Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 58 (2017) 305 – 310

16th CIRP Conference on Modelling of Machining Operations

The effect of material parameters on chip formation in orthogonal cutting

simulation of Ti-5553 Alloy
Melih Ozkutuk,Yusuf Kaynak*
Marmara University, Department of Mechanical Engineering, Goztepe Campus, Kadikoy-ISTANBUL-TURKEY
Marmara University, Department of Mechanical Engineering, Goztepe Campus, Kadikoy-ISTANBUL-TURKEY
* Corresponding author. Tel.: +90 216 336 5770; Fax: +90 216 337 89 87. E-mail address: yusuf.kaynak@marmara.edu.tr

Abstract

It is a well-known fact that Titanium alloys are commonly used materials in various industries, in particular aerospace and biomedical
applications. Among titanium alloys, Ti-5553 is a new generation high temperature near beta alloy. As it has superior properties such as high
tensile strength and fatigue life, it has a potential to be used in aerospace industries for structural components to replace Ti-6Al-4V. This study
provides investigation on the effect of material parameters including material model parameters and frictional conditions on simulating chip
formation in orthogonal cutting process of this alloy. Experimental data are used to compare forces, contact length, and to predict chip form in
the cutting process of Ti-5553 alloy. This study illustrates that with calibrated frictional condition and critical damage value, it is possible to
predict forces, contact length and chip morphology that can show good agreement with experimental data.
©2017
© 2017TheTheAuthors.
Authors. Published
Published by Elsevier
by Elsevier B.V.is an open access article under the CC BY-NC-ND license
B.V. This
Peer-review under responsibility of the scientific committee of The 16th CIRP Conference on Modelling of Machining Operations, in the
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
person of the
Peer-review Conference
under Chairs
responsibility of theProf. J.C.
scientifi Outeiro and
c committee Prof.
of The 16thG.CIRP
Poulachon.
Conference on Modelling of Machining Operations

Keywords: Finite Element Simulation, Ti-5553 Alloy, Orthogonal Cutting

1. Introduction are available at the open literature relevant to the modeling

process of this Ti-5553 alloy considering machining process.
Titanium alloys primarily stand out due to two properties: Germain et al. [9]determined the Johnson-Cook constants for
high specific strength and excellent corrosion resistance. This this material using dynamic shear tests. Another study
also explains their preferential use in the aerospace sector, the contributes the modeling effort of this particular alloy is
chemical industry and medical engineering [1]. Ti-5553 (Ti– presented by Sun et al. [10]. Authors presented turning
5Al–5Mo–5V–3Cr) is a recently developed near beta Ti alloy processes of Ti-5553 alloys under dry and cryogenic
that is gathering increasing interest in aircraft structural machining conditions. Besides, they presented modeling of
applications, especially in the landing gear components [2] cutting process of this alloy by utilizing Johnson-Cook
because of their higher yield strength, excellent fatigue crack material model to predict cutting forces [10]. Although these
growth resistance and good hardenability [3]. studies are good starting point to model cutting of this alloy
However, this new refractory material presents a poor under various conditions, it requires further attempt to be able
machinability compared to steel or the more known titanium to predict deformation response of chip formation of this alloy
alloy: Ti-6Al-4V (Ti-64) due to various reasons addressed by under wide range of cutting conditions and to better predict
researchers at limited studies [4, 5] including low thermal response resulting from various cutting parameters.
conductivity and for this reason it is categorized as heat As Ti-5553 alloys have been used to make landing gear
resistance material [6]. Some initial studies showed an components in aerospace industries and machining processes
important point to improve machining performance that might needed to produce these components, the development of
be useful for industry utilizing this work material [4, 7, 8]. adequate tooling and machining strategies for such alloys
However, to the best knowledge of authors, only two studies requires a good understanding of the thermal, mechanical and

2212-8271 © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of The 16th CIRP Conference on Modelling of Machining Operations
doi:10.1016/j.procir.2017.03.324
306 Melih Ozkutuk and Yusuf Kaynak / Procedia CIRP 58 (2017) 305 – 310

microstructural aspects involved in the chip formation process cutting tool [11,12]. The workpiece is modeled as plastic and
[5]. meshed 6122 quadrilateral elements. Mesh and boundary
Thus this study presents modeling of orthogonal cutting conditions are given in Fig. 2. Considering the cutting process,
process of Ti-5553 alloy and exhibit simulation of cutting workpiece top and right side were allowed to change the heat
process at two different cutting speeds. Besides, it shows the with environment. The cutting tool’s rake angle side and
effort made to understand the effect of frictional condition and clearance side were allowed to change the heat with the
material parameters on some outputs such as forces, contact environment. The convection coefficient was 20 W/m2K
length, and chip form in orthogonal cutting simulation of Ti- which is the default value for free air convection in DEFORM
5553 alloy. Predicted forces and chip forms show good 2D. Heat transfer coefficient, hint at tool-chip interface was
agreement with experimental data but predicted contact length taken as 1000 kW/m2K [13]. All other sides of work material
shows some variation with data obtained from experimental and cutting tool as shown in Fig. 2 were assumed to have
study. room temperature.
Table 1. Properties of the tool material [12].
2. Experimental parameters and conditions
Properties Carbide Tool
The work material used in this research was a Ti-5553 (Ti– Expansion (µm /m°C) 4.7
5Al– 5Mo–5V–3Cr) alloy in the form of cylindrical disk of 63 Density (g/cm3 ) 15.0
mm diameter and 3 mm thickness. Cutting speeds of 60 and
Poisson’s ratio 0.20
120 m/min and the uncut chip thickness of 0.15 mm were
Specific heat (J/kg/°C) 203
used in the experimental work. Ti-5553 orthogonal machining
tests were carried out on a Doosan Puma GT2100 lathe Conductivity (W/m°C) 46.0
machine with 18 kW power and maximum spindle speed of Young’s modulus (GPa) 800
4500 rpm. It was used an uncoated carbide tool which ISO
number is TCMW16T308 H13A, under dry cutting conditions
to calibrate finite element model. The STNCN 2525 M16 -2
tool holder with rake angle, α= -2 degree, was used and
mounted on dynamometer to measure cutting forces. Cutting
forces were recorded by KISTLER 2129AA dynamometer.
Experimental setup for orthogonal cutting process of Ti-5553
alloys is illustrated in Fig 1.
Tool holder Dynamometer

Fig.2. Mesh and boundary conditions for (a) work material, and (b) cutting
tool.
Cutting tool
The thermal conductivity of the Ti-5553 alloy material is
7.1 W/m°C [14] and the Young’s modulus is defined as
temperature-dependent variable and given to the Table 2.
Table 2. Young’s modulus of Ti-5553 alloy [14].
Ti-5553
disc T (°C) 21.11 204.44 426.66 648.88
Young’s modulus (GPa) 117.2 106.8 95.1 82.7

3.1. Material modeling

Fig. 1. Experimental setup.
To model plastic behavior of Ti-5553 alloy, Johnson-Cook
3. FE Model for Orthogonal Machining constitutive equation is employed, which can be represented
by the following equation [15]:
DEFORM-2D commercial software (V11.0.2) is used to
ºª § · º
m
ª
simulate orthogonal cutting process. The created tool and its V ª A  B  H n º «1  C ln H » «1  ¨ T  To ¸ » (1)
¬ ¼ H 0 ¼ « © Tm  To ¹ »
dimensional properties used in the simulation were determined ¬ ¬ ¼
considering the tool used in the experimental work. Rake
angle of -2°, the cutting tool edge radius of 25 µm and a
clearance angle of 7 ° are employed in the tool geometry. The In the Johnson-Cook constitution equation V refers to
cutting tool is considered to be rigid material and meshed flow stress. In this equation ߝҧ, ߝҧሶ and ߝҧ଴ሶ are the strain, strain
containing 2969 elements. Also the cutting tool is modeled as rate and reference strain rate(ߝҧ଴ሶ ൌ ͳ‫ି ݏ‬ଵ )[16, 17] respectively.
uncoated and without chip breaker. The values shown in Table Workpiece temperature is expressed as ܶ , the workpiece's
1 are used for thermomechanical properties of the carbide melting temperature ܶ௠ (1630 °C)[10], and room temperature
ܶ଴ (20 °C). n is the hardening coefficient, m is the thermal
 Melih Ozkutuk and Yusuf Kaynak / Procedia CIRP 58 (2017) 305 – 310 307

softening coefficient. Coefficient A is the yield strength, B is

the hardening modulus, and C is the strain rate sensitivity Simulaiton 60 m/min Experiment 60 m/min
coefficient.[15] In the Ti-5553 alloy cutting simulation, J-C Simulation 120 m/min Experiment 120 m/min
1200
was used in the model by taking the constants from the work

Cutting Force, Fc (N)

1000
reported by Germain et al. [9]. These constants are shown in
the Table 3. 800

600
Table 3. Johnson-Cook Equation constants of Ti-5553 [9]. 400

A(MPa) B(MPa) n m C 200

1175 728 0.26 1.0 0.047 0

0.6 0.7 0.8
Shear Friction Factor, mf
3.2. Fracture criterion and friction tool/chip interface
Fig.3. Experimentally measured and predicted cutting forces at two
The friction between the workpiece-tool contact is different cutting speeds.
modelled using shear friction factor,݉௙ [18];
Table 4 shows the quantitative comparison between predicted
߬ ൌ ݉௙ Ǥ ݇ (2) forces including cutting forces and thrust forces at 60 m/min
and 120 m/min and experimentally measured force data.
Here, τ is the maximum shear stress on the rake face, k is
shear flow stress of the work material at the tool chip
Simulation 60 m/min Experiment 60 m/min
interface. Simulation 120 m/min Experiment 120 m/min
The fracture criterion for chip formation was the model 700

proposed by Cockcroft and Latham[19]; 600

Trust Force, Ft (N)

500
ఌത
‫׬‬଴ ߪଵ ݀ߝҧ ൌ‫ܦ‬ (3) 400
300

ε is the effective strain; ߪଵ the maximum principal stress; D is 200

critical damage value that is a material constant. 100

Cockroft and Latham’s criterion equation indicates that 0

0.6 0.7 0.8
chip segmentation has started when the damage expression Shear Friction Factor, mf
reaches critical damage value.
In order to calibrate, critical damage value and frictional Fig.4. Experimentally measured and predicted average thrust forces at two
conditions, various values of these were employed and cutting different cutting speeds.
process of Ti-5553 is simulated to see their effect on
measured outputs, particularly forces and chip form. The largest difference in between experimental data and
predicted results considering thrust force is approximately
4. Result and discussions 18.6 percent when mf =0.8 is used at 60 m/min cutting speed.
Table 4.Comparison between experimentally and predicted forces.
4.1. Cutting force Fc (N) Ft (N)
Experimental Results (60 m/min) 971 490
In cutting process, one of the notable outputs is force
components and predicting force components can be Simulation mf=0.6 939 (-3.3%) 482 (-1.6%)
considerable criteria to assess the simulating capability of Simulation mf=0.7 991 (2.1%) 509 (3.8%)
model utilized for cutting simulation. Figure 3 shows variation Simulation mf=0.8 1120 (15.3%) 581 (18.6 %)
of the predicted average cutting force at two different cutting Experimental Results (120 m/min) 897 475
speed values as shear friction factor changes in between 0.6 to Simulation mf=0.6 862 (-3.9%) 463 (-2.5%)
0.8. While shear friction factor varies, critical damage value
Simulation mf=0.7 914 (1.9%) 484 (1.9%)
(D) was kept constant as 150. Predicted forces at 120 m/min
are smaller than forces predicted at 60 m/min. The variation at Simulation mf=0.8 932 (3.9%) 503 (5.9%)
60 m/min cutting speed is approximately 181 N in between Besides, the largest difference in between experimental data
0.6 to 0.8 shear friction factor, while at 120 m/min it is and predicted results considering main cutting force is
approximately 70 N. Variation is much higher at low cutting approximately 15.3 percent when mf =0.8 is used at 60 m/min
speed. cutting speed. These experimental values are the average
Predicted thrust forces at two different cutting speed shows values of forces measured and it has error bars ( as shown in
similar trend to cutting forces, as shown in Figure 4. At lower Figures 3 and 4) and considering this fact, it is possible to
cutting speed, the predicted thrust forces are larger than forces reach conclusion that model is capable of predicting
at higher cutting speed. The variation at 60 m/min cutting experimental conditions in terms of force components.
speed is larger than the variation at 120 m/min when shear
friction factor is varied in between 0.6 to 0.8.
308 Melih Ozkutuk and Yusuf Kaynak / Procedia CIRP 58 (2017) 305 – 310

4.2. Contact Length Experimentally measured contact length generally larger than
predicted contact length this is mainly because of the wear
Contact length is an important parameter in metal cutting occurs during experiment and enlarges contact length but in
process as it is related to frictional conditions in between simulation, wearing of cutting tool is not taken into account
cutting tool and chip. The images of cutting tool used in
experiment and measured contact length can be seen in Figure 4.3. Chip Morphology
5(a). Predicted contact length is shown in Figure 5(b).
Prediction of chip morphology, namely chip thickness (tc) and
chip valley (b) resulting from three different critical damage
value (DCR) at two different cutting speeds are presented in
this section. The predicted values of chip thickness are shown
in Figure 7. It is obvious that critical damage value affects the
lc lc
predicted chip thickness in simulation the orthogonal cutting
(a) (b) process of Ti-5553 alloy. At lower cutting speed, predicted
chip thickness is larger as compared to higher cutting speed.
Increased critical damage value leads to producing ticker
Fig.5. Demonstration of contact length a) experimental measurement b)
chips.
prediction.

Figure 6 shows the effects of critical damage value (DCR) on

contact length. At lower critical damage values, predicted

Predicted Chip Thickness, t c (μm)

60 m/min 120 m/min
250
contact length at higher cutting speed is much larger than the
contact length at lower cutting speed. But at the largest value 200
of critical damage, different trend is obtained.
150

100
Predicted Contact Length, lc (μm)

300 60 m/min 120 m/min

50
250
0
200 100 150 200

150 Critical Damage Value, Dcritical

100
Fig.7. Predicted average chip thickness, mf =0.7.
50
The chip valley results obtained from simulation at two
0 different cutting speeds shows similar results with the
100 150 200
Critical Damage Value, Dcritical
predicted chip thickness. As shown in Figure 8, increased
cutting speeds results in reduced chip valley at different DCR.
Fig.6. Predicted contact length, mf = 0.7. Besides, increased critical damage value leads to increased
chip valley.
Table 5 shows the comparison between predicted contact
lengths vs. experimentally measured contact length at two
60 m/min 120 m/min
different cutting speeds. At lower cutting speed, the difference
Predicted Chip Valley, b (μm)

120
in between experimentally measured contact length and 100
predicted contact length is smaller than 49 percent and the
80
difference gets reduced as DCR decreases. At higher cutting
speed, the difference is much larger at in particular larger DCR. 60

Table 5. Comparison between experimentally and predicted contact length. 40

Contact Length, lc (µm) 20

Experimental Results (60 m/min) 303.4 0

100 150 200
Simulation DCR =100 194 (-30.1%) Critical Damage Value, Dcritical
Simulation DCR =150 156 (-48.6%)
Fig.8. Predicted average chip valley, mf = 0.7.
Simulation DCR =200 163 (-46.2%)
Experimental Results (120 m/min) 329.1 Experimentally generated chips at two different cutting
Simulation DCR =100 218 (-33.7%) speeds and predicted chips at these cutting speeds with
various DCR values are shown in Figure 9. It is obvious that
Simulation DCR =150 252 (-23.4%)
segmented chips are obtained from experimental study.
Simulation DCR =200 146 (-55.6%)
Simulation is capable of predicting segmented chips as well,
as shown in Figure 9. The quantitative comparison of
 Melih Ozkutuk and Yusuf Kaynak / Procedia CIRP 58 (2017) 305 – 310 309

experimentally measured chip thickness and valley and between experimental result and simulation is lower than five
numerically predicted chip thickness and valleys resulting percent. This show that chip thickness and valley is
from two different cutting speeds are illustrated in Table 6. significantly affected from critical damage value in simulation
60 m/min 120 m/min of chip formation process of Ti-5553. Besides, cutting speed
also plays an important role on generated chip morphology
during simulation.
Experimental data

5. Conclusions

This study presents the preliminary effort to understand

material parameters and frictional conditions on prediction
capability of simulation of orthogonal cutting of Ti-5553
alloy. Johnson-Cook material model is utilized to simulate
DCR=100

cutting process of Ti-5553 alloy. The predicted outputs

resulting from simulation were cutting force components,
contact length, and chip morphology. To improve capability of
the model to predict mentioned outputs, critical damage value
(DCR) and frictional constant in between cutting tool and chip,
mf varied. Varying both parameters results in changes
predicted output to some extent. But within selected range,
simulation is well capable of capturing experimental data. In
DCR=150

particular, within selected value of variables and using

Johnson-Cook, it is possible to have very good agreement with
experimental data considering forces and chip form. But it
should be also noted that outputs are affected from variation of
DCR and mf, but in the meantime, cutting speed also play major
role to determine sensitivity of DCR and mf on predicted output.
DCR=200

Considering overall outputs, this study shows that selecting

DCR = 150 and mf =0.7 to simulate cutting process of Ti-5553
alloy can provides well capturing experimental force data.
However, it is still possible to modify critical damage value to
obtain much closer prediction capability for the experimental
Fig.9. Comparison between experimentally obtained and predicted chip
contact length and chip morphology data.
morphology, t0 = 0.15 mm.

Simulation is capable of well predicting chip thickness and Acknowledgements

chip valley, as shown in Table 6. The largest difference in
between experimentally measured chip thickness and Financial support from TUBITAK (The Scientific and
numerically predicted one is 19.6 percent when DCR = 100. Technological Research Council of Turkey) under project
The largest difference in between experimentally measured number 214M068 is gratefully acknowledged.
chip valley and numerically predicted one is 30.2 percent
when DCR = 100. References
Table 6. Comparison between experimentally and predicted chip geometry.
[1] Leyens C, Peters M. Titanium and titanium alloys: Wiley Online Library;
Chip thickness Chip valley 2003.
(tc), µm (b), µm [2] Strunk Kar SK, Ghosh A, Fulzele N, Bhattacharjee A. Quantitative
microstructural characterization of a near beta Ti alloy, Ti-5553 under
Experimental Results different processing conditions. Materials Characterization. 2013;81:37-
215.6 114.1
(60 m/min) 48.
Simulation, DCR =100 173.3 (-19.6%) 92.7 (-18.7%) [3] Hua K, Xue X, Kou H, Fan J, Tang B, Li J. Characterization of hot
deformation microstructure of a near beta titanium alloy Ti-5553. Journal
Simulation, DCR =150 188.4 (-12.3%) 100 (-12.6%) of Alloys and Compounds. 2014;615:531-7.
Simulation , DCR =200 208 (-3.5%) 108.6 (-4.7%) [4] Arrazola P-J, Garay A, Iriarte L-M, Armendia M, Marya S, Le Maitre F.
Machinability of titanium alloys (Ti6Al4V and Ti555. 3). Journal of
Experimental Results
183.8 121.8 materials processing technology. 2009;209(5):2223-30.
(120 m/min)
[5] Ugarte A, M'Saoubi R, Garay A, Arrazola P. Machining Behaviour of Ti-
Simulation , DCR =100 165.4 (-10.0%) 85 (-30.2%) 6Al-4V and Ti-5553 All0oys in Interrupted Cutting with PVD Coated
Simulation , DCR =150 173.6 (-5.5%) 86 (-29.4%) Cemented carbide. Procedia CIRP. 2012;1:202-7.
[6] Wagner V, Baili M, Dessein G, Lallement D. Experimental
Simulation, DCR =200 179.6 (-2.3%) 89 (-26.9%) characterization of behavior laws for titanium alloys: application to
Ti5553. Key Engineering Materials: Trans Tech Publ; 2010. p. 147-55.
Considering chip thickness and valley, it is possible to well [7] Wagner V, Baili M, Dessein G. The relationship between the cutting
predict chip thickness at DCR =200 where the difference in speed, tool wear, and chip formation during Ti-5553 dry cutting. The
International Journal of Advanced Manufacturing Technology.
310 Melih Ozkutuk and Yusuf Kaynak / Procedia CIRP 58 (2017) 305 – 310

2015;76(5-8):893-912. [14]Manual D. User Manual Deform-User Manual SFTC-Deform

[8] Baili M, Wagner V, Dessein G, Sallaberry J, Lallement D. An V1102,Scientific Forming Technologies Corporation Ed, Columbus, OH,
experimental investigation of hot machining with induction to improve USA2015.
Ti-5553 machinability. Applied mechanics and Materials: Trans Tech [15] Johnson GR, Cook WH. A constitutive model and data for metals
Publ; 2011. p. 67-76. subjected to large strains, high strain rates and high temperatures.
[9] Germain, G., A. Morel, and T. Braham-Bouchnak. Identification of Proceedings of the 7th International Symposium on Ballistics: The
material constitutive laws representative of machining conditions for two Hague, The Netherlands; 1983. p. 541-7.
titanium alloys: Ti6Al4V and Ti555-3. Journal of Engineering Materials [16] Nemat-Nasser, S., et al., Dynamic response of conventional and hot
and Technology, 2013. 135(3): p. 031002. isostatically pressed Ti–6Al–4V alloys: experiments and modeling.
[10] Sun Y, Huang B, Puleo D, Jawahir I. Enhanced Machinability of Ti-5553 Mechanics of Materials, 2001. 33(8): p. 425-439.
Alloy from Cryogenic Machining: Comparison with MQL and Flood- [17] Khan, A.S., Y.S. Suh, and R. Kazmi, Quasi-static and dynamic loading
cooled Machining and Modeling. Procedia CIRP. 2015;31:477-82. responses and constitutive modeling of titanium alloys. International
[11] Kaynak Y, Manchiraju S, Jawahir, IS. Modeling and Simulation of Journal of Plasticity, 2004. 20(12): p. 2233-2248.
Machining-induced Surface Integrity Characteristicsof NiTi Alloy. [18] Ozel T. The influence of friction models on finite element simulations of
Procedia CIRP.2015;31: 557-562. machining. International Journal of Machine Tools and Manufacture
[12] Özel T, Zeren E. Finite element modeling the influence of edge 2006; 46(5): 518-530.
roundness on the stress and temperature fields induced by high-speed [19] Cockcroft M, Latham D. Ductility and the workability of metals. J Inst
machining. The International Journal of Advanced Manufacturing Metals. 1968;96(1):33-9.
Technology. 2007;35(3-4):255-67.
[13] Filice, L., et al., On the finite element simulation of thermal phenomena
in machining processes, in Advanced Methods in Material Forming.
2007, Springer. p. 263-278.