NIPPON STEEL & SUMITOMO METAL TECHNICAL REPORT No.
112 APRIL 2016
UDC 621 . 746 . 047 : 681 . 3
Technical Report
Numerical Simulation of the Continuous Casting Process and
the Optimization of the Mold and the Strand
Norimasa YAMASAKI* Shozo SHIMA
Keiji TSUNENARI Satoru HAYASHI
Masahiro DOKI Yuichiro KATO
Daisuke MIKI Takeo NAKANISHI
Abstract
The phenomenon of the uneven solidification in the continuous casting process was ex-
amined by the numerical simulations. The simulation model which analyze the distortion
behavior of the solidifying steel shell was developed. The influence of the mold shape to the
growth of the solidifying shell was studied. It was found that the multiple taper design is
useful to the even solidification. The numerical model that simulate the free surface flow of
the spray water in the secondary cooling of the strand was developed using a particle-based
method. The uneven water flow in the width direction was examined, and the over-cooling
around the center in width direction was found.
1. Introduction the solidifying shell grows and thermally shrinks. In a slab rectan-
Continuous casting in the steel making process is a process that gular in shape, the shell on the wide face mold (in the width direc-
continuously solidifies molten steel in a liquid state to a solid state tion of a slab) shrinks more than the one on the narrow face mold (in
called slab. Molten steel is poured into a mold built up of cooled the thickness direction of a slab). Therefore, in the conventional slab
cupper plates on its four sides through refractory-made submerged continuous casting machine, mold cupper plates of narrow face
nozzles and solidification starts at the periphery. Solidified periph- mold determining the width are tapered and designed to reduce the
ery, called as the solidifying shell, is very thin as compared to the slab width toward the lower end of the mold.
width and the thickness of the slab. Under this state, though periph- However, it is difficult to match the taper to the shrinkage of the
ery of the four faces is thinly solidified, molten steel in liquid state solidifying shell completely and a gap tends to be developed be-
exists in the inner part. Length of the mold in the casting direction is tween the shell and the mold in the neighborhood of the corner of a
about one meter and a plurality of roll is installed right below the slab, wherein the narrow face mold plate edge contacts the wide
mold to protect the thin solidifying shell. In order to solidify com- face mold plate. Thermal resistance between the solidifying shell
pletely, spray nozzles are installed in between the rolls for cooling and the mold grows, decreasing heat transfer. This means that as
and cooling by water or mist of mixture of water and air is applied. heat transfer in the neighborhood of the slab corner decreases, un-
The mold is termed as the primary cooling zone and the spray cool- even solidification tends to be developed more easily. An inappro-
ing zone right below the mold is termed as the secondary cooling priate taper in particular promotes solidification unevenness and
zone. triggers breakage in the solidifying shell or a break-out in worst cas-
In the mold of primary cooling zone, the solidifying shell is thin es. In order to make the heat transfer uniform in the mold, mold ta-
and in case a shell thinner than the one in other parts of periphery is per design is crucial; however, observation of the state of solidifying
formed due to uneven solidification, the solidifying shell becomes shell within a mold is practically impossible therefore clarification
unbearable to the static pressure of the molten steel and the shell of the phenomenon by means of numerical analysis is effective.
may break. Breakage of the shell causes flow-out of molten steel, Then, a numerical analysis model that analyses the distortion behav-
triggering a break-out in worst case, which means leaking out of ior of a solidifying shell within a mold was developed and the effect
molten steel from right below the mold. As temperature decreases, of the mold shape over the development of uneven solidifying shell
* General Manager, Maintenance Planning Dept., Plant Engineering and Facility Management Center
2-6-1 Marunouchi, Chiyoda-ku, Tokyo 100-8071
- 64 -
NIPPON STEEL & SUMITOMO METAL TECHNICAL REPORT No. 112 APRIL 2016
was studied.
In the secondary cooling, it is important to ensure the following:
to solidify the cast slab completely before it reaches the last support
roll, to solidify uniformly in width direction, and to prevent the oc-
currence of transverse cracks at edge positions in width direction. In
order to obtain a uniform cooling in width direction, optimization of
spray nozzle arrangement, cooling area of a single nozzle, and water
flow rate distribution in width direction were conducted; however,
in recent years, because rolls separated in width direction are used
generally, uneven cooling in width direction has been caused more
by the water collected on rolls since the water flows out and drops at
an interspace in the width direction wherein the bearings of the sep-
arated roll are installed. Fig. 1 FEM 4-nodes thick shell element
However, as observation of the water flow is difficult, analysis
of spray water flow by means of flow analysis is effective. In addi-
.
tion, attention was paid to particle method as a means of flow analy- mal strain rate, and ε m is transformation strain rate.
sis to realize visualization of free moving surfaces flow of spray wa- As for viscoplastic strain, a model proposed by Wang 2) et al. was
ter and visualization of the phenomenon was attempted. Further- used. Relational expression between stress and strain rate obtained
more, if visualization of the flow is realized, prediction of heat from the result of high temperature tensile test result 5) was used to
transfer coefficient in spray cooling becomes possible. By an experi- determine physical properties for calculation.
ment simulating the flow obtained by analysis with particle method, Next, formularization with the finite element method is stated.
heat transfer coefficient was measured and by inputting the data to Since shell thickness is small as compared to cast slab size (width,
boundary conditions in the solidification calculation, the mechanism thickness, and casting length in longitudinal direction) and the tem-
of development of uneven solidification in width direction was stud- perature distribution within the solidifying shell can be expressed
ied. relatively in a simple manner (approximated by a quadric), while
analyzing the distortion with the finite element method (FEM), for-
2. Optimization of Mold Shape by Solidifying shell mularization using 4-nodes thick shell elements as calculating ele-
Distortion Analysis ment was executed. Figure 1 shows the shell elements used in the
2.1 Analysis of solidifying shell distortion within mold FEM.
In almost all solidification distortion analyses, a non-steady Regions of temperature distribution and plastic distortion were
method of fixing the coordinate system to an intercept of a cast defined by several integration points of elements in thickness direc-
steel 1) is used. This method calculates temperature and stress by tion. With this FEM calculation using the elements, 3D analysis tak-
changing the peripheral boundary conditions in accordance with ing into account material transfer, phase transformation including
elapse of time. However, the method cannot take into account the solidification, and viscoplastic behavior in high temperature regions
temperature gradient and stress gradient in casting direction. To also became possible. Procedure of the calculation is to allot tem-
solve the problem, a method of analysis in a velocity field on the co- perature distribution obtained with the aforementioned solidification
ordinate system fixed to a space was proposed, taking into account calculation to each integration point in the element (four points on a
the viscoplastic behavior and likewise the rigid plasticity analysis plain and a plural of n point in shell thickness direction) and to re-
used in rolling analysis.2, 3) However, the model is a two dimensional peat the calculation by changing loading rate until the displacement
model based on the assumption of generalized plain strain and was rate converges, while maintaining rigid matrix constant. When the
insufficient to represent the cracks that are practically developed in displacement rate converges and the distortion state is calculated,
actual slabs. It can be noted that since the solidifying shell thickness state of contact of mold with solidifying shell is known. Heat trans-
is very thin compared to cast slab size within a mold and that the fer becomes low where a gap is developed and heat flux becomes
temperature distribution in the shell thickness direction can be ex- smaller. Taking into account the smaller heat flux, solidification is
pressed in a relatively simple distribution form, a 3D finite element recalculated and the abovementioned distortion analysis is repeated.
analysis model was developed by formularization using shell ele- This procedure is repeated until the temperature distribution and the
ments. The developed model takes into account substance move- shell distortion behavior converge ultimately.
ment and phase transformation, including solidification, thermal 2.2 Effect of narrow face mold taper shape on development of
shrinking, and viscoplastic behavior. solidifying shell
As for the method of solidification calculation, there are various While studying the effect of the taper shape of the narrow face
methods such as enthalpy method and equivalent specific heat meth- mold on the development of the solidifying shell using the analysis
od; however, in the model, assuming that the temperature distribu- model, solidification uniformity ratio was newly proposed as an as-
tion within the shell can be approximated by a quadric, method of sessment index. Extent of delay in solidification at a corner can be
solving the heat conduction equation with Runge-Kutta 4) method assessed, judging from the shape of a white band observed on a sec-
was employed. Since the model handles viscoplastic behavior, for- tion perpendicular to the direction of casting, which is considered to
mularization of the model by means of strain rate was applied so be produced at the location wherein the discharge flow impinges on
that deformation rate can be taken into account. Overall strain rate is the narrow face mold and inverts. Solidification uniformity ratio is
expressed by the following formula. defined as the value of smallest shell thickness at the delayed solidi-
. . . . .
ε = ε e + ε vp + ε t + ε m (1) fication location in the neighborhood of a corner divided by the shell
.e . .
where ε is elastic strain rate, ε vp is viscoplastic strain rate, ε t is ther- thickness at normal locations. Using this index, effect of the narrow
- 65 -
NIPPON STEEL & SUMITOMO METAL TECHNICAL REPORT No. 112 APRIL 2016
Fig. 2 Comparison between single and multiple taper narrow face mold
face mold taper shape on the development of the solidifying shell
was studied. It is generally known that the narrow face mold taper
shape improves the state of contact of the cast slab with the mold,
equalizing solidification by making the mold closely match the ther-
mal shrinkage in width direction.
As shown in Fig. 2, a narrow face mold is used generally in an
inclined state to match shrinkage and a mold with a single taper of a
constant gradient from the top end to the bottom end (the narrow
face mold on the right side in Fig. 2) is conventionally used. How-
ever, since the amount of the thermal shrinkage is larger in the initial
stage of casting and becomes smaller in later stage, a multiple taper
(the narrow face mold on the left side in Fig. 2) with larger taper on
the upper stage has been proposed to provide better matching.6) As
shown in Fig. 2, with the multi-tapered narrow face mold, the gap
between the solidifying shell and the mold is considered to become
smaller.
Fig. 3 Calculated shell thickness and gap between shell and mold
In order to verify the effect of the multi-tapered narrow face
mold, casting under the following condition was analyzed using the
developed numerical analysis model: casting speed of 1.5 m/min,
cast slab width of 1 200 mm, and cast slab thickness of 250 mm. A
1/4 model of half of the width and the thickness, taking into account
symmetricity was used. Mesh size was 5 mm in width and thick-
ness, and in the shell thickness direction, thickness of 20 mm was
divided into 20 elements. Figure 3 (a) shows the solidifying shell
thickness (in the upper figure) and the gap between the mold and the
cast slab (in the lower figure) about the case of a single taper narrow
face mold with a constant gradient. In the representation of the gap,
the extend of contact force is represented by the color changing in
the contact area of the mold with the cast slab. Since the gap is de-
veloped in the neighborhood of a corner, delayed solidification is Fig. 4 Comparison between calculated and measured uniformity ratio
of the shell
occurs at the corner section of the wide face mold edge.
Figure 3 (b) shows the result of calculation in the case of the
multi-tapered mold having a larger taper in the upper section and a
smaller taper in the lower section. The gap at the corner section be-
comes smaller and shell uniformity ratio was improved. This is re-
lated to the phenomenon that thermal shrinkage rate along with the
progress of solidification is larger at the upper section of the mold
and smaller at the lower section of the mold. Since the gradient from
top to bottom is constant for a single-tapered mold, the taper is in-
sufficient in the initial stage of casting wherein shrinkage is large. Fig. 5 Measured shell thickness near the corner
Therefore, a gap is developed at a corner and solidification is de-
layed. In contrast, in case of a multiple taper mold, a sufficient taper both in calculation result and measured values by employing a mul-
provides shrinkage in the initial stage and the gap at the corner be- tiple taper mold. It has been proved that the accuracy is sufficient
comes smaller and the heat sink effect is improved. for taper designing. The narrow face mold shape of the mold was
Comparison of calculation and measured values of solidification optimized using the developed model and was applied to an actual
uniformity ratio is shown in Fig. 4. An example of a measured value machine.
of shell thickness is shown in Fig. 5. Uniformity ratio is improved
- 66 -
NIPPON STEEL & SUMITOMO METAL TECHNICAL REPORT No. 112 APRIL 2016
3. Numerical Simulation of Secondary Cooling thermometer installed at 18 m downstream from the meniscus under
Spray Water Flow by Particle Method the condition of constant casting speed of 1.0 m/min for the slab of
In secondary cooling, cooling is conducted by a plurality of noz- 300 mm × 2 200 mm in size. The secondary cooling zone of a con-
zle installed in between rolls in the width direction and uniform tinuous casting machine is generally composed of segments each of
cooling in the width direction by a single nozzle was directed.7) which consists of a plurality of roll, and the surface of a cast slab
However, there are no reports that have sufficiently analyzed in a can only be observed in between the segments. Since roll pitch is
quantitative manner the spray water flow of the water that falls small, measurement was conducted above the upstream location of
through the space opened for installing complicatedly laid-out split the unbending segment wherein a relatively large spacing is provid-
roll bearings and the water collected between a roll and the cast slab ed for the ease of extracting a segment for maintenance work. A
(collected water). Therefore, the cooling uniformity in the width di- scanning type monochromatic radiation thermometer (measurement
rection is not sufficient. In recent years, in order to support the cast wave length: 1.0 μm) was used.
slab of thin solidification thickness from being cast at a higher As shown in Fig. 6, it was found that the temperature at the cen-
speed, the pitch of rolls of a continuous casting machine has been ter of the width is lower by more than 100°C as compared to edge
made smaller. In accordance with this, another problem of larger roll section. The following two causes have been considered as the cause
deflection due to smaller roll diameter and deteriorated rigidity has of such unevenness in the width direction. The first is the effect of
emerged. Then, a long roll was separated to a plurality of short rolls reducing the water flow rate at the edge position in the secondary
in the width direction and the deflection of a roll has been lessened. cooling to prevent over cooling at the edge,11) and the second is the
However, installation of a plurality of bearings in the width direc- effect of the flow of the molten steel in the mold. However, the con-
tion became necessary and a phenomenon that casts slab cooling ventional way of thinking has failed to explain the reason of the ex-
spraying water drops unevenly along and over such bearings treme temperature drop in such an example. Therefore, to seek for a
emerged. cause of uneven cooling, effect of arrangement of split roll bearings
In order to analyze the effect of the flow on uneven solidifica- and flow of water that flows thereon was studied.
tion, it is important to grasp the flow of the spray water in a quanti- 3.2 Formularization of water flow model
tative manner. Although flow analysis is effective, the conventional In order to analyze spray water flow by calculation, the particle
method using a mesh cannot comply with the flow of complicated method (MPS) 12, 13) was applied and analysis was conducted using
free moving surfaces and application of a meshless particle method 8) the all-purpose particle method fluid analysis software “Particle-
was considered. Using the flow of spray water obtained by the anal- works” 14) with addition of a function. The particle method was em-
ysis, the heat transfer coefficient between cooling water and the cast ployed because with generally applied method, such as finite differ-
slab was measured under various conditions. Solidification analysis ence method or finite volumetric method using a calculation mesh
of the cast slab was conducted by inputting the measured heat coef- in a space, sufficient analysis cannot be conducted for flow analysis
ficients to boundary conditions. In the conventional solidification as spray water flow consists of mainly free-moving surfaces. The
analysis,9, 10) one-dimensional calculation at the center in the width particle method is one of the meshless methods and division of a
direction of a cast slab and/or two-dimensional analysis on a cross calculation space to mesh is unnecessary. Therefore, it is excellent
section perpendicular to the direction of casting are generally con- in representing free-moving surfaces. The applied MPS method is a
ducted; however, analysis that takes into account the unevenness in method that uses particles in a continuum as calculation points for
the width direction was not conducted sufficiently in a quantitative solutions. To solve the formula of continuity and Navier-Stokes for-
manner. Therefore, a model was developed that can take into ac- mula (Formula (2)), which are also calculated in finite volumetric
count the effects of spray, each roll, location of a bearing, dropping method, the gradient term is discretized using the interparticle inter-
water, and collected water. The cause of development of uneven action model, as shown in Fig. 7.
cooling in the width direction was studied using the model.
Dρ Du 1
3.1 Actual state of cast slab temperature unevenness = 0, =− p+v u+f (2)
∆ ∆ 2
Dt Dt ρ
Figure 6 shows an example of cast slab surface temperature
within a strand of a general vertical bending type slab continuous where u: flow speed, p: pressure, ρ: density, ν: dynamic viscosity
machine measured by a radiation thermometer. The surface temper- coefficient, f: external force (gravity), and t : time.
ature was measured by a width-direction-scanning type radiation Figure 7 shows the discretization of gradient vector of physical
quantity φ at the position of a particle i. Where r: the position of a
particle, d: number of dimension of space, n0: particle density, w:
weight-related function, set so as to exert less effect as interparticle
distance becomes larger, and < > is a symbol that denotes the identi-
ty of the interparticle interaction model.
In the calculation, for formularization of the model, particle di-
ameter of 3 mm, 30° of contact of angle of water with roll-slab, and
oval nozzle spray pattern were assumed.15)
Fig. 6 Comparison between the measured and the calculated surface
temperatures at 18 m below the meniscus Fig. 7 MPS method and formulation of gradient Φ
- 67 -
NIPPON STEEL & SUMITOMO METAL TECHNICAL REPORT No. 112 APRIL 2016
A typical bearing arrangement within a strand was taken up and rate side; since the diameter of the acrylic pipe is slightly different
three rows of rolls and two rows of spray zone in between were also from the diameter of the pipe used in the calculation, amount of wa-
taken up for forming the model for analysis. Figure 8 shows the ter collected on a roll became different, causing difference in the
model. The upper spray consists of eight nozzles and the lower amount of water overflowing outside the system.
spray consists of seven nozzles and they are arranged in a staggered Based on the analysis, it was found that the amount of water col-
manner with respect to casting direction. There are three separated lected on the middle separated roll of the whole split rolls is large,
rolls and two intermediate bearings installed in the center position the water collected on the roll drops at the bearing position, and the
of width. Distance between the spray nozzle and the slab is 155 mm
and the spray angle of a nozzle was 100° in width direction and 30°
in casting direction.
3.3 Spray water flow analysis by particle method and verification
by actual measurement
Figure 9 shows the result of analysis based on the model shown
in Fig. 8. It is a view from the slab side at five minutes after the start
of spray.
Figure 9 shows the result with the spray of water flow rate of 20
L/min per spray nozzle and it was found that the cooling water
Fig. 10 Measured spray water flow from the water model (view from
sprayed to the cast slab drops unevenly via over the bearings and the slab side)
water is collected on the upper side of the downstream rolls, over-
flowing backward. The accuracy of the model was confirmed by a
water model. In the water model, pipes made of acrylic resin were
pressed onto an acrylic plate simulating a slab, and water was
sprayed from nozzles laid in between pipes. An example of the view
of the result of the experiment is shown in Fig. 10. Flow rate of a
nozzle was 20 L/min, which is same with that of the calculation
shown in Fig. 9.
It was found that flow of spray water dropping via over a bear-
ing and the flow of water collected over the roll are represented well
by calculation. In order to assess the analysis result more quantita-
tively, calculation and experiment results of the quantity of the water
flowing out at bearing section and slab edge were compared. The re-
sult is shown in Fig. 11.
In regions 1–8 defined in Fig. 8, water flow rates obtained by
calculation and measurement quantitatively show good agreement.
Three graphs show each result in cases wherein the water flow rates
from a spray nozzle were 5 L/min, 10 L/min and 20 L/min. In re-
gions 6 and 7, the difference between the calculated value and the
measured value in water flow rate appears on the high water flow
Fig. 8 Simulation model for spray water flow between rolls (unit: mm)
Fig. 11 Comparison between the calculated and the measured water
flow rates
(a) Water flow rate of each nozzle is 5 L/min, (b) 10 L/min, (c)
Fig. 9 Calculated spray water flow (view from slab side) 20 L/min
- 68 -
NIPPON STEEL & SUMITOMO METAL TECHNICAL REPORT No. 112 APRIL 2016
flow of cooling water causes over cooling in the neighborhood of into account.
the slab center in the width direction. Therefore, solidification analy- As a result, overcooling at the center of the slab width was found
sis needs to be conducted taking into account the spray water flow. and uneven distribution of solidification in the width direction was
3.4 Assessment of uneven temperature distribution in width direc- proved (Fig. 14). Figures 14 (a), 14 (b), and 14 (c) show surface tem-
tion with solidification calculation in secondary cooling zone perature, heat transfer coefficient, and solid faction at the center of
In order to grasp the solidification condition in the secondary the slab thickness respectively. Heat transfer coefficient in Fig. 14 (b)
cooling zone, solidification calculation taking into account the measured at the position apart from the meniscus by 5 m is higher in
abovementioned dropping water and the like is necessary. Experi- the neighborhood of the center of the width due to the aforemen-
ment for measuring heat transfer coefficient in spray cooling was tioned interference of collected water with sprayed water. With this
conducted 16) by heating a steel plate installed with a plurality of effect, surface temperature at the position apart by 5 m from the me-
thermocouple and simulating the state of dropping water and col- niscus becomes lower, and from the center solid fraction distribution
lected water. Figure 12 shows an image of an experiment for meas in Fig. 14 (c), it is found that at width center portion, solidification is
uring heat transfer coefficient. A steel plate heated up to 900°C in an completed earlier than at a near edge portion.
atmosphere-controlled furnace was cooled by a single spray nozzle Figure 6 shows the comparison of results of surface tempera-
immediately after being extracted from the furnace and the heat tures calculated and measured by the aforementioned radiation ther-
transfer coefficient was sought for by inversely analyzing the meas mometer. The trend of temperature at the width center was lower by
ured temperature.17) more than 100°C as compared to the temperature at the near edge
Experiments were conducted for various spray nozzle water flow portion. Cause of overcooling at the width center is considered to be
rates and steel plate temperatures and the heat transfer coefficient attributed to spray dropping water being collected on the center split
was arranged as a function relative to spray water flow density on roll, interfering with spray water and thereby increasing the heat
the surface of the plate, steel surface temperature, and spray imping- transfer coefficient.
ing pressure.
Effect of dropping water on heat transfer coefficient was meas
ured by providing water flow from above the spray simulating drop-
ping water over a bearing. Flow rate of dropping water was decided
by referring to the water flow rate obtained by the result of the parti-
cle method (Fig. 11). The effect of the collected water was studied
in a cooling experiment wherein a plate simulating a roll made con-
tact with the plate. Water that was simply collected and stayed on
the roller does not exert big cooling effect; however, it was found
that when collected water interfered with sprayed water, cooling Fig. 13 Boundary conditions for the simulation model of solidification
was promoted in the interference region as the water was stirred. As
heat coefficients of a spray nozzle itself, dropping water, collected
water, interference region, and so on could be measured from these
experiments; they were used as boundary conditions of solidifica-
tion calculation within a strand. The boundary condition of heat
transfer was divided to four regions in between rolls (Fig. 13).
I is the roll cooling region, II is the air cooling or cooling by
dropping water region, III is the spray cooling region, and IV shows
the collected water cooling or cooling region by spray dropping wa-
ter. As for the slab width direction, the heat transfer coefficient, pro-
vided to the region wherein dropping water interferes with spray
water (II), was 1.1 times higher than the value of the case of drop-
ping water only. Similarly, heat transfer coefficient 1.5 times higher
than the value of the case of collected water only was provided to
the region wherein collected water interfered with spray water (IV).
These magnification rates of coefficients were calculated from the
result of the aforementioned heat transfer coefficient measurement
experiment wherein dropping water and collected water were taken
Fig. 14 Calculated results of solidification
(a) Surface temperature, (b) Heat transfer coefficients, (c) Solid fraction
Fig. 12 Experimental image (water is sprayed on the heated steel plate) at the center of the slab in thickness
- 69 -
NIPPON STEEL & SUMITOMO METAL TECHNICAL REPORT No. 112 APRIL 2016
4. Conclusion evenness of solidifying shell in width direction. The tempera-
• A 3D model that analyses the solidifying shell distortion be- ture distribution in width direction obtained by the analysis
havior in a continuous casting mold was developed and the ef- showed good agreement with the values measured with a radi-
fect of the mold narrow face taper on the development of shell ation thermometer. The extreme drop in temperature at the
was quantitatively assessed. It was found that a gap is devel- center of the width is considered to be influenced by over cool-
oped between the solidifying shell and the mold in the neigh- ing by interference of collected water on the middle part of a
borhood of a corner and a delay in solidification takes place roll with spray water.
when a single taper narrow face mold is used. It was found that
by employing a multiple taper (larger taper on upper section), References
the gap at the corner decreased, and solidification becomes 1) Thomas, B.G. et al.: Modeling of Casting and Welding Processes IV.
Metals & Materials Society, 1988, p. 287
uniform. Measured values of solidification uniformity ratio
2) Wang, Z. et al.: Transactions of the JSME, Series A. 53 (492), 1739 (1987)
show good agreement with the quantitatively calculated values 3) Tatsumi, N. et al.: Transactions of the JSME, Series A. 55 (514), 1389
and the accuracy of analysis was confirmed. By using the (1989)
model, narrow face mold shape design was optimized and ap- 4) Ohnaka, I.: Introduction of Computer Simulation of Heat Transfer and
plied to an actual machine. Solidification. Maruzen, 1985, p. 184
5) Matsumiya, T. et al: Tetsu-to-Hagané. 68, 1782 (1982)
• In order to research the cause of the phenomenon of uneven 6) Matoba, Y. et al.: Transactions of the JSME, Series A. 66 (646), 1127
solidification in the cast slab width direction developed in sec- (2000)
ondary cooling in continuous casting, flow of spray water 7) Matsukawa, T. et al: Kawasaki Steel Giho. 19, 7 (1987)
within a strand was numerically analyzed by the particle meth- 8) Koshizuka, S.: Particle Method. Maruzen, Tokyo, 2005
9) Petrus, B. et al.: Metall. Mater. Trans. 42B, 87 (2011)
od. With the analysis of spray water flow with particle method,
10) Kimura, M. et al.: CAMP-ISIJ. 2, 1154 (1989)
states of spray dropping water over bearings and spray water 11) Yoshida, T. et al.: CAMP-ISIJ. 1, 168 (1988)
being collected on a roll were clarified. An experiment measur- 12) Koshizuka, S. et al.: Nucl. Sci. Eng. 123, 421 (1996)
ing heat transfer coefficients by simulating the interference of 13) Koshizuka, S. et al.: Fluid Dynamics J. 4, 29 (1995)
dropping water and collected water made clear by the flow 14) Particleworks Ver.4.0.1 User’s Manual. Prometech Software Inc., 2012
15) Yamasaki, N. et al.: Tetsu-to-Hagané. 99, 593 (2013)
analysis was conducted. From the result of solidification calcu- 16) Kubori, S. et al.: AISTech. 2010
lation using measured heat transfer coefficients, it was found 17) Beck, J.V.: Inverse Heat Conduction. Wiley-Interscience Publication,
that dropping water over the bearing in the secondary cooling Hoboken, 1985
within a strand and collected water on a roll effected in the un-
Norimasa YAMASAKI Masahiro DOKI
General Manager Senior Manager
Maintenance Planning Dept. Process Engineering Dept.
Plant Engineering and Facility Management Center Equipment Div.
2-6-1 Marunouchi, Chiyoda-ku, Tokyo 100-8071 Kimitsu Works
Shozo SHIMA Yuichiro KATO
General Manager Senior Manager
Continuous Casting Plant Engineering Dept. Steelmaking Quality Control &Technical Dept.
Plant Engineering Div. Steelmaking Div.
Plant Engineering and Facility Management Center Kimitsu Works
Keiji TSUNENARI Daisuke MIKI
Senior Manager Senior Manager
Equipment & Construction Procurement Dept. Production & Technical Control Div.
Machinery & Materials Procurement Div. Nagoya Works
Satoru HAYASHI Takeo NAKANISHI
Researcher Manager
Heat-Processing Technology Dept. Steelmaking Technical Dept.
Process Technology Div. Steelmaking Div.
Process Research Laboratories Oita Works
- 70 -
