Bifilm Inclusions in High Alloyed Cast Iron

Home , Draft (engineering), Pattern (casting)

materials

Article Biﬁlm Inclusions in High Alloyed Cast Iron

Marcin Stawarz * and Malwina Dojka

Department of Foundry Engineering, Silesian University of Technology, 7 Towarowa Street, 44-100 Gliwice, Poland; [email protected] * Correspondence: [email protected]; Tel.: +48-32-338-5532

Abstract: Continuous improvement in the quality of castings is especially important since a cast without defects is a more competitive product due to its longer lifecycle and cheaper operation. Producing quality castings requires comprehensive knowledge of their production, crystallization process, and chemical composition. The crystallization of alloyed ductile iron (without the addition of magnesium) with oxide bifilm inclusions is discussed. These inclusions reduce the quality of the castings, but they are a catalyst for the growth of spheroidal graphite that crystallizes in their vicinity. The research was carried out for cast iron with a highly hyper-eutectic composition. Scanning electron microscopy and EDS analysis were used in the research. A detailed analysis of the chemical composition was also carried out based on the spectrometric method, weight method, etc. Based on the obtained results, a model of spheroidal graphite crystallization near bifilm inclusions was proposed. The surface of the analyzed graphite particles was smooth, which suggests a primary crystallization process. The phenomenon of simple graphite and bifilm segregation towards the heat center of the castings was also documented.

Keywords: biﬁlms; spheroidal graphite; alloyed cast iron; crystallization

Citation: Stawarz, M.; Dojka, M. Bifilm Inclusions in High Alloyed Cast Iron. Materials 2021, 14, 3067. 1. Introduction https://doi.org/10.3390/ Foundry engineering processes are prone to many issues during casting manufactur- ma14113067 ing that may influence the final casting quality. Many are often ignored by manufacturers or even unknown. It is extremely important to realize the cause of poor casting quality Academic Editor: before buying expensive additives for metal improvement and waste money on defective Francesco Iacoviello castings. In many works, researchers explain that the design of a proper gating system, the way that liquid metal is poured into a mold, and melt treatments are important factors that Received: 24 May 2021 affect a casting’s quality [1–6]. Most of this research is based on Professor John Campbell’s Accepted: 2 June 2021 bifilm theory [6–8]. In short, as one of the researchers explains [5], according to this theory, Published: 4 June 2021 turbulence during the pouring generates defects inside the alloy as air bubbles and oxide films form on the surface of the liquid metal. The oxide film on the surface may fold over on Publisher’s Note: MDPI stays neutral itself. The double oxide ‘bifilm’ acts as a crack in the liquid metal, leading to the initiation with regard to jurisdictional claims in of additional cracks and hot tears in the casting. Figure1 presents a schematic illustration published maps and institutional affil- iations. of surface turbulence causing the entrainment of bifilms and associated bubbles. Dispinar D. et al. [1] and G. Gyarmati et al. [2] explained that the bifilm content in the melt is much more important than the dissolved gas content such as hydrogen, which has always been considered to be the major source of defects. They proved that with proper degassing, bifilm inclusions can be eliminated from the melt, and regardless of the Copyright: © 2021 by the authors. hydrogen content, aluminium castings without defects can be manufactured. Similarly, Licensee MDPI, Basel, Switzerland. Tiryakio˘gluM. [4] concluded that if air entrainment defects are eliminated, there is no This article is an open access article reason to measure or control the hydrogen levels in the melt. He clarified that due to the distributed under the terms and conditions of the Creative Commons presence of bifilms with two unbonded interfaces, nucleation is bypassed during pore Attribution (CC BY) license (https:// formation. Hydrogen diffuses to oxide bifilms and inflates them, therefore, hydrogen creativecommons.org/licenses/by/ serves as an agent to make entrainment defects visible. What is even more interesting, 4.0/). Uluda˘gM. et al. [3] studied Al-7%Si-Mg alloy and discovered that the best melt quality

Materials 2021, 14, 3067. https://doi.org/10.3390/ma14113067 https://www.mdpi.com/journal/materials Materials 2021, 14, 3067 2 of 11

was obtained when the melt was degassed without additives. Moreover, their research showed that all examined melt additions decreased the melt quality. They explained that Ti-containing intermetallic compounds were nucleated on biﬁlms. On the other hand, the addition of Sr to A356 increased the number of biﬁlms in the melt and also formed Materials 2021, 14, x FOR PEER REVIEWlarger pores. Sr reacts with Al2O3, and the formation of complex compounds results in2 of the 11

fracturing of oxides into smaller pieces through breakaway oxidation.

FigureFigure 1.1. SchemeScheme ofof biﬁlmsbifilms andand bubblesbubbles formationformation [[8].8].

MostDispinar research D. et in al. this [1] area and has G. beenGyarmati conducted et al. [2] on nonferrousexplained that alloys, the but bifilm bifilm content defects in arethe alsomelt a is cause much of more problems important with castthan iron.the dissolved Our previous gas content work showed such as thathydrogen, bifilm inclu-which sionshas always may cause been difficulties considered with to be the the inoculation major so ofurce high-chromium of defects. They cast proved iron [9 ].that During with inoculationproper degassing, with Ti addition,bifilm inclusions the TiC compoundscan be eliminated that serve from as the underlays melt, and for regardless M7C3 carbides of the arehydrogen transported content, on the aluminium bifilms with castings the crystallization without defects front can to thebe manu shrinkagefactured. cavities. Similarly, Thus, theyTiryakio are noğlu longer M. [4] anconcluded active nucleus that if inair this entrainm location,ent which defects is are very eliminated, unfortunate there from is anno economicreason to viewpoint.measure orAccording control the to hydrogen previous leve investigations,ls in the melt. bifilm He defectsclarified are that also due present to the inpresence the microstructure of bifilms with of greytwo castunbonded iron. Professorinterfaces, J. nucleation Campbell wroteis bypassed [10] that during graphite pore mayformation. growon Hydrogen bifilm inclusions. diffuses to He oxide explains bifilms that and the silicainflates film, them, when therefore, entrained hydrogen into the alloyserves as as a bifilm,an agent first to attractsmake entrainment the oxy-sulfide defects nuclei visible. to nucleate What is on even it. Then, more these interesting, nuclei serveUluda asğ M. the et underlay al. [3] studied for the Al-7%Si-Mg nucleation ofalloy flake and and discovered nodular graphite.that the best This melt theory quality of graphitewas obtained nucleation when is the also melt mentioned was degassed in the works without of I. additives. Riposan et Moreover, al. [11]. F. Hsutheir et research al. [12] andshowedR. Dojka that all et al.examined [13] noted melt the additions importance decreas of a gatinged the melt system quality. for manufacturing They explained good- that qualityTi-containing castings intermetallic according tocompounds John Campbell’s were nucleated bifilm theory. on bifilms The conducted. On the other research hand, onthe ductileaddition irons of Sr revealed to A356 that increased turbulent the fillingnumber from of bifilms the top in results the melt in oxide and also bifilms formed entrained larger inpores. ductile Sr ironreacts castings. with Al F.2O Hsu3, and et the al. [formation12] concluded of complex that floating compounds bifilms results are the in sources the frac- of intrinsicturing of cracks oxides in into castings. smaller Ductile pieces ironthrough castings breakaway may be oxidation. weakened by bifilm inclusions, resultingMost in research their premature in this area fracture. has been These conducted results show on nonferrous that turbulent alloys, filling but bifilm significantly defects influencesare also a cause the elongation of problems of ductile with cast iron. iron The. Our authors’ previous recent work work showed presented that inbifilm the paper inclu- confirmssions may the cause results difficulties of cast iron with investigations the inoculation in theof high-chromium area of bifilm defects. cast iron In [9]. the During paper, theinoculation crystallization with Ti of addition, alloyed ductile the TiC iron compounds (without the that addition serve as of underlays magnesium) for with M7C3 oxide car- bifilmbides are inclusions transported is discussed. on the bifilms The phenomenon with the crystallization of simple graphite front to and the bifilmshrinkage segregation cavities. towardsThus, they the are heat no center longer of an the active castings nucleus was in also this documented. location, which is very unfortunate from 2.an Materials economic and viewpoint. Methods According to previous investigations, bifilm defects are also present in the microstructure of grey cast iron. Professor J. Campbell wrote [10] that graphite Tests were carried out based on a two-stage metallurgical process of a liquid metal. may grow on bifilm inclusions. He explains that the silica film, when entrained into the Experimental melting was carried out in a medium-frequency induction furnace (PI25, alloy as a bifilm, first attracts the oxy-sulfide nuclei to nucleate on it. Then, these nuclei ELKON Sp. z o.o., Rybnik, Poland) with a capacity of 25 kg. Steel scrap with a low sulfur serve as the underlay for the nucleation of flake and nodular graphite. This theory of content was used as a charge. Table1 presents the results of the charge material chemical graphite nucleation is also mentioned in the works of I. Riposan et al. [11]. F. Hsu et al. content analysis. Steel scrap metal was used as a charge material for the test. [12] and R. Dojka et al. [13] noted the importance of a gating system for manufacturing good-quality castings according to John Campbell’s bifilm theory. The conducted research on ductile irons revealed that turbulent filling from the top results in oxide bifilms entrained in ductile iron castings. F. Hsu et al. [12] concluded that floating bifilms are the sources of intrinsic cracks in castings. Ductile iron castings may be weakened by bifilm inclusions, resulting in their premature fracture. These results show that turbulent filling significantly influences the elongation of ductile iron. The authors’ recent work presented in the paper confirms the results of cast iron investigations in the area of bifilm defects. In the paper, the crystallization of alloyed ductile iron (without the addition of magnesium)

Materials 2021, 14, 3067 3 of 11

Table 1. Chemical analysis of steel scrap used in the research.

Chemical Composition, % of Weight C Cr Si Mn Ni Mo S 0.080 0.037 0.001 0.594 0.024 0.01 0.006 Co Cu Al Sb As B P 0.010 0.004 0.046 0.009 0.108 0.00 0.016 Pb Nb Sn Ti W V Fe bal 0.002 0.031 0.009 0.001 0.011 0.006 98.995

The remaining components added during melting included: Ferrosilicon FeSi75 and synthetic graphite with a carbon content above 99.35%. The steel scrap was prepared appropriately for each melting. Before the weighing process, the steel scrap was cleaned of oxides and other impurities. Then, it was dried at 250 ◦C for 2 h inside a resistance furnace (DW40, Mario Di Maio SpA, 11-20122 Milano, Italy). For each melt, the measured portion of ferrosilicon FeSi75 was annealed at 650 ◦C for 2 h in a resistance chamber furnace (Curing Furnace F-120, Mario Di Maio SpA, 11-20122 Milano, Italy). Another stage included initial melting, which consisted of a melt of steel scrap with added graphite carburizer and ferrosilicon. After removing the slag, the prepared and melted material was poured into a steel casting mold. In the next step, the material prepared this way was used in the proper melt. Double melting was conducted to eliminate gas dissolved in the liquid alloy, which is recommended in the literature [14,15]. During the initial melting, samples for carbon content analysis were collected. The planned carbon content was 0.5%. The second stage of melting consisted of melting a previously prepared charge in an induction furnace and correcting the carbon content. During the melting of the charge, the metal bath was degassed [14]. This method was based on overheating liquid metal to 1400 ◦C, followed by a gradual reduction in the temperature in the furnace to approx. 1200 ◦C to remove gases from the bath, which is described in [16]. After the liquid alloy reached the temperature of 1200 ◦C, it was heated to approx. 1350 ◦C. Then, the liquid metal was poured into a ladle, the bottom of which was covered with FeTi67 foundry alloy to degas the metal bath. The chemical content analysis of samples was done in several stages, and the results are presented in Table2. During the ﬁrst stage, an analysis of the chemical composition was performed using a Leco GDS 500 spectrometer (Model No 607-500, Leco Corporation, 3000 Lakeview Ave, St. Joseph, MI, USA). Due to the signiﬁcant quantity of alloying additives, which in this case was silicon, the estimated Si content exceeded the allowable limits for the measured content of this element and for the used standard (maximum Si level circa 5%). Taking the above into account, it was decided to further analyze the Si content using the gravimetric method. Additional analysis for carbon and sulfur content was also conducted using a CS125 carbon-sulfur analyzer (Leco Corporation, 3000 Lakeview Ave, St. Joseph, MI, USA). The results of the aforementioned analyses are presented in Table2.

Table 2. Chemical composition of high-silicon cast iron.

Chemical Composition, % of Weight (1) (2) (2) (3) Si C S P Mn Mo Cu Mg Ti Fe bal Ce HSCI 18.70 0.52 0.003 0.022 0.301 0.022 0.064 0.00 0.027 80.34 6.32 (1) Si analysis by weight method, (2) carbon and sulfur analysis by CS 125 Leco, and (3) eutectic carbon equivalent calculated for C, P, and Si.

The samples for the metallographic tests were cast from the analyzed melt. These samples were cast in molds made of resin-covered sand, whose shape corresponds to the standardized samples for testing impact strength. Metallographic tests were performed using scanning electron microscopy (Phenom Pro-X with EDS system–Phenom-World B.V. Dillenburgstraat 9T Eindhoven, 5652 AM, The Netherlands). Cross-sections of the analyzed samples were used for testing. Materials 2021, 14, x FOR PEER REVIEW 4 of 11

Table 2. Chemical composition of high-silicon cast iron.

Chemical Composition, % of Weight (1) Si (2) C (2) S P Mn Mo Cu Mg Ti Fe bal (3) Ce HSCI 18.70 0.52 0.003 0.022 0.301 0.022 0.064 0.00 0.027 80.34 6.32 (1) Si analysis by weight method, (2) carbon and sulfur analysis by CS 125 Leco, and (3) eutectic carbon equivalent calculated for C, P, and Si.

The samples for the metallographic tests were cast from the analyzed melt. These samples were cast in molds made of resin-covered sand, whose shape corresponds to the standardized samples for testing impact strength. Materials 2021, 14, 3067 Metallographic tests were performed using scanning electron microscopy (Phenom4 of 11 Pro-X with EDS system–Phenom-World B.V. Dillenburgstraat 9T Eindhoven, 5652 AM, The Netherlands). Cross-sections of the analyzed samples were used for testing.

3.3. ResultsResults 3.1.3.1. SEMSEM AnalysisAnalysis FigureFigure2 a2a presents presents the the contraction contraction cavity cavity cross-section, cross-section, where where the dendritic the dendritic macrostruc- macro- turestructure of silicon of silicon ferrite ferrite is visible. is visible. Figure Figure2b–d 2b–d presents presents the cross-section the cross-section of the of contraction the contrac- cavitytion cavity with disorganizedwith disorganized biﬁlm bifilm inclusions inclusions inside. inside. Inclusions Inclusions of this of type this tend type to tend increase to in- thecrease degree the ofdegree expansion. of expansion. Figure2 e,fFigure shows 2e–f that shows a non-expanded that a non-expanded biﬁlm formed bifilm a formed distinct a complexdistinct complex shape. shape.

FigureFigure 2.2.Contraction Contraction cavitycavity (a(a)) with with a a visible visible dendritic dendritic structure. structure. Contraction Contraction cavities cavities ( b(b––dd)) filled filled withwith bifilm-type bifilm-type oxide oxide inclusions. inclusions. A A single single bifilm bifilm rolled rolled into into a a tube tube ( e(e,f–).f).

An interesting phenomenon was captured in Figure3, which showed that besides bifilm inclusions, the contraction cavities also contained spheroidal graphite precipitates. The external surface of the graphite precipitates suggests that these are primary precipitates [16–19]. These precipitates were pushed out of the liquid alloy towards the contraction cavity. Bifilm inclusions that were also floating in the liquid alloy reached the contraction cavity on the crystallization front. Materials 2021, 14, x FOR PEER REVIEW 5 of 11

An interesting phenomenon was captured in Figure 3, which showed that besides bifilm inclusions, the contraction cavities also contained spheroidal graphite precipitates. The external surface of the graphite precipitates suggests that these are primary precipi- Materials 2021, 14, 3067 tates [16–19]. These precipitates were pushed out of the liquid alloy towards the contrac-5 of 11 tion cavity. Bifilm inclusions that were also floating in the liquid alloy reached the contraction cavity on the crystallization front.

FigureFigure 3. 3.Contraction Contraction cavities cavities ( a(a–f–f)) ﬁlled filled with with biﬁlms bifilms inclusions inclusions and and nodule nodule graphite graphite precipitates. precipitates.

The surface of spheroidal graphite precipitates was smooth. According to previous studies [17,18,20–22], such a structure suggests that these are primary graphites that crystallized directly from the liquid. In such a case, the graphite crystallization process was stopped due to the discharge of spheroids from the liquid alloy. That is why the surface of precipitations is smooth. They were not disturbed by the secondary crystallization process. The chemical content of the analyzed alloy is strongly hypereutectic, which supports the hypothesis that this is primary graphite. Materials 2021, 14, x FOR PEER REVIEW 6 of 11

The surface of spheroidal graphite precipitates was smooth. According to previous studies [17,18,20–22], such a structure suggests that these are primary graphites that crystallized directly from the liquid. In such a case, the graphite crystallization process was stopped due to the discharge of spheroids from the liquid alloy. That is why the surface Materials 2021, 14, 3067 of precipitations is smooth. They were not disturbed by the secondary crystallization6 ofpro- 11 cess. The chemical content of the analyzed alloy is strongly hypereutectic, which supports the hypothesis that this is primary graphite.

3.2.3.2. EDSEDS AnalysisAnalysis TheThe EDSEDS analysisanalysis waswas conductedconducted forfor aa selectedselectedcross-section cross-section of of a a fragment fragment with with a a flat flat surface.surface. FigureFigure4 4 shows shows an an image image of of the the measurement measurement area. area. The The composition composition of of several several elementselements suchsuch as oxygen, silicon, aluminum aluminum,, carbon, carbon, and and iron iron was was determined. determined. Based Based on onthese these analyses, analyses, it itcan can be be concluded concluded that that the the bifilm bifilm layer layer was was primarily rich inin oxygen,oxygen, aluminum,aluminum, andand silicon.silicon. DueDue toto possiblepossible errorserrors duringduringthe theanalysis, analysis, the the chemical chemical content content of of thisthis layerlayer cannotcannot bebe unambiguouslyunambiguously specified.specified. TheThe preparationpreparation ofof metallographicmetallographic micro-micro- sectionssections was was hindered hindered in in this this case. case. The The spheroidal spheroidal graphite graphite precipitates precipitates may may have have been been lost duringlost during polishing, polishing, and theand oxide the oxide inclusions inclusions (bifilms) (bifilms) may may have have been been damaged damaged as well. as well.

FigureFigure 4.4. TheThe EDSEDS analysisanalysis ofof thethe contractioncontraction cavity’scavity’s surfacesurface coveredcovered inin biﬁlmsbifilms with with spheroidal spheroidal graphitegraphite extrusions.extrusions.

Figure5(1) shows the results of a local spectral analysis of the spheroidal graphite precipitates. The result explicitly conﬁrms that the precipitation consists primarily of carbon. The oxide inclusion was also analyzed, as shown in Figure5(2). The results of the oxide ﬁlm analysis did not provide a clear assessment of the chemical composition, but the layer was rich in oxygen, aluminum, silicon, carbon, and titanium. Materials 2021, 14, x FOR PEER REVIEW 7 of 11

Figure 5(1) shows the results of a local spectral analysis of the spheroidal graphite precipitates. The result explicitly confirms that the precipitation consists primarily of car- Materials 2021, 14, 3067 bon. The oxide inclusion was also analyzed, as shown in Figure 5(2). The results of7 ofthe 11 oxide film analysis did not provide a clear assessment of the chemical composition, but the layer was rich in oxygen, aluminum, silicon, carbon, and titanium.

FigureFigure 5. ChemicalChemical content content local local analys analysisis of the dark extrusion ( (1)–graphite,1)–graphite, and the light extrusion (2)–biﬁlm.(2)–bifilm.

4.4. Discussion Discussion InIn addition addition to to clusters, the liquid cast iron submicroscopic graphite particles were formedformed from thethe graphitegraphite precipitates precipitates in in the the metallic metallic charge charge material. material. The The graphite graphite structure struc- turetype type caused caused the individual the individual flat layers flat layers of carbon of carbon atoms toatoms retain to their retain bonds their at bonds temperatures at tem- peraturesas high as as 2000 high◦C, as while 2000 the°C, bondswhile betweenthe bond thes between layerswere the layers easily were broken. easily The broken. destruction The destructionof graphite crystalsof graphite during crysta meltingls during and melting overheating and ofoverheating the alloy began of the with alloy a began loss of with bonds a lossbetween of bonds the packets between of the flat packets layers. These of flat packets layers. thenThese dissociated packets then into dissociated individual flatinto layers, indi- vidualwhich wasflat layers, caused which by the iron,was caused oxygen, by and the hydrogen iron, oxygen, ions enteringand hydrogen them. Dependingions entering on them.the charge Depending material, on overheating the charge temperature,material, over etc.,heating the liquid temperature, cast iron et consistedc., the liquid of blocks cast ironof graphite consisted packets, of blocks packets of graphite of flat layers packets, or areas packets covered of flat with layers melted or areas packets covered in the formwith meltedof individual packets thin in the graphite form of molecules individual forming thin graphite ionic bondsmolecules between forming carbon ionic atoms bonds andbe- tweensolvent carbon atoms, atoms as well and as othersolvent elements atoms,and as well individual as other carbon elements atoms. and The individual aforementioned carbon atoms.graphite The formations aforementioned maintained graphite an unstable formations equilibrium maintained in a liquidan unstable alloy. Theequilibrium temperature in a liquidand time alloy. that The the temperature alloy can withstand and time arethat decisive the alloy factors can withstand for the size are of decisive homonymous factors forgroupings the size andof homonymous atoms. groupings and atoms. AsAs the liquid alloy cooled down to the liquidus temperature, the flat flat layers began compoundingcompounding intointo packets,packets, which which then then compounded compounded into into blocks blocks that that eventually eventually combined com- binedwith each with other. each other. InIn addition addition to to homogeneous homogeneous germs, germs, heterogeneous heterogeneous germs germs in the in form the form of various of various non- metallicnon-metallic inclusions inclusions can also can play also an play important an important role in role the ingraphitization the graphitization process. process. These These types of germs are most frequently attributed to oxide inclusions, especially SiO types of germs are most frequently attributed to oxide inclusions, especially SiO2 com-2 poundscompounds [20,23]. [20 ,The23]. literature The literature shows shows oxides oxides identified identified in the interior in the interior of spheroidal of spheroidal graphite,graphite, including including magnesium, magnesium, silicon, aluminum, silicon, aluminum,and titanium and oxides titanium [20,23]. John oxides Campbell [20,23]. [10]John wrote Campbell about [10 the] wrote role of about oxide the films role (silica of oxide bifilms) films in (silicathe growth bifilms) of invarious the growth forms of various forms of graphite. The presence of oxides (MgO, SiO2, Fe2O3, MnO) or sili- graphite. The presence of oxides (MgO, SiO2, Fe2O3, MnO) or silicates (Mg2SiO4, Fe2SiO4) cates (Mg SiO , Fe SiO ) in liquid ductile iron or the slag bound within has been docu- in liquid ductile2 4 iron2 or4 the slag bound within has been documented [10,23]. Complex mented [10,23]. Complex compounds, such as (Mg2SiO4, Fe2SiO4) have been recognized compounds, such as (Mg2SiO4, Fe2SiO4) have been recognized in cast iron [24]. Campbell in cast iron [24]. Campbell proposed that these oxides are not taut marbles, cubes, rods, proposed that these oxides are not taut marbles, cubes, rods, etc., but rather films or bi- etc., but rather films or bifilms with a lower Stokes velocity, which lets them remain in the films with a lower Stokes velocity, which lets them remain in the alloy suspension for a alloy suspension for a long time. Other researchers came to similar conclusions [25,26]. long time. Other researchers came to similar conclusions [25,26]. When there is no Mg, the When there is no Mg, the oxysulfide particles nucleate on oxide bifilms rich in silica, and then graphite nucleates and grows on oxysulfide inclusions, forming flake graphite. The magnesium additives eliminate bifilms rich in silica, and then graphite spheroids grow on oxysulfides, as it is assumed that graphite naturally grows spherically. Non-metallic inclusions, along with a set of exogenous graphite particles, heavily influence the cast iron’s ability to nucleate individual components of the microstructure, Materials 2021, 14, x FOR PEER REVIEW 8 of 11

oxysulfide particles nucleate on oxide bifilms rich in silica, and then graphite nucleates and grows on oxysulfide inclusions, forming flake graphite. The magnesium additives

Materials 2021, 14, 3067 eliminate bifilms rich in silica, and then graphite spheroids grow on oxysulfides, as8 ofit 11is assumed that graphite naturally grows spherically. Non-metallic inclusions, along with a set of exogenous graphite particles, heavily influence the cast iron’s ability to nucleate individual components of the microstructure, mainlymainly graphite. Graphite Graphite spheroids spheroids start start fo formingrming via via regular nucleation through very heavyheavy supercooling supercooling or or irregular irregular nucleation, nucleation, within within inclusions inclusions [23]. [23]. We We noted noted an an interest- interest- inging phenomenon phenomenon associated associated with with the the formation formation of of an an oxide film film or bifilm bifilm and graphite primary crystallization. The The results results are are in in close close agreement agreement with theories presented by otherother researchers mentioned in the paper. A Accordingccording to the conducted studies presented inin Figures Figures 22–5–5 andand recentrecent worksworks [[8],8], the the following following mechanism mechanism was was proposed proposed for for the the ob- ob- servedserved phenomena. The scheme in Figure 6 6 showsshows thatthat inin stagestage 11 afterafter pouring,pouring, onlyonly oxideoxide bifilmbifilm inclusions are are present in the liquid allo alloy.y. Then, in stage 2, graphitegraphite crystallization occurs,occurs, and and primary primary graphite graphite nodules nodules form form in in the the liquid liquid alloy, alloy, accompanied accompanied by by inclusions. inclusions. SpheroidalSpheroidal graphite is the main phase in the investigated alloy.

FigureFigure 6. BifilmsBiﬁlms in in the the liquid liquid alloy ( (stagestage 1 1).). Graphite crystallization directly from the liquid (stage(stage 2).2).

AccordingAccording to to the the theories theories of of D. D.M. M. StefanescuStefanescu and J. Campbell [10,23], [10,23], graphite graphite pre- pre- cipitatescipitates crystallize crystallize on on the substrates. Graphi Graphitete growth growth in in a a liquid solution requires not onlyonly the the intensive intensive diffusion diffusion of of carbon carbon atoms atoms towards towards the the growing growing crystal crystal but but also also the thein- tensiveintensive movement movement of ofiron iron atoms atoms in inthe the opposite opposite direction. direction. The The graphite graphite eutectic eutectic system system in spheroidalin spheroidal cast cast iron iron crystallizes crystallizes outside outside the the area area of of simultaneous simultaneous growth growth [27]. [27]. Graphite andand austenite austenite crystallize crystallize separately separately under under significant significant supercooling supercooling conditions. conditions. During During the firstthe firststage, stage, graphite graphite spheroids spheroids grow growin direct in directcontact contact with the with liquid. the However, liquid. However, the eutectic the austeniteeutectic austenite growth does growth not does occur not simultaneous occur simultaneously.ly. Both phases Both grow phases separately, grow separately, and graph- and itegraphite spheroids spheroids grown grown via austenite via austenite dendrites dendrites are eventually are eventually surrounded surrounded by austenite by austenite and separatedand separated from fromthe metallic the metallic fluid fluid[17]. [A17 simi]. Alar similar cast iron cast ironcrystallization crystallization pattern pattern was pre- was sentedpresented by M. by Zhu M. Zhu et al., et al.,but butwithou withoutt the theparticipation participation of bifilms of bifilms [27]. [27 ]. BasedBased on reported theories andand resultsresults [[17,27],17,27], we we can can predict predict the the presence presence of of austenite austen- itein stagein stage 3 (Figure 3 (Figure7). 7). Due to a low carbon solubility in the liquid alloy (with a high Si content), the graphite precipitates may be transported on the crystallization front. In addition to graphite nodules, an oxide layer and light bifilms were also pushed by the growing crystals, which led to stage 4 (Figure7). In this stage, graphite precipitates might be covered and mixed into the double oxide film. The spheroidal graphite nucleation on the surface of the oxide film may be the reason for the accumulation of graphite nodules in the contraction cavity.

Materials 2021, 14, x FOR PEER REVIEW 9 of 11

Materials 2021, 14, 3067 9 of 11 Materials 2021, 14, x FOR PEER REVIEW 9 of 11

Figure 7. Graphite eutectic system growth in the bifilm environment step by step (stage 3 and satge 4).

Due to a low carbon solubility in the liquid alloy (with a high Si content), the graphite precipitates may be transported on the crystallization front. In addition to graphite nodules, an oxide layer and light bifilms were also pushed by the growing crystals, which led to stage 4 (Figure 7). In this stage, graphite precipitates might be covered and mixed into FigureFigure 7. GraphiteGraphitethe eutectic eutectic double system system oxide growth growth film. in The the spheroidal bifilm bifilm envi environment ronmentgraphite step nucleation by step (stage( stageon the 3 andsurface satge of 4).4 ).the oxide film may be the reason for the accumulation of graphite nodules in the contraction cavity. IfDueIf we we to look look a low again again carbon at at the solubility oxide inclusions in the liquid in in alloy the companion (with a high of Si graphite content), presented the graphite in FigureprecipitatesFigure 33,, wewe canmay can observe observebe transported the the fragile fragile on and the and torn crystall torn character characterization of front. the of oxidethe In oxide addition film. film. The to sameThe graphite same torn layer nod-torn layerules,of the anof oxide theoxide oxide film layer isfilm visibleand is lightvisible in Figurebifilms in Figure5 were, which 5, also which may pushed bemay the bybe result the growingresult of mixing of mixingcrystals, and turbulenceand which turbu- led lencetoduring stage during filling 4 (Figure filling the mold.7). the In mold.this It may stage, It be may that graphite be when that precipitates the when bifilms the seizedbifilms might and seizedbe covered mixed and with mixedand the mixed with graphite into the graphitetheprecipitates, double precipitates, oxide and graphitefilm. and The graphite nodulesspheroidal torenodules graphite apart tore the nucleation apart fragile the film. fragileon the film.surface of the oxide film may Moreover,beMoreover, the reason the for film film the may accumulation also be damaged of graphite by graphite nodules growth in the contractionon its surface. cavity. In stage 55 (Figure (FigureIf we 8)8 )look it can can again be be noticed, noticed, at the oxidehow how graphite inclusions graphite nodules nodules in the mixed companion mixed with with the of the bifilmsgraphite bifilms are presented are pushed pushed onin theFigureon thegrowing growing3, we dendrites.can dendrites. observe Due the Due to fragile the to thepresented and presented torn crystallization character crystallization of the mechanisms, oxide mechanisms, film. a The portion a portionsame of torn the of precipitatedlayerthe precipitated of the oxide graphite graphite film nodules is visible nodules is in forced Figure is forced out 5, towardswhich out towards may the becontraction the the contraction result cavitiesof mixing cavities which and which isturbu- pre- is sentedlencepresented during in stage in stagefilling 6 (Figure 6 the (Figure mold. 8). 8). It may be that when the bifilms seized and mixed with the graphite precipitates, and graphite nodules tore apart the fragile film. Moreover, the film may also be damaged by graphite growth on its surface. In stage 5 (Figure 8) it can be noticed, how graphite nodules mixed with the bifilms are pushed on the growing dendrites. Due to the presented crystallization mechanisms, a portion of the precipitated graphite nodules is forced out towards the contraction cavities which is presented in stage 6 (Figure 8).

Figure 8. CreationCreation processprocess of of shrinkage shrinkage cavity cavity (stage (stage 5). Shrinkage5). Shrinkage cavity cavity with primarywith primary graphite graphite nodules nodules mixed andmixed trapped and trapped with the bifilm inclusions (stage 6). with the biﬁlm inclusions (stage 6).

5. Conclusions This work provided the following observations and conclusions:

Figure 8. Creation process of shrinkage cavity (stage 5). Shrinkage cavity with primary graphite nodules mixed and trapped with the bifilm inclusions (stage 6).

Materials 2021, 14, 3067 10 of 11

• Spheroidal precipitates crystallized directly from the liquid and may nucleate on the substrates created by the bifilms. This hypothesis was indirectly supported by a large number of spheroidal graphite precipitates in the presence of bifilms. • Spheroidal graphite precipitates were forced out of the liquid. In the discussed case, the precipitates were forced in the direction of the contraction cavities. • Contraction cavities with no visible bifilm inclusions, as well as spheroidal graphite precipitates, were observed in the sample. • The smooth surface of spheroidal graphite precipitates suggests that primary graphite crystallized directly from the liquid. • Spheroidal graphite precipitates occurring during the mixing in the liquid alloy ripped the bifilms. • Some bifilm precipitates occurring in the alloy did not have the time to develop and formed tube-like shapes. • Primary graphite was forced out of the metallic fluid. There are oxide inclusions in the way of the graphite spheroids. Spheroid graphite precipitates were captured in oxide lattices. Both were discharged from the liquid alloy towards the contraction cavity.

Author Contributions: Conception of research, M.S. and M.D.; methodology, M.S. and M.D.; investigation, M.S. and M.D.; data curation, M.S.; writing—original draft preparation, M.S. and M.D.; writing—review and editing, M.S. and M.D.; visualization, M.D.; supervision, M.S. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Data sharing is not applicable to this article. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Dispinar, D.; Campbell, J. Critical assessment of reduced pressure test. Part 1: Porosity phenomena. Int. J. Cast Met. Res. 2004, 17, 280–286. [CrossRef] 2. Gyarmati, G.; Fegyverneki, G.; Mende, T.; Tokár, M. Characterization of the double oxide film content of liquid aluminum alloys by computed tomography. Mater. Charact. 2019, 157, 109925. [CrossRef] 3. Uluda˘g,M.; Çetin, R.; Dispinar, D.; Tiryakio˘glu,M. Characterization of the Effect of Melt Treatments on Melt Quality in Al-7wt %Si-Mg Alloys. Metals 2017, 7, 157. [CrossRef] 4. Tiryakio˘glu,M. The Effect of Hydrogen on Pore Formation in Aluminum Alloy Castings: Myth Versus Reality. Metals 2020, 10, 368. [CrossRef] 5. Dojka, R.; Jezierski, J.; Campbell, J. Optimized gating system for steel castings. J. Mater. Eng. Perform. 2018, 27, 5152–5163. [CrossRef] 6. Campbell, J. A Personal View of Microstructure and Properties of Al Alloys. Materials 2021, 14, 1297. [CrossRef] 7. Campbell, J. The consolidation of metals: The origin of bifilms. J. Mater. Sci. 2016, 51, 96–106. [CrossRef] 8. Campbell, J. Complete Casting Handbook, 2nd ed.; Elsevier: Oxford, UK, 2015. 9. Dojka, M.; Stawarz, M. Bifilm Defects in Ti-Inoculated Chromium White Cast Iron. Materials 2020, 13, 3124. [CrossRef] 10. Campbell, J. The Structure of Cast Irons. Mater. Sci. Forum 2018, 925, 86–89. [CrossRef] 11. Riposan, I.; Chisamera, M.; Stan, S.; Toboc, P.; Ecob, E.; White, D. Al, Zr–FeSi preconditioning of grey cast irons. Mater. Sci Technol. 2008, 24, 579–584. 12. Hsu, F.; Wang, K.; Li, C. Bifilm defects in ductile-iron support bracket castings. Int. J. Cast Metal. Res. 2017, 30, 148–158. [CrossRef] 13. Dojka, R.; Jezierski, J.; Tiedje, N.S. Geometric Form of Gating System Elements and Its Influence on the Initial Filling Phase. J. Mater. Eng. Perform. 2019, 28, 3922–3928. [CrossRef] 14. Stawarz, M. Intermetallic phases in alloyed cast iron with 18% si addition. Arch. Foundry Eng. 2020, 2, 37–42. [CrossRef] 15. Stawarz, M.; Kajzer, W.; Kajzer, A.; Dojka, M. Physicochemical properties of silicon cast iron. Arch. Foundry Eng. 2017, 2, 101–106. [CrossRef] 16. Sakwa, W. Castings with Enhanced Properties, 1st ed.; Silesian University of Technology: Gliwice, Poland, 1973. 17. Stefanescu, D.M.; Alonso, G.; Larranaga, P.; De la Fuente, E.; Suarez, R. On the crystallization of graphite from liquid iron-carbon– silicon melts. Acta Mater. 2016, 107, 102–126. [CrossRef] Materials 2021, 14, 3067 11 of 11

18. Stefanescu, D.M.; Alonso, G.; Larranaga, P.; De la Fuente, E.; Suarez, R. Reexamination of crystal growth theory of graphite in iron-carbon alloys. Acta Mater. 2017, 139, 109–121. [CrossRef] 19. Stefanescu, D.M.; Alonso, G.; Larranaga, P.; De la Fuente, E.; Suarez, R. Reassessment of crystal growth theory of graphite in cast iron. Mater. Sci. Forum 2018, 925, 36–44. [CrossRef] 20. Alonso, G.; Stefanescu, D.M.; De la Fuente, E.; Larranaga, P.; Suarez, R. The influence of trace elements on the nature of the nuclei of the graphite in ductile iron. Mater. Sci. Forum 2018, 925, 78–85. [CrossRef] 21. Górny, M.; Stefanescu, D.M. Thin-Wall Ductile Iron Castings. In Cast Iron Science and Technology; Stefanescu, D.M., Ed.; ASM International: Columbus, OH, USA, 2017. 22. Górny, M.; Kawalec, M.; Sikora, G.; Olejnik, E.; Lopez, H. Primary structure and graphite nodules in thin walled high-nickel ductile iron castings. Metals 2018, 8, 649. [CrossRef] 23. Alonso, G.; Stefanescu, D.M.; Gonzalez, R.; Suarez, R. Effect of magnesium on the solid-state nucleation and growth of graphite during annealing of white iron. Int. J. Met. 2020, 14, 728–735. [CrossRef] 24. Stawarz, M.; Janerka, K.; Dojka, M. Selected phenomena of the in-mold nodularization process of cast iron that influence the quality of cast machine parts. Processes 2017, 5, 68. [CrossRef] 25. Bruna, M.; Galˇcík, M. Casting quality improvement by gating system optimization. Arch. Foundry Eng. 2021, 1, 132–136. [CrossRef] 26. Gyarmati, G.; Fegyverneki, G.; Tokár, M.; Mende, M. Investigation on double oxide film initiated pore formation in aluminum casting alloys. Int. J. Eng. Manag. Sci. 2020, 5, 141–153. [CrossRef] 27. Zhu, M.; Zhang, L.; Zhao, H.; Stefanescu, D.M. Modeling of microstructural evolution during divorced eutectic solidification of spheroidal graphite irons. Acta Mater. 2015, 84, 413–425. [CrossRef]