 718 was chosen.This material is a heat-resistant Ni-Fe alloy, which is hardened by precipitation of secondary phases into the metal matrix [25], achieving in this case an average value of hardness 42 HRC, with strain hardening affected areas on several surface points due to the initial saw cuts.The chemical composition, mechanical, and physical properties of tested materials are shown in Table 1.The alloy offers good resistance to fatigue and creep combined with high corrosion resistance under high temperatures [28].Nevertheless, these properties involve high cutting forces during machining, low material removal rates, adhesions, and other issues, which enhance premature tool wear [28][29][30][31][32][33].In the light of these concerns, selecting this material for experimental case of study implies aggressive machining conditions and becomes suitable in order to test SAM behavior and establish a comparison with conventional milling process.
Regarding tools used during the tests, three types of tools were chosen.In the case of roughing by milling, four teeth carbide tool coated with AlTiN of 16 mm of diameter and 40 mm of cutting length was used.On the other hand, for the finishing operations the tool diameter used was of 6 mm with six teeth, 26 mm of cutting length and coated with TiN/TiAlN.In the case of the first tool, AlTiN coating was chosen because presents higher surface hardness than TiAlN coating and the high presence of aluminum offers better resistance to oxidation in comparison with other coatings [34].However, the finishing mill was chosen with TiN/TiAlN multi-layer coating with the aim of obtaining a balance between surface roughness and cutting temperature resistance [34].Nevertheless, it must be taken into account that recently other PVD coatings based on  The alloy offers good resistance to fatigue and creep combined with high corrosion resistance under high temperatures [28].Nevertheless, these properties involve high cutting forces during machining, low material removal rates, adhesions, and other issues, which enhance premature tool wear [28][29][30][31][32][33].In the light of these concerns, selecting this material for experimental case of study implies aggressive machining conditions and becomes suitable in order to test SAM behavior and establish a comparison with conventional milling process.
Regarding tools used during the tests, three types of tools were chosen.In the case of roughing by milling, four teeth carbide tool coated with AlTiN of 16 mm of diameter and 40 mm of cutting length was used.On the other hand, for the finishing operations the tool diameter used was of 6 mm with six teeth, 26 mm of cutting length and coated with TiN/TiAlN.In the case of the first tool, AlTiN coating was chosen because presents higher surface hardness than TiAlN coating and the high presence of aluminum offers better resistance to oxidation in comparison with other coatings [34].However, the finishing mill was chosen with TiN/TiAlN multi-layer coating with the aim of obtaining a balance between surface roughness and cutting temperature resistance [34].Nevertheless, it must be taken into account that recently other PVD coatings based on nc-AlCrN/a-Si 3 N 4 , AlTiCrN with a nanocomposite top layer, TiAlCrN or AlTiCrSiYN/AlTiCrN with 55% of Al in the Si + Y layers were successfully tested for nickel based alloys machining [35][36][37][38].
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2. nc-AlCrN/a-Si3N4, AlTiCrN with a nanocomposite top layer, TiAlCrN or AlTiCrSiYN/AlTiCrN with 55% of Al in the Si + Y layers were successfully tested for nickel based alloys machining [35][36][37][38].
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations ( 1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ1 = 1, µ2 = µ4 = 0.1, and µ3 = 0.001; Equations (2) and (4) correspond to the two components of the objective function that represent the point-surface proximity.Equation (5) controls the rigidity and Equation ( 3) is the fairness of the two boundary curves to achieve a fair motion.nc-AlCrN/a-Si3N4, AlTiCrN with a nanocomposite top layer, TiAlCrN or AlTiCrSiYN/AlTiCrN with 55% of Al in the Si + Y layers were successfully tested for nickel based alloys machining [35][36][37][38].
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations (1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ1 = 1, µ2 = µ4 = 0.1, and µ3 = 0.001; Equations (2) and (4) correspond to the two components of the objective function that represent the point-surface proximity.Equation (5) controls the rigidity and Equation ( 3) is the fairness of the two boundary curves to achieve a fair motion.nc-AlCrN/a-Si3N4, AlTiCrN with a nanocomposite top layer, TiAlCrN or AlTiCrSiYN/AlTiCrN with 55% of Al in the Si + Y layers were successfully tested for nickel based alloys machining [35][36][37][38].
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations (1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ1 = 1, µ2 = µ4 = 0.1, and µ3 = 0.001; Equations (2) and (4) correspond to the two components of the objective function that represent the point-surface proximity.Equation (5) controls the rigidity and Equation (3) is the fairness of the two boundary curves to achieve a fair motion.nc-AlCrN/a-Si3N4, AlTiCrN with a nanocomposite top layer, TiAlCrN or AlTiCrSiYN/AlTiCrN with 55% of Al in the Si + Y layers were successfully tested for nickel based alloys machining [35][36][37][38].
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations (1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ1 = 1, µ2 = µ4 = 0.1, and µ3 = 0.001; Equations (2) and (4) correspond to the two components of the objective function that represent the point-surface proximity.Equation (5) controls the rigidity and Equation (3) is the fairness of the two boundary curves to achieve a fair motion.
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations (1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ1 = 1, µ2 = µ4 = 0.1, and µ3 = 0.001; Equations (2) and (4) correspond to the two components of the objective function that represent the point-surface proximity.Equation (5) controls the rigidity and Equation (3) is the fairness of the two boundary curves to achieve a fair motion.nc-AlCrN/a-Si3N4, AlTiCrN with a nanocomposite top layer, TiAlCrN or AlTiCrSiYN/AlTiCrN with 55% of Al in the Si + Y layers were successfully tested for nickel based alloys machining [35][36][37][38].
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations (1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ1 = 1, µ2 = µ4 = 0.1, and µ3 = 0.001; Equations (2) and (4) correspond to the two components of the objective function that represent the point-surface proximity.Equation (5) controls the rigidity and Equation (3) is the fairness of the two boundary curves to achieve a fair motion.nc-AlCrN/a-Si3N4, AlTiCrN with a nanocomposite top layer, TiAlCrN or AlTiCrSiYN/AlTiCrN with 55% of Al in the Si + Y layers were successfully tested for nickel based alloys machining [35][36][37][38].
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations ( 1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ1 = 1, µ2 = µ4 = 0.1, and µ3 = 0.001; Equations ( 2) and ( 4) correspond to the two components of the objective function that represent the point-surface proximity.Equation ( 5) controls the rigidity and Equation ( 3) is the fairness of the two boundary curves to achieve a fair motion.nc-AlCrN/a-Si3N4, AlTiCrN with a nanocomposite top layer, TiAlCrN or AlTiCrSiYN/AlTiCrN with 55% of Al in the Si + Y layers were successfully tested for nickel based alloys machining [35][36][37][38].
In the case of SAM, the same tool was used in both operations; a PCBN grinding tool of 16 mm of diameter and 20 mm of cutting edge was used.On the other hand, concerning cutting conditions, in order to allow for using SAM technology in this type of machining centers, they were adapted to the spindle capacities.Finally, the cutting fluid chosen to be able to use the grinding tool was a synthetic emulsion with a concentration of 20%.Tools characteristics, cutting conditions, and operation times are summarized in Table 2.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations ( 1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ1 = 1, µ2 = µ4 = 0.1, and µ3 = 0.001; Equations ( 2) and ( 4) correspond to the two components of the objective function that represent the point-surface proximity.Equation ( 5) controls the rigidity and Equation ( 3) is the fairness of the two boundary curves to achieve a fair motion.In relation to machining strategies, in the case of roughing with milling tools several passes were used to avoid tool break failure.However, with SAM technology it is possible to perform this operation in one only pass/step.On the other hand, in order to carry out finishing strategies, a mathematical algorithm developed in 2016 by Pengbo Bo et al. [39] was used.The use of this algorithm is needed to obtain the surface dimensional requirements in this type of parts where it not allowed any section break/jump between different tool paths along blade surface.These applications occur frequently and this algorithm offers a full solution to perform the machining operations just with one tool path.
This algorithm focuses on approximation of free form surfaces by envelopes of motions of surfaces of revolution (milling tools).This algorithm is perfectly suited for flank milling purposes.However, their approach optimized both the shape of the milling and its trajectory.In this set up, the shape of the (virtual) milling tool was fixed according to the existing (physical) counterpart.The equations that control this algorithm are shown in Equations (1)-( 5).In particular, taking "a" and "b" as the two boundary curves and "d" as the unknown distance, Equation (1) represents the whole objective function to obtain this approximation where µ 1 = 1, µ 2 = µ 4 = 0.1, and µ 3 = 0.001; Equations ( 2) and ( 4) correspond to the two components of the objective function that represent the point-surface proximity.Equation ( 5) controls the rigidity and Equation ( 3) is the fairness of the two boundary curves to achieve a fair motion.
Metals 2018, 8, 24 6 of 11 In this way, the full length of tool cutting edge is constantly in contact with the non-ruled surface designed and layers are confined to a single one if tool dimension covers all of the surface height.The accuracy of the process is controlled by the error of the objective function (see Equation ( 1)).Therefore, to take this into account, the surface height used for these tests was 20 mm, that is, the same height dimension of the cutting tool edges.
Finally, after machining tests were conducted, in order to validate SAM technique from an industrial point of view, the blades obtained were scanned with an ATOS GOM and compared with the Computer-aided design (CAD) model.This equipment is based on the triangulation effect with two cameras with a resolution of 17 µm and an accuracy of 35 µm.Afterwards, besides three-dimensional (3D) surface topography of the walls were obtained using a Leica confocal microscope with a resolution of 0.1 nm.

Results and Discussion
Each blade was made five times with both techniques in order to validate the results obtained, and the test finishing criteria was taken the finalization of each blade.In Figure 4, the IBR blades machined are shown during roughing and finishing operations with milling and SAM operations.In this this section, the analysis of different results obtained during the experimental tests are set out.In particular, cutting forces and surface roughness are compared between both operations.Afterwards, based on these analyzed results, a scanned IBR blade obtained by SAM is compared to CAD model.The aim of analyzing these parameters is to demonstrate that SAM process is not only a process in which manufacturing time is reduced, but also is technologically a suitable option to be used in the industry because it improves the current milling process.
Metals 2018, 8, 24 In this way, the full length of tool cutting edge is constantly in contact with the non-ruled surface designed and layers are confined to a single one if tool dimension covers all of the surface height.The accuracy of the process is controlled by the error of the objective function (see Equation (1)).Therefore, to take this into account, the surface height used for these tests was 20 mm, that is, the same height dimension of the cutting tool edges.
Finally, after machining tests were conducted, in order to validate SAM technique from an industrial point of view, the blades obtained were scanned with an ATOS GOM and compared with the Computer-aided design (CAD) model.This equipment is based on the triangulation effect with two cameras with a resolution of 17 µm and an accuracy of 35 µm.Afterwards, besides three-dimensional (3D) surface topography of the walls were obtained using a Leica confocal microscope with a resolution of 0.1 nm.

Results and Discussion
Each blade was made five times with both techniques in order to validate the results obtained, and the test finishing criteria was taken the finalization of each blade.In Figure 4, the IBR blades machined are shown during roughing and finishing operations with milling and SAM operations.In this this section, the analysis of different results obtained during the experimental tests are set out.In particular, cutting forces and surface roughness are compared between both operations.Afterwards, based on these analyzed results, a scanned IBR blade obtained by SAM is compared to CAD model.The aim of analyzing these parameters is to demonstrate that SAM process is not only a process in which manufacturing time is reduced, but also is technologically a suitable option to be used in the industry because it improves the current milling process.

Cutting Forces
As it was exposed above, forces were registered during IBR blade machining through roughing and finishing operations with milling and SAM techniques.Figure 5 shows total average and maximum total force modulus obtained during the machining processes, as well as estimated errors, which are between ≈2% and ≈7%.Due to the fact that obtained values are below 10%, they are considered admissible [40].

Cutting Forces
As it was exposed above, forces were registered during IBR blade machining through roughing and finishing operations with milling and SAM techniques.Figure 5 shows total average and maximum total force modulus obtained during the machining processes, as well as estimated errors, which are between ≈2% and ≈7%.Due to the fact that obtained values are below 10%, they are considered admissible [40].As it can be appreciated on the plot, most significant differences between these techniques are recorded during roughing process.It is due to the fact that cutting conditions are more aggressive under these circumstances.SAM medium values overpass 1100 N with a maximum peak over 2500 N.This implies an increase ≈40% and ≈100%, respectively, against traditional milling.This behavior is a result of using one tool pass for Flank SAM when compared to conventional milling with several tool passes; what can be translated into higher material removal rates during SAM and higher forces, but at the same time, a decrease of machining time.In particular, the material removal rate achieved with SAM was 240 mm 3 /s what supposes an increase of ≈375% in comparison with conventional milling.These MRR values were calculated in the CAM stage taking into account the material to remove and tool feed as the volume to remove in the operation time.In the non-distant future, it could be calculated by the integral blade rotor weight reduction divided by production time.Nevertheless, finishing process the difference between the two techniques was inverted, obtaining lower force values with SAM than with milling.Concretely, this process presents medium forces around 600 N with a peak of 1200 N. Thus, in comparison with conventional milling, forces are reduced 20% and 15%, respectively.

Surface Roughness
Regarding surface roughness, the topography and data obtained with the confocal microscope are shown in Figure 6.
Topography in both cases shows a typical surface milling and grinding pattern perfectly generated.This implies stable and controllable cutting process in both cases.However, the differences between both of the technologies are caused by the values obtained in the surface roughness.In the case of milling, the blade surface presents an average roughness (Ra) of 4.85 µm.This value is drastically reduced when SAM technique is used.In this case, the value obtained is 2.66 µm, that is, ≈45% less average roughness in comparison with the milling operation.This behavior is preserved in the mean values of five consecutive maximum heights between peak-valley (Rz).In particular, the values obtained were 49.79 µm and 36.03µm in the case of milling and SAM machining, respectively.This represents a reduction of ≈28% when SAM is used as machining technology.
Therefore, combining the results obtained regarding cutting forces and surface roughness, SAM technology is suitable for being used to machine IBR blades.With this technology, not only machining times are improved, but also smaller cutting forces during finishing process and better As it can be appreciated on the plot, most significant differences between these techniques are recorded during roughing process.It is due to the fact that cutting conditions are more aggressive under these circumstances.SAM medium values overpass 1100 N with a maximum peak over 2500 N.This implies an increase ≈40% and ≈100%, respectively, against traditional milling.This behavior is a result of using one tool pass for Flank SAM when compared to conventional milling with several tool passes; what can be translated into higher material removal rates during SAM and higher forces, but at the same time, a decrease of machining time.In particular, the material removal rate achieved with SAM was 240 mm 3 /s what supposes an increase of ≈375% in comparison with conventional milling.These MRR values were calculated in the CAM stage taking into account the material to remove and tool feed as the volume to remove in the operation time.In the non-distant future, it could be calculated by the integral blade rotor weight reduction divided by production time.Nevertheless, finishing process the difference between the two techniques was inverted, obtaining lower force values with SAM than with milling.Concretely, this process presents medium forces around 600 N with a peak of 1200 N. Thus, in comparison with conventional milling, forces are reduced 20% and 15%, respectively.

Surface Roughness
Regarding surface roughness, the topography and data obtained with the confocal microscope are shown in Figure 6.
Topography in both cases shows a typical surface milling and grinding pattern perfectly generated.This implies stable and controllable cutting process in both cases.However, the differences between both of the technologies are caused by the values obtained in the surface roughness.In the case of milling, the blade surface presents an average roughness (R a ) of 4.85 µm.This value is drastically reduced when SAM technique is used.In this case, the value obtained is 2.66 µm, that is, ≈45% less average roughness in comparison with the milling operation.This behavior is preserved in the mean values of five consecutive maximum heights between peak-valley (R z ).In particular, the values obtained were 49.79 µm and 36.03µm in the case of milling and SAM machining, respectively.This represents a reduction of ≈28% when SAM is used as machining technology.
Therefore, combining the results obtained regarding cutting forces and surface roughness, SAM technology is suitable for being used to machine IBR blades.With this technology, not only machining times are improved, but also smaller cutting forces during finishing process and better surface roughness values are obtained.However, another parameter that has to be taken into account is the strict accuracy needed to satisfy the requirements in this type of aeronautical components.surface roughness values are obtained.However, another parameter that has to be taken into account is the strict accuracy needed to satisfy the requirements in this type of aeronautical components.

Dimensional Deviation
In Figure 7   The results registered present tiny differences, in terms of approximation quality, between the IBR blade obtained by SAM and the CAD model.It should be noted that the tolerances in this type of turbomachinery components need values below 50 µm [41].In this case, deviations that are obtained are below this value.Nevertheless, overcut is shown in a small area located near the corner.This behavior was found due to the fact that it consists of a non-developable surface, which has to be dealt with a free-form tool to adequate the flank to the defined surface [42].Thus, the process presented is able to satisfy the requirements in this type of turbomachinery components.
Therefore, results obtained in these experiment show, from a technical point of view related with cutting forces, surface roughness, and dimensional accuracy, which SAM technology does not present a limitation for being used with conventional machines (similar to the one used for this case); as long as cutting conditions were adequately adapted to spindle rotary capacity.

Conclusions
In this work, Inconel 718 IBR blades were manufactured with the main objective of comparing milling and SAM techniques in conventional machining centers using oil emulsions instead of mineral cutting oils, reducing its use in ≈80%.For this, both techniques were compared from a  surface roughness values are obtained.However, another parameter that has to be taken into account is the strict accuracy needed to satisfy the requirements in this type of aeronautical components.

Dimensional Deviation
In Figure 7   The results registered present tiny differences, in terms of approximation quality, between the IBR blade obtained by SAM and the CAD model.It should be noted that the tolerances in this type of turbomachinery components need values below 50 µm [41].In this case, deviations that are obtained are below this value.Nevertheless, overcut is shown in a small area located near the corner.This behavior was found due to the fact that it consists of a non-developable surface, which has to be dealt with a free-form tool to adequate the flank to the defined surface [42].Thus, the process presented is able to satisfy the requirements in this type of turbomachinery components.
Therefore, results obtained in these experiment show, from a technical point of view related with cutting forces, surface roughness, and dimensional accuracy, which SAM technology does not present a limitation for being used with conventional machines (similar to the one used for this case); as long as cutting conditions were adequately adapted to spindle rotary capacity.

Conclusions
In this work, Inconel 718 IBR blades were manufactured with the main objective of comparing milling and SAM techniques in conventional machining centers using oil emulsions instead of mineral cutting oils, reducing its use in ≈80%.For this, both techniques were compared from a The results registered present tiny differences, in terms of approximation quality, between the IBR blade obtained by SAM and the CAD model.It should be noted that the tolerances in this type of turbomachinery components need values below 50 µm [41].In this case, deviations that are obtained are below this value.Nevertheless, overcut is shown in a small area located near the corner.This behavior was found due to the fact that it consists of a non-developable surface, which has to be dealt with a free-form tool to adequate the flank to the defined surface [42].Thus, the process presented is able to satisfy the requirements in this type of turbomachinery components.
Therefore, results obtained in these experiment show, from a technical point of view related with cutting forces, surface roughness, and dimensional accuracy, which SAM technology does not present a limitation for being used with conventional machines (similar to the one used for this case); as long as cutting conditions were adequately adapted to spindle rotary capacity.

Conclusions
In this work, Inconel 718 IBR blades were manufactured with the main objective of comparing milling and SAM techniques in conventional machining centers using oil emulsions instead of mineral cutting oils, reducing its use in ≈80%.For this, both techniques were compared from a technical point of view.In particular, cutting forces, surface roughness and dimensional accuracy were analyzed.The main conclusions obtained are listed below:

•
Flank SAM technique applied to roughing operations presented higher material removal rates when compared with conventional milling process, reaching 240 mm 3 /s.This means that the process has been optimized through manufacturing time reduction.

•
Conversely, this increase of material removal rates is governed by the increase of cutting forces during SAM roughing in a ≈40% and ≈100% for medium and maximum force values, respectively.Nonetheless, in respect to finishing tasks, where material removal rates are similar for both techniques, SAM showed lower cutting forces.In particular, the values are reduced ≈20% and ≈15% for medium and maximum force values, respectively.

•
Regarding surface roughness, the topography obtained with both techniques presents a regular pattern that is associated to each technology.However, the surface roughness values present high differences between them.Specifically, the use of SAM technique implies a reduction of ≈45% and ≈28% in R a and R z , respectively.

•
Concerning dimensional accuracy, the use of SAM generates deviations below the current aeronautical requirements in this type of components.
Therefore, the use of SAM as machining technology implies better results than conventional milling obtaining a suitable, controllable and predictable process to manufacture aeronautical critical components in heat-resistant super alloys, such as Inconel 718, IBR blades presented in this work.Moreover, with this work, the two main restrictions for using SAM as machining process in conventional machining centers with oil emulsions has been avoided.

Figure 3 .
Figure 3. Experimental Setup on Ibarmia machining center, showing the IN718 circular sector.

Figure 3 .
Figure 3. Experimental Setup on Ibarmia machining center, showing the IN718 circular sector.

Figure 5 .
Figure 5. Average and maximum force modulus obtained during blades manufacturing.

Figure 5 .
Figure 5. Average and maximum force modulus obtained during blades manufacturing.

Figure 6 .
Figure 6.Surface roughness in both processes.
results obtained of comparing the IBR blade manufactured by SAM technology with the CAD model are shown.

Figure 6 .
Figure 6.Surface roughness in both processes.

3. 3 .
Dimensional Deviation In Figure 7 results obtained of comparing the IBR blade manufactured by SAM technology with the CAD model are shown.Metals 2018, 8, 24 8 of 11

Figure 6 .
Figure 6.Surface roughness in both processes.
results obtained of comparing the IBR blade manufactured by SAM technology with the CAD model are shown.

Table 2 .
Tools and cutting conditions.

Table 2 .
Tools and cutting conditions.

Table 2 .
Tools and cutting conditions.

Table 2 .
Tools and cutting conditions.

Table 2 .
Tools and cutting conditions.

Table 2 .
Tools and cutting conditions.

Table 2 .
Tools and cutting conditions.

Table 2 .
Tools and cutting conditions.

Table 2 .
Tools and cutting conditions.