2018 Volume 59 Issue 9 Pages 1477-1482
In this study, we reproduced the hyper-eutectic cast iron made at the end of the Edo period, and considered its new possibility.
It was found that foundry men of that period made non-chill hyper-eutectic cast iron even at very low silicon content. As for ϕ50 × 250 mm T.P, the tensile strength of C4.53% flake graphite cast iron is 107 MPa which is very low for cast iron. It is thought that the produced thick cannon most probably burst in actual combat with this material.
The melt’s flow ability is the best in the eutectic composition but deteriorates suddenly when entering the hyper-eutectic composition, due to the formation of Kish graphite. Plate type carbide forms when silicon content is low and with quick cooling, in the hyper-eutectic composition. It is thought to form, due to the very slow C diffusion from the matrix to the free graphite, during the eutectoid transformation. Test pieces having plate type carbides shows excellent damping capacity.
This Paper was Originally Published in Japanese in J. JFS 89 (2017) 695–700.
It was found that foundry men of Edo period made non-chill cast iron cannon even at very low silicon content and machined it. Figure 1 shows the cast iron cannon in Anori shrine and its micro structure, which was made by foundry man Y. Masuda.1) The composition of this cast iron is hyper eutectic composition (C4.48 mass% (hereafter %), Si0.13%, CE about 4.5%) and chill (Ledeburite) is not shown even at very low silicon as 0.13%. It is thought that the foundry men at that period knew this technology, even though foundry men nowadays do not know it. In the Fig. 1, enlarged graphite is shown which forms directly from the molten metal at first and is called Kish graphite. This enlarged graphite shows high damping capacity, compared to that of common gray cast iron.2)
Cast iron cannon of Anori shrine and its microstructure. (C4.48%–Si0.13%–Mn < 0.001%–P0.117%–S0.034%–Ti0.005%)
In this study, we reproduced the low silicon hyper eutectic cast iron which was made at the end of the Edo period, and considered its new possibility.
With high frequency induction furnace (Electric power 50 kW, 3000 Hz, melting capacity 50 kg), 3.9–5.0% of carbon molten metal was prepared. (Basic composition: Si0.13%, Mn0.05%, P0.114%, S0.037%, Ti0.003%) As shown in Fig. 2, carbon amount was adjusted with covering the molten metal’s surface with carburizer (carbon layer) and holding at different temperature. For example, as shown in Fe–C diagram of Fig. 3, if the molten metal is held at 1250°C, its carbon content becomes 4.5% naturally, following the solidus line of graphite. If the molten metal is held at 1490°C, its carbon content becomes 5.1%.
Carbon layer (carburizer) on the molten metal’s surface.
Relationship between C amount and liquidus line.
Table 1 shows the relationship between the holding temperature and carbon amount. Table 2 shows the change of silicon amount in silicon addition test. Like as the above method, gray cast iron samples of 0.01%–1.40% silicon were cast. (Basic composition: C4.5%, Mn0.05%, P0.114%, S0.037%, Ti0.003%) The shape of the gray cast iron sample was ϕ30 mm × 250 mm or ϕ50 × 250 mm. The micro structure and mechanical properties were investigated.
The damping capacity was measured with the following method. Firstly, ϕ25 × 250 mm test piece mas machine from the holding part of tensile test piece. Secondly, the sound speed (V, m/s) transmitting the ϕ25 × 250 mm test piece was measured with ultra sound making instrument. Thirdly, we calculated the Young Ratio (E, GPa) with eq. (1). In this equation, we set the density (ρ) as 7200 (kg/m3), parameter (α) as 1.116. Fourth, following the equation of Abe,3) we calculated the damping capacity (Q−1) from the Young Ratio (E, GPa).
| \begin{equation} V = 1.116 \times \sqrt{10^{9}\times E/\rho} \end{equation} | (1) |
| \begin{equation} Q^{-1} = 0.1404 \times \exp (-0.03163\,\text{E}) \end{equation} | (2) |
Spiral sand mold for melting flow test.
Figure 5 shows the microstructure of ϕ30 × 250 mm test piece (hereafter T.P.). In hypo eutectic composition T.P. (a) (C4.14%), only chill (Ledeburite) is shown and graphite is not shown. In eutectic composition T.P. (b) (C4.31%), graphite begins to crystallize. In the connection, I may add that the old kettle was poured with eutectic composition (about C4.3%), in order to prevent crack formation during cooling of casting, through the graphite expansion.4)
Microstructure of ϕ30 × 250 mm T.P. (Si0.13%–Mn0.05%–P0.114%–S0.037%)
In hyper eutectic composition T.P. (c) (C4.53%), chill is not shown and only flake graphite is shown. In the hyper eutectic composition, primary graphite becomes the nucleus of eutectic graphite and stabilizes the eutectic graphite, so chill does not form.5) The graphite of hyper eutectic composition shows star type which is peculiar to hyper eutectic cast iron.
As for the hypo eutectic composition T.P. (a) (C4.14%) or eutectic composition T.P. (b) (C4.31%), as it was impossible to machine it, we could not do tensile test.
As for the hyper eutectic composition T.P. (c) (C4.53%), as it was possible to machine it, we could do tensile test. However, the tensile strength value is only 117 MPa.
Figure 6 shows the micro structure of 50 × 250 mm T.P. Table 3 shows their mechanical properties. Kish graphite get thickened and lengthened with increasing carbon amount. Tensile strength and hardness decrease with increasing carbon amount. As for the T.P. (b) (C4.53%) whose composition is similar to that of Anori shrine cannon, its tensile strength is only 107 MPa. Even though it was used for test firing, it is thought, it could not be used in actual combat as cannon. However, as Kish graphite is formed in the matrix, high damping capacity is expected.
Microstructure of ϕ50 × 250 mm T.P. (Si0.13%–Mn0.05%–P0.114%–S0.037%)
Figure 7 shows the SEM of low silicon hyper eutectic cast iron for ϕ30 × 250 mm T.P. As pointed with arrows, plate type micro structures are shown, which are different from graphite or matrix.
SEM of low Si hyper eutectic cast iron. (C4.53%, Si0.13%)
Figure 8 shows the difference of Vickers hardness between the plate and the matrix. The Vickers hardness of pearlite matrix is 199 HV. That of the plate is very hard as 628 HV. From the result, it can be considered that the plate is not ferrite but cementite.
Plate type microstructure and its matrix hardness. (C4.53%, Si0.13%)
Figure 9 shows the relationship between silicon amount and micro structure in hyper eutectic cast iron for ϕ30 × 250 mm T.P. On the condition under (b) Si0.41%, as pointed with arrows, plate type carbides form. However, on the condition over (c) 0.62% Si, the plate type carbides do not form. From these result, it can be said that plate type carbides do not form on the condition of over 0.5% Si. The reason will be explained later.
Relationship between Si amount and micro structure in hyper eutectic cast iron. (ϕ30 × 250 mm T.P.)
Figure 10 shows the influence of carbon and silicon amount on the formation of plate type carbide for ϕ30 × 250 mm T.P. The plate type carbide is shown only in the low silicon hyper eutectic cast iron (a). It is not shown in the high silicon hyper eutectic cast iron (b) or in the common cast iron (c).
Influence of C and Si on the formation of plate type carbide. (ϕ30 × 250 mm T.P.)
Figure 11 shows the comparison of damping capacity between low silicon hyper eutectic cast iron (Fig. 5(c), C4.53%) and other materials. For reference, the value of damping capacity obtained from FC300 and FCD400 are also showed.
Comparison of damping capacity between low Si hyper eutectic cast iron and other materials.
The order of excellent damping capacity value is as follow;
| \begin{align} &\text{Low Si hyper eutectic cast iron} \\ &\quad> \text{hyper eutectic cast iron} > \text{FC300} > \text{FCD400} \end{align} | (3) |
As for the effect of hyper eutectic cast iron on damping capacity, Matsui explains as follows: In the hypo eutectic cast iron, each eutectic cells are independent each other. In the hyper eutectic cast iron, as Kish graphite connects eutectic cells, graphite shares stress easily and the effect of hyper eutectic graphite is stronger than that of hypo eutectic graphite, so the damping capacity becomes large.
Let’s confirm the result of damping capacity obtained in this study. In this study, the damping capacity of the ultra low silicon cast iron is higher than that of hyper eutectic cast iron. The difference between the two materials are the plat type carbide. Owing to the plat type carbide, the damping capacity for scattering the vibration maybe become higher.
3.3 Relationship between carbon amount and melt flow abilityTable 4 shows the relationship between carbon amount and melt flow length. The flow length in hypo eutectic composition is 62 cm and that in eutectic composition is 83 cm which is the longest. In the hyper eutectic composition, the flow ability becomes bad with increasing carbon amount of the molten metal. It is considered that the melt’s flow ability becomes bad proportional to the amount of Kish graphite.
As mentioned above, the plate type carbide forms in hyper eutectic composition. This carbide also forms in hyper eutectoid steel, as shown in Fig. 12(a).6) There are something in common between them as follows; (a) shape of carbide, (b) low silicon composition, (c) formation timing. However, there are something different. In the hyper eutectoid steel, the plate type carbide grew from grain boundary to inner grain, which is called Widmannstatten structure. In the hyper eutectic cast iron, the plate type carbide forms only in the grain. Comparing this two type, the mechanism of the plate type carbide formation is considered as follow.
Comparison of plate type carbide between hyper eutectoid steel and low silicon hyper eutectic cast iron.
Figure 13 shows the mechanism of plate type carbide formation in the hyper eutectoid steel and low silicon hyper eutectic cast iron. First, the mechanism of plate type carbide formation in the hyper eutectoid steel is as follow. Let’s see an example in C1.2% hyper eutectoid steel (Fig. 13(a)). During cooling between Acm (900°C) and eutectoid temperature (738°C), carbon saturation degree decreases from 1.2% to 0.7%, and the difference is 0.5%. On the condition of enough slow cooling, saturated carbon diffuses to grain boundary and net-work type carbide (Fe3C) forms. On the condition of quick cooling, as the carbon diffusion from grain matrix to grain boundary is slower than eutectoid transformation, plate type carbide (Fe3C) forms also in grain matrix.7) As the carbide grows easily from grain boundary to inner grain in the viewpoint of energy, plate type carbide grows from grain boundary to inner grain, and the Widmannstatten structure forms.
Mechanism of plate type carbide formation.
Secondly, the mechanism of plate type carbide formation in the low silicon hyper eutectic cast iron is as follow. During cooling between eutectic solidification ending and just before eutectoid transformation, the carbon saturation degree in the matrix decreases from 2.1% to 0.7%, and the difference between them is 1.4%. On the condition of slow cooling (III and IV), this 1.4% carbon diffuses to free graphite and becomes graphite. However, on the condition of quick cooling (III′ and IV′), the carbon diffusion from matrix to free graphite is slower than eutectoid transformation. As carbon composition is the highest in the matrix center between graphite and graphite, the plate carbide (Fe3C) forms between graphite and graphite. Unlike the hyper eutectoid steel plate carbide does not form on the grain boundary. It forms only in the inner grain.
Figure 14 shows the comparison of micro structure between D30 × L250 mm T.P and D50 × L250 mm T.P. Plate carbide forms only in the D30 × L250 mm T.P. As cooling ratio of D50 × L250 mm T.P is very slow enough for carbon diffusion, like the above case (III and IV), plate type carbide does not form.
Comparison of microstructure ϕ30 × 250 mm T.P. and ϕ50 × 250 mm T.P.
As explained in Fig. 9, plate carbide does not form, if silicon amount exceeds 0.5%. The reason is considered as follow. Figure 15 shows the influence of silicon amount on carbon solubility at eutectic temperature and eutectoid temperature. This was calculated from the relationship between silicon amount and section of phase diagram, which was proposed by H. Hanemann.8) For example, in the case of 0% silicon, carbon solubility in austenite matrix at eutectic temperature is about 2.1% and that at eutectoid temperature is 0.7%. As the result, about 1.4% carbon should diffuse from matrix to the free graphite. In the case of 0.5% Si, C solubility in austenite matrix at eutectic temperature is about 1.7%, and that at eutectoid temperature is about 0.6%. As the result, 1.1% carbon should diffuse from matrix to the free graphite.
Influence of Si amount on C solubility in austenite at eutectic temp. and at eutectoid temp.
That is, as the silicon amount increases to 0.5%, carbon diffusion amount decreases from 1.4% to 1.1%, so the excessive carbon amount decreases in the matrix. From this reason, in the case of 0.5% Si, there is little excessive carbon in the matrix. Therefore, even though in the ϕ30 × 250 mm T.P whose cooling ratio is quick comparatively and whose diffusion time is short, the plate type carbide does not form.
In this study, we reproduced the hyper-eutectic cast iron made at the end of the Edo period, and considered its new possibility. The results obtained are as follows;
