DRILLING OF CARBON/EPOXY COMPOSITES BY ...

The 1st International Conference on Industrial, Systems and Manufacturing Engineering (ISME’14) -----------------------------------------------------------------------

DRILLING OF CARBON/EPOXY COMPOSITES BY ELECTRICAL DISCHARGE MACHINING Jamal Y. Sheikh-Ahmad, [email protected]. Mechanical Engineering Department, The Petroleum Institute, Abu Dhabi, UAE Sandeep R. Shinde, [email protected]. Department of Industrial & Manufacturing Engineering, Wichita State University, Wichita, Kansas, USA

ABSTRACT This study investigates the feasibility of electrical discharge machining process (EDM) for drilling carbon fiber reinforced composites (CFRP). EDM drilling of CFRP was conducted on a small hole machining system using dionized water as the dielectric fluid. Copper and graphite rods were used as the electrode materials. The effect of gap current, pulse-on time and electrode material on material removal rate, tool wear and delamination was investigated. It was found that the highest material removal rates were obtained with graphite electrode at the highest current and highest pulse-on time (highest energy input). On the other hand, these conditions also corresponded to the highest electrode wear rate and most delamination damage. Because if this, an optimum set of conditions for EDM drilling could not be found within the ranges of parameters tested. However, due to the interaction between process parameters it is found that intermediate levels of pulse-on time cause a slight reduction in delamination damage. Keywords: EDM, CFRP, Drilling, Material removal rate, Delamination, Wear

1. INTRODUCTION Drilling of carbon fiber reinforced polymer composites (CFRPs) is an important secondary machining process in the aerospace industry and it almost accounts for 50% of the total cost of machining composites. This is due to the fact that even a smaller private jet has up to 250,000 to 400,000 holes and a bomber or a transport aircraft has 1,000,000 to 2,000,000 holes [1]. CFRPs are difficult to machine by conventional processes because of their inhomogeneous structure. The defects observed in drilling these materials include hole taper, hole roundness error, surface roughness, splintering, fiber pullout, matrix cracking and delamination. The existence of these defects poses a threat to aircraft damage tolerance, survivability and reliability. Therefore there is a need for better machining processes that can reduce or eliminate machining damage. Electrical discharge machining (EDM) is a non-traditional machining process typically used to cut hard-to-machine materials with high accuracy [2]. This is due to the fact that the workpiece is never in contact with the tool and hence there are no contact forces acting between the tool and the workpiece. Machining is done by the effect of spark erosion which is generated in a gap between an electrode and the workpiece material. The gap, and sparking action, are stabilized by a servo control mechanism. A few studies in the literature have addressed the feasibility of EDM of CFRPs. These studies confirmed the feasibility of this process and pointed out potential problems arising from its application, specifically thermal damage to the workpiece.


Strong et al. [3] performed EDM drilling of AS4/3501-6 CFRP using a graphite electrode at different levels of gap voltage, pulse-on time and pulse-off time. Material removal rate (MRR) was evaluated as a measure of machinability. It was reported that MRR is highest at highest energy levels. However, the size of heat affected zone was twice as large as the hole diameter. Guu et al. [4] performed drilling of plain weave carbon/phenolic composite with copper electrode. Delamination damage and surface roughness were evaluated. It was pointed out that delamination has its highest levels at highest current setting. The experiments carried out by George et al. [5] and Lau et al. [6] addressed the issue of electrode wear and the effects of current and polarity. Wang et al. [7] investigated the effect of carbon fiber direction on material removal rate, electrode wear and surface roughness when copper and graphite electrodes were used. This study was further clarified by Habib et al. [8] who concluded that the material removal mechanism depends on the direction of the carbon fibers. The intent of this study is to investigate the use of electrical discharge machining for machining small holes in aircraft carbon fiber composites and to analyze the effect of process parameters like pulse-on time, interval time, gap current and tool material on tool wear rate, material removal rate and hole quality.

2. MATERIALS AND METHODS 2.1 Workpiece The workpiece material used in this work was a rectangular multidirectional CFRP coupon measuring 46mmx23.5mmx1.8mm. The coupons were cut with a diamond saw from a laminate with the stacking sequence [(PW/XW)2PW]s, where PW represents a single ply with the fiber directions aligned coincident with the coupon axes and XW is a single ply aligned at ±45o with the coupon axes. A layer of nonconductive epoxy resin was removed from the top surface of the coupon with 220 grit sand paper to expose the carbon fibers and allow for current to flow in the workpiece. The resistivity of the sanded surface ranged for 2.5 to 4.5  cm. Figure 1 shows an optical image of the coupon with top surface shown after sanding.

Figure 1. Top surface of CFRP sample.

2.2 EDM Electrodes Copper and graphite were used as electrode materials in this experiment. The electrodes were cylindrical in shape with a diameter of 6mm, which is the nominal hole diameter. The electrodes were received in long rods, which were cut to proper size with a diamond saw. The ends of the electrodes were squared by face grinding. Sets of four electrodes of each graphite and copper were used for the machining. The end of each electrode was ground flat after EDM machining of one hole so that it can be reused for further experiments.


2.3 Machine Setup The machine used for this experiment was a 5-axes RAYCON small hole EDM system. The system consisted of an SH10 small hole EDM generator to provide the necessary DC power cycle for machining, Fanuc GN 6 series numeric controller to control the machine axes, Barnstead 210 distilled water generator to produce deionized water for the dielectric medium and a Lauda RE106 series refrigerating circulator of 16-liter capacity to circulate and cool the dielectric medium throughout the system. Figure 2 shows a schematic of the EDM system and machining setup. Initial setup of the SH10 generator consisted of adjusting pulse-on time, interval time, gap voltage and DC current. The pulse-on time was adjusted to produce various degrees of surfaces finish as well as to achieve regular and efficient machining conditions. The interval time was adjusted to provide maximum efficiency with stable machining. Interval time also provided a little time between arcs to remove the debris from the machined hole. The electrode was attached to the Z-axis of the machine and the gap between the workpiece and electrode was controlled by a servomechanism, which maintained the required gap by comparing reference or set voltage value and online voltage value. Spark erosion took place during the pulse-on time, which removed the material from the workpiece and electrode. During pulse-off time the dielectric fluid cooled the electrode and workpiece and flushed away debris from the gap. In this experiment the electrode was attached to the positive pole of the power generator and the target gap voltage was set at 65V.

Figure 2. Schematic of the EDM machining setup.

2.4 Response Measurement The responses measured in each experiment were material removal rate (MRR), tool wear rate (TWR) and delamination factor (DF). Material removed was calculated by weighing the workpiece before and after machining. An ACCULAB L-series weighing scale is used to measure the weight in grams. The difference in the weight of workpiece before and after the machining is divided by machining time to give material removal rate (MRR) in grams/min. (1) The machining time was recorded by using a stopwatch.


A similar procedure was used to calculate the tool wear rate (TWR). The electrodes are weighed before and after machining and the difference was used to give electrode wear in grams. This difference in weight was divided by the machining time to give the electrode wear rate in grams/min. The graphite electrodes were dried for sufficient time before weighing them. This is done to avoid any misinterpretation in weighing the graphite electrodes, as they tended to absorb small amounts of dielectric fluid during machining Delamination factor gives the extent of hole edge damage and it is defined by DF 

d max d

(2)

where, DF is delamination factor, dmax is maximum damaged diameter d is the hole diameter as shown in the Figure 3.

2.5 Parametric Study Factors that were taken into consideration while designing the experiments were pulse-on time, pulse-off time, gap current, dielectric temperature, dielectric flow rate, gap between electrode and work piece, and electrode material. An initial screening experiment determined that gap current and pulse-on time were the two most significant parameters affecting machining performance. Further experiments were conducted only with these two parameters and electrode material as a categorical factor. A face centered central composite design was selected to run at three levels of gap current and pulse-on time and two levels of categorical factor electrode material. A central composite design commences with a factorial or fraction factorial design with center points and added axial runs also called as “star runs”. Addition of star runs along with center points allows estimation of curvature in the model. This is due to that fact that addition of axial points basically includes the quadratic terms in the model [9]. This design produced a set of 26 runs with 5 repetitions of the center point. The entire experimental design was repeated twice to ensure repeatability of the results. Experiments were conducted in random order to eliminate the effects of unknown factors. Pulse-off time is the time between two sparks, also called as interval time was held constant at 100sec. Gap between electrode and work piece was maintained constant by servomechanism and the gap voltage was maintained at 65V. Table 1 shows the different parameters of the experiment and their levels.

Table 1. Experimental parameters and their levels Parameter Gap Current Gap Voltage Pulse-on Time Pulse-off Time Electrode material Dielectric Temperature Dielectric Flow Rate

Values 0.4, 1.2, 2.0 A 65 V 20, 105, 190 µsec 100 µsec Copper, Graphite o 25 C 17 l/min


3. RESULTS AND DISCUSSION 3.1 Hole Characteristics It was observed during EDM machining of CFRP that material is removed in solid and gaseous states. Carbon fibers are electrically conductive but the bonding epoxy resin is not electrically conductive. Therefore, carbon fibers were removed by spark erosion while bonding material was decomposed due to extreme high temperature and presence of high intensity sparks during machining. Eroded carbon fibers were removed by flushing with the dielectric fluid in the gap between electrode and workpiece whereas, decomposition of resin layers caused release of gases in form of intermediate puffs. The appearance of the machined hole was studied for presence of delamination, fiber pullout and surface finish. Figure 3 shows an optical image of the entry (top) and exit (bottom) of the machined hole. It can be seen from these images and other similar images of other pieces that entry and exit of the hole do not show any fiber pullout. Also, the walls of the hole were clean and no fiber pullout was observed on the cross-section of the hole. However, significant delamination was observed on the hole entry (top) which resulted in the loss of one or more of the top plies as shown in Figure 3. Very little or no delamination was observed at hole exit (bottom).

dmax

d

Delamination

Figure 3. Optical images of the EDMed hole top surface (left) and bottom surface (right).

Delamination on top surface of the workpiece was observed due to high localized temperature and high spark intensity. During machining, base surface of the workpiece was not exposed to these high intensity sparks. Also, temperature at the base surface was very low as compared to the top surface. Delamination on base surface was hence very minute and in some cases not observed. Quality of the machined hole was quantified by delamination factor as defined by equation (2). 3.2 Electrode Wear In EDM machining material removal takes place due to spark erosion of both workpiece and electrode materials. During each spark the electrode is bombarded with either electrons or positive ions depending upon electrode polarity. This bombardment causes localized heat generation at the surface of the electrode which causes localized


melting and/or vaporization of the electrode material. This material removal, which occurs at sparking surface of the electrode, is called as electrode wear. Figure 4 shows before and after machining optical images of a graphite electrode. It can be seen from these pictures that the electrode shows higher amounts of wear on the corner of the circular electrode. This is observed because of the higher concentration of sparks at the corner. The number of sparks required to produce the 3D shape or cavity in a workpiece is higher when compared to number of sparks require to produce a flat surface. Since each spark removes some material from electrode, edge of the circular electrode shows more wear than end of the electrode. In EDM, low electrode wear is desired. As the machined surface is an exact replica of the electrode, excessive electrode wear results in poor dimensional accuracy of the hole and poor surface finish.

(a)

(b)

(c)

Figure 4. Optical images of graphite electrode (a) side view of new electrode, (b) end view of worn electrode, (c) side view of worn electrode

3.3 Analysis of Variance (ANOVA) ANOVA of the experimental results was performed using Design-Expert software. Face centered central composite design was used with two numerical factors and one categorical factor. These are namely gap current, pulse-on time and type of electrode, which are noted as A, B, and C respectively in terms of coded factors. The results of ANOVA showed that all three factors significantly affect the outcomes of MRR and TWR whereas delamination factor was mainly affected by current and pulse-on time. Second order polynomial was fitted to the data and the resulting models are shown in Table 2. The table also shows the goodness of fit in terms of adjusted R 2. It can be seen that the models for MRR and TWR are of good fit while the model for DF is of moderate fit. The model equations obtained from ANOVA were used to generate graphs that show the nature of interaction between all the process parameters. The experimental results were also plotted on these graphs to understand effect of current, pulse-on time and tool material on all the three responses.

Table 2. Regression models for responses (A is current and B is pulse-on time). Electrode Graphite Copper Graphite Copper Graphite & Copper

Model MRR = 2.712E-03 + 0.011 A + 2.938E-05 B - 1.926E-07 B2 + 6.265E-05 A·B MRR = - 8.264E-04 + 0.013 A + 4.723E-05 B - 1.923E-07 B2 + 1.528E-05 A·B TWR = 4.762E-04 + 4.492E-03 A - 3.263E-06 B - 8.480E-04 A2 + 9.851E-06 A·B TWR = 1.417E-03 + 5.263E-04 A - 1.241E-05 B + 8.248E-04 A2 + 9.851E-06 A·B DF = 1.16474 + 0.072151 A - 6.84265E-004 B + 3.83894E-006 B2

R2 (adj) 0.983 0.983 0.925 0.925 0.780


Material removal rate (MRR): Figure 5 shows the effects of current and pulse-on time on MRR for copper and graphite electrodes. These graphs are plotted for low, medium and high levels of pulse-on time. It can be concluded form these plots that both electrodes show high MRR at high settings of current and low MRR at low settings of current. It is also interesting to notice that at low current settings there is no significant increase in the MRR for corresponding increase in pulse-on time for both copper and graphite electrodes. As current setting increases, any increase in pulse-on time shows significant increase in MRR for both types of electrodes. This is evidenced by the presence of the interaction term A·B in the regression models for MRR. It can also be inferred from the plots that as compared to copper electrodes, graphite electrode shows higher MRR and are considerably affected by pulse-on time at higher current settings.

0.05

0.05 20 sec Model 105 sec Model 190 sec Model 20 sec Exp 105 sec Exp 190 sec Exp

0.03

0.04

MRR (g/min)

MRR (g/min)

0.04

20 sec Model 105 sec Model 190 sec Model 20 sec Exp 105 sec Exp 190 sec Exp

0.02

0.01

0.03

0.02

0.01

Copper Electrode

Graphite Electrode

0.00

0.00 0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

0.2

0.4

0.6

0.8

1.0

Current (Amp)

1.2

1.4

1.6

1.8

2.0

2.2

Current (Amp)

Figure 5. Effect of current on material removal rate.

Delamination factor: Figure 6 shows the effect of current and pulse-on time on delamination factor for copper and graphite electrodes. It is clear from these graphs that, for copper as well as for graphite, an increase in the gap current results in a proportional increase in delamination factor. The effect of pulse-on time on delamination factor is not significant. DF appears to be slightly higher for low and high pulse-on time and slightly lower for medium pulse-on time. The highest DF is caused by the highest current and highest pulse-on time (highest discharge energy level) and the lowest delamination is caused by the lowest current and the intermediate pulse-on time. The graphs also show that different current and pulse-on time settings give almost same amount of delamination factor for copper and graphite electrodes. This means that DF is independent of electrode material and the same regression model applies for both electrodes. Tool wear rate (TWR): Figure 7 shows the effects of current and pulse-on time on tool wear rate for copper and graphite electrodes. It is clearly seen from this Figure that the response of TWR to process parameters is different for copper and graphite electrodes. The differences are in concavity of curvature of the curves and the interaction between current and pulse-on time. Furthermore, wear rates for the copper electrodes are generally lower than those for graphite. For copper electrodes TWR increases with an increase in the current with an upward concavity. For low level of current, an increase in pulse-on time corresponds to a decrease in TWR. This trend is inverted for the highest current setting. The effect of pulse-on time on TWR at intermediate current setting is not significant.


The graphite electrode TWR also increases with an increase in current. However, the concavity of the curve is down and which suggests that there exists an optimum value (outside the range of parameters of these experiments) for TWR after which any increase in current will result in decrease of the TWR. The effect of pulse-on time on TWR is not significant for the lowest current setting, and it is the highest for the highest current setting. At the intermediate and high current settings an increase in pulse-on time leads to an increase in TWR.

1.5


1.4

1.4

1.3

DF

DF

1.3

20 sec Model 105 sec Model 190 sec Model 20 sec Exp 105 sec Exp 190 sec Exp

1.2

1.2

1.1

1.1

Copper Electrode

Graphite Electrode

1.0

1.0 0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

0.2

0.4

0.6

0.8

Current (Amp)

1.0

1.2

1.4

1.6

1.8

2.0

2.2

Current (Amp)

Figure 6. Effect of current on delamination factor.

0.010


0.006

0.008

TWR (g/min)

TWR (g/min)

0.008

0.004

0.002

0.006


0.002

Copper Electrode

0.000 0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Graphite Electrode

0.000 2.2

0.2

0.4

0.6

0.8

Current (Amp)

1.0

1.2

1.4

1.6

1.8

2.0

2.2

Current (Amp)

Figure 7. Effect of current on tool wear rate.

4. CONCLUSIONS EDM drilling of plain weave carbon fiber reinforced polymer composite was conducted using copper and graphite electrodes under different levels of gap current and pulse-on time and machinability was evaluated by material


removal rate, electrode wear rate and delamination damage. The following conclusions can be drawn from this study: 1.

2. 3.

4.

Material removal rate is influenced by all the three factors namely gap current, pulse-on time and electrode material. Both copper as well as graphite electrodes show high material removal rates for higher settings of current and pulse-on time (high discharge energy level). As compared to copper, graphite electrodes are significantly influenced by any changes in pulse-on time and current. Also at different settings graphite electrodes give higher MRR as compared to copper. Tool wear rate is also influenced by gap current, pulse-on time and type of electrode material. For copper and graphite electrodes, decrease in gap current and pulse-on time (lower energy level) shows significant decrease in tool wear rate. As compared to graphite, copper electrodes show less amount of tool wear at different settings of pulse-on time and current. Delamination factor is only influenced by gap current and pulse-on time. High current and high pulse-on time (high energy level) result in more delamination rate. Low delamination can be achieved by keeping current at low level and pulse-on in the rage of 80µs to 105µs.

REFERENCES 1. Hubber, J and Micillo ,C., Innovated manufacture for automated drilling operation, proceeding of Autofact west, V2, p292-316,(1980). 2. Elamn C., J. (2001). Electrical Discharge Machining. Michigan, Society of Manufacturing Engineers. 3. Strong , A. B.,Crockett, T., Merrell, R. S., & Dredge, J.D. (1991). Cutting cured carbon fiber composites using electrical discharge machining. S.A.M.P.E. (Society of Aerospace Materials and Process Engineers) quarterly, 22, no3, 40-45. 4. Guu, H.Y., Hocheng, H., Tai, H.N.,& Liu, Y.S.(2001) Effect of electrical discharge machining on the characteristics of carbon fiber reinforced carbon composite. Journal of Material Science, 36, 2037-2043. 5. George, P. M., Manocha, L. M., Ragunath, B.K. & Warrier, A. M. (2004). EDM machining of carbon-carbon composite- a Teguchi approach. Journal of Material Processing Technology, V145, 66-71. 6. Lau, S.W., Wang, M., & Lee, B.W. (1990). Electrical discharge machining of carbon fiber composite materials. International Journal of Machine Tools Manufacture, 30, no2, 297-308. 7. Wang, H., Habid, S., Okada, A., Uno, Y. (2010) EDM charactersitics of carbon fiber reinforced plastic. Proc. Of the 16th International Symposium on ElectroMachining, 65-68. 8. Habib, S., Okada, A. and Ichii, S. (2013) Effect of cutting direction on machining of of carbon fibre reinforced plastic by electrical discharge machining process. Int. J. Machining and Machinability of Materials, V. 13, 414427. 9. Motgomery, C., D. (1997). Design & Analysis of Experiments. New York, John Wiley & Sons, Inc.