Numerical and experimental investigation of process

Numerical and experimental investigation of process parameters in non-isothermal forward extrusion of Ti–6Al–4V S. Javid Mirahmadi & Mohsen Hamedi

The International Journal of Advanced Manufacturing Technology ISSN 0268-3768 Int J Adv Manuf Technol DOI 10.1007/s00170-014-6108-9

1 23

Your article is protected by copyright and all rights are held exclusively by SpringerVerlag London. This e-offprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Int J Adv Manuf Technol DOI 10.1007/s00170-014-6108-9

ORIGINAL ARTICLE

Numerical and experimental investigation of process parameters in non-isothermal forward extrusion of Ti–6Al–4V S. Javid Mirahmadi & Mohsen Hamedi

Received: 10 November 2013 / Accepted: 25 June 2014 # Springer-Verlag London 2014

Abstract Ti–6Al–4V is the most common Ti alloy that may be worked at supra or subtransus temperatures, conventionally or isothermally. In this article, the effect of extrusion process parameters and die geometry on the extrusion force and adiabatic temperature rise is investigated. The process is considered to be non-isothermal. The results show that the nonuniform effective strain field inside the workpiece generates more elongated grains near the surface of the extrudate. The results also indicate that the ram velocity has the highest effect on both extrusion force and adiabatic temperature rise. The study suggests that in non-isothermal extrusion of Ti–6Al–4V, die and billet temperatures have stronger effect on the extrusion force compared with the die angle. However, effect of the die angle on the adiabatic temperature rise is considerable and cannot be neglected. Effect of the billet temperature on adiabatic temperature rise is more than the effect of die angle. The simulation model and analytical results are verified by experimental investigations that shows good agreement between the corresponding results. The provided insights using the approach and the results presented in this article gives a better understanding of forward extrusion of Ti–6Al–4V material. The results can be used by the tool designers and process planners in developing the tools and the processes that gives better yield and reduces the manufacturing costs. S. J. Mirahmadi (*) School of Mechanical Engineering, College of Engineering, University of Tehran, Tehran, Iran e-mail: [email protected] S. J. Mirahmadi R&D Department, Mavadkaran Engineering Company (MAPNA Group), Tehran, Iran M. Hamedi MAPNA R&D, Tehran, Iran

Keywords Non-isothermal extrusion . Ti–6Al–4V . Adiabatic temperature rise . Extrusion force

1 Introduction Ti–6Al–4V is the most common Ti alloy. It is a two-phase alloy, with low density, high specific strength and attractive corrosion resistance that make it an ideal choice for many aerospace applications [1]. Process of ingot breakdown takes place at temperatures above β transus and is followed by a combination of supra and subtransus operations and appropriate heat treatment to achieve desired microstructure and mechanical properties. Above the β transus temperature, Ti–6Al– 4V has a single phase of β (bcc), and below it, two phases of α (hcp) and β are present [2, 3]. Cooling from single phase of β, produces the lamellar microstructure that might be unsuitable in some applications; so the process of forming should be designed in a way to prevent the material temperature exceeding the β transus temperature. Flow stress of Ti–6Al–4V is very sensitive to temperature and strain rate change [4] and with the presence of die-chilling in non-isothermal processes causes some operational difficulties. Extrusion is a basic manufacturing process that is used to fabricate the parts with constant cross section. In the extrusion of Ti alloys, glass is applied to the workpiece as a lubricant that also acts as a thermal insulation to reduce the effect of diechilling. The results of a few studies about the non-isothermal forward extrusion of Ti–6Al–4Vare published. Udagawa et al. investigated the effect of process parameters upon Ti–6Al–4V tubes quality during extrusion numerically [5]. Srinivasan and Venugopal studied the warm open-die extrusion of Ti-64. Open-die was used to reduce the workpiece-die contact and reduce the process force [6]. Investigations of Zhang et al.

Author's personal copy Int J Adv Manuf Technol

showed that non-isothermal backward extrusion of Ti-15-3 alloy was not successful and instead of it, isothermal backward extrusion was used [7]. Li et al. studied extrusion of Ti– 6Al–4V numerically and experimentally. They concluded that the heat generated in Ti alloy extrusion is significant [8]. Bergamini et al. studied the extrusion of T- and U-shape cross section Ti-64 extrudates. Their conclusion is that the temperature of hot plastic deformation should be slightly higher than the β transus [9]. Considering extrudate distortion after hot extrusion of titanium billets, Damodaran and Shivpuri have performed a sensitivity analysis of the effect of process parameters on the distortion of Ti–6Al–4V billet in hot asymmetric extrusion [10]. Shin et al. studied the non-isothermal backward extrusion of Ti–6Al–4V and optimized the process to prevent the defect formation [11]. Considerable refinement of the microstructure is the result of extrusion of Ti–6Al–4V in the range 600–700 °C with a lamellar microstructure that is reported by Zherebtsov et al. [12]. Surveying the published literature reveals that the number of publications concerned with extrusion of Ti–6Al–4V is limited. There is not any thorough study of the effect of process parameters on the extrusion force and adiabatic temperature rise and interaction of process parameters. In this paper, the effect and significance of process parameters and die geometry on the extrusion force and adiabatic temperature rise is presented. For this purpose, a 2D axisymmetric FEM model of forward extrusion is developed and validated by experimental work. Then this model is used to simulate the non-isothermal forward extrusion of Ti–6Al–4V with various parameters. The investigated parameters include two groups, i.e., process parameters (ram velocity, die, billet and punch temperature) and die geometry parameters (die semi-angle (α), entrance and exit fillet radius). Effect of each parameter on the extrusion force and adiabatic temperature rise is analyzed and discussed.

Fig. 1 Die parameters

the workpiece, 2D axisymmetric modeling is used. Simplification of 3D model to 2D axisymmetric is shown in Fig. 2. The flow stress of material is determined as a function of strain rate, temperature, and strain by isothermal hot compression tests. The flow data are corrected for the effect of friction and adiabatic temperature rise and are implemented in the simulations. Variation of true stress as a function of true strain, strain rate, and temperature is presented in Fig. 3. Heat transfer coefficient at tool/workpiece interface as a function of contact pressure [13], friction factor as a function of temperature [14], and material properties are extracted from the literature [15, 16] and presented in Table 1. The geometry of workpiece, die and punch are modeled in a CAD package and imported to the FEM software via the standard format

2 Modeling and simulations 2.1 FEM model development A coupled thermomechanical modeling based on the rigidviscoplastic finite element method (FEM) is developed to simulate the extrusion process with different design parameters. Some of these parameters are related to die, i.e., die angle, entrance and exit fillet radius of the die zone and others are related to process, i.e., die, punch, and workpiece temperature and the ram velocity. Die-related parameters are depicted in Fig. 1. A commercially available FEM package is utilized to perform the simulations. Because of axisymmetric geometry of

Fig. 2 Simplification of 3D model to 2D axisymmetric: a initial billet, b final extrudate

Author's personal copy Int J Adv Manuf Technol 100

100 870 °C

60

890 °C 910 °C 930 °C 950 °C

40 20

80 True Stress (MPa)

True Stress (MPa)

0.03 s-1

80

60 0.003 s-1

40 20

0

0.0003 s-1

0 0

0.2

0.4 True Strain

0.6

(a) −1

Fig. 3 Stress–strain curves at a 0.003 s

0

0.2

0.4 True Strain

0.6

(b)

and b 910 °C

IGES. The die and the punch are considered to be rigid and the workpiece is rigid-viscoplastic. The 2D axisymmetric model used for extrusion simulations is shown in Fig. 4. At the symmetry axis, the nodes of the workpiece are constrained to move only in vertical direction. 2.2 Parameter selection to study their effects Minimization of extrusion force results in more die life leading to lower capacity press machine, and therefore merits proper investigation. Effect of die angle on extrusion force is investigated by considering constant die and process parameters and varying die angle. The results illustrated in Fig. 5

show that force history corresponding to die angles 20° and 30° are coincided. Increasing the die angle results in requiring more extrusion force; therefore, optimum die angle to minimize the extrusion force lies between 20° and 30°. Since different die angles result in different stress state, each die angle determines evolution of ductile damage within the extruding workpiece. In order to predict the probability of void or crack initiation, several damage models are developed [17–20]. Cockcroft-Latham model is a damage model that is extensively used in process simulation to predict the failure occurrence [21–28]. Cockcroft-Latham damage is based on the tensile strain energy per unit of volume and can be expressed as [20]:

Table 1 The parameters implemented in FEM model

Heat capacity (Nmm−2 K−1)

H-13

Ti–6Al–4V

5.60

3.10 at 325 °C 3.81 at 625 °C 5.09 at 925 °C

Thermal conductivity (Ns−1 K−1)

Punch

28.6 at 215 °C 7.6 at 325 °C 28.4 at 350 °C 12 at 625 °C

Initial billet

28.4 at 475 °C 16 at 925 °C Emissivity 0.15 0.6 Heat transfer coefficient of die/ 0.24 at 200 MPa Friction factor at die/workpiece 0.27 at 750 °C interface 0.23 at 800 °C 0.20 at 850 °C

Axis of symmetry

Die

0.14 at 950 °C 0.10 at 1000 °C Fig. 4 2D axisymmetric model adopted for extrusion simulations


16 80° 70° 60° 50° 40°

Extrusion Force (Ton-Force)

14 12 10 8 30° 20°

6 4 2 0 0

10

20 Stroke (mm)

30

40

Fig. 5 Effect of die semi-angle on extrusion force, die and punch temperature=300 °C, billet initial temperature=940 °C and ram velocity=50 mm/s

Z¯εf σ σ¯ 0

dε¯ ¼ C 1

ð1Þ

where εf is the fracture strain, σis the equivalent stress, σ* is the highest tensile stress, is the equivalent strain, and C1 is the critical damage value. Distribution of normalized Cockcroft-Latham damage for different die angles, shown in Fig. 6, indicates that increasing die angle increases level of damage within the workpiece. Moreover, for die angles 20° and 30°, the maximum damage occurs on the external surface of the extrudate and is observable, but for other die angles, the maximum damages is located at center of the extrudate; i.e., by increasing the die angle, probability of failure at the center of the extrudate increases. In Fig. 7, evolution of effective strain field inside the extrudate during the process is shown. As shown, the effective strain field at the tip of the workpiece is

Fig. 6 Effect of die semi-angle on normalized Cockcroft–Latham damage distribution for die and punch temperature=300 °C, billet initial temperature=940 °C and ram velocity=50 mm/s

α=20°

α=30°

α=40°

α=50°

α=60°


Max. strain Fig. 7 Evolution of effective strain field inside the extrudate

different from other portions of the workpiece. After the workpiece tip, the effective strain distribution in any transversal cross section becomes steady. In any transversal cross section, the maximum effective strain is near the surface of the workpiece and reduces towards the center of the workpiece. The maximum effective strain occurs near the tip of the workpiece. Because of non-uniform distribution of the effective strain in the workpiece, the microstructure of the workpiece is expected to be non-uniform. Given the effect of die angle on extrusion force and damage distribution, evaluating the effect and significance of die angle and other parameters on the process is performed. Level selection of each parameter and analysis of its effect is based on a ½ fraction twolevel factorial design resolution VII with principle fraction [29]. Considered parameters and their values are shown in Table 2. Geometrical die parameters and the pertaining values are depicted in Fig. 1.

Table 2 The parameters and their values used in FEM simulation

3 Validation of simulation model with experimental work In order to experimentally investigate the extrusion process of Ti-6A-4V, an experimental set-up including an industrial vertical 630-tonne hydraulic press is utilized that is shown in Fig. 8. The press is equipped with a closed-loop servo-hydraulic circuit to control the position-velocity relationship during the forming process. A pressure transmitter is used to measure the pressure and consequently the force during forming stroke. The die consists of two halves that are hold together by a special designed die holder made of H-13 hot working tool steel. The die and the punch are also made of H-13 and are heat treated to gain hardness of 48 HRC. Heating up the die holder and the die is done by a heating unit which consists of a resistant electrical heater. A type K thermocouple is inserted in the die holder body to observe the temperature. To prevent the heat damage to the press bed, two water cooling plates are placed on the press bed.

Die parameter

Values

Process parameter

Values

A: Die semi-angle B: Entrance fillet radius (mm) C: Exit fillet radius (mm) –

20, 40 2, 5 2, 5 –

D: Ram velocity (mm/s) E: Die temperature (°C) F: Billet temperature (°C) G: Punch temperature (°C)

7, 50 150, 300 940, 980 150, 300


Punch assembly Heating unit Die holder

Cooling plates

Thermocouple

Fig. 8 Experimental setup

Fig. 9 Microstructure of initial billet (×1,000)

The initial billet of Ti–6Al–4V with diameter of 20 mm has a chemical composition and microstructure shown in Table 3 and Fig. 9 correspondingly. To evaluate the results of the simulations, a die with semi-angle of 40° and entrance and exit fillet radius of 5 mm is made and a set of test are performed with different temperatures of billet, die, and punch. The extrusion forces measured in experimental tests are compared with simulation results in Fig. 10 with billet temperature of 940 °C and die and punch temperature of 300 °C. Cases with ram velocity of 7 mm/s and die temperature of 150 °C having the maximum extrusion force more than 26-ton-force lead to plastic deformation of the punch. The experimental results are in good agreement with FEM results. An extruded workpiece and its corresponding FEM simulation are displayed in Fig. 11. The simulation result predicted that non-uniform distribution of effective strain causes non-uniformity in the microstructure of the workpiece. Metallographic studies are done to investigate this phenomenon. Figure 12 shows that in the non-extruded portion of the workpiece the grains remain equiaxed; but in the extruded section, more elongated grains exist due to higher effective strain.

4 Investigation of critical process parameters and their interaction effects As mentioned in the previous section, because of high sensitivity of flow stress of Ti alloys to temperature, unsuitable selection of process parameters may lead to excessive increase in extrusion force and consequently producing plastic deformation of the punch. Moreover, in α + β forming, adiabatic temperature rise may put the workpiece in single-phase region of β and results in undesirable microstructure. So extrusion force and workpiece temperature both should be assessed in planning the process. A set of investigations using parameters given in Table 2 are done and two criteria of extrusion force and adiabatic temperature rise are studied. 4.1 Effect of process parameters on extrusion force To access the effect of each considered parameters given in Table 2 on extrusion force, analysis of variance is done on the obtained results. According to Pareto chart of the standardized effects on extrusion force shown in Fig. 13, ram velocity, die temperature, billet temperature, entrance fillet radius, die angle, and exit fillet radius have significant effect on the

Table 3 Chemical composition of initial billet Element

Al

V

C

N

Fe

H

O

Y

Each

Total

Ti

%

6.54

4.13

0.024

0.006

0.20

0.002

0.180

15 20 25 30 35 35

Hold Values Entrance Radius Exit Radius Ram Velocity Die Temperature Punch Temperature

5 5 7 300 150

Fig. 21 Contour plot of adiabatic temperature rise vs. billet temperature and die angle

Author's personal copy Int J Adv Manuf Technol Adiabatic Temperature Rise (°C)

100

Ram Velocity (mm/s)

80 16 24 32 40

60

40

30

40

50 60 Die Angle

70

6 24 32 40 48 48

Hold Values Entrance Radius Exit Radius Die Temperature BilletTemperature PunchT emperature

20

0 20

< – – – – >

&

80

5 5 300 940 150

Fig. 22 Contour plot of adiabatic temperature rise vs. ram velocity and die angle for initial billet temperature of 940 °C

force and adiabatic temperature rise were investigated. The results may be summarized as the following: &

& & &

& &

Non-uniform distribution of the effective strain inside the workpiece results in non-uniform microstructure in the extrudate. Greater effective strain near the surface of the workpiece causes more elongated grains in that zone. Despite usage of glass lubricant as a thermal barrier on the surface of the workpiece, dwell time may lead to excessive die-chilling and increase of extrusion force. Ram velocity has the most effect on the extrusion force. Increasing ram velocity decreases the extrusion force. Die and billet temperatures have high effect on the extrusion force, but die angle and entrance and exit fillet radius of die have a low effect on the extrusion force. Increasing die and billet temperatures decreases the extrusion force. In Ti–6Al–4V extrusion, adiabatic temperature rise is significant and should be considered in process design. Ram velocity and billet temperature have the most effect on the adiabatic temperature rise. Increasing ram velocity causes raise in the rate of adiabatic temperature rise, but increasing billet temperature decreases this rate. 50

Ram Velocity (mm/s)

40

Excessive adiabatic temperature rise Hold Values Enterance Radius Exit Radius Die Temperature Billet Temperature Punch Temperature

30

20

10

5 5 300 980 300

OK

0 20

30

40 50 Die Angle

60

70

80

Fig. 23 Plot of suitable die angles and ram velocities to prevent excessive adiabatic temperature rise for initial billet temperature of 980 °C

Die angle and entrance fillet radius also have effect on the adiabatic temperature rise, but entrance fillet radius has a low effect.

The methodology and results presented in this article provide more insight and better understanding of forward extrusion of Ti–6Al–4V material. The results help the tool designers and process planners to develop the tools and the processes that give better yield and therefore reducing the cost of the manufactured parts. Acknowledgements The authors extend their gratitude toward R&D Deputy of MAPNA Group for financial support of this project.

References 1. Momeni A, Abbasi SM (2010) Effect of hot working on flow behavior of Ti–6Al–4V alloy in single phase and two phase regions. Mater Des 31:3599–3604 2. Donachie MJ (2000) Titanium: a technical guide. 2nd edn. ASM International, pp. 13-24. 3. Semiatin SL, Seetharaman V, Weiss I (1997) The thermomechanical processing of alpha/beta titanium alloys. JOM 49(6):33–39 4. Cai J, Li F, Liu T, Chen B, He M (2011) Constitutive equations for elevated temperature flow stress of Ti–6Al–4V alloy considering the effect of strain. Mater Des 32:1144–1151 5. Udagawa T, Kropp E, Altan T (1992) Investigation of metal flow and temperatures by FEM in the extrusion of Ti–6Al–4V tubes. J Mater Proc Tech 33(1):155–174 6. Srinivasan K, Venugopal P (1993) Warm open-die extrusion of Ti– 6Al–4V. J Mater Proc Tech 38(1):265–277 7. Zhang S, Xu Y, Shang Y, Yang K, Wang Z (2001) Lubrication in hot tube extrusion of superalloys and Ti alloys. J Mater Sci Tech 17(1): 113–114 8. Li LX, Rao KP, Lou Y, Peng DS (2002) A study on hot extrusion of Ti–6Al–4V using simulations and experiments. Int J Mech Sci 44(12):2415–2425 9. Bergamini R, Mapelli C, Venturini R (2003) Hot extrusion experiments performed on Ti–6Al–4V for the production of special cross section. Met Sci Techno 21(2):18–25 10. Damodaran D, Shivpuri R (2004) Prediction and control of part distortion during the hot extrusion of titanium alloys. J Mater Proc Tech 150(1):70–75 11. Shin TJ, Lee YH, Yeom JT, Chung SH, Hong SS, Shim IO, Park NK, Lee CS, Hwang SM (2005) Process optimal design in non-isothermal backward extrusion of a titanium alloy by the finite element method. Comput Method Appl 194(36):3838–3869 12. Zherebtsov S, Mazur A, Salishchev G, Lojkowski W (2008) Effect of hydrostatic extrusion at 600–700 °C on the structure and properties of Ti–6Al–4V alloy. Mat Sci Eng A-Struct 485(1):39–45 13. Hu ZM, Brooks JW, Dean TA (1998) The interfacial heat transfer coefficient in hot die forging of titanium alloy. J Mech Eng Sci 212(6):485–496 14. Li LX, Peng DS, Liu JA, Liu ZQ, Jiang Y (2000) An experimental study of the lubrication behavior of A5 glass lubricant by means of the ring compression test. J Mater Proc Tech 102(1): 138–142 15. Boivineau M, Cagran C, Doytier D, Eyraud V, Nadal MH, Wilthan B, Pottlacher G (2006) Thermophysical properties of


16.

17. 18. 19. 20. 21.

22.

solid and liquid Ti–6Al–4V (TA6V) alloy. Int J Thermophys 27(2):507–529 Zhou J, Li L, Duszczyk J (2003) 3D FEM simulation of the whole cycle of aluminium extrusion throughout the transient state and the steady state using the updated Lagrangian approach. J Mater Proc Tech 134(3):383–397 Freudenthal AM (1950) The inelastic behavior of solids. Wiley. Oyane M (1972) Criteria of ductile fracture strain. Bull JSME 15(90): 1507–1513 Rice JR, Tracey DM (1969) On the ductile enlargement of voids in triaxial stress fields. J Mech Phys Solids 17(3):201–217 Cockcroft MG, Latham DJ (1968) Ductility and the workability of metals. J Inst Metals 96(1):33–39 Landre J, Pertence A, Cetlin PR, Rodrigues JMC, Martins PAF (2003) On the utilisation of ductile fracture criteria in cold forging. Finite Elem Anal Des 39(3):175–186 Quan GZ, Wang FB, Liu YY, Shi Y, Zhou J (2013) Evaluation of varying ductile fracture criterion for 7075 aluminum alloy. T Nonferr Metal Soc 23(3):749–755

23. Takuda H, Mori K, Hatta N (1999) The application of some criteria for ductile fracture to the prediction of the forming limit of sheet metals. J Mater Process Technol 95:116–121 24. Ozturk F, Lee D (2004) Analysis of forming limits using ductile fracture criteria. J Mater Process Technol 147(3):397–404 25. Nicolaou PD, Semiatin SL (2005) An analysis of cavity growth during open-die hot forging of Ti–6Al–4V. Metall Mater Trans A 36(6):1567–1574 26. Kukuryk M (2012) Analysis of deformation and damage evolution in hot elongation forging. Arch Metall Mater 55(2):417–424 27. Quan GZ, Wang FB, Liu YY, Shi Y, Zhou J (2013) Evaluation of varying ductile fracture criterion for 7075 aluminum alloy. T Nonferr Metal So 23(3):749–755 28. Mirahmadi SJ, Hamedi M, Ajami S (2014) Investigating the effects of cross wedge rolling tool parameters on formability of Nimonic® 80A and Nimonic® 115 superalloys. Int J Adv Manuf Technol. doi: 10.1007/s00170-014-6047-5 29. Montgomery DC (2001) Design and Analysis of Experiments. 5th edn. John Wiley & Sons, pp. 218-276.