The materials investigated were produced in a normal production process of El Nasr Castings Co. in Egypt; Fatahalla et al. [10] . Pig iron, commercially pure iron and other alloys were melted in a basic high frequency induction furnace with a melt capacity of 6 ton/hr. The normal charge material, which produced conventional DI, was subjected to two austempering treatments, which produced ADI. Additions of the following alloys: FeNi (95% Ni), FeMo (70% Mo), FeCr (70% Cr), and FeMn (70% Mn) were done to obtain LADI in the as-cast condition while the main addition alloy was the Nickel which was controlled at a level of 2 mass %. The melt was then superheated and held at 1773 K for 0.6 ks. It was then tapped into a preheated ladle at 1733 K with a pocket in the base (Sandwich Technique) [10] . It was then inoculated (16% by weight of charge) using Fe-Si-Mg alloy (45, 50 and 5 mass % respectively) having a grain size of 15 - 50 mm. thereafter, inoculated (0.6% by weight of charge) using Fe-Si alloy (80% Si) having a grain size of 0.2 - 3 mm. The inoculant was added to the stream of molten iron as the ladle was being filled for a period of 0.3 ks. The temperature of molten iron before treatment was measured to be 1693 K and after treatment was 1663 K for a period of 0.24 ks. An immersion thermocouple, digital meter stick was used to monitor the melt temperature in the induction furnace and pouring ladle. Spectrographic samples were cast in a copper mould for analysis by emission spectrograph. Table 1 shows the chemical compositions of the conventional DI and LADI.

After spheroidisation treatment, the melt was cast into a Y-block, using green sand moulds. All ingots of conventional DI and LADI, were cast into the same dimensions to maintain the same solidification cooling rate (SCR) and also the casting conditions. The conventional DI ingots were poured from a single preheated ladle at a pouring temperature in the range 1623 - 1643 K for 0.12 ks. The LADI was cast solely under almost the same conditions. Upon completion of pouring, all castings were allowed to cool overnight in the moulds before shake-

out and sandblasting.

It is recommended [1] that most of machining be carried out before heat treatment, depending on the ADI components which will have either a high hardness or high work-hardening rate. The conventional DI Y-block samples were machined into specimens with 0.5 mm finishing allowance. The specimens were cut from the bottom of Y-blocks while the top sections were not used in order to avoid variations in nodule characteristics and porosity. After heat treating, the specimens were machined to final dimensions. The LADI castings got their dimensions by sectioning and grinding in cold working operations.

The microstructural evaluations (including nodules and matrix characteristics) were conducted on small cubes, 20 mm side, taken from each Y-block of the as-cast and as-heat treated castings. Standard experimental techniques for metallography were used [11] . 3% nital was used as an etchant for 8 s to reveal the structural characteristics of the investigated alloys. Observation of the microstructures was carried out using Reichert Me F2 universal optical microscope.

Micro- and macro-hardness tests were conducted. The latter type was conducted using three indentation hardness tests, 1) Vickers, 2) Brinell, and 3) Rockwell. All the hardness tests were carried out at ambient temperature (300 K) using Otto Wolpert-Weker universal hardness testing machine. Standard microhardness Vickers test was applied on the specimens of each type to obtain the hardness of its matrix constituents. It was also carried out on longitudinal section perpendicular to the worn surface at different distances. Wolpert high quality digital microhardness tester set at 200 g load for 15 s was used. Preparation of samples for microhardness tests was the same as those for microstructure.

Tensile tests were performed in accordance with ASTM-E8-1990 and the round tensile specimens were used. The tensile tests were conducted on a motor driven tensiometer machine type W at ambient temperature (300 K) at a strain rate of 4 ´ 10⁻⁴ s⁻¹ up to fracture. Data collected included ultimate tensile strength (R_m), 0.2% proof stress (Rp_0.2), percentage elongation (% A₄) and toughness. Each value was the average of at least three successful tensile tests.

For Charpy impact tests, smooth (unnotched) specimens were used in accordance with ASTM-E23-1990. The capacities of Charpy impact testing machine type RKP300 were 300 J. Impact tests were performed at 300 K and all impact test results were averaged from, at least, three successful tests.

Wear-resistance tests were carried out with a pin-on-ring type apparatus using the TNO-Tribometer under dry sliding conditions in the ambient air. ADI specimens in the form of cylindrical pin of 8 mm diameter and 12 mm height were used. The pin was pressed against the circumference of the ring under a constant load of 265 N for 2.7 ks duration at a constant sliding distance of 2500 m. Mass losses were measured in grams before- and after wear tests with a precision to the fourth decimal by an analytical balance (Mettler). At least three successful tests were performed for each set of conditions and the average mass loss was calculated. After testing, the worn surfaces of pins were examined by optical microscope. Friction coefficient as F_n = normal force, and F_f = friction force was calculated.

Analysis of fracture surfaces illuminates the role of various microconstituents, and under different heat treatment conditions, in promoting or limiting tensile (static) and impact (dynamic) toughness. Caution was taken for appropriate handling of specimens after fracture, former to any fractographic examination. They were instantly examined, in the SEM, to avoid any accumulation of under humidity on the fracture surfaces. OM Axiovert 405 M-ZEISS was used to explain the fracture behaviour of the two types; tensile and impact fracture.

Nodule characteristics: As polished specimen-photographs were examined for nodule-characteristics. The differences among the nodule characteristics (count, size and nodularity) were, therefore, detected. Figure 1(a), Figure 1(b) illustrates, the as-polished microstructure for alloy 1 (conventional DI) and 6 (as-cast LADI), resprectively. Higher nodule count (87 nodule/mm²) and smaller nodule size (50 μm) was observed in Figure 1(a) for alloy 1 (DI) relative to that in Figure 1(b) (69 nodule/mm² and nodule size 62 μm) for alloy 6 (LADI). Comparing the two types of alloys 1 and 6 with ADI and as-treated (LADI) specimens, respectively, it was observed that, there is no remarkable change in the graphite nodule characteristics due to austempering heat treatment. Therefore, alloys 1 and 6 were taken to represent the other types of ADI and LADI specimens, respectively. The value of % nodularity of alloy 1; 95.5% and alloy 6; 93.5% are close to each other. These results can be attributed to the same SCR. These variations of nodule count may stem from the variation in chemical compositions between the two types of ductile-irons [9] .

Also, Jenkins [12] stated that, nodule count affects graphite size and shape. Increasing nodule count results in a decrease in nodule size, which improves tensile and fracture properties. The present result of % nodularity satisfies the ASM specification [13] , which indicates that for most purposes, a % nodularity of 85% - 100% should be in a fully nodular form. Additionally, these results are confirmed with the results obtained by Grech [14] . He concluded that, the morphology of graphite influences the final properties of an ADI. Whereas, for optimum mechanical properties, the % nodularity of the graphite in castings with a low nodule count should be as high as possible.

Figure 2(a) and Figure 2(b) illustrates the matrix micro-constituents after etching the specimens for as-castal- loys 1 and 6, respectively. The micostructure in Figure 2(a) is a typical bull’s eye type [11] with ferrite surrounding the graphite nodules. The microstructure delineates, generally, graphite nodules embedded in a ferritic- pearlitic matrix (48% pearlite). Figure 2(b) represents the microstructure of alloy 6 with Ni content 2.22 mass %.

The matrix micro-constituents are mainly pearlite (P), ledeburite (L), and carbide (C) (cf. photo) occurring in the inter-nodular regions. Graphite nodules in a matrix of higher volume fraction of ferrite (52%) provide a ductile- iron with good ductility, impact resistance and tensile strength, as will be discussed later [15] . The microstructure of alloy 1 resembles those obtained previously by Fatahalla [15] and Jenkins [12] . The latter [12] indicated that, increasing nodule count decreases the pearlite content, decreasing strength and increasing elongation. The presence of carbide in the structure of alloy 6 (cf. Figure 2(b)) confirms the results of K. M. Ibrahim et al. [16] who suggested that the low Ni-alloy additions lead to the formation of intercellular carbides. The previous investigators [16] indicated that repetition of production of bainitic ductile-iron in the as-cast condition based on Ni and Mo could be achieved. While, for the present LADI, Mn and Cr were added in the as-cast condition. Moreover, there was slight difference in volume fraction (V_f) of pearlite, which has a higher value than that of ledeburite and carbide in the matrix of alloy 6.

This maybe imputed to the presence of elements that favour pearlite and carbides in the composition, such as Mn and Cr. Figure 3 reveal the matrix constituents of the ADI after etching for alloys 2, 3 (Figure 3(a), Figure 3(b)) and Figure 4 for alloys 5, 6. Using micro-hardness testing, and comparing the micro-constituents, observed in Figure 3 and Figure 4, with those in reference [11] it could be deduced that the grey regions are the bainite while the RA is the bright phase. Lower bainite structure is observed in Figure 3(a) and Figure 4(a) for austempering temperature of 573 K, while upper bainite structure is observed in Figure 3(b) and Figure 4(b) for austempering temperature of 673 K. Reducing austenitising temperatures from 1203 K to 1143 K will refine the bainitic structure, as observed comparing Figure 3 and Figure 4 respectively which confirms the previous explanation [6] [15] . The transformation of austenite into ferrite and carbon saturated austenite during austempering of ductile-iron occurs by nucleation and growth process [6] [17] . Consequently, the microstructure of ADI produced by austempering process is strongly dependent on the transformation temperature. Due to the slow rate of carbon diffusion at low austempering temperature, the growth rate of the ferritic platelets will be rather slow. Moreover, these ferrite platelets will be fine in nature and called then lower bainite [5] as is shown in Figure 3(a) and Figure 4(a). On the other hand, at higher austempering temperature, the carbon diffusion rate is higher and, consequently, the growth rate of these ferritic platelets is rather rapid. Consequently, these ferrite platelets will be large or coarse in nature and is called the upper bainite. Microhardness measurements and comparison with literature [17] revealed the different phases in the present investigation.

The matrix micro-constituents for alloys 7 and 8 (LADI) are shown in Figure 5(a) and Figure 5(b), respectively. Figure 5(a) revealed the phases being; fine LB, acicular ferrite (AF), and RA. However, in Figure 5(b) coarse UB replaced the fine LB and additionally, carbideis observed in small areas as indicated by “C” in the lower-left region of Figure 5(b). Comparing Figure 5(a), Figure 5(b) for LADI with Figure 4(a), Figure 4(b) for ADI, it is noticed that acicular ferrite (AF) and carbide do exist in the latter (both are subjected to the same austenitising temperature of 1143 K and same austempering temperatures of 573 K for Figure 4(a) and Figure 5(a), and 673 K for Figure 4(b) and Figure 6(b). The present results shown in Figure 5 confirm that obtained previously by Bartosiewicz et al. [17] in which they observed coarse ferrite and, additionally, they revealed more austenite on the expense of ferrite at higher austempering temperature. Their conclusion is clearly noticed, in the present investigation, if we compare Figure 5(a) and Figure 5(b). It is also to be noticed that their austenitising temperature was 1130 K and their austempering temperatures were 533 and 658 K [17] .

Macrohardness: Table 3 presents a macrohardness conversion table for the 3 types of ductile-iron measured by three hardness-indentation methods; Vicker’s (HV), Brinell (HB) and Rockwell (HR_A) and (HR_C). The macrohardness values of ductile-iron investigated showed dependence on its matrix constituents. The lowest hardness is observed for alloy 1 (conventional DI) referring to ferritic-pearlitic matrix (cf. Figure 2(a)). This result can be explained since in the as-cast condition, the matrix would consist of varying proportions of pearlite and ferrite, and as the amount of pearlite increases, the macrohardness of the iron also increases [14] . Table 3 also, indicates a slight increase in hardness for ADI due to increasing the austenitising temperature from 1143 K (alloys 2, 3) to 1203 K (alloys 4, 5). The effect of austempering temperature is more pronounced in Table 3. Comparing the hardness of alloys 3, 5 with those of alloys 2, 4; a significant decrease is observed.

It is believed that the hardness of the UB observed in Figure 4(b) and Figure 5(b) together with the more amount of RA resulted in drastic decrease in hardness compared to Figure 4(a) and Figure 5(a). The fine structure of LB in Figure 4(a) and Figure 5(a) is believed to raise the hardness. The slight decrease in hardness for the LADI due to raising the austempering temperature refers to the increase in RA at 673 K relative to its amount at 573 K (cf. Figure 5(a), Figure 5(b)). The results of hardness, in the present investigation, confirms those obtained by Korichi Priestner [19] and Abouelala [20] with slight differences. It is to be taken into consideration that the slight differences in chemical composition, austenitising and austempering temperatures between the present and previous investigations are the cause of slight differences in results. Results in Table 3 for LADI specimens are supported by those obtained previously [21] . He indicated that the macrohardness values decrease with increasing the austempering temperatures. The lower hardness values are obtained at the austempering temperature of 673 K. The higher hardness of the as-cast LADI given in Table 3 (4770 HV) refers to the microstructure of Figure 2(b) revealing hard pearlite and carbide constituents. This extremely high hardness value of as-cast LADI confirms the results obtained previously by Velez et al. [22] . The present results of hardness values for alloys 7 and 8 (LADI) show consistency with the results of Volkov et al. [23] . They suggested that by increasing austempering temperatures from 572 K to 663 K, the Brinell hardness values decreased by about 16% to 20% on the average (This decrease per cent is 22% in the present investigation as detected from Table 3).

Microhardness: Table 4 lists the microhardness values of the different matrix micro-constituents for the 3 types of alloys under investigation. The hardness of the ferrite, and pearlite phases of alloy 1 confirmed the previous results [24] which showed values of 1700, 2900 MPa for the ferrite and pearlite respectively. J. Aranzabal et al. [25] indicated that, the relative microhardness values range of common micro-constituents in cast iron are: Ferrite 1000 - 2500 MPa, austenite 1000 - 4000 MPa, acicular ferrite 3000 - 5000 MPa, austempered (Bainite) 2500 - 6000 MPa and the carbide ranges from 10,000 to 20,000 MPa. The values of microhardness obtained in Table 4 for the different constituents fall within the indicated ranges of [25] .

Table 5 shows the average values of ultimate tensile strength (R_m), 0.2% proof stress (R_p0.2), maximum plastic strain (A₄ %) and toughness, which is calculated as the area under the tensile stress-strain diagram in MPa. Table 5 delineates that alloy 1 has the lowest R_m and R_p0.2; whereas, it exhibits highest A₄% and toughness “E” compared to ADI and LADI alloys. This high toughness of alloy 1 may stem from the microplasticity associated with ferritic-pearlitic matrix and good % nodularity [28] . One of the interesting results in the present investigation can be noticed from Table 5 being; comparable R_m, and R_p of ADI (alloys 3, 5) to values of LADI (alloy 8). Moreover, alloys 3, 5 (ADI) showed superior ductility and toughness compared to those of LADI (alloy 8). Adding to the previous advantages the high hardness of alloys 3, 5 throws the light on possibility of using this relatively cheap ADI replacing the more expensive LADI in industrial applications such as producing rolling mills and others. Furthermore, the present tensile results confirm those obtained previously [11] - [13] .

To evaluate the performance of the present alloys, the quality index (QI), which have been introduced by [12] implying both strength and elongation, is used:

where QI is a constant, and defined as Quality Index.

The calculated values of QI for the 3 types of ductile-iron are listed in Table 5. The highest values of QI is for alloy 3 (ADI); 119.9 and alloy 5 (ADI); 114.5 which, indicates a combination of higher strength and elongation and therefore, higher material performance. The segregation is known to reduce the toughness and ductility [6] . Table 5 which showed low ductility and toughness for the LADI suggests that the addition of alloying elements like Ni and Mo is not needed as far as the toughness is concerned.

Table 6 shows the impact fracture toughness (energy absorbed in J) of the 3 types investigated of ductile-iron. The highest value is observed for the conventional DI (alloy 1) and the lowest values refer to LADI (alloys 6, 7, and 8). For the sake of comparison between the energy absorbed until fracture in a static test (tensile) and a dynamic test (impact) the values of toughness (area under stress strain diagram in MPa) was multiplied by the impact-test-volume to get the toughness in Joule units. The values listed in Table 6 indicate that the energy required to fracture the tensile specimen (static test) is higher than that obtained by the impact specimen (dynamic test) for the 3 types of ductile-iron. This decrease in energy through impact testing reflects the high rate of loading in impact test compared to tensile test. It is also interesting to point out that although the impact toughness of LADI (alloys 7, 8) was relatively very low, however, its toughness showed values 10 times or more the impact toughness (cf. Table 6).

This phenomenon reflects the sensitivity of these LADI alloys to strain rate testing [30] . The present results of energy absorbed for ADI specimens can be explained by that obtained by Ratto et al. [31] . They indicated that, in the low range of austempering temperatures ductile-iron develops moderate toughness and low impact

Hardness units are MPa.

^*Toughness is the area under the curve of the tensile stress-strain diagram (dimensions of the tensile test specimen: gauge diameter = 4 mm, and gauge length = 20 mm).

^*E impact = Actual energy measured in impact test in J. ^**E = Area under tensile stress-strain diagram (MPa) × volume of impact specimen (0.010 × 0.010 × 0.055 cubic metre) (J).

strength. It is believed that, the low value of energy absorbed for alloy 6 (as cast LADI-cf. Table 6) referred to the effect of alloy additions; Ni, Mo, Mn, and Cr and its segregation, and the presence of carbides. Kobayashi [27] concluded that, the increase in Mn content resulted in dropping of impact value, due to increasing of the V_f of pearlite and segregation of Mn. Moreover, the carbides increase rapidly with increasing Mo addition, i.e.; the carbides and the segregation of Mo to cell boundaries cause significant loss of impact energy. Marrow and Çetinel [32] indicted that, the decline in impact toughness maybe due to the segregation of carbides around the ferrite/austenite grain boundaries as the austempering temperature exceeds 623 K. Authors of [33] noted that, the impact energy values are more sensitive to these structural changes than, for example, % elongation.

Wear characteristics: Table 7 lists the values of mass loss and coefficient of friction (μ) of the 3 types of iron;

DI, ADI, and LADI under investigation. Conventional DI (alloy 1) showed the highest value of mass loss and hence, the lowest wear resistance material. This may refer to the relatively high volume fraction of ferrite being 52% (cf. Figure 2(a)) and in consistence with the suggestion given before by Ping et al. [33] . In general, reviewing the results in Table 7 we can observe inverse proportionality between the values of hardness and those for mass loss. This confirms that wear resistance is a hardness dependent property in consistence with results and conclusion of [34] .

It can also be seen (cf. Table 7) that although the LADI (alloys 7, 8) showed the highest wear resistance, however, moderate wear resistance are observed for the ADI (alloys 2, 4). Therefore, LADI showed about twice higher values of wear resistance compared to ADI (alloys 2, 4), however, it is a compromise between required properties and alloy price. M. C. Jeng [34] stated that, the wear resistance absolutely depends on the hardness of the matrix structure, thus confirming the present results in Table 7. The coefficient of friction (µ) of alloy 1 is high because of the tendency to adhesion of soft ferrite matrix. The lower value of µ for ADI and LADI specimens maybe attributed to the difficulty of adhesion of the bainite matrix in ADI and hard carbide matrix in LADI. This phenomenon of decreasing “μ” is supported by Hemanth [35] , who reported that, it should be possible to reduce (µ) by increasing the flow stress of the base metal by alloying or heat treatment.

Work-hardened layers due to wear: The work-hardened layers in the different alloys of the present investigation are seen in Figure 6. It is noticed that the higher the microhardness value the less thickness of the work- hardened layer is observed. For instance, for the conventional DI-alloy 1-(lowest hardness) the thickness of the work-hardened layer is about 260 μm while for the LADI-alloy 7-(highest hardness) the thickness is about 100 μm. Other investigated alloys showed intermediate values of work-hardened layer thicknesses. It is also interesting to note that the work-hardened layer of the ADI (alloy 2) is more or less very similar in behaviour to that of the LADI (alloy 8) with a thickness of about 160 μm for both. Again it is suggested that ADI with special heat treatments (austenitising and austempering) can be a material of industrial use.

Fractographic study was carried out on selected-samples of the investigated alloys. For the sake of comparison alloys 2, 3 (ADI) and alloys 7, 8 (LADI) were selected for fractography since they were subjected to the same austenitising temperature of 1143 K and the same austempering temperatures (573, 673 K). Additionally, the as- cast DI (alloy 1) and as-cast LADI (alloy 6) fracture surfaces were investigated.

Failure in tension: Scanning electron microscopy was used to reveal the fracture surfaces for all investigated alloys (from Figure 7 to Figure 10). Figure 7 reveals the features of the fracture surface of the conventional DI (alloy 1). The ferrite areas, in the interior of the matrix, showed ductile patterns (symbol “F”). Sometimes small areas of dendritic lobes are also seen on the fracture surface (symbol “DL”). The graphite-nodules are pulled out leaving behind a vacant (symbol “V”). Graphite nodules (symbol “G”) are observed in their sites surrounded by ferrite. It is interesting to note that there is an observable clearance between the nodules and the surrounding ferrite (symbol “C”). This clearance “C” reflects the amount of plastic deformation that occurred prior to fracture. The cavities were generated due to decohesion occurred at graphite/ferrite interface at very low strain level due to the fact that the graphite nodules are weakly bonded to the ferrite matrix surrounding them [36] .

Figure 8 reveals the fractographs of ADI type (alloys 2, 3). Figure 8(a) shows cleavage pattern in the brittle area indicated, and small clearance “C” between the graphite-nodule and its cavity. These features reflect the relatively brittle mode of failure. Again vacant sites are seen “V”.

Small area of ductile features can be seen in the Figure 8(a) occurring in the RA regions. The phenomenon of white small particles on the surface of graphite nodules in DI was related to carbon precipitation during heat

treatment as was indicated by N. Fatahalla et al. [37] . The white precipitates on graphite nodules shown in Fig- ure 8(b) are due to the excess carbon of retained austenite in the matrix. The change of lower bainite to upper bainite due to raising the austempering temperature from 573 K to 673 K was reflected on the fracture characteristics in Figure 8(b) (alloy 3) relative to that observed in Figure 8(a) (alloy 2). Since the amount of RA was increased and, additionally, the ductility of upper bainite is higher than that of lower bainite, the ductile mode of fracture covered much larger areas on the fracture surface in Figure 8(b) relative to Figure 8(a). Brittle patterns areas are less than that in Figure 8(a) and it is to be noted that dimples can clearly be seen at the RA area neighbouring to nodule (cf. Figure 8(b)) reflecting moderate ductile mode of fracture in consistance with literature [38] .

Figure 9(a), Figure 9(b) reveals the fractography of the LADI (alloys 7, 8). Both photos indicate a brittle mode of fracture. However, the clearance between the graphite-nodule and its corresponding cavity is much smaller in Figure 9(a) (alloy 7) relative to that of Figure 9(b) (alloy 8). This corresponds to the higher hardness

and strength of alloy 7 relative to that of alloy 8. Features in Figure 9(a) reveal cleavage pattern and faceted regions (symbol “S”) indicating a brittle mode of fracture. In Figure 9(b) the features showed quasi-cleavage and pearlitic areas, and the clearance was relatively wider than that for alloy 7. The relatively larger amount of RA and the more ductile upper bainite do affect the fracture pattern. The effect of some alloying elements on the fracture behaviour of DI was previously investigated [39] supporting the present results.

Figure 10 shows the fracture surface of the as-cast LADI (alloy 6). Fragmentation of graphite can be observed, large areas of quasi-cleavage “Q” and facetted pattern “S”, and pearlite regions “P”. Very tight clearance “C” (indicated by a white arrow) between graphite-nodules and its corresponding cavities reflecting the brittleness of the alloy.

Failure in impact: Figure 11 shows the fracture surface of the conventional as-cast DI (alloy 1) after failure in impact testing (high rate of loading). The surface features still resemble a ductile mode of fracture. Smooth ripple pattern surrounding the graphite nodules in the ferrite regions is observed (cf. Figure 11). Additionally, areas of dimples are also shown in the figure. Again we can observe vacant sites “V” referring to pulled nodules and some fragmentation of graphite. The relatively large clearance between nodule and its cavity indicates the degree of plasticity prior to fracture, this result confirms that obtained previously by Fatahalla et al. [40] .

Figure 12 shows the fracture surfaces of ADI (alloys 2, 3) after failure in impact testing. Both photos reveal a mixed mode of fracture (ductile + brittle). However, in Figure 12(a), cleavage “CL” pattern is revealed in the upper portion (brittle characteristic) while smooth ripple pattern “R” is seen in the lower area of the same photo (ductile feature). In Figure 12(b) most of the fracture surface is characterised by a smooth ripple pattern “R” plus regions of dimples “D” reflecting higher ductility. The clearance “C” between the nodules and their corresponding cavities are much tolerated in Figure 12(b) rather than that observed in Figure 12(a), reflecting plastic-

ity and higher ductility in the latter. Fragmented graphite nodules are observed in both photos of Figure 12. Wolkov et al. [41] examined the behaviour of ADI under dynamic load in terms of the fracture mechanics. They found that, as “brittleness” is considered in terms of energy, they ought to classify high strength bainitic ductile- irons as brittle fracture materials.

Figure 13 shows the fracture surfaces of LADI (alloys 7, 8) fractured in impact. All features in both photos of Figure 13 reveal tendency to brittleness. However, Figure 13(b) reveals a bit of ductile fracture features such as

ripple “R” and dimple “D” patterns in the ductile features region of Figure 13(b). Although a small area in Fig- ure 13(a) (alloy 7) showed dimple pattern, however, most of the fracture surface is covered with cleavage “CL”, carbide “S”, and hard lower bainite regions. On the other hand, facets and carbide areas were also detected in Figure 13(b). The effect of high strain rate loading (impact) on the fracture phenomenon of LADI (alloys 7, 8) can be clearly seen if we compare the photos in Figure 13 with those in Figure 9 (tension).

Figure 14 shows the fracture surface of the as-cast LADI (alloy 6) after failure in an impact test. The river pattern “RV” of ausferrite in between two graphite nodules, can be seen. Also, quasi-cleavage features “Q” that occur on unspecified planes of pearlite and the flat facets “S” of carbide indicate the brittleness of the alloy. Tightness between the nodules and their corresponding cavities, i.e. small clearance “C” is another evidence of less plastic deformation prior to fracture. Alloy 6 (LADI) has better tensile strengths, macrohardness and lower impact energy than that of alloy 1 which refers to the blocky carbides located around the graphite nodules in the former. Brittleness and low toughness of alloy 6 (LADI) is believed to refer to segregation of the alloying elements; Ni, Mn, Mo, and Cr and formation of carbides that occurred in the matrix.