14 RESEARCH AND INNOVATION Finally, surface finish is measured for finished component (roughing and semi-finishing surface finish values are not obtai- ned). ATOS 5M GOM scanning system is used for surface quality measurement. This equipment is based on the triangulation effect with two cameras. In this case surface finish values are 11 μm for blade finishing and 17 μm for hub finishing. The machining process time decreases in a considerable way in comparison with the conventional milling, from 10 minutes for each blade to 2 minutes. Besides, the obtained surface finish qua- lity it is better, being 7 μm. 3. Super Abrasive Machining (SAM) In order to consider SAM process as a competitive alternative against milling, in previous work made by the team, an IBR blade was manufactured using this process in both roughing and finis- hing operations. Figure 7 shows the comparison between the CAD model and the blade manufactured by SAM [20]. Figure 7: Dimensional deviation from nominal geometry with SAM technology [13]. Owing to analyse tool wear mechanisms when using different cooling techniques (Minimum Quantity of Lubrication (MQL) and a conventional oil emulsion coolant) in the machining of Inconel718, full slots were machined. On the one hand, using MQL technique was not enough for the machining of the full slots in Inconel718. The high temperature achieved during the process due to the lack of refrigeration of this lubrication technique caused a colour modification that changed the mechanical properties of the core tool. In fact, it caused the breakage of the bottom part of the tool. Therefore, it was conclu- ded that SAM technology with MQL is not a feasible solution On the other hand, during the machining with the oil emulsion coo- lant, the main wear mechanisms were heavy loadings of machined material in the bottom of the grinding tool and some grain detach- ment all over the tool. Despite the aggressiveness of the process, the benefits of lubrication and refrigeration of the coolant help the grinding tool from reaching an excessive temperature harmful for it. In the next table, the three case studies mentioned above are summarised: Figure 5:(a) blisk semi-finish process; (b) finished blisk component. 2. Algorithm based milling Due to programming limitations generally associated to CAM software, and, in order to be able to obtain an optimized milling strategy for finishing operation, algorithm based strategy is deve- lop in this section. The algorithm is based on the approximation between the desig- ned surface and the tool conical envelope. Automatic detection of conical envelopes [19] is adopted, showing that this initialization strategy, when incorporated to real manufacturable process, reduces the milling time significantly by detecting large envelopes within fine machining tolerances. Milling simulations are conducted in a five-axis KONDIA HS1000 machining center, being numerically controlled by Heidenhain iTNC530. Manufactured geometry are Ti6Al4V blisk blades samples (Figure 6b). In this case, roughing and semi-finishing ope- rations do not present research interest, being the blade finishing strategy the studied operation. Operations Tool Process parameters (F,S, ap, ae) Strategy Finishing Tapered ball nose end ø1.5 Ø 3° 500 mm/min, 6000 rpm, 24mm (tool length), 0.2 mm Flank milling (algorithm) Technology Finishing time Total cost Surface quality Conventional milling 3hour 3minutes 2280 € 11 μm Algorithm based milling 36 minutes 2090 € 6 μm Super Abrasi- ve machining 4,2 minutes 1430 € 7 μm Figure 6: a) tool axis motion and color-coded approximation; (b) manufactured blisk blades. Table 3: Comparison of the three cases analysed