© 2012 ISIJ 1848

ISIJ International, Vol. 52 (2012), No. 10, pp. 1848–1855

Influence of Magnesium Content and Coating Type on GraphiteDegeneration in Surface Layer of Iron Castings in Resin Sand –P-Toluol Sulphonic Acid (PTSA) Moulds

Nicoleta IVAN, Mihai CHISAMERA and Iulian RIPOSAN*

POLITEHNICA University of Bucharest, Materials Science and Engineering Faculty, 313. Spl. Independentei, RO-060042,Bucharest, Romania.

(Received on March 17, 2012; accepted on June 7, 2012)

The performance of coatings on furan resin sand moulds [P-Toluol Sulphonic Acid (PTSA) as hardener][FRS-PTSA], was compared to ‘no sulphur’ Novolak resin coated sand [NRS] moulds by analysing thegraphite characteristics in the surface layer of Mg-treated irons (0.020–0.054%Mg) and over the entiresection. The extent of an abnormal surface layer is influenced by the different Mg contents for compactedor nodular graphite irons, the sulphur content of the moulding system and the type of mould coating (con-taining sulphur or desulphurization materials). With lower Mg content, more graphite degeneration wasapparent in the cast surface layer, especially at less than 0.03%Mgres [typical Mg content for vermicular/compacted graphite cast irons] with increasing differences between FRS-PTSA and NRS moulds, anduncoated and coated moulds results. Un-coated FRS-PTSA moulds prompted five times more degenerategraphite in the surface layer than NRS moulds. Sulphur bearing coatings increased the layer up to fivetimes the thickness in NRS moulds and up to 50% in FRS-PTSA moulds, especially for less than0.030%Mg residual. Desulphurization type coatings based on MgO appear to be efficient for high Mg-content iron (0.05%Mg, ductile iron) but not active enough for less than 0.03%Mg (compacted graphiteiron type). The change in graphite characteristics in the centre of samples evolved in a clear relationshipwith degenerate graphite in the surface layer, for the experimental solidification conditions.

KEY WORDS: compacted graphite iron; ductile iron; graphite degeneration; abnormal surface layer; furanresin sand mould; S-bearing coating; MgO-based coating.

1. Introduction

Chemically bonded sand moulds (self-set, no-bake, coldbox) are commonly used for producing a large size range ofiron castings. Recently, it was estimated that more than 60%of USA foundries use no-bake (42%) and cold box (21%)processes.1) Furan resin–acid and phenolic resin–acid no-bake systems appear to be very popular in ductile ironfoundries because the strong moulds contribute to fewercontraction defects during solidification (shrinkage andmicro-shrinkage). In this case only the contraction of thecooling liquid iron needs to be compensated for, hence, few-er and smaller risers can be employed, or even riserlessdesigns can be considered (>2.0 cm cooling modulus).2)

By using equipment specifically developed for simultane-ous recording of cooling and contraction curves, mouldrigidity was observed to have an obvious effect on ductileiron shrinkage tendency.3–5) The highest levels of both con-centrated and total shrinkage were recorded in green sandmould solidification, where a much higher level of the initialeutectic expansion, as compared to furan resin mould, was

also recorded. In all cases the apparent density of ductileiron samples was higher in furan resin mould versus greensand mould. It was found that green sand moulds, being lessrigid moulds, encourage the formation of contractiondefects, not only because of high initial expansion values,but also because of a higher cooling rate during solidifica-tion and, consequently, increased undercooling below themetastable equilibrium temperature up to the end of solidi-fication.

Graphite degeneration, in terms of losing its degree ofcompactness, in the surface layer of Mg-treated cast irons isa common defect, which can occur with any moulding tech-nique, but each with characteristics affected by various fac-tors. Although resin mould technology is suitable for ductileiron casting it frequently contributes to graphite degenera-tion in the surface of spheroidal graphite cast iron, forminga flake (lamellar) graphite seam.

With furan and phenolic resin-bonded moulding, P-ToluolSulphonic Acid (PTSA) is usually used as the hardener, butit is an important source of sulphur in this system, affectingthe formation of a degenerate graphite layer.6–10) Not surpris-ingly, oxygen has an influence in resin mould technology,including systems with sulphur built in, especially if turbu-lent flow occurs, water–based no-bake binder, Mg-silica

* Corresponding author: E-mail: i_riposan@rectorat.pub.roDOI: http://dx.doi.org/10.2355/isijinternational.52.1848

ISIJ International, Vol. 52 (2012), No. 10

1849 © 2012 ISIJ

reaction or oxidation of reactive elements causing dross for-mation. The MgS–O reaction can occur resulting in a freshrelease of sulphur, which is then able to react with and causea loss of effective Mg. Nitrogen bearing resins have a pro-found effect on the occurrence and severity of surface pin-holes, but have a limited influence on surface graphitedegeneration.11–13)

Various coating materials have been tested to determinewhether surface deterioration can be prevented by mouldcoating including conventional graphite-based coatings,inorganic materials expected to act as desulphurizers(Al2O3, CaCO3, Basic slag, CaF2, Talc), and sinterable mate-rials expected to act as protective layers. By employing CaOor MgO desulphurization type coatings, to restrict the roleof SO2, the surface layer effect has been significantlyreduced. Dressing with a CaO/MgO/Talc composition isviewed as being particularly effective. Generally, mouldcoatings based on materials such as CaO and MgO, to devel-op a desulphurizing environment, are more effective thansimpler protective coatings.6,9,11–14)

The objective of the present paper is to evaluate theoccurrence of a graphite degenerated layer at the surface of

Mg-treated iron castings in a furan resin sand-PTSA mould-ing system as influenced by residual magnesium content andmould coating type.

2. Experimental Procedure

The base irons were produced in an acid lined corelessinduction furnace (100 kg, 2 400 Hz, 100 kW). The meltswere superheated to 1 550°C and maintained for 8 minutesand thereafter tapped into the pre-heated tundish magnesiumtreatment ladle [1 530°C]. FeSiCaMgRE alloy [(wt.%):46Si, 0.8Al, 1.2Ca, 6.05Mg, 0.96RE-rare earth, Fe-bal] wasused, at three addition levels [1.0, 1.5 and 1.9 wt.%], toobtain different residual magnesium levels in the final iron.Treated irons were inoculated in all cases during transfer ofMg-treated iron to the pouring ladle [0.5 wt.% Ca,Ce,S,O–FeSi alloy: (wt.%) 70–76Si, 0.75–1.25Ca, 0.75–1.25Al,1.5–2.0Ce, less than 1.0 sulphur and oxygen, Fe-bal with0.2–0.7 mm particle size].15)

The composition of the final ductile iron (Table 1) wasslightly hypo-eutectic (3.9–4.0%CE carbon equivalent),including (wt.%): 3.06–3.23C, 2.54–2.76Si, 0.15–0.18Mn,

Table 1. Chemical composition of tested cast irons.

IronTreatment, wt.% Chemical composition*, wt.%

Carbon Echiv. CE (%)FeSiCaMgRE Ca, Ce, S, O–FeSi C Si Mn Mg Ce La

I 1.0 0.5 3.06 2.76 0.15 0.020 0.0062 0.0022 3.90

II 1.5 0.5 3.23 2.54 0.18 0.027 0.0082 0.0033 3.99

III 1.9 0.5 3.15 2.54 0.17 0.054 0.0106 0.0050 3.92

*Other elements (wt.%): 0.017–0.018P, 0.015–0.018S, 0.039–0.042Cr, 0.0034–0.0087Mo, 0.07–0.08Ni, 0.012–0.015Al,0.033–0.036Cu, 0.006–0.008Ti, 0.008–0.013V, 0.0023–0.0029W, 0.0031–0.0033Sn, 0.0030–0.0032As, 0.0021–0.0032N.

Fig. 1. Experimental set including semi-cylindrical samples arrangement in the resin sand mould (a, b), cut sample forstructure analysis (c), the analysis procedure to evaluate the surface layer thickness (d), the structure parameters inthe surface layer (e), their variation on the sample section (f) and in sample centre (g).

© 2012 ISIJ 1850

ISIJ International, Vol. 52 (2012), No. 10

0.017–0.018P, less than 0.02S and different levels of Mg(0.020–0.054%), Ce (0.006–0.011%) and La (0.0022–0.0050%) reflecting the different additions of FeSiCaMgREalloy, with other residuals maintained at low levels. Thechemistries of the three final irons (I, II, III) represent dif-ferent graphite nodularities, from vermicular/compactedgraphite iron, through mixed graphite morphologies (ver-micular + nodular) up to the fully nodular graphite iron.

Furan resin (3.0 wt.%)–P-Toluol Sulphonic Acid (PTSA)(1.5 wt.%)–Silica Sand (95.5 wt.%) [FRS-PTSA] mouldingsystem was used. Typically, the thermo-physical properties ofthese moulds include: 1 306 W.s1/2/m2.K thermal diffusivity,1170 J/kg.K specific heat, 0.94 W/m.K thermal conductivityand 1 550 kg/m3 density. Each mould included four identicalsemi-cylindrical samples [28 mm wall thickness, 0.56 kg,cooling modulus CM = 7.1 mm] (Figs. 1(a) and 1(b)).

Standard Quik-cupTM cups [35 mm wall thickness, 0.35 kgiron casting and CM = 7.3 mm] were also used, as referencesulphur free ceramic moulds: phenol-formaldehyde resin(Novolak) coated sand [NRS] (Croning process). Typically,the thermo-physical properties of these ceramic cups include:1 487 W.s1/2/m2.K thermal diffusivity, 1 280 J/kg.K specificheat, 1.08 W/m.K thermal conductivity and 1 600 kg/m3

density.Two types of coating were applied on the concave surface

of these sample moulds (see Fig. 1(a)). The coatings wereprepared in the same conditions, with powder materials upto 0.01 mm grain size at the same quantities (0.25 g FeS2 or0.25 g MgO), with the same binder (4.0 g Polystyrene and10 ml Toluene) and applied onto the moulds surface in thesame procedures (0.35–0.40 mm coating thickness). Thecoating incorporating FeS2 powder (49–52 wt.%S) as a sup-plementary source of sulphur at the metal–mould interface,was used to evaluate if this element migrates into the ironmelt to cause graphite degeneration in the surface layer of

Mg-treated cast irons. A second coating incorporated a MgObased material, which is anticipated to perform as a desul-phurizer.

3. Results and Discussion

The semi-cylindrical samples (Fig. 1(c)) and cup sampleswere cut and prepared for metallographic analysis understandard conditions. The thickness of the surface layer of thesemi-cylindrical samples was evaluated along the 30 mmtotal length (25 mm for cup samples), with 100 μm separat-ing the analysed points, avoiding the end (corner) effects forboth types of samples (Fig. 1(d)). The thickness of the sur-face layer was expressed as an average and a maximum lev-el from a total of 300 and 250 measurements, respectively.The structure was analysed on 10 individual directions inthe surface layer (Fig. 1(e)). A total area of 800 μm2 wasanalyzed, including the surface layer, the transition zone andadjacent “good” area. The structure variation on the semi-cylindrical samples section was evaluated along 3 radii(Fig. 1(f)), up to the centre of samples, with 0.1 mm sepa-rating the analyzed points up to 2.0 mm distance from thesurface, and thereafter 1.0 mm interval between analyzedpoints. The structure was also analysed in the centre of thesemi-cylindrical sample (Fig. 1(g)) and centre of the cupsample, along 3 directions, each one including 5 analysispoints with 2 mm between them; the averages of the struc-ture parameters were considered.

The cast iron microstructures in the casting surface regionare presented in Fig. 2 for FRS-PTSA moulds and in Fig. 3for NRS moulds. The severity of the abnormal surface layeris influenced by the residual magnesium content (Mg-treatediron type), the moulding system (with and without sulphurcontent), and whether the coating increased the availablesulphur, or removed it by desulphurization (Table 2, Fig. 4).

Fig. 2. The influence of residual magnesium content and coating type on the iron structure in the surface layer of semi-cylindrical cast iron samples solidified in Furan Resin Sand-PTSA [FRS-PTSA] mould.

ISIJ International, Vol. 52 (2012), No. 10

1851 © 2012 ISIJ

In uncoated mould samples, there is a big differencebetween solidification in FRS-PTSA moulds and NRSmoulds as they affect the abnormal graphite types at thesurface layer, including degenerate graphite, vermicular/

compacted and/or lamellar morphologies. In NRS moulds,the occurrence of degenerate graphite in the surface layer isless likely: samples showed a limited layer less than 100 μmaverage thickness, comparing to 185–532 μm in FRS-PTSAmoulds and up to 215 μm maximum, compared to 370–880μm in FRS-PTSA moulds.

FRS-PTSA moulds evidently promoted more degenerategraphite in the surface layer of the ductile iron test castings.The thickness of this layer increased more than five timescompared to NRS moulds, depending on the residual magne-sium content. At lower magnesium content, there was greatergraphite degeneration in the cast surface layer, especially atless than 0.03%Mgres, typical of vermicular/compactedgraphite cast irons (Fig. 4).

The application of different mould coatings had a stronginfluence on the graphite morphology in the cast surfacelayer, prompting deterioration to lamellar graphite, or con-versely minimizing the thickness of the surface layer byaffording some protection.

Fig. 3. The influence of residual magnesium content and coating type on the iron structure in the surface layer of cast ironcup samples solidified in Novolak Resin Coated Sand [NRS] mould.

Fig. 4. Influence of residual magnesium content, mould and coating type on the average surface layer thickness of Mg-treated iron castings [a) layer thickness; b) layer thickness–magnesium content relationship].

Table 2. Abnormal Surface Layer Thickness.

MouldType*

Iron Abnormal Surface Layer Thickness,

Type Mgres[%]

Average Maximum

UC FeS2 MgO UC FeS2 MgO

FRS-PTSA

I 0.020 531.47 595.11 472.95 879.07 1133.02 709.24

II 0.027 281.60 417.83 231.94 572.90 839.64 165.92

III 0.054 185.49 194.06 24.56 371.79 452.88 129.20

NRS

I 0.020 85.90 220.06 5.30 214.23 318.20 65.60

II 0.027 64.60 88.00 1.0 124.80 178.80 1.0

III 0.054 15.50 70.90 1.0 68.00 170.00 1.0

*FRS-PTSA–Furan Resin Sand-PTSA Mould; NRS–Novolak Resin SandMould.

© 2012 ISIJ 1852

ISIJ International, Vol. 52 (2012), No. 10

Incorporating sulphur in the coating material had anadverse effect on the degenerate graphite surface layer, inboth moulding systems. Comparing the effects of an inten-tional sulphur content in the test mould coating, the averagelayer thickness increased up to five times in NRS moulds(no mould sulphur) and up to 50% in the FRS-PTSAmoulds, which have sulphur in the binder system. Thebehaviour of this sulphur bearing coating was attributable tothe newly introduced sulphur in the NRS moulds and to thesulphur supplementing that already in the binder system(PTSA), in the second case.

This suggests that even though sulphur migration occursin both these examples the intentionally added sulphur dif-fuses more readily than the sulphur type in the FRS-PTSAbinder. It is assumed that sulphur supplied by the sulphur–bearing coating added to the sulphur available in the mould(PTSA contribution) and diffused into the outer layer ofMg-treated iron to react with Mg in the mould cavity. As aconsequence of the localized, elevated sulphur levels, mag-nesium is partially consumed prior to solidification, and thenodulising potential of the treated iron decreases.

It is assumed that sulphur supplied by the S-bearing coat-ing added to the sulphur available in the mould (PTSA con-tribution) and diffused into the outer layer of Mg-treatediron to react with Mg in the mould cavity. As a consequenceof the localized, elevated sulphur levels, magnesium is par-tially consumed prior to solidification, and the nodulising

potential of the treated iron decreases. Some micro-inclusionswere detected in the surface layer of castings, typicallyresulted from Mg–S reactions (Fig. 5).

CH3C6H4SO3H [PTSA] + HOH [water steam]= C6H5CH3 [Toluene] + H2SO4

= C6H5CH3 [Toluene] + {SO2} + {H2O}........... (1)

{SO2} + 3[Fe] = [FeS] + 2[FeO] ............... (2)

[Mg] + [FeS] = (MgS) + [Fe] .................. (3)

[Mg] + [FeO] = (MgO) + [Fe] ................. (4)

3[Mg] + {SO2} = (MgS) + (MgO) .............. (5)

Sulphur bearing coating contributed to the increased sur-face layer thickness for all residual magnesium contents, butwith increased effect at lower Mg content. For both FRS-PTSA and NRS moulds, using the 0.054%Mgres treated ironas a reference, decreasing magnesium content led to anincrease in layer thickness, double for 0.027%Mgres andthree times for 0.020%Mgres, respectively.

The coatings based on desulphurizer type materials, suchas MgO, provided some protection at the metal–mould inter-face against graphite degeneration, with a decrease in theaverage thickness of the layer compared to the reference(uncoated moulds). Practically, no graphite degeneration inNRS moulds and beneficial effects in the FRS-PTSAmoulds, with residual magnesium content as a strong factor:10–30% decrease for 0.020–0.027%Mgres and more thanseven times decrease of the layer for 0.054%Mgres, respec-tively. It is assumed that this coating contributed to localdesulphurization thereby reducing the negative effect of sul-phur released by the mould.

The residual magnesium content, mould type and appliedcoatings also had an influence on graphite characteristics inthe centre of the castings, in the test solidification conditions(Table 3, Figs. 6 and 7). The graphite characteristics wereevaluated with Automatic Image Analysis [analySIS® FIVEDigital Imaging Solutions software], which evaluated thefollowing parameters:

Graphite Nodularity (NG)= 100 [Σ Anodules + 0.5 × Σ Aintermediates]/Σ Aall particles.... (6)

Shape Factor (SF) = 4πAG/PG2 ................. (7)

Aspect Ratio (AR) = b/a ...................... (8)

Fig. 5. Typical micro-inclusions in the surface layer of semi-cylindrical cast iron samples solidified in Furan ResinSand-PTSA [FRS-PTSA] mould [Nital 2% etched].

Table 3. Graphite Characteristics in the Centre of Test Castings.

MouldType*

Iron Graphite Characteristics

Type Mgres[%]Nodularity, % Shape Factor Aspect Ratio

UC FeS2 MgO UC FeS2 MgO UC FeS2 MgO

FRS-PTSA

I 0.020 55.03 53.21 61.03 0.65 0.65 0.74 1.88 1.91 1.70

II 0.027 73.23 68.14 78.57 0.76 0.75 0.82 1.53 1.66 1.50

III 0.054 83.08 82.45 84.39 0.83 0.83 0.87 1.36 1.36 1.36

NRS

I 0.020 47.73 46.16 55.98 0.38 0.35 0.73 2.01 2.15 1.83

II 0.027 57.42 51.17 61.87 0.74 0.72 0.76 1.99 1.95 1.83

III 0.054 75.73 71.21 77.60 0.84 0.81 0.85 1.51 1.53 1.48

*FRS-PTSA–Furan Resin Sand-PTSA Mould; NRS–Novolak Resin Sand Mould.

ISIJ International, Vol. 52 (2012), No. 10

1853 © 2012 ISIJ

where:Anodules is the area of particles classified as nodules;

Aintermediates – area of particles classified as intermediates; Aall

particles – area of all graphite particles; AG – area of the graphiteparticle in question; PG – perimeter of the graphite particlein question; AR – the maximum ratio of height (b) and width(a) of a rectangular boundary for the measured particle.

As expected, in the reference uncoated FRS-PTSA andNRS mould samples, at higher magnesium residual, graphitenodularity was high (Figs. 6(a) and 6(b)). In the iron serieswith the highest residual magnesium (0.054%) the graphitenodularity was at 83% in the FRS-PTSA mould, and 10%lower in the cup-type sample, due to higher Mg loss at theopen sample surface. Graphite nodularity decreased at lowermagnesium content, but proportionately for the two mouldmedia (NRS: FRS-PTSA): 57% vs 73% for 0.027%Mgresand 48% vs 55% for 0.020%Mgres, respectively.

The selection of thermal analysis cups to represent moulds

with low sulphur content introduced a different pouring ratewhile filling the mould, followed by open top solidification.The filling rate determined by cooling curves parameterswere 0.04 kg/s for semi-cylindrical samples and only 0.028kg/s for Quik-cup samples (especially lower, to protect theincluded thermocouple). Open top solidification conditionsfavour the gaseous Mg loss, similarly to the open ladlebahaviour. For that reason the Mg loss at the exposed surfacewould be expected to be higher than in enclosed moulds rep-resenting the high sulphur moulding media, FRS-PTSA.

The application and the type of coating appear to alsoinfluence graphite nodularity into the center of these cast-ings, decreasing up to 10% with a sulphur bearing coating orincreasing by up to 20% with a MgO type coating, more soat lower magnesium content, compared to uncoated moulds.

With increasing thickness of the surface layer, the graphitenodularity decreased in the casting centre. The two character-istics have a close relationship, as Fig. 7 shows. The data show

Fig. 6. Influence of the residual magnesium content, mould and coating type on the graphite characteristics in the centreof Mg-treated iron test castings [a, b–nodularity; c, d–shape factor; e, f–aspect ratio].

© 2012 ISIJ 1854

ISIJ International, Vol. 52 (2012), No. 10

that the assumed greater loss of Mg from the open surface inthe NRS test coupled with a slower fill rate consistently gavelower nodularity than that found with the FRS-PTSA test.

Residual magnesium level, the mould type and coatingconditions all affect other graphite phase characteristics inthe central area of the test castings, such as shape factor(Figs. 6(c) and 6(d)) and aspect ratio (Figs. 6(e) and 6(f)).Higher nodularity and shape factor and lower aspect ratio ofthe graphite phase represent a preferred ductile iron struc-ture. Higher magnesium content, higher shape factor andlower aspect ratio are at more favourable levels for FRS-PTSA type solidification, as is graphite nodularity. Similareffects of the applied coatings were found, with either neg-ative results using a sulphur bearing coating or positive with

an MgO-bearing coating. At lower magnesium residual inthe cast irons, there was a greater difference between resultsfor the FRS-PTSA and NRS moulds and uncoated and coat-ed moulds, respectively.

The graphs in Fig. 8 show the variation of the graphitecharacteristics as compactness (nodularity, shape factor,aspect ratio) across the section of the semi-cylindrical sam-ples (focused on 2 mm from the surface), solidified in FRS-PTSA moulds, depending on the final magnesium content inthe iron samples.

The parameters describing graphite characteristics arerelated to the thickness of the degenerate graphite layer.Considering the distance from the surface, where the maxi-mum level of each parameter was obtained, it typicallyoccurred between 0.1–1.1 mm: 1.0–1.1 mm from the surfacefor 0.02–0.03%Mgres and only 0.2 mm for the 0.054%Mgres

irons.Considering the structure in the surface layer, there is a

difference between low magnesium (0.02–0.03%) and highmagnesium (0.054%) treated irons. In low magnesium irons,the surface layer includes two sub-layers. Increasing theresidual magnesium led to a decrease in the first sub-layerthickness from 0.6 mm at 0.020%Mg, to 0.3 mm at0.027%Mg with the sub-layer disappearing at 0.054%Mg.The minimum degree of compactness for the graphite shapei.e. the lowest nodularity and shape factor, and the highestaspect ratio, was found at the boundary of these two sub-layers. Despite the graphite degeneration in the surface layer

Fig. 7. Graphite nodularity in the casting centre and relationship tosurface layer thickness.

Fig. 8. Influence of the resin sand-PTSA mould coating on variations in the semi-cylindrical sample section on graphitenodularity (a, b, c), shape factor (d, e, f) and aspect ratio (g, h, i) [a, d, g–0.020%Mgres; b, e, h–0.027%Mgres; c, f, i–0.054%Mgres].

ISIJ International, Vol. 52 (2012), No. 10

1855 © 2012 ISIJ

of these test castings to a lamellar morphology, the highergraphite compactness parameters at the metal to mould con-tact (the casting surface) were due to a higher cooling rate,because of the high heat transfer rate at the initial metal tomould contact.

4. Summary

•　The relative performance of coatings for furan resinsand moulds [P-Toluol Sulphonic Acid (PTSA) ashardener] [FRS-PTSA], was compared to no sulphur,Novolak resin coated sand [NRS] moulds byanalysing the degenerate graphite surface layer inMg-treated iron with 0.020 to 0.054%Mgres in the testcastings, and determining the graphite characteristicsover the entire section.

•　The extent of an abnormal surface layer is influencedby the different Mg contents for compacted or nodulargraphite irons, the sulphur content of the mouldingsystem and the type of mould coating (containing sul-phur or desulphurization materials).

•　Uncoated FRS-PTSA moulds with sulphur in thebinder promoted degenerate graphite in the surfacelayer of the test castings, with the thickness of thislayer increasing more than five times compared toNRS moulds (no mould sulphur), but also stronglydependent on residual magnesium content.

•　Sulphur bearing coatings increased the surface layerthickness by up to five times in NRS moulds and by50% in FRS-PTSA moulds. The layer was twice asthick at 0.027%Mgres and three times for 0.020%Mgres

irons, compared to 0.054%Mgres treated iron, due tointentionally added sulphur in the NRS moulds and byaugmenting the sulphur already present in the FRS-PTSA moulds.

•　The coatings based on desulphurizer type materials,such as MgO, provided some protection at the metalto mould interface. No graphite degeneration wasobserved in the NRS moulds, with 10–30% decreasein the layer at 0.020–0.027%Mgres and more than sev-enfold decrease in the layer at 0.054%Mgres in FRS-PTSA moulds. It is assumed that this coating behavedas a desulphurizer locally and countered the negativeeffect of sulphur released by the mould.

•　With lower Mg content, more graphite degenerationwas apparent in the cast surface layer, especially atless than 0.03%Mgres [typical Mg content for vermic-ular/compacted graphite cast irons] with increasingdifferences between FRS-PTSA and NRS moulds,and uncoated and coated moulds results.

•　The application and the type of coating appear to haveinfluence on the graphite nodularity even into the cen-ter of these castings. Nodularity decreased by up to10% with a sulphur bearing coating and increased byup to 20% with a MgO-bearing coating, with moreeffect at lower Mg content, compared to uncoatedmould conditions.

•　The graphite characteristics (nodularity, shape factor,aspect ratio) in the centre of the analyzed samplesevolved in a clear relationship with the changes in thedegenerate graphite surface layer, for the prevailing

solidification conditions.•　In low Mg irons, the surface layer contained two sub-

layers with different graphite characteristics. Increas-ing the Mg content led to a decrease in thickness ofthe first sub-layer from 0.6 mm at 0.020%Mg, to 0.3mm at 0.027%Mg and total elimination at 0.054%Mg.The degree of graphite shape compactness was at aminimum (the lowest nodularity and shape factor, andthe highest aspect ratio) at the boundary of these twosub-layers. The graphite compactness parameterswere higher at the direct metal to mould contact as aresult of the higher cooling rate in this area.

•　Comparing the graphite morphologies in irons solidi-fied against different mould coatings it appears thatapplication of the coatings mainly controlled thechemical interactions between S/Mg rather thanaffecting heat transfer.

•　A MgO-based coating appears to be efficient for highMg content ductile iron at 0.05%Mg, but does notprovide enough protection for low Mg compactedgraphite irons with <0.03%Mg. An effective coatingdesigned to protect these irons must go beyond therole of desulphurizing with supplementary Mg to pre-serve graphite nodularity. This dual activity isachieved with coatings containing active Mg, derivedfrom (Mg)FeSi materials.15–17)

AcknowledgementThe work has been funded by the Sectoral Operational

Programme Human Resources Development 2007-2013 ofthe Romanian Ministry of Labour, Family and SocialProtection through the Financial Agreement POSDRU/6/1.5/S/19. The authors would like to recognize and thankMichael Barstow for reviewing and editing this paper.

REFERENCES

1) A. Dahlouist: Mod. Cast., (2011), No.12, 20.2) DUCTILE IRON TECHNIQUES, Suggestions for Ductile Iron

Production. R.T.I.T Inc., Montreal, Canada, (1971).3) I. Riposan, M. Chisamera, S. Stan, C. Gadarauteanu and T. Skaland:

Proc. of 2003 Keith Millis Symp. on Ductile Cast Iron, DIS,Strongsville, OH, USA, (2003), 125.

4) M. Chisamera, I. Riposan, S. Stan, P. Toboc, T. Skaland and D.White: Proc. of 8th Int. Symp. on Sci. and Proc. of Cast Iron [SPCI-8], Tsinghua University Press, Beijing, China, (2006), 16; Int. J. CastMet. Res., 24 (2011), 28.

5) M. Chisamera, I. Riposan, S. Stan and M. Barstow: AFS Proc. 2012,American Foundry Society, Schaumburg, IL, USA, (2012), Paper 12-071.

6) J. Baier and M. Koppen: Manual of Casting Defects. Incidence andAvoidance of Defects Attributable to Molding Sands, IKO-Erbsloh,Marl, Germany, (1994), 32.

7) N. A. Golovan, K. J. A. Dudni and V.V. Dubrov: Russ. Cast. Prod.J., (1977), No.7, 35.

8) H. Xiaogan: AFS Trans., 100 (1992), 9.9) F. Marti and S. I. Karsay: AFS Trans., 87 (1979), 221.

10) W. Bauer: Giesserei-Praxis, (1982), No.11, 175.11) I. Riposan, M. Chisamera and S. Stan: China Foundry, 7 (2010), 163.12) I. Riposan, M. Chisamera, S. Stan and T. Skaland: Proc. of 8th Int.

Symp. Sci. and Proc. of Cast Iron [SPCI-8], Tsinghua UniversityPress, Beijing, China, (2006), 110; Tsinghua Sci. Techn. J., 13(2008), 157.

13) I. Riposan, M. Chisamera and S. Stan: Int. J. Metalcasting, accepted.14) S. Morita and N. Inoyama: AFS Cast. Met. Res. J., 5 (1969), 109.15) N. Ivan: PhD Thesis, POLITEHNICA University of Bucharest,

Romania, (2011).16) N. Ivan, M. Chisamera and I. Riposan: Mater. Sci. Technol., (2012),

accepted.17) N. Ivan, M. Chisamera and I. Riposan: Int. J. Metalcasting, accepted.