Gray Iron (part one)

Moderatore: Moderatore Area Tecnica

Avatar utente
Messaggi: 5773
Iscritto il: 31/03/2010, 18:26
Località: Milano - Chieti

Gray Iron (part one)

Messaggio da Aldebaran »

CAST IRONS are alloys of iron, carbon, and silicon in which more carbon is present than can be retained in solid solution
in austenite at the eutectic temperature. In gray cast iron, the carbon that exceeds the solubility in austenite precipitates as
flake graphite. Gray irons usually contain 2.5 to 4% C, 1 to 3% Si, and additions of manganese, depending on the desired
microstructure (as low as 0.1% Mn in ferritic gray irons and as high as 1.2% in pearlitics). Sulfur and phosphorus are also
present in small amounts as residual impurities.
Certain important but low-tonnage specialty items in this family of cast metals (notably the austenitic and other highly
alloyed gray irons) are not dealt with here; instead the emphasis is on the properties of gray irons used most often and in
the largest tonnages.
Classes of Gray Iron
A simple and convenient classification of the gray irons is found in ASTM specification A 48, which classifies the
various types in terms of tensile strength, expressed in ksi. The ASTM classification by no means connotes a scale of
ascending superiority from class 20 (minimum tensile strength of 140 MPa, or 20 ksi) to class 60 (minimum tensile
strength of 410 MPa, or 60 ksi). In many applications strength is not the major criterion for the choice of grade. For
example, for parts such as clutch plates and brake drums, where resistance to heat checking is important, low-strength
grades of iron are the superior performers. Similarly, in heat shock applications such as ingot or pig molds, a class 60 iron
would fail quickly, whereas good performance is shown by class 25 iron. In machine tools and other parts subject to
vibration, the better damping capacity of low-strength irons is often advantageous.
Generally, it can be assumed that the following properties of gray cast irons increase with increasing tensile strength from
class 20 to class 60:
· All strengths, including strength at elevated temperature
· Ability to be machined to a fine finish
· Modulus of elasticity
· Wear resistance

On the other hand, the following properties decrease with increasing tensile strength, so that low-strength irons often
perform better than high-strength irons when these properties are important:
· Machinability
· Resistance to thermal shock
· Damping capacity
· Ability to be cast in thin sections
Gray iron is used for many different types of parts in a very wide variety of machines and structures. Like parts made
from other metals and alloys, parts intended to be produced as gray iron castings must be evaluated for the specific
service conditions before being approved for production. Often a stress analysis of prototype castings helps establish the
appropriate class of gray iron as well as any proof test requirements or other acceptance criteria for production parts.
Successful production of a gray iron casting depends on the fluidity of the molten metal and on the cooling rate (which is
influenced by the minimum section thickness and on section thickness variations). Casting design is often described in
terms of section sensitivity. This is an attempt to correlate properties in critical sections of the casting with the combined
effects of composition and cooling rate. All these factors are interrelated and may be condensed into a single term,
castability, which for gray iron may be defined as the minimum section thickness that can be produced in a mold cavity
with given volume/area ratio and mechanical properties consistent with the type of iron being poured.
Fluidity. Scrap losses resulting from misruns, cold shuts, and round corners are often attributed to the lack of fluidity of
the metal being poured.
Mold conditions, pouring rate, and other process variables being equal, the fluidity of commercial gray irons depends
primarily on the amount of superheat above the freezing temperature (liquidus). As the total carbon (TC) content
decreases, the liquidus temperature increases, and the fluidity at a given pouring temperature therefore decreases. Fluidity
is commonly measured as the length of flow into a spiral-type fluidity test mold.
The significance of the relationships between fluidity, carbon content, and pouring temperature becomes apparent when it
is realized that the gradation in strength in the ASTM classification of gray iron is due in large part to differences in
carbon content (~3.60 to 3.80% for class 20; ~2.70 to 2.95% for class 60). The fluidity of these irons thus resolves into a
measure of the practical limits of maximum pouring temperature as opposed to the liquidus of the iron being poured.
These practical limits of maximum pouring temperature are largely determined by three factors:
· The ability of both mold and cores to withstand the impact of molten iron, an ability that decreases as
the pouring temperature increases, thereby favoring low pouring temperatures
· The fact that metal tap temperatures seldom exceed 1550 °C (2825 °F). Because ladling and reladling to
the point of pouring generally accounts for temperature losses of 55 to 85 °C (100 to 150 °F), the final
pouring temperatures seldom exceed 1450 to 1495 °C (2640 to 2720 °F), and in most instances
maximum pouring temperatures in the range 1410 to 1450 °C (2570 to 2640 °F) are considered more
· The necessity to control the overall thermal input to the mold in order to control the final desired
The usual microstructure of gray iron is a matrix of pearlite with graphite flakes dispersed throughout. Foundry practice
can be varied so that nucleation and growth of graphite flakes occur in a pattern that enhances the desired properties. The
amount, size, and distribution of graphite are important. Cooling that is too rapid may produce so-called chilled iron, in
which the excess carbon is found in the form of massive carbides. Cooling at intermediate rates can produce mottled iron,
in which carbon is present in the form of both primary cementite (iron carbide) and graphite. Very slow cooling of irons
that contain large percentages of silicon and carbon is likely to produce considerable ferrite and pearlite throughout the
matrix, together with coarse graphite flakes.
Type A flake graphite (random orientation) is preferred for most applications. In the intermediate flake sizes, type A flake
graphite is superior to other types in certain wear applications such as the cylinders of internal combustion engines. Type
B flake graphite (rosette pattern) is typical of fairly rapid cooling, such as is common with moderately thin sections (about
10 mm, or 3
in.) and along the surfaces of thicker sections, and sometimes results from poor inoculation. The large flakes
of type C flake graphite are typical of kish graphite that is formed in hypereutectic irons. These large flakes enhance
resistance to thermal shock by increasing thermal conductivity and decreasing elastic modulus. On the other hand, large
flakes are not conducive to good surface finishes on machined parts or to high strength or good impact resistance. The
small, randomly oriented interdendritic flakes in type D flake graphite promote a fine machined finish by minimizing
surface pitting, but it is difficult to obtain a pearlitic matrix with this type of graphite. Type D flake graphite may be
formed near rapidly cooled surfaces or in thin sections . Frequently, such graphite is surrounded by a ferrite matrix,
resulting in soft spots in the casting. Type E flake graphite is an interdendritic form, which has a preferred rather than a
random orientation. Unlike type D graphite, type E graphite can be associated with a pearlitic matrix and thus can produce
a casting whose wear properties are as good as those of a casting containing only type A graphite in a pearlitic matrix.

There are, of course, many applications in which flake type has no significance as long as the mechanical property
requirements are met.
Solidification of Gray Iron. In a hypereutectic gray iron, solidification begins with the precipitation of kish graphite
in the melt. Kish grows as large, straight, undistorted flakes or as very thick, lumpy flakes that tend to rise to the surface
of the melt because of their low relative density. When the temperature has been lowered sufficiently, the remaining
liquid solidifies as a eutectic structure of austenite and graphite. Generally, eutectic graphite is finer than kish graphite.
In hypoeutectic iron, solidification begins with the formation of proeutectic austenite dendrites. As the temperature falls,
the dendrites grow, and the carbon content of the remaining liquid increases. When the increasing carbon content and
decreasing temperature reach eutectic values, eutectic solidification begins. Eutectic growth from many different nuclei
proceeds along crystallization fronts that are approximately spherical. Ultimately, the eutectic cells meet and consume the
liquid remaining in the spaces between them. During eutectic solidification, the austenite in the eutectic becomes
continuous with the dendritic proeutectic austenite, and the structure can be described as a dispersion of graphite flakes in
austenite. After solidification, the eutectic cell structure and the proeutectic austenite dendrites cannot be distinguished
metallographically except by special etching or in strongly hypoeutectic iron.
With eutectic compositions, obviously, solidification takes place as the molten alloy is cooled through the normal eutectic
temperature range, but without the prior formation of a proeutectic constituent. During the solidification process, the
controlling factor remains the rate at which the solidification is proceeding. The rapid solidification favored by thin
section sizes or highly conductive molding media can result in undercooling. Undercooling can cause the solidification to
start at a temperature lower than the expected eutectic temperature for a given composition (Fig. 3). This can result in a
modification of the carbon form from A to E type or can completely suppress its formation and form primary carbides
Room-Temperature Structure. Upon cooling from the eutectic temperature, the austenite will decompose, first by precipitating some of the
dissolved carbon and then, at the eutectoid temperature, by undergoing
complete transformation. The actual products of the eutectoid transformation
depend on rate of cooling as well as on composition of the austenite, but
under normal conditions the austenite will transform either to pearlite or to
ferrite plus graphite.
Transformation to ferrite plus graphite is most likely to occur with slow
cooling rates, which allow more time for carbon migration within the
austenite; high silicon contents, which favor the formation of graphite rather
than cementite; high values of carbon equivalent; and the presence of fine
undercooled (type D) flake graphite. Graphite formed during decomposition
is deposited on the existing graphite flakes.
When carbon equivalent values are relatively low or when cooling rates are
relatively fast, the transformation to pearlite is favored. In some instances, the microstructure will contain all three
constituents: ferrite, pearlite, and graphite. With certain compositions, especially alloy gray irons, it is possible to produce
a martensitic matrix by oil quenching through the eutectoid transformation range or an austempered matrix by appropriate
isothermal treatment (Ref 1). These treatments are often done deliberately in a secondary heat treatment where high
strength or hardness is especially desired, such as in certain wear applications. The secondary heat treatment of gray iron
castings is of great value in producing components that must be hard when machining requirements prohibit the use of
components that are cast to final shape in white iron.
Reference cited in this section:
Reference cited in this section
1. M.D. VanMaldegiam and K.B. Rundman, On The Structure and Properties of Austempered Gray Cast Iron,
Trans. AFS, 1986, p 249
Section Sensitivity
Fig. 3 Undercooling from rapid cooling
of a eutectic composition
In practice, the minimum thickness of section in which any given class of gray iron may be poured is more likely to
depend on the cooling rate of the section than on the fluidity of the metal. For example, although a plate 300 mm (12 in.)
square by 6 mm (0.24 in.) thick can be poured in class 50 as well as in class 25 iron, the former casting would not be gray
iron because the cooling rate would be so rapid that massive carbides would be formed. Yet it is entirely feasible to use
class 50 iron for a diesel engine cylinder head that has predominantly 6 mm (0.24 in.) wall sections in the water jackets
above the firing deck. This is simply because the cooling rate of the cylinder head is reduced by the "mass effect"
resulting from enclosed cores and the proximity (often less than 12 mm, or 0.47 in.) of one 6 mm (0.24 in.) wall to the
other. Thus the shape of the casting has an important bearing on the choice of metal specification.
It should be recognized that the smallest section that can be cast gray, without massive carbides, depends not only on
metal composition, but also on foundry practices. For example, by adjusting silicon content or by using graphitizing
additions called inoculants in the ladle, the foundryman can decrease the minimum section size for freedom from carbides
for a given basic composition of gray iron.
The mass effect associated with increasing section thickness or decreasing cooling rate is much more pronounced in gray
iron than in cast steel. The mass effect in cast steel results in increased grain size in heavy sections. This also applies to
gray iron, but the most important effects are on graphite size and distribution, and on amount of combined carbon.
For any given gray iron composition, the rate of cooling from the freezing temperature to below about 650 °C (1200 °F)
determines the ratio of combined to graphitic carbon, which controls the hardness and strength of the iron. For this reason
the effect of section size in gray iron is considerably greater than in the more homogeneous ferrous metals in which
cooling rate does not affect the form and distribution of carbon throughout the metal structure.
Typical Effects of Section Size. When a wedge-shape bar with about a 10° taper is cast in a sand mold and sectioned
near the center of the length, and Rockwell hardness determinations are made on the cut surface from the point of the
wedge progressively into the thicker sections, the curves so determined show to what extent continually increasing section
size affects hardness.
The tip of the wedge
is white iron (a mixture of carbide and pearlite) with a hardness greater than 50 HRC. As the iron becomes mottled (a
mixture of white iron and gray iron), the hardness decreases sharply. A minimum is reached because of the occurrence of
fine type D flake graphite, which usually has associated ferrite in large amounts. With a slightly lower cooling rate, the
structure becomes fine type A flake graphite in a pearlite matrix with the hardness rising to another maximum on the
curve. This structure is usually the most desirable for wear resistance and strength. With increasing section thickness
beyond this point, the graphite flakes become coarser, and the pearlite lamellae become more widely spaced, resulting in
slightly lower hardness. With further increase in wedge thickness and decrease in cooling rate, pearlite decomposes
progressively to a mixture of ferrite and graphite, resulting in softer and weaker iron.
Volume/Area Ratios.
It is extremely difficult to predict with accuracy the cooling rate for castings other than fairly
simple shapes. However, because minimum limitations are involved here, the problem can be resolved through
comparisons of the casting design with ratios of volume to surface area or with minimum plate sections.
The volume/area (V/A) ratios for round, square, and plate sections provide a fairly accurate indication of the minimum
casting sections possible in simple geometrical shapes. The V/A ratios can be reported in either English metric units and can be converted simply by treating them as length measurements or
Comparison of the ratios of volume to surface area for different shapes gives good agreement with the actual cooling rates
of castings made in the same mold material. For long round bars and infinite flat plates, V/A is diameter/4 for bars and
thickness/2 for plates; that is, a large plate casting would have the same cooling rate as a round bar with a diameter twice
the plate thickness. Most castings, however, freeze somewhat faster than an infinite flat plate, and rather than establishing
a 2-to-1 ratio of bar to plate, a smaller ratio will often give a better correlation with the cooling rate. Metric units and can be converted simply by treating them as length measurements.
Similar comparisons have been made for production castings. In one study, the properties of a flat section from a 0.6 m
(24 in.) cross pipe fitting having a nominal thickness of 29.5 mm (1.16 in.) were compared with the properties of a 50 mm
(2 in.) diam cylindrical test bar cast from the same heat. The tensile strengths of the test bars were within about 16 MPa
(2.3 ksi) of the tensile strengths of the cross pipe fittings for eight heats ranging in strength from about 205 to 310 MPa
(30 to 45 ksi), an average variation of less than 8%. These results from production castings correlate well with the
calculated equivalence given in Table 3. Other examples of this type of correlation are given in Ref 3.
Relationships developed for various specific castings are valid when an iron of controlled composition, and therefore of
similar section sensitivity, is used consistently. For instance, with a copper-molybdenum iron of well-controlled
composition, a tensile strength of 450 MPa (65 ksi) in the ASTM B test specimen has been found to ensure 345 MPa (50
ksi) tensile strength in a cast crankshaft 2.13 m (7 ft) long with sections thicker than 30.5 mm (1.2 in.). Such translation of
properties of a small test bar to properties expected in a larger section cannot be done indiscriminately, because different
irons may vary widely in section sensitivity.
References cited in this section
2. R. Schneidewind and R.G. McElwee, Composition and Properties of Gray Iron, Parts I and II, Trans. AFS,
Vol 58, 1950, p 312-330
3. H.C. Winte, Gray Iron Castings Section Sensitivity, Trans. AFS, Vol 54, 1946, p 436-443
Prevailing Sections
Although the ASTM size B test bar (30.5 mm, or 1.2 in., diam) is the bar most commonly used for all gray irons from
class 20 to class 60, ASTM specification A 48 provides a series of bar sizes from which one that approximates the cooling
rate in the critical section of the casting can be selected. In practice, it is customary to be somewhat more definite
regarding the prevailing values of minimum casting section considered feasible for the various ASTM classes of cast iron.
Test Bar Properties
Mechanical property values obtained from test bars are sometimes the only available guides to the mechanical properties
of the metal in production castings. When test bars and castings are poured from metal of the same chemical history,
correlations can be drawn between the thermal history of the casting and that of the test bar. The strength of the test bar
gives a relative strength of the casting, corrected for the cooling rate of the various section thickness. Through careful
analysis of the critical sections of a casting, accurate predictions of mechanical behavior can be achieved.
Usual Tests. Tension and transverse tests on bars that are cast specifically for such tests are the most common methods
used for evaluating the strength of gray iron.
Yield strength, elongation, and reduction of area are seldom determined for gray iron in standard tension tests. The
transverse test measures strength in bending and has the additional advantage that a deflection value may be obtained
Data can usually be obtained faster from the transverse test
than from the tension test because machining of the specimen is unnecessary. The surface condition of the bar will affect
the transverse test but not the tension test made on a machined specimen. Conversely, the presence of coarse graphite in
the center of the bar, which can occur in an iron that is very section sensitive, will affect the tension test but not the
transverse test.
ensure minimum statistical variation between bars. By its nature gray iron behaves as a brittle material in tension, with no
measurable elongation after fracture. This characteristic can be exaggerated by imposing a nonaxial load during tensile
testing, resulting in statistical variations, which may not be a true measure of the quality of the iron. To overcome this
tendency, many shops use self-aligning nonthreaded grips in the performance of tensile tests on gray iron tests bars.
Typical Specifications. ASTM A 48 is typical of specifications based on test bars. In practice, one of three different
standard sizes of separately cast test bars is used to evaluate the properties in the controlling section of the castings. After
manufacturer and purchaser agree on a controlling section of the casting, the size of test bar that corresponds,
approximately, to the cooling rate.
Most gray iron castings for general engineering use are specified as either class 25, 30, or 35. Specification A 48 is based
entirely on mechanical properties, and the composition that provides the required properties can be selected by the
individual producer. A manufacturer whose major production is medium-section castings of class 35 iron will find, for
heavy-section castings for which the 50 mm (2 in.) test bar is required for qualifying, that the same composition will not
meet the requirements for class 35. It will qualify only for some lower class, such as 25 or 30. As the thickness of the
controlling section increases, the composition must be adjusted to maintain the same tensile strength.
Compressive Strength. When gray iron is used for structural applications such as machinery foundations or supports,
the engineer, who is usually designing, to support weight only, bases his calculations on the compressive strength of the
The compressive strength of gray iron is typically three to four times that of the tensile strength.
If loads other than dead weights are involved (unless these loads are constant), the problem is one of dynamic stresses,
which requires the consideration of fatigue and damping characteristics.
Tensile strength is considered in selecting gray iron for parts intended for static loads in direct tension or bending.
Such parts include pressure vessels, autoclaves housings and other enclosures, valves, fittings, and levers. Depending on
the uncertainty of loading, safety factors of 2 to 12 have been used in figuring allowable design stresses.
Transverse Strength and Deflection. When an ASTM arbitration bar is loaded as a simple beam and the load and
deflection required to break it are determined, the resulting value is converted into a nominal index of strength by using
the standard beam formula. The value that is determined is arbitrarily called the modulus of rupture. The values for
modulus of rupture are useful for production control, but cannot be used in the design of castings without further analysis
and interpretation. Rarely does a casting have a shape such that those areas subject to bending stress have a direct
relationship to the round arbitration bar. A more rational approach is to use the tensile strength (or fatigue limit) and, after
determining the section modulus of the actual shape, apply the proper bending formula. However, because the difficulty
of obtaining meaningful values of tensile strength in tests of small specimens, the load computed in this manner is usually
somewhat lower than the actual load required to rupture the part, unless unfavorable residual stresses are present in the
finished part.
Elongation of gray iron at fracture is very small (of the order of 0.6%) and hence is seldom reported. The designer
cannot use the numerical value of permanent elongation in any quantitative manner.
Torsional Shear Strength. Most gray irons have high torsional shear strength. Many grades
have torsional strength greater than that of some grades of steel. This characteristic, along with low notch sensitivity,
makes gray iron a suitable material for shafting of various types, particularly in the grades of higher tensile strength. Most
shafts are subjected to dynamic torsional stresses, and the designer should carefully consider the exact nature of the loads
to be encountered. For the higher-strength irons, stress concentration factors associated with changes of shape in the part
are important for torque loads as well as for bending and tension loads.
Modulus of Elasticity. Typical stress-strain curves for gray iron are shown in Gray iron does not obey Hooke's
law, and the modulus in tension is usually determined arbitrarily as the slope of the line connecting the origin of the
stress-strain curve with the point corresponding to one-fourth the tensile strength (secant modulus). Some engineers use
the slope of the stress-strain curve near the origin (tangent modulus). The secant modulus is a conservative value suitable
for most engineering work; design loads are seldom as high as one-fourth the tensile strength, and the deviation of the
stress-strain curve from linearity is usually less than 0.01% at these loads. However, in the design of certain types of
machinery, such as precision equipment, where design stresses are very low, the use of the tangent modulus may
represent the actual situation more accurately.
Hardness of gray iron, as measured by Brinell or Rockwell testers, is an intermediate value between the hardness of the
soft graphite in the iron and that of the harder metallic matrix. Variations in graphite size and distribution will cause wide
variations in hardness (particularly Rockwell hardness) even though the hardness of the metallic matrix is constant.
If any hardness correlation is to be attempted, the type and amount of graphite in the irons being compared must be
constant. Rockwell hardness tests are considered appropriate only for hardened castings (such as camshafts), and even on
hardened castings, Brinell tests are preferred. Brinell tests must be used when attempting any strength correlations for
unhardened castings.
References cited in this section
5. D.E. Krause, Gray Iron--A Unique Engineering Material, in Gray, Ductile, and Malleable Iron Castings--
Current Capabilities, ASTM STP 455, American Society for Testing and Materials, 1969, p 3-28
6. C.F. Walton and T.J. Opar, Iron Castings Handbook, Iron Casting Society, 1981, p 235
7. R.B. Gundlach, The Effects of Alloying Elements on the Elevated-Temperature Properties of Gray Irons,
Trans. AFS, 1983, p 389
Fatigue Limit in Reversed Bending
Axial loading or torsional loading cycles are frequently encountered in designing parts of cast iron, and in many instances
these are not completely reversed loads. Types of regularly repeated stress variation can usually be expressed as a
function of a mean stress and a stress range. Wherever possible, the designer should use actual data from the limited
information available. Without precisely applicable test data, an estimate of the reversed bending fatigue limit of
machined parts may be made by using about 35% of the minimum specified tensile strength of the particular grade of
gray iron being considered. This value is probably more conservative than an average of the few data available on the
fatigue limit for gray iron. Tensile strength is plotted on the horizontal axis to represent fracture strength under static load (which corresponds to a 0 stress
range). Reversed bending fatigue limit is plotted on the ordinate for 0 mean stress, and the two points are joined by a
straight line. The resulting diagram yields a fatigue limit (maximum value of alternating stress) for any value of mean
Few data available are applicable to design problems involving dynamic loading where the stress cycle is predominantly
compressive rather than tensile. Some work done on aluminum and steel indicates that for compressive (negative) mean
stress. Gray iron is probably at least as strong as this for loading cycles resulting in negative
mean stress, because it is much stronger in static compression than in static tension. It is therefore a natural assumption
that the parallel behavior is conservative.
If, prior to design, the real stress cycle can be predicted with confidence and enough data are available for a reliable S-N
diagram for the gray iron proposed, the casting might be dimensioned to obtain a minimum safety factor of two based on
fatigue strength. (Some uses may require more conservative or more liberal loading). The safety factor is determined by the distance from the origin to the fatigue limit
line along a ray through the cyclic-stress point, divided by the distance from the origin to that point. On this diagram, point P' represents a stress cycle having a negative mean stress. In other words, the maximum
compressive stress is greater than the tensile stress reached during the loading cycle. In this instance, the safety factor is
the distance OF'/OP'. However, this analysis assumes that overloads will increase the mean stress and alternating stress in
the same proportion. This may not always be true, particularly in systems with mechanical vibration in which the mean
stress may remain constant. For this condition, the vertical line through P would be used; that is, DK/DP would be the
factor of safety.
Most engineers use diagrams such as Fig. 14 mainly to determine whether a given condition of mean stress and cyclic
stress results in a design safe for infinite life. The designer can also determine whether variations in the mean stress and
the alternating stress that he anticipates will place his design in the unsafe zone. Usually the data required to analyze a
particular set of conditions are obtained experimentally. It is emphasized that the number of cycles of alternating stress
implied is the number normally used to determine fatigue limits, that is, approximately 10 million. Fewer
cycles, as encountered in infrequent overloads, will be safer than indicated by a particular point plotted on a diagram for
infinite life. Too few data are available to draw a diagram for less than infinite life.
Fatigue Notch Sensitivity. In general, very little allowance need be made for a reduction in fatigue strength caused
by notches or abrupt changes of section in gray iron members. The low-strength irons exhibit only a slight reduction in
strength in the presence of fillets and holes. That is, the notch sensitivity index approaches 0; in other words, the effective
stress concentration factor for these notches approaches 1. This characteristic can be explained by considering the
graphite flakes in gray iron to be internal notches. Thus, gray iron can be thought of as a material that is already full of
notches and that therefore has little or no sensitivity to the presence of additional notches resulting from design features.
The strength-reducing effect of the internal notches is included in the fatigue limit values determined by conventional
laboratory tests with smooth bars. High-strength irons usually exhibit greater notch sensitivity, but probably not the full
theoretical value represented by the stress concentration factor. Normal stress concentration factors (see Ref 9) are
probably suitable for high-strength gray irons.
References cited in this section
8. W.L. Collins and J.O. Smith, Fatigue and Static Load Tests of a High-Strength Cast Iron at Elevated
Temperatures, in Proceedings of ASTM, Vol 41, 1941, p 797-807
9. R.E. Peterson, Stress Concentration Factors, John Wiley, New York, 1974
Pressure Tightness

Gray iron castings are used widely in pressure applications such as cylinder blocks, manifolds, pipe and pipe fittings,
compressors, and pumps. An important design factor for pressure tightness is uniformity of section. Parts of relatively
uniform wall section cast in gray iron are pressure tight against gases as well as liquids. Most trouble with leaking
castings is encountered when there are unavoidable, abrupt changes in section. Shrinkage, internal porosity, stress
cracking, and other defects are most likely to occur at junctions between heavy and light sections.
Watertight castings are considerably less challenging to the foundryman than gastight castings. A slight sponginess or
internal porosity at heavy sections will not usually leak water, and even if there is slight seepage, internal rusting will
soon plug the passages permanently. For gastightness, however, castings must be quite sound.
Lack of pressure tightness in gray iron castings can usually be traced to internal porosity, which also is called internal
shrinkage. In gray iron this seems to be a phenomenon distinctly different from the normal solidification shrinkage that
often appears on the casting surface as a sink or draw, which can be cured by risering. Internal porosity or shrinkage (very
difficult to prevent, even by the use of very heavy risers) is usually associated with poor feeding and lack of directional
solidification. It can also be associated with heavy inoculation with calcium to promote a high graphite cell count. On the
other hand, the use of strontium-bearing inoculants does not reduce cell size and can help control internal shrinkage in
marginally gated castings.
Visible internal porosity may appear at centers of mass when the phosphorus content exceeds 0.25%. In critically gated
castings, visible internal porosity may appear when the phosphorus content is as low as 0.09%. Chromium and
molybdenum accentuate this effect of phosphorus, while nickel has a slight mitigating influence.
The effect of phosphorus may be caused in part by the fact that lowering phosphorus content also lowers the effective
carbon equivalent. Lowering carbon equivalent by reducing carbon or silicon, or both, instead of phosphorus, might
similarly reduce the leakage of pressure castings, but other foundry problems (such as increasing amounts of normal
shrinkage) would be encountered. ASTM A 278 for pressure castings requires a carbon equivalent of 3.8% (max), a
phosphorus content of 0.25% (max), and a sulfur content of 0.12% (max) for castings to be used above 230 °C (450 °F).
In addition to composition control, good overall foundry practice is required for consistently producing pressure-tight
castings. Sand properties and gating must be controlled to avoid sand inclusions. Pouring temperature must be adequate
for good fluidity, and heavy sections should be fed wherever possible.
Mold properties have been found to interact with composition to influence internal soundness. Generally, molds rammed
tightly produce the soundest castings because of freedom from mold wall movement during solidification.
Impact Resistance
Where high impact resistance is needed, gray iron is not recommended. Gray iron has considerably lower impact strength
than cast carbon steel, ductile iron, or malleable iron. However, many gray iron castings need some impact strength to
resist breakage in shipment or use.
There is incomplete agreement on a standard method of impact testing for cast iron. Two methods that have been used
successfully are given in ASTM A 327. Most impact testing of cast iron has been used as a research tool.
Most producers of cast iron pressure pipe use a routine pipe impact test as a control. Impact resistance in pipe helps avoid
breakage in shipping and handling.
The machinability of most gray cast iron is superior to that of most other cast irons of equivalent hardness, as well as to
that of virtually all steel. The flake graphite introduces discontinuities in the metal matrix, which act as chip breakers. The
graphite itself serves as a lubricant for the cutting tool. However, economical cutting depends on more than inherent
machinability alone. Often, trouble in machining gray iron can be traced to one or more of several factors: the presence of
chill of corners and in light sections, the presence of adhering sand on the surface of the casting, swells, usually the result
of soft molds, shifted castings, shrinks, and phases included in the matrix as a result of melting practice.
Chill at corners and in light sections is more likely to be encountered with small castings, with higher-strength irons, and
with designs that have light sections in the cope (or top) of the mold. Most foundries control iron with a chill test that
gives an indication of the tendency of the iron to form white or mottled iron in light sections. The foundryman may treat
his iron with a small amount (0.5 to 2.5 kg/Mg, or 1 to 5 lb/ton) of a graphitizing alloy such as calcium-bearing
ferrosilicon or other proprietary inoculant, which effectively decreases chill. Inoculation to achieve control of the
tendency to chill usually does not result in significant changes in the composition or the physical properties of the iron,
although it does produce changes in the mechanical properties.
Light sections (5 mm, or 3
in.) usually cannot be cast in gray iron of higher than class 25. Class 30 iron can be cast in 6
mm ( 1
in.) sections. These values are different for different designs, depending on how the casting is made and gated.
The important thing to understand is that the cooling rate in the mold at the time of freezing determines whether the iron
will be gray, white, or mottled. If the thin section is in the drag or near the gate, the flow of hot metal heats the mold,
thereby decreasing the rate of cooling and enhancing the formation of the gray iron.
If chill is encountered, it is generally best to correct the trouble at its origin. It is usually uneconomical to anneal castings
to remove chill because, in addition to heat treating for 2 h at 900 °C (1650 °F) for unalloyed irons, recleaning may be
required for removing scale. Distortion beyond tolerance often occurs, and there are sacrifices in hardness and strength.
Adhering sand usually can be removed by effective cleaning, but sand present as the result of penetration of the iron
into the mold wall is extremely difficult to blast clean. This is a foundry defect that is best corrected at the source.
Slowing the speed of machining and increasing the rate of feed is the best approach to salvaging castings of this type.
Carbide tools are better than high-speed tools for resisting the extreme abrasion.
Swells are most troublesome in operations such as broaching and in other setups tooled for high production. The
additional metal often places an excessive load on the tool, which may chip or dull but not actually fail until some time
after the troublesome parts have been machined. In highly automated machining centers, this casting defect can result in
significant statistical variation in dimensions and high scrap rates.
Shifted castings are similar to swells in their action on cutting tools. Shifts or swells also may cause excessive tool
loading if the locating points are affected. It is important to consider the positions of such locating points when designing
the castings and also to avoid indiscriminate grinding of locating points in the foundry cleaning room.
Shrinks are not present but can be troublesome when encountered in operations such as drilling. For example, the drill
may tend to drift from its intended path to follow the shrink, which offers less resistance to the drill. Sometimes a drill
may break because it encounters a region of higher hardness, which often is associated with an area of shrink. Cast iron is
the easiest metal to cast without internal shrink. Eutectic freezing is accompanied by expansion due to the precipitation of
low-density graphite, which aids in obtaining internal soundness.
Machinability rating is complex and is discussed in the article "Machinability of Steels" in this Volume. Criteria such
as power per unit volume in unit time are of greatest importance in selecting a machine tool and the size of its motor.
Machinability ratings based on tool life under standard test conditions are helpful but are not readily interpreted into the
economics of machining.
One way to indicate the effect on machinability of changing from one grade of iron to another ,
which is based on an experiment in which metal was removed from four types of gray iron using single-point carbide
tools. For each type of iron, cutting speed was varied until the removal of 3300 cm3 (200 in.3) of iron resulted in a 0.75
mm (0.030 in.) wear land on the tool. The values of cutting speed thus obtained serve as qualitative evaluations of
machinability, but not as qualitative indices. Optimum cutting speeds, tool materials, cutting fluids, feeds, and finish
requirements must be studied for any given machine tool setup. Tool life is an important factor, because the machine must
be stopped to change tools and the tools must be resharpened. Progress has been made in decreasing both machine
downtime for tool changes and resharpening cost by the use of solid carbide inserts or bits and by other means.
Annealing. The greatly improved machinability obtainable in gray iron by annealing has been advantageous to
automotive and other industries for many years. Annealing is usually of the subcritical type, such as 1 h at 730 to 760 °C
(1350 to 1400 °F). Some employ a cycle of heating to 785 to 815 °C (1450 to 1500 °F) and cooling at 22 °C/h (40 °F/h) to
about 595 °C (1100 °F). These treatments graphitize the carbide in the pearlite and result in a ferritic matrix. Finding it
uneconomical to graphitize primary carbide, most users try to avoid obtaining it. Annealing for improved machinability is
most economical when the casting is small and the amount of machining is large.
The annealing treatments described result in sacrifices in hardness and strength. A typical class 35 iron will be
downgraded to about class 20 in strength by this treatment. In applications in which wear resistance is important, such as
cylinder blocks, gray iron is not annealed because of the unsatisfactory performance obtained with a ferritic matrix.
Reference cited in this section
10. U.S. Air Force Machinability Report, Vol 1, 1950, p 135
Gray iron is used widely for machine components that must resist wear. Different types of iron, however, exhibit great
differences in wear characteristics. These differences do not correlate with the commonly measured properties of the iron.
In the discussion that follows, the general conclusions are related to metal-to-metal wear during sliding contact under
conditions of normal lubrication. Although most of the supporting illustrations are for engine cylinders, the results have
wider applicability.
The published data on gray iron wear are somewhat inconsistent; accelerated-wear tests often do not correlate with field
service experience, nor does field experience in one application necessarily agree with that in another application. In
many applications, properties of the surface are at least as important as properties of the metal; for instance, wear
resistance may frequently be enhanced by lapping the wear areas. Relative hardness between mating parts also may be
important to optimum wear resistance. Frequently, a hardness difference no greater than 10 points on the Brinell scale is
considered optimum.
For components in sliding contact, such as engine cylinders, valve guides, and latheways, the recognized types of wear
are cutting wear, abrasive wear, adhesive wear, and corrosive wear. In well-designed machinery, wear proceeds slowly
and consists of combinations of very mild forms of these four types of wear. The predominant characteristic of this socalled
normal wear is the development of a glazed surface during break-in. The primary objective in any wear application
is to establish conditions that produce and maintain this glaze. A secondary objective is to minimize normal wear.
Cutting wear is caused by the mechanical removal of surface metal as a result of surface roughness and is similar to the
action of a file. It usually occurs during the breaking in of new parts.
Abrasive wear is caused by the cutting action or loose, abrasive particles that get between the contacting faces and act
like a lapping compound. Under some circumstances, abrasive particles embedded in one or both of the contacting faces
can produce a similar action.
Adhesive wear is caused by metal-to-metal contact, resultant welding, and the breaking of those welds. When this
happens on a large scale (a process known as galling), the metal is smeared, and severe surface damage results. Even in
properly operating equipment, some wear occurs by adhesion on a microscopic scale. An intermediate form of adhesive
wear called scuffing is less destructive than galling, but still results in an abnormally high rate of metal removal. In all
probability, accelerated-wear tests between clean surfaces (with or without the presence of a lubricant) are predominantly
tests of galling or scuffing.
Corrosive wear is a special type of wear that combines abrasion or adhesion with the chemical action of the
environment. In engine cylinders, it is caused by condensed acidic products of combustion during low-temperature
operation. Usually this kind of wear cannot be corrected by modifications in ordinary types of gray iron.
Resistance to Scuffing
Assuming reasonable design and operating conditions that minimize cutting and abrasive wear, scuffing or mild galling is
the abnormal wear condition most likely to defeat the objective of obtaining and maintaining a glazed surface due to
normal wear. Several alloy combinations with widely varied microstructures were tested by Shuck (Ref 11). A brake shoe
type of specimen was held against a rotating gray iron drum for 1 h, after which both specimen and drum were checked
for weight loss. The conditions of this test indicate that wear was caused primarily by adhesion (scuffing), although
cutting wear may have had a contributing effect. The tests covered almost all conceivable microstructures and a wide
variety of compositions having hardnesses below 300 HB. The conclusions, which basically agreed with other less
comprehensive investigations, were:
· Microstructure determines wearing characteristics
· As the graphite structure becomes coarser and tends toward type A, scuffing decreases
· Interdendritic type D graphite and its associated ferrite give very poor results
· Secondary ferrite associated with random type A graphite is less damaging than that associated with
type D
· Pearlitic, acicular, or tempered martensitic structures in the same hardness range are equal in wear
· For a given type of graphite, as the matrix becomes more pearlitic and harder, wear resistance increases
These results were substantiated in scuff tests of cylinder sleeves in a diesel engine. The engine, equipped with the test
sleeves, was operated at constant speed, and the horsepower was increased by increments above normal until scuffing
occurred. The scuffing horsepower was expressed in a ratio with normal horsepower.
Effect of Graphite Structure. To eliminate differences in matrix, four types of gray iron were hardened and tempered
at 205 °C (400 °F) to produce a uniform martensitic matrix. The changes in casting method in tests 3 and 4 caused no
significant change in scuffing. The iron in test 5 is one used formerly in high-performance brake drums to give maximum
resistance to scoring.
The combination of type D graphite and ferrite in test 1 is especially poor even though the rest of the
structure is pearlite. Pearlitic type A iron with small amounts of free ferrite gave ratios greater than 1.33 in the same tests,
showing that the results obtained in test 1 are peculiar to the combination of type D graphite and ferrite. When type D
graphite occurs in as-cast irons, it is usually associated with ferrite. Consequently, foundry practices that avoid the
formation of type D graphite should be adopted when producing castings for wear applications.
Several investigators have noted similar, apparently inconsistent, results, which can be summarized as follows: increasing
hardness from 160 to 260 HB increases resistance to scuffing; however, further increasing hardness above 260 HB
decreases resistance to scuffing, even in heat-treated structures.
These tests, along with those of Ref 11, apparently apply only to scuffing and not to normal wear. As indicated in the
section "Resistance to Normal Wear," wear resistance is different for different modes of wear.
Other investigators have observed that a fine network of steadite (iron phosphide-iron carbide eutectic) increases scuff
resistance, while massive carbide reduces it.
Effect of Chemical Composition. The diesel cylinder scuff tests for evaluating the effect of graphite used five
different compositions from five foundries for tests 3 and 4 and three compositions from two foundries for test 2. Despite
the variations in source and in composition, all fine type A irons performed nearly alike in the scuff tests.
Similarly, the three type D materials performed essentially the same. No other composition was checked with coarse type
A graphite. To cover the subject of scuffing more completely, other factors were evaluated during the cylinder sleeve
tests. They help to clarify some of the additional controversial aspects of wear.
Surface finish effects were determined by running similar tests on cylinder sleeves with honed finishes of 0.75 and 2.3
μm (30 and 90 μin.) rms. In each, the sleeves were made from type D graphite iron, hardened, and tempered at 205 °C
(400 °F).
The coarser finish gave about 20% greater scuff resistance; however, this effect was lost if the finish became finer under
normal operating conditions. The loss of scuff resistance by the smoothing effect of normal wear may account for some of
the failures that occur in certain applications after break-in has apparently been successful.
Other important factors affecting scuff resistance are the material and finish of the mating part and the type of lubricant
Reference cited in this section
11. A.B. Shuck, A Laboratory Evaluation of Some Automotive Cast Irons, Trans. AFS, Vol 56, 1948, p 166-

Torna a “Iron & Steel”