Manganese, Sulfur and Manganese-Sulfur Ratio Effects in Gray Cast Iron

Paper 03-072 (05).
pdf, Page 1 of 30
AFS Transactions 2003 © American Foundry Society, Des Plaines IL USA
AFS Transactions 03-072 Page 1 of 30
Manganese, Sulfur and Manganese-Sulfur Ratio Effects in Gray Cast Iron
G. M. Goodrich
Professional Metallurgical Services, Buchanan, Michigan
T.G. Oakwood
Professional Metallurgical Services, Buchanan, Michigan
R.B. Gundlach
Climax Research Services, Wixom, Michigan
Copyright 2003 American Foundry Society
ABSTRACT
A study was undertaken to examine the role of sulfur and manganese on the microstructures and mechanical properties of
gray cast iron. A series of heats of gray iron containing varying amounts of sulfur and manganese were cast as step blocks at
several foundries under commercial production conditions. The microstructures and mechanical properties were determined
as functions of composition and casting variables. It was found that graphite flake length decreased and eutectic cell count
increased as solidification cooling rate increased. A correlation was established between measured tensile strength and
predictions of tensile strength using a regression analysis based on composition and section size. Over the range of sulfur
levels evaluated, it was found that tensile strength decreased with increasing sulfur level. Over the range of manganese levels
studied, tensile strength first decreased and then increased with increasing manganese level. Tensile strength also decreased
and then increased with increasing excess manganese. Tensile strength was correlated with the precipitation of manganese
sulfide prior to the onset of solidification. Maximum tensile strength was achieved when sulfur or manganese content was
less than that required for MnS precipitation at the liquidus. When MnS precipitation occurred prior to the onset of
solidification, tensile strength was reduced.
INTRODUCTION
Many investigators have studied the individual roles of sulfur and manganese in gray iron. They have attempted to explain
the effects of these two elements in terms of manganese level, sulfur level or manganese to sulfur ratio. All of these
investigators agree that sulfur and manganese have a significant influence on both the microstructure and mechanical
properties of gray iron. As will be discussed below, however, some of the results have led to confusion as to the exact roles
of sulfur and manganese in the metallurgy of gray cast iron.
This study is based on the hypothesis that the effects of sulfur and manganese on the microstructure and properties of gray
iron are related to the dissolved or “effective” amounts of both elements, which remain dissolved in the liquid at the onset of
solidification. The effective amounts of sulfur and manganese can be expressed in terms of solubility product data. In this
paper, the known effects of sulfur and manganese, which are described in the literature, will be briefly reviewed. The results
of a current experimental program aimed at furthering the understanding of the roles of sulfur and manganese on the
properties of gray iron will then be described. Data from both the literature and the current investigation will be analyzed in
terms of the role of dissolved sulfur and manganese as expressed by equilibrium solubility product data.
HISTORICAL REVIEW
A review of the literature has identified several microstructure variables in gray iron, as well as malleable iron, that are
affected by changes in sulfur and manganese content. Some of these variables have been related to the amount of manganese
or sulfur, whereas others have been related to the ratio of manganese to sulfur. These findings are summarized in Table 1.
The nomenclature used in the table is the same as that used in the literature. A manganese to sulfur ratio (Mn/S) of 1.7 is
considered “balanced”. Excess manganese refers to Mn/S being greater than 1.7, and excess sulfur refers to Mn/S being less
than 1.7.
As can be seen from the table, increases in sulfur content at constant manganese level result in an increase in chill and mottle
depth (1). As manganese content is increased up to a manganese to sulfur ratio slightly above 1.7, the effect of sulfur on chill
and mottle depth is minimized. When the Mn/S ratio exceeds 1.7 by larger values, the excess manganese results in an
increase in chill and mottle depth (1,2,3). In addition, the form of graphite developed in gray iron is influenced by Mn/S ratio.
At excess manganese flake graphite is found, and as the ratio is reduced, graphite flakes become finer. In some instances
Paper 03-072 (05).pdf, Page 2 of 30
reference has been made to formation of a “mesh” graphite. When an excess of sulfur exists, abnormal graphite forms have
been reported. Terminology such as “spiky” graphite has been used in the literature (3).
Table 1. Effects of Sulfur and Manganese on Microstructure
Condition Result
Chill and Mottle Depth

Excess S Increased
Balanced to Slight Excess Mn Minimized
Excess Mn Increased
Graphite Form
Excess Mn Normal Flake Graphite
Slight Excess Mn Fine Graphite
Excess S Abnormal Graphite
Matrix Structure
Excess Mn Ferrite Plus Pearlite
Balance to Slight Excess Mn Pearlite
Excess S Hypereutectoid Carbides
Eutectic Carbides, Pearlite
Eutectic Cell Count

Increased S Increases
Increased Mn Decreases
Mn/S ratio also affects the matrix structure of gray iron. At excess manganese levels, the formation of ferrite is favored,
whereas as manganese is reduced, and the Mn/S ratio is lowered, pearlite formation is favored. (This effect is counter to what
is observed in ductile iron where higher manganese levels favor pearlite formation.) At excess sulfur levels, hypereutectoid
carbides have been observed, which are often associated with iron sulfide particles. With increasing amounts of sulfur
eutectic carbides have also been found (3). Finally, sulfur has been found to increase eutectic cell count, whereas manganese
decreases it. (4).
A number of investigators have examined the effects of varying sulfur and manganese contents on mechanical properties
(3,5,6,7)
. The effect of sulfur and manganese content on tensile strength is shown in Figures 1 and 2 for different manganese to
sulfur ratios (5). It should be noted that the same mechanical property data is used in both figures. As can be seen, tensile
strength decreases as both sulfur and manganese content increase.
As is evident from Table 1 along with Figures 1 and 2, the effects of sulfur and manganese on the structure and properties of
gray iron are complex, and, in some cases, confusing. Reducing manganese to sulfur ratio can increase strength, but
unwanted microstructures could possibly form. Furthermore, tensile strength decreases as both manganese and sulfur
increase. In the particular case of manganese, this is somewhat surprising since, in ferrous metals manganese is usually
considered a solid solution strengthener, and also increases hardenability. Figure 1 shows a clear increase in strength, at all
sulfur levels, as the manganese to sulfur ratio is reduced. In Figure 2, however, the effects of manganese to sulfur ratio are
not as clear. At 0.2% Mn and 0.5 % Mn comparable tensile strengths can be obtained at different manganese to sulfur ratios.
The effects of sulfur and manganese on the microstructure and properties of gray iron have often been attributed to the
presence (or absence) of manganese sulfide formation. Manganese and sulfur, in solution in liquid iron at a particular
temperature, are in equilibrium with manganese sulfide as follows:
Mn (in soln) + S (in soln) ↔ MnS (1)
A similar expression describes the equilibrium between iron and sulfur:
Fe (liq) + S (in soln) ↔ FeS (2)

Combining Equations 1 and 2 algebraically results in the following expression, which is often cited in the literature (3,5).
Mn (in soln) + FeS ↔ Fe (liq) + MnS (3)
Since this reaction is reversible, an excess of manganese is needed to drive the reaction to the right to assure formation of
manganese sulfide, and avoid the formation of unwanted iron sulfide. Equations 1-3 assume that manganese sulfide and iron
sulfide form as distinct particles. There is a continuum however where both manganese and iron can be present together in
an individual sulfide in the form (Fe,Mn S). The stoichiometric amount of manganese needed to just combine with the sulfur
is 1.7(%S), i.e. the ratio of manganese to sulfur in a manganese sulfide particle is 1.7. A guideline has therefore evolved
aimed at assuring that all of the sulfur present in gray iron forms manganese sulfide. This guideline is given by the following
expression (2,8):
Mn% = 1.7(S%) + 0.3 % (4)
330
Mn=1.7S
Mn=1.7S+0.1
310
Mn=1.7S+0.2
Ultimate Tensile Strength, MPa
Mn=1.7S+0.3
290
270
250
230
210
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 0.24
% Sulfur
Fig. 1. Ultimate Tensile Strength vs Sulfur Content. After Reference 5.
The extra manganese in Equation 4 is aimed at assuring freedom from iron sulfide and producing optimum microstructures.
While Equation 4 has been used to avoid unwanted matrix structures, graphite forms and chilling tendencies, it stops short of
defining the optimum combinations of manganese and sulfur. As was shown in Figures 1 and 2, strength can be increased by
adjusting manganese and sulfur levels to lower manganese to sulfur ratios, provided the levels are optimized to avoid
unwanted microstructures. Furthermore, Equation 4 simply defines a conservative relationship that assures that all sulfur is
present as manganese sulfide. It does not describe the conditions under which manganese sulfide forms in gray iron. The
work to be described in this paper is based upon the use of an equilibrium solubility product approach to describe the
formation of manganese sulfide, and the effects of manganese and sulfur on mechanical properties of gray iron.
USE OF SOLUBILITY PRODUCT DATA TO DESCRIBE EFFECTS OF SULFUR AND MANGANESE

It has been proposed that a principal factor influencing the properties of gray iron is the “free” sulfur, which is present in the
liquid and is constantly changing during solidification. (9). A way of describing free sulfur (or manganese) is to define it as
the amount present in uncombined form at the onset of solidification. To do so requires the use of solubility product data.
The solubility of manganese sulfide in liquid iron will vary with composition and temperature, and is described by the
following equation (2):
Log10 (% Mn x % S) = −1920/T (5)

(T = Deg. K )
330
Mn/S = 1.7
310 Mn/S = 1.7 + 0.1

Ultimate Tensile Strength, N/sq. mm
Mn/S = 1.7 + 0.2

Mn/S = 1.7 + 0.3
290
270
250
230
210
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Wt. % Manganese
Fig. 2. Ultimate Tensile Strength vs. Manganese Content. After Reference 5.
Figure 3 shows graphs of Equation 5 for two temperatures, 1200°C and 1400°C. The data show the various combinations of
manganese and sulfur that are in equilibrium with each other in liquid iron at each temperature. For example, Point A
designates a point at which 0.6% Mn is in equilibrium with 0.12% S at a temperature of 1400°C. If the temperature is
reduced, MnS reaction products will form, while at higher temperatures manganese and sulfur will remain in solution. If an
iron containing this manganese and sulfur level is cooled from 1400°C to 1200°C, both manganese and sulfur will be
removed from solution to form MnS, and some manganese and sulfur will remain in solution in the liquid iron. The arrow
drawn parallel to the stoichiometric line in Figure 3 depicts this. Once the temperature of 1200°C is reached, new
equilibrium percentages of dissolved manganese and sulfur are established as shown at Point B. In this case the new
percentages are 0.092% sulfur and 0.55% Mn, and are the percentages remaining in the liquid iron at 1200°C. Assuming
solidification commences at 1200°C, these values would define the free sulfur and manganese remaining in solution in liquid
iron.
A consequence of this approach can be seen in Table 2. The liquidus temperatures for gray irons with three different carbon
equivalents are shown (10). Also shown is whether or not MnS precipitation can be expected prior to the onset of
solidification of these irons assuming a composition containing 0.11% sulfur and 0.49% Mn. This combination of
manganese and sulfur was obtained by assuming 0.11% sulfur and calculating the amount of manganese normally required
using Equation 4. The equilibrium temperature for MnS precipitation for this combination of manganese and sulfur was then
calculated from Equation 5 to be 1237°C.
Table 2. MnS Formation During Solidification (0.11% S, 0.49% Mn)
Carbon Equivalent Liquidus Temperature Ppt. Of MnS Before

°C (Reference 10)* Solidification?
3.50 1252 No
3.86 1213 Yes
4.13 1168 Yes
*It is recognized that the data used to calculate liquidus temperature as a function of carbon equivalent in Table 10 as well as in the subsequent
development of Figure 21 has been updated in recent years. The data were selected since numerical values were available to develop an equation
for use in this work. The use of more recent liquidus data could enhance the findings resulting from this work.
The data in Table 2 indicate that no MnS precipitation occurs in the iron with a 3.50 carbon equivalent prior to the onset of
solidification. A small amount of MnS would be expected to form prior to the onset of solidification of the iron with the 3.86
carbon equivalent, and a larger amount in the iron with the 4.13 carbon equivalent. All of the MnS in the iron with the 3.50
carbon equivalent precipitates between the beginning of pro-eutectic solidification, defined by the liquidus, and the end of the
1.2
0.8 Mn = 1.7S
Wt. % Manganese
A
0.6
B
0.4
0.2 1400C
1200C
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Wt. % Sulfur
Fig. 3. The Effect of Temperature on the Equilibrium Relationship Between Manganese and Sulfur.
eutectic solidification. For the irons with carbon equivalents of 3.86 and 4.13, some of the MnS will form prior to the onset
of pro-eutectic solidification, and the balance between the liquidus and the end of the eutectic solidification. While the
manganese addition determined by Equation 4 predicts that all of the sulfur will be effectively combined as MnS, Equation 4
does not describe when MnS will form during solidification. It is apparent that the amount of free sulfur and manganese
present at the onset of solidification will vary with carbon equivalent even at constant manganese and sulfur levels.
Furthermore, while manganese is soluble both in solid and liquid iron, sulfur has limited solubility in the solid, and, as
solidification proceeds, sulfur is rejected into the remaining liquid. As a result, the sulfur distribution and the morphology of
manganese sulfide changes as solidification proceeds up to the end of the eutectic solidification. It would thus be expected
that influence of sulfur and manganese on structure and properties would differ as a result of these variations.
An example of changing sulfide distribution, and the rejection of sulfur to the remaining liquid can be seen in Figure 4 (11).
The right side of the photograph shows clusters of small manganese sulfide particles in eutectic regions. The microstructure
results from sulfur being rejected to the last liquid remaining before solidification is complete.
Fig. 4. Microstructure of Gray Cast Iron Showing Clusters of MnS Particles in Eutectic Regions (Right Side of
Photograph). Original Magnification 112X. Unetched. After Reference 11.
As was noted in Figures 1 and 2, the mechanical properties of gray iron are functions of both manganese and sulfur content.
It needs to be established if there is a correlation between mechanical properties and the formation of MnS prior to
solidification. Tables 3 and 4 give the manganese contents and tensile strengths, which have been reported in the literature,
for two groups of gray iron with carbon equivalents of 3.80 and 4.16 (3.5). Parts of the data shown were used to develop
Figures 1 and 2. The sulfur contents were 0.12 and 0.11 respectively, and the liquidus temperatures were estimated as
1220°C and 1168°C (10). An analysis of the data was devised, which allowed tensile strength to be evaluated as a function of
the difference between the actual sulfur and the equilibrium amount of sulfur needed for MnS precipitation prior to the onset
of solidification. The tensile strengths of each group of irons are plotted in Figures 5 and 6 as functions of the difference
between the actual sulfur in the iron, and various equilibrium values at each respective liquidus. The equilibrium values of
sulfur for each group were calculated from Equation 5 for the various manganese levels. A negative value on the horizontal
axis indicates that the equilibrium sulfur, determined by a particular level of manganese, is greater than the actual sulfur
level. This means that all of the sulfur is in solution at the onset of solidification. As manganese is added, the equilibrium
level of sulfur is decreased and the difference between the actual sulfur and the equilibrium values is reduced. At some point,
sufficient manganese is added to initiate MnS precipitation, and the difference value becomes zero, and eventually becomes a
positive value. As can be seen from Figures 5 and 6, as the difference value approaches zero, tensile strength drops sharply.
Table 3. Tensile Strengths For Various Manganese Levels (References 3,5)

Carbon Equivalent: 3.80
Sulfur: 0.12%
Liquidus Temperature: 1220°C
Manganese Tensile Strength

Wt. % MPa
0.06 307
0.08 298
0.12 312
0.17 301
0.21 295
0.23 292
0.29 289
0.38 264
0.44 236
0.54 227
Table 4.Tensile Strengths For Various Manganese Levels (References 3,5)

Carbon Equivalent: 4.16
Sulfur: 0.11%
Liquidus Temperature: 1168°C
Manganese Tensile Strength

Wt. % N/sq. mm
0.03 269
0.06 267
0.09 269
0.14 255
0.19 253
0.24 257
0.28 248
0.32 240
0.36 227
0.45 219
0.67 213
0.84 223
Figures 7 and 8 are companion graphs to Figures 5 and 6 respectively, and show the tensile strengths of the two series of gray
irons described above as a function of the difference between the actual and equilibrium manganese levels. In each case a
single equilibrium value of manganese was calculated using Equation 5 based on the liquidus temperature and the sulfur
content. This value was then subtracted from the various manganese contents to develop the horizontal axis in Figures 7 and
8. Again, as the difference value approaches zero, the tensile strength is reduced significantly.
The foregoing analysis establishes a correlation between mechanical properties and the formation of MnS. If MnS
precipitation occurs prior to the onset of solidification, tensile strength is reduced. Maximum tensile strength is obtained if
MnS precipitation can be avoided prior to the onset of solidification.
EXPERIMENTAL VERIFICATION
The discussion above indicates that there is a direct relationship between the observed effects of sulfur and manganese in
gray iron and the amounts of these elements remaining in liquid solution just prior to the onset of solidification. This concept
however requires further experimental verification. The mechanical property data in Tables 3 and 4 were developed with
non-commercial iron compositions melted under laboratory conditions. It needs to be established if the analysis used to
evaluate this data can also explain the effects of sulfur and manganese in commercial iron compositions melted under
conditions typical of a production foundry.
A series of heats of gray iron, containing varying amounts of sulfur and manganese, were cast as step blocks at several
foundries. The mechanical properties were determined and the microstructures were analyzed. The purpose of this report is
to summarize these results and to evaluate the role of dissolved sulfur and manganese on microstructure and properties.
EXPERIMENTAL PROCEDURE
TEST CASTING
To evaluate the variables thought to influence the relationship between manganese, sulfur, and mechanical properties, a step
block was chosen to evaluate microstructure and properties. The test variables evaluated were:
Cooling Rate Carbon Equivalent

Pouring Temperature Manganese Level
Inoculation Effect Sulfur Level
To obtain different cooling rates, the step block had four steps. Each step was 6” wide. The step size graduated from 3/8” X
2” to 3/4” X 2-3/8” to 1-1/2” X 3”, to 3” X 4”. The step blocks were poured during routine production. Each step block
weighed about 30 pounds. Interference with production was minimized by using a no bake mold that had no gating system.
The mold was poured with an open face like a chill wedge. Shakeout of the step blocks took place after the disappearance
of any red color. Figure 9 shows the mold.
320
310
300

290
280
270
260
250
CE = 3.80
240 Liquidus: 1220C
% S: 0.12
230
%Mn: 0.06-0.54
220
210
200
-0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2
Departure From Equilibrium Sulfur
(Total Sulfur - Equilibrium Sulfur @ Liquidus), Wt. %
Fig. 5. Ultimate Tensile Strength vs. Total Sulfur Minus Equilibrium Sulfur at the Liquidus for a Carbon Equivalent
of 3.80. Data From References 3,5.
300
290
280
UltimateTensile Stength, MPa
270
260
250
240
230
CE = 4.16
Liquidus: 1168C
220 % S: 0.11
% Mn: 0.03-0.84
210
200
-1.6 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2
(Total Sulfur - Equilibrium Sulfur @ Liquidus), Wt. %
Fig. 6. Ultimate Tensile Strength vs. Total Sulfur Minus Equilibrium Sulfur at the Liquidus for a Carbon Equivalent
320
310
300
290
Ultimate Tensile Strength, N/sq.mm
280
270
260
250
240 CE = 3.80
Liquidus: 1220C
230 % S: 0.12
% Mn: 0.06-0.54
220
210
200
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Departure From Equilibrium Manganese
(Total Manganese - Equilibrium Manganese @ Liquidus),Wt. %
Fig. 7. Ultimate Tensile Strength vs. Total Manganese Minus Equilibrium Manganese for a Carbon Equivalent
300
290
280
270
260
250
240
230 CE = 4.16
Liquidus: 1168C
220 % S: 0.11
% Mn: 0.03-0.84
210
200
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
(Total Manganese - Equilibrium Manganese @ Liquidus), Wt. %
Fig. 8. Ultimate Tensile Strength vs. Total Manganese Minus Equilibrium Manganese at a Carbon Equivalent
Paper 03-072 (05).pdf, Page 10 of 30
Fig. 9. Step Block Mold.
FOUNDRY SELECTION CRITERIA

Seven production foundries were selected to provide a combination of carbon equivalents, manganese levels and sulfur
levels. The table below provides a summary of the nominal levels. Appendix I provides the details.
Table 5. Nominal Composition Comparison
Foundry Excess
I.D. % C.E. %S %Mn % Mn
A 3.87 0.072 0.43 0.308
B 4.22 0.090 0.49 0.337
C 4.09 0.054 0.64 0.548
D 4.28 0.080 0.68 0.544
E 4.14 0.018 0.70 0.669
F 3.70 0.100 0.60 0.430
G 4.32 0.076 0.22 0.091
The intent was to produce the step block castings under conditions experienced during actual production. All of the foundries
are successful commercial foundries. Some unusual nuances that separate two of the foundries from the others (other than
the unusual sulfur level for Foundry E and the unusual manganese level for Foundry G) that should be kept in mind when
reviewing the results include:
Foundry A did not employ any inoculation practice.

Foundry F produced high strength gray iron and had a heavy alloy level compared to the other foundries.
To evaluate the effects of pouring temperature and to introduce inoculation effects, each foundry was instructed to pour three
step blocks. To vary pouring temperature, the first step block was to be poured with the first metal out of the ladle, and the
second and third step blocks were to be poured with the last metal in the ladle after the ladle was used to pour production
castings. To introduce a variation in the effects inoculation, one of the last two molds received approximately 4gms of the
Paper 03-072 (05).pdf, Page 11 of 30
loose “fines” from the inoculant that the foundry normally used. The fines were placed on the bottom of the mold and the
metal poured over them.
The pouring temperatures were recorded for four of the foundries and are indicated in Appendix I. For the most part, the
effects were considered from a generic hot versus cold condition since the pouring temperature requirements varied for each
foundry in the study.
The last two molds were used to study an inoculation versus no inoculation effect. Appendix I documents the composition of
the inoculated mold with the second analysis, A2. The last mold poured without additional inoculation was considered to
have a low inoculation effect due to fade as compared to the last mold poured with loose fines in the bottom. This
comparison, like the temperature comparison was considered from a qualitative standpoint rather than a quantitative
standpoint in order to determine if an effect existed that was worth studying further.
The foundries were also instructed to provide chilled spectrographic samples, pins for combustion carbon and sulfur analysis,
and “B” bars for routine mechanical testing. In some instances the foundries provide an analysis and their mechanical
properties and in other instances, they did not. It was not a requirement since it was desired to have all of the testing done at
one location to remove a source of variability. The data in Appendix I reflects the composition determined on the chill disc
as the Lab 1 data. This testing was conducted at a laboratory that later reported difficulties with their combustion carbon and
sulfur analyzer.
Since the carbon equivalent, sulfur and manganese levels were critical to understanding of the data, it was decided to conduct
chemical analysis at a second laboratory. The compositions on the first step block poured and on the inoculated step block
poured last were determined at a second laboratory. The composition obtained from the second laboratory was used in this
analysis. This second analysis more closely reflected the results that the foundries provided than the results from the first
laboratory.
To address apparent gaps in the data, a second round of step block castings was poured at two of the selected foundries,
Foundry B and Foundry E, with the nominal compositions shown in Table 6.
Table 6. Nominal Compositions of Supplementary Castings
Foundry Excess
I.D. % C.E. %S %Mn % Mn
B 4.25 0.12 0.57 0.366
E 4.18 0.025 0.55 0.508
These second round step blocks were poured from the first metal out of the ladle and were intended to take advantage of the
flexibility that the specific foundries had in providing different manganese levels. The intent was to determine if the trends
observed in the first round of testing carried through when the manganese ranges and the manganese/sulfur relationship
ranges were extended.
MECHANICAL TESTING
Tensile bars were machined from the center of each of the steps from each of the test blocks. These test bars along with the
“B” bars were tested to determine ultimate tensile strength. Figure 10 documents the cutting to obtain the test bars from the
step blocks. Brinell Hardness per ASTM E10-01 was determined from the cross-sections of each step.
METALLOGRAPHIC ANALYSIS
Metallographic analysis was conducted on each test bar. To conduct this analysis, a longitudinal sample was removed from
each test bar at the fracture. This cross section was mounted in appropriate mounting material and was ground and polished
in accordance with standard metallographic procedures per ASTM E3-95. The resulting samples were then examined
unetched and etched with a metallurgical microscope at magnifications up to 1000X. The purpose for this examination was
to evaluate the microstructure at the fracture so that correlations between the structure at the fracture and the fracture results
could be realistically evaluated.
Paper 03-072 (05).pdf, Page 12 of 30
Factors evaluated metallographically at the fracture included:
Maximum Graphite Flake Length ASTM Graphite Size Class Eutectic Cell Count
Graphite Flake Branching* Ferrite Content Steadite Content
Carbide Content MnS Particle Count**
* Since one contention is that sulfur causes graphite branching and that the branching interacts with mechanical properties, an attempt was made to attach
significance to the branching issue with an estimate of the number of graphite flakes determined per cell. The values determined were identified as “Degree
of Branching”.
**Since MnS was a major part of this investigation, metallographic analysis included a qualitative assessment of the presence of MnS particles. The MnS
particle population was arbitrarily divided into five levels. Each metallographic sample was rated with level 5 being the highest population and level 1 being
the lowest.
Fig. 10. Step Block After Sectioning for Mechanical Testing.
EXPERIMENTAL RESULTS
Appendix I gives a complete compilation of all of the metallographic and mechanical property data obtained for the step
blocks poured at each foundry in this study. For reference this data is shown along with chemical composition and
processing information for each foundry.
METALLOGRAPHIC RESULTS
The principal metallographic features, which were affected by process variables, were maximum graphite flake size and
eutectic cell count. Metallographic evaluation showed that as step size increased, which resulted in a decrease in
solidification cooling rate, maximum graphite flake length increased and eutectic cell count decreased. Typical examples of
these results are shown for maximum graphite flake length (Foundries A, B,C,and D) in Figure 11, and for eutectic cell count
(Foundries E, F and G) in Figure 12.
Paper 03-072 (05).pdf, Page 13 of 30
Step Size, mm
0 20 40 60 80
120
100
Maximum Graphite Flake Length at 100X, mm
80
60
Foundry A
40 Foundry B
Foundry C
Foundry D
(Foundry A)
20
(Foundry C)
(Foundry D)
(Foundry B)
0
0 0.5 1 1.5 2 2.5 3 3.5
Step Size, in.
Fig. 11. Maximum Graphite Flake Length vs. Step Size.
Step Size, mm
0 20 40 60 80
10000
9000 14
8000 Foundry E 12
Eutectic Cell Count, #/sq.mm

Foundry F
Eutectic Cell Count, #/sq.in.
7000 Foundry G
(Foundry E) 10
(Foundry F)
6000 (Foundry G)
8
5000
4000 6
3000
4
2000
1000
0 0.5 1 1.5 2 2.5 3 3.5
Step Size, in.
Fig. 12. Eutectic Cell Count vs. Step Size.
There was no correlation between the other metallographic features noted above and cooling rate. Furthermore, none of the
metallographic features were influenced by pouring temperature or the addition of inoculant in the third mold.
Paper 03-072 (05).pdf, Page 14 of 30
MECHANICAL PROPERTIES
Mechanical properties were influenced primarily by solidification cooling rate. No correlations were observed between
mechanical properties and either pouring temperature or addition of inoculant to the third mold. Example data obtained from
the step blocks cast at Foundries A, B, C and D is shown in Figure 13 where tensile strength is plotted versus step size.
Step Size, mm
0 20 40 60 80
45 300
40
250
35

Ultimate Tensile Strength, KSI
30
200
25
150
20
Foundry A
15 Foundry B
100
Foundry C
Foundry D
10
(Foundry A)
50
(Foundry B)
5 (Foundry C)
(Foundry D)
0 0
0 0.5 1 1.5 2 2.5 3 3.5
Step Size, in.
Fig. 13. Ultimate Tensile Strength vs. Step Size.
As step size increased, and solidification cooling rate was reduced, tensile strength decreased. All of these trends were
considered characteristic of gray iron behavior as reported in handbooks and the technical literature. The results obtained in
this study were thus considered consistent with earlier data, and could therefore be used for further analysis.
The measured tensile strengths, however, depended not only on step thickness but also on the composition of each of the
irons. In order to evaluate the effects of sulfur and manganese alone on tensile strength, a mathematical procedure was
adopted to adjust the measured tensile strengths to compensate for variability due to differences in the amounts of other
elements. To accomplish this adjustment, calculated values of tensile strength were first obtained for each of the step blocks
using an equation, developed by Bates et. al., which relates tensile strength to composition and section size(12). This equation
is as follows:
UTS (KSI) = Factor A X Factor B (6)
Where:
Factor A = 101.1193 X (%TC + Si/4 + %P/2) + 4.3887/cast bar radius
Factor B = 1.00
+ 0.1371 X %Si
- 0.0021 X (%Mn-1.7X%S)
- 0.3131 X % S
+0.3562 X % Cr
+0.0282 X % Ni
+0.1107 X % Cu
+0.6297 X % Mo
-5.2985 X % Ti
-0.2305 X % Sn
Paper 03-072 (05).pdf, Page 15 of 30
The coefficients for each of the terms in Factor A and B were developed from a regression analysis. Measured values of
tensile strength obtained from the step blocks are plotted against calculated values in Figures 14 through 16. Figure 14 was
developed for the first two molds as described in the procedures above, Figure 15 for molds where additional inoculant was
added, and Figure 16 represents all of the data obtained. It is recognized that some of the data points in each of the figures
are from iron compositions that fell outside the limits of the regression analysis used to develop Equation 6. However the
Predicted Tensile Strength, MPa
120 180 240 3000 360 420
70
420
60
Measured Tensile Strength, MPa

Measured Tensile Strength, KSI
360
y = 0.5504x + 11.505
50 R2 = 0.6597
300
40
240
30
180
20
120
First Two Molds
10
10 20 30 40 50 60 70
Predicted Tensile Strength, KSI
Fig. 14. Measured Tensile Strength vs. Predicted Tensile Strength for the First Two Molds.

120 180 240 300 360 420
70
420
60 y = 0.7859x + 3.1909
Measured Ultimate Tensile Strength, KSI
R2 = 0.7437
360
50
300
40
240
30
180
20
120
Third Mold-Additional Inoculation
10
10 20 30 40 50 60 70
Predicted Ultimate Tensile Strength, KSI
Fig. 15. Measured vs. Predicted Tensile Strength for the Third Mold with Additional Inoculation.
Paper 03-072 (05).pdf, Page 16 of 30
120 180 240 300 360 420

70
420
60

All Data 360
Measured Tensile Strength, KSI
50
300
40
240
30
y = 0.6314x + 8.6141 180
R2 = 0.6672
20
120
10
10 20 30 40 50 60 70
Predicted Tensile Strength, KSI
Fig. 16. Measured vs. Predicted Tensile Strength for All Data.
correlation between measured and predicted values was considered sufficient to permit extending Equation (6) to include all
the data. Figures 14 through 16 were thus used to adjust values of measured tensile strength from adjusted calculated values.
To adjust the measured tensile strengths due to variations in the amounts of elements other than sulfur and manganese, the
tensile strength values for each of the step blocks were calculated from Equation 6, excluding the contributions from sulfur
and manganese. These results were then “normalized” to strengths calculated from the composition of Foundry B (C.E.
4.22), which was arbitrarily selected for this purpose. The differences in calculated tensile strengths between each of the
other foundries and Foundry B were determined. The measured tensile strengths for each of the foundries, other than
Foundry B, were “corrected” by an amount equal to these differences in calculated tensile strengths multiplied by the slope of
Figure 14 for data from the first two molds poured at each foundry or Figure 15 for data from the third mold poured at each
foundry where additional inoculant was added. Remaining variations in the “corrected” values of measured tensile strength
were attributed only to the effects of sulfur and manganese.
Figures 17-19 are graphs of corrected tensile strengths versus sulfur, manganese and excess manganese. All of the data were
obtained from the first two molds cast at each of the foundries. No correlation between tensile strength and sulfur,
manganese, or excess manganese was observed for the third mold where additional inoculant was added. This was probably
due to the ineffective nature of the process of casting the third mold which resulted in low metal temperature and poor
distribution of the inoculant.
Figure 17 is a graph of corrected tensile strength versus sulfur content for all of the foundries for the 3.0 inch step size.. A
reduction in corrected tensile strength occurs with increasing sulfur content. This effect was observed for all step sizes, but
was more pronounced at the larger step sizes.
Figure 18 shows the effect of manganese content on corrected tensile strength. The data shown are for the 1.6 inch step size.
The graph shows higher strengths at both low and high manganese levels, with a shallow minimum in strength at manganese
concentrations of 04-0.5 weight percent. This behavior was observed for all of the step sizes.
Figure 19 shows how corrected tensile strength varied with excess manganese. The data shown are again for the 1.6 inch step
size, but are representative of all four step sizes. The highest values of corrected tensile strength were observed at both low
and high levels of excess manganese. The lowest strengths were observed at excess manganese levels of 0.3-0.5 weight
percent.
Paper 03-072 (05).pdf, Page 17 of 30
35
220
Corrected Ultimate Tensile Strength, MPa

Corrected Ultimate Tensile Strength, KSI 30
180
25
20 140
15 100
3.0 Inch Step Size

10 60
First Two Molds
5
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Sulfur Content, Wt. %
Fig. 17. Corrected Ultimate Tensile Strength vs. Sulfur Content.
40
260

35
Corrected Ultimate Tensile Strength, KSI
220
30
180
25
140
20
1.6 Inch Step Size

15 100
First Two Molds
10
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Manganese Content, Wt.%
Fig. 18. Corrected Ultimate Tensile Strength vs. Manganese Content.

Paper 03-072 (05).pdf, Page 18 of 30
40
260

35
Corrected Ultimate Tensile Strength, KSI
220
30
180
25
140
20
1.6 Inch Step Size

15 100
First Two Molds
10
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Mn-1.7%S, Wt. %
Fig. 19. Corrected Ultimate Tensile Strength vs. Excess Manganese.
DISCUSSION
COMPARISON OF MEASURED AND PREDICTED TENSILE STRENGTHS
The plots of measured versus predicted tensile strength shown in Figures 14 through 16 show that the data obtained in this
study are reasonably consistent with the predictions of Equation 6. While a 1:1 relationship between measured and predicted
tensile strength was not observed, the correlation was quite reasonable. This is despite the fact that the data were obtained
under a variety of melting conditions at different foundries, and that some of the iron compositions were outside the limits of
the regression analysis used to obtain Equation 6.
COMPARISON OF RESULTS WITH OTHER STUDIES

The results of this study were compared with the findings of two other investigations, which also examined the effects of
sulfur and manganese on the mechanical properties of gray iron. The first was a study carried out by the British Cast Iron
Research Association (5). The BCIRA study consisted of melting several series of laboratory heats. Each series was melted to
a specific carbon equivalent and a specific sulfur level. For each carbon equivalent and sulfur level, heats were melted with
manganese levels which varied from <0.1 wt.% to as high as 0.84 wt.%. Some of the data obtained in this investigation are
given in Tables 3 and 4 above. A second study used for comparison was conducted by Prabhu et. al. (13). This investigation
also consisted of melting several laboratory heats to varying sulfur and manganese contents.
Figure 20 shows a combination plot of corrected tensile strength versus excess manganese for both of the aforementioned
investigations as well as for the current study. The 0.75 inch data from the current study are shown. Both the BCIRA data
and the data obtained by Prabhu et. al. were corrected to this section size using Equation (6). The carbon equivalents of the
BCIRA data selected were 4.16 and 4.29. The carbon equivalents for the data developed by Prabhu et. al. ranged from 4.03-
4.11. As was noted earlier the data from the current investigation was corrected to a composition with a carbon equivalent of
4.22. The data from the current investigation correlated best with the BCIRA results at a carbon equivalent of 4.16. As can
be seen, the trend of the tensile strengths from all three investigations as a function of excess manganese is the same.
Excellent correlation is achieved between the BCIRA data at a carbon equivalent of 4.16 and the results of the current
investigation and those of Prabhu et. al. These comparisons indicate that, after correcting for iron composition variations, the
effects of manganese and sulfur on mechanical properties observed in this study are consistent with other investigations.
Further comparisons will be made in subsequent discussion related to dissolved sulfur and manganese. It should be noted
that tensile strength has been observed to increase and decrease with increasing manganese or excess manganese content (5,13).
Figure 20 shows that an increase or decrease in strength depends on the range of manganese levels being investigated.
Paper 03-072 (05).pdf, Page 19 of 30
350
BCIRA C.E. 4.16
This Study 0.75 Inch Step Size, Corrected to C.E. 4.22
BCIRA C.E. 4.29
300
Prabhu et.al. C.E. 4.03-4.11
250
200
150
100
50
-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Mn -1.7S, Wt.%
Fig. 20. Comparative Plots of Corrected Tensile Strength vs. Excess Manganese.
ROLE OF DISSOLVED SULFUR AND MANGANESE

As was noted earlier, analysis of results from the literature indicate that there is a direct relationship between the observed
effects of sulfur and manganese in gray iron and the amounts of these elements remaining in liquid solution just prior to the
onset of solidification. The analysis of the BCIRA study, which is depicted in Figures 5 through 8, gives examples of this
relationship. A similar analysis was carried out for the results of the current investigation
1260
1240
1220
Liquidus Temperature, Deg. C
1200
1180
1160 y = -92.593x2 + 573.15x + 380.24
1140
1120
1100
3.4 3.5 3.6 3.7 3.8 3.9 4 4.1 4.2
Carbon Equivalent
Fig. 21. Liquidus Temperature vs. Carbon Equivalent. After Reference 10.
Paper 03-072 (05).pdf, Page 20 of 30
The liquidus temperatures for each of the iron compositions melted at the various foundries was estimated from Figure 21.
This figure was constructed from the data in Table 2. Using the procedures described in the Historical Review section above,
the equilibrium sulfur levels at the liquidus for each iron composition were determined. The corrected tensile strengths for all
of the iron compositions were then plotted versus the differences between total sulfur contained in each iron and the
equilibrium value. An example is shown in Figure 22 which shows data from all the foundries at a step size of 0.75 inches.
To fully develop the analysis, data from the BCIRA study at a carbon equivalent of 4.16 is included. Again, the BCIRA data
was adjusted for step size. As can be seen strength falls as the sulfur differential goes from a negative to a positive values.
The correlation of the data from the current investigation, involving “normal” production iron, with the BCIRA results was
quite good in all four step sizes. However, all of the data from this investigation fall in the narrow regime on the horizontal
axis where the sulfur differential is near zero, and where tensile strength is decreasing rapidly. The horizontal axis of Figure
22 was expanded in this region resulting in Figure 23. This graph presents a more detailed picture of the change in tensile
strength which occurs as the sulfur differential approaches a positive value.
Figure 24 shows the graph from Figure 22 with the addition of the data developed by Prabhu et.al. The results of the work by
Praghu et.al.also fall in the narrow regime where tensile strength is falling rapidly. Figure 25 shows a plot of Figure 24 where
the horizontal axis is again expanded to reveal increased detail regarding the change in tensile strength as the sulfur
differential approaches positive values. Reasonably good correlation among the three investigations can be observed.
Again using the procedures described in the Historical Review section above, the equilibrium manganese levels at the
liquidus for each iron composition were determined. The corrected tensile strengths for all of the iron compositions were
300
This Study
280 BCIRA Data C.E. 4.16
260
240
220
200 0.75 Inch Step Size
180
-1.6 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2
(Total Sulfur - Equilibrium Sulfur @ Liquidus), Wt.%
Fig. 22. Corrected Ultimate Tensile Strength vs. Total Sulfur Minus Equilibrium Sulfur.
then plotted versus the differences between total manganese contained in each iron and the equilibrium value. An example of
these results, along with the corresponding BCIRA data, is shown in Figure 26. While there is more data scatter than what
was observed in the sulfur analysis, the trends for both sets of data are similar. As the manganese differential goes from
negative to positive values, strength is decreased.
Paper 03-072 (05).pdf, Page 21 of 30
300
280

260
240
220
200
180
This Study
160
BCIRA Data C.E. 4.16
Linear (BCIRA Data C.E. 4.16)
120
Linear (This Study)
100
-0.15 -0.1 -0.05 0 0.05 0.1
(Total Sulfur - Equilibrium Sulfur @ Liquidus), Wt%
Fig. 23. Corrected Ultimate Tensile Strength vs. Total Sulfur Minus Equilibrium Sulfur.
Finally, Figure 27 shows, with an expanded horizontal axis, a comparison of the corrected tensile strengths for the 0.75 inch
step size data from this investigation with data from the BCIRA study and the results of the work by Prabhu et. al. Corrected
tensile strength falls as the manganese differential goes from a negative to a positive value.
280
260
240
220
200
This Study 0.75 Inch Step Size
Prabhu et.al.
180
160
-1.6 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2
(Total Sulfur-Equilibrium Sulfur @ Liquidus), Wt.%
Fig.24. Corrected Ultimate Tensile Strength vs. Total Sulfur Minus Equilibrium Sulfur.
Paper 03-072 (05).pdf, Page 22 of 30

300
280

260
240
220
200
180
160
This Study 0.75 Inch Step Size
140 BCIRA Data C.E. 4.16
Prabhu et. al.
120
100
-0.15 -0.1 -0.05 0 0.05 0.1
(Total Sulfur - Equilibrium Sulfur@ Liquidus), Wt.%
Figure 25. Corrected Ultimate Tensile Strength vs. Sulfur Differential.
300
280
260
240
220
200
180
160
140
This Study
100
-2 -1.5 -1 -0.5 0 0.5
(Total Manganese - Equilibrium Manganese @ Liquidus), Wt. %
Fig.26. Corrected Ultimate Tensile Strength vs. Total Manganese Minus Equilibrium Manganese.
Paper 03-072 (05).pdf, Page 23 of 30
300
280

260
240
220
200
180
160
This Study
140
Prabhu et.al.
100
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
(Total Manganese - Equilibrium Manganese @ Liquidus), Wt.%
Fig. 27. Corrected Ultimate Tensile Strength vs. Total Manganese Minus Equilibrium Manganese..
MAXIMIZING TENSILE STRENGTH BY CONTROLLING MANGANESE SULFIDE PRECIPITATION

The information developed in this study can be used to optimize manganese and sulfur levels to obtain maximum tensile
strength. The results show that maximum tensile strengths are obtained when MnS precipitation is suppressed to as low a
temperature as possible prior to the onset of solidification. The BCIRA laboratory data in Tables 3 and 4, and in Figures 5
and 6, shows that the reduction in tensile strength due to MnS precipitation varies with carbon equivalent, and ranges from
46-80 MPa (6.6-11.6 KSI). A practical estimate from the data in Figures 22 and 23 for the current investigation indicates that
a reduction in tensile strength of up to 50 MPa (7.2 KSI) could be anticipated due to MnS precipitation if it occurs before
solidification starts.
The ideal situation would be to suppress MnS precipitation to below the liquidus temperature. The liquidus temperature of
gray iron is largely dependent on carbon equivalent, and can thus be estimated from Figure 21. This temperature can then be
used in a series of calculations in Equation 5 to develop a curve that shows combinations of manganese and sulfur that are in
equilibrium at the onset of solidification (See Figure 3). If a manganese content is selected for a given gray iron composition,
sulfur level which is in equilibrium with this manganese level can be obtained from the curve. A sulfur level slightly less
than this value should maximize tensile strength since MnS precipitation would not occur. Conversely, if average sulfur
content is known, a manganese level could be obtained which would avoid MnS precipitation.
As an example, a gray iron with a carbon equivalent of 4.1 and a sulfur content of 0.09% can be considered. From Figure 21,
the liquidus temperature of 1175 °C (1447°K) is obtained. Using Equation 5 the solubility product curve for the liquidus,
shown in Figure 28, can be obtained. For the example sulfur level of 0.09%, equilibrium is established with a manganese
level of 0.52%. If sulfur is held constant at 0.09%, strength will be maximized if manganese is below 0.52%.
Correspondingly, at manganese content of 0.52%, reducing sulfur to below 0.09% will maximize strength. Under these
conditions, MnS precipitation will occur at a temperature below the liquidus. This temperature can be calculated from
Equation 5 using a manganese level and a sulfur level slightly below 0.52% and .09% respectively. The temperature at the
end of eutectic freezing can be considered a theoretical lower limit to MnS precipitation. The MnS solubility product curve
for the temperature at the end of eutectic freezing, 1081°C (14), is also shown in Figure 28. For the example being considered,
0.09 % sulfur is in equilibrium with 0.43% manganese, and 0.52% manganese is in equilibrium with 0.074% sulfur. A range
of manganese and sulfur levels can be identified between the two temperatures that can be used to define region A in Figure
28. Combinations of manganese and sulfur levels within this region would maximize strength. Conversely, combinations of
manganese and sulfur levels extending into region B in Figure 28 would result in MnS precipitation above the liquidus
temperature and a reduction in strength.
Figure 29 shows MnS solubility product curves for the liquidus temperature for each of several carbon equivalents. These
curves can be used to determine manganese and sulfur levels that avoid precipitation of MnS prior to the onset of
Paper 03-072 (05).pdf, Page 24 of 30
solidification using the procedures described above. It should be noted that, since liquidus temperature decreases with
increasing carbon equivalent, the maximum manganese level to avoid MnS precipitation above the liquidus decreases at any
given sulfur level. For example, the manganese level in equilibrium with 0.08% sulfur at a carbon equivalent of 3.7 is
0.66%. At a carbon equivalent of 4.3, the manganese level in equilibrium with 0.08 sulfur is reduced to 0.54 %.
0.75
C.E.: 4.1 Liquidus:
0.7 1174C (1447K)
0.65
0.6
Region B
Manganese, Wt.%
0.55
0.5 Region A
0.45
0.4
0.35
0.3 End of Eutectic Freezing:

1081C (1354K)
0.25
0.06 0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.14 0.15 0.16
Sulfur, Wt.%
Fig. 28. Manganese-Sulfur Solubility Product Data for a Carbon Equivalent of 4.1.
CONCLUSIONS
Based on the results of this investigation, the following conclusions can be drawn:
1. The data obtained exhibited the expected effects of solidification cooling rate on mechanical properties and
microstructure. Decreasing cooling rate by increasing step thickness decreased tensile strength, increased maximum
graphite flake length, and decreased eutectic cell count. No effects were observed on microstructure or properties as
a result of changes in pouring temperature or the addition of additional inoculant to the mold.
2. A reasonably good correlation was obtained between measured tensile strength and tensile strength predicted by a
regression analysis based on section size and chemical composition.
3. After normalizing measured tensile strength for variations in chemical composition other than manganese and sulfur,
it was found that, over the range of sulfur levels evaluated, tensile strength decreased with increasing sulfur level.
Furthermore, over the range of manganese levels evaluated, tensile strength first decreased and then increased with
increasing manganese level. In addition, tensile strength first decreased and then increased with increasing amounts
of excess manganese.
4. Tensile strength could be correlated with the precipitation of MnS prior to the onset of solidification. Maximum
tensile strength was achieved when sulfur or manganese content was less than that required for MnS precipitation at
the liquidus. When sulfur or manganese content was high enough to result in MnS precipitation prior to the onset of
solidification, tensile strengths were reduced.
5. The results of this work advances the understanding of the role of manganese, sulfur and Mn/S ratio on the
mechanical properties of gray iron, and help explain some of the variations observed to date.
Paper 03-072 (05).pdf, Page 25 of 30
Carbon Equivalent
3.7
1.3 3.8
3.9
4.0
4.1
4.2
1.1 4.3
Manganese, Wt.%
0.9
0.7
End of Eutectic Freezing
0.5
0.3
0.02 0.04 0.06 0.08 0.1 0.12
Sulfur, Wt.%
Fig. 29 Manganese-Sulfur Solubility Product Data at the Liquidus for Several Carbon Equivalents.
RECOMMENDATIONS FOR FUTURE WORK

The results of this work established that a correlation exists between MnS precipitation and tensile strength. The underlying
mechanism responsible for the correlation needs to be established. The precipitation of MnS after the onset of solidification
needs to be studied, and the partitioning of sulfur and manganese between solid and liquid during solidification needs to be
examined. As solidification proceeds, both manganese and sulfur will be rejected from the solid to the liquid. Manganese
however is soluble in the solid whereas sulfur is extremely limited. The compositions of the manganese sulfide particles
formed prior to the onset of solidification need to be compared with those formed during the solidification process. The role
of MnS precipitation on the solidification process itself, and the effects of MnS precipitation on elements of the
microstructure that control strength, such as graphite flake length and volume fraction as well as eutectic cell count, need to
be determined.
This work was undertaken to evaluate the effects of manganese and sulfur on tensile properties only. Brinell hardness values
were obtained, however many factors can influence hardness and tensile strength to hardness ratio. The effects of manganese
and sulfur on hardness and tensile strength to hardness ratio need to be investigated further.
Finally, adjusting manganese and sulfur to maximize tensile strength could also affect machinability. A study investigating
the effects of manganese sulfide formation on machinability should be undertaken. Results from such a study would help
avoid adverse effects of increased tensile strength on machinability.
ACKNOWLEDGEMENTS
The authors wish to extend their sincere appreciation to the following foundries that provided the step block castings for this
investigation. The participating foundries included M&H Valve Co., Blackhawk Foundry and Machine Co., Rochester Metal
Products Corporation, Kohler Co., Brillion Iron Works, Grede Foundries Inc. Iron Mountain Division and Accurate Castings.
The authors also wish to thank Anderson Laboratories, Inc. and Climax Research Services for providing chemical analysis
and mechanical property testing. The authors wish to thank Mr. Dale McGuffin of Professional Metallurgical Services for
his efforts in preparing samples for chemical, mechanical property and metallographic analysis. Finally the authors wish to
acknowledge the American Foundry Society Cast Iron Division (Divison 5), Gray Iron Reseach Committee (5H) for their
overall support of this program.
Paper 03-072 (05).pdf, Page 26 of 30
REFERENCES
1.) Boyes, J.W., “Influence of Manganese and Sulfur Contents on Chill and Mottle Formation”, Report No. 664,
BCIRA J., 1962, p. 568.
2.) “The Effects of Alloying Elements in Cast Iron-2: Sulfur, Report No. 1229, BCIRA J., May, 1976, p. 271.
3.) Fuller, A.G., “The Effect of Manganese and Sulfur on Mechanical Properties and Structure of Flake Graphite Cast
Irons”, AFS Transactions, 1986, p. 863.
4.) Fuller, A.G., “Further Studies on Some Factors Influencing Sinking Under Bosses on Thin Plates: The Role of
Eutectic Nucleation”, Report No.478, BCIRA J., 1958, p. 157.
5.) Alderson, A., “The Influence of Manganese and Sulfur on the Structure and Mechanical Properties of Grey Cast
Iron”, Report No. 1512, BCIRA J., 1983, p. 59.
6.) Bates, C.E., “Alloy Element Effects on Gray Iron Properties: Part II”, AFS Transactions, 1986, p. 889.
7.) Muzumdar, K.M. and Wallace, J.F., “Effect of Sulfur in Cast Iron”, AFS Transactions, 1973, p. 412.
8.) “The Effects of Alloying Elements in Cast Iron-4: Manganese, Report No. 1245, BCIRA J., November, 1976, p.570.
9.) Gundlach, R.B., Climax Research Services, Private Communication.
10.) Heine, R.W. and Rosenthal, P.C. “Principles of Metal Casting”, McGraw-Hill, 1955, p. 190.
11.) Goodrich, G.M. and Oakwood, T.G., Professional Metallurgical Services, Unpublished Research.
12.) Bates, C.E., Tucker, J.R. , and Starrett, K.S., “Composition, Section Size and Microstructural Effects on the Tensile
Properties of Pearlitic Gray Cast Iron”, AFS Research Report No.5, Sept. 1991.
13.) Prabhu, K., Lover, C.R. Jr.,Cornell, H.H., “Influence of Manganese and Sulfur on the Strength, Hardness and
Microstructure of Gray Cast Iron, AFS Transactions, 1988, p.287.
14.) “Cast Irons”, ASM Specialty Handbook, Davis, J.R., ed., ASM International, 1996.
Paper 03-072 (05).pdf, Page 27 of 30
APPENDIX 1
PROCESSING, COMPOSITIONS, PROPERTIES AND METALLOGRAPHIC RESULTS
Foundry A Chemical Composition, Wt. %
C S P Si Mn Cr Ni Mo Al Cu Sn V Mg C.E.
Lab 1 2.88 0.078 0.09 1.87 0.37 0.16 0.07 0.02 <0.005 0.16 3.53
Lab 2-1 3.04 0.072 0.09 2.4 0.43 0.19 0.07 <0.01 <0.01 0.14 <0.01 0.01 <0.01 3.87
Lab 2-2 3.07 0.07 0.089 2.48 0.38 0.19 0.07 0.01 <0.01 0.14 <0.01 <0.01 <0.01 3.93
Other Elements (Foundry A): None
Tensile Properties Processing Metallography

Mold # Step Size UTS (KSI) BHN UTS/BHN* Pour Temp. Extra Mold Inoc. MFL @ Fr* Gr SC Deg. Br. MnS Rat. % Pear. % Ferrite % Steadite %Carbide Cell Ct.**
1 0.41 36.5 197 185 Deg C 20.32 4 43 1 99+ Tr Tr 0 4477
1 0.75 39.26 207 190 Hot No 45.72 3 20 1 98 1 1 0 2819
1 1.6 35.8 201 178 50.8 2 20 2 98 1 1 0 4181
1 3 33.78 207 163 40.64 3 14 2 99 1 Tr Tr 3368
2 0.41 36.83 183 201 Cold Yes 22.86 4 15 1 85 15 Tr 0 9142

2 0.75 35.94 187 192 35.56 3 20 1 89 10 1 0 4394
2 1.6 35.46 187 190 45.72 3 16 1 85 15 Tr 0 7432
2 3 35.56 179 199 63.5 3 16 1 84 15 1 0 4297
3 0.41 39.52 183 216 Cold No 22.86 3 14 1 98 2 Tr 0 5645

3 0.75 38 192 198 45.72 2 17 1 98 1 1 0 3019
3 1.6 35.3 187 189 48.26 2 20 2 96 3 1 0 2439
3 3 31.21 179 174 60.96 2 14 2 94 5 1 0 2348
*Per SAE J431 (Aug 96)
B Bar 37.31
Note: No initial Inoculant *mm, 100X **No./Sq in.
Chem. Anal.: Lab 1 Lab 2-1 Lab 2-2
Calc. Liquidus Temp: Deg. C 1250 1212 1203
Equil. MnS Temp: Deg. C 974 999 945
Equl. %S at Liquidus 0.148 0.118 0.132
Equil. % Mn at Liquidus 0.703 0.71 0.071
Foundry B Chemical Composition, Wt. %
C S P Si Mn Cr Ni Mo Al Cu Sn V Mg Ti Pb C.E.
Lab 1 3.23 0.11 0.033 2.2 0.45 0.07 0.05 0.01 <0.005 0.19 3.97
Lab 2-1 3.5 0.09 0.028 2.16 0.49 0.08 0.06 <0.01 <0.01 0.18 <0.01 <0.01 <0.01 4.22
Lab 2-2 3.48 0.094 0.026 2.29 0.47 0.07 0.06 <0.01 <0.01 0.18 <0.01 0.01 <0.01 4.25
Lab 2-3S 3.47 0.12 0.021 2.33 0.57 0.08 0.08 0.02 <0.01 0.23 <0.01 <0.01 <0.01 0.01 <0.01 4.25
Other Elements (Foundry B): None

1 0.41 30.67 170 180 Deg. C 38.1 2 17 2 97 3 0 0 6929
1 0.75 32.05 179 179 1438 No 53.34 2 27 3 98 2 0 0 4529
1 1.6 28.07 167 168 58.42 1 18 2 98 2 0 0 2658
1 3 24.33 163 149 91.44 1 32 3 99 1 0 0 2406
2 0.41 17.42 156 112 1331 Yes 30.48 3 22 1 25 75 0 0 6116

2 0.75 28.64 163 176 55.88 2 22 3 80 20 0 0 4703
2 1.6 24.99 163 153 81.28 1 30 2 85 15 Tr 0 2425
2 3 19.28 149 129 99.06 2 40 3 75 25 Tr 0 2465
3 0.41 21.38 149 143 1331 or 1262 No 33.02 2 28 2 50 60 Tr 0 5619

3 0.75 28.85 156 185 104.14 2 26 3 80 20 0 0 2645
3 1.6 23.97 156 154 106.68 1 32 2 60 40 Tr 0 2671
3 3 16.43 140 117 106.68 2 23 2 40 60 0 0 2845
B Bar 35.31
*mm, 100X **No./Sq in.
Supp. 0.41 29.78
0.75 27.92
1.6 25.21
3 20.78
Supp. B Bar 37.19 37.19
Chem Anal: Lab 1 Lab 2-1 Lab 2-2 Lab 2-3 Supp.
Calc. Liquidus Temp: Deg. C 1196 1150 1143 1143
Equil. MnS Temp: Deg. C 1198 1143 1144 1375
Equil. % S at Liquidus 0.109 0.09 0.094 0.077
Equil. % Mn at Liquidus 0.45 0.5 0.47 0.37

Paper 03-072 (05).pdf, Page 28 of 30
Foundry C Chemical Composition, Wt. %
C S P Si Mn Cr Ni Mo Al Cu Mg Ti Ce Sn V C.E.
Lab 1 3.1 0.057 0.056 1.98 0.6 0.07 0.03 0.01 0.005 0.32 <0.005 0.006 0.0023 <0.005 3.79
Lab 2-1 3.42 0.054 0.05 1.97 0.64 0.06 0.04 <0.01 <0.01 0.29 <0.01 <0.01 0.01 4.09
Lab 2-2 3.4 0.052 0.048 1.98 0.63 0.07 0.04 <0.01 <0.01 0.29 <0.01 <0.01 0.01 4.08
Additional Elements (Foundry C): None

1 0.41 34.71 212 164 25.4 3 (B) 34 1 99+ Tr Tr 0 2819
1 0.75 38.63 212 182 1444 No 50.8 2 17 1 100 0 Tr 0 2426
1 1.6 35.42 212 167 86.36 2 24 2 99+ Tr Tr 0 1510
1 3 30.27 197 154 63.5 2 21 1 99+ Tr Tr 0 1755
2 0.41 35.64 207 172 1265 No 22.86 3 (A&D) NA 1 85 15 Tr 0 3561

2 0.75 34.56 207 167 38.1 4(B) 40 1 90 10 Tr 0 1632
2 1.6 30.33 192 158 50.8 2 17 1 85 15 Tr 0 1329
2 3 26.98 179 151 53.34 2 23 1 85 15 Tr 0 2000
3 0.41 39.03 201 194 1265 Yes 27.94 3 23 1 95 5 Tr 0 5284

3 0.75 37.71 207 182 48.26 2 16 1 99 1 Tr 0 2252
3 1.6 34.19 197 174 43.18 2 19 1 95 5 Tr 0 1942
3 3 25.29 187 135 60.96 2 14-17 2 90 10 Tr 0 1903
B Bar 35.66 *Per SAE J431 (Aug 96)

*mm, 100X **No./Sq in.
Chem Anal: Lab 1 Lab 2-1 Lab 2-2
Equil. % S at Liquidus 0.087 0.073 0.075
Foundry D Chemical Composition, Wt. %
Lab 1 3.16 0.092 0.11 2.29 0.64 0.05 0.04 0.02 <0.005 0.57 <0.005 0.011 0.0059 0.006 3.96
Foundry D 3.42 0.072 0.098 2.375 0.65 0.057 0.04 0.014 0.003 0.575 0 0.013 0.0153 0.007 0.008 4.24
Lab 2-1 3.46 0.08 0.1 2.35 0.68 0.05 0.06 <0.01 <0.01 0.59 <0.01 <0.01 0.01 4.28
lab 2-2 3.45 0.082 0.11 2.45 0.69 0.05 0.05 <0.01 <0.01 0.59 <0.01 <0.01 0.01 4.3
Other Elements (Foundry D): Pb: 0.00072, As: 0.007, Sb: 0.003 , Co: 0.004, B: 0.0004, La: 0.0012, Nb: 0.0019, W: 0.0017
Tensile Properties Metallography

1 0.41 30.93 163 190 Deg. C 27.94 2 26 3 94-96 3.0-5.0 1 IC 0 7754
1 0.75 29.44 187 157 1387 No 48.26 3 16 1 98 0 2 0 3239
1 1.6 25.66 174 147 63.5 2 17 2 98 Tr 2 0 1510
1 3 26.4 170 155 55.88 2 25 1 99+ Tr Tr 0 1858
2 0.41 29.58 167 177 1347 No 27.94 3 30 1 93 7 Tr 0 4794

2 0.75 29.14 187 156 38.1 2 18 2 95 3 2 0 2639
2 1.6 24.9 179 139 73.66 2 21 1 99+ Tr Tr 0 1677
2 3 22.28 170 131 58.42 1 27 2 97 3 Tr 0 1361
3 0.41 31.1 163 191 1347 Inoculoy 63 20.32 3 12 1 93 7 Tr 0 8471

3 0.75 29.65 179 166 73.66 2 22 1 99+ Tr Tr 0 4684
3 1.6 25.72 174 148 76.2 1 21 1 100 0 Tr 0 1677
3 3 24.65 167 148 68.58 1 24 2 99+ Tr Tr 0 2052
* Per SAE J431 (Aug 96)
B Bar 30.65 Note: Mold #1 Pour Short *mm, 100X **No./Sq in.
B Bar 32.251
BHN 197
Chem. Analysis: Lab 1 Foundry D Lab 2-1 Lab 2-2
Equil. S at Liquidus: 0.077 0.068 0.064 0.062

Paper 03-072 (05).pdf, Page 29 of 30
Foundry E Chemical Composition, Wt. %
C S P Si Mn Cr Ni Mo Al Cu Mg Ti Ce Sn Pb Sb V C.E.
Lab 1
Foundry E 3.32 0.016 0.016 1.97 0.75 0.168 0.116 0.074 0 0.236 0.003 0.0148 0 0.007 0 0.003 0.0121 3.98
Lab2-1 3.48 0.018 0.017 1.97 0.7 0.16 0.12 0.05 <0.01 0.22 <0.01 <0.01 0.02 4.14
Lab 2-2 3.5 0.018 0.017 2.04 0.72 0.17 0.12 0.05 <0.01 0.23 <0.01 <0.01 0.02 4.19
Lab 2-3S 3.5 0.025 0.019 2.03 0.55 0.11 0.11 0.03 <0.01 0.44 <0.01 <0.01 0.01 <0.01 0.02 4.18
Other Elements (Foundry E): None

Mold # Step Size UTS (KSI) BHN UTS/KSI* Pour Temp. Extra Mold Inoc. MFL @ Fr* Gr SC Deg. Br. MnS Rat. % Pear. % Ferrite % Steadite %Carbide Cell Ct.**
1 0.41 37.65 170 221 Deg. C 30.48 3 24 1 99+ Tr 0 Tr 7606
1 0.75 31.37 187 168 1454 No 66.04 2 18 1 97 3 0 Tr 3768
1 1.6 35.12 187 188 58.42 2 23 1 97 3 0 Tr 1632
1 3 30.27 174 174 76.2 1 22 1 99+ Tr 0 Tr 1400
2 0.41 34.7 174 199 1387 No 27.94 3 24 1 99+ Tr 0 Tr 5355

2 0.75 32.69 187 175 53.34 2 14 1 98 2 Tr Tr 4039
2 1.6 35.72 197 181 50.8 2 14 1 98 2 Tr 0 1839
2 3 30.47 179 170 50.8 2 30 1 99+ Tr Tr 0 1439
3 0.41 34.51 163 212 1387 Inoc 90 35.56 2 25 1 99 1 0 Tr 6568

3 0.75 27.86 187 149 50.8 2 18 1 99 1 Tr Tr 4148
3 1.6 29.26 187 156 53.34 2 19 1 98 2 0 Tr 1981
3 3 27.89 179 156 83.82 2 36 1 99 1 0 0 1800
* Per SAE J431 (Aug 96)
B Bar 41.4 *mm, 100X **No./Sq in.
Note: B Bar From Foundry E

Note: Foundry E "Inside BHN": 187
Supp. 0.41 26.66

0.75 30.72
1.6 27.1
3 21.26
Chem. Analysis Foundry E Lab 2-1 Lab2-2 Lab 2-3 Supp.
Equil. % S at Liquidus 0.066 0.066 0.063 0.082
Foundry F Chemical Composition, Wt. %
Lab 1 2.79 0.1 0.042 2.11 0.56 0.12 0.07 0.26 0.007 0.23 <0.005 0.011 0.0028 0.062 3.51
Lab 2-1 2.93 0.1 0.036 2.27 0.6 0.13 0.07 0.31 <0.01 0.21 <0.01 0.02 0.01 3.7
Lab 2-2 2.92 0.1 0.038 2.3 0.59 0.12 0.07 0.28 <0.01 0.21 <0.01 0.03 0.01 3.7
Other Elements (Foundry F): None

1 0.41 49.03 235 209 Deg. C 30.48 3 4 to 6 4 100 0 Tr Tr 7716
1 0.75 40.67 241 169 No 38.1 3 7 4 100 0 Tr Tr 2497
1 1.6 36.17 241 150 30.48 3 8 2 100 0 Tr Tr 1735
1 3 36.07 229 158 58.42 2 10 3 100 0 Tr Tr 1774
2 0.41 46.63 229 204 No 22.86 4 8 2 100 0 Tr Tr 5897

2 0.75 40.92 235 174 40.64 2 7 3 99+ Tr lTr Tr 3187
2 1.6* 31.46 235 134 43.18 2 18 3 96 3 Tr 1 1961
2 3* 33.61 229 147 58.42 3 8 2 99 0 Tr 1 1652
3 0.41 48.62 229 212 Yes 30.48 3 8 3 100 0 Tr Tr 5890

3 0.75 47.74 241 198 27.94 2 10 2 100 0 Tr Tr 3484
3 1.6 48.15 235 205 33.02 2 9 3 100 0 Tr Tr 2045
3 3 45.03 229 197 45.72 2 9 3 100 0 Tr Tr 1877
B Bar 58.22 *mm, 100X **No./Sq in.
* Sample Broke at Large Internal Flaw in Grip Area

** Sample Broke at Flaw
Chem. Analysis: Lab 1 Lab 2-1 Lab 2-2

Paper 03-072 (05).pdf, Page 30 of 30
Foundry G Chemical Composition, Wt. %
Lab 1 3.58 0.085 0.041 2 0.19 0.02 0.02 0.01 <0.005 0.02 <0.005 0.015 0.0016 <0.005 <0.01 4.26
Lab 2-1 3.57 0.076 0.07 2.2 0.22 0.01 0.03 <0.01 <0.01 0.03 <0.01 <0.01 4.32
Lab 2-2 3.56 0.076 0.035 2.69 0.21 0.01 0.03 <0.01 <0.01 0.02 <0.01 <0.01 <0.01 4.47

1 0.41 32.45 167 194 Deg. C 30.48 3 15 1 100 0 Tr 0 8264
1 0.75 32.73 183 179 No 55.88 2 10 2 99+ Tr Tr 0 5432
1 1.6 26.59 174 153 114.3 1 14 3 75 25 Tr 0 2265
1 3 21.93 163 135 96.52 2 16 2 98 2 Tr 0 2142
2 0.41 33 179 184 No 33.02 2 15 3 100 0 Tr Tr 5032

2 0.75 32.31 192 168 58.42 1 14 3 100 0 Tr 0 3477
2 1.6 29.03 183 159 66.04 2 20 2 99+ Tr Tr 0 1877
2 3 23.69 179 132 73.66 2 17 2 99+ Tr 0 0 1503
3 0.41 28.63 170 168 Yes 33.02 2 18 1 100 0 Tr 0 7329

3 0.75 25.08 174 144 55.88 2 19 2 99+ Tr Tr Tr 2865
3 1.6 20.72 170 122 66.04 2 22 3 90 10 Tr Tr 2103
3 3 20.94 159 132 78.74 0 19 2 98 2 Tr 0 1716
B Bar 34.27 *mm, 100X **No./Sq in.
Chem. Analysis Lab 1 Lab 2-1 Lab 2-2
Calculated Liquidus Temp: Deg. C 1141 1128 1092

Manganese, Sulfur and Manganese-Sulfur Ratio Effects in Gray Cast Iron

Uploaded by

Copyright:

Available Formats

Manganese, Sulfur and Manganese-Sulfur Ratio Effects in Gray Cast Iron

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Manganese, Sulfur and Manganese-Sulfur Ratio Effects in Gray Cast Iron

Uploaded by

Copyright:

Available Formats

Paper 03-072 (05).

AFS Transactions 03-072 Page 1 of 30

Manganese, Sulfur and Manganese-Sulfur Ratio Effects in Gray Cast Iron

Copyright 2003 American Foundry Society

AFS Transactions 03-072 Page 2 of 30

Table 1. Effects of Sulfur and Manganese on Microstructure

Chill and Mottle Depth

Eutectic Cell Count

Mn (in soln) + S (in soln) ↔ MnS (1)

A similar expression describes the equilibrium between iron and sulfur:

Fe (liq) + S (in soln) ↔ FeS (2)

AFS Transactions 03-072 Page 3 of 30

Mn (in soln) + FeS ↔ Fe (liq) + MnS (3)

Mn% = 1.7(S%) + 0.3 % (4)

Fig. 1. Ultimate Tensile Strength vs Sulfur Content. After Reference 5.

USE OF SOLUBILITY PRODUCT DATA TO DESCRIBE EFFECTS OF SULFUR AND MANGANESE

AFS Transactions 03-072 Page 4 of 30

Log10 (% Mn x % S) = −1920/T (5)

310 Mn/S = 1.7 + 0.1

Mn/S = 1.7 + 0.2

Fig. 2. Ultimate Tensile Strength vs. Manganese Content. After Reference 5.

AFS Transactions 03-072 Page 5 of 30

Table 2. MnS Formation During Solidification (0.11% S, 0.49% Mn)

Carbon Equivalent Liquidus Temperature Ppt. Of MnS Before

AFS Transactions 03-072 Page 6 of 30

Table 3. Tensile Strengths For Various Manganese Levels (References 3,5)

Manganese Tensile Strength

AFS Transactions 03-072 Page 7 of 30

Table 4.Tensile Strengths For Various Manganese Levels (References 3,5)

Manganese Tensile Strength

Cooling Rate Carbon Equivalent

AFS Transactions 03-072 Page 8 of 30

Ultimate Tensile Strength, MPa

AFS Transactions 03-072 Page 9 of 30

AFS Transactions 03-072 Page 10 of 30

Fig. 9. Step Block Mold.

FOUNDRY SELECTION CRITERIA

Table 5. Nominal Composition Comparison

Foundry A did not employ any inoculation practice.

AFS Transactions 03-072 Page 11 of 30

Table 6. Nominal Compositions of Supplementary Castings

AFS Transactions 03-072 Page 12 of 30

Factors evaluated metallographically at the fracture included:

Fig. 10. Step Block After Sectioning for Mechanical Testing.

AFS Transactions 03-072 Page 13 of 30

Fig. 11. Maximum Graphite Flake Length vs. Step Size.

Eutectic Cell Count, #/sq.mm

Fig. 12. Eutectic Cell Count vs. Step Size.

AFS Transactions 03-072 Page 14 of 30

Ultimate Tensile Strength, MPa

Fig. 13. Ultimate Tensile Strength vs. Step Size.

UTS (KSI) = Factor A X Factor B (6)

AFS Transactions 03-072 Page 15 of 30

Measured Tensile Strength, MPa

Predicted Tensile Strength, MPa

AFS Transactions 03-072 Page 16 of 30

Predicted Tensile Strength, MPa

120 180 240 300 360 420

Measured Tensile Strength, MPa