gupta2014 Study of Cutting Speed Variation in the Ultrasonic Assisted Drilling of Carbon fibers composites

The following research was focused on the examination of exit delamination, machined surface and cutting temperature measurement in UAD and a comparison to that in CD at the cutting spee

Study of Cutting Speed Variation in the Ultrasonic Assisted Drilling of Carbon

Fibre Composites

Aniruddha Gupta

WMG, University of Warwick Coventry, West Midlands, UK

Stuart Barnes

WMG, University of Warwick Coventry, West Midlands, UK

Iain McEwen

School of Engineering, University of Warwick Coventry, West Midlands, UK

Nadia Kourra

WMG, University of Warwick Coventry, West Midlands, UK

Mark A Williams

WMG, University of Warwick Coventry, West Midlands, UK

ABSTRACT

Ultrasonic assisted drilling (UAD) has been proven

effective for the thrust force reduction as compared to

conventional drilling (CD) for same machining parameters The

following research was focused on the examination of exit

delamination, machined surface and cutting temperature

measurement in UAD and a comparison to that in CD at the

cutting speeds of 0.942 m/min, 9.42 m/min, 94.2 m/min and

282.6 m/min at a constant feed rate of 0.05 mm/rev in the

through-hole drilling of CFRP material X-Ray computed

tomography (CT) was used to identify the exit delamination,

internal damage, circularity and center deviation in CD and

UAD A maximum of 82.8% reduction in the center deviation

and 33.2% reduction in circularity of the holes were found

when drilled in UAD as compared to those in CD Furthermore,

the cutting temperature in the drilling of CFRP has been

measured and compared for both the cases of CD and UAD

Ultrasonic assistance produced 10°C higher cutting temperature

than that in CD at the cutting speed of 282.6 m/min while at

lower cutting speeds (0.942 m/min and 9.42 m/min), the cutting

temperatures with and without ultrasonic assistance did not

have a significant difference (2°C and 4°C respectively)

1 INTRODUCTION

Carbon fibre reinforced plastics (CFRPs) are attractive as

aircraft and aerospace structural components due to their lighter

weight and higher specific strength as compared to metals

[1-7] which has led to greater weight savings resulting in greater

pay load, longer range and fuel saving in comparison to the

metals [8] In spite of curing to the final shape, the machining

of composites is required at several stages of the production

e.g cutting and drilling of holes The mechanical drilling

process is one of the final processes in the manufacturing of the composite components [2] The conventional drilling (CD) of CFRPs has been proven difficult due to delamination and short tool life According to a study [1], 60% of all the part rejections

in air craft industry is associated with delamination in drilling For carbon fibre composites, high thrust force is considered to be the most detrimental to CFRP drilling due to delamination and damage inside the holes in a drilling process [9, 10] Some researchers have reported that the longer tool life, lower thrust force and segmented and smaller chip lengths have been observed when drilling metals by introducing vibrations in the drilling [11, 12] The effects have been observed typically when the frequency of the vibrations has been kept above

20 kHz because of achieving an intermittent cutting action [13] Such type of drilling is known as ‘Ultrasonic assisted drilling (UAD)’ [7, 13, 14] In the recent years, Makhdum et al [15, 16] have conducted experiments on CFRP to develop an understanding for ultrasonic assisted drilling and found an almost 90% reduction in the thrust force and generation of the longer chips during UAD of CFRP However, the effect on the damage during the drilling was not a focus in their work A similar experimental study has also been conducted by Mehbudi et al [17] for glass fibre laminate (GFRP) material They conducted experiment at a low cutting speed (maximum cutting speed = 31.4 m/min) and found UAD to be effective in the reduction of thrust force and exit delamination Sadek et al [18] have also reported obtaining zero exit delamination between 113 m/min (6000 rpm) to 226 m/min (12000 rpm) spindle speed having 0.025 mm/rev feed rate with 6 mm diameter drill but no images having zero exit delamination were presented in their work Also, they used the oscillation

Proceedings of the ASME 2014 International Mechanical Engineering Congress and Exposition

IMECE2014 November 14-20, 2014, Montreal, Quebec, Canada

IMECE2014-37046

frequencies of 30 Hz and 60 Hz which are not in ultrasonic

range

For cases when the feed rate of the cutting tool is not

negligible in comparison to the cutting speed, the rake and

clearance angles are affected [19] In the case of drilling, where

the feed rate is in the perpendicular direction to the cutting

speed, the feed rate becomes more dominating as compared to

cutting speed in terms of cutting and thrust forces when the

ultrasonic oscillations are imposed on the drill [19] Because of

these ultrasonic modulations in the feed rate, the effective

normal rake angle of the twist drill at all the points on the

cutting edge varies rapidly (~114° angle variation within

26.9 µs) This research was focused on the effect of the

effective normal rake angle variation on the thrust force,

internal damage of the machined surface of the holes, exit

delamination and the cutting temperature in the ultrasonic

assisted drilling of CFRP material Four cutting speeds were

used – 0.942 m/min, 9.42 m/min, 94.2 m/min and 282.6 m/min

at a constant feed rate of 0.05 mm/rev and comparison with that

in CD

CALCULATION FOR CD AND UAD

In order to visualize the tool-material interaction in UAD, a

calculation was performed to identify the variation in the

effective normal rake angle during the drilling for a simple

two-flute twist drill Though similar calculation has also been

performed by Zhang et al [20] and Wang et al [21] for the

vibration assisted drilling of CFRP materials while calculating

the mean thrust force and torque for CFRP materials, some of

the intermediate steps involving geometrical relationships for

the tool geometry were not clear even in the mentioned

references in their work Therefore, it becomes essential to

mention the relevant equations used in the present research for

developing a fundamental understanding about the variations

generated because of ultrasonic oscillations on the effective

normal rake angle Figure 1 shows the axial projection of the

chisel and cutting edges with the visualization of the cutting

velocity, feed velocity and the effect on axial rake and

clearance angles with the introduction of the feed velocity in

the work reference system for an arbitrary point ‘i’ of the

cutting edge

Figure 1: Visualization of the rake and clearance angle in the work

reference system [19]

In order to calculate the effective normal rake angle, the effect of feed rate on the orthogonal rake would have to be considered at every point of the cutting edge [19]

As shown in the Figure 1 for the point ‘i’ –

μi= tan−1(VV

Effective feed rate for CD,

Effective feed rate in UAD,

Vf = S0N + aω cos ωt (5) From the machining theory, [19, 20],

γXD = tan−1(rrtan θ) (6) The orthogonal rake angle in the work reference system is given by [19] –

tan 𝛾𝑜𝑊 = tan 𝛾 + tan 𝜆sin 𝜙 cos 𝜙

Where, tan 𝜙𝑊 = tan 𝜙cos 𝜇 ; tan 𝜙𝐷 = tan ∅cos 𝛽

∅𝑠 = cone angle of the twist drill,

𝛽𝑖= sin−1(𝑟𝑟) , Putting the respective values from Equation 1 to 6 into Equation 7, the orthogonal rake angle for

UAD in the work reference system for point ‘i’ of the cutting

edge in the Figure 1 is –

tan 𝛾𝑜𝑊 =tan[tan

( tan )+ tan (

)]+ tan 𝜆 cos 𝜙

(8) Hence, the effective normal rake angle in UAD [19]

tan 𝛾 𝑊 = tan 𝛾cos 𝜆

The effective normal rake angle variation has been plotted and compared for the various cutting speeds in Figure 2 for CD and UAD It can be observed that the wave form of the effective normal rake angle in UAD changes from step wave (Figure 2(a)) to sinusoidal wave (Figure 2(d)) when cutting speed is increased from 0.942 m/min (50 rpm) to 282.6 m/min (15000 rpm) because as the cutting speed is increased, the point

of concern on the cutting edge moves a longer distance before the ultrasonic oscillation completes an oscillation cycle It can also be observed that the value of the maximum rake angle in

an ultrasonic cycle reduces from 63.19°, Figure 2(a), to 33.01°, Figure 2(d), when the speed is increased from 0.942 m/min (50 rpm) to 282.6 m/min (15000 rpm) while the effective normal rake angle in CD remains at 28.95° at all the cutting speeds Also, Figure 2(d) shows that the maximum normal rake angle in the UAD varies from 33.01° to 26.33° which is close

𝜇 𝑖 𝜸 𝑿𝑫 𝒊

𝜇 𝑖

to the maximum effective normal rake angle in CD at

282.6 m/min (28.95°) cutting speed

Figure 2: Effective normal rake angle variation in CD and UAD at

cutting speeds of (a) 0.942 m/min, (b) 9.42 m/min, (c) 94.2 m/min

and (d) 282.6 m/min at with feed rate of 0.05 mm/rev having

ultrasonic frequency of 37.22 kHz and 2.9 µm amplitude (peak to peak) in UAD

In the work of Zhang et al [20], because of using low vibration frequency of 130 Hz, they obtained a small difference

of 4° between the effective normal rake angles in conventional and vibration assisted drilling

Arola et al [22] have mentioned about the localization of the damage while increasing the rake angle in the orthogonal trimming of graphite/epoxy composite material In order to identify the effect of these rapid variations in the effective normal rake angle in UAD, the drilling experiment was performed further

3 EXPERIMENTAL SETUP AND PROCEDURE

Drilling was performed on an ULTRASONIC 65 monoBLOCK® machine with an ultrasonic actuator built into the tool holder This machine works on the concept of the reverse piezoelectric effect, wherein the transmission of the high-frequency electrical signal to the piezo-elements in the tool holder to the spindle is executed without contact (inductive) An oscillation is generated in the axial direction and is superimposed on the conventional tool rotation and feed rate This particular machine has a maximum capability of 18,000 rpm spindle speed and 40,000 mm/min feed rate The ultrasonic oscillation parameters depend upon the tool – tool holder combination Once the drilling tool is located within the tool holder, the optimum oscillation frequency is determined by the machine Also, the maximum oscillation amplitude is fixed

The amplitude can be varied from 0% to 100% of the maximum oscillation amplitude of tool – tool holder combination In the present case, the frequency of the oscillation of the tool was

37220 Hz and the maximum oscillation amplitude was 2.9 µm (peak to peak)

The carbon fibre composite material was provided by BAE Systems It was a 7.4 mm thick unidirectional stack of -45°, 0°, +45° and 90° orientations consecutively with woven carbon fibre layer on the top and bottom having bismaleimide (BMI) matrix The drilling tool used in the experiment was uncoated tungsten carbide, 3 flute twist drill having 6.0 mm diameter, 66.0 mm overall length and 150° point angle having flat rake and clearance surfaces The constant axial rake and clearance angles (due to flat rake & clearance surfaces) were measured in the laboratory, (rake angle = 6.40° and clearance angle = 10.56°) This type of drill was recommended from the tool manufacturer for good surface finish, high hole-accuracy, reduced tool wear and longer tool life due to three cutting edges

as compared to 2-flute twist drill The experiment was

conducted under dry-machining conditions

The final experimental setup including the drill, the ultrasonic actuator, dynamometer (Type – 9257B), fixture and the CFRP strip is shown in the Figure 3 To find out the reproducibility, the experiment was repeated three times further with the same drill It was found that the variation of the values was similar whereas the particular values were getting higher due to the incremental tool wear after each drilling

-60

-40

-20

0

20

40

60

80

Time (micro seconds)

Normal rake angle in CD and UAD in work reference system

r = 1.0 mm UAD

r = 1.4 mm UAD

r = 1.8 mm UAD

r = 2.2 mm UAD

r = 2.6 mm UAD

r = 3.0 mm UAD

r = 1.0 mm CD

r = 1.4 mm CD

r = 1.8 mm CD

r = 2.2 mm CD

r = 2.6 mm CD

r = 3.0 mm CD

-60

-40

-20

0

20

40

60

80

Time (micro seconds)

Normal rake angle in CD and UAD in work reference system

r = 1.0 mm UAD

r = 1.4 mm UAD

r = 1.8 mm UAD

r = 2.2 mm UAD

r = 2.6 mm UAD

r = 3.0 mm UAD

r = 1.0 mm CD

r = 1.4 mm CD

r = 1.8 mm CD

r = 2.2 mm CD

r = 2.6 mm CD

r = 3.0 mm CD

-30

-20

-10

0

10

20

30

40

50

Time (micro seconds)

Normal rake angle in CD and UAD in work reference system

r = 1.0 mm UAD

r = 1.4 mm UAD

r = 1.8 mm UAD

r = 2.2 mm UAD

r = 2.6 mm UAD

r = 3.0 mm UAD

r = 1.0 mm CD

r = 1.4 mm CD

r = 1.8 mm CD

r = 2.2 mm CD

r = 2.6 mm CD

r = 3.0 mm CD

0

5

10

15

20

25

30

35

Time (micro seconds)

Normal rake angle in CD and UAD in work reference system

r = 1.0 mm UAD

r = 1.4 mm UAD

r = 1.8 mm UAD

r = 2.2 mm UAD

r = 2.6 mm UAD

r = 3.0 mm UAD

r = 1.0 mm CD

r = 1.4 mm CD

r = 1.8 mm CD

r = 2.2 mm CD

r = 2.6 mm CD

r = 3.0 mm CD

b

a

d

c

Figure 3: Final experimental set-up having dynamometer, fixture

holding CFRP strip, ultrasonic tool holder and the 6 mm diameter

drilling tool

4 RESULTS AND DISCUSSION

4.1 Thrust force

The thrust force was recorded using a Type 9257B Kistler

dynamometer and the average thrust force was considered as a

reading for each drilling as shown in the Figure 4

The average was taken between the readings where the

thrust force was stable and away from the entrance/exit effects

of the hole Distance drilled from 2 mm to 5 mm during drilling

was considered for the average for each drilling, e.g the

average thrust force in the Figure 4 was measured between the

time instances of 120s to 192s for 0.942 m/min cutting speed as

the depth from 2 mm to 5 mm was drilled between these time

instances (considering 3 mm depth travelled by the tool prior to

drilling) The overall average values of thrust force

measurements for each parameter are plotted in the Figure 5

The thrust force was found to be the maximum at the

lowest cutting speed of 0.942 m/min in CD and UAD, Figure 5

Unlike reported by Mehbudi et al [17] and Makhdum et al [15,

16], thrust force was not found to be reduced because of

ultrasonic assistance at every cutting speed The maximum

reduction in the thrust force because of ultrasonic assistance

was found to be 9.1 N (although a small difference) at

9.42 m/min cutting speed as compared to CD For the rest of

the cutting speeds, the thrust force was found to be similar in

CD and UAD Also, the minimum thrust force was found to be

at 282.6 m/min cutting speed in both CD and UAD

Figure 4: Thrust force profile during the drilling of one hole (0.942

m/min, 0.05 mm/rev), average thrust force = 231.8 N

Figure 5: Average thrust force recorded CD and UAD 4.2 X-Ray Computed Tomography (CT)

In order to examine the extent of the damage accurately in the drilled-holes, the non-destructive technique of X-ray CT scanning has been used and a methodology has been developed [23] Each drilled hole was scanned separately by Nikon Metrology system XTEK 320 In order to identify the physical center of the hole, four straight marks were machined around where the hole was going to be drilled, Figure 6 A 2 mm diameter, 2-flute, Q-coat tungsten carbide slot drill with flat end was used to machine the marks First, the four marks were machined and then the tool was changed to the drilling tool before the hole was drilled keeping the work-piece fixed The center of the hole was identified by the point of intersection of the lines joining the midpoints of the machined marks around the hole through image processing, (Figure 6 (b))

Figure 6: (a) Straight machined marks and the drilled hole, (b) center identification and the dimensions of hole and marks in mm

The utilization of this method lies to the quality of the machined marks around the hole and the identification of the edges of the slots in the reconstructed data Hence, several parameters were tried and finally, at the cutting speed of

113 m/min and feed rates of 250 mm/min cross-feed and 100 mm/min in-depth feed, the marks were machined with two passes – first 0.4 mm deep and then further 0.3 mm making the final depth of the machined marks to be 0.7 mm

The data provided by CT scanning was reconstructed to a 3D representation of the hole with back-projection reconstruction algorithms in CT Pro software, provided by Nikon Metrology The reconstructed model was exported from

VG Studio Max 2.2 software in DICOM images The exported slides had the thickness of a voxel, 3D reconstructed pixel,

150 170 190 210 230 250

Cutting speed (m/min)

Drill

Ultrasonic actuator

CFRP strip

Fixture Dynamometer

having a thickness of approximately 14 µm Every DICOM

image was analysed by a MATLAB code that utilizes image

processing tools to provide complete internal information of

each drilled hole

With the help of the image processing, the exit

delamination and the dimensional metrological results on the

deviation of radius, circularity and distance between actual and

ideal centers (center deviation) were calculated for each layer

The entrance delamination could not be measured through this

method because the depth of the machined marks was 0.7 mm

and the analysis of the image processing was started once the

center of the hole was determined at the depth of 0.7 mm

4.2.1 Exit delamination

In general, two types of the exit delamination factors have

been considered i.e the diameter delamination (Dmax/D0) and

the area delamination (Amax/A0) [17, 24] Where Amax is the

damaged area at the exit and Dmax is the maximum diameter of

the damaged area (Amax) at the exit and D0 and A0 are the

diameter and circular area of the original drilled-hole As the

diameter delamination implies the whole area to be affected

inside the maximum diameter of the damage whereas the area

delamination signifies the actual affected area due to

delamination as the case of spalling mentioned by Davim at al

[24], therefore, the area delamination was considered for the

analysis in the present research work

The damaged area of the exit delamination for each drilled

hole was calculated from X-Ray CT scanning data of the layers

near to the exit of the hole The maximum area amongst all the

areas of the layers near to the exit was considered as the area

for exit delamination of a particular hole The area delamination

factors were then calculated by the ratio of the exit delaminated

area and the circular area of the original intended hole for all

the parameters The respective area delamination factors are

plotted in the Figure 7

The exit delamination at 0.942 m/min cutting speed was

found to be higher in UAD as compared to that in CD because

the mean thrust force at the exit-laminates was found to be

higher in UAD (222.2 N) as compared to that in CD (217.4 N),

even if the average thrust forces at 0.942 m/min are same in CD

and UAD (difference of only 0.5 N) One possible reason for

this could be the ultrasonic impacts at the exit in UAD as the

cutting speed is very slow The maximum reduction in the exit

delamination because of ultrasonic assistance was found at

9.42 m/min cutting speed in UAD in comparison to that in CD

One of the reasons for this reduction in the exit delamination

corresponds to the maximum reduction of thrust force (9.1 N)

at 9.42 m/min because of ultrasonic assistance, Figure 5 The

exit area delamination at 94.2 m/min was found to be lower in

UAD as compared to that in CD even if the average thrust

forces are same at in both At 282.6 m/min cutting speed, the

exit delamination was found to be higher in UAD in

comparison to that in CD as also shown in their corresponding

average thrust force data in the Figure 5

Figure 7: Area delamination factor for all the cutting speeds 4.2.2 Internal Damage

The internal damage was measured by measuring the maximum radius at each layer of the X-ray CT scanned hole through image processing in MATLAB Further, the average was taken between the drilled depths from 2 mm to 5 mm This average value was considered for the analysis for a particular hole The average values of the maximum radius for all the parameters are plotted in Figure 8 It was found that the average maximum radius (and hence internal damage) was lower in the case of UAD in comparison to that in CD at all the cutting speeds except 9.42 m/min At 9.42 m/min cutting speed, the internal damage was found to be similar in both the processes

of CD and UAD

The reduction in the average maximum radius because of ultrasonic assistance was found to be more at 94.2 m/min (20 µm) than that at 282.6 m/min cutting speed (6 µm) even if

no reduction in the thrust force because of ultrasonic assistance

at these cutting speeds was observed This happened because the effective normal rake angle at 94.2 m/min in UAD (39°) was more than that obtained at 282.6 m/min in UAD (33°), Figure 2(c) & (d), leading to a localization and reduction in the damage as mentioned by Arola et al [22]

Figure 8: Comparison of the average maximum radius in CD and UAD

1 1.05 1.1 1.15 1.2 1.25 1.3 1.35

Cutting speed (m/min)

CD UAD

3.05 3.06 3.07 3.08 3.09 3.1 3.11 3.12

Cutting speed (m/min)

CD UAD

4.2.3 Circularity

The circularity of each hole was calculated from the

difference between the maximum and the minimum radius in a

particular layer including the damage of the layer The average

of the circularity of all the layers between the drilled depths

from 2 mm to 5 mm was considered as the average circularity

of a hole The lower is the circularity value, the more circular

the hole is and the less damage has been produced The

circularity values for the various cutting speeds have been

plotted in the Figure 9 It was found that the average circularity

values were lower in the case of UAD as compared to those in

CD at all the parameters except 9.42 m/min cutting speed At

9.42 m/min cutting speed the circularity was found to be lower

in CD than that in UAD Also, at 282.6 m/min, the average

circularity for CD was found to be the highest amongst all the

cutting speeds which also corresponds to the highest ‘average

maximum radius’ of the internal damage at 282.6 m/min cutting

speed as shown in the Figure 8

Figure 9: Average circularity in CD and UAD at various cutting

speeds

The lowest circularity was found at 0.942 m/min cutting

speed in UAD A maximum of 33.8% reduction in circularity

was obtained in UAD at the cutting speed of 282.6 m/min in

comparison to CD Based on the circularity data, it can be

concluded that the holes drilled by UAD at 0.942 m/min cutting

speed have the lowest internal damage as also shown in the

Figure 8 Also, similar average maximum radius, Figure 8, and

higher circularity in UAD, Figure 9, indicate higher internal

damage in UAD at 9.42 m/min cutting speed as compared to

that in CD

4.2.4 Center deviation

During the drilling, because of the natural vibrations and

chattering, the drill sometimes deviates from its original center

which is known as deviation of the centers The ideal center of

the hole was fixed at the point of cross-section of the lines

joining the mid-points of the four machined marks around the

hole as shown in the Figure 6 The actual center of the hole in a

layer was found by the MATLAB itself by identifying a circular

loop The distance between the ideal and actual centers was

calculated for each layer of a hole Further, the average of the center deviations of the layers between the drilled depth of

2 mm and 5 mm was calculated This average was considered

as the average center deviation of a hole

The deviation in the centers in UAD was found to be lower than that in CD at all the cutting speed parameters except 9.42 m/min cutting speed, Figure 10 At 9.42 m/min cutting speed, the deviation was found to be similar (3 µm difference) in both the drilling processes of conventional drilling and UAD Figure

10 shows that the highest deviation in the center of the drilled hole was found at 282.6 m/min cutting speed in CD while the lowest deviation was found at 282.6 m/min cutting speed in UAD A reduction of 82.8% in center deviation was obtained at 282.6 m/min cutting speed

Figure 10: Average center deviation in CD and UAD at various cutting speeds

It can be concluded from the center deviation data that the drill was more stable in UAD than that in CD at all the cutting speeds Some selective cutting speeds in CD (in this case 9.42 m/min) may have similar center deviation as that in UAD

A similar metrological study of circularity, cylindricality, surface roughness and hole oversize was also performed by Akbari et.al [13] in the drilling of Inconel 738-LC alloy in UAD They also found reduced circularity and lower surface roughness in UAD as compared to those in CD at the hole-exits However, their study was limited to the hole-exits only unlike the current analysis of throughout the hole

4.3 Machined surface and SEM analysis

In order to analyze the internal machined surface of a hole, the material was cut diametrically opposite with a diamond coated precision saw and the internal cylindrical surface of the holes was analyzed in the scanning electron microscope (SEM) The machined surfaces in both the processes of CD and UAD

“look” similar at all the parameters when observed at a low magnification of 40 x The machined surfaces at 0.942 m/min

in CD and UAD are shown in the Figure 11 and Figure 12 respectively However, at the exit side, fibre pull-out was found

in the conventionally drilled hole, Figure 11 Apart from the damage in the laminates near exit, the entire machined surface

0

0.05

0.1

0.15

0.2

0.25

Cutting speed (m/min)

CD UAD

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

Cutting speed (m/min)

CD UAD

display similar damage Also, when observing at a higher

magnification (60,000 x), plastic deformation of the smeared

matrix was found in CD and UAD at 0.942 m/min cutting

speed

Figure 11: Machined surface in CD (0.942 m/min, 0.05 mm/rev)

Figure 12: Machined surface in UAD (0.942 m/min, 0.05 mm/rev)

At 9.42 m/min cutting speed, regular, sharp cracks were

found all over the smeared matrix in conventionally drilled

hole, Figure 13, while the matrix-softening and edge-rounding

was found at the cracked edges in the matrix of the machined

surface in UAD, Figure 14, at a higher magnification of

100,000x Also, the crack-ends were found to be round in UAD

which suggest the softening of the smeared matrix, Figure 14

This kind of softening of the matrix could happen either in the

case of temperature rise or if the property of the matrix would

have been transited from brittle to ductile Similar transition

from brittle to ductile of CFRP material in UAD was also

reported by Makhdum et al [7] They reported larger lengths of

the chips generating in UAD of CFRP material as compared to

that in CD and the reason behind formation of such larger chips

in UAD was asserted as transition from brittle to ductile

behavior This phenomenon is further discussed at the end of section 4.5 of this paper

Figure 13: Sharp and brittle cracks in CD at 9.42 m/min

Figure 14: Softened Matrix in UAD at 9.42 m/min cutting speed

At 94.2 m/min, Figure 15 and Figure 16, the machined surface in CD was found to have debris on throughout the machined surface while in UAD, the surface was machined smooth showing plastic deformation in the smeared matrix On the other hand, at 282.6 m/min, clean machined surface was found in CD, Figure 17, and a lot of debris was found on the machined matrix surface in UAD, similar to the one shown in Figure 15 One possible reason for generating the debris on the machined surface due to ultrasonic assistance at 282.6 m/min could be the intermittent cutting mechanism in UAD; however, there is no evidence to support this argument

Figure 15: Debris formation on the machined matrix surface in

CD (94.2 m/min, 0.05 mm/rev)

Figure 16: Clean machined matrix surface in UAD (94.2 m/min,

0.05 mm/rev)

Figure 17: Clean machined matrix in CD (282.6 m/min, 0.05

mm/rev)

4.4 Chip morphology

The chips were collected after drilling each hole on a double-sided carbon tape Further analysis was performed in SEM after applying the gold coating In order to have overall information of the average chip formation, 3 random samples were analyzed out of every drilled hole

The chips obtained in both the processes of CD and UAD were found to be similar In general, three types of chip-fragments were found in both the processes –

 Large (1.4 mm) fragments of chip

 Small (100 µm to 300 µm length) fragments of chip

 Small broken fibres (10 to 200 µm length) The large fragments of the chips were obtained at 0.942 m/min and 282.6 m/min cutting speeds in CD, Figure 18

Figure 18: Large fragment of the chip obtained in CD (0.942 m/min, 0.05 mm/rev)

The short broken chip-fragments with the length between

100 µm to 300 µm were the most common fragments types obtained in all the cutting speeds in both the processes of CD and UAD, Figure 19 These types of chip fragments were found

in a large in portion at 282.6 m/min cutting speed in CD and UAD Because of the fibre-matrix debonding, the broken loose fibres were also obtained at every cutting speed in both the processes of CD and UAD as shown in Figure 21

At every cutting speed, there was a portion of chips having plastic deformation in the matrix, Figure 22 and 23 These types

of chips were obtained in a large portion at the cutting speeds of 94.2 m/min and 282.6 m/min in both CD and UAD Though a small portion of chips having plastic deformation were also found at the cutting speed of 0.942 m/min in CD and UAD Hence, there was a mixture of chips found at every cutting speed in CD and UAD and in particular, no clear difference was found in the chips formed in CD and UAD at the current machining parameters

Figure 19: Long and short, broken chip-fragments in UAD at

0.942 m/min cutting speed

Figure 20: Long chip-fragments obtained in UAD at 282.6 m/min

cutting speed along with short and regular chip-fragments

Figure 21: Loose fibres obtained in the drilling of CFRP in CD as

well as UAD (9.42 m/min, 0.05 mm/rev)

Figure 22: Plastic deformation in the matrix of the chip in UAD at

0 942 m/min cutting speed

Figure 23: Definite chip formation and plastic deformation in the matrix obtained at 94.2 m/min cutting speed

Makhdum et al [15] reported generation of larger chips in UAD in comparison to that in CD whereas in the present research there was no difference found in the chips A possible explanation for this could be the different ultrasonic oscillation amplitude and different matrix of the CFRP material; however, the matrix used in their work was not reported

4.5 Cutting temperature measurement

The cutting temperature is still a point of discussion amongst the researchers Some researchers have claimed the cutting temperature to be higher in UAD as compared to CD [14] In the present study, the cutting temperature was measured using the “surface mount” thermocouple by sticking it between two CFRP plates as shown in the Figure 24 The other end of the thermocouple was connected to the thermocouple data logger connected to a computer PicoLog software was used for data acquisition and storage To avoid damaging the welded joint of thermocouple, drilling was performed within a distance

of 1 mm from the welded joint The dimensions of the CFRP

plates for temperature measurement were 70 mm x 25 mm x

7.4 mm and the rest of the setup was similar to Figure 3

Figure 24: Thermocouple setup for cutting temperature

measurement

The cutting temperature variation in CD and UAD is

plotted in Figure 25.The cutting temperature was found to be

11°C higher in UAD at 282.6 m/min than that in CD while at

94.2 m/min cutting speed, it was 20°C lower in UAD than that

in CD At the lower cutting speeds of 0.942 m/min and 9.42

m/min, the cutting temperature did not show a significant

difference between CD and UAD (2°C and 4°C respectively)

Figure 25: Cutting temperature measurement at various cutting

speeds in CD and UAD 1

The glass transition temperature of bismaleimide matrix

lies between 200°C to 288°C based upon the composition [25]

The cutting temperature at 9.42 m/min was 126.56°C in UAD

(<200°C) Hence, the softening of the smeared matrix shown in

the Figure 14 confirms the smeared matrix to be transiting from

brittle to ductile in UAD This explains the reason behind the

higher internal damage in UAD than that in CD even after a

reduction of 9.1 N in thrust force at 9.42 m/min cutting speed

as in a recent research by Barnes et al [26], it has been

identified that softer matrix enhances fibre pull-out generating

higher internal damage even if the thrust force is reduced

However, in their work, they also found higher exit

delamination and lower thrust force in drilling of softer matrix

material contradicting the current case of the maximum

reduction in the exit delamination at 9.42 m/min due to

ultrasonic assistance, Figure 7, as compared to that in CD A

possible explanation for this could be the higher resistance to

crack propagation in ductile materials as compared to that in

1 For the reproducibility of the cutting temperature results, the repetition

of the cutting temperature measurement experiment is planned soon

brittle materials in conjunction to lower thrust force causing reduction in the exit delamination [27]

5 CONCLUSIONS

Based upon the experimental study of ultrasonic assisted drilling without coolant at various cutting speeds with a constant feed rate (mm/rev), following conclusions can be drawn –

a) Similar values of thrust force are obtained in CD and UAD at all the cutting speeds at constant feed rates Though a small reduction in thrust force is obtained (9.1

N in the present case) at some cutting speeds (in the present case 9.42 m/min)

b) A reduction in the average thrust force causes the reduction in the exit delamination (as the case of 9.42 m/min cutting speed) However the internal damage

of the hole was not found to be related to the average thrust force when comparing the CD and UAD As at 9.42 m/min cutting speed the average thrust force reduced but not the internal damage Also, at 94.2 m/min cutting speed, the average thrust force was similar in CD and UAD while the internal damage was lower in UAD c) Higher effective normal rake angles obtained in UAD, Figure 2, is one of the factors causing the reduction in the internal damage during the drilling of CFRP material Also, The internal damage increases with increasing the cutting speed in both CD and UAD (an increase of 46 µm

in CD and 44 µm in UAD in the present case when increasing the cutting speeds from 0.942 m/min to 282.6 m/min)

d) The holes are more circular when drilled through UAD in comparison to CD because of having lower internal damage as compared to that in CD A maximum of 33.8% reduction in circularity was obtained in the present case at the cutting speed of 282.6 m/min

e) Less center deviation and hence more stable drilling is obtained in UAD as compared to CD A maximum of 82.8% reduction was obtained in the center deviation in UAD as compared to that in CD at cutting speed of 282.6 m/min

f) In order to obtain the minimum damage, the cutting speed would have to be variable in though out the drilling depth

It will have to be high at the exit (282.6 m/min in the present case) and low during the drilling of intermediate layers of CFRP material

g) The chip fragments generated in CD and UAD are not different at current machining parameters

h) The cutting temperature is similar in CD and UAD at lower cutting speeds (0.942 m/min and 9.42 m/min in the present case) and it is different in CD and UAD at higher cutting speeds Further research is required to develop a thorough understanding about the deviation of cutting temperature because of ultrasonic assistance

6 NOMENCLATURE

𝛾 𝐷 = dynamic axial rake angle at point i,

Surface mount

thermocouple

1 mm

Drilling

CFRP Plates

0.942, 64.61

9.42, 122.59

94.2, 90.39

282.6, 70.76 0.942, 62.54

9.42, 126.56

94.2, 70.25

282.6, 81.66

0

20

40

60

80

100

120

140

Cutting Speed (m/min)