On the education of the real geometry of mechanical parts through the tolerancing and the tridimensional metrology

A comparative study on conventional and orbital

drilling of woven carbon fiber reinforced epoxy

laminates

A. Sadek

(1)

, M. Meshreki

(2)

, Z. Shi

(2)

and H. Attia

(1,2)*

(1)

Department of Mechanical Engineering, McGill University, Montreal, QC, Canada

(2)

Aerospace Manufacturing Technology Centre, Institute for Aerospace Research,

National Research Council Canada, Montreal, QC, Canada

*helmi.attia@nrc.ca / helmi.attia@mcgill.ca

Abstract:  Drilling  of  Carbon  Fiber  Reinforced  Polymers  (CFRPs)  is  a  challenging
process of major economic and safety concerns for the aerospace industry. This research
work  presents  an  experimental  investigation  of  the  orbital  drilling  technique  versus
conventional  drilling  of  woven  CFRP  laminates.  Moreover,  the  tribological  aspect  of
conventional  drilling  is  investigated  through  studying  the  effect  of  the  dry  machining,
air  cooling,  and  Minimum  Quantity  Lubrication  (MQL)  conditions  on  the  quality  of
produced holes. This paper studies the attributes of hole quality for a range of material
removal rates (MRR): 11.4 to 22.8 cm³/min, using a three flute solid carbide 9.53 mm
drilling  tool  for  conventional  drilling.  Corresponding  conditions  were  employed  to
produce 9.53 mm diameter holes using a 6.35 mm four flute end-mill for Conventional
Orbital  Drilling  (COD).  The  preliminary  results  of  Superabrasive  Orbital  Drilling
(SAOD)  using  superabrasive  diamond-coated  steel  tools  are  also  presented.  MQL  and
air  cooling  resulted  in  reduced  cutting  temperatures,  compared  to  conventional  dry
drilling. COD produced delamination-free holes with enhanced hole surface quality, in
addition to the pronounced reduction in cutting temperatures and thrust forces.
Key  words:  Conventional  and  orbital  drilling,  Carbon  Fiber  Reinforced  Polymers
(CFRPs), Minimum Quantity Lubrication (MQL), Machining-induced defects.

INTRODUCTION

The  continuous  evolution  of  usage  of  polymer  composites  in  the  aerospace  industry
attracts more research efforts on different post-processing techniques applied to polymer
composites. Various novel drilling techniques e.g. Conventional Orbital Drilling (COD)
and  Superabrasive  Orbital  Drilling  (SAOD)  are  being  used  by  today‟s  aerospace
manufacturers.  However,  not  much  is  known  about  the  effects  of  employing  such
processes on the quality attributes of the produced part.

Delamination of Carbon Fiber Reinforced Polymer (CFRP) layers is one of the

most critical defects associated with drilling of composite laminates. Delamination
could take place at the hole entry and/or exit, or through the layers of the laminate
thickness. The amount of delamination is controlled by the process parameters and the

quality  of  the  bonding  between  the  layers  of  the  material.  The  entry  delamination
becomes  dominant  in  the  case  of  high  speed  drilling,  while  the  exit  delamination
remains  the  most  common  type  to  take  place  during  intermediate  and  low  drilling
speeds [Rawat et al., 2009].

COD and SAOD are two emerging orbital drilling techniques that have

demonstrated  a  potential  for  eliminating  defects  associated  with  drilling  of  CFRP
laminates [Lindqvist et al., 2001] [Persson et al., 1997]. Such potential stems from the
nature of these processes, which is based on eliminating the axial dwell of the drilling
tool  over  the  uncut  material  thickness  [Iyer  et  al.,  2006].  This  is  achieved  through
rotating a cutting tool of a smaller diameter about its own axis and simultaneously about
the axis of the desired cylindrical hole, located at an offset distance from the tool axis.
The cutting tool used for COD is usually an end-mill or a cutting head with PCD inserts,
while a diamond coated mandrel is used for SAOD (Figure 1).

Figure 1 Description of tool types and motion in (a) COD, and (b) SAOD processes

The research work published in [Lindqvist et al., 2001] showed the results of using

COD to produce holes in sandwich constructions with aluminum and carbon
honeycomb core for aerospace applications. This process produced delamination-free
holes with allowable geometric accuracy. The tool experienced, however, considerable
chipping on the main cutting edge, due to the lack of tool dynamic stability.

Using fluted diamond-coated abrasive cutters for edge trimming of graphite/epoxy

laminates  was  investigated  in  [Colligan  et  al.,  1999].  Delamination-free  surfaces  were
successfully  produced.  The  surface  quality  and  the  cutting  forces  were  reported  to  be
directly proportional to the grit size and the material removal rate (MRR), respectively.
The  feed  rate  was  reported  to  be  of  a  minor  effect  on  the  surface  quality.  The  results
reported in [Persson et al., 1997] compared the effect of producing holes via SAOD and
other  conventional  drilling  processes  on  the  behavior  of  the  drilled  CFRP  laminates
under  fatigue  loading.  A  diamond  coated  mandrel  rotating  at  a  rotational  speed  of
25,000  rpm  and  an  orbital  speed  of  125  rpm  were  employed.  An  enhanced  static  and
fatigue behaviors were reported for plates with holes produced  by SAOD compared to
the behavior of plates with holes produced by conventional drilling.

(a)

(b)

The effect of air cooling and Minimum Quantity Lubrication (MQL) on the

produced surface finish, tool wear, cutting forces, and torques in drilling of Aluminum
was investigated in [Tasdelen et al, 2008]. However, no such study has been carried out
for drilling of CFRPs up to the present.

The objective of this research work is to conduct a comparative study between

SAOD, COD, and conventional drilling under dry, air cooled, and MQL conditions.
Such comparative study aims at ranking these processes, in terms of their capabilities
and limitations.

EXPERIMENTAL SET-UP

All the drilling tests were performed on a 5-axis Makino A88  machining centre, having
a 50 kW spindle, maximum spindle speed of 18,000 rpm and maximum feed rate of 50
m/min. The COD and SAOD tests were conducted under dry conditions. Conventional
drilling  tests  were  conducted  under  dry,  air  cooled,  and  MQL  conditions.  The  MQL
tests  were  performed  using  Vogel  LubriLean  Vario  system  connected  through  the
spindle.  The  aerosol  was  supplied  through  the  coolant-through-holes  of  the  tool.  Air
pressure of 0.6 MPa and oil lubricant flow of 9 ml/hr were used for MQL drilling tests.
Figure 2(a) shows the experimental set-up used for the drilling experiments. Fixture (1)
was  used  to  hold  the  CFRP  laminates.  The  fine  chips  were  evacuated  using  a  special
vacuum  arrangement  (2).  The  cutting  forces  were  recorded  using  a  3-component
dynamometer  (3),  Kistler  9255B,  and  5070A  Kistler  charge  amplifier.  Cutting
temperatures were measured using a FLIR ThermoVision A20M Infrared camera (4) at
the hole exits.

Figure 2; (a) Experimental set-up, (b) Four flute end-mill, diamond coated steel

mandrel and three flute drill bit used for drilling of woven CFRP laminates

The test material used was a quasi-isotropic laminate comprising 35 plies of 8-

harness  satin  woven  graphite  epoxy  prepreg.  The  laminate  was  manufactured  by
autoclave  molding  with  a  cure  time  of  60  min  at  127  °C  under  516.75  kPa  autoclave
pressure  to  produce  a  final  cured  thickness  of  6.35  ±  0.02  mm.  For  conventional
drilling, a three-flute 9.53 mm diameter drill with a 30° helix angle, 150° point angle, a

(a)

(b)

fluted length of 63 mm and an overall length of 100 mm was used. The COD tests were
performed using a four flute 6.35 mm end-mill with a fluted length of 19 mm and a total
length of 63 mm. The SAOD tests were performed using a (100 Grit) diamond coated
6.35 mm

steel mandrel tool, with a coating length of 6.35 mm and an overall length of

76 mm. Figure 2(b) shows the three types of tools.

Delamination at the entry and exit of holes were investigated using the Olympus

Model GZX 12 optical microscope. Hole circularity and hole size errors along the depth
of the hole were measured using a „„Mitutoyo-Mach 806‟‟ coordinate measuring
machine (CMM). Surface roughness measurements were done using a Form Talysurf
series 2 surface profilometer.

The test matrix for this research work is given in Table 1. Two sets of experiments

were  conducted  in  order  to  perform  this  comparative  study.  The  first  group  of  tests
compared  the  conventional  and  COD  tests  that  were  performed  at  the  same  range  of
rotational speeds; 6,000 to 12,000 rpm, and MRRs of 11.4 to 22.8 cm³/min. Such MRR
range  was  achieved  by  a  single  feed  rate  of  0.06  mm/rev  for  the  conventional  drilling
tests, versus a range of helical feed rates of 1,200 to 2,400 mm/min with a helical pitch
of  3  mm  for  the  COD  tests.  The  selection  of  this  operating  range  was  limited  by  the
capability of the machine tool to perform such an intricate helical motion at high helical
feeds.  The  second  set  of  experiments  was  conducted  to  compare  the  performance  of
COD  and  SAOD  techniques  at  the  same  rotational  speed  of  16,000  rpm,  with  orbital
feed  rates  600  to  1,200  mm/min  and  a  helical  pitch  of  1  mm  to  produce  a  range  of
MRRs 1.9 to 3.8 cm³/min.

Table 1 Test matrix for conventional, COD, and SAOD drilling

Set

Conventional (Dry,

Air, and MQL)

COD

SAOD

MRR

(cm³/min)

Speed

(rpm)

Feed rate

(mm/rev)

Orbital feed

(mm/min)

Pitch

(mm)

Orbital feed

(mm/min)

Pitch

(mm)

6,000

0.06

1,200

11.4

9,000

1,800

17.1

12,000

2,400

22.8

16,000

600

1.9

900

2.85

1,200

3.8

RESULTS AND DISCUSSION

In this section, the quality attributes of the holes produced using the SAOD, COD, and
conventional drilling under dry, air cooled, and MQL conditions are compared over the
ranges of MRR specified in Table 1. The MRR is a common parameter for conventional
and orbital drilling techniques that also has an indication on the productivity of the
process. For conventional drilling experiments, the MRR was increased by increasing
the rotational speed while the feed rate was fixed. For COD and SAOD the MRR was

increased by increasing the orbital feed while the rotational speed remained constant,
(Table 1).

1.1. Delamination

Hole entry and exit delamination were investigated. None of the holes produced by the
entire  set  of  experiments  has  shown  considerable  entry  delamination.  The  exit
delamination  for  holes  produced  at  MRR  22.8  cm³/min  using  dry,  air  cooled,  MQL
conventional drilling, and COD are shown in Figure 3 (a), (b), (c), and (d) respectively.
The delamination-free holes produced at MRR 3.8 cm³/min using COD and SAOD are
shown in Figure 3 (e), and (f), respectively.

The holes produced by conventional drilling showed considerable amounts of exit

delamination, while COD and SAOD resulted in delamination-free holes. The
delamination damage could be quantified by a delamination factor (ф

max

nominal

where “D

max

” is the maximum diameter that contains the observed delamination zone,

and “D

nominal

” is to the nominal hole diameter, as shown in Figure 3 (a).

Figure 3;Exit Delamination for holes produced at MRR 22.8 cm³/min using (a) Dry, (b)

Air cooled, (c) MQL conventional drilling, (d) COD at MRR 22.8 cm³/min, (e) COD at

MRR 3.8 cm³/min, and (f) SAOD at MRR 3.8 cm³/min

Figure 4 compares the exit delamination factors associated with dry, air cooled, and

MQL  conventional  drilling  techniques.  Two  main  trends  could  be  concluded  from
Figure  4.  First,  as  the  MRR  is  increased  from  11.4  to  22.8  cm³/min  the  exit
delamination  factor  increases  by  an  average  of  7%,  23%,  and  13%  for  dry,  air,  and
MQL, respectively. Second, the MQL resulted in about  6%  and  8.5% reduction in  the
delamination factor compared to the dry and air conditions, respectively.

The results shown in Figure 4 could be explained through investigating the drilling

thrust forces for each condition. Figure 5(a) and (b) show an increase in thrust forces in
all of the drilling techniques as the MRR increased, as a consequence of the higher feeds

(a)

(b)

(c)

(d)

(e)

(f)

max

hole

required to achieve higher MRRs. Figure 5(a) shows that the MQL resulted in a thrust
force  average  reduction of  17%  and  20%  compared  to  the  dry  and  air  cooled  drilling,
respectively, which resulted in the relatively lower delamination factors exhibited by the
MQL drilling. The thrust forces associated with COD are 40% to 50% lower than those
obtained  by  conventional  drilling  under  dry,  air  cooled,  and  MQL  conditions.  Such
percentage  of  reduction  in  thrust  forces  of  the  COD  is  attributed  to  the  helical  tool
motion  that  eliminates  the  stationary  tool  center  [Persson  et  al.,  1997],  and  is  in
agreement with the percentage reported in [Langella et al., 2001].

Figure 4; Exit delamination factors of conventional drilling under

dry, air cooled, and MQL conditions

The thrust forces for the SAOD and COD at lower MRRs are compared in Figure

5(b). SAOD resulted in 60% to 75% increase in thrust forces compared to the COD.
This increase can be attributed to (i) the absence of the real oblique cutting condition in
the SAOD process where higher specific cutting forces are required due to the smaller
equivalent chip size, and (ii) the higher effect of frictional forces generated as a result of
the engagement of the abrasive particles with the epoxy matrix, as well as the particle‟s
shallow clearance [Ahmad et al., 2009][Colligan et al., 1999].

Figure 5; Drilling thrust forces for (a) COD and conventional drilling under

dry, air, and MQL conditions. (b) COD and SAOD at lower MRR levels

During the drilling process, the remaining thickness of material near the hole exit

starts to fail experiencing delamination above a certain threshold drilling thrust force

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

11.4

17.1

22.8

Exi

lam

ation

Fact

(ɸ

)

Material Removal Rate (cm³/min)

Dry

Air

MQL

100

150

200

250

300

350

11.4

17.1

22.8

(

)

Material Removal Rate (cm³/min)

Dry

Air

MQL

COD

100

150

200

250

300

350

1.9

2.85

3.8

(

)

Material Removal Rate (cm³/min)

SAOD

COD

(a)

(b)

level F

t,threshod

[Ho-Cheng et al., 1990]. The results in Figure 5(a) and (b) suggest that the

level of the drilling thrust force in the COD and SAOD did not exceed this threshold
value, since no delamination was associated with these techniques.

1.2. Drilling Temperature

Thermal  damage  during  machining  is  a  critical  aspect  that  must  be  considered  while
studying  the  impact  of  different  process  parameters  on  hole  quality  attributes.  CFRPs
are sensitive to thermal damage that causes material deterioration and decomposition at
the machined surface and/or within the heat affected zones. Figure 6 shows a schematic
representation of the temperature measurement using the IR camera. The temperature is
monitored for the point of interest (point A), which is the intersection of the axis of the
hole  to  be  drilled  and  the  outermost  layer  on  the  exit  side  of  the  CFRP  laminate.  The
maximum temperature takes place when point (A‟) on the tool tip coincides with point
(A) at the instant of the exit penetration of the tool tip. The results shown in Figure 7(a)
and (b) represent the maximum recorded temperature for each drilling condition.

Figure 6; Schematic of temperature measurement using the IR camera

Figure 7; Maximum temperature at the tool exit for (a) COD and conventional drilling

under dry, air, and MQL conditions, and (b) COD and SAOD at lower MRR levels

Figure 7(a) shows the exit temperatures for COD and for conventional drilling

under  dry,  air  cooled, and  MQL  conditions.  In all  of  the  involved  techniques,  the  exit
temperature  increased  as  the  MRR  increased.  This  is  a  result  of  the  friction-induced
temperature rise associated with higher rotational speeds. Compared to conventional dry
drilling,  the  use  of  MQL  and  air  cooling  resulted  in  temperature  reduction  by  40%  to
45% and 25% to 30%, respectively. COD resulted in a significant temperature reduction
of  45%  to  60%  compared  to  conventional  dry  drilling  due  to  the  intermittent  cutting
nature of COD that creates a cyclic tool cooling effect during cutting.

100

150

200

250

300

350

11.4

17.1

22.8

Exi

Tem

ratu

(

°C)

Material Removal Rate (cm³/min)

Dry

Air

MQL

COD

100

150

200

250

300

350

1.9

2.85

3.8

Exi

Tem

ratu

(

°C)

Material Removal Rate (cm³/min)

SAOD

COD

(a)

(b)

Material decomposition temperature

Figure 7(b) compares the exit temperatures obtained in COD to SAOD at lower

MRR levels. COD resulted in temperature levels 40% to 45% lower than SAOD, which
involves higher specific cutting energies and consequently significant effect of friction-
induced  temperature  rise.  Thermogravimetric  Analysis  (TGA)  test  was  performed  in
order to determine the decomposition temperature of the CFRP material being used.  In
Figure  7(a)  and  (b),  only  the  dry  and  SAOD  processes  at  the  highest  MRR  slightly
crossed the decomposition temperature limit.

1.3. Hole Surface Roughness

The  arithmetical  mean surface  roughness „Ra‟  was  measured  along  the  hole  depth  in
order to  represent  the quality of the produced hole surface.  Figure  8(a) shows  that  the
surface  quality  produced  by  MQL  drilling  was  practically  better  or  equal  to  that
produced  by  dry  and  air  cooled  drilling.  The  COD  produced  better  surface  quality
compared to conventional drilling at low and medium MRR. However, the effect of the
helical path errors and tool dynamics at high MRR resulted in a lower surface quality.

Figure 8(b) shows a pronounced increase in surface roughness of the holes

produced by SAOD compared to the enhanced surface quality produced by COD. This
is  attributed  to  the  relatively  coarse  tool  coating  grains  (Grit  100).  The  relatively  low
tool  rotational  speed  could  be  another  factor  for  the  high  surface  roughness  values  as
shown  in  Figure  8(b).  Finer  grains  could  have  been  used  to  produce  better  surface
quality but this would have led to higher cutting forces. Tool design considerations, e.g.
spiral  flutes  for  chip  flow,  could  dramatically  expand  the  capabilities  of  the  SAOD
technique [Colligan et al., 1999].  Further investigation of the effect of tool design and
the grit size on the cutting forces and surface quality is underway.

Figure 8;Hole surface roughness along the hole depth for (a) COD and conventional

drilling under dry, air, and MQL conditions. (b) COD and SAOD at lower MRR levels

1.4. Hole Size Error

The hole size error is defined by the ratio of the difference between the actual hole
diameter “D

actual

”, and the nominal hole diameter “D

nominal

” to the nominal hole

diameter (Hole Size Error (%) = (D

actual

– D

nominal

)/D

nominal

,). A negative error indicates

that the produced hole is smaller than the nominal size. The limits of allowable
tolerance were defined based on the dimensional tolerance of the drill bit which is

0.0

1.0

2.0

3.0

4.0

5.0

6.0

11.4

17.1

22.8

(μ

Material Removal Rate (cm³/min)

Dry

Air

MQL

COD

0.0

1.0

2.0

3.0

4.0

5.0

6.0

1.9

2.85

3.8

(μ

Material Removal Rate (cm³/min)

SAOD

COD

(a)

(b)

±0.0127 mm that represents ±0.13% hole size error. Figure 9(a) and (b) show that all the
drilling tests produced hole sizes within acceptable tolerance limits, except the COD at
MRRs  11.4  and  17.1  cm³/min.  Such  excessive  errors  could  be  a  consequence  of  the
helical path errors generated by the controller of the machine tool and/or the deflection
of  the  tool  under  the  action  of  the  tangential  force.  However,  the  holes  produced  by
COD  at  MRR  22.8  cm³/min  were  within  tolerance.  This  could  be  due  to  the
compensation effect of tool dynamics. Figure 9 (b) shows that COD produced hole sizes
within  acceptable  tolerance  limits  at  MRR  of  1.9  to  3.8  cm³/min  due  to  lower  helical
path errors and tool dynamics effects at this range of MRR.

Figure 9; Hole size error for (a) COD and conventional drilling under dry, air, and

MQL conditions. (b) COD and SAOD at lower MRR levels

CONCLUSIONS

The following conclusions can be drawn from this study:

1. Using MQL with conventional drilling resulted in a drilling thrust force reduction

of 17% to 20%, compared to the dry and air cooled conventional drilling.
However, such reduction was not sufficient to eliminate the exit delamination. On
the other hand, the COD and SAOD produced delamination-free holes.

2. Using MQL and air cooling with conventional drilling resulted in up to 45% and

30%  temperature  reduction,  respectively.  This  suggests  that  using  MQL  and  air
cooling becomes more essential at high MRR ranges where the exit temperatures
of  dry  conditions  exceed  the  material  decomposition  temperature.  The  COD
resulted  in  a  temperature  reduction  of  up  to  60%  compared  to  conventional  dry
drilling  at  high  MRRs.  The  SAOD  resulted  in  significantly  higher  temperatures
compared to COD at low MRR ranges. However, further design considerations of
the superabrasive tool can significantly enhance the performance of the process.

3. The COD conducted at low MRR resulted in superior hole surface quality. On the

other  hand,  the  COD  at  high  MRR  produced  lower  surface  quality  compared  to
conventional  drilling  due  to  helical  path  errors  and  tool  dynamics  effect.  MQL
and  air  cooling  did  not  show  a  significant  improvement,  in  terms  of  surface
quality. The SAOD produced significantly lower surface quality as a result of the
relatively coarse diamond grains used.

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

Size

rro

)

Material Removal Rate (cm³/min)

Dry

Air

MQL

COD

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

Size

rro

)

Material Removal Rate (cm³/min)

SAOD

COD

(a)

(b)

1.9 2.85

3.8

11.4

17.1 22.8

Upper limit

Lower limit

Upper limit

Lower limit

4. All the drilling tests produced hole sizes within acceptable tolerance limits, except

the COD tests conducted at the medium MRR range. This was attributed to the
helical path errors and tool deflection under the action of the tangential force.

5. The COD produced delamination-free holes, with superior surface quality and

significantly reduced temperatures. However, multi-axis machine centers have
limitations on performing the COD at high MRR.

ACKNOWLEDGEMENTS

This work was conducted under the partial financial support of the Natural Sciences and
Engineering  Research  Council  of  Canada  (NSERC),  which  the  authors  greatly
appreciate.  The  authors  would  like  to  acknowledge  the  support  of  the  Aerospace
Manufacturing  Technology  Centre  (AMTC),  Institute  for  Aerospace  Research  (IAR),
National Research Council Canada (NRC), where the experiments were conducted.

REFERENCES

[Lindqvist et al., 2001] Lindqvist, R.; EriKsson I.; Wolf M.; "Orbital drilling of

sandwich constructions for space applications"; In: Proceedings of the 2001 SAE
Aerospace Automated Fastening Conference; Seattle 2001

[Persson et al., 1997] Persson E.; Eriksson I.; Zackrisson L.; "Effects of hole

machining defects on the strength and fatigue life of composite laminates";
Composites Part A: Applied Science and Manufacturing; Vol. 28 (2); pp. 141-151;
1997; ISSN 1359-835X

[Iyer et al., 2007] Iyer R.; Koshy P.; Ng E.; "Helical milling: An enabling technology

for hard machining precision holes in AISI D2 tool steel"; International Journal of
Machine Tools and Manufacture; Vol. 47(2); 2007; pp. 205-210; ISSN 0890-6955

[Colligan et al., 1999] Colligan K.; Ramulu M.;"Edge trimming of graphite/ epoxy with

diamond abrasive cutters"; vol. 121(4); pp. 647-655; ISSN 1087-1357

[Tasdelen et al., 2008] Tasdelen B.; Wikblom T.; Ekered S.; "Studies on minimum

quantity lubrication (MQL) and air cooling at drilling"; Journal of Materials
Processing Technology; Vol. 200(1-3); pp. 339-346; ISSN 0924-0136

[Rawat et al., 2009] Rawat S.; Attia H.; "Characterization of the dry high speed drilling

process of woven composites using Machinability Maps approach"; CIRP Annals -
Manufacturing Technology, Vol. 58(1); 2009;pp. 105-108; ISSN 0007-8506

[Langella et al., 2005] Langella A.; Nele L.; Maio A.; "A torque and thrust prediction

model for drilling of composite materials"; Composites Part A: Applied Science
and Manufacturing; Vol. 36(1); 2005; pp. 83-93; ISSN 1359-835X

[Ho-Cheng et al., 1990] Ho-Cheng H.; Dharan C.; "Delamination During Drilling in

Composite Laminates"; Transactions of ASME Journal of Engineering for
Industry; Vol. 112(3); 1990; pp. 236–239

[Ahmad et al., 2009] Ahmad J.S.; "Conventional Machining of FRPs"; in: Machining

of Polymer Composites; Springer 2009; ISBN 978-0-387-35539-9