Solidification of Thin Wall Ductile Iron Castings with Hypereutectic Composition

Transcription

1 , pp Solidification of Thin Wall Ductile Iron Castings with Hypereutectic Coposition Marcin GÓRNY AGH University of Science and Technology Chair of Cast Alloys and Coposites Engineering, , Reyonta 23, Krakow, Poland. E-ail: (Received on Noveber 24, 2009; accepted on March 30, 2010) Nuerical calculations are presented describing the solidification of a thin wall ductile iron castings with a hypereutectic coposition. Nuerical odel was ipleented in Matlab Siulink environent. The odel takes into account the presence of off-eutectic austenite as well as priary graphite. Experiental verification was ade using casting with the shape of Archiedes spiral. Theral analysis showed that there is high teperature drop of liquid etal due to intensive heat transfer between the flowing etal strea the ould aterial. Theral analysis along with icrostructure observations were ade and show reasonable agreeent of nuerical calculations with experiental easureents. KEY WORDS: ductile iron; thin wall castings; theral analysis; nuerical siulation. 1. Introduction Ductile iron characterize high sensitivity on cooling rate what in consequence leads to structural gradients. 1,2) As a result there are continuous changes of structural features of cast iron with changes of its properties accordingly. It has been proved 3,4) that it is possible to produce thin wall ductile iron (TWDI) with wall thickness even below 3 (without chills, cold laps and isruns). TWDI can be lighter than their substitute ade of aluinu alloys 5 7) and characterized siilar or better echanical properties, definitely better duping capacity. Fro an econoics point of view costs involved in producing ductile iron is uch lower than the ones corresponding to Al alloys. 5,6) All the technological aspects involved in the production of thin wall ductile iron castings, should have been worked out before considering the developent aluinu alloys castings as cast iron substitutes. Nuerous studies have been published on thin wall ductile iron, particularly on the solidification orphologies, 8) icrostructure characterization, 3,9,10) echanical properties, 11,12) carbide foration factors, 1,13,14) production, 15,16) ould filling 17,18) and theral analysis. 19) Moreover, various experiental relationships have been developed between the cheical coposition, 10,14,16) pouring teperature, 14) spheroidization and inoculation practice, 14,15) casting geoetry, 20) plate thickness, 2,10,14) and ould aterials. 21) Yet, ost of these works are liited to siple plate shaped castings. Casting with the shape of Archiedes spiral can be used to analyze technological features of ductile iron as well as the kinetics of solidification. In TWDI castings the first stage of etal cooling is of great iportance. This stage ebraces etal cooling fro pouring teperature to the onset of the solidification process. Pouring teperature (in ould represented by initial teperature of liquid etal) and also its further drop as a result of intensive heat transfer ould-flooding etal are responsible for gradient structure, exhibited by variations in graphite nodule count, ferrite and ceentite fractions in a cross section of a casting. Siulation of solidification of ductile iron can be helpful for understanding the echanis of gradient structure foration in TWDI. The ai of this work is to perfor nuerical siulation of TWDI with hypereutectic coposition and its experiental verification. 2. Experiental A odeling lay-out was designed for thin wall castings. Modeling lay-out, which is shown in Fig. 1(a), consist of gating syste and Archiedes spirals with 1.5 length and , and sections, respectively. Coon gating syste enabled siultaneously filling spiral cavities with different wall thickness. Ductile iron eployed in the present work was produced in a ediu-frequency induction furnace of 15 kg capacity. The raw aterials were Soreletal, coercially pure silicon, and steel scrap. The etal was preheated at C and then poured into the ould. Mould was ade of cheically bounded silica sand. Spheroidization and inoculation processes were ade in the ould, which was equipped with a reaction chaber containing a ixture of 0.85% spheroidizer (44 48% Si, 5 6% Mg, % La, % Al, % Ca) and 0.5% of inoculant (73 78% Si, % Ca, % Ba, % Al) connected to a ixing basin. In addition, post-inoculation occur in the ixing basin by introducing 0.1% of inoculant. The role of the ixing basin is to ensure that co ISIJ

2 ISIJ International, Vol. 50 (2010), No. 6 Fig. 1. a) Casting lay-out (spiral diensions in ), b) photographs of Archiedes spirals with different wall thickness. Fig. 2. Microstructure of ductile iron in spiral with wall thickness of 3 at different distance fro the beginning of spiral: a) 0.01, b) 0.1, c) 0.2, d) 0.3, e) 0.4, f) 0.5, g) 0.6. are presented icrophotographs of structure as a function of spiral length. Results of etallographic experients along with results fro theral analysis are suarized in Table 1. Fro analysis of the graphite distribution (an exaple of typical histogra is shown in Fig. 3) result that biodal histogra represents graphite orphology. Fro etallographic analysis estiation of average radii of eutectic (Re) and priary (large) graphite (Rg) were ade (see Table 1). The presence of large nodules indicates that they have nucleated before the eutectic part of the solidification and by that had longer tie to grow. As the growth is controlled by diffusion, a higher cooling rate will require a higher nuber of priary graphite nodules. This eans that priary graphite nodule count increases as distance fro the inlet increases (higher cooling rate). As a result of turbulent fluid flow of liquid etal larger nodules can be dragged by etal strea and in consequence nuber of priary graphite nodules can increase ore pronounced with increasing distance fro the inlet to the ould cavity. The ain group of the nodules has nucleated during the eutectic part of the solidification. As the distance fro the inlet increases initial teperature of etal decreases (see Table 1). As a consequence cooling rate (near eutectic equilibriu teperature) and axiu undercooling at the onset of eutectic solidification increases, cause increase in graphite nodule count (Ng and Ne). Suing up fluid flow plete ixing of the liquid iron occurs after dissolution of the agnesiu and inoculant alloys. Just after filling the ixing basin, a graphite plug is reoved to enable etal flow into the ould cavity reproducing Archiedes spirals with 1.5 length and different wall thickness (Fig. 1(a)). The cheical coposition of the produced ductile irons was 3.60% C; 3.10% Si; 0.03% Mn; 0.025% P; 0.01% S; and 0.039% Mg. Teperature of etal in ould cavity was estiated using unsheathed theroeleents wires in regular 0.1 distances. Flooding etal strea in old cavity closes circuit of unsheathed theroeleents wires (K type) with thickness of 0.2 connected to the digital data acquisition syste (AGILENT A). In Fig. 1(b) it is shown castings of Archiedes spirals with wall thickness of 0.001, and Castability of tested cast iron for wall thickness 3 was Microstructure Characterization of graphite orphology and atrix icrostructure was perfored on cross sections at different distance fro the beginning of the spiral. The graphite orphology was characterized using iage analysis software Leica QWin (v 3.5.0). The two diensional spatial size distribution of nodules was converted to a three diensional size distribution using Wiencek22) equation. In Fig. 2 there 2010 ISIJ 848

3 Table 1. Results fro etallographic exainations and theral analysis. below liquidus equilibriu teperature of graphite. This teperature is given by Eq. (2) in the for 23) : C Si 0.33P Tg...(2) The odel describing the growth of graphite in liquid is given in work. 23) This odel is based on the assuption that the diffusion area is a space liited by two concentric spheres. The inner sphere coincides with the surface of a sphere of radius R g (t) and the outer radius aounts R l. Growth rate equation of graphite spheres in a liquid has the for of: can affect nodule count distribution (Fig. 3) especially by increasing second group of nodules that is priary graphite. Microstructure is of pearlitc ferritic atrix, free fro chills. Near inlet there is alost ferritic atrix as a result of theral heating fro down-gate. As distance fro inlet increases ferrite fraction ( f f ) decreases (see Table 1). Fro icrostructure observations there are seen oval shaped spaces surrounded by graphite nodules. They are believed to be reaining fro austenite dendrites. 4. Nuerical Siulation Nuerical siulation uses well known heat balance equation it the for of: k Fig. 3. ρ c πt Histogra of graphite nodule count. ( T T A c dt o) vρ vρδh...(1) where: T, teperature; T o, initial teperature of the ould; t, tie; v, volue of the casting; A, surface area of the casting; r, density of the etal; c, specific heat of etal; DH f, latent heat; df s /, evolution of solid fraction; k, theral conductivity of the ould aterial; r, density of the ould aterial; c, specific heat of the ould aterial. The process of ductile iron solidification with the hypereutectic coposition is divided into the following stages: Stage I: The solidification of priary graphite. Stage II: Eutectic solidification. Stage III: The solidification of austenite in the for of dendrites. Stage I Solidification in ductile iron with hypereutectic coposition (carbon equivalent 4.26) start with nucleation and growth of graphite spheres in the liquid after undercooling f df s drg 1 2D C o C e...(3) 2 t ρ ( ) g Where: D, diffusion coefficient of carbon in liquid; and r g, density of graphite. After integration of Eq. (3) radius of the graphite is given by:...(4) Using the relationship between the degree of undercooling (DT) and the concentration difference (DC C o C e ) in the for (Fig. 4) DC DT/ g we have: and: R g D T Rg 2 Δ ρ g g dr 2D C o C e t ρ ( ) g 2D g g g...(5)...(6) In a odel, it is assued instantaneous nucleation. 27) Evolution of graphite fraction can be expressed by Eq. (7): dfg NR dr g g 2 g 4π...(7) where: N g, is the nodule count (of priary graphite). Evolution of graphite fraction covers the period fro undercooling below equilibriu teperature for graphite liquidus until the teperature of eutectic transforation. In Stage I the concentration of carbon in the liquid varies fro the value of C o (initial carbon content) up to C e (carbon ρ 2 t t ISIJ

4 ipingeent. This is called the correction factor due to slowdown in the growth ipact of growing grain. 26) content in graphite eutectic). Stage II This stage involves nucleation and growth in ters of undercooling below eutectic equilibriu teperature (T e ) 24) : T e Si 14.88P...(8) Stage II will be divided into two periods. Period 1 After reaching the eutectic coposition, just below the eutectic equilibriu teperature there is nucleation of austenite envelope on a priary graphite nodules and start the growth of globular eutectic on priary graphite. Growth of austenite envelope is given by the equation 25) : Re ( C3 C2)...(9) ( R R ) R D Where: C 2, C 3 : carbon content in austenite, at austenite/ graphite and austenite/liquid interface respectively, C 4, C gr : carbon content in bulk liquid and graphite, R g, R e : radii of austenite envelope and eutectic graphite, respectively. Fro the ass balance in the volue of (4/3)pR 3 g the radius of graphite can be calculated fro 25) : R 3 ( C4 C3) Rg 3 ( C4 Cgr ) Re 3 ( C2 Cgr )...(10) Concentrations C 2, C 3, and C 4 can be expressed as a function of undercooling. Assuing that the JE, E S and BC lines for the Fe C syste are straight, the copositions in Eq. (10) can be given by C 2 Fig. 4. C, C3 C, C4 Ce 2 Fragent of the Fe C equilibriu syste. ( C C ) dr (11) where: C g, C e, they are respectively the concentration of carbon in the points E and C of Fe C Si syste; 2, 3, 4, coefficients of directional lines, respectively E S, JE and BC in the Fe C Si syste. Evolution of austenite fraction can be given by Eq. (12): df NR dr 2 4π e ( 1 fs)...(12) where: N e, is the nodule count (eutectic graphite nodules). Equation (12) includes (1 f s ) ter to account for grain 3 e 4 Period 2 Period 2 in Stage II involves nucleation and growth of graphite eutectic nodules and nucleation and growth of austenite envelopes. Nucleus of eutectic graphite growth freely in liquid up to the diension of the R o. In this period their growth is calculated using a siilar procedure as for the priary graphite nodules taking into account undercooling below the extrapolated liquidus for graphite. Once the graphite is the radius aounted r o there is diffusion controlled growth of graphite though austenite shell using Eqs. (9) and (10). Austenite envelope growth is calculated using an Eq. (12). Stage III Stage III involves nucleation and growth of austenite dendrites. This phase occurs after undercooling below extrapolated liquidus for austenite (T g ). It is assued instantaneous nucleation of austenite. Austenite liquidus is given by 24) : T ( C 0.25Si 0.5P)...(13) In this paper, the growth of equiaxed austenite dendrites will be described by the relationship 27) : where...(14)...(15) R d, radius of austenite dendrite; D L, diffusion of carbon in austenite; G, Gibbs Thopson paraeter;, liquidus slope lines for austenite in Fe C syste; C L, carbon concentration in liquid. Evolution of austenite dendrite fraction can be given by Eq. (16): dfd dr d μ T Δ 2 DL μ 2π 2 Γ k ( 1) C 4π gnr dr d d d 2 d ( 1 fs )...(16) Where: N d, nuber of austenite dendrites; dr d /, growth rate of austenite dendrites. Equation (16) includes (1 f s ) to account for grain ipingeent. Spheres 28) are not copletely filled by the network-type dendrites. It is assued that internal fraction of solid aounts g d Matlab-Syulink TM (version R2009a) was used for nuerical calculations of the solidification of a spiral-shaped TWDI casting with a wall thickness of 3. With access to high-perforance coputing algoriths and echaniss for analysis of Matlab-Siulink enable quickly and efficiently carry out coplex calculations. Here are the ethod of nuerical solving of differential equations and linear integration, differentiation, interpolation and approxiation of functions and any others. Nuerical calculations were perfored on the basis on data taken fro experients that are the initial teperatures of etal in ould and the results fro etallographic L 2010 ISIJ 850

5 Table 2. Selected values used in siulation. studies (such as the nuber of spheres of graphite). Experiental studies ade it possible to obtain the actual cooling curves of ductile iron. Physical properties used in nuerical odeling are suarized in Table Experiental Results and Discussion In Fig. 5 there are shown cooling curves resulting fro nuerical odelling copared with experiental curves obtained at different distances fro the beginning of the spiral. Results of coputer siulation show fairly good copliance with the experiental results. Both nuerical calculated and experiental curves show recalescence, that is, the teperature difference between the highest and lowest teperatures. However, the predicted recalescence is soewhat higher than deterined by theral analysis. Flowing etal strea through the ould cavity heats it up. In consequence conditions of heat exchange along the flowing path are changing. Increasing distance fro the inlet is accopanied by a shorter contact tie of liquid etal with a ould, which increases the cooling rate. Cooling rate (see Table 1) in turn affects and increase the axiu undercooling at the onset of graphite eutectic. Maxiu undercooling estiated fro theral analysis and fro siulation are graphically shown in Fig. 6. The preheating during filling can has an influence on the teperature easureent so as to reduce DT. Maxiu undercooling obtained fro siulation shows rather good confority with experiental easureents. This is especially iportant because undercooling is the driving force easure for the nucleation stage during solidification. Not all substrates in the undercooled elt play an active role in the nucleation process. The iniu substrate sizes, which becoe active nucleation sites, decrease continually at increasing degrees of undercooling. In consequence nodule count increases. An influence of undercooling on eutectic nodule count estiated by etalographic exainations is show in Fig. 7. Undercooling started fro 48 C (at x 0.01 ) and if distance fro inlet increases it goes up to the value of 65 C. When axiu degree of undercooling increases, below the ceentite eutectic foration teperature, chills can be fored in the structure. Below in Fig. 8 there are shown results of siulation both priary and eutectic graphite radii along with their austenite envelopes. Calculated graphite radii can be copared with experiental. An average eutectic radius for x 0.10 aounts R e 3.42 (see Table 1). Siulation gives radius at the end of solidification aounted R e In case of priary graphite results are as follows: R g (exp.) 9.03 and R g (si.) Siulation show a little lower radii in coparison to experiental results. The differences are connected with the effect of fluid flow, teperature drop and heating of the ould by flowing etal strea. The longer tie of flowing etal strea the lower teperature drop and the lower cooling rate as a result of change in theral paraeters of a ould with teperature. In this connection fluid flow has an effect to the teperature distribution. Mainly it can be anifested by different slope of teperature tie curve before axiu undercooling (see Fig. 5). During fluid flow teperature can decreases below liquidus teperature for graphite. Fro this tie flowing etal strea can have already nucleated priary graphite nodules, which growth in this period is not taken into account in nuerical calculations. Moreover graphite keep growing even after the end of solidification, which is also not included in siulation. Fro these reasons siulated radii are understated. It is worth to note that thin wall castings, which solidify with high cooling rate cause the elt undercooled below extrapolated liquidus line for austenite. As a result nucleation and growth of austenite dendrites takes place. Figure 9 presents undercooling below the extrapolated lines for austenite, which is the driving force easure for the solidification of austenite dendrites. Influence of the solidification of austenite dendrite on cooling curve is pronounced and it is shown in Fig. 10. Fro Fig. 10 follows that nucleation and growth of austenite dendrites have an iportant theral effect. It is visible by change in the slope on the cooling curve and also decreases in both undercooling and recalescence. In odel ISIJ

6 Fig. 5. Coparison of the cooling curves predicted by the odel and the experientally easured cooling curves for the different therocouple position (x). Solid lines-siulation, dotted lines-experiental. Fig. 6. Maxiu undercooling as a function of spiral length. Fig. 7. Eutectic nodule count versus axiu undercooling. ling of the solidification of ductile iron austenite dendrites should be taken into account. Especially in thin wall castings, because reducing wall thickness fraction of austenite dendrites increases. 19) Such solidification behavior anifested by presence both priary graphite and austenite dendrites can be taken into account by nuerical siulation 2010 ISIJ 852

7 paraeters can be observed. It usually applies to nodules count and to a saller extent to the atrix. Fro work 1) results that risers or ultiply inlets can significantly reduce structure inhoogeneity. Fig. 8. Radii of eutectic (R e ) and priary (R g ) graphite, respectively, radii of austenite envelopes for eutectic (R g ) and priary (R gg ) graphite, respectively and solid fraction denoted by f s (results for x 0.10 ). 6. Conclusions (1) It has been adopted the odel describing the solidification of ductile cast iron with hypereutectic coposition in Matlab-Siulink environent. The odel takes into account the presence of off-eutectic austenite as well as priary graphite. Both phases are typical for thin wall ductile iron castings. (2) Experiental verification using casting with the shape of Archiedes spiral was done. Theral analysis along with icrostructure observations showed that cooling curves predicted with the presented odel gives reasonable agreeent with experiental easureents. (3) Theral analysis showed that there is high teperature drop of liquid etal due to intensive heat transfer between flowing etal strea the ould aterial. Teperature drop can have a pronounce effect on structure and in consequence casting properties. REFERENCES Fig. 9. Fig. 10. Siulated undercooling below the extrapolated lines for austenite (x 0.10 ). Modelling with including and excluding the possibility of austenite dendrites nucleation (x 0.10 ). given in this work. Moreover the database of experiental data in the for of theral analysis and the etallographic investigation for different initial teperatures akes odeling of TWDI reliable. Casting with the shape of Archiedes spiral is designed to the easure of fluidity. Aside fro technological aspect in production TWDI (how is the fluidity of ductile iron for a given wall thickness) it represents different cooling conditions as the result of the fact that the etal flowing in the ould channel cavity heated it, and thus changing conditions for the exchange of heat flowing strea the ould aterial. It has a pronounce effect on structure and in consequence on casting properties. Inhoogeneous of structure 1) E. Fras, M. Górny and H. F. Lopez: AFS Trans., 116 (2008), ) M. Górny: Arch. of Foudry Eng., 7 (2007), 73. 3) E. Fras, M. Gorny and H. F. Lopez: AFS Trans., 116 (2008), ) M. Lessiter: Eng. Cast. Solid., 2 (2000), 37. 5) E. Fraś, M. Górny and W. Stachurski: Fou. Rev., 56 (2006), ) S Katz: Fou. Manag. Tech., (1997), 34. 7) M. Górny: Arch. Foudry Eng., 9 (2009), ) C. Yeung, H. Zhao and W. B. Lee: Mater. Charact., 40 (1998), ) O. Dogan, K. Schres and J. Hawk: AFS Trans., 111 (2003), ) R. E. Ruxanda, D. M. Stefanescu and T. S. Piwonka: AFS Trans., 110 (2002) ) F. R. Juretzko, L. P. Dix, R. E. Ruxanda and D. M. Stefanescu: AFS Trans., 112 (2004), ) J. W. Torrance and D. M. Stefanescu: AFS Trans., 112 (2004), ) A. Javaid, J. Thopson, M. Sahoo and K. G. Davis: AFS Trans., 107 (1999), ) A. Javaid, J. Thopson and K. G. Davis: AFS Trans., 110 (2002), ) C. Labrecque and M. Gagne: Int. J. Cast Met. Res., 16 (2003), ) J. F. Cuttino, J. R. Andrews and T. S. Piwonka: AFS Trans., 107 (1999), ) F. Mapaey and Z. A. Xu: AFS Trans., 105 (1997), ) M. Górny: Arch. Foudry Eng., 9 (2009), ) K. M. Pedersen and N. Tiedje: Mater. Sci. Foru, 508 (2006), ) D. M. Stefanescu, R. E. Ruxanda and L. P. Dix: Int. J. Cast Met. Res., 16 (2003), ) R. E. Showan, R. C. Aufderheide and N. Yeoas: Mod. Cast., (2006), ) K. Wiencek and J. Rys: J. Mater. Eng., 3 (1998), ) E. Fraś: Solidification of Cast Iron, AGH, Cracow, (1981), ) E. Fraś: Theoretical Basis for Crystallization, AGH, Cracow, (1986), ) E. Fraś: Solidification of Metal and Alloys, PWN, Warsaw, (1992), ) A.N. Kolgoorov: Bulletin de l Acadeie des Sciences de l URSS, 3 (1937), ) D. M. Stefanescu: Science and Engineering of Casting Solidification, 2nd ed., Springer, NY, (2009), ) I. Dustin and W. Kurz: Z. Metallkd., 77 (1986), ISIJ