
International Journal of ChemTech Research CODEN (USA): IJCRGG ISSN: 09744290 Vol.8, No.12 pp 304311, 2015

Pulsed CO2 Laser Cutting of Al/SiCp Composite sheets
P.L.Arun*, D.Elancheziyan, H.Ravi kumar, S.Thileepan
Department of Mechanical Engineering, Saveetha Engineering College, Chennai602105, Tamilnadu, India
Abstract: Metal matrix composites (MMCs) are widely used in aerospace and automotive industries. Attaining a decent surface texture while cutting these advanced materials is challenging and hence the research attention is focused towards the application of pulsed CO2 laser cutting process on Al/SiCp composite. The process parameters in laser cutting like power, frequency, cutting speed and gas pressure affect the quality of cut surface. Surface finish and kerf width are observed as the quality characteristics for various combinations of input parameters with the experimental trials planned as per Taguchi’s L9 orthogonal array. A combined technique of grey desirability analysis (GDA) is presented for multi response optimization. Significant improvements in the responses are observed with the optimal setting of parameters permitting the usage of GDA technique in experimental welding optimization. Laser power and pulsing frequency are found to significantly affect the quality characteristics studied in the process.
Keywords: Grey relational analysis; Desirability analysis; Laser cutting; Optimization; Al/SiCp composite.
1 Introduction
Aluminium and its alloys are the widely used materials, next to steel in automotive and maritime applications. Ceramic reinforced aluminium exhibit good strengthtoweight ratio and wear resistance. Aluminium and its alloys can be cut by traditional methods but ceramic reinforcement in MMC was found to create tool wear in conventional machining.
Wide spread industrial applications are impossible without solution to stringent design requirements and problems associated with cutting. Laser beam (noncontact mode) was generally employed to generate complex cut profiles faster than other methods. However the material properties govern the selection of laser system [1]. Pulsed CO2 laser beam can be employed to cut materials like metals, ceramic, plastics and composites [2]. The mechanism of metal removal by lasers involves melting, vaporizing and degrading. The vaporized material was removed by the assisting gas producing no mechanically induced damage to the work material. Investigations with CO2 lasers while handling aluminium had indicated the presence of heat affected zone in the cut surface [3]. While cutting polymers, it was observed that cutting speed plays an important role in determining the dimensional accuracy and finish of cut surface, while the relationship between speed of cutting and surface finish was observed to be nonlinear [4]. Nitrogen was found to produce smooth cut surface with smaller kerf, when used as an assist gas on austenitic stainless steels [5]. From the literature it was found that the cutting parameters play a vital role in deciding the quality characteristics of the surface machined [15].
Artificial neural network (ANN), grey relational analysis (GRA), principal component analysis (PCA), response surface methodology (RSM), simulated annealing (SA), fuzzy logic, technique for order of preference by similarity to ideal solution (TOPSIS) and genetic algorithm (GA) were generally employed to solve multi response optimization problems [612]. The tool wear was found to be more in traditional machining processes, while handling stronger and advanced materials like metal matrix composites. The increased tool wear was a major concern as it could spoil the finish of the cut surface [13, 14]. The by similarity to ideal solution (TOPSIS) and genetic algorithm (GA) were generally employed to solve multi response optimization problems [612]. The tool wear was found to be more in traditional machining processes, while handling stronger and advanced materials like metal matrix composites. The increased tool wear was a major concern as it could spoil the finish of the cut surface [13, 14]. The
The technique of GRA was employed to compute the grey relational grade as the performance index. The grade values could be used to sort out the near optimal parameter setting. The GRA was found to be very effective in an integrated format along with PCA and RSM [1720]. The desirability analysis involving simple computational efforts, when compared to techniques like PCA, TOPSIS, ANN, SA was also found to yield optimal solutions in various manufacturing processes. The desirability method was integrated with RSM and Taguchi techniques to improve its effectiveness [2124].
From the available literature, it was understood that not enough work was carried out in laser cutting of aluminium based composite. Hence the present work was focussed towards developing a new methodology (GDA), which combines the merits of grey theory and desirability analysis to predict the optimal set of machining parameters in pulsed CO2 laser cutting process.
2 Experimental Trial Design and Observations
2.1. Material
Al6061 matrix
Silicon carbide
2.2. Machine
2.3. Experimentation
The dominant cutting parameters like cutting velocity, laser beam power, assist gas pressure and pulsing frequency were found to affect the quality of the cut surface [13]. Preliminary cutting trials were performed to identify the acceptable upper and lower bounds of parameters for which the quality of cut surface remained within acceptable limits without dross and burning effect. These cutting parameters were varied at three levels and Taguchi’s L9 orthogonal array was used to design the cutting trials. The various levels of input parameters chosen for the laser cutting trials are shown in Table 1.
Table 1 Chosen levels of cutting parameters
Symbol 
Input Parameters 
Level 1 
Level 2 
Level 3 
P 
Laser power (W) 
1800 
2000 
2200 
F 
Pulse frequency (Hz) 
5 
10 
15 
V 
Cutting velocity (mm/sec) 
4 
8 
12 
Table 2 Quality characteristics observed during various trials
Trial 
Input 
Responses 

A 
B 
C 
Ra (μm) 
KW (mm) 

1 
1 
1 
1 
7.012 
0.602 
2 
1 
2 
2 
6.211 
0.572 
3 
1 
3 
3 
7.825 
0.527 
4 
2 
1 
2 
6.775 
0.598 
5 
2 
2 
3 
7.183 
0.575 
6 
2 
3 
1 
6.817 
0.588 
7 
3 
1 
3 
5.505 
0.586 
8 
3 
2 
1 
7.334 
0.508 
9 
3 
3 
2 
6.979 
0.525 
3 Multi Response Optimization Using Grey Desirability Analysis (GDA)
The demand for good surface finish (requiring little post processing) poses a huge challenge to metal cutting industries. The grey theory used in finding the optimal operating condition compensates the drawback of regression analysis by identifying the link between parameters based on the amount of difference or similarity of trends among those elements [13]. Generally the Taguchi techniques use signaltonoise (S/N) ratio to compare the responses. The GDA method involves following steps.
Step 1: Calculate the S/N ratio (yij) for both the responses treated as the lowerthebetter characteristic using Equation (1). The lowerthebetter analysis tends to minimize the responses, improving them significantly [3].
(1)
Where r = number of replications; m = number of trials; yij = observed response values; i = 1,2,3..r and j = 1,2,…m.
Step 2: Calculate the normalized S/N ratio (Zij) using Equation (2) to decrease the effect of variability among S/N ratio [6]. The normalized S/N ratio varies between 0 and 1.
(2)
Step 3: Compute the grey relational coefficient (GRC (γ)) for normalized S/N ratio values [13] using Equation (3).
(3)
Where; is the referential sequence; is the comparative sequence; and. ξ is the distinguishing coefficient whose value is chosen as 0.5.
Step 4: Compute the individual grey desirability (dij) value for the quality characteristics [7, 11], using the desirability function (largerthebetter type) represented by Equation (4).
(4)
Li and Ti are the lower and target values of the responses respectively.
Step 5: Calculate the composite performance measure (CPM) by taking the geometric mean of individual desirability values using Equation (5). The CPM value lies between 0 and 1.
(5)
Step 6: Find the main effect (εi) of various parameters using Equation (6) to identify the optimal level.
(6)
Step 7: Calculate the predicted S/N ratio at the selected optimal level [6, 11], using Equation (7) and perform ANOVA to find the contribution of individual parameters.
(7)
= Average S/N ratio and = Average S/N ratio corresponding to the ith factor at the f th level
Step 8: Conduct confirmation test for validation.
4. Results and Discussion
The grey generating technique was applied to transform the disordered raw data to regular series useful for measuring the relationship between different data elements [13]. The target for lowerthebetter characteristic is zero and a linear normalization of the observed data was performed to find the normalized S/N ratio values. The GRC and individual desirability values along with the CPM values are listed in Table 3. The CPM values offer the single representation for the two responses and a larger value of CPM is preferred irrespective of the nature of responses. The CPM value plotted for different cutting trials is shown in Fig. 3. The noteworthy variations in the computed CPM values signify the selected levels of different cutting parameters. The peak value of CPM was obtained for the seventh trial and hence the operating condition corresponding to that trial could be closer to the optimum parameter setting and would offer better responses.
Table 3 Data preprocessing and CPM values
Trial 
S/N ratio 
Normalized S/N ratio 
Grey relational coefficient 
Individual grey desirability 
CPM 

SR 
KW 
SR 
KW 
SR 
KW 
SR 
KW 

1 
16.97 
4.408 
0.312 
0.000 
0.421 
0.333 
0.131 
0.000 
0.000 
2 
15.83 
4.852 
0.657 
0.301 
0.593 
0.417 
0.390 
0.126 
0.221 
3 
17.80 
5.564 
0.000 
0.784 
0.333 
0.698 
0.000 
0.547 
0.000 
4 
16.68 
4.466 
0.410 
0.039 
0.459 
0.342 
0.188 
0.013 
0.050 
5 
17.16 
4.807 
0.243 
0.270 
0.398 
0.407 
0.097 
0.110 
0.103 
6 
16.62 
4.612 
0.392 
0.139 
0.451 
0.367 
0.177 
0.051 
0.095 
7 
14.85 
4.642 
1.000 
0.159 
1.000 
0.373 
1.000 
0.059 
0.243 
8 
17.37 
5.883 
0.184 
1.000 
0.380 
1.000 
0.070 
1.000 
0.265 
9 
16.86 
5.597 
0.325 
0.806 
0.426 
0.721 
0.138 
0.581 
0.284 
Figure 3 Variation of CPM for various trials
4.2. Optimal Laser Cutting Parameters and Analysis of Variance
The main effect of the laser cutting parameters on the CPM was calculated for each level and listed in Table 4, from which the optimal parameter level was found as A3B2C2. The result of pooled ANOVA on CPM is listed in Table 5, which could be used to identify the contribution of various cutting parameters in affecting the responses. Generally aluminium alloys are difficult to cut by lasers due to their high reflectivity and high power requirements in continuous wave mode of operation. Hence a pulsed laser beam using nitrogen as assist gas was employed to cut the material. Generally a higher degree of melting was found at the top surface of work material than at the bottom surface. A higher level of pulse frequency improves the value of CPM. This was due to the fact that high instantaneous energy in pulses at high frequency results in quicker melting and blowoff resulting in improved responses. A higher level of laser power was found to increase the energy content of the beam, which was essential to melt the matrix to be ejected by gas pressure removing a portion of reinforcements along with it.
Table 4 Main effect of parameters on CPM
Parameters 
Level 1 
Level 2 
Level 3 
MaxMin 
P 
0.074 
0.083 
0.264 
0.190 
F 
0.098 
0.196 
0.126 
0.099 
V 
0.120 
0.185 
0.115 
0.070 
Table 5 Result of ANOVA on CPM
Source of variance 
Sum of squares 
Degrees of freedom 
Mean sum of square 
Fratio 
% Contribution 
P 
0.0690 
2 
0.0345 
6.5601 
66.31 
F 
0.0154 
2 
0.0077 
1.4669 
14.83 
V 
0.0091 
2 
0.0046 
0.8658 
8.75 
Error 
0.0105 
2 
0.0053 

10.11 
Total 
0.1040 
8 


100 
4.3 Confirmation Experiment
A confirmation test becomes essential to authorize the methodology of GDA for optimization. The responses observed for the initial parameter setting were compared those obtained with the optimal parameter setting predicted by the GDA method (Table 6). It was found that GDA approach had improved the responses significantly.
Table 6 Comparison of the responses obtained for the initial and optimal parameter setting
Responses 
Initial parameter setting 
Optimal Setting using GDA 
% Improvements 

Calculated S/N ratio 
Response Value 
Predicted S/N ratio 
Response Value 
S/N ratio 
Response Value 

SR (μm) 
16.5767 
6.846 
15.2029 
5.5365 
1.3738 
1.3095 
KW (mm) 
5.6408 
0.531 
5.3613 
0.487 
0.2795 
0.044 
Parameter settings 
A3 B3 C2 
A3 B2 C2 

5 Conclusion and Future Research
In the present work Al/SiCp composite plates of thickness 4 mm were cut using the pulsed CO2 laser cutting process and a combined approach (GDA) was revealed for predicting the optimal cutting condition. The following conclusions were drawn.
The investigation findings will offer the necessary guideline and database for cutting Al/SiCp composite plates using pulsed CO2 laser cutting process, hence widening the scope of application of the MMCs.
References
*****