Design of Double Cone Twist Drill Geometry to Improve the ...

DESIGN OF DOUBLE CONE TWIST DRILL GEOMETRY TO IMPROVE THE HOLES QUALITY WHILE DRILLING IN MULTI-STACK MADE OF CFRP/AL

Redouane Zitoune Institut Clément Ader

Toulouse, France Krishnaraj Vijayan

PSG College of Technology Coimbatore, India

Sofiane Alam bouacif Institut Clément Ader

Toulouse, France

Collombet Francis Institut Clément Ader

Toulouse, France

ABSTRACT

In this paper, experimental and numerical studies on

drilling of multi-material made of CFRP laminate sandwiched with aluminum part have been carried out. These tests have been conducted using carbide drills (K20) to study the influence of spindle speed, feed rate and lip length of the double cone drill on cutting forces and holes quality. From the experimental study it was found that the double cone drills used for the experimental analysis encountered less thrust force compared to the standard twist drill. In addition the quality of the holes after drilling was evaluated, and found that no delamination (entry and exit of the hole of the composite part) even at higher feed rate (> to 0.1 mm/rev). By numerical modeling, it was noticed that, after machining with standard twist drill at higher feed rates, the thrust force generated induces a small opening at the CFRP/Al interface.

INTRODUCTION

Drilling and fastening of the hybrid materials in one-shot operation reduce manufacturing time. One of the most common problems encountered during automatic drilling of sandwich structures are that (i) the continues chips of aluminum passing through composite materials can damage them, and (ii) the high thrust force of drilling can separate CFRP and aluminum plates which contributes to the accumulation of the of aluminum chip and carbon dust at the interface of CFRP/Al. LATECIS a French company in Toulouse area, developed a system called OPERA, dedicated for one shot drilling. Ideally the OPERA system is expected to handle high production rates (1 attachment mounted approximately at every 12 seconds) [1-2]. The automation of these tasks must also enable greater mounting precision, improved ergonomics, health and safety of the operators, particularly for the new hybrid materials like composite/titanium or composite/aluminum assemblies. Due to the different mechanical properties of materials constituting the

hybrid panels, their machinability remains an open problem and can prevent the automation of these tasks.

From the literature review, it is found that composite materials especially Carbon Fibre Reinforced Polymers (CFRP) are often used to a larger extent due to their downsizing and rightsizing impacts. However, due to their laminated structure and the anisotropy of the carbon fiber reinforce plastics several damage are seen during drilling. The damages are peel up delamination at the hole entry, thermal alteration, fibrepull-out and fuzzing on the wall of the hole, and exit delamination and uncut fibre at the hole exit [3-5]. Moreover, the mechanism of material removal in composite materials is strongly influenced by the relative angle between the direction of the cutting speed and the direction of fiber [6]. Sakuma et al. [4] drilled holes using four drill materials and investigated drill wear pattern, flank wear width and cutting forces. Many researchers investigated the effect of tool geometry on drilling of polymer composite materials. Many of the modified geometries (Zhirov point, Brad and Spur, step drill, candle stick, saw drill, etc) are difficult to regrind [5]. Krishnaraj et al. [7] drilled composites at high spindle speed and studied the influence of tool geometry. They reported that double cone drill offers better surface finish when compared to standard twist drill and Brad & Spur drill. In particular double cone drills which the length of the principal cutting edge is equal to the length of the secondary cutting edge offer better finish [7-8].

Unlike composites, the material removal during machining of aluminum is mainly done by shearing the material. In the case of drilling of aluminum and its alloys, the main problem is the adhesion of aluminum on the main cutting edges (BUE), on the rake face and on the flutes of the drill (BUL). This bonding is responsible for premature wear of the cutting tool, for the poor surface finish of the hole and the variations in the diameter of the hole. Several authors showed that, with low cutting speeds

1 Copyright © 2014 by ASME

Proceedings of the ASME 2014 International Mechanical Engineering Congress and Exposition IMECE2014

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

IMECE2014-36526

Vacuum

Sensors to monitor temperature

Rigid mould

(up to 25 m/min), bonding of aluminum occurs at the rake face and at the main cutting edges [9–11]. One way to overcome problems related to the machining of aluminum and its alloys is to increase cutting speeds. However, machining at high cutting speeds (e.g. 300 m/min) causes a significant increase in the cutting temperature (above 300°C) [11]. At this temperature a chemical reaction between aluminum and cobalt occurs to form micro welding of aluminum on the cutting edges of the tool by diffusion. In the experimental works of [12,13] the authors during the dry turning of aluminum alloy have shown that, in the first time of machining, BUL is caused by thermo-mechanical mechanisms. Once the BUL is formed, the initial cutting conditions change thus enabling the BUE formation through mechanical adhesion. However, during continuous machining, the BUE continues to grow until a critical thickness is reached, and once it reaches the point, it is plastically extended over the BUL due to the action of the mechanical forces. Additionally, the chips shape and length of the chip passing through the hole as well as built up edges of aluminium at the primary cutting edges combined with increased tool wear affect the hole quality [1-2, 14]. In [15-16], authors have studied drilling of Gr/Bi-Ti in the context of process conditions and cost optimization. The obtained results show that, with the carbide twist drill, it is preferable to drill with a small spindle speed (inferior to 1000 rpm) and a less average value of feed rate (around 0.08 mm/rev).

Among the various tool geometries investigated, double cone drill was found to offer many advantages compared to the modified geometries. Only a few investigations on drilling of CFRP laminates using double cone drill have been reported. In literature, the double cone drills are optimized for drilling of metallic materials (steel or aluminum). However, there is little information about the influence of the geometry of the double cone drill on the quality of holes machined.

In this paper, experimental study on drilling of multi-material made of CFRP laminate sandwiched with aluminum part has been carried out. These drilling tests have been conducted using carbide drills (K20) to study the influence of spindle speed, feed rate and lip length of the double cone drill on cutting forces and holes quality. The objective is to identify the better machining parameters and the tool geometry, which are able to machine with respect of the industrial criteria imposed by LATECIS Company. These criteria are discontinuous aluminum chips, roughness of the wall of the holes is inferior or equal to 1mm and 3mm respectively on the aluminum and the composite part. In order to predict the mechanical behavior of the interface CFRP/Al vs. the applied load (measured thrust force) and the position of the tool inside the material, a numerical model is proposed. The main aim of this model is to predict the force critical load reasonable of the separation of interface CFRP/Al as well as the accumulation of the carbon dust in the interface.

EXPERIMENTATION PROCEDURE

MATERIALS

The sandwich plates studied are composed of a carbon/epoxy and aluminium. The CFRP composite specimen used in the investigation was 4.2 mm thick. The laminate was made out of 16 unidirectional plies of 0.26 mm thickness each. The 16 unidirectional plies are made of carbon/epoxy prepreg and manufactured by Hexcel Composite Company with the reference Hexply T700-M21. The following was the staking sequence [90/45/0/-45]2s. These materials were compacted using a vacuum pump in a controlled atmosphere. A mold for the laminate was prepared and placed in a vacuum bagging and evacuated to 0.7 bar (Ref. Figure 1). Curing was then carried out at 180 °C for 120 min during which the pressure was maintained at 7 bar in an autoclave. Figure 1-b shows the laminate used for conducting the experiments. The aluminum alloy used in this study is extensively used in aviation industry have referenced as Al 2024. The percentage of alloying elements is as following: Al 93.5% Si 0.5%, Cu 3.8–4.9%, Mg 1.2–1.8%, 0.1% Cr.

Fig. 1. Manufacturing of the composite part by the autoclave process.

DRILLING DETAILS

Drilling is done on a CNC machine developed by

LATECIS Company (Toulouse, France) under the research project OPERA (French acronyms which means Automated drilling and riveting of aircraft structures). OPERA machine is shown in Figure 2-a. The acquisition of cutting forces is carried out using a four-component Kistler dynamometer as shown in Figure 2-b. The dynamometer is connected to a Kistler charge amplifier type 5019. The output of the amplifier is transformed into a cutting force through a computer that stores the force signals versus cutting time. The sandwich panel to be drilled is clamped on a dedicated support (see Figure 2-b). On the latter, a hole of 18 mm is machined to allow the drill bit and to prevent the bending of the sandwich plate. The drilling tests performed are based on full factorial experimental design using three feed rates (0.05 mm/rev, 0.1 mm/rev and 0.15 mm/rev) and two spindle speeds (2020 rpm and 2750 rpm).


Characteristics geometric of the reference tool (ref-tool) marketed by French industry for drilling of composite materials for Airbus industries is presented in Fig 3-a, and the same has been modified to double cone geometry. The modification of the tool geometry is achieved using a 5-axis grinding machine.

(a)

(b) Fig. 2. Experimental device for drilling tests with (a) CAD

model showing the OPERA System developed by Latecis company and(b) OPERA head and CRRP/Al parts.

Fig. 3. CAD models showing the active portion (point) of the drilling tools with (a) reference tool and (b) double cone tool.

The double cone tools measure 90° and 132° point angles

with cutting edge lengths, namely L1 and L2 (Ref. Figure 3) while L1 represents the size of the principal cutting edge number 1 which is characterized by a point angle of 132°, L2 represents the size of principal cutting edge number 2 which is characterized by a point angle of 90°. The double cone drills

tested are ground with different L1/L2 ratios, in which M1, M2 and M3 represent L1/L2 ratios of 0.33, 1 and 3.1 respectively. Drilling trials were carried out using a 6.35 mm of diameter made of tungsten carbide (K20). For the confidentiality reasons is not possible to give more details about the geometry of the tools. Each experimental condition was repeated 3 times in order to get consistent values. To remove the influence of tool wear, each experiment was performed with a new drill. The quality of the machined surface (wall of the hole) is quantified using the SEM observation and the surface roughness. The surface roughness (Ra) of the hole was measured by surface roughness tester (MitutoyoSJ 500) with a sampling length (cut-off) of 0.8 mm. In case of CFRP the length of measurement through the hole was 4 mm (0.8 x 5 = 4 mm).

RESULTS AND DISCUSSION

ANALYSIS OF CUTTING FORCE

From the average value of cutting forces for all tested tools at various feed speed, it can be observed that, drilling with the reference tool induces higher thrust force compared to the double cone tools (Ref. Figure4). When the feed rate has been increased from 0.05 mm/rev to 0.15 mm/rev, drilling with the reference tool is subjected to an increase of 60 % of the total thrust force in the composite (with the maximum value of 126 N) and 158 % in the aluminum (with the maximum value of 481 N). This result is observed while drilling at a spindle speed of 2020 rpm. Similar increase of thrust force values are observed when drilling using double cone tools as well. However for the same machining parameters, drilling with double cone tools produce 15 % to 30% lesser thrust force in the composite when compared to the thrust produced while machining using reference tool. This can be explained by the fact that adding a secondary point angle at 90° reduces the theoretical average chip thickness by 15%. It is well established by the literature, that (orthogonal cutting on unidirectional composite material) increasing the depth of cut (chip thickness) lead to increased cutting forces. During drilling at a spindle speed of 2750 rpm, similar results are obtained compared a spindle speed of 2020 rpm. The evolution of the thrust force generated during drilling of aluminum vs. feed rate show that, the level of the recorded forces induced by the reference drill and the double cone drill M2 and M3 are similar. However, the double cone drill M1 induces a less thrust force.


(a)

(b)

Fig. 4. Influence of tool geometry and feed rate on thrust force during drilling at a spindle speed of 2020 rpm with (a) thrust

force in composite and (b) thrust force in aluminum.

Fig. 5. Influence of the machining parameters and the tool

geometry on the torque during drilling aluminum.

This difference can be attributed to the fact that, adding a secondary point angle (90°) on the one hand reduces the

theoretical average chip thickness and on the other hand increase the surface contact between the drill bit and the aluminum part. This increasing of the surface contact leads to increased friction. This can be seen by the torque recorded vs feed rates for various tool geometries (Ref. Figure5).

QUALITY OF THE MACHINED HOLES

Figure 6 and 7 show the effect of drill geometry on surface finish at various feed rates, for a spindle speed of 2020 rpm in the CFRP and aluminum respectively. Experimental results can reveal that at a low feed rate (< 0.1 mm/rev), the quality of the machined surface is better for all the drill used. In this case, values of the measured roughness are smaller (< 3 µm). Further it can also be observed that, for all machining parameters used, values of the surface roughness obtained with the reference tool are inferior compared to those obtained with double cone tools. This difference can be linked to the interaction between the size of chip thickness and the drill point angle. In addition, double cone drills favor the apparition of continuous chips during drilling of aluminum. This small roughness is obtained on the wall of aluminum holes when drilling is conducted with double cone drills compared to the reference drill.

With the increasing of the length of second edge of the double cone drills, we reduce the chip thickness and we improve the machining quality (cf. Figure 7). Drilling with double cone drills offers stable surface roughness (inferior to 1µm) in the aluminum when machining is conducted with feed rates between 0.05 and 0.1 mm/rev and for any spindle speed (2020 rpm or 2750 rpm). However in the same machining parameters, machining with reference drill gives a roughness values superior to 1 µm. In addition for all tools used and drilling with 0.15 µm, the roughness measured in the aluminum is superior to 1 µm. Generally, drilling with double cone drill M2 gives better machining quality. Specially, drilling with a feed rate of 0.1 mm/rev and using the double cone tool (M2) offers a small roughness at both the spindle speeds on the composite and aluminum holes (roughness in the CFRP inferior to 3 µm and in the aluminum inferior to 1 µm). In addition drilling with a feed rate of 0.1 mm/rev favors the apparition of broken aluminum chips. The stable surface roughness value beyond 0.1 mm/rev could be due to the wiping effect of the primary cutting edge 2. It is important to mention that, the drilling with a small feed rate (0.05 mm/rev) favors the formation of continuous chips for all used tools. With these continuous chips, two problems have been observed when the OPERA system is used.

0

0.2

0.4

0.6

0.8

1

1.2

0.05 0.1 0.15

Torq

ue (N

.m)

Feed rate (mm/rev)

Al - reference drillAl - double cone - M1Al - double cone - M2Al - double cone - M3


Fig. 6. Evolution of the roughness in CFRP vs. feed rate for

reference and double cone drills (spindle speed of 2020 rpm).

Fig. 7. Evolution of the roughness in the aluminum vs. feed for reference and double cone drills (spindle speed of 2020 rpm).

Fig. 8. Different nature of damages observed while drilling at

feed rate of 0.05 mm/rev, spindle speed of 2020 rpm with (a) chip flow and (b) damages on the wall of the CFRP hole.

The first problem is the interaction of the Al chips with the composite and leads to the degradation of the first ply of the

CFRP part located at the hole entry as well as the damages apparition on the wall of the hole (Ref. Figure 8). The second problem is associated with aspiration system. Continuous chips reduce the efficiency of the aspiration system.

NUMERICAL MODELING

For this study, two numerical models are proposed in order to analyze the mechanical behavior at the interface during the displacement of the tool inside the materials. More precisely, our objectives are to analyze on the one hand, the stress distribution in contact zone (tool/parts) and on the other hand the local deflection of the CFRP and aluminum parts. In the first model, one ply under the tool is considered. In this case, contact tool/CFRP corresponds to the uniformly distributed load and located on the area of the chisel edge of the tool.

This area is considered as circular with 1.2 mm of diameter. For the second model, when we have 0 ply under the tool, drill starts the contact with aluminum (Ref. Figure 9). In this case, the deflection of the aluminum parts under the loading induced by the chisel edge of the tool is analyzed. In this situation, the contact tool/aluminum is modeled by a load uniformly distributed and applied on a circular surface characterized by a chisel edge radius of the tool (1.2 mm). Amplitudes of the applied loads for each model correspond of those measured experimentally. For this, two cases are considered: the first one corresponds to the case of lower feed rate (0.05 mm/rev) and the second case corresponds to the higher feed rate (0.1 mm/rev).

In the Figure 10-a and Figure 10-b, cartographies of displacements((OZ) direction) are illustrated for the applied loads of 80 N and 120 N respectively. These forces correspond to the two feed rates 0.05 mm/rev and 0.15 mm/rev respectively. The analysis of the interface CFRP/Al shows that, for a small feed rate and when one ply under the tool, the considered damage is located in the vicinity of the chisel edge and the deflection of the CFRP part is limited thanks the aluminum part. However when the tool starts machining the aluminum part (0 ply under the tool), negligible amount of opening is observed at the interface CFRP/Al. The maximum size of this opening is around 0.24μm when the applied load is equal to 200 N. However when the load reaches to 400 N (feed rate of 0.15 mm/rev), the size of this opening increases 0.34μm. From the work of Haddad et al. [17] when machining of Hexply T700/M21 carbon/epoxy composite material, the major particles generated by the end milling have a size of around 0.45μm. In this case when the applied load is around 450 N, the carbon dust cannot pass through the CFRP/Al interface. Therefore even in the most unfavorable situation (high feed rate and thrust force), it is not necessary to separate the parts (CFRP

3 mm

0.5 mm

0

1

2

3

4

5

6

0.05 0.1 0.15

Rou

ghne

ss R

a (µ

m)

Feed rate f (mm/rev)

CFRP - Reference drillCFRP - Double cone - M1CFRP - double cone - M2CFRP - double cone - M3

0

1

2

3

4

5

6

0.05 0.1 0.15

Rou

ghne

ss R

a (µ

m)

Feed rate f (mm/rev)

Al - Reference drillAl - Double cone - M1Al - Double cone - M2Al - Double cone - M3


and rivet

mm

aluminum) ints.

Fig. 9.mesh and bounmodels with (a

mod

u

Con

n order to clea

.Cross sectionndary conditiona) model with del with zero p

u =

Co

u = v = w = 0

Fz (N)

ntact zone

an the interfac

(a)

(b) n of the 3D mons for the propone ply underply under the

= v = w = 0

Fz (N)

ontact zone

ce before to p

odel showing posed numerir the tool and tool.

put the

the ical (b)

F

Fig. 10Cartogwhen one p

loading

graphies of theply under the tg of 80 N and

(a)

(b) e displacementool is conside(b) applied lo

(a)

nt in the directered with (a) aoading of 120 N

tion (OZ) applied N.


Fig. when

CON

desigforcemateFromprocconc

•

•

•

11. Cartograpn 0 ply under

of 200N

NCLUSION In this paper,gn of the geoes and the merial made ofm the experimess using d

clusions can be When drillinforces measuthose measureference driused, the thrcompared tothe double co When the mdrill M2, thethe CFRP andrill type Mwith referenc(M1 and M3drill M2, feediscontinuoualuminum anand inferior t

From the nuany value of the size of

(phies of the dithe tool is con

N and (b) appl

the impact ometry of the

machining quaf CFRP com

mental analysidiamond-coatee drawn:

ng is conducteured in the Cured when dill. However, wrust forces me those generaone drill M1 a

machining is ce measured avnd in the Alum

M2 (L2/L1 = 1ce twist drills 3). In additioned speed of 0.us aluminum nd CFRP are ito 3 μm in the

umerical modef the thrust for

the opening

(b) isplacement innsidered with lied loading o

of machining double cone lity when dri

mposite have is carried out ed twist dril

ed using doubCFRP are infdrilling is cowhen the doubeasured in theated when theand M2 are us

conducted witverage roughn

minum obtaine1) are inferior

and the othern, drilling wit.1 mm/rev an

chips and inferior to 1 μe composite.

eling, it has brce measured (g at the inter

n the direction(a) applied lo

of 400 N.

parameters andrill on the cilling a multibeen investigduring the dr

lls, the follo

le cone drill,ferior comparonducted withble cone drill aluminum ar reference too

sed.

th the doubleness “Ra” valued with doubler to those obtr double coneth the doubled 2020 rpm inroughness in

μm in the alum

been found tha(smaller or hirface of CFR

n (OZ) oading

nd the utting -stack gated. rilling owing

thrust red to h the M3 is

re less ol and

e cone ues in e cone tained drills

e cone nduce n the

minum

at; for gher), RP/Al

A

TL

R

[

[2

[3

[4

[5

[6

[7

[

[9[

[

[

[

[

[

[

[

remains smechanis

• Finally usystem wmm/rev operationchips in composit

ACKNOWLE

This work wLatecis Compa

REFERENCE

1] Zitoune R., coated tool Composites

2] Zitoune R.,Kmaterial and1255, 2010.

3] Abrate S. “book, Ed. by1997.

4] Sakuma Y. SRelation betVol. 27, No.

5] Tsao C.C. anwith varioustools & Mfr.

6] Zitoune, R., cutting condphase“, Com

7] Krishnaraj Vdrilling of GVol. 12, pp 1

8] Zitoune R., Double Conmesh/CFRP/

9] Chamberlain10] List G, Nou

Wear behavialloy. Wear 2

11] Nouari M, Lcoating on wMach Tools

12] Carrilero MMarcos M. Ain the turnin2002;42:215

13] Sanchez JMMicrostructuthe dry mach2005;164-66

14] Brinksmeierconsisting oaluminum al

15] Kim D. et atitanium allo

16] Ramulu M. stacks”, Com

17] Haddad M. Trimming of

small comparsm of material

using the dowhen drilling i

and spindle n is made bealuminum an

te and Ra < 1

EDGMENTS

was supported any.

ES

et al. “Influencon drilling per

Part B: EngineerKrishnaraj V., Cod aluminum stack

“Machining of cy Mallick P.K, M

S.”Study on drilltween tool mater228, pp. 1237-12

nd Hocheng H. ”s drill bits in dri., Vol. 44, pp. 108Collombet, et al

ditions representmp. Sc. and Tech. V., Vijayarangan GFRP”, Indian jou189-196. 2005.

El Mansori M.,ne Drill Implicati/Woven ply”. Wen B. Metals Handuari M, Géhin D, iour of cemented 2005;259:1177–8List G, Girot F, Gwear mechanismManuf 2005;45:1

MS, BienvenidoA SEM AND EDng processes of A5–20. M, Rubio E, Aural characterizathining of aerospa65:911–8. r E. et al. “Drof carbon fibre lloys’, Annals of al. “Drilling procoy stacks”, Comp

et al. “A studymposite structures

et al. “Surface f CFRP”, Applied

red to the dustl removal [17

uble cone dris conducted speed of

e possible by nd surface rouµm in the Alu

under the F

ce of machining rformance of CFring. 43(3)1480-1ollombet F. “Stuk”, Composite S

composites”, CoMarcel Deckker In

ling of reinforcedrial and wear be244. 1984. ”Taguchi analysiilling of composi85-1090. 2004. l. “Experiment-catative of the lon65 (3-4), 455-46S. et al., “An i

urnal of Enginee

, Krishnaraj V. “ons in Tool Wea

ear, 302(1-2), 156dbook.ASM; 1998

Gomez S, Manacarbide tools in

89. Géhin G. Effect o

ms in dry drilling1436–42. R, Sanchez JM

DS insight into thAA2024 Al-Cu all

Alvarez M, Setion of material aace aluminum allo

rilling of multireinforced plastithe CIRP, Vol. 51cess optimization

posite structures, py on the drilling s, Vol. 54 pp. 67-Quality and Du

d Mechanics and

t size generat].

rill M2 withat a feed spe2020 rpm ogetting disco

ughness Ra <uminum.

FUI project

parameters and FRP/Aluminium488. 2012.

udy of drilling ofStructures Vol. 92

omposites enginenc., New York, p

d plastics (GFRPehavior”, Bulletin

s of delaminationite material”, Int

alculation compang fibre compos

66. 2005. investigation on ring and Materia

“Tribo-functionalar During Drilling60-1567. 2013. 8. aud P, Le Petitcodry machining of

of machining parag of aluminium a

, Alvarez M, Ghe BUL and BUEloy. Int J Mach T

ebastien MA, Madhered over cutoys. J Mater Proc

-layer compositics (CFRP)”, Ti1, pp. 87-90. 2002n for graphite/bispp. 101- 114. 200of composite an

77. 2001. ust Analysis in HMaterials 232, 57

ted by the

h OPERA eed of 0.1 one shot, ontinuous

< 3 µm in

involving

new nano-sandwich”.

f composite 2, pp.1246–

eering hand pp. 777-807.

P & CFRP)-n of JSME,

n associated t. J.Machine

arison of the site drilling

High speed als Sciences,

l Design of g of Copper

orps Y, et al. f aluminium

ameters and alloys. Int J

Gonzalez A, E differences Tools Manuf

Marcos M. tting tool in cess Technol

e materials tanium and 2. smaleimide-04. nd titanium

High Speed 7-62. 2012.


Design of Double Cone Twist Drill Geometry to Improve the ...

Documents