International Conference on Trends in Mechanical and Industrial Engineering (ICTMIE'2011) Bangkok Dec.

, 2011

Finite Element Modeling of Secondary Phase Grain Size on the Stress Strain Curve of Ti alloy
Srinivasu.G, NarasimhaRao. R, and Nandy. T. K.

AbstractThe stress-strain curve of 0.16 of low cost beta (LCB) Ti alloy (Ti10V4.5Fe3Al) was measured and then it was simulated by using finite element method (FEM) based on the individual stress-strain curves of the single-phase and , then the measured result compared with the simulated curve. The results showed that the simulated result closely fits with the measured one, which proves that this FEM built in this work is acceptable. This FEM model was used for calculation of stress-strain curve at three different grain sizes at constant volume fraction of is 16%. The results showed that finer grain size having higher yield strength compared with the coarser one. The results showed that the stress gradient exists in both the phases and the distributions of stress are non-homogeneous. The stress in the phase is lower than the stress in the phase, whereas the strain in the phase is higher than that in the phase. KeywordsGrain size, Finite Element Method, Stress-Strain Curve, Titanium alloy. I. INTRODUCTION ITANIUM (Ti) alloys having a unique combination of mechanicaland physical properties and excellent corrosion resistance to sea water,which make them popular for a variety of applications [1].Titanium alloys are excellent candidates for aerospace applications owing to their high strength to weight ratioand excellent corrosion resistance [2]. The expansion oftitanium applications to non-aerospace industries (e.g, automotive,chemical, energy, marine, biomedical, sports, and architecture)entails improvements in the understanding of titaniummetallurgy, advances in processing methods, ability to manufacturecomponents without defects, and development of lowcostalloys [3].Beta titanium alloys offer a wide range of mechanical and chemical properties and can be thermo mechanically processed to achieve high strength, toughness and fatigue resistance even in large sections [4].The development of a low cost beta (LCB) alloy [5] like Timetal LCB (Ti4.5Fe6.8Mo1.5Al) makes use of ferromolybdenum addition to reduce the cost so that it becomes competitive for automobile and other aerospace

applications. The alloy possesses excellent workability and mechanical properties which are comparable to that of high strength steel [6]. Mechanical properties of titanium alloys are important criteria of material service capabilities both in aerospace and industrial applications. Microstructure of the alloy is one of the important factors controlling both the tensile strength and the fatigue strength. With different percentages of and phases, several metastable microstructures can be formed. Thus, it is formed with integration of several physical properties, functional performances, and strain in different phases and even in different grains is inhomogeneous, it is hard to express the distribution of stress and strain in the titanium alloy clearly [7].Therefore, anyinnovation in the design and predicting the properties with the help of microstructure will lead to reduce experimentation and large overall cost savings. The key to these innovations is athorough understanding of the materials constitutive responseto the imposed conditions and detailed process analysis. This isachieved by effective material modeling and integration of microstructuralinformation with process simulation and optimizationmethodologies. Many investigators [8, 9] calculated the stress-strain curve and then compared with the measured one, by using the stressstrain curves of the individual phases. In this paper the stressstrain curve of 0.16 of Ti alloy calculated based on the individual stress-strain curves of single and phase alloys using finite element method (FEM). The simulated result was compared with the measured one and then, the distribution of stress and strain of the 0.16 of Ti10V4.5Fe3Al alloy was analyzed in this work. Finally, with the FEM model the effect of secondary phase () grain size on the stress-strain curve is calculated. II. MATERIALS AND METHODS A. Materials and Methods The LCB Ti alloy i.e., Ti-10V-4.5Fe-3Al has been used in this work. The chemical composition of the alloy was analyzed using spectrometer and the chemical composition shown in Table 1. The Ti alloy was produced by using vacuum arc remelting process. The alloy was processed thermo-mechanically in order to get the equiaxed microstructure. Thermo-mechanical process includes forging and rolling. The ingot was forged at 900 C (beta forging) in order to remove the defects in the
175

Srinivasu.G is with the National Institute of and Technology, Warangal, 506004, INDIA (corresponding author to provide phone:+91-9493040920; fax::+918702459547 e-mail:srinivasu@nitw.ac.in. NarasimhaRao. R, was with the National Institute of and Technology, Warangal, 506004, INDIA (phone: +91-9441569066; email:rnraonitw@gmail.com). Nandy.T.K., Scientist-Gis with the, Defense Metallurgical Research Laboratory (DMRL)Hyderabad.

International Conference on Trends in Mechanical and Industrial Engineering (ICTMIE'2011) Bangkok Dec., 2011

ingot. Further it rolled into 10mm thick slab at 750 C. For microstructural examination, samples were cut from the 10 mm thick slab, and heat treated with the tensile specimens at required conditions, polished metallographically using standard metallographic technique and finally etched with Krolls reagent (2ml of HF, 6ml of HNO 3 , 92ml of Distilled Water). Etched samples were examined under scanning electron microscopy and optical microscope.
TABLE I. Chemical composition of the Ti-10V-4.5Fe-3Al alloy
Elements V (wt %) Al (wt %) 3 Fe (wt %) 4-5 O
(PPM)

according to the ASTM standard.The stress-strain curves of experimentally measured , 0.16 and literature were shown in Fig. 2.

N (PPM) 100

H (PPM) 19

Ti

Ti10V4.5Fe -3Al

9.6-10

800

Rest

Fig. 2. Stress-strain curves of full and 0.16 volume fraction of , and full of Tialloy.

a
Phase (white)

Phase (Black)

B. Finite element modeling From the Fig. 1 it is observed that the phase is distributed homogeneously on the phase. A 2D model was built by using FEM replicates the microstructure. The FEM model consists of 800 triangular elements, which are connected by 1681 nodes (Fig. 3). The violet elements represent the phase, whereas the green elements represent the phase. The volume fractions of phase in the microstructure as well as in the FEM model are same of 16%, while remaining is beta phase. Loads areapplied in the vertical direction in terms of displacement are as follows: 0.005, 0.05, 0.1, and 0.3mm.

TABLE II. Parameters of stress and strain for simulation

b
No 1 2 3 4 5 6

Single Phase Strain Stress (MPa)

Single Phase Strain Stress (MPa)

0.017 0.022 0.035 0.061 0.1 0.2

296 318 341 349 355 369

0.015 0.02 0.05 0.1 0.15 0.2

1187 1231 1195 1181 1170 1156

Fig. 1. Microstructure of Ti10V4.5Fe3Al alloy (a) at 0.16 volume fraction of . (b) Full beta

The stress-strain curves of single phase and are required in order to simulate the stress-strain curve. Hence the stressstrain curve of full was calculated experimentally and is taken from the literature [7-8]. The stress-strain curves of full and 0.16 volume fraction of specimen weremeasured by using hydraulic servo universal testing machine. The tensile samples are prepared
176

The Youngs modulus of the single and phase alloys are 17412 MPa and 79133MPa respectively, and Poissons ratio is 0.31. The measured parameters of stress and strain are used in FEM model shown in Table 2.

International Conference on Trends in Mechanical and Industrial Engineering (ICTMIE'2011) Bangkok Dec., 2011

III. SIMULATION RESULTS

Phase (Green)

A. Comparison between the measured and simulated results of the Ti alloy 1200 1000 Stress (Mpa) 800 600 400 200 Experimental Curve Calculated Curve

Phase (Pink)

Fig. 3.2D FEM model for the simulation (finer grains)

0 0 5 Strain (%) 10 15
Fig. 6. Comparison between measured and calculated stress-strain curves

Fig. 3 indicates the FEM model fine grain size of secondary phase (), Two more FEM model are developed with different size i.e, medium and coarser grain sizes of secondary phase of Ti alloy (Fig. 4 and 5).

Fig. 5. FEM model for the simulation of medium grain size of (at constant volume fraction of at 16%)

Fig. 4. FEM model for the simulation of coarser grain size of (at constant volume fraction of at 16%)

By using the ANSYS software, the stressstrain curve of the0.16 Tialloy was calculated based on the individual stressstrain curves of single-phase and single-phase , then compared with the measured results. The measured stress strain curve and the simulated one by calculating with the twodimension model are shown in Fig. 6. As there is little error between the measured result and the simulated one, but the curve fit is good. Therefore, the FEM used in this work is valid, which can be used for further simulation to analyze the effect factors and discussion for the distribution of microstress and strain of the 0.16 Ti10V4.5Fe3Al. The reasons for error between the measured curve and the simulated one are because of following assumptions. (1) Generally the microstructure is in three-dimensional form, but it is very difficult to build the 3D form hence the model is restricted to two-dimensional and this model adopted for simulation. (2) The regular elements are built by linking nodes, which is different from the real microstructure. The result will be also affected by the number of elements. (3) The shape and distribution of phase in the model are not absolutely the same as the real one as well as its fraction; meanwhile, the errors can be caused by the inhomogeneous distributions in the microstructure. (4) The influence of grain boundary and interphase boundary is ignored in the model. After the validation, the same FEM model used to for the simulation to know the effect of grain size on the stress strain curve. After the simulation the stress strain curves are calculated at two different grain sizes and compared with the finer grain size. It is observed that the finer grain size curve is higher as compared with the other two coarser grain curves (Fig. 7).

177

International Conference on Trends in Mechanical and Industrial Engineering (ICTMIE'2011) Bangkok Dec., 2011

1200 1000 800 Stress (Mpa) 600 400 200 0 5 Strain (%) 10 15 Fig. 7. Comparison between the calculated stress-strain curves at different grain sizes of (by kept volume percent of is constant at 16%) B. The analysis of the distribution of the stress and strain of the Tialloy 0 Coarser size Medium size Finer size

The stress distributions (First principle stress) are non uniform in all cases and also noted that the stress in the beta phase is higher than in alpha phase. Fig. 8.shows the stress distributions of Ti alloy at a displacement of 0.1mm for different grain sizes of secondary phase alpha. The maximum stress decreases with the increase of grain size, where as the minimum stress increases with the increase of grain size of secondary phase .

a b

Fig. 9.First principleelastic strain distributions at displacement of 0.1mm of Ti alloy at different grain sizes of (at constant volume fraction is 16%) (a) Finer (b) Medium& (c) coarser Fig. 8.First principle stress distributions at displacement of 0.1mm of Ti alloy at different grain sizes of (at constant volume fraction is 16%) (a) Finer (b) Medium& (c) coarser

178

International Conference on Trends in Mechanical and Industrial Engineering (ICTMIE'2011) Bangkok Dec., 2011

Fig. 9.represents the first principle elastic strain distributions at 0.1mm of Ti alloy for different grain sizes of alpha phase. It is observed that the elastic strain is more in alpha phase as compared with the strain in the beta phase because the beta phase is harder than the alpha phase. It is also observed that the elastic strain increases with the increase in load.

beta phase is harder than the alpha phase is softer. It is also observed that the elastic strain increases with the increase in load. The first principle total strain is the sum of plastic strain and elastic strain. The distributions of first principle total strain are similar to the distributions of first principle plastic strain. It also observed that the first principle total strain is more in alpha phase as compared to that of beta phase. IV. CONCLUSIONS 1. The measured stressstrain curve of volume fractions of 0.16 is matches with the calculated curve, so the work is preferable and can be used for further simulation to analyze the effect of factors and discuss the distribution of micro-stress and strain of this alloy. 2. The distributions of first principle stress and strain in the and phases of the Ti-10V-4.5Fe-3Al alloy were simulated in all cases. The stress increases with an increase in the load, the maximum stress is in the phase, where as the minimum stress is in the phase. The strain increases with an increase in the load too, whereas the maximum strain is in the phase, and the minimum one in the phase. 3. The stress strain curve of finer grain size is higher than the coarser one. 4. The maximum stress developed in the FEM model increases with the decrease in grain size of secondary phase i.e. . References
[1] R. Boyer, G.Welsch, and E.W. Collings, ed., Materials Properties Handbook: Titanium Alloys, ASM International, Materials Park,OH, 1994. [2] R.R. Boyer, An overview on the use of titanium in the aerospace industry, Materials Science and Engineering, vol.A213, 1996, pp. 103114. [3] Makoto Yamada, An overview on the development of titanium alloys for non-aerospace application in Japan, Materials Science and Engineering, vol. A213, 1996, pp.8-15. [4] D. Eylon, in: ShiroFujishiro, D. Eylon, T. Kishi (Eds.), Metallurgy and Technology of Practical Titanium Alloys, TMS, Warrendale, PA, 1994, pp. 2936. [5] P.J. Bania, in: D. Eylon, R.R. Boyer, D.A. Koss (Eds.), Beta Titanium alloys in 1990s, TMS, Warrendale, PA, 1993, pp. 314. [6] P.J. Bania, A.J. Hutt, R.E. Adams, W.M. Parris, in: F.H. Froes, I.L. [7] ZHAO Xiqing, ZANG Xinliang, WANG Qingfeng, Park Joongkeun, and YANG Qingxiang, Numerical simulation of the stress-strain curve and the stress and strain distributions of the titanium-duplex alloy, Rare Metals, Vol. 27, No. 5, Oct 2008, p. 463. [8] Bo Liao, ChunLing Zhang, Jing Wu, DayongCai, Chunmei Zhao, XuejunRen, Qingxiang Yang, Numerical simulation of the stressstrain curve of duplex weathering steel, Materials and Design, vol. 29, 2008, pp. 562567. [9] JindrichJinoch, SreeramamurthyAnkem And Harold Margolin, Calculations of Stress-Strain Curve and Stress and Strain Distributions for an - Ti-8Mn Alloy, Materials Science and Engineering, vol. 34, 1978, pp. 203 211.

Fig. 10.First principleplastic strain distributions at displacement of 0.1mm of Ti alloy at different grain sizes of (at constant volume fraction is 16%) (a) Finer (b) Medium& (c) coarser

Fig. 10.indicates the first principle plastic strain distributions at 0.1mm of Ti alloy for different grain sizes of alpha phase. It is observed that the plastic strain is more in alpha phase as compared with that of beta phase because the
179