Development of Ingot Casting Technology for ESR Electrode Casting
Zbyněk Hudzieczek, Material and Metallurgical research, Ltd, Pohraniční 693/31, Ostrava -Vítkovice, The Czech
Republic. zbynek.hudzeiczek@mmvyzkum.cz
Vladislav Kurka, Material and Metallurgical research, Ltd, Pohraniční 693/31, Ostrava -Vítkovice, The Czech
Republic. vladislav.kurka@mmvyzkum.cz
Jaroslav Pindor, Material and Metallurgical research, Ltd, Pohraniční 693/31, Ostrava -Vítkovice, The Czech
Republic. jaroslav.pindor@mmvyzkum.cz
Material and Metallurgical Research, Ltd is equipped by new commercial simulation software intended for a 3D
simulation of liquid metals flow during filling and the subsequent steel solidification with prediction of soundness,
nyiama criterion, microporosity, hot spots, segregation, etc. Correct parameter settings of simulation (material
properties, heat transfer coefficients, pouring rates, initial temperatures) are necessary for obtaining relevant results
from numerical simulation.
This work is focused on improvement of the special ingot casting (electrode for electro-slag remelting) in Magma
software. Study is focused on influences of casting speed and on the inner integrity of the ingot – shrinkage size, porosity
and segregation. The electrode is special ingot with high h/d ratio (h/d= 9.8).
1. Introduction
MAGMA5 software was obtained by Material and Metallurgical research, Ltd. (MMR) for simulation of casting and
solidification of castings of steel, cast iron and non-ferrous metals. This software can simulate behaviour of metal in full
3D environment.
The objective of this article is to publish the results, which are focused particularly on the relation of the speed of casting
and internal discontinuities of material, possibly segregation when casting the "ELECTRODE" ingot. This ingot is
further remelted at the equipment for electro-slag remelting.
From metallurgical point of view, the ELECTRODE ingot is one of the most internally discontinuous ingots. Its internal
discontinuities are given particularly by the fact that it is an ingot without conical features, the H/D ratio is highly
slender (ca 9.8) and the head of ingot is not insulated.
P92 steel was selected as the casting material for numerical examination.
As the above-mentioned parameters (taper, slenderness and insulation use) may not be changed due to subsequent use on
electro-slag remelting, one of the few possibilities of changing the quality of the final casting is the change of parameters
of the actual casting process. The results of this numerical examination will result in the reduction of internal
discontinuities and achieving minimum segregation along the height of the ingot.
2. Experimental procedure
2.1 Definition of geometry and computing network
Using MAGMA5 environment [1], basic parameters were defined for the casting equipment according to the actual
conditions and the materials used in the pilot plant MMR ELECTRODE mould consists of 2.5 X moulds on top of each
other. This was disregarded for SW needs and 2.5 small X moulds were supplanted with one high cylindrical mould with
the same casing thickness as X moulds. The weight of mould material was maintained for the needs of dissipation of the
same heat. Figure 1 shows the full casting assembly.
89
Figure 1: Electrode casting assembly
The top red part of the ingot is the ingot head. Its length corresponds to the half of the height of the X mould.
In order to shorten the computation time, a half symmetry was selected in the symmetry plane of the whole plant. The
mesh consisting of hexahedrons had 54,169 "cavity cells" (sprues, ingot body, ingot head) and 717,025 overall cells
including all materials. Figure 2 shows computation mesh including:
 Sprue pin, stool and mould - in blue,
 Heat-resisting ceramics – green,
 Sprue and sprue pin – brown,
 Ingot body – grey and
 Ingot head – red.
Figure 2: Computational mesh
Figure 2 confirms that the casting plant is not provided with head insulation, which would be depicted in orange colour.
At places where intensive flow of liquid metal occurred and at narrow places (sprues) the mesh had to be refined.
90
2.2 Definition of computation conditions
In order to be able to compare the results, the initial and marginal conditions were identical for all model variants
including the initial temperature (casting temperature), which was 1577 °C. In all, six casting variants were simulated.
Description of variables at individual variants is provided in table 1.
Variant
5+1
5+3
5+5
7+1
7+3
9+1
Table 1: Casting parameters for individual variants
Ingot body Ingot body Ingot head Ingot head
casting time casting
casting time casting
(min)
speed
(min)
speed
(kg.min-1)
(kg.min-1)
5
315
1
336
5
315
3
112
5
315
5
67
7
225
1
336
7
225
3
112
9
175
1
336
Overall
casting time
(min)
6
8
10
8
10
10
To describe the appropriate temperature distribution on the mould, most authors use for individual interface the different
coefficients of heat transfer [2].
3. Results and Discussion
The objective of individual variants is to ascertain the impact of intensity of metal casting into ingot body and intensity
of head filling for the particular conditions of this slender ingot. The assumption is that the slow filling of the ingot head
will compensate for the changes of the melt volume due to its solidification, and in this way shrinkage cavity size will be
reduced.
The main objective of the variant with fast filling of the ingot head (1 min.) is to verify whether the melt has sufficient
flow dynamics in intensive casting so that the hot part of the melt is cast directly into the head and thus improves filling
properties.
Even during simulation calculation, it is possible to view continuous results. Figure 3 illustrates a temperature field of
metal after 200 seconds from the beginning of the casting process for the 5+1 variant.
Figure 3: Temperature field during ingot casting
91
Technological use of the ELECTRODE ingot follows after remelting on an electro-slag remelting plant. Therefore, the
shrinkage porosity in the ingot body does not have such a negative effect on the final use. Much more important is the
size of the shrinkage cavity in the middle of the ingot, as it affects the possibility of ingot use. The longer the cavity the
lesser the possibility of ingot being used, as the part containing the cavity needs to be removed.
Table 2 sums up and compares the shares of liquid (red), solid (blue) and semi-liquid (green) part at the end of casting,
share of the already solidified material and the overall solidification time of the ingot.
Table 2: Summary of results – solidification character
5+3
5+5
7+1
7+3
Variant
picture immediately
after casting
5+1
9+1
percentage
of
solidified metal at
the end of casting
(%)
time
of
full
solidification (s)
14.21
31.22
47.48
21.04
37.47
28.21
1626
1713
1729
1633
1918
1878
The results presented in table 2 clearly show the impact of head casting time to the share of solidified part but also
distribution of molten material. When comparing the variants when the ingot body is cast for 5 minutes, the filling of the
head has a significant role. In the case of the 5+1 variant, the molten metal (red) is distributed along the whole length of
the ingot and the solidified part makes only 14 % at the end of the casting, which results not only in creation of a deep
cavity but also porosity in the ingot body. In the case of the 5+3 variant, the molten metal is distributed very
unfortunately with regard to the fact that a spout is created in the middle and the liquid part in the bottom half of the
ingot tends to close and thus create conditions for more intensive porosity. Conversely, in the case of 5+5 variant the
portion of the liquid part is clearly visible particularly in the ingot head. Moreover, no neck is created in the middle of
the ingot not preventing the filling effects of the head which is the main prerequisite for minimization of creation of
internal discontinuities - porosity.
In the case when we compare the variants when the overall time of casting reaches 10 minutes (variants 5+5, 7+3 and
9+1), the 5+5 variant always turned out as best both in terms of solidification time and the solidification character. In
terms of solidification character the 7+3 variant was the worst, when a neck is created which will result in creation of
larger porosities under this neck.
The comparison of the share of the already solidified part for the variants with the overall time of casting of 10 minutes
is also worth mentioning. In the case of the 5+5 variant, the share of the solidified part is almost 50 %, whereas in the
case of 9+1 variant it is not even one third.
Table 3 compares discontinuities of ingots for the individual studied variants. The pictures include both shrinkage
cavities (including their depth) and micro-porosity.
92
Variant
share and
distribution of
porosity
shrinkage cavity
depth (mm)
Table 3: Comparison of internal discontinuities for the studied variants
5+1
5+3
5+5
7+1
7+3
651
672
431
441
294
9+1
357
The calculation of the size of the central shrinkage, or the central porosity is also influenced by the calculation of the
formation of the air gap between the mould wall and ingot body during the solidification and cooling due to the
shrinkage [3, 4].
Ingot discontinuities need to be viewed as a complex problem particularly in the case of slender ingots. Pictures in table
2 show that small shrinkage cavities tend to occur particularly for 5+5 and 7+3 variants, as these are the variants with
the largest portion of solidified or semi-liquid parts at the end of casting. Therefore, it could be assumed that the
volumetric changes during ingot solidification would not be that prominent as in the case of the other variants with a
smaller portion of solid and semi-liquid part at the end of casting.
This assumption was confirmed only partially because the extent of internal discontinuities in the case of 5+5 variants is
not the smallest.
The 7+3 variant showed the smallest depth of shrinkage cavity. However, this occurs at the expense of creation of a neck
which prevented filling by ingot head. It is particularly clear in the picture provided in table 2. The negative impact of
prevention of casting into ingot body is shown especially by the increased extent of internal discontinuities in the bottom
half of the ingot body.
Variants 5+1 and 5+3 were evaluated as the worst ones in terms of internal discontinuities, when the depth of shrinkage
cavities reaches up to 1.7 times the depth of the most shallow shrinkage cavity.
In the case of selection of the best variant, a compromise would have to be found between the size of the shrinkage
cavity and the extent of micro-porosity. However, all the best variants have the same denominator in terms of internal
discontinuities, namely the overall time of casting of 10 minutes.
Within the studied variants, the scope of segregation was also examined. The results are summarised in table 4.
93
Variant
Carbon segregation
(0.04-0.12 %)
5+1
Table 4: Carbon segregation
5+3
5+5
7+1
7+3
9+1
The pictures contained in table 4 show the extent of segregation in ingot section. In order that the pictures were
comparable, the range of segregation was unified for the extent of 0.04 hm. % to 0.12 hm. %. The target content of
carbon for this steel grade is determined to be 0.08 hm. %. If the segregation nears the lower values, the colour of this
area is blue, if it nears higher values, it turns orange and then yellow. The mean (and therefore target) value is shown in
red.
All variants share the following attributes. Increased carbon content in ingot axis and decreased carbon content in ingot
bottom and head. In terms of ingot homogeneity, the 5+5 variant seems to be the best one. The main credit for this
homogeneous section should be given particularly to the long time of ingot head casting.
4. Conclusions
In this work were simulated casting and solidification of P92 steel in ELECTRODE ingot (intended for electro-slag
remelting) casting plant. 6 different casting variants were examined. In terms of internal discontinuities, the selection of
the best variant is a matter of compromise. However, in general the best results were achieved by the group that had the
longest casting time of 10 minutes, i.e. the 5+5, 7+3 and 9+1 variants.
In terms of the extent of segregation and the effort to minimize them, the 5+5 variant proved to be the best, which had
the smallest segregation distribution in its section.
Overall, it may be stated that of the examined variants the best variant of casting of P92 steel into the electrode ingot is
the variant with 5 minutes of casting ingot body and 5 minutes for ingot head.
Acknowledgement
Complete evaluation of suitability of used methodology was realised as a part of the project No. CZ.1.05/2.1.00/01.0040 "Regional Materials
Science and Technology Centre" within the frame of the operation programme "Research and Development for Innovations" financed by the
Structural Funds and from the state budget of the Czech Republic
Literature
[1] MAGMA5, Release Notes, Installation Guide & Stamdard Manuals. 2012, MAGMA GmbH.
[2] BOGDAN, O. Numerical Analysis of Casting Technology and A-segregation Prediction in AISI 4340 Forgings
Products. Industrial Soft, Montreal, Canada, 2010.
[3] WEITAO LIU: Finite Element Modelling of Macrosegregation and Thermomechanical Phenomena in Solidification
Processes. These, Ecole Des Mines De Paris, 2005, 194 pages.
[4] ZJUBAN, N. A. aj.: Issledovanije defektov litoj struktury osevoj zony slitkov stali 38XHNZ3MFA i ich povedenije vo
vremja osadki. STAL, C.4, S. 24-27, 2011
94