Effect of machining parameters on surface finish of Inconel 718 in end milling

Surface finish is an important criteria in machining process and selection of proper machining parameters is important to obtain good surface finish. In the present work effects of the machining parameters in end milling of Inconel 718 were investigated. Central composite design was used to design the total number of experiments. A Mathematical model for surface roughness has been developed using response surface methodology. In this study, the influence of cutting parameters such as cutting speed, feed rate and depth of cut on surface roughness was analyzed. The study includes individual effect of cutting parameters on surface roughness as well as their interaction. The analysis of variance (ANOVA) was employed to find the validity of the developed model. The results show that depth of cut mostly affected the surface roughness. It is also observed that surface roughness values are comparable in both dry and wet machining conditions.


Introduction
Inconel 718 is a well known nickel-based super alloy used in structural materials because of its high strengthto-weight ratio, fracture toughness, fatigue strength, corrosion resistance and ability to operate at high temperatures. Nickel alloys are broadly used in petrochemical, automotive, aerospace, marine, nuclear sectors and biomedical [1,2]. Although these materials are widely preferred, their machinability is not at the desired level. The reason for this is due to their thermal and mechanical properties, which limit the machinability of Nickel based alloys during the machining process [3,4]. Surface roughness is significantly influenced by cutting speed, feed rate and depth of cut. Therefore, it is necessary to develop predictive mathematical models in order to determine the surface roughness, in advance (before the actual machining is performed). There are a large number of tools available to develop predictive models for machining processes, namely Response Surface Methodology (RSM), Artificial Neural Network (ANN), Taguchi method, etc [5][6][7][8][9]. Among all predictive model technique, RSM is a widely used tool for designing experiments and analyzing data to predict model and to select the best mathematical model that represents the real life experimental results [10]. ANN model produced more percentage error as compare to RSM model in turning of Al 7050/10/SiCp and Al 7075 composites [11]. Choudhary and Baradie [12] developed response model for tool life surface roughness and cutting force with central composite method using RSM. They found it very useful for assessing the maximum tool life and surface finish. A good design of the experiment and an accurate predictive model need to be developed to minimize the number of experiments optimizing the cutting parameters for desired machinability.
The cutting parameters are an important criterion to obtain good machinability of Nickel-based super alloys [13][14][15]. Liaoa et al. [16] experimented end milling of Inconel 718 under various cutting speeds by cemented carbide tools for slot and side milling. The results show that cutting speed in the range of 90 to 110 m/min is appropriate for slot milling, and 55 to 135 m/min is appropriate for side milling. Devillez et al. [17] Performed wet and dry turning tests at various cutting speeds using TiCN/Al2O3 coated carbide tool inserts. The results show an acceptable surface quality for a cutting speed of 60 m/min in the dry condition. At the same cutting speed, the result also shows the same order of residual stress and micro hardness for both dry and wet conditions. Zhaopeng et al. [18] studied tool wear pattern in the dry machining of Inconel 718 with PVD coated cemented carbide tool. Their study shows that build up edge (BUE) was formed at a low cutting speed of 20 m/min which resulted in chipping and tool wear. While at a high cutting speed range of 45 m/min to 50 m/min, the wear debris diffused between the tool and the work piece, which accelerated the BUE and peeling wear debris.
Coolants and its influence on machinability have been discussed thoroughly in available literature [19,20]. Yazid et al. [21] investigated the effect of cutting parameters and machining conditions on surface integrity for turning Inconel 718. They reported that at low speed of 90 m/min and 120 m/min, 50 ml/h of lubricant shows a better surface roughness compared to 100 ml/h lubricant and the dry condition. Kamata and Obikawa [22] applied MQL while using three different coated carbide tools TiCN/Al2O3/TiN (CVD), TiN/AlN superlattice (PVD) and TiAlN (PVD) for Inconel 718 in turning and reported that air pressure is an important parameter influencing tool life and surface finish. Zhang et al. [23] investigated the effects of different cutting conditions on the tool life and cutting forces in end milling Inconel 718 with a combination of TiAlN and TiN coated tools. They reported a combination of cryogenic compressed air and the microdroplets of biodegradable vegetable oil significantly reduced the tool wear compared to the dry cutting condition. Pusaveca et al. [24] examined the effects of cutting fluid on surface integrity in turning of Inconel 718. The result shows that the combination of cryogenic machining and cutting fluid provides a better surface finish compared to cutting fluid alone.

Experimental designs
In all conventional machining process, the most influential machining parameters are cutting speed, feed rate and depth of cut. A proper combination of machining parameters is very important to achieve good machinability. The design of experiments is done by using Central Composite Design (CCD) technique to minimize the number of experiments For CCD, the number of experiments required can be determined from Eq. 1.
Where n represents the number of experimentations, k is the total number of parameters considered.
The experimental results are used to develop predictive models by using response surface method for surface roughness. The proposed relationship between the output response and input variables can be represented by Eq. 2 R=b 0 +b 1 ×A+b 2 ×B+b 3 ×C+b 11 ×A 2 +b 22 ×B 2 +b 33 ×C 2 ×b 12 ×AB+b 23 ×BC+b 13 ×AC (2) Where R is the response and A, B, and C are the different machining factors. The regression coefficients bi=1, 2, 3 is computed by the regression method using experimental results. The significance of factors and their interactions is computed using statistical analysis. Optimum response model and setting of parameters are produced by using the response model in Eq. 2. A total 20 run or set of experiments were conducted with different cutting parameters on basis of CCD and the surface finish is recorded for each run. Table 1 shows the range of cutting parameters used in the experiments. The adequacy of the quadratic model evaluated using the analysis of variance (ANOVA). Design expert 9.0 was used to establish the design matrix and to analyze the experimental data.

Experimental result and discussion
To predict the surface roughness in end milling of Inconel 718, the surface roughness value of each run was input and RSM was performed. Table 2 and Table 3 shows the CCD matrix, with variable cutting parameters and experimental surface roughness for a dry condition and wet conditions. Based on the experimental results second order predictive models have been developed for surface roughness by using Response Surface Method. The obtained equations for the response factors are given below in equation 3. The Fit and summary test suggested that model is significant for dry condition.
The developed model is verified by using ANOVA and results are shown in Table 3 Figure 2. Indicates that the surface roughness increases with an increase of speed and feed rate. Figure 3. Shows the influence of cutting speed and depth of cut on surface roughness. The trend of the graph clearly indicates that the surface roughness increases with the cutting speed and depth of cut. But surface roughness increase rapidly with increase in depth of cut compare to cutting speed. Figure 4 shows the effect of feed rate and depth of cut on surface roughness. It can be observed that the surface roughness increases with the feed rate and depth. More specifically, the depth of cut has more effect on surface roughness as compared to the feed rate. Figure 5 and 6 shows plot surface roughness Vs A, B at 0.3 depth of cut. Figure 7 and 8 shows plot surface roughness Vs A,C at 125 mm/min feed rate. Surface roughness is low in intermediate cutting speed but is can be observed it increases with increase in depth of cut. Figure 9 and 10 shows plot surface roughness Vs B, C at 1850 rev/min. The trend is simillar when feed rate and depth of cut increases surface roughness also increases.
The surface roughness is highly affected by cutting parameters. For surface roughness, it is observed that depth of cut is the most influential parameter followed by cutting speed and feed rate. The result shows surface roughness is comparable while machining in wet condition and dry machining condition. The lowest surface roughness observed at cutting speed 1850 rev/mm, depth of cut 0.3 μm and feed rate 125 mm/min under wet condition. Whereas lowest surface roughness at cutting speed 1850 rev/mm, depth of cut 0.3 μm and feed rate 85 mm/min under dry condition.

Conclusions
In the present work, the effect of machining parameters (cutting speed, feed rate, depth of cut) on surface roughness was experimentally investigated. RSM used to develop the predictive mathematical model for machining of Inconel 718. The validity of the model is verified using ANOVA. The results show low surface roughness at lower cutting parameters in the selected range. The depth of cut has a higher influence on surface roughness compared to other two parameters (cutting speed and feed rate). Experiential results show, obtained surface roughness values are comparable in dry machining with coolant machining.