Microstructural Features and Erosion Wear Resistance of Friction Stir SurfaceHardened Spheroidal Graphite Cast Iron

Tun-Wen Cheng, Truan-Sheng Lui+ and Li-Hui Chen

Department of Materials Science and Engineering, National Cheng Kung University,No. 1 University Road, Tainan 701, Taiwan, R. O. China

A ferritic spheroidal graphite cast iron (also named as SG cast iron, ductile cast iron, ductile iron) was treated with friction stir process(FSP) to harden the surface layer owing to a unique microstructure into which the ferritic structure transforms after high temperature deformationand subsequent direct cooling. When the friction stirred surface experiences thermomechanical cycle during FSP (here named friction stir surfacehardening, FSSH), a non-traditional bainite structure can be obtained through subsequent cooling process. The bainite structure primarilyconsists of iron carbide (Fe3C), acicular ferrite and martensite with retained austenite aggregates. It is evident that the FSSH structure caused bydeformation at austenite temperature has resulted in a significant increase in the microhardness of about 1000HVyielding a primarily martensiticaccompanying bainitic phase transformation. The experimental results also show that the process has resulted in significant improvement inerosion resistance at low angle impingement than that of ferritic specimens. In addition, the maximum erosion rate of ferritic specimens occurs at20­25°of impact while the peak of the FSSH specimen shifts to higher angle resulting from the formation of continuously cooled martensitic andbainitic structure. [doi:10.2320/matertrans.M2011224]

(Received July 27, 2011; Accepted October 25, 2011; Published December 7, 2011)

Keywords: spheroidal graphite cast iron, friction stir process, surface hardening, erosion wear, thermomechanical

1. Introduction

Spheroidal graphite cast iron (also known as SG cast iron,ductile cast iron, ductile iron) is a common engineeringmaterial, because of its excellent castability, tensile ductilityand thermal properties. In addition, the mechanical propertiesof SG cast iron can be varied by performing post heattreatment to obtain a suitable matrix microstructure accordingto the desired mechanical properties. However, a wide rangeof strength and ductility combinations can be achieved bysubsequent thermal processing. On the other hand, ferritic SGcast iron suffers from particle erosion when it is used in manyapplications. In order to increase its erosion resistance,various surface modification processes such as high fre-quency induction,1) laser beam,2) plasma transfer arc(PTA),3,4) etc., have been adopted to produce a hardenedsurface layer. The formation of a large account of carbides oreutectic cementite during subsequent rapid solidification isthe major cause of the hardening. For the improvement ofwear resistance, Rigsbee and coworkers5) have shown thata laser surface remelted layer can improve the erosionresistance of SG cast iron. However, this experimental resultwas based only on certain oblique impact conditions. Incontrast, our previous investigations4) which used PTA forsurface remelting at a normal impact angle yielded adiametrically opposite result due to the existence of graphitenodules in the surface hardening layer. However, thisconclusion implies that the microstructural modification ofthe surface layer under an identical chemical composition ofhigh ductility ferritic SG cast iron can actually affect theerosion behavior.

In the present study, friction stir process (FSP) which refersto friction stir surface hardening (FSSH) here was used forthe surface hardening of ferritic spheroidal graphite cast iron(SG cast iron). Compared to the abovementioned surface

hardening method, though surface remelting and resolidifi-cation do not occurr in this process, the surface hardeningeffect can still be obtained and possess the flexibility of beingperformed on local desired positions. However, a uniquemicrostructure can be achieved by this process due to thefrictional heat input and severe deformation. The effect ofthese microstructural features on erosion resistance under anoblique angle was clarified in this investigation.

FSP is a spinoff of friction stir welding6) (FSW)technology, and the microstructural features are affected bythe frictional heat and severe plastic deformation to acquire asurface hardening layer. As for ferrous materials, few studieshave investigated the FSW of interstitial free steel7) (IF steel:20 ppmC), carbon steel,8) and stainless steel.9) They reportedthat the FSW caused grain refinement in the stir zone (SZ) ofthe steel. Complex phase transformation also occurred duringFSW due to the various welding parameters affecting thepeak temperature which are above or below A3 temperature.

There are very few studies with respect to FSP performedon cast iron. Fujii et al.10) reported the possibility ofmartensitic transformation emerging in FCD700 (ductilecast iron) and FC300 (gray cast iron) after FSP. To gaina further understanding, this study aims at clarifying themicrostructural features introduced by FSP pertaining to theimprovement in erosion wear resistance.

2. Experimental Procedures

2.1 Material preparationFerritic SG cast iron was selected as the experimental

material. Its chemical composition was Fe­2.88C­2.47Si­0.43Mn­0.029P­0.018S­0.035Mg and it was melted in ahigh frequency induction furnace, and spheroidized withFe­45Si­3Mg nodulizer. Then the melt was inoculated withFe­75mass% ferrosilicon. The molten iron was poured andcast in a Y-shape mould to obtain a Y block. All materialswere first ferritized by holding at 930°C for 5 h and furnace+Corresponding author, E-mail: z7408020@email.ncku.edu.tw

Materials Transactions, Vol. 53, No. 1 (2012) pp. 167 to 172©2011 The Japan Institute of Metals

cooled to 730°C for another 3 h, followed by furnace coolingdown to room temperature. The full ferritic structure isshown in Fig. 1(a) with tensile elongation of 20% and thecharacteristics of the ferritic specimens are shown in Table 1.

For FSP surface hardening (hereafter referred to as frictionstir surface hardening, FSSH), the processing tool was ofdiameter ¤15mm, with no pin and was made of tungstenmaterial with a tilt angle of 3° under load control at a rotation

speed of 2200 rpm and traveling speed of 1mm·s¹1. AfterFSSH, the microstructure of the surface layer was determinedquantitatively by scanning electron microscopy (SEM).Electron Probe X-ray Micro Analyzer (EPMA) wasperformed to identify the distribution of various elements,and XRD analysis with copper X-ray tube and step size of0.001 degrees was used to identify each existing phase from20­100°. The microhardness profile from the surface was

(a) (b)

(c)

(e)

(d)

Fig. 1 Morphologies of ductile iron: (a) base metal (BM); (b) macrostructure of the FSSH specimen; (c) TMAZ1 of the FSSH specimen;(d) HAZ1 of the FSSH specimen; (e) Hard eye structure of HAZ2 region where ¡¤: martensite, ¡b: bainitic or acicular ferrite and £:austenite.

T.-W. Cheng, T.-S. Lui and L.-H. Chen168

measured through a cross section using a micro Vickerstesting machine.

2.2 Erosion testThe erosion specimens were cut from FSSH samples

which were 50mm © 4mm wide and 4mm thick rectangularplates with a side containing a FSSH layer were polishedwith 600-grit SiC sand paper before the erosion test. Forcomparison, the ferritic SG cast iron specimens as the basematerial (BM) were also tested. The erosion test wasconducted at room temperature using a sand blast typeerosion test rig shown schematically in Fig. 2. The obliqueimpact angle ª as indicated in Fig. 2 was varied from 15 to45°. The SEM micrograph of the SiO2 erodent is shown inFig. 3. The particles were ejected by compressed air flow,impacting the specimens at about 16.5 cm/s (estimated by asingle-shot high-speed photography).

After the erosion test, a microbalance with 0.01mgaccuracy was used to measure the weight loss, and allspecimens were ultrasonically cleaned in acetone beforeweighing. The average erosion rate was taken as the weight

loss per 300 g weight of the impacted erodent and performedon at least three specimens.

3. Experimental Results

3.1 Microstructural features and the microhardnessprofile of the friction stir surface hardening layer

Friction stir surface hardening (FSSH) was adopted todenote the hardening surface layer on ferritic SG cast iron.Meanwhile, the frictional heat input of any given hardeninglayer can be specified under a constant rotation speed andmoving rate of the friction tool in this investigation.However, the surface hardening layer possesses a uniquemacroscopic appearance. The macrostructure of FSSHspecimens as shown in Fig. 1(b) can be defined as 4 differentregions respectively. (1) The top surface layer which containsdeformed graphite nodules due to high temperature stirringflow is defined as TMAZ, as shown in Fig. 1(c). (2) Themicrostructure contains acicular phase [Fig. 1(d)] is definedas HAZ1 where the matrix structure is similar to TMAZ. Inthis region, the graphite nodules remain unchanged. (3) Theregion which contains chunk-like phases surrounding thegraphite nodules also known as a hard eye structure as shownin Fig. 1(e), is defined as HAZ2. And (4) the base materialregion contains a fully ferritic structure.

Figure 4 shows the microhardness profile of the FSSHspecimen which is measured from the cross section verticalto the processing direction. From the microhardness profileshown in Fig. 4(a), the TMAZ region defined as (1) layer

Table 1 Characteristics of experimental ductile irons.

Mean particle size of spheroidal graphite (µm) 28

Mean grain size of ferrite (µm) 50

Volume fraction of spheroidal graphite (%) 10.5

Hardness of ferritic matrix (HV) 215

Nodular fraction (%) 85

Nodular counts (No./mm2) 260

Fig. 2 Schematic illustration of the erosion test.

Fig. 3 Morphology of the SiO2 erodent.

(a)

(b)

Fig. 4 Microhardness profile of the FSSH specimen: (a) cross section;(b) from the surface.

Microstructural Features and Erosion Wear Resistance of Friction Stir Surface Hardened Spheroidal Graphite Cast Iron 169

is significantly greater than the near regions. Figure 4(b)indicates that TMAZ has greater hardness than that of HAZ1and HAZ2. In addition the (2) layer containing an acicularstructure also shows an extremely high hardness value ofover 800HV. And the hardness decreases progressively awayfrom the friction surface.

To understand the evolution of phase transformationduring FSP, Fig. 5(a) is a diffractometer trace to distinguishthe difference between the BM and FSSH specimens. Ferritic

peaks are observed in a fully ferritic specimen while FSSHspecimens possess the peaks of iron carbide (Fe3C) andaustenite due to the thermomechanical effect. Figure 5(b)demonstrates the XRD diffraction pattern (20­30°) of theferritic specimen (BM) on three different planes (ND, RD,and TD) respectively. It shows that (002) graphite peak isinsignificant around 26 degrees. It is suitable to suggest thatthe peak around 26 degrees is the (002) cementite peak. Asa result, this implies that heat and severe deformation willbe induced due to FSSH and can actually cause subsequentphase transformation. From Fig. 6, the mapping resultsconfirm a tendency towards extremely low carbon contentwithin the areas with acicular phase as shown in Fig. 6(a).However, the FSSH matrix shows a mixture of bainiticstructures that primarily consist of bainitic ferrite, martensiteand retained austenite aggregates.

3.2 Erosion rate and morphologies of wear surfaceFigure 7 shows the wear rate of the materials as a function

of angles of impingement. It is noted that the wear rate ofboth specimens varies with the oblique impact angles ofimpingement. The base metal (BM) specimens exhibit amaximum wear rate at an impact angle of 20­25° while theFSSH specimen exhibits a maximum wear rate at an angleof about 30° impact. It is noted that the wear rate of theFSSH specimens is smaller than that of the BM specimensirrespective of these angles of impingement.

Figure 8 shows the wear surface morphologies of bothspecimens, and reveals typical ductile wear mechanisms suchas grooving, lip formation and craters. An oblique impactgives rise to more groovings while a large-angle impactcauses more craters. The FSSH specimen was eroded underoblique impact, and little graphite particles could berecognized, while the BM specimen demonstrates thelocations of graphite nodules, and the cave-in feature canbe recognized [see those marked by “G” in Fig. 8(a)]. TheFSSH specimen which contains a fair amount of bainitic andmartensitic structures is shown in Fig. 8(b). For the FSSH

(a)

(b)

Fig. 5 X-ray diffraction analysis of: (a) BM and FSSH specimen (2ª: 20­100°); (b) the BM specimen on three different planes (ND, RD, and TD)respectively.

(a)

(b)

Fig. 6 EPMA analysis of: (a) TMAZ; (b) HAZ2.

T.-W. Cheng, T.-S. Lui and L.-H. Chen170

specimen, Fig. 9 also illustrates the subsurface features whenthe specimen is eroded at a 30° impact. As depicted, flow ofthe materials along the erosive direction can be distinguishedfrom the bending of graphite in the vicinity of the wearsurface. A smaller grooving size is more frequently observedunder an identical impact angle.

4. Discussion

Regarding the extremely high microhardness value on thefriction stir surface hardening (FSSH) layer, from the Fe­C­Si phase diagram, it is evident that the eutectic point at

1147°C corresponding to 3.6% C refers to the microstructuralfeatures as shown in Fig. 1(c) and it does not show anydendritic phase in the near surface region. Accordingly itshould be noted that the peak temperature during FSSHdid not exceed the eutectic temperature, and in particularthe embedded graphite particles within the hardened layerremained undissolved [Figs. 1(c), 1(d)]. Figure 4(b) showsthe variation of microhardness as a function of depth fromthe surface. It is evident that FSSH can actually acquirea significant increase since the experimental data are fairlyclose to the hardness of regular martensitic phase of asquenched SG cast iron.

In this study, friction stir process (FSP) is conducted as ahot deformation method at austenitizing temperature, but theeffect of FSP parameters such as rotation speed, feeding rateor subsequent cooling rate on microstructural features need tobe clarified, and further detailed examination is still needed.From Figs. 1(c) and 1(d), a refined microstructure of thematrix consisting of acicular ferrite and deformed graphitenodules can be observed. FSSH can actually be employedto refine the acicular structure through the refinement ofaustenite grain size that resulted from the occurrence ofdynamic recrystallization by thermomechanical deformation.High temperature deformation at the austenitizing stage willproduce deformed and/or dynamic recrystallized austenitegrain depending on the temperature, and deformation wasinvolved, resulting in an increased number of nucleation sitesfor acicular ferrite during the subsequent cooling stage. Itshould be noted that though the austenitizing temperaturewithin the stirring zone could not be maintained uniformly,a previous report has mentioned that the deformation ofthis prior austenite will increase the rate of phase trans-formation.11)

During FSSH of the ferritic SG cast iron, the high coolingcan lead to sufficient undercooling that will suppress thecarbon diffusion rate and introduce a thermal mechanicaleffect. The top surface which underwent FSP presents anextremely high hardness due to the directly cooled bainiticand martensitic transformation. Cui et al.12) have reported thethermal cycle of high carbon steel (S70C) during FSP, andconfirmed that the peak temperature of SZ would exceed A3temperature with proper FSP parameters. Friction stirring

Fig. 7 Erosion rate at different impact angles.Fig. 9 Observation of 30° impact subsurface of FSSH specimens where

ED is the direction of particle ejection.

(a)

(b)

Fig. 8 Wear surface morphologies: (a) BM, 30°; (b) FSSH, 30° where EDis the direction of particle ejection.

Microstructural Features and Erosion Wear Resistance of Friction Stir Surface Hardened Spheroidal Graphite Cast Iron 171

causes the work-hardened austenite phase to transform easilythrough a diffusional process on the surface to promote phasetransformation which yields a large amount of hardeningphase since the subsequent cooling rate is significantly high.Therefore the friction stir hardening surface consists ofacicular ferrite or bainite, and a fair amount of high carboncontent austenite phase is retained in the microstructure dueto silicon being a strong graphitizer that retards carbideprecipitation in SG cast iron. The transformation area of HAZis shown in Fig. 6(b) where the carbon content is much greatthan ferrite. From experimental results as shown in Fig. 1(c),rapidly heated austenite was found to have partially trans-formed into martensite in the stage after the retreating flow ofFSSH. From the microhardness profile as shown in Fig. 4(b),extremely high hardness could be acquired as the mixturestructure consisting of bainitic and martensitic structure couldbe recognized in the vicinity of the surface. It should be notedthat a hard eye structure consisting of bainite and retainedaustenite could be recognized in the HAZ2 layer and isprobably correlated with a certain period of rapid heatingaustenization surrounding the graphite nodules due tosubsequent frictional heat input.

From the experimental results as shown in Fig. 7, theFSSH specimen possesses a relatively high wear resistancecompared to BM. In particular, FSSH introduced rapidheating in the surface without melting as shown in Fig. 1(c),and this was followed by quenching to enhance a highhardness layer accompanied by a mixed microstructure[Fig. 4(a)] without changing the chemical composition ofmaterial. Consequently, the wear resistance could beimproved by FSSH predominantly owing to the micro-structural change which is induced by solid-state trans-formation of austenite to martensite phase or nontraditionalaustempering bainite phase.

From wear surface morphologies of the TMAZ layer andferritic SG cast iron (BM), as those illustrated in Fig. 8, allspecimens reveal typical ductile wear mechanisms such asgrooving, lips and craters.

On the other hand, both BM and FSSH specimensdemonstrate that the plastic flow governs the materialremoval rate when the SiO2 particles eroded under obliqueimpact, and no brittle wear phenomena can be observed. Inthis ductile erosion process, it is suitable to suggest thatbainitic and martensitic phases will reinforce the matrix andsuppress the wear loss resulting from the cave-in effectaround graphite nodules. Consequently, it is reasonable tosuggest that FSSH is effective in improving the erosionresistance at oblique impact, and moreover, the improvementis more pronounced with reducing the impact angles asshown in Fig. 7.

Regarding the base material as the ferritic SG cast iron,when it was eroded under all impact angles it showed nobrittle cracking, and possessed the maximum erosion rate at20­25° impact (Fig. 7). This impact angle corresponds to thefrequently observed angle of around 20­30° in the wellknown ductile erosion mechanism for ductile metals.13) It issuitable to suggest that the retained austenite phase will playan important role on wear behavior of the FSSH specimens,and in particular, phase transformation of retained austenitecan be caused by particle erosion.14) However, during erosion

testing, particle impingement can raise the surface temper-ature which may lead to ¾-carbide precipitation that alsoaffects the increasing impact angle of maximum erosion rate.

5. Conclusions

The fully ferritic matrix transforms into a multiphasestructure when FSSH is performed on the ferritic SG castiron. The FSSH specimen consists of 4 layers which areTMAZ, HAZ1, HAZ2 and base metal respectively. Theferritic SG cast iron and TMAZ of the FSSH specimen weregiven an erosion test and the test results are summarizedbelow:

(1) Crashed and elongated graphite is observed in TMAZ.TMAZ and HAZ1 both consist of acicular ferrite, Fe3C, andmartensite with retained austenite aggregates. HAZ2 shows ahard eye structure and contains bainite and retained austenite.

(2) The microhardness decreases from TMAZ to basemetal progressively. A significant increase of microhardnessup to 1000HV can be achieved due to subsequent rapidcooling after FSSH.

(3) FSSH can improve erosion wear resistance when theparticle erosion is under an oblique impact. The remarkableimprovement is due to the presence of bainite and martensitephases which harden the matrix and suppress the cave-ineffect caused by the presence of graphite nodules.

Acknowledgements

The authors would like to thank Prof. Chin Pao Cheng,Department of Mechatronic Technology, NTNU for produc-ing FSSH specimens. And the authors are also grateful to theChinese National Science Council for its financial support tothis work (Contract: NSC 100-2221-E-006-093).

REFERENCES

1) M. N. Yoozbashi, S. Yazdani and T. S. Wang: Mater. Des. 32 (2011)3248­3253.

2) H. Yan, A. H. Wang, Z. T. Xiong, K. D. Xu and Z. W. Huang: Appl.Surf. Sci. 256 (2010) 7001­7009.

3) W. S. Dai, T. S. Lui and L. H. Chen: Int. J. Cast Met. Res. 12 (1999)119­125.

4) W. S. Dai, T. S. Lui and L. H. Chen: Int. J. Cast Met. Res. 12 (1999)233­240.

5) C. H. Chen, C. J. Altstetter and J. M. Rigsbee: Metall. Trans. A 15(1984) 719­728.

6) W. M. Thomas, E. D. Nicholas, J. C. Needham, M. G. Murch, P.Templesmith and C. J. Dawes: GB Patent Application No. 9125978.8(1991).

7) H. Fujii, L. Cui, N. Tsuji, M. Maeda, K. Nakata and K. Nogi: Mater.Sci. Eng. A 429 (2006) 50­57.

8) R. Ueji, H. Fujii, L. Cui, A. Nishioka, K. Kunishige and K. Nogi:Mater. Sci. Eng. A 423 (2006) 324­330.

9) S. H. C. Park, Y. S. Sato, H. Kokawa, K. Okamoto, S. Hirano and M.Inagaki: Scr. Mater. 49 (2003) 1175­1180.

10) H. Fujii, Y. Yamaguchi, S. Kiguchi and K. Nogi: Mater. Trans. 49(2008) 2837­2843.

11) J. Achary and D. Venugopalan: Metall. Trans. A 31A (2000) 2575­2585.

12) L. Cui, H. Fujii, N. Tsuji and K. Nogi: Scr. Mater. 56 (2007) 637­640.13) I. Finnie, J. Wolak and Y. Kabil: J. Mater. 2 (1967) 682­700.14) F. Y. Hung, L. H. Chen and T. S. Lui: Mater. Trans. 45 (2004) 2981­

2986.

T.-W. Cheng, T.-S. Lui and L.-H. Chen172