DYNAMIC MATERIALS MODELING TECHNIQUE FOR OPTIMIZATION OF WARM WORKABILITY OF ARMCO IRON AND BINARY IRON ALLOYS G. S. Avadhani Department of Metallurgy, Indian Institute of Science, Bangalore 560 012, India. ABSTRACT Warm working characteristics of alpha iron, Fe-5Si and Fe-5Co alloys are studied in the temperature range of 400-900C and strain rate range of 0.001-100s-1 using compression tests. Processing maps are developed using Dynamic Materials Model (DMM). The results show that alpha iron undergoes Dynamic Re-crystallization (DRX) with a peak efficiency of power dissipation of 35% at 800C and 0.001s-1. Silicon addition increases the efficiency of power dissipation in the DRX domain to 55% and this is attributed to a decrease in the Curie temperature. Cobalt addition, on the other hand, increases the Curie temperature of iron and shifts the DRX domain to higher temperatures. Alpha iron exhibits flow instabilities in the temperature range of 400-700C when strain rate is above 10s-1. The flow instability in Fe-5Si alloy is more intense than in alpha iron, and the instability regimes are wider in both Fe-5Si and Fe-5Co alloys. The Processing Maps thus developed facilitate in avoiding these regimes during warm working of these materials. INTRODUCTION Processing of materials at temperatures in the range 0. 4 – 0.6 Tm (Tm is melting point temperature) is termed as warm working. It offers the advantage of obtaining a combination of increased strength and ductility in addition to good surface finish and close dimensional tolerance. In view of these advantages, critical components for automobiles are forged in the warm working regime(1). Hitherto, processes are developed using trial and error techniques which are expensive as well as time consuming and may not always lead to a successful solution or optimization. In recent years, however, the trial and error techniques are replaced by modeling techniques, which are developed on the basis of science based principles. The processing map approach(2) developed on the basis of dynamic materials model(3) has been found to be a very useful tool for understanding the hot deformation mechanisms in a variety of materials and optimize the hot workability (4). In this approach, the power dissipation efficiency of the material is defined as: = 2m/ (m + 1) ….. (Eq.1) 1-1 where m is the strain rate sensitivity of flow stress. The variation of with temperature and strain rate represents the characteristics of power dissipation occurring through microstructural changes in the work piece, which constitutes a power dissipation map. A continuum instability criterion (5) developed on the basis of extremum principles of irreversible thermodynamics, which is given by: ln( m / m 1) m 0 …..(Eq. 2) ( ) = ln is used to delineate the regimes of flow instability. The instability parameter ( ) is evaluated as a function of temperature and strain rate and plotted to obtain an instability map. It is superimposed onto the power dissipation map to obtain a processing map. Deformation of pure iron in the warm working range at 1s-1 using high speed torsion was studied(6) and the peak ductility occurring at about 860C was associated with a recrystallized structure during deformation. At lower strain rates, however the ductility peak was observed(7) at about 800C. Dynamic re-crystallization (DRX) at lower stresses and dynamic recovery (DRY) at higher stresses in hot torsion were reported by Glover and Sellars(8). While diffusion controlled thermally activated glide was observed in alpha iron and Fe-2.8 Si alloy in compression(9), one of the most important observation is the effect of ferro-and para-magnetism on the creep deformation of alpha iron(10). It is observed that the creep rates were lower and the activation energy for creep was higher in the ferromagnetic regime due to lower diffusion rates which are also strongly influenced by magnetic forces. Also, it has been shown that, re-crystallization behavior in pure iron is affected significantly by Curie temperature(11). The aim of the investigation is to evaluate the deformation behavior of alpha iron, Fe-5Si and Fe-5Co alloys in the warm working range of temperature and in a wide range of strain rate. It is well known that silicon addition to iron decreases the Curie temperature while cobalt addition raises it(12). The study will therefore throw light on the influence of magnetic forces on the deformation characteristics of alpha iron. The approach of processing maps has been used in this study since the power dissipation characteristics are sensitive to phase transformations and therefore will effectively reveal the influence of ferromagnetism on deformation. EXPERIMENTAL The chemical composition of armco iron, Fe-5Si and Fe-Sco alloys and their initial grain size are given in Table-1. Armco iron was forged at 900C and annealed for two hrs. at 750C followed by furnace cooling. The Fe-5Si and Fe-5Co alloys were hot forged at 700C, annealed at 875C and furnace cooled. 1-2 Cylindrical specimens of 10mm and 15 mm height were used for compression resting. Care was taken to obtain closely parallel load bearing surfaces for the specimen. In addition, grooves were provided on these surfaces so that effective lubrication (MoS2 up to 700C and molten glass at higher temperatures) was ensured during compression. The temperature of the specimen was monitored with the aid of a chromel/alumel thermocouple embedded in a 0.8 mm diameter hole machined at half the specimen height. This thermocouple was also used for the measurement of the adiabatic temperature rise in the specimen during deformation. A computer controlled servo-hydraulic testing machine (Custom built by DARTEC, UK) was used for the compression tests. The machine was equipped with an exponentially decaying actuator speed, enabling constant true strain rates in the range 0.001 – 100s-1 to be imposed on the specimen. Isothermal tests were conducted by surrounding the specimen, platens and push rods with a resistance furnace. The temperature was controlled to within 2oC. The adiabatic temperature rise was recorded using a Nicolet transient recorder. The tests were conducted over a temperature range of 400-900C at 100oC intervals for various true strain rates ranging from 0.001 to 100s-1. In each test, the specimen was compressed to about half of its original height and the load-displacement data was obtained. These were converted into true stress-true plastic strain curves using standard equations. The flow stress data as functions of temperature, strain rate and strain were obtained from these curves and used to construct the power dissipation maps. Similarly, the instability parameter was plotted as a function of temperature and strain rate to get the instability map which was superimposed on the power dissipation map to get the processing map for the materials investigated. The deformed specimens were air cooled and examined using standard metallographic techniques. Tensile testing was carried out on the sample in the DRX domain for alpha iron using standard specimen dimensions to validate the results exhibited by the processing map. RESULTS AND DISCUSSION Alpha iron Typical true stress-true strain curves recorded at 500C and 800C at different strain rates for alpha iron are shown in Fig. 1(a) and Fig. 1(b), respectively. At 500C, the material shows strain hardening at all strain rates. At 800C, the curves corresponding to strain rates of 0.001 and 0.01 s-1 showed steady-state behavior while at higher strain rates, strain hardening is observed. The variation of flow stress () with temperature (T), strain rate ( ) and strain () for alpha iron is shown in Table-2. The processing map obtained at a strain of 0.5 for alpha iron is shown in Fig. 2. The maps obtained at other strains are similar indicting that strain effect is not significant. The processing map of alpha iron exhibit two domains. The first domain occurs in the temperature range 600-850C and strain rate range 0.001 – 0.1s-1 with a peak efficiency of 35% at 800C. This may be interpreted to represent the process of DRX on the basis of the shapes of the stress strain curves in the domain and 1-3 following observations on the variations of grain size, ductility and efficiency of power dissipation with temperature. Typical microstructure of -iron specimen deformed at 0.001s-1 & 800C (corresponding to the peak in the DRX domain) is shown in Fig. 3(a). The variation of the average grain diameter with temperature at a strain rate of 0.001s-1 is shown in Fig. 4(b). The grain size increases with temperature up to the temperature for the peak efficiency (800C) beyond which, there is an abnormal grain growth. The measured tensile ductility values at a strain rate of 0.001 s-1 at different temperatures from the present investigation are plotted in the Fig 4(a). The ductility data of Robbins et al.(7) at a strain rate of 0.5 s -1 is also shown. Both the profiles show a ductility peak at 800C, which matches with the temperature for peak efficiency in the DRX domain [Fig. 4(c)]. As the grain size increases, ductility drops sharply beyond 800C. All these observations are typical of DRX (13,14) and confirm that DRX occurs in this domain. The lower efficiency value of DRX may be attributed to the magnetic domain structures in alpha iron. The grain boundary migration which is essential for dynamic recrystallization is slowed down by the presence of magnetic domains below the Curie temperature (-770 C). The migrating boundary has to overcome the strong electron spins in the direction of magnetization. The efficiency of power dissipation is lower as some energy is spent to reorient the magnetic domains by the migrating grains(15). This is also the reason for higher activation energy observed for self diffusion of iron in ferrite region(10). As the temperature is increased beyond the Curie temperature, the grain boundaries can migrate uninhibited by magnetic domain structure giving rise to abnormal grain growth, lowering the ductility and strength. The occurrence of DRX in alpha iron was also reported by Glover et al.(8) on the basis of microstructural study and kinetic analysis of hot torsion data. The second domain occurs in the temperature range 400 –450C and strain rate range 0.001 – 0.01s-1 with a peak efficiency of 27% at 400oC and 0.001s-1, exhibits dynamic recovery(14). In view of the BCC structure of -iron a large number of slip systems are available which promote easy cross-slip of screw dislocations causing DRY. Thus the lower temperature domain may be interpreted to represent dynamic recovery of -iron. The processing map of alpha iron also exhibits flow instability regime in the temperature range 400-700C when the strain rate is above 10s-1. Microstructural examination of the specimen deformed at 500C and 100s-1 [Fig. 3(b)] showed that alpha iron exhibits flow localization in this regime which should be avoided in processing. Fe-5Si Alloy: The processing map for Fe-5Si alloy is shown in Fig.5 at a strain of 0.5. The processing map for this alloy is very similar to that for alpha iron (Fig. 2). The domain occurring 1-4 with a peak efficiency of 56% at 800oC/0.001s-1 may be interpreted to represent dynamic re-crystallization(14). In comparison with alpha iron, the peak DRX efficiency is higher by about 20%. The DRX efficiency in Fe-5Si has increased because the Curie temperature is lowered by Si addition. For a 10% addition of Si to iron, the Curie temperature decreases from 770C to 600C(12). The decrease in the magnetization of the Fe-Si alloys will enhance the grain boundary migration compared to alpha iron. The instability regime in this alloy is also similar to alpha iron which is not good for processing. Fe-5Co Alloy: The processing map for Fe-5Co alloy is shown Fig.6, which exhibits two domains. The first domain in the temperature range 600-900C and strain rate range 0.001-1s-1 with a maximum efficiency of 33% occurring at 900C and 0.01s-1 represents DRX of the alloy(14). It is interesting to note that the DRX temperature corresponding to the peak efficiency has increased to 900C in comparison with that in alpha iron and Fe-5Si (800C) and the strain rate increased from 0.001 to 0.01s-1. Also the maximum efficiency (33%) is similar to that in alpha iron (35%) but lower than in Fe-5Si alloy. These effects may be attributed to the effect of Cobalt additions on the Curie temperature of alpha iron. The Curie temperature of alpha iron increases from about 770C to about 900C with the addition of 10% Co(12). Thus Cobalt additions strengthen the magnetic domains of alpha iron, restrict the grain boundary migration and reduce the efficiency of DRX. Higher temperatures are therefore required to achieve dynamic re-crystallization. The other domain represents dynamic recovery of the alloy. The processing map of Fe5Co alloy also exhibits flow instability in a wider range of temperature and strain rate as compared with alpha iron. All these instability regimes should be avoided in processing. SUMMARY AND CONCLUSIONS The warm working characteristics of alpha iron, Fe-5Si and Fe-5Co alloys were studied in the temperature range 400-900C and strain rate range 0.001-100s-1. On the basis of the flow stress data obtained as a function of temperature and strain rate in compression, processing maps were developed, using Dynamic Materials Modeling technique. The following conclusions are drawn from this investigation. 1. Alpha iron undergoes dynamic re-crystallization in the temperature range600-850 C and strain rate range 0.001-0.1s-1 with a maximum efficiency of 35% occurring at 800 C and 00.1s-1. At these conditions, the ductility reaches a peak value. 1-5 2. The lower than expected efficiency value DRX in alpha iron is attributed to the restrictive effect of magnetic domains to the migration of grain boundaries. 3. Addition of silicon increases and that of cobalt decreases the efficiency of power dissipation for DRX in alpha iron. This result is attributed to the lowering and increasing the Curie temperature by Silicon and Cobalt additions, respectively. 4. Alpha iron exhibits adiabatic shear bands in the temperature range 400700 C when the strain rate is above 10s-1. This instability regime becomes wider in Fe-5Si and Fe-5Co alloys and should be avoided in warm working of these materials. REFERENCES 1. J.H. Reynolds and D.J. Naylor, “Microstructure and properties of warm worked medium carbon steels”, Mater. Sci. Tech., 1988, 4, 586-602. 2. G.S.Avadhani, “Optimization of process parameters for the manufacturing of rocket casings: A study using processing maps”, JMEPEG, 2003,12(6), 609- 622. 3. Y.V.R.K. Prasad, H.L. Gegel, S.M. Doraivelu, J.C. Malas, J.T. Morgan, K.A. Lark and D.R. Barker, “Modeling of dynamic materials behavior in hot deformation: Forging of Ti-6242”, Metall. Trans., 1984, 15A, 18831892. 4. Y.V.R.K. Prasad and S. Sasidhara, Eds. Hot Working Guide – a compendium of Processing Maps, ASM Intl. Materials Park, Ohio, 1997. 5. Y.V.R.K.Prasad and T.Seshachryulu, “Modeling of hot deformation for microstructural control”, Intl. Mater Rev. 1998, 43(6), 243-258. 6. R.A. Reynolds and W.J.McG Tegart, “The deformation of some pure irons by high – speed torsion over the temperature range 700-1250 oC”, J.Iron and Steel Inst., 1962, 200, 1044-1059. 7. J.L. Robbins, O.C. Shephard and O.D. Sherby, “Role of crystal structure on the ductility of pure iron at elevated temperature”, J. Iron and Steel Inst., 1961, 199, 175-180. 8. G. Glover and C.M. Sellers, “Recovery and re-crystallization during high temperature deformation of alpha iron”, Metall. Trans., 1973, 4, 765-775. 9. J.L. Uvira and J.J. Jonas, Trans. Metall. Soc. “Hot compression of Armco iron and silicon steel”, AIME, 1968, 242, 1619-1626. 1-6 10. S. Karashima, H. Oikawa and T. Watanable, “Creep deformation of iron and its alloys in ferro-and paramagnetic temperature regions”, Acta Met., 1966, 14, 791-792. 11. Yoshinori Murata and Masahiko Morinaga, “Re-crystallization behavior of pure iron at curie temperature”, Scripta Mater.2000, 43, 509-513. 12. O. Kubaschewski : IRON-Binary Phase diagrams, Springer-Verlag, Berlin, 1982. 13. G.S.Avadhani, “Hot Deformation Mechanisms and Microstructural Evolution during upset forging of: -Fe, Fe-5Ni, Fe-5Co & Fe-5Mo alloys and Maraging Steel” Ph.D. Thesis, Indian Institute of Science, Bangalore, India, 2001. 14. G.S.Avadhani, “Warm Working Behavior of Alpha Iron, Fe-Si, Fe-Co and Fe-Ni Alloys: A Study using Processing Maps”, .M.Sc.(Engg.) Thesis, Indian Institute of Science, Bangalore, India, 1996. 15. Alan Cottrell: An Introduction to Metallurgy p. 498, ELBS, UK, 1982. LIST OF TABLES Table 1 Chemical composition of armco iron, Fe-5Si and Fe-Sco alloys and their initial grain size Table 2 Flow stress values (in MPa) of Armco iron at different temperatures and strain rates for various strains (corrected for adiabatic temperature rise). 1-7 LIST OF FIGURES Fig. 1 True Stress-True plastic strain curves for alpha iron obtained in compression at (a) 500C and (b) 800C at different strain rates. Fig. 2 Processing map obtained for Armco iron at a strain of 0.3. Contour numbers represent percent efficiency of power dissipation. Shaded region corresponds to flow instability. Fig. 3 Microstructures of alpha iron in the DRX domain deformed at: (a) 800C/0.001 s-1 and (b) 500C/100s-1 in the instability regime showing localized flow Fig. 4 (a) Ductility, (b) Grain size and (c) Efficiency of power dissipation vs Temperature profile in the DRX domain, obtained for alpha iron. Fig. 5 Processing map obtained for Fe-5Si alloy at a strain of 0.3. Contour numbers represent percent efficiency of power dissipation. Shaded region corresponds to flow instability. Fig. 6 Processing map obtained for Fe-5Co alloy at a strain of 0.3. Contour numbers represent percent efficiency of power dissipation. Shaded region corresponds to flow instability. Table 1: Chemical composition (Wt.%) and initial grain size of the materials used Material C Mn S P Si Co Fe Avg.Grain Dia.(m) Armco iron 0.007 <0.03 <0.005 0.003 -- -- Bal. 118 Fe-5Si 0.007 <0.03 <0.005 0.003 5 -- Bal. 140 Fe-5Co 0.007 <0.03 <0.005 0.003 -- 5 Bal. 125 Table 2: Flow stress values (in MPa) of Armco iron at different temperatures and strain rates for various strains (corrected for adiabatic temperature rise). 1-8 Strain 0.1 0.2 0.3 0.4 0.5 Temperature, oC Strain rate, s-1 400 500 600 700 800 900 0.001 0.010 0.100 1.000 10.00 100.0 261.8 322.3 452.0 475.4 387.1 386.0 143.8 185.7 256.9 336.3 361.7 377.4 86.1 92.1 144.8 206.0 285.8 316.7 48.6 64.1 88.9 121.2 176.9 226.5 22.8 34.6 50.9 71.8 102.2 135.7 20.1 25.0 36.9 52.5 72.9 88.5 0.001 0.010 0.100 1.000 10.00 100.0 299.0 368.3 507.1 536.2 482.0 452.4 166.1 209.4 282.7 376.9 415.6 438.1 97.0 112.2 164.3 225.9 319.4 364.5 54.2 73.1 100.1 135.8 196.6 261.9 24.4 37.5 58.0 82.9 115.5 163.6 21.3 28.2 41.6 60.9 83.1 108.7 0.001 0.010 0.100 1.000 10.00 100.0 326.5 394.7 537.7 581.8 525.0 526.7 185.3 224.1 299.0 395.5 437.3 482.6 103.6 121.5 176.4 238.2 332.7 389.6 54.9 77.6 109.2 147.9 205.7 285.9 24.8 39.1 60.5 90.0 126.9 177.6 21.9 29.5 43.5 66.4 92.3 120.7 0.001 0.010 0.100 1.000 10.00 100.0 334.3 414.1 556.0 604.7 563.2 572.1 200.0 235.4 305.9 400.8 447.3 502.6 106.4 128.1 185.7 244.8 335.4 403.9 56.5 80.4 115.6 156.5 214.2 295.4 24.1 40.2 63.3 94.3 133.6 185.7 22.4 30.3 44.3 67.6 95.7 128.0 0.001 0.010 0.100 1.000 10.00 100.0 345.6 425.4 582.2 622.3 611.7 628.1 209.4 244.0 317.1 404.8 453.8 516.3 111.0 131.0 193.0 247.6 335.5 395.1 57.9 80.2 120.8 161.4 217.9 290.4 23.6 40.3 63.1 95.9 137.2 176.6 22.7 30.7 44.6 68.8 100.4 129.6 1-9 Fig.1 (a) Fig.1 (b) Fig.3 (a) Fig.2 Fig.3 (b) 1 - 10 Fig. 5 Fig.6 1 - 11