Effects of cutting parameters during turning 100C6 steel

The objective of the paper is to evaluate the e ffects of cutting parameters in terms of cutting speed, depth of cut and feed rate on the influence of the s urface roughness, consumed power, cutting time and tool vibrations during turning process. The material chosen in this case was 100C6 steel in dry conditions. The e ffects of the selected process parameters have been investigated using full factorial design of e xperiments (33) and the multiple linear regression (MLR). Thus, first-order empirical models w re established. Analysis of variance (ANOVA) was employed to check the validity of the dev eloped models within the limits of the factors that were being investigated and to test the signifi cance of the above parameters. Results indicate that the feed rate is the only signifi cant factor affecting the surface roughness. The cutting speed and feed rate are the m ost influential factors on cutting time. Estimated tool vibrations are functions of cutting spee d, f d rate and depth of cut in decreasing order. Finally, the models obtained can be used for determination of optimal settings of cuttings parameters and this methodology shou ld elp us to obtain the best process parameters for dry turning of 100C6 steel.


Introduction
Even though the manufacturing industry has made an important progress, the metal cutting industries continue to suffer from a major drawback of not running the machine tools at their optimum operating conditions.Indeed, turning is one of the most commonly used machining processes in machining industry.Thus, the choice of optimized cutting parameters is very important to control the surface quality or surface roughness, a dimensional accuracy of a workpiece, the cutting time, tool life or the tool wear rates, to limit the consumed power of machine tool, to avoid high levels of cutting forces and tool vibrations.The problem of arriving at the optimum levels of the cuttings parameters has attracted the attention of the researchers and practicing engineers for a very long time.Unfortunately, the impact of the research work in this area does not seem to have reached a large majority of manufacturing engineers with the result that the operating conditions continue to be chosen on the basis of handbook values, manufacturers' recommendations or worker experience.Therefore, for lack of well defined rule, conservative cuttings conditions are usually chosen resulting in lower metal removal rates and a loss of productivity.A large number of analytical and experimental studies on surface roughness [1,2], cutting tool wear [3], cutting forces [4] and tool vibrations [5], related to turning operations has been conducted.However in these studies, the interactions between the cutting parameters were not taken into consideration in the modeling process.But the key element for achieving high quality at low cost is Design of experiments (DOE).In this paper full factorial design of experiments approach is used to evaluate the effects of cutting parameters like cutting speed, depth of cut and feed rate on the resulting surface roughness, consumed power, cutting time and tool vibrations during a dry turning EPJ Web of Conferences of 100C6 steel.Finally, this methodology should help us to obtain the best cuttings parameters for the optimization of the machining process.

Turning process parameters
In order to identify the cuttings parameters that may affect the machining characteristics of turned parts, an Ishikawa cause-effect diagram was constructed as illustrated in figure fig. 1.The process parameters affecting the characteristics of turned parts are: cutting tool parameters, workpiece properties, cutting phenomena and cutting parameters [6].Cutting tests were carried out on a CNC SOMAB model 500 lathe machine (24kW power) under dry cutting conditions.Machining was performed dry because it has been considered as the machining of the future due to concern regarding the safety of the environment [7].Because of its wide application, the 100C6 steel has been selected as the workpiece material in this study.The cutting power and time has been study to take into account the capacity of the lathe and productivity.
Process parameters selected for study are: depth of cut, feed rate and cutting speed.The range of the selected process parameters was selected after considering the recommendations given in the tool manufacturer's catalogue [8].The process parameters, their designated symbols and ranges are given in Table 1.The following parameters were kept constant during entire experimentation: work material, cutting tool insert product by Safety Company of the standard designation CNMG 1204 08 5B (OR2500) and mounted on DCLNL 2525M 12 tool holder (standard designation), tool overhang and dry environment.
14th International Conference on Experimental Mechanics 3 Methodology

Full factorial design of experiments
Experimental design is very important engineering tool for improving of process.Designed experiments can often be applied in the product design processes.This will produce information concerning which factors are most influential one and through use of this information the design can be improved [9].
In this case, a full factorial experimental design 3 3 has been selected with all combinations of the factors at three levels as illustrated in figure 2. Thus, 27 experiments were conducted at parameter levels as shown in Table 1 on the four measured dependant variables surface roughness, consumed power, cutting time and tool vibrations.The resolution of this full factorial design allows us to estimate all the main effects, factor interactions.Note that run orders are used randomly during the experiments.

Multiple linear regressions
Multiple linear regressions attempt to model the relationship between two or more explanatory variables and a response variable by fitting a linear equation to observed data.By conducting experiments and applying regression analysis, a mathematical model of the response (Y) with independent input variables (X) can be obtained.Usually a lower-order polynomial in the region of interest is employed, like the first-order model seen in equation (Eq.1): where a 0 is constant, a i,i=1,2,3 represent the coefficients of linear terms and ε is the experimental error.X i,i=1,2,3 reveals the coded variables that correspond to the selected turning parameters.The coded variables are obtained from the following transformation equations: where X 1 , X 2 and X 3 are the coded values of a p , f and V c respectively.a p 0 , f 0 , V c 0 and a p +1 , f +1 , V c +1 are the values of a p , f and V c are zero level and +1 level respectively.The surface roughness, consumed power, cutting time and tool vibrations were analyzed as responses.
The ordinary least squares method is used to estimate the parameters a i,i=0,1,2,3 within : where the calculation matrix of independent variables is X and the variance matrix is (X T X) −1 a.The calculated coefficients or the model equation need to however be tested for statistical significance.In this respect, the following tests are performed. 13004-p.3

Test of significance
Tests of significance of the regression models are performed as an Analysis of variance (ANOVA) procedure by calculating the F-ratio, which is the ratio between the regression mean square and the mean square error.The F-ratio is the ratio of variance due to the effect of the factor (in this case the model) and variance due to the error term.This ratio is used to check the validity of the models under investigation with respect to the variance of all terms included in the error term at the desired significance level α.A significant model is desired.
Test for significance on individual model coefficients is the basis for model optimization.It involves the determination of the P-value or probability value, usually relating the risk of falsely rejecting a given hypothesis.For example, a Prob.> F value on an F-test tells the proportion of time you would expect to get the stated F-value if no factor effects are significant.The Prob. > F value determined can be compared with the desired probability or α-level.
The checks performed here include determination of the correlation coefficients R 2 .These R 2 coefficients have values between 0 and 1.In addition to the above, the adequacy of the models is also investigated by examination of residuals [9].

Results and discussion
The results from the machining runs performed as per experimental plan is shown in Table 2.For a best discussion of results we have reorganized the 27 experiments (N • exp) to (run) in 3 blocks (ap = 1, 1.5, 2).

surface roughness model
The predictive linear surface roughness model Ra in terms of coded factors (Eq.4) was been transformed using (Eq.2) to provide the predictive surface roughness as a function of cutting parameters a p , f and V c (Eq.5) as follows: In the table 3, the value of Prob.> F for the term of models are less than 0.05 (95% confidence), this indicates that the obtained model are considered to be statistically significant within the limits of factors studied, which is desirable.The other important regression coefficient of the model R 2 adjusted was found to be 0.6598.This shows that the model can explain the variation in surface roughness to the extent of 65.98% and it can be concluded that the first order model was adequate to represent this process.The analysis of variance results shows that the only significant factor on the surface roughness is the feed rate f .The depth of cut a p and the cutting speed V c don't impact the surface roughness in the studied range, which could be used to improve productivity if it would not worsen the surface microstructure of the workpiece and the dimensional accuracy.Thus, the insignificant terms can be removed so as to adjust the fitted first-order model (Eq.6) which is in accordance with machining theory.

Cutting time models T
To observe the influence of cutting parameters on productivity, we have studied the models of the cutting time T : As expected, the cutting time T model is function of cutting speed V c (m/min) and feed rate f (mm/rev).Depth of cut a p is not significant.

Consumed power models Pu
The evolution of the consumed power Pu models in function of cutting parameters have studied : In accordance with the machining theory, the consumed cutting power Pu is affected by the selected parameters: cutting speed V c , feed rate f and depth of cut a p in decreasing order.

RMS accelerations models
The three models of RMS cutting axial (X), radial (Y) and tangential (Z) accelerations are described below: In these models, the order of effects is respected: first the cutting speed, then the feed rate and finally the depth of cut.The tool vibrations are greater in the tangential direction compared with the others directions axial and radial as illustrated in figure 6.The tangential accelerations A z are analyzed in table 6, the effect of cutting speed (56%) is more important than the two other effects: feed rate ( f ) and dept of cut (ap) respectively (18%) and (4%).

Discussions
In this study, globally, depth of cut is not a significant parameter.It's due to the range chosen.Thus, depth of cut has not a significant effect on surface roughness, thus indicating that a minimum cutting tool vibration can be obtained by decreasing the depth of cut without significant change in the value of surface roughness, but the decrease of depth of cut has an effect on cutting time and consumed cutting power, so the operator must choose this parameters depending on the surface roughness and productivity desired.Feed has major contribution on surface roughness (68%) and cutting time (around 60%).The others process parameters are affected at the lowest level (around 25%).The cutting speed although highly significant for the tool vibrations on X and Z directions (around 50%) and minor significance on Y direction and consumed power of machine tool (35% each one).The lowest significant level is obtained for the cutting time (24%).
13004-p.7 In the Fig. 7, it is clear that each selected parameters affect one or more process parameters in a significant manner.

Conclusions
Regression analysis is performed to find out the effects of selected cutting parameters like cutting speed, depth of cut and feed rate on the resulting surface roughness, consumed power, cutting time and tool vibrations during a dry turning of 100C6 steel.
The results of ANOVA have proved that the proposed mathematical models could adequately describe the performance indicators within the limits of the factors that are being investigated with 95% confidence interval.
The analysis of models results show that the surface roughness is influenced only by the feed rate, the cutting time increase principally with the feed rate, the consumed power is affected by a combination of cutting parameters and the tool vibration increase principally with cutting speed.The models found in this study will be used in a second part for optimizing of cutting parameters using a genetic algorithm.
This study should help the operator to choice the cutting parameters depending on the surface quality (surface roughness Ra) desired productivity (Cutting time T ) and used machine (cutting time T and consumed power Pu) with controlling vibrations (tool accelerations A x , A y , A z ).
The results of this study are valid for 100C6 steel and selected parameters and their specified ranges.Any extrapolation must be confirmed through further experimentation.

Fig. 7 .
Fig. 7. Percentage contribution of cutting parameters on various process parameters (Ra, T, Pu, A x , A y , A z ).

Table 1 .
levels of independant variables

Table 2 .
Design layout and experimental results.

Table 3 .
Analysis of variance of surface roughness model.

Table 4 .
Analysis of variance of cutting time model T .

Table 5 .
Analysis of variance of consumed cutting power model Pu.

Table 6 .
Analysis of variance of RMS acceleration model in tangential directionA z .