Even after a period of 100 years, there remains some mystery about the Iron Carbon diagram. Not only because of the existence of a double system, or the fact that structure formation in high carbon alloys can hardly be explained by this diagram, but also because of the variety of diagrams that are available, as is shown by fig.1 and 2.
Figure 1 "European" version of the Fe-C diagram . Figure 2 "American" version of the Fe-C diagram.
The "safest" way to live with the existing iron carbon diagram(s) is just to follow the general explanations and opinions that are given on this matter. Why take the risk to burn your hands on something that is so complicated and thoroughly researched by so many scientists? The outcome of my research work, which is published on this Website, however, revealed so many new facts, that it seems to be worthwhile to take the risk to be condemmed forever for having attacked this "sacred" diagram! .
In The Iron-Carbon double diagram, a fundamentally wrong concept?, it was shown that the so-called ledeburite "eutectic" actually consists of solid iron-carbide in which small primary austenite dendrite-branches are embedded. After grinding, polishing and etching, sections through these small dendrite branches appear as isolated islands surrounded by an iron-carbide phase. It was also shown that in the case of a hypereutectic white structure, the typical shape of primary carbide needles is only the result of an optical effect caused by needle-shaped austenite dendrites that are found alongside the carbide needles. Once the austenite needle ends, the "primary" carbide needle just flows over this austenite tip and merges with the rest of the carbide phase. Actually, there appears to be no difference between "eutectic" and "primary" carbides. The final shape of all carbide phases is forced upon by the primary austenite which means that the carbide phase was the last to solidify. As primary austenite continues to grow untill the solidus temperature is reached, all carbides must also form at the solidus temperature. This implies that right at the solidus temperature, local liquid pools, situated between primary austenite, must have been present, which can transform directly into the chemical compound Fe3C. These liquid pools must have had a carbon content of this composition, i.e. 6,7% Carbon!!
If this statement is correct, it means that a real liquidus line should be drawn from 15400C at 0% carbon to 1150OC at 6,7% carbon. This new liquidus line is shown in red in figure 3. The region beyond 6,7% carbon remains open for further research.
This major change in the Iron-carbon diagram makes it possible:
To replace the double diagram by a single one
To explain the huge variations found in the positions of carbide liquidus lines
To account for the contradictionary results obtained in establishing the graphite liquidus
To account for the numerous graphite formation mechanisms that exsist.
Comparison of the New Liquidus line with the existing Fe-C diagram.
Regions within the "old" diagram that will undergo major changes :
Austenite liquidus
Carbide liquidus
Graphite liquidus
"Eutectic" points
1-Austenite liquidus
If the position of the new liquidus line is correct, it means that under normal conditions, primary austenite always forms with a high degree of undercooling. If this is true, it also means that austenite formed in this way will be supersaturated with carbon!. The consequences hereof on graphite formation in cast irons, will be dealt with on a separate Webpage.
Does literature provide information indicating a tendency for austenite to undercool? Although scattered, many publications on this subject, covering a range from pure iron up to the hypereutectic region can be found.
-Undercooling of pure iron.
When pure iron is melted under a slag layer, a carrier melt or a combination of both, undercooling of the austenite liquidus temperature of 3000 C can be reached [1], as is shown in fig.4.
-Austenite undercooling in Steel
Melting conditions appear to influence the liquidus point in all iron-carbon alloys. Especially the presence of oxygen has a marked effect on the actual liquidus temperature. In steel production it has even be used to estimate the oxygen level of the steel. Liquidus temperatures of base material and aluminium deoxidized samples are determined and the difference related to the oxygen content of the base material [2]. This correlation is shown in fig.5.
-Austenite undercooling in malleable Cast iron.
A similar method was used by Jacobs [3]to measure the degree of oxidation in malleable iron. Actual liquidus temperatures were compared with theoretical values calculated from the chemical composition of the material. Oxidized melts always showed higher liquidus temperatures as compared to deoxidized ones [3,4,5,6].
Attempts to correlate liquidus temperatures with chemical compositions, lead to the construction of various equations. When compared to each other, differences in liquidus temperatures of 500C are obtained for the same composition [7], see figure 6.
Very small bismuth additions are capable of lowering the liquidus temperature with as much as 450C [8]!.
-Austenite undercooling in gray cast iron.
Oxidizing and non-oxidizing melting conditions show a similar influence on austenite liquidus temperature in this range of compositions as they do in the lower carbon alloys [4,9], attempts to correct for austenite undercooling failed (read: did not show any practical advantage)[10]. And the issue was abandonned. See fig.7,
-Effect inoculation on austenite undercooling.
Although the effect of inoculation on cast irons has always been related to graphite formation during the "eutectic" part of solidification, Lux [11] showed that its effect is twice as large on the liquidus arrest as compared to undercooling changes in the "eutectic" part in the case of pure Fe-C alloys, as is shown by figure 8.
This publication seems to be a typical example of the way we have been looking at cast iron solidification: not a single word was spend to mention this rather spectacular behaviour in the liquidus arrest region!
Only a single publication points to the effect of inoculation on the position of liquidus temperatures [12].
Interpretation of a liquidus arrest in cooling curves taken on cast iron is not always easy. Cooling curves recorded in cups when poured successively, can reveal considerable differences in the shape and position of the liquidus arrests , as is shown in figure 9. Observation of the metallographic structure does not allow an explanation of these differences [13].
Figure 9, Liquidus arrest recorded in pairs of simultaneously poured commercial cups. For clarity, cooling curves have been shifted along x-axis.
In figure 10, variations of austenite liquidus lines are added and indicated in green.
-Muhlberger effect.
Measurements of the specific volume of liquid cast iron showed that some strange effects occurred between 60 to 80 above the actual liquidus temperature [14]. Within this region, liquid iron suddenly expands, which disappears shortly afterwards. This behaviour is assumed to be the result of a pre-orientation on a molecular scale, or a pre-crystallisation of austenite dendrites in the hypereutectic region[15]. In figure 11, this "Muhlberger effect" is added in blue.
Russian research work [16]showed density changes in a temperature range about 150-1600C above the liquidus line. It is assumed that this represents the formation of carbon-enriched colonies.Drawn as a purple line in figure 11.
-The Eutectic point.
Probably the most characteristic and at the same time most discussed feature of cast iron is the presence of a "eutectic".The fact that even two completely different "eutectics" can form of which one does not look like a eutectic at all has, kept many researchers puzzling for decades. Only after the acceptance of the concept of an "abnormal eutectic", discussions ceased, although even this flexible concept is not capable of explaining the hughe shape differences between nodular and flake graphite.
Incidently, however, publications can be found which put some question marks at existing thoughts:
Dodonov [17] as a result of his investigation of austenite formation in cast iron:
"The general impression is that the entire system has been permeated by fine disperse graphite formations, some of which were captive in the residual austenite and did not grow subsequently, i.e. during the eutectic transformation. It should be pointed out that these findings conflict with the basic principles of the traditional metallography of multi-component alloys."
De Sy[18] in his work to find the mechanism behind nodular graphite formation:
"In a shrinkage cavity of a definitely hypereutectic nodular iron, we observed well formed macroscopic dendrites(fig.2).The composition of a small dendrite itself was checked and found almost identical to that of the casting.Does a eutectic form dendrites? We don't believe it does; sometimes small dendrites of one of the phases but never dendrites of the eutectic itself. Therefore, and for many other reasons, it is believed that nodular iron solidifies without eutectic but in a mixed cristals system, forming supersaturated austenite in which graphite precipitation occurs, beginning at a certain degree of supersaturation, and spherolutic growths follows by a peritectic transformation or reaction."
In their attempt to follow solidification of cast iron with a High-temperature microscope, Siepman et al [19] had to admit that in no case they were able to observe direct formation of flake graphite.
Lohberg and Reihim [20], using directional solidification techniques, establish the fact that their experiments point to a solidification mechanism of a binary alloy of 5.3% carbon rather than of a eutectic solidification. This point has also been added as a red + sign figure 12.
-The Hypereutectic region. The Hypereutectic region of the Iron-Carbon diagram.,
has dealt extensively with this subject. Although this page was entirely composed of publications from all over the world, it's conclusions are completely different from the general accepted idea's on what really happens in this region.
From all publications that tried to proove the existance of a graphite liquidus, roughly 50% was unable find any indication in their cooling curves.
Twenty-five percent found some sign of liquidus arrest and assume that this points to a graphite liquidus.
The remaining 25% also found liquidus arrests, but after microscopical examination it always turned out to point to the formation of austenite dendrites!
An interesting aspect in this region is the fact that after magnesium additions, the traditional graphite liquidus seems to shift towards higher carbon equivalents with as much as 550 C.
Figure 13
This liquidus shift has been added in orange in figure 14.
Similar effects can be found with additions of Ce and Calcium [21]. The graphite liquidus in the hyper-eutectic region is lowered, whilst the austenite liquidus in the hypoeutectic range is raised!
Figure 14.
The hypereutectic metastable region has been dealt with in:
The Iron-Carbon double diagram, a fundamentally wrong concept?,
From all liquidus lines presented over the years, the one positioned near the solidus [23] appears to be the most justifiable one.
In figure 15, this line has been added in yellow.
Figure 15.
Concluding remarks.
If all facts and ideas are put together, it is possible to replace the existing Iron carbon double diagram by a new diagram of which the principle is shown in figure 16 below.
The actual shape of the liquidus (concave,convex or straight) is not known as is the minimum temperature at which liquid iron can still contain 6.7% carbon. It is known for instance that carbides can form at temperatures as low as 1080 C by alloying with elements such as silicon that only dissolve in the austenite without changing the carbide composition!
With this new diagram, it is nescessary and possible to take another look at the reason behind graphite formation in cast iron.
References:
[1]Unterkuhlungsmessungen an reinem Eisen,Kobalt und Nickel sowie Eisen-Kohlenstoff-, Nickel-Kohlenstoff- und Kobalt-Kohlenstoff-Legierungen Teil 2.
W.Patterson, S.Engler, R.Moser,
Giessereiforschung, Vol.21 (1969), pp.51-63.
[2] Schnellbestimmung des Sauerstoffes im flussigen Stahl mit Hilfe der thermischen Analyse.
G.Mahn,B.Chandra,
Stahl und Eisen, Vol.89 (1969), pp.1314-15.
[3] Practical Application of Liquidus Control for Malleable Iron Melting.
F.W.Jacobs,
AFS Transactions, Vol. (1981), pp.261-76.
[4] Relationship of Casting Defects to Solidification of Malleable Base White cast Iron.
C.Leon, U.Ekpoom, R.W.Heine,
AFS Transactions Vol. (1981), pp.323-44.
[5]
The Carbon Equivalent Fe-C-Si diagram and its Application to cast irons.
R.W.Heine,
AFS Cast Metals Research journal 1971. pp.49-54.
[6] Metallurgical Processing Variables Affecting the Solidification of Malleable Base White cast Iron,
AFS Transactions, Vol. (1981), pp.1-14.
[7] Liquidus and Eutectic Temperatures and Solidification of White Cast Irons.
R.W.Heine,
AFS Transactions Vol. (1977), pp.537-44.
[8]Zum Einfluss der Rohstoffe,der Schmelzfuhrung und der Schmelzbehandlung auf das Erstarrungsverhalten von Eisen-Kohlenstoff-Legierungen, besonders Temperguss.
W.patterson, R.Dopp,
Giesserei technisch wissensch. Beihefte, Vol.16 (1964), pp.49-86.
[9] Carbon Equivalent vs. Austenite Liquidus: What is the Correct Relationship for Cast irons?
A.Alagarsamy, F.W.jacobs,G.R.Strong,R.W.Heine,
AFS Transactions, Vol. (1984), pp.871-80.
[10] Working-group Thermal Analysis.Dutch foundry association NVvGT (1976).
[11] Inoculation effect on graphite formation in pure Fe-C and Fe-C-Si,
B.Lux, H.Tannenberger,
Intern. Foundry Congress 1962.
[12] Effects of inoculation and oxidation on the shape of austenite dendrite in cast iron.
N.Kayama,K.Murai,K.Suzuki,Y.Kataoka,
Report no.33 (1982) of the Casting research Laboratory, Waseda University.
[13] Application of Austenite Dendrite Growth Model to Analyze Liquidus Temperature Measurements in Cups.
F.Mampaey,
AFS Transactions, Vol. (1998), pp. 461-67.
[14]Volumenanderung und Gefugebildung bei der Kristallisation von Gusseisen.
H.Muhlberger,
Archiv fur das Eisenhuttenwesen, Vol.33 (1962), pp.681-98.
[15]Gelenkte Erstarrung von Gusseisen,
R.Wlodawer,
pp. 453-54
[16]The microheterogeneous structure of molten cast iron.
V.A.Izmailov,A.A.Vertman, A.M.Samarin.
Russian Castings production 1971, pp.30-33.
[17] Influence of components on structure formation in cast iron between liquidus and solidus.
A.A.Dodonov,Ri Khosen,A.A.Zhukov,
Russian castings production 1974, pp.206-067.
[18]Graphite Spherulite Formation and Growth.
A.L.De Sy,
Foundry, november 1953, page 367.
[19] Die Eignung der Hochtemperaturmikroskopie zur Untersuchung von Schmelz-und Kristallisationsvorgangen bei eisen-Kohlenstoff-und Eisen-Kohlenstoff-X-Legierungen (X:Chrom;Vanadium)
Teil II:Versuchsergebnisse an eutektikumhaltigen Fe-C-X-Legierungen und deren Deutung.
H.Siepmann,K-E Honer,W-D Schneider,
Giessereiforschung,Vol.23 (1971), pp.151-58.
[20]Beitrage zur gerichteten Erstarrung von Eisen-Kohlenstoff-Silicium-Legierungen.
Teil2.Konzentrationsverschiebungen bei gerichteter Erstarrung unter-und ubereutektischer Legierungen.
K.Lohberg,A.Reihim,
Giessereiforschung, Vol.31 (1979), pp.107-113.
[21]Principles Involved in the Use of Cooling Curves in Ductile Iron Process Control.
M.D.Chaudhari, R.W.Heine, C.R.Loper,jr.,
AFS Cast Metals Research Journal, Vol. (1975), pp.520-60.
[22] Crystallization of irons treated with cerium, tellurium and calcium.
D.N.Khudokormov, O.S.Komarov,
Russian Castings production, pp.171-173.
[23]A Revision of the Fe-C-Si System,
K.R.Olen,R.W.Heine,
AFS Cast metals Research Journal,March 1968,pp28-43.
