You are on page 1of 6

Journal of Materials Processing Technology 96 (1999) 42±47

A note on the control of the solidi®cation front in


the continuous casting of copper tubes
J.M. Rodriguez*, A. Esteva, S. Meza
Metallurgical and Mechanical Department of Engineering, Ponti®cia Universidad Catolica de Chile,
Vicuna Mackenna 4860, Santiago, Chile

Received 17 February 1998; received in revised form 6 October 1998

Abstract

This research studies the position of the solidi®cation front in a continuous casting process of copper tubes to prevent the main failures
derived from a lack of control on this front. Speci®cally, the point is to prevent the leaking of liquid copper in the exit zone of the mould, to
prevent mandrel failures or casting detention due to its thermal contraction, and ®nally, to prevent failures in the granular structure of the
tubes. # 1999 Elsevier Science S.A. All rights reserved.

Keywords: Continuous casting of copper tubes

1. Introduction It is advisable to use a crucible that is as tight as possible


due to the cast metal bath being very susceptible to pene-
In this research, the position of the solidi®cation front in a tration of non-desirable elements such as hydrogen and
continuous casting process of copper tubes is studied to oxygen. Furthermore, at the exit of the tubes there is a
prevent the main failures derived from a lack of control on nitrogen seal that prevents the penetration of humidity and
this front. Speci®cally, the point is to prevent the leaking of environmental oxygen, improving the mechanical properties
liquid copper in the exit zone of the mould, to prevent of the cast tubes [1,2].
mandrel failures or casting detention due to its thermal The surface of separation between the liquid and the solid
contraction, and ®nally, to prevent failures in the granular is the solidi®cation front. This front is in¯uenced by numer-
structure of the tubes. ous factors, such as the extraction speed, the cooling mould
Continuous casting is a process in which there is a feed of design and the refrigerant ¯ow. These act simultaneously
liquid metal coming from a melting furnace to a mould during all of the process, in terms of a heat-transfer coef®-
which, supplied with a heat extraction system, produces the cient. It is for this reason that it is important to look for
metal solidi®cation. This metal, when extracted in a con- measurements that allow the detection, as far as is possible,
tinuous form, helps to establish a process of melting and of this front, apart from obtaining ranges in the variables in
continuous solidi®cation. This process involves the melting an experimental way that in¯uence the front, so as to be able
operations, solidi®cation, and extraction of the metal simul- to obtain a certain degree of control of the solidi®cation
taneously. front.
At the present time, the most common continuous casting Due to the requirement of high volume and good quality
methods are horizontal continuous casting and vertical production, an excellent global factor of heat transfer is
continuous casting. The selection of one or the other type needed for the purpose of increasing the casting speed and
is based on the production requirements, and the material, obtaining homogeneous solidi®cation. It has been deter-
the form, and the dimensions of the product. mined that the greatest resistance to heat transference is
that of the gas and air gaps [2±5].
The solidi®cation system is formed by two parts, an
*
Corresponding author. Tel.: 56-2-686-4627; fax: 56-2-686-5828 exterior part called the mould and an interior part called
E-mail address: jrodrigu@ing.puc.cl (J.M. Rodriguez) the mandrel. They must have certain characteristics both in

0924-0136/99/$ ± see front matter # 1999 Elsevier Science S.A. All rights reserved.
PII: S 0 9 2 4 - 0 1 3 6 ( 9 9 ) 0 0 3 1 4 - 3
J.M. Rodriguez et al. / Journal of Materials Processing Technology 96 (1999) 42±47 43

form and in material to be able to be used in the continuous di®cation front has to change. Thus it is very interesting to
casting process. know how this front varies, and in what amount, due to
The main characteristics that the mould and mandrel must changes in the variables previously mentioned.
have are the following: high thermal conductivity, chemical This is the reason why the casting tube cross-section had
inertness with respect to the material that is meant to be to be measured every time a variable was changed. The
melted, good machinability, long life, wear and tear resis- mandrel dimensions must also be known in order to establish
tance, a high melting temperature, thermal shock resistance, in which part of it the copper solidi®ed.
thermal stability, low heat capacity; they must keep their
mechanical properties at high temperatures; they must be
self-lubricating, or have a low friction coef®cient; and must 3. Experimental results
have high resistance to burn-out and tension.
According to all of the previous characteristics, the In Table 1 the different internal tube diameters obtained
material used generally to manufacture the mould and the through experimental measurements are shown, as well as
mandrel is graphite. the different locations over the mandrel of the solidi®cation
The mould and mandrel design is of great importance due front; this last information was obtained simply from a
to the removal of the tube to be cast needing to be made geometrical calculation starting from the previous measure-
easier. To obtain this, the mandrel must have a converging ments, and from the mandrel conicity. Fig. 2 shows the
conicity due to tube contraction. The mould conicity must be mandrel and over it the area where the solidi®cation fronts
diverging in order to minimize the gap between the tube and are formed when the operation conditions are varied (mainly
the mould, and also to facilitate tube extraction. the casting speed and the refrigerant ¯ow).
Previously it was found that the removal stage was Within the interval of the front position, the casting speed
simpli®ed enormously if a conicity of 0.66218 was used, (cm/min) was increasing with distance from the front origin,
converging for the graphite mandrel, and 0.22718 diverging giving as a result approximately an increasing straight line
for the graphite mould [4]. with a gradient of 0.61 (shown in Fig. 3).
A cooler refrigerated by water is used to remove the Fig. 4 shows some of the different pieces of tubes where
solidi®cation heat from the copper tubes. The cooler is a heat the measured internal diameters were a little different
exchanger in reverse ¯ow, the function of which is to remove (included in Table 1), whilst Fig. 5 shows the moment when
the heat coming from the mould through a liquid refrigerant. such copper tubes were obtained through the continuous
The copper cooler is made of a mechanized stage with the casting process.
form of a coil in its interior, to increase the turbulence of the
refrigerant (water), and in this way to increase its ef®ciency 4. Analysis of results
[6].
As a general tendency it was observed that when the
2. Experimental procedure casting speed was increased, the solidi®cation front was

In Fig. 1, which corresponds to the design of a solidi®ca- Table 1


tion system tested successfully in a research work previously Position of the solidification front
published in the Giessereiforschung magazine [2], the geo- Average inner diameter (mm) Front location (mm)
metry of each component of the system is illustrated, as well
as the type of material used. 18.10 81.73
18.06 83.01
The objective of the quality tests that are applied to the 18.03 84.28
copper tubes is to determine if the tube that is produced is of 17.99 85.56
good condition. Its ductility, structure and chemical com- 17.95 86.84
position is studied against changes of variables within a 17.93 87.69
17.91 88.11
casting; and the characteristics amongst castings are also
17.85 90.24
compared. This is accomplished by means of the following 17.76 93.22
tests: (i) the tensile test; (ii) the hydrogen test; (iii) the 17.75 93.65
sulphur test; and (iv) micrography and macrography. 17.74 94.07
One of the objectives of this research is to control the 17.73 94.50
17.70 95.35
solidi®cation front. It is very important to have information
17.65 97.05
about the location of the front in the mould, and especially in 17.64 97.48
the mandrel. For this reason the dimensions must be taken 17.61 98.33
into consideration; both the mandrel dimensions as well as 17.60 98.76
the copper tube cross-sectional dimensions. 17.58 99.61
17.55 100.46
When varying the removal speed or the ¯ow of the
17.45 103.86
refrigerant in the continuous casting experiment, the soli-
44 J.M. Rodriguez et al. / Journal of Materials Processing Technology 96 (1999) 42±47

Fig. 1. The conical mould and conical mandrel for the production of tubes (dimensions: mm).

ing the inner tube diameters, because each one of


them corresponds to the diameter that the mandrel has in
the place where the inner tube wall is formed by solidi®ca-
tion.
Despite the differences in the inner diameter, as shown in
Table 1, being small, they are measurable and they serve to
detect the advances and retreats of the solidi®cation front
Fig. 2. Solidification zone of the casting tubes. during the casting process. Excessive advances can produce
spilling of liquid copper or can cause the obtaining of a bar
formed further from the mandrel origin, i.e. there was instead of a tube. On the other hand excessive retreats
an advance of the front. This change in the solidi®ca- produce casting stops due to obstructions produced by
tion front position was detectable when rigorously measur- premature solidi®cation [1].
J.M. Rodriguez et al. / Journal of Materials Processing Technology 96 (1999) 42±47 45

Fig. 3. Smoothed graphical representation of Table 1.

Fig. 4. Different pieces of copper tubes obtained by means of continuous casting.

In this way the known conicity of the graphite mandrel The properties of graphite seem to be adequate to create
serves as a means of control during the continuous casting the mandrel in this process of the continuous casting of
process of copper tubes, because the inner diameter of the copper tubes, according to the experience acquired [1±
tubes indicates the exact position of the solidi®cation front. 3,5,7,8].
On the other hand, even when there is certain wear of the Fig. 4 presents three different samples of copper tubes that
graphite mandrel, its conicity is maintained suf®ciently well solidi®ed in different mandrel places, but relatively close to
to operate in a controllable productive process. each other. As might be expected, the granular structure
46 J.M. Rodriguez et al. / Journal of Materials Processing Technology 96 (1999) 42±47

Fig. 5. The moment in which the samples of Fig. 3 were obtained by means of continuous casting.

revealed by a cross-sectional etch is quite similar in the three casting stops due to obstructions produced by premature
cases. solidi®cation.
On the other hand, even when there is certain wear of the
graphite of the mandrel, its conicity is retained suf®ciently
5. Conclusions well for the mandrel to be used in the continuous casting
process of copper tubes. Furthermore, graphite as a material
The conicity of the mandrel, apart from being necessary has other very useful properties for this purpose.
for the continuous casting of copper tubes because the inner
tube walls are formed on top of it, becomes very useful in
order to control the position of the solidi®cation front Acknowledgements
through rigorous measurements of the inner diameter of
the casted tubes. When the solidi®cation front is excessively The authors would like to thank the Fondo Nacional de
advanced, spills of liquid copper may occur, or the product Desarrollo Cienti®co Y Tecnnologico (Fondecyt) through
can be a bar instead of a tube. On the other hand the the project 1960364, for making this technological research
excessive retreat of the solidi®cation front can produce work possible.
J.M. Rodriguez et al. / Journal of Materials Processing Technology 96 (1999) 42±47 47

References [4] J.M. RodrõÂguez, Influence of the conicity of the mould, Influence of
the conicity of the mould and the core on continuous casting of
copper pipes, Wire 43(2) (1993) 113±115.
[1] J.M. RodrõÂguez, Continuous casting of copper tubes, Giesserei- [5] J.M. RodrõÂguez, R. Maturana, J von Storch, Continuous casting of
forschung 46(4) (1994) 122±124. copper tubes in a reducing furnace under controlled atmosphere wire-
[2] J.M. RodrõÂguez, J.P. Montecinos, Continuous casting of copper tubes drawability of the tubes, Wire 39(2) (1989) 125±128.
using an inert gas seal at the exit zone, Giessereiforschung 49(2) [6] J.M. RodrõÂguez, J.P. Montecinos, A study of different cooler designs in a
(1997) 68±70. continuous copper bar casting process, Wire 40(5) (1990) 533±537.
[3] J.M. RodrõÂguez, J.P. Montecinos, Continuous automated [7] R. Thomson, Continuous casting from graphite moulds, AFS Trans
horizontal casting of copper in an induction furnace with 79 (1971) 161±168.
selected cooling and mould designs, Wire 41(5) (1991) 409± [8] J.G. Campbell, Graphite dies for non-ferrous continuous casting,
414. Light Metals and Metal Ind. 28(329) (1965) 42±47.

You might also like