Procedia CIRP 3 ( 2012 ) 91 – 96

2212-8271 © 2012 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibility of Professor D. Mourtzis and Professor G. Chryssolouris. doi: 10.1016/j.procir.2012.07.017

45th CIRP Conference on Manufacturing Systems 2012

Size Effects in Micro Drilling Ferritic-Pearlitic Carbon Steels M. Abouridouanea,*, F. Klockea, D. Lunga, O. Adamsa

aLaboratory for Machine Tools and Production Engineering (WZL), Chair of Manufacturing Technology, RWTH Aachen University, Steinbachstrasse 19, Aachen 52074, Germany

* Corresponding author. Tel.: +49-241-8028176; fax: +49-241-8022293. E-mail address: m.abouridouane@wzl.rwth-aachen.de

Abstract

The main objective of this research work is to investigate size effects by down scaling the twist drilling process into the micro range (drill diameter: d = 50 μm – 1 mm). Therefore, experimental micro drilling tests in ferritic-pearlitic carbon steel C45 are performed with different cutting parameters and compared with data obtained from conventional drilling. Based on the concept of a representative volume element (RVE) and constitutive material modelling, as well as using the Lagrangian formulation proposed in the implicit FE code Deform 3DTM, a 3D multiphase FE computational model was successfully applied to predict size effects in micro drilling. © 2012 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibility of Professor D. Mourtzis and Professor G. Chryssolouris.

Keywords: Size effects; Micro drilling; Multiphase FEM

1. Introduction

Compared with laser and lithographic techniques, mechanical micro machining is a high potential competitive process to manufacture 3D micro components or macro parts with micro features from various materials at high accuracy, efficiency and low costs. Therefore, mechanical micro machining is a central topic for applications such as electronics, telecommunications, sensor technology, optics, medical engineering, fine mechanics, biotechnology and environmental engineering [1-3].

Although conventional machining and micro machining show many similarities, the cutting parameters of conventional machining, like cutting speed, feed and depth of cut, cannot be offhand down scaled into the micro range due to size effects [4-5]. When the uncut chip thickness is on the same order as the material grain size, the workpiece material cannot any more be assumed as homogeneous and isotropic. Furthermore, the tool edge radius influences the cutting mechanism in micro machining significantly with regard to the effective rake angle and the ploughing effect [6-8].

These phenomena, known as size effects, have a high impact on the cutting forces, process stability, and resulting surface finish in micro cutting. It is therefore of great economic interest to analyse these size effects more closely and to introduce appropriate counter-measures.

The main goal of this research work is to analyse size effects by down scaling the twist drilling process into the micro range (diameter: d = 50 μm – 1 mm). Therefore, experimental micro drilling tests in ferrite-pearlite two-phase carbon steel C45 were performed with different cutting conditions (drill diameter, feed, cutting speed) and compared with data obtained from conventional drilling. A new piezoelectric force and torque measuring system was developed for micro drilling and assayed in terms of suitability and accuracy. This measuring system makes it possible for the first time to measure the torque generated during micro drilling with drill diameters down to 100 μm. Based on the concept of RVE and constitutive material modelling, as well as using the Lagrangian formulation proposed in the implicit FE code Deform 3DTM, a 3D multiphase FE computational model was successfully applied to simulate micro drilling ferritic-pearlitic carbon steel C45.

Available online at www.sciencedirect.com

© 2012 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibility of Professor D. Mourtzis and Professor G. Chryssolouris. Open access under CC BY-NC-ND license.

Open access under CC BY-NC-ND license.

92 M. Abouridouane et al. / Procedia CIRP 3 ( 2012 ) 91 – 96

2. Experimental size effects identification

In order to identify size effects by down scaling the drilling process into the micro range, drilling tests were carried out on carbon steel C45 using various drill diameters.

2.1. Materials, tools and process parameters

The main machining material of this research work is the two-phase ferritic-pearlitic carbon steel C45 (chemical composition: 0.45% C, 0.22% Si, 0.63% Mn, 0.016% P, and 0.026% S). For the multiphase FE modelling, the mechanical flow behaviour of normalized ferritic-pearlitic carbon steel C45 and its phases ferrite and pearlite was determined by means of compression tests. Applied ultra-fine grain carbide (ISO HW-K20: grain size 0.5-0.7μm) micro twist drills tools for this investigation were not coated and not web thinned. A high performance digital microscope and a scanning electron microscope were applied to measure and to check the micro drills geometry. Exemplarily, the geometry of 100 μm drill is shown in Fig. 1. The used micro drills (50 μm – 1 mm) as well as their comparative macro drills (d > 1 mm) had the same point angle, clearance angle and helix angle.

The drilling tests were conducted without the application of coolant on the ultra-precision CNC Machining centre KERN Evolution (position accuracy of ± 0.5 μm, maximal spindle speed of 160,000 rpm, non-contact measurement of the micro drills with a laser beam) at various cutting speeds in the range of 35 – 95 m/min. Due to different mechanical loading capacities of the drills and to ensure the comparability of the micro and conventional drilling results, the feed was selected depending on the drill diameter f = 0.012·d and the drilling depth amounted 2·d for all drilling tests. To demonstrate ploughing effects in the micro scale for a selected micro drill d = 200 μm and a cutting speed of vc = 35 m/min, micro drilling tests were performed with progressively reduced feed. In order to guarantee reproducible results, the drill tests were repeated 3 times for each test condition. Additionally, a new drill was used for each test parameters combination to reduce the influence of tool wear.

2.2. Micro measurement setup

For drilling with diameters larger than 3 mm, 4-component measuring dynamometers for force and torque measurement are available. The requirements on a measuring system for micro drilling with diameters below 1 mm are on the one hand the acquisition of low feed forces below 35 N and torque values below 15 Nmm and good dynamical properties on the other hand.

Fig. 1. Tool geometry of a 100 μm micro drill

In drilling operations the run-out deviation introduces a stimulation with the rotational frequency. Due to this fact, the sensor system used for the trials in this paper has to match high requirements in terms of dynamical behaviour especially for the torque measurement.

A Kistler 9256B dynamometer (working range of ± 250 N, response threshold of 2 mN) consists of four 3-component piezoelectric sensors was used to measure feed force. An additional high sensitive torque sensor (Kistler 9329: working range of ± 1 Nm, response threshold of 30 μNm) is mounted on top of the force measuring dynamometer. The workpiece is mounted directly on the torque sensor in order to keep the mass on the sensor low. Fig. 2 shows the resulting measurement setup. The used torque sensor has very good dynamical properties. A natural frequency larger than 53 kHz guaranties a proper measurement even at rotational speeds of 160,000 rpm. The low threshold of 0.03 Nmm and the high sensitivity of more than 2,000 pC/Nm allow a measurement with diameters down to 100 μm. Charge amplifiers are used to convert the sensor’s charge output signals to standard voltage signals that are digitized with a 16 bit National Instruments data acquisition card. The data acquisition on a Windows computer is realised with NI LabVIEW and the data analysis is carried out with NI DIAdem.

Fig. 2. Measurement setup for the micro drilling tests

300 μm

Tool material: ultra fine grain carbide Point angle: = 118° Front clearance angle: = 10° Helix angle: = 35°

Workpiece

Torque sensor Kistler 9329

Force measuring dynamometer Kistler 9256B

93 M. Abouridouane et al. / Procedia CIRP 3 ( 2012 ) 91 – 96

2.3. Experimental results und discussion

Size effect chisel edge length To compare the obtained micro drilling results of this

work with earlier macro drilling tests (d = 1 - 10 mm and the same twist geometry) in the normalized steel C45 [9], the related feed force and torque to the cross section of the uncut chip (d·f/2) are plotted versus drill diameters between 50 μm und 10 mm in Fig. 3.

A higher increase of the related feed force was observed by down scaling of the drill diameter in the micro range, see Fig. 3 (a). This size effect on the related feed force can be attributed to the dependency of the chisel edge length dc.e. on the diameter of the selected drills, as shown in Fig. 3 (b). The technical qualified increase of the part of the chisel edge of the micro drill is to ensure the rigidity and stability of the tool, leading to a significant influence on the micro drilling process reactions. The related torque behaves proportionately to the drill diameter in the micro scale like in the case of the conventional macro drilling, Fig. 3 (c).

Size effect ploughing

Fig. 4 shows the experimental results of the micro drilling tests (drill diameter d = 200 μm, rß = 0.86 μm) with reduced uncut chip thickness and a constant cutting speed of vc = 35 m/min. A higher increase of the related feed force kf and torque tr with decrease in the corresponding uncut-chip-thickness to cutting-edge-radius ratio h/rß could be observed in the conducted drilling tests. This size effect in the process reactions can be explained by the greater relative contribution of the parasitic force arising from ploughing caused by the cutting edge radius. Below the so-called minimum uncut chip thickness, a transition from a cutting to ploughing dominant process takes place. This result agrees with the experimental investigations on orthogonal cutting by Schimmel [10].

20

30

40

50

60

0 1 2 3 4 5 6 7 8 9 10drill diameter d / mm

ratio

(dc.

e./d

) / %

kf = 2*Ff,max / (d*f)

ddc.e.

tr = 2*Tz,max / (d*f)

micro drilling

0

4

8

12

16

20

0 1 2 3 4 5 6 7 8 9 100

4

8

12

16

20Workpiece material: C45 normalized # Cutting material: HW-20Cutting speed: vc = 35 m/min # Feed: f = 0.012*d # Cooling: dry

Drill diameter d, mm

Rel

ated

feed

forc

e k f , (

kN/m

m2 )

Rel

ated

torq

ue t r , (

kNm

m/m

m2 )

Fig. 3. The size effect of the drill micro geometry

6

10

14

18

22

0 0.2 0.4 0.6 0.8 1.0 1.20.8

1.2

1.6

2.0

2.4

kf

tr

Cutting > ploughingPloughing > cutting

The minimum uncut chip thickness

Workpiece material: C45 normalizedCutting material: HW-20Drill diameter: d = 200 μmCutting edge radius: rß 0.86 μmCutting speed: vc = 35 m/min

Ratio of the uncut chip thickness to edge radius h/rß

Rel

ated

feed

forc

e k f , (

kN/m

m2 )

Rel

ated

torq

ue t r , (

kNm

m/m

m2 )

Fig. 4. The size effect ploughing

Size effect microstructure Fig. 5 shows scanning electron microscope (SEM)

pictures of deformed chips during micro drilling two-phase ferritic-pearlitic carbon steel C45 with two different drill diameters d = 500 μm and d = 1 mm. On the cut chips, micro holes are clearly identifiable, see Fig. 5. These micro holes could be attributed to the size of uncut chip thickness h (h = 2.6 μm for d = 500 μm and h = 5.2 μm for d = 1 mm) which was on the same order as the harder pearlite grains by micro drilling. This microstructure size effect has a high impact on the surface finish of the workpiece.

Size effect built-up edge

The influence of the cutting speed on the feed force and torque during micro drilling tests is shown in Fig. 6. An increase of cutting speed reduces the feed force and torque until a minimum value is reached. This phenomenon is related to thermal softening of the material caused by the temperature increase with the cutting speed. At higher cutting speeds, feed forces and torque show an increase. This augmentation results from the formation of built-up edges at higher cutting speed contrary to conventional machining, see Fig. 6 (b).

Fig. 5. The size effect of the microstructure

(a)

(b)

(c)

d = 500 μm

vc = 35 m/min f = 6 μm C45 normalized

vc = 35 m/min f = 12 μm C45 normalized

Micro holes d = 1 mm

94 M. Abouridouane et al. / Procedia CIRP 3 ( 2012 ) 91 – 96

1.5

2.0

2.5

3.0

3.5

4.0

30 40 50 60 70 80 90 1000

0.1

0.2

0.3

0.4

0.5

T

Ff

Workpiece material: C45 normalizedCutting material: HW-20Feed: f = 2.4 μmDrill diameter: d = 200 μmCooling: dry

Cutting speed vc, m/mim

Feed

forc

e F f, N

Torq

ue T

, Nm

m

Fig. 6. The influence of the cutting speed on the feed force and torque by micro drilling steel C45

This tendency can also be explained by the strain rate hardening of the material in the primary shear zone. Similar behaviour of the feed force with the cutting speed was observed in high speed cutting of the same workpiece material [11].

3. FE modelling of size effects

Due to the very complicated cutting process at the micro scale, most developed FE models for micro cutting heterogeneous materials are still limited at present to two dimensional orthogonal cut and only give a qualitative prediction of simple plane strain cutting processes [12-15]. To increase the understanding of size effects occurring in micro cutting, a 3D thermo-mechanically coupled two-phase FE model of micro drilling carbon steel C45 was applied in this work.

3.1. 3D FE computation model for micro drilling

The 3D simulation of micro drilling tests was performed by using the commercial implicit FE code Deform 3DTM. The representation of tool and workpiece in the 3D FE simulation requires the input of geometrical, thermal and mechanical data. Since the geometry of the tool influences strongly the micro drilling process, the design of the selected micro drills has to be accurate. Only then, the FE model is able to predict real process behaviour during cutting. The micro drills geometries were designed by the CAD program SolidWorks, whereas the detailed geometric parameters of the tool were made available by the tool manufacture. For more efficient computing, the volume of the

modelled workpiece was selected as small as possible. The CAD models have been transferred in Deform 3DTM and meshed using 3D tetrahedron elements. After the meshing of workpiece, two-phase microstructure was generated using the developed multiphase material model in [16]. Exemplarily, the generated 3D two-phase FE models for micro drilling with d = 100 μm and d = 1 mm are shown in Fig. 7.

In the FE computation, a gradient continuous auto-remeshing is used for generating the mesh of the workpiece. To minimize the interpolation error and to improve the convergence of the solution, the mesh density in areas with high gradients of plastic strain has to be increased. Because of the extensive computation time by continuous remeshing, the FE computation has been started by the full contact of the cutting edge of the drill and the adjusted workpiece (see Fig. 7). The micro drills are modeled as rigid. They have a constant mesh, which is finer at the cutting part of the tool to calculate the temperature distribution within the tool.

The cutting process in micro drilling is modeled as forming operation and the chip formation is simulated by continuous remeshing, so that no material fracture criteria are needed. The movement of the micro drill (rigid body with mesh) is specified by its translation and angular velocities in z-direction, while the workpiece (deformable) was constrained on the bottom and the round surfaces in the x, y, and z directions. Friction at the objects interfaces, tool-workpiece and chip-workpiece is governed by the Coulomb friction model (μ = 0.2). For the thermal boundary conditions, conduction and convection of the generated heat were applied. The gap conductance and the thermal convection coefficient between two contacting surfaces were assumed to be 107 W/m2K and 20 W/m2K, respectively. The workpiece and tool temperatures were initially set to room temperature (20°C). In order to adapt the energy balance in the modeled workpiece to the experiment, the nodes temperature at the bottom and the round surfaces was kept constant at a value of 20°C. The implicit 3D simulation of the micro drilling process requires an enormous amount of CPU time. Therefore, parallel computing (2 x 3.2 GHz) was necessary to solve such a complicated problem.

Drilling FE model (d = 100 μm) Drilling FE model (d = 1000 μm)

Workpiece

Micro drill

Fig. 7. 3D two-phase FE models for micro drilling

(a)

(b)

(a) (b) 20 μm 20 μm

95 M. Abouridouane et al. / Procedia CIRP 3 ( 2012 ) 91 – 96

3.2. Computation results and discussion

The developed 3D two-phase FE model for micro drilling was validated using the experimental results of the present work. Under the same machining conditions of the conducted drilling tests (vc = 35 m/min, f = 0.012·d), micro drilling was simulated for two different drill diameters d = 100 μm and d = 1 mm. In order to investigate the influence of the microstructure in micro drilling, two types of simulation were performed, one with isotropic material behaviour and the second with the mixture material model discussed above. Furthermore, drill CAD models with sharp as well as rounded cutting edge were employed in the FE simulation to demonstrate the influence of the cutting edge rounding. Three of the four size effects identified in micro drilling tests were predicted successfully with the developed FE model. To simulate the size effects due to built-up edge, the two-phase FE model needs to be enhanced by tool wear and workpiece material adhesion in future investigation.

Prediction of the size effect of drill micro geometry

Fig. 8 shows a comparison between experiment and simulation results for drilling with d = 100 μm and d = 1 mm. The predicted values of the feed force and torque with the mixture FE model are in good agreement with the measured results (average deviation less than 10%). On the contrary, the deviation by the isotropic model, which takes no influence of the microstructure, is up to 30%. In this respect, the developed two-phase FE model for micro drilling can realistically predict the size effect of chisel edge length on the related feed force.

Prediction of the size effect ploughing

For the prediction of ploughing, when micro drilling with drill diameter d = 100 μm, FE computations with the mixture FE model were performed using sharp and rounded cutting edges (rß = 0 μm and rß = 0.2 μm). These simulations were made in two successive steps: First, plunging the drill into the workpiece for one half of the feed without drill rotation and second, starting the drilling process. Fig. 9 (a) and (b) show the computed chip form after one half revolution of the drill when drilling with sharp and rounded cutting edge, respectively. Contrary to the simulation with sharp cutting edge, the deformed chip in front of the rounded main cutting edge became thinner and thinner until it disappeared. During simulation, the used FE program Deform 3DTM deletes elements with very small thickness. This computed effect demonstrates ploughing caused by the small ratio between the uncut chip thickness and cutting edge radius in micro cutting. In the same manner, ploughing effect could also be predicted with the isotropic FE model. SEM photographs of

deformed chips in micro drilling tests on C45 with drill diameter d = 100 μm indicate a ploughing dominant cutting, as illustrated in Fig. 9 (c) and (d).

Prediction of the size effect of the microstructure

The validation of the mixture FE model for micro drilling is exemplarily represented in Fig. 10 for a drill diameter of d = 1 mm. The predicted values of the feed force and torque with the developed mixture model are in good agreement with the measured results (average deviation about 7% and 3% respectively). On the contrary, the deviation by the isotropic model, which takes no influence of the microstructure, is about 20%, see Fig. 10 (b). Additionally, a realistic prediction of pearlite holes on deformed chip was also obtained with the applied 3D multiphase FE model, as shown in Fig. 10 (d). Hence, the size effect of the microstructure can be predicted by means of the multiphase FE model.

0

4

8

12

16

20

0 1 2 30

2

4

6

8

10

12

Simulation(isotropic FE macro model [9])

tr

kf

Simulation (isotropic FE model, d = 100 μm)

Experiment

Simulation (mixture FE model, d = 100 μm)

Workpiece material: C45 normalized # Cutting material: HW-20Cutting speed: vc = 35 m/min # Feed: f = 0.012*d # Cooling: dry

Drill diameter d, mm

Rel

ated

feed

forc

e k f , (

kN/m

m2 )

Rel

ated

torq

ue t r , (

kNm

m/m

m2 )

Fig. 8. FE model validation - feed force and torque (rß = 0.2 μm for d = 100 μm, rß = 1 μm for d = 1 mm)

Fig. 9. FE model validation - ploughing in micro drilling C45 with d = 100 μm (vc = 35 m/min, f = 1.2 μm)

Experiment

Simulation

rß = 0.2 μm rß = 0 μm

(a) (b)

(c)

(d)

Chip Chip

96 M. Abouridouane et al. / Procedia CIRP 3 ( 2012 ) 91 – 96

50

40

30

20

10

N22%

3% 19%7%

Mix

ture

mod

el

vc = 35 m/min, f = 12 μm/rev

TestNmm

Isot

ropi

cm

odel

Isot

ropi

cm

odel

Mix

ture

mod

el

Test

00

4

8

12

16

20Feed forceTorque

Chip

Drill

Workpiece

Microstructure

Chip form

Holes

Fig. 10. FE model validation - chip form in micro drilling C45 with d = 1 mm (rß = 1 μm)

4. Conclusions

In this paper some of the experimental and computation results of recent research into micro cutting carbon steel C45 have been presented. Progress has been made mainly by identifying and 3D multiphase modelling size effects in micro drilling. A new piezoelectric measuring system has been established for the measurement of force and torque generated during micro drilling with drill diameters down to 100 μm. Some of the significant findings from this investigation on micro drilling can be summarized as follows:

The size effect caused by the technical qualified increase of the chisel edge length is identified by down scaling the drilling process into the micro range.

Relation of the size effect ploughing to the ratio between uncut chip thickness and cutting edge radius has been established.

Ploughing leads to a high increase of the related feed force and torque in micro drilling.

Detected pearlite micro holes on the deformed chip point out the microstructure size effect on the surface finish of the workpiece.

Contrary to conventional machining, formation of built-up edges occurs in micro drilling at higher cutting speed and causes an increase of feed force and torque and hence deteriorates the machined surface quality.

A 3D multiphase FE computational model is used to simulate micro drilling.

Three of the four size effects identified in micro drilling test were predicted successfully with the applied multiphase FE model.

Acknowledgements

The results presented in this paper were determined within the framework of the Projects “3D FE Micro drilling simulation of multiphase materials” assisted by the research work of the DFG (Deutsche Forschungs-gemeinschaft). The authors wish to record their gratitude to the DFG for its financial support.

References

[1] Alting, L., Kimura, F., Hansen, H.N., Bissacco, G., 2003, Micro Engineering, Annals of the CIRP, 52/2: 635-657.

[2] De Chiffre, L., Kunzmann, H., Peggs, G.N., Lucca, D. A., 2003, Surfaces in Precision Engineering, Microengineering and Nanotechnology, Annals of the CIRP, 52/2:561-577.

[3] Weule, H. et al., 2004, International State of the Art of Micro Production Technology, Production Engineering, XI/1:29-36.

[4] Kopalinsky, E. M., Oxley, P.L.B., 1984, Size Effects in Metal Removal Processes, Institute of Physics Conference Series, n70:389-396.

[5] Klocke, F., Lung, D., Abouridouane, M., Sangermann, H., Skalierungseffekte bei der Mikrozerspanung, VDI-Z Integrierte Produktion, 150:42-43.

[6] Dornfeld, D., Min, S., Takeuchi, Y., 2006, Recent Advances in Mechanical Micro machining, Annals of the CIRP, 52/2:745-768.

[7] Klocke, F., Gerschwiler, K., Abouridouane, M., 2007, Size Effects of the Tool Edge Radius on Specific Cutting Energy and Chip Formation in Drilling, Proceedings of the 2nd International Conference on New Forming Technology, Vollertsen, F. and Yuan, S. (Eds.), 499-509, Bremen.

[8] Kotschenreuther, J., 2008, Empirische Erweiterung von Modellen der Makrozerspanung auf den Bereich der Mikrobearbeitung, Dissertation, University of Karlsruhe.

[9] Klocke, F., Lung, D., Gerschwiler, K., Risse, K., Abouridouane, M., 2006, Modelling and Scaling Effects in Drilling, the 9th CIRP International Workshop on Modelling of Machining Operations, 263-269, Bled Slovinien.

[10] Schimmel, R.J., Endres, W.J., Stevenson, R., 2002, Application of an internally consistent material model to determine the effect of tool edge geometry in orthogonal machining, Trans. ASME, 124:536-543.

[11] Hoppe, S., 2004, Experimental and Numerical Analysis of Chip Formation in Metal Cutting, Dissertation, RWTH Aachen.

[12] Chuzhoy, L., DeVor, R.E., 2003, Machining Simulation of Ductile Iron and Its Constituents, Part 2: Numerical Simulation and Experimental Validation of Machining, Journal of Manufacturing Science and Engineering, Transactions of the ASME, 125/2:192-201.

[13] Park, S., Kapoor, S.G., DeVor, R.E., 2004, Mechanistic Cutting Process Calibration via Microstructure Level Finite Element Simulation Model, ASME J. Manuf. Sci. Eng., 126/4:706-709.

[14] Simoneau, A., Ng, E., Elbestawi, M.A., 2007, Grain Size and Orientation Effects When Microcutting AISI 1045 Steel, Annals of the CIRP, 56/1:57-60.

[15] Schulze, V., Zanger, F., Michna, J., Ambrosy, F., Pabst, R., 2011, Investigation of the machining behaviour of metal matrix composites (MMC) using chip formation simulation, 13th CIRP Conference on Modelling of Machining Operations, Sintra Portugal, 20-29.

[16] Mustapha Abouridouane, Fritz Klocke, Dieter Lung, Oliver Adams, 2012, A new 3D multiphase FE model for micro cutting ferritic–pearlitic carbon steels, CIRP Annals - Manufacturing Technology, Vol. 61/1: 71-74.

(a)

(b)

(c) (d)