Room Temperature Mechanical Properties and Tribology of Nicralc by hcj


									 Room Temperature Mechanical Properties and Tribology of Nicralc and Stellite as Cast Alloys.
                Silva, W. S.1; Souza, R.M.2; Mello, J.D.B3; Goldenstein*, H. 1
            Universidade de São Paulo, Escola Politécnica, PMT –
                       Universidade de São Paulo, Escola Politécnica, PME.
            Universidade Federal de Uberlândia, Faculdade de Engenharia Mecânica.


        NICRALC is a family of intermetallic compound-based casting alloys proposed as a substitute
for abrasion and erosion-resistant cobalt alloys. NICRALC microstructures contain hard chromium
carbides dispersed in a continuous Ni-Al-Cr ' (L12) ordered intermetallic matrix. Five as-cast
NICRALC alloys with different carbon contents, different carbide types and morphological features
were compared with Stellite 6 alloy and high-chromium-white-cast-iron (HCWCI), with regard to their
mechanical and tribological behaviour at room temperature.
        Mechanical properties were evaluated by measurements of compression, hardness and fracture
toughness; tribological properties were evaluated by sand-rubber-wheel wear, reciprocating sliding
wear, block-on-ring lubricated sliding wear and erosion wear tests at room temperature.
        Yield stress, hardness and fracture toughness of NICRALC alloys were found to be dependent
upon carbon content. However, their wear behaviour was found not to be directly related to hardness,
yield stress and fracture toughness, but instead dependent upon the type and distribution of carbides.
Eutectic NICRALC alloys demonstrated wear resistances similar to those shown for STELLITE.


        The NICRALC family of foundry alloys combines a some characteristics of Ni-based
superalloys with features found in high-chromium-white-cast-irons (HCWCI); associating the
anomalous behaviour of the ordered intermetallic phase Ni3Al (’) (which increases in strength
following increases in temperature) with a dispersion of hard chromium carbides [1, 2, 3, 4]. The
motivation of the development of the NICRALC alloys is the search for an alternative to iron-based
wear resistant alloys at temperatures where they loose significantly their strength, as well as to search a
substitute for cobalt-based high temperature wear resistant alloys.
        The mechanical strength of intermetallic Ni3Al can be improved by the addition of hard phases
[5]. Bullock et al. [6], studying Ni-Ni3Al-Cr3C2 in situ directionally solidified composites, found that
the composite presented better creep resistance than the matrix-only material. Nemoto et al. [7],
studying polycrystalline Ni3(Al,Cr) alloys with low carbon content, observed fine dispersions of M23C6
carbides after solution annealing and aging treatments and these carbides increased the yield strength at
high temperatures. As-cast alloys developed following the NICRALC concept have been explored with
regard to their microstructures and solidification sequence [2, 3, 4], thermodynamic modeling of the
equilibrium phases [2, 3, 8], castability [4, 9] and mechanical behavior at different temperatures [4, 10].
Simultaneously, the tribology of spray-formed NICRALC alloys was studied [11] using the pin-on-disc
test method at room temperature.
        With regard to mechanical behavior of NICRALC alloys, previous works show that room
temperature yield strength is directly related to carbon content [4, 10]. Mechanical tests at different
temperatures showed an increase in yield strength up to temperatures between 600 and 800°C. The
maximum strength also increased with increasing carbon contents. The temperatures where maximum
strength were found decreases with the carbon content. Cast irons and STELLITE materials, however,
always demonstrated decreases in yield strength with increasing temperature [4, 10].

        Ni-based alloys have applications in various complex engineering systems, yet there is little
published data regarding the wear behaviour of these alloys. At room temperature, it was not possible
to establish some correlation between mechanical properties and the erosion resistance [12]. In this
case, the erosion wear of different materials of similar hardness can be correlated to the strain
hardening behaviour.
        Experiments with nickel-aluminide plasma spray coatings deposited over a Fe-based superalloy
showed that the impact erosion rate, measured at room temperature and at velocity of 40 m/s, was
determined by the ductile behavior of the coating, characterized by “platelet and lips mechanisms”
[13]. On the other hand, single impacts at high velocity (~300 m/s) caused a brittle response in
thermally sprayed Cr3C2–NiCr coatings, with both the carbide grains and matrix being cleaved by the
erodent particle [14]. Heat treatment of the coating significantly reduced this erosion rate, meaning that
the ductility of the alloys controls the erosion rate of the coatings under high velocity tests. High
temperature intermediate velocity erosion tests using the same materials studied in this paper showed
that NICRALC alloys were more wear resistant than STELLITE and HCWCI at high temperatures [4].
        Cavitation erosion experiments run at room temperature showed that the damage in NICRALC
alloys was concentrated in the disordered phase of the matrix, while in STELLITE alloys the damage
was concentrated at the interface between the matrix and Cr carbides [4]. The literature shows that the
cavitation and erosion resistance of STELLITE is related to a strain hardening behavior of the Co
matrix due to either twinning or strain-induced-shear phase transformation from FCC to HC martensite
        Pin-on-disk abrasion wear tests using SiC sandpaper disks of various abrasive grain sizes
showed that spray-formed NICRALC and as-cast NICRALC alloys showed wear resistance similar to
or better than as-cast HCWCI, despite their considerably lower hardness [11]. NICRALC alloys have a
hardness of 550 HV, while HCWCI has a hardness of 650 HV. The behaviour of NICRALC varied
with carbon content, carbides volume fraction and size, and, consequently with the hardness as well
        The sliding wear characteristics of superalloys and STELLITE are usually explained by two
main mechanisms [16]: (i) “oxidative wear”, in which the rate of wear is controlled by the kinetics of
oxidation as well as the thermomechanical properties of the oxides formed in conditions of mild wear;
ii) nucleation and propagation of cracks followed by chip removal during severe wear. Other authors
suggest that oxide film spalling is essentially a fatigue process, implying that there are critical
temperatures and contact loads above which the wear becomes severe, and oxidation is no longer the
controlling factor [17]. In the case of room temperature lubricated wear, the process is controlled by
contact characteristics, which depend upon surface roughness, fluid viscosity, contact loads as well as
the microstructure, mechanical properties and the relative speed of the bodies in contact [18].
        Some results of the literature show relations between fracture toughness and the wear resistance.
The abrasive wear of heat-treated steels is dependent on the size of abrasive particle and the mode II
fracture toughness of the wear surface. The wear rate becomes significant for low fracture toughness
values [19]. The sliding wear resistance of high-speed steel (HSS) is determined by the substructure,
the morphology, and the size and the type of carbides, all of which influence the initiation and
propagation of cracks [20]. The new wear resistant metal matrix composites produced by powder
metallurgy developed to present a bimodal dispersion of hard particles shows an improved fracture
toughness while maintaining a high level of wear resistance without the need to decrease the
mechanical strength [21]. The results obtained from spray formed NICRALC alloys tested with a pin-
on-disc abrasion test showed that mass loss decreases with increases in hardness [11], except for the
material with the highest carbide volume fraction, more refined carbide distribution and hardness. This
exception can be determined by the expected lower fracture toughness of this alloy.

Experimental Details

        This work aims to characterize the mechanical properties and the wear resistance of NICRALC
alloys, with different carbon contents, different amounts of carbides and different as-cast
microstructures, evaluated at room temperature. The work will also compare these obtained results to
those measured for conventional, industrial grade materials, including as-cast Co-based STELLITE 6
alloy and HCWCI, with a view toward replacing STELLITE as a non-magnetic bearing material, used
for submerged pumps used in oil extraction. The NICRALC alloys, STELLITE and HCWCI were
investigated by erosion, abrasion, sliding (dry reciprocating and lubricated FALEX test), compression,
indentation and chevron notched fracture toughness mechanics tests.
        The NICRALC alloys used in this work were designed using computational thermodynamic
tools developed previously [2, 3, 8]. The alloys studied were cast with different levels of carbon
content so as to obtain hypoeutectic, eutectic and hypereutectic as-cast alloys, each containing Cr
carbides (M7C3 and/or M3C2) distributed in a predominantly ’ (Ni3Al) matrix.
        Table 1 presents the chemical composition of the analyzed materials in this work. NICRALC
alloys C1, 05, 07, 10 and 13 as well as STELLITE 6 were obtained by investment casting under a flow
of Argon in an open induction furnace [9], while NICRALC OSP was spray-formed after being melted
under an Argon atmosphere [11]. Raw materials used include electrolytic Ni, commercial Al,
electrolytic Cr and graphite. Boron was added using Al-4.3%B as a master alloy. Pouring was effected
at 1550°C. The HCWCI was produced by conventional sand casting.
        NICRALC alloys 05, 07, 10 and 13 were tested after solution annealing at 1200°C for 24 h
followed by aging heat treatment at 760°C for 24 h. NICRALC C1, NICRALC OSP and STELLITE
were tested in as-cast condition. HCWCI was solution hardened at 950°C for 30 minutes and twice
tempered at 550°C for 1 hour.

Table 1 – Chemical Composition of Studied Alloys.
                                                   Weight %
                    C    Ni      Cr     Al     B     Si    Fe            Others
 NICRALC C1       0.840 Bal    12.180 7.630    -   2.290 1.44               -
  NICRALC 05      0.565 Bal     7.097 10.081 0.014 0.030    -           Mn=0.040
  NICRALC 07      0.666 Bal    11.075 9.207 0.039 0.045     -           Mn=0.040
  NICRALC 10      0.860 Bal    12.012 8.547 0.019 0.052     -           Mn=0.045
  NICRALC 13      1.312 Bal    11.100 9.284 0.045 0.047     -           Mn=0.040
 NICRALC OSP      0.500 Bal     8.000 9.440 0.060    -      -               -
  STELLITE 6      1.200 3.0    30.000    -     -   2.000 3.00 Co=Bal; W=4.5; Mo=1.5; Mn=2.0
    HCWCI         2.620 1.25   16.320    -     -   0.450 Bal             Mn=1.0
                                                                Mn=1.03; Mo=1.19; Nb=1.85

        Typical NICRALC as-cast structures are presented in figure 1. The NICRALC alloys have a
matrix essentially composed of ordered ’ with a complex dispersion of carbides. NICRALC C1 (fig. 1-
e) and 05 (fig. 1-a) are clearly hypoeutectic, while alloys 07 (fig. 1-b) and 10 (fig. 1-c) demonstrate
eutectic carbide distribution. These alloys also display rosette shaped /’ agglomerates [4]; and
NICRALC 13 is hypereutectic (fig. 1-d). Spray-formed NICRALC OSP has an ordered matrix showing
dispersed fine primary carbides, and islands of ’+  (black areas in figure 1-f).
        Moreover, in the as-cast condition NICRALC alloys display the coexistence of two different
carbides, frequently found as “duplex” double carbides (probably the result of the incomplete  + M7C3
 ’ + M3C2 reaction), as can be seen in figure 2. EDS microanalysis near the carbides indicate the
existence of a Cr-depleted, Al-enriched halo around the carbides, with a Ni/Al ratio approaching the

binary stequiometric ordered ’ composition. X-ray analysis of carbides extracted from NICRALC 05,
07, 10 e 13 [4] show the coexistence of M7C3 and M3C2 carbides; Rietveld analysis [4] was used to
determine the predominant carbide in each NICRALC alloy as summarized in Table 2.

                          (a)                                        (b)

                          (c)                                        (d)

                          (e)                                         (f)
Fig. 1 – Optical images of alloys microstructures, showing the chromium carbides morphologies and
matrix. (a) NICRALC 05; (b) NICRALC 07; (c) NICRALC 10; (d) NICRALC 13; (e) NICRALC C1;
                        (f) NICRALC OSP (dark areas are formed by ’+).

        Carbides present in the STELLITE and HCWCI were mostly M7C3, with small amounts of MC
[4]; this carbide corresponds to WC in the STELLITE and NbC in the case of the HCWCI. Typical
microstructure is shown in figure 3.



                              ’ around the
                                              Ordered Matrix

  Fig. 2 – As-cast microstructure of typical NICRALC alloy showing the double eutectic carbide
type Cr7C3 and type M3C2 as well as the ’ around the carbides. SEM-BSE Image. Marble etched.

Table 2 – Carbides characteristics of the studied alloys [4].
       Alloy           Carbides Type         Extracted Carbides (Vv %)       Microstructure
   NICRALC 05 M7C3 (>85%) & M3C2                        4.86                  Hypoeutectic
   NICRALC 07 M7C3 (>85%) & M3C2                        6.79                  Near-Eutectic
   NICRALC 10 M7C3 (>85%) & M3C2                        6.38                    Eutectic
   NICRALC 13 M3C2 (>90%) & M7C3                        7.40                  Hypereutectic
  NICRALC C1           M7C3 & M3C2                        -                   Hypoeutectic
  NICRALCOSP           M7C3 & M3C2                 3.0 +/-1.2 [11]                  -
    STELLITE            M7C3 & MC                         -                   Hypoeutectic
      HCWCI             M7C3 & MC                         -                   Hypoeutectic

Fig. 3 – Optical images of alloys microstructures, showing the chromium carbides morphologies
          and matrix. (a) STELLITE 6 (Marble etching); (b) HCWCI (NITAL etching).

Mechanical Properties Tests
        Cylindrical compression specimens of NICRALC alloys, 5 mm in diameter and 5 mm in length
and prismatic compression specimens of STELLITE and HCWCI, 3.5 x 3.5 mm2 square and 5 mm in
length, were conventionally machined. Compression tests were performed with a computer-controlled

universal testing machine at a strain rate of 6.7x10-3 s-1. Tests were run at room temperature in air. The
offset compressive yield stress was measured at 0.2% plastic strain.
        Fracture toughness was measured using the chevron notch method, following prior results [22,
23, 24]. This method greatly simplifies the determination of the fracture toughness of high strength
materials, following mathematical and experimental works performed in the late 1970’s and early
1980’s [25], which made possible the determination of the fracture toughness values from the
maximum load measured during testing [26].
        The Young modulus and matrix hardness was determined by instrumented micro-hardness,
using the Oliver-Pharr [27] model. During the tests was used normal load of 0.98 N.

Wear Tests

        Hard particle erosion tests were performed on a stationary fluidized bed developed by Kunioshi
[28], driven by room temperature atmospheric air compressed at 2.4 bar for 300 minutes. The hard
particles used were electro-melted, faceted Al2O3, with sizes between 150 and 212 m. Impact angles
may be approximated at 90°, and the maximum velocity was 7.5 m/s. Specimen dimensions were 2.0 x
30.0 x 9.0 mm3 and their surfaces were standardized by grinding at 600 meshes, such that the initial
surface average roughness (Ra) was close to 0.10 m.
        Sand-rubber-wheel abrasion tests were performed in equipment developed by Ramos [29], with
the normal load kept at 130 N using a servomotor. The tangential forces, the number of rotations and
the temperature were all monitored throughout the test. The wear tests were run for 600 s using a wheel
speed of 200 rpm and abrasive quartz foundry-sand according to specifications AFS 100. For each set
of conditions, three runs were performed, each for a length of 2,800 m. All samples were finished by
grinding with a 120# abrasive wheel. Samples with 10.5 x 21.0 x 42.0 mm3 were fixed to a steel
specimen holder so as to conform to the specifications of [29].
        The lubricated wear tests were performed using FALEX equipment (block-on-ring
configuration). Parameters used for this wear include the following: a) lubricating oil AGECOM 60 PS
with viscosity of 4.8 x 10-3 Pa.s at 100ºC; b) testing temperature maintained at 100°C; testing speed
400 rpm; d) normal load 600 N; e) 20000 cycles, corresponding to a sliding distance of 2200 m.
        Reciprocating sliding tests were performed on a Plint machine, using the following parameters:
a) steel sphere counter bodies made of quenched and tempered SAE 52100, 4.95 mm in diameter, and
WC-Co sphere counter bodies, 4.76 mm in diameter; b) sliding frequency of 6.0 Hz; sliding amplitude
of 6.0 mm with normal load (dead weight) of 9.806 N; c) 60 minutes test time, amounting to a total
sliding distance of 260 m; d) room temperature and without any lubricant; e) room relative humidity
was kept below 50%; f) prior to testing, samples were ground with 220 grinding paper for 5 min,
leaving surface roughness (Ra) around 0.10 m. This specific test was run only with STELLITE,
NICRALC C1 and NICRALC OSP, all in their as-cast condition [30].
        Debris formed both during reciprocating sliding and lubricated block-on-ring tests were
collected (by filtering the lubricant in the case of block-on-ring) and analyzed in the SEM-EDS, after
being cleaned and coated with Au.


        Table 3 summarizes the mechanical properties results at room temperature. Fracture toughness
for all materials increases with decreasing microhardness in a manner similar to correlations found for
tool steels [22, 31]. STELLITE, compared to NICRALC 05, 07 and 10, showed a higher microhardness
corresponding to a lower toughness; the results for HCWCI show an even stronger correlation [31].
The only exception was the spray-formed NICRALC OSP, probably due to heterogeneities introduced

by the fabrication route and the refined microstructure and carbide distribution, and NICRALC 13,
which the predominance of brittle M3C2 carbides, instead of the tougher M7C3, promoted a higher
damage concentration during fracture.
        Room temperature yield strength for NICRALC alloys tested in compression increased
monotonically with carbon content and carbide volume fraction, except alloy NICRALC 10; carbide
morphology and distribution did not play an important role in this case, but recent work with
directionally solidified samples showed significant improvement over conventionally cast alloys when
samples were tested with loads parallel to the direction of solidification [4].

Table 3– Summary of room temperature mechanical properties from compression, instrumented
microhardness and chevron notch fracture toughness tests
   Liga      Matrix Hardness - HV0.01 [GPa]                        E [GPa]                YS [MPa]         KICV [MPa . m1/2]
  Nicralc C1          5.5 +/- 0.18                             235. 24 +/- 1.48              610.0              55.2 +/- 3.1
Nicralc OSP          4.13 +/- 0.01                              225.39 +/- 4.49              948.3              35.5 +/- 3.6
STELLITE 6           7.07 +/- 0.37                             234.47 +/- 12.13              927.0              39.1 +/- 0.8
    HCWCI            6.13 +/- 0.09                              211.72 +/- 3.70             2.185.0             21.4+/- 0.3
  Nicralc 05         3.69 +/- 0.11                              222.13 +/- 3.40              507.8             65.8+/- 3.4 *
  Nicralc 07               -                                           -                     576.1              52.6+/- 4.5
  Nicralc 10         3.45 +/- 0.08                              216.94 +/- 3.40              520.6                59.8**
  Nicralc 13         4.99 +/- 0.35                              235.82 +/- 9.12             1021.9              49.6+/- 4.9
* Values of fracture toughness for alloy 05 are conditional only, because the results did not satisfy all requisites of Linear Elastic
Fracture Mechanics as defined by ASTM E-1304 97.
** Individual valid result.

        The results of wear resistance tests are summarized in table 4. Generally, there is no direct
relationship between the volume loss during wear tests and the hardness and other mechanical
properties of the NICRALC alloys. Wear resistance appear to be more directly related to phase
morphologies and distributions, mainly the eutectic (for alloys NICRALC C1, 05, 07 e 10) and primary
hypereutectic carbides (for alloy NICRALC 13). The best results regarding wear resistance were
obtained from alloys NICRALC 07 and 10, both with typical eutectic carbide morphology
homogeneously distributed.

Table 4 – Summary of the wear tests results obtained in the erosion, abrasion, FALEX and
reciprocating sliding tests.
                                                                            Reciprocating Sliding
                                 Erosion                 Abrasion                                           Lubricated Sliding
                      Hardness                                               Volume Loss [10-
      Alloys                   Volume Loss             Volume Loss                4                         Volume Loss [10-
                       [HV0.5]                                                      .mm3/m]
                                [mm3/m2]               [10-3mm3/m]                                              6.
                                                                            WC-Co        SAE 52100
  NICRALC C1            426.1            0.84               19.2             10.0          37.0                      50.5
 NICRALC OSP            449.4            3.85               43.6             12.0          24.0                     351.5
    STELLITE            500.9            0.96               9.2               9.0           1.0                       4.2
       HCWCI            741.3            4.46               1.9                        -                             23.8
  NICRALC 05            407.9            3.04               22.3                       -                            507.7
   NICRALC07            462.1            3.84               10.7                       -                             83.1
  NICRALC 10            456.7            0.93               7.4                        -                             35.8
  NICRALC 13            482.2            4.27               15.1                       -                            382.3

       Table 5 presents the friction coefficient obtained for each of the different wear tests. For the
block-on-ring (Falex) test, the friction coefficients obtained were in the range predicted by the IRG

diagram [32] for boundary lubrication regime, which is consistent with the test conditions. The higher
friction coefficients were associated with bigger mass losses due to the sliding for the NICRALC
alloys. For the reciprocating sliding test the friction coefficient varied significantly based on type of
counter-body used: average values measured for SAE 52100 steel showed to be larger than those
measured for WC-Co. In the sliding tests, the relationship between mass loss and friction shows that
the wear depends on the matrix behaviour and the carbide distribution, so wear is determined by the
adherence mechanisms. On the other hand, in the abrasion tests, there was little or no correlation
between friction coefficient and mass loss for the wear tests. In this case, the severity of the wear turns
negligible the friction mass loss relation.

Tabela 5 – Friction values measured for abrasion, lubricated sliding (FALEX) and reciprocating
sliding wear tests.
      Alloys                         Reciprocating Sliding         Lubricated
                                     WC-Co          SAE 52100       Sliding
  NICRALC C1        0.35 +/- 0.01 0.46 +/- 0.04 0.56 +/- 0.06 0.11 +/- 0.010
 NICRALC OSP        0.43 +/- 0.01 0.45 +/- 0.04 0.53 +/- 0.05 0.13 +/- 0.005
    STELLITE        0.35 +/- 0.01 0.36 +/- 0.05 0.56 +/- 0.09 0.09 +/- 0.005
   FOFO CONV        0.40 +/- 0.01       -               -        0.12 +/- 0.005
   NICRALC 05       0.38 +/- 0.01       -               -        0.15 +/- 0.010
   NICRALC 07       0.39 +/- 0.01       -               -              -
   NICRALC 10       0.39 +/- 0.01       -               -        0.10 +/- 0.010
   NICRALC 13       0.35 +/- 0.01       -               -        0.16 +/- 0.005


Mechanical Properties

        The yield strength behaviour of the NICRALC alloys is determined by the carbon amount and
the volume fraction of carbides. In this case the mechanical properties increased with the carbon
content. The NICRALC OSP is an exception as it shows higher yield strength, despite its lower carbon
content. In this case, it is reasonable to attribute this behaviour to a more refined carbide microstructure
and grain refinement.
        In general, the fracture toughness of each of the alloys studied was inversely proportional to
hardness and yield strengths. Fracture toughness was controlled by the distribution of second (hard)
phases; for NICRALC C1 and STELLITE, the continuous interdendritic distribution of carbides
provided a preferential path for crack growth (figures 4-a and 4-b). In those cases, the fracture
toughness was defined by both deviations in the crack growth path and by the intrinsic toughness of the
carbide phases [31].
        In alloys where the carbides are more heterogeneously distributed throughout the matrix, such
as NICRALC 05 and 13, the matrix characteristics dominate. The finer and more homogeneously
distributed ’ ordered domains in the matrix improved the higher fracture toughness of alloy NICRALC
05 (figure 4-c), while the larger carbides M3C2 (with their accompanying stress concentration) and
coarser ordered domains in NICRALC 13 (figure 4-d) reduced its toughness. Furthermore, there is
published evidence [30], that the M3C2, carbides are more brittle than M7C3. The micrograph pictured
in figure 4-d clearly shows that the carbides in NICRALC 13 suffer cleavage not only along the main
fracture line but along the plastic zone as well, resulting in the creation of secondary cracks.

                         Ordered Matrix                       M7C3


                                                             Co-rich matrix

                              (a)                                         (b)

          Secondary crack growth by                          Cracked M3C2 Carbide
            interdendritic carbides

                         Pro-eutectic Matrix                             Crack growth by
                                                                       disordered boundary

                              (c)                                         (d)
Fig. 4 – Crack path in the NICRALC C1 (a), STELLITE (b), NICRALC 05 (c) and NICRALC 13
          (d) after chevron fracture toughness tests. SEM-BSE Images. Marble etching.

Wear Tests
        The results from the lubricated sliding block-on-ring test (FALEX) showed that phase
distribution, not hardness, was the main determinant of rate of wear. NICRALC alloys generally
suffered concentrated damage in the matrix (figure 5-a and 5-c), while the reference alloy STELLITE
suffered damage at the carbide-matrix interface (figure 5-b). Figure 5-a also shows cracks formed on
M3C2-type carbides. In the case of alloy NICRALC 07 (figure 5-c) wear concentrated heavily on
carbide-free regions, mainly in the disordered  films surrounding ’ ordered domains. Similar results
were found in cavitation-erosion tests [4]. The areas around the carbides (’ halos) shows no wear
damage. This behaviour can be explained as protection by the carbides associated to the absence of
ordered/disordered phase boundaries. No evidence of micro-cracks formation at the matrix was
observed. The dominant wear mechanism was determined to be plastic collapse due to adhesion
followed by failure of the adhered patch.
        Co-based STELLITE was supposed to offer wear resistance superior to that offered by
NICRALC in both block-on-ring FALEX and reciprocating slide tests due to a tribochemical reaction
between the Ni3Al phase and the steel counter bodies. However, subsequent SEM-EDS analysis of the
debris collected from both tests caused this hypothesis to be discarded. Figure 6 shows grooves and
gouges caused by hard particles abrading the worn surface. EDS analysis of these particles suggested a
complex structure formed by a mixture of carbides, oxides and matrix matter. There was no evidence of
damage on the layers below the surface. Alloys with eutectic or near-eutectic microstructures with
more homogeneous carbide distribution demonstrated improved performance in the block-on-ring
lubricated sliding test, a result analogous to the findings of the sand-rubber-wheel abrasion test.
                                                                     Damage concentrated at Matrix
                                                                          Carbides Interface

                                  M3C2 Cracks

                  Worn matrix
                                                                             Worn matrix

                           (a)                                               (b)


                                                  ’+ 


     Fig. 5 - Wear micromechanisms on the worn surface in lubricated sliding tests (FALEX).
 (a) NICRALC C1; (b) STELLITE; (c) NICRALC 07. Sliding direction is from the bottom to the
        Figure 7 shows the strong presence of Fe in the EDS analysis of debris filtered out of the
lubricating oil. There is also a small Al peak, possibly an indication of the formation of Fe aluminates
through a tribochemical reaction similar to mechanical synthesis. However, this hypothesis is
contradicted by the fact that smaller volume losses of the block specimen are associated with higher
absolute mass loss of the steel bearing ring counter body. The only exception to this trend is alloy
NICRALC 13, in which a large loss of volume of the block correlated to even greater mass loss in the
ring. This anomaly was most likely due to the strongly abrasive effect of the large hypereutectic M3C2
type carbides in this particular alloy.
        Results obtained in the reciprocating sliding wear tests show a strong dependence on the type of
counter body used. In table 4 it can be seen that NICRALC alloys experience significant wear during
the use of SAE 52100 spheres, while STELLITE suffered no wear at all. When WC-Co spheres were
used, there was almost no difference between the level of wear suffered by the two materials. Two
distinct phenomena can explain this paradox: first, the worn surface of STELLITE tested against SAE
52100 (figure 8-c) shows some indications of strain-induced transformation (CoCo) or twinning, as
described in the literature [15]. The increase in hardness caused by this transformation could help to
inhibit wear in the STELLITE by concentrating the damage on the steel spheres. The fact that an
increased amount of iron was found on the collected debris from the STELLITE-SAE 52100
experiments (figure 9-c), as compared to that found on the debris that originated from the NICRALC-
SAE 52100 (figure 9-b). Experiments seem to confirm this hypothesis.

  FIG. 6 – Worn surface after lubricated sliding (FALEX) test of NICRALC 07 alloy. Abrasion
particle determined by EDS to be a mixture of oxides, matrix debris and carbides, responsible for
                     micro cuts on the surface. Vertical direction of sliding.

 Fig. 7 - Worn surface debris from lubricated sliding (FALEX) (left) and EDS spectrum (right)
                         that shows the iron and aluminum presence.

        Secondarily, there is no sign of twining or phase transformation in the reciprocating sliding
experiments of STELLITE against WC-Co. In this case, the surface of the STELLITE most likely
wears too quickly for a strain-induced transformation to occur. The same phenomenon has been
attributed to similar STELLITE subjected to long term cavitation-erosion conditions [15].
        The presence of iron in the results of the EDS analysis of debris from NICRALC/ SAE52100
experiments (figure 9-a) could mean that a tribochemical reaction is again responsible for the
accelerated wear. However, EDS analysis of debris from the NICRALC/WC-Co experiments (figure 9-
c) show an Al peak intensity matching that found for NICRALC/SAE52100 (figure 9-b). Even if there
is some tribochemical reaction, it is not significantly responsible for the worst result found for the
NICRALC/ SAE52100 contact. And the presence of O and Cr in the results of the EDS analysis of
debris (figure 9), also supports the hypothesis that the formation and spalling of oxide films both play
an important role in the operating wear mechanism, as seen in figure 8, although the test is conducted
at room temperature, the microscopic contact region of contact between the body and counter body
would most likely reaches high temperatures as the test progressed. The wear resistance of both
STELLITE and NICRALC as demonstrated in this test is heavily dependent upon the nature of the
counter body and on the formation of oxides.


                          Al and Cr Oxides
                                                                                  Cr Oxides

                           (a)                                              (b)

                                 Fresh Matrix
                                                         Cr Carbides

                                 Slip Planes

                                     Worn Damage

Fig. 8 – Reciprocating Sliding worn surface. (a) NICRALC C1/WC-Co; (b) STELLITE/WC-Co;
  (c) STELLITE/SAE52100, where can be seen evidences for strain induced twinning or phase
            transformation, a strengthening mechanism that can explain the results.

        STELLITE, which may be strain-hardened through a phase transformation or a twinning
mechanism, and HCWCI, which demonstrated higher initial hardness, both showed less weight loss
during the abrasive wear test. NICRALC alloys showed a similar trend to that described in previous
tests; with wear rates depending more on the distribution and type of carbides than on the strength and
hardness of the alloy. NICRALC 05 alloy, hypoeutectic, with smaller hardness and a lower volume
fraction of carbides, and NICRALC 13, hypereutectic alloy with higher hardness and a higher volume
fraction of M3C2 carbides (table 3) both suffered elevated mass losses. The alloys NICRALC 07, 10
and C1, with intermediate hardness, carbon content and carbide volume fraction, with near-eutectic
microstructures, showed less mass loss during wear test.
        In terms of the sand-rubber-whell tests, micrographs of the worn STELLITE sample (figure 10-
a) shows more homogeneous abrasion than that was seen for the NICRALC alloys (figure 10-b). The
worn surface of NICRALC alloys shows carbides in high relief, characterizing the lower wear
resistance of the NICRALC matrix with respect to the STELLITE matrix. As can be seen from the
worn surface of NICRALC 07 alloy (figure 10-b), the homogeneous distribution of carbides gives this
alloy and the NICRALC 10 increased wear resistance similar to that shown for STELLITE, and
significantly improved over the other NICRALC alloys. In the micrograph for alloy 13 (figure 10-d), it
is possible to see fracture and decohesion at the interface of the more fragile M3C2 carbides, signal of
evidence of the wear acceleration.
        The sand-rubber-wheel abrasion tests results obtained in this work are very different from those
obtained by pin-on-disk abrasive wear results for both cast and spray-formed NICRALC alloys,
published early [11]. In those experiments NICRALC showed wear resistance greater than or similar to

HCWCI, when tested as pins against SiC sandpaper disks, a fact which provides evidences of the
strong influence of the tribological system is evidenced. Sand-rubber-wheel abrasion tests, in which
silica particles (softer than some of the carbides) are free to roll, present conditions much different from
those of the pin-on-disk wear test, in which much harder SiC particles are glued to the disk paper.
Additionally, the geometry of the pin-on-disk test is such that the pin must pass many times over the
same spot on the disk [33].

                           (a)                                                  (b)

                           (c)                                                  (d)
Fig. 9 – SEM/EDS of wear residues from alternating sliding experiments without lubrication. (a)
NICRALC C1/AISI 52100 steel; (b) NICRALC C1/WC-Co; (c) STELLITE/AISI 52100 steel; (d)
              STELLITE / WC-Co. The Fe peaks are indicated by the arrows.

        Results for the erosion tests were consistent with the other tests in that the wear resistance of the
NICRALC alloys was heavily dependent on the carbide distribution and carbide type present. Alloys
NICRALC 10 (eutectic) and C1 showed the best results and similar to STELLITE behaviour,
confirming the poor wear resistance of the +’ matrix. Finer carbide distribution provided better
protection for the matrix, thus minimizing wear. Figure 11 shows the eroded surfaces of some
materials tested in this work, further demonstrating that the main erosion mechanism is plastic
deformation of the matrix. Due to the relatively small energy of the individual impacts of abrasive
particles on the surface studied in this work, the craters described in the literature [13] are not evident.
It is possible to see mechanisms of plateau formation (plate-like), lips and cracks in all materials.
Cracks evidences are indicated by arrows. These cracks link wear with crack propagation, a fact which
could help explain the high wear rate of the HCWCI, a low fracture toughness material.

                  Stellite Worn Surface                                  Eutectic Carbides

                                                             Larger M7C3 carbide

                                                                      Eutectic Carbides

                             (a)                                           (b)

                                    M3C2 Carbides

                                   Worn Matrix                       M3C2
                                                                    Carbide        Worn Matrix

                             (c)                                           (d)
 FIGURE 10– Details of the worn surfaces after sand rubber wheel abrasion tests (a) STELLITE; (b)
   NICRALC 07; (c) NICRALC 13; (d) details of the damage suffered by M3C2 carbides in alloy
                          NICRALC 13. SEM-backscattered electrons.

        With regard to NICRALC the link between fracture toughness and loss of mass during erosion
tests is not as obvious as is in the case of HCWI. Alloys NICRALC C1 and NICRALC 10, with
intermediate fracture toughness were less eroded. Alloys as distinct as NICRALC 05 and 13
(respectively the larger and the smaller fracture toughness) presented mass losses of the same
magnitude, as did NICRALC 07 and NICRALC OSP, two alloys with distinct microstructures and
mechanical properties. It appears that the larger mass loss are associated with the smaller fracture
toughness, except for alloy NICRALC 5, but the determining factor is the carbide distribution , not the
fracture toughness.
        For the other wear tests it seems reasonable to assume that there is no correlation between
fracture toughness and wear rate, as wear occurs mainly by a cutting and deformation of the ordered
phase mechanisms, independent of the carbon content. There was some evidences of carbide cracking
during wear (figure 4-a), but no direct evidence that those cracks contribute to the overall wear rate.

                    Crack                                                      s

                            (a)                                             (b)


      Fig. 11 – Erosion wear surfaces. (a) NICRALC C1 alloy; (b) STELLITE; (c) HCWCI.

        Hypereutectic NICRALC 13 alloy, despite having greater volume fraction of carbides and
higher hardness, presented a higher mass loss than all other NICRALC alloys except NICRALC 5 in
wear as well as erosion and sliding tests. This result is related not only to the smaller fracture toughness
of the alloy and the intrinsic brittleness of the M3C2 type carbides, but mainly to the coarse distribution
of carbides, which leaves unprotected matrix areas unlike the eutectic alloys.

        As-cast NICRALC alloys with different C contents and different primary carbide distributions
(hypo, eutectic and hypereutectic), were compared with a spray-formed NICRALC alloy, an as-cast
STELLITE 6 and a conventionally as-cast high chromium white cast iron.
        The yield strength of NICRALC alloys was directly dependent of carbon content and volume
fraction of carbides. The spray-formed alloy is an exception, due mainly to its more refined
microstructure. The fracture toughness of NICRALC was inversely proportional to both hardness and
yield strength. The carbide distribution plays an important role, like in the case of the NICRALC OSP.
The carbide type has a role too, as shown in the case of NICRALC 13 alloy which contains
predominantly M3C2 carbides, weaker than M7C3.
        In general manner, STELLITE alloy showed higher wear resistance than NICRALC alloys at
room temperature. This result can be explained by the higher matrix hardness of STELLITE at room
temperature and the phase transformation induced strengthening of its matrix.
        In the reciprocating sliding tests of NICRALC and STELLITE the wear resistance varied with
the counter body material; tests performed using WC-Co spheres as counter bodies resulted in similar

results for both NICRALC and STELLITE alloys, while when using AISI 52100 steel spheres the
STELLITE was much more wear resistant.
        In NICRALC alloys wear damage is primarily controlled by the matrix resistance, while in
STELLITE wear is controlled by damage to the carbides.
        Under abrasion, erosion and lubricated sliding wear tests, the best results of NICRALC alloys
were obtained for the alloys which presented more homogeneous carbide distributions (eutectic alloys).
For those eutectic alloys the results obtained were similar to the results obtained for STELLITE.
        There is no evidence of a strict relation between fracture toughness and wear resistance of
NICRALC alloys. On the other hand, the lower fracture toughness can be associated with the higher
erosion mass loss of the HCWCI. For NICRALC the larger erosion mass loss was associated with the
worst fracture toughness (NICRALC OSP and NICRALC 13).

     The authors would like to tanks the financial support of FAPESP, CNPq and CAPES. IPT for
HCWCI donation and Professors Cláudio Kiminami, Claudemiro Bolfarini e Dr. Aroldo Mourisco (in
memoriam) for NICRALC OSP preparation.
     This work is dedicated to Dr. Mourisco .


[1] YOSHIMURA, H.N.; MATSUBARA, Y.; GOLDENSTEIN, H., “WCI – White Cast Intermetallic Compound – A New
     High Temperature Foundry Material”. Acta Microscopica. Vol.6, Supl. A, p. 174-175, September, 1997.
[2] GOLDENSTEIN, H., SILVA, Y.N., YOSHIMURA, H.N., “Designing a New Family of High Temperature Wear
     Resistant Alloys Based on Ni3Al IC - Experimental Results and Thermodynamic Modelling”, Intermetallics, 2001.
     NI”,Dissertação de Mestrado em Engenharia Metalúrgica e de Materiais, PMT/EPUSP, Julho de 2003, São Paulo/SP –
[4] Silva, W.S., Avaliação das Propriedades Mecânicas e Tribológicas de Ligas NICRALC, Tese de Doutorado, EPUSP
     Materials Science Engineering, Vol A 107, 1989, pp. 81.
[6] BULLOCK, E.; McLEAN, M.; MILES, D.E., “Creep behaviour of a Ni-Ni3Al-Cr3C2 eutectic composite”,. Acta Met.,
     v.25, p.333-344, 1977.
[7] NEMOTO, M.; TIAN, W.H.; HAN, C.S.; SANO, T., “Dislocation-particle interactions in precipitation strengthened
     Ni3Al and NiAl”. In: Structural Intermetallics, eds. R.Darolia et al, TMS, 1993. Proceedings. p.551-560.
[8] FARINA, A. B.; GOLDENSTEIN, H.; SILVA, W. S.; BENEDUCE NETO, F., “Progressos no desenvolvimento do
     banco de dados termodinâmicos para o sistema Ni-Cr-Al-C”. In: XVI CBECIMAT, 2004, Porto Alegre - RS. v. XVI, p.
[9] HISATO, M. K.,” Estudos de Fundibilidade de uma Liga NiCrAlC”. Trabalho de Formatura, PMT/EPUSP, 2005.
[10] Silva, W.S., Ferrandini, P.L., Caram, R., Goldenstein, H., Estudo da Resistência Mecânica de Ligas NICRALC em
     diferentes temperaturas, XVII CBECIMAT, Foz do Iguaçu, 2006
[11] MOURISCO, A.; SILVA, Y.N.; GOLDENSTEIN, H.; KIMINAMI, C.S.; BOLFARINI, C., “Microstructure of Wear
     Resistant Nickel Alloys Produced by Spray Forming”, In: EUROMAT’2001, 2001, Rimini. Proceedings of the
     EUROMAT 2001. Associazone Italiana di Metallurgia / FEMS, 2001.
[13] S.B. MISHRA A, K. CHANDRA A, S. PRAKASH A., B. VENKATARAMAN, “Characterisation and erosion
     behaviour of a plasma sprayed Ni3Al coating on a Fe-based superalloy”, Materials Letters 59 (2005) 3694 – 3698
[14] S.J. Matthews, B.J. James , M.M. Hyland, Microstructural influence on erosion behaviour of thermal spray coatings,
     Materials Characterization 58 (2007) 59–64.
     DURING CAVITATION EROSION OF A Co STAINLESS STEEL. Materials Science & Engineering. A, Structural
     Materials: properties, microstructure and processing, v. 358, n. 1-2, p. 199-204, 2003

     DE ARCO TRANSFERIDO, Dissertação de Mestrado – UFPR, 2003.
[17] So, H.; Chen, C. T.; Chen, Y. A. Wear behaviours of laser-clad stellite alloy 6. Wear, vol. 192, p. 78 – 84, 1996.
[18] MARU, M.M., Estudo do Desgaste e Atrito de um par metálico sob deslizamento lubrificado, Tese de Doutorado,
     EPUSP, 2003.
[19] I. Sevim, I. Eryurek, Effect of fracture toughness on abrasive wear resistance of steels, Materials and Design 27 (2006)
[20] Xu Liujie, Xing Jiandong, Wei Shizhong, Zhang Yongzhen, Long Rui, Investigation on wear behaviors of high-
     vanadium high-speed steel compared with high-chromium cast iron under rolling contact condition, Materials Science
     and Engineering A 434 (2006) 63–70
[21] Hans Berns, H., Comparison of wear resistant MMC and white cast iron, Wear 254 (2003) 47–54
[22] Silva, W.S., Estudo da tenacidade à fratura do aço rápido M2 fundido, modificado e tratado termicamente, Dissertação
     de Mestrado, EPUSP, 2001.
[23] MENDANHA, A., Tenacidade à fratura de aço para trabalho a frio, Dissertação de Mestrado, EPUSP, 2004.
[24] W.S. da Silva, J.T.N. Medeiros, H. Goldenstein, fracture toughness of as-cast, conventional and sintered from powder
high-speed tool steels, 7TH TOOLING CONFERENCE, Torino, 2006.
[25] Newman, J. C. "A Review of Chevron-Notched Fracture Specimens", in Chevron Notched Specimens – Testing and
     Stress Analysis. Editors: Underwood, Freiman & Barata, ASTM Committee E-24 on Fracture Testing, Louisville, Ky,
     21 April 1983.
[26] ASTM E1304 – 97, “Standard Test Method for Plane-Strain (Chevron-Notch) Fracture Toughness of Metallic
     Materials”, ASTM Committee E-24, Annual Book of ASTM Standards, 1989, p 962-972
[27] OLIVER W.C., PHARR, G.M., “A new improved technique for determining hardness and elastic modulus using load
     and sensing indentation experiments”, Journal of Materials Research, v7.n6, pp.1564-1582, 1992.
     DE NiCr COM WC E Cr3C2. Tese de Doutorado, IPEN-CNEN, 2003.
     EPUSP, 2005.
[30] Mello, J.D.B, Tribological Behavior of multi-component ferrous alloys, International Conference Abrasion 2005, São
     Paulo, Brazil, p. 19-45.
[31] BROECKMANN, C., “Fracture of Tool Steels on a Microscopical Scale”. Proc. Conf. Tooling – Progress in Tool
     Steels, 1996, p.491-499.
 [32] [31]GEE, A.W.J., BEGELINGER, A., SALOMON, G., “Failure mechanisms in sliding lubricated contacts”, in: Mixed
     lubrication and lubricated wear, Proceedings of the 11 Leeds-Lyon Symposium, England, p.108.
[33] SINATORA, A., ALBERTIN, E., Personal communication, Apud, (2002)


To top