JPH0292413A - Forced cooling method for shape - Google Patents

Abstract

PURPOSE:To improve shape and quality of the shapes by setting density of cooling water of each cooling nozzle of plural numbers disposed lengthwise to cool the vertical flange surface of the shapes in a certain process. CONSTITUTION:The plural cooling water injecting nozzles 9 are disposed lengthwise opposing to the flange 1a vertical to the horizontal web 1b of the H shaped steel. By these nozzles 9, the shapes are cooled forcibly. Thereby, effective water quantity density is obtained previously through a test piece for cooling. Through the effective quantity density corresponding to the cooling width of each nozzle and the effective cooling water density of the uppermost nozzle, an influence coefficient of down flowing water is obtained. Then, through the relation between the ratio of the down flowing water volume density to the cooling water volume density and the influence coefficient of the down flowing, the cooling water volume density to satisfy the target effective water volume density of each cooling width of the shapes 1 is set successively to the cooling water injecting nozzles from the upper to the lower parts to perform the cooling. By this method, the shape and quality of the shapes are improved.

Description

[Detailed Description of the Invention] [Industrial Application Field] At least one surface to be cooled of the rolled section steel is placed almost perpendicularly to a horizontal surface, such as a flange surface of an H section steel, a ■ section steel, etc. The present invention relates to a forced cooling method for suppressing vertical bending by water cooling during or after hot rolling.

(Prior Art) Generally, as shown in FIG. 8, if the temperature distribution in various parts of the H-section steel 1 becomes uneven during rolling manufacturing or after rolling, upward or downward bending may occur in the H position. Are known. If the degree of bending is large, it may become difficult to convey the steel on the roller table 2, or the H-shaped steel may not be smoothly introduced into the roller straightening machine during the straightening process.

Such bending occurs when the temperature of the lower flange becomes higher than the temperature of the upper flange due to the influence of the closed space surrounded by the roller table 2, the lower part of the flanges on both sides, and the quebe tb. This occurs when forced cooling from the outer surface of the flange is not uniform on the upper and lower sides of the flange to reduce residual stress due to the unique cross-sectional shape in which the web 1b is thinner than the thickness of the web 1b. Conventionally, various cooling means have been proposed to make the temperature distribution uniform within the cross section of the rolled material.
Japanese Patent Publication No. 30378 proposes a water cooling device that injects cooling water onto a surface inside a lower flange of an H-shaped steel that does not include the lower surface of the web. In addition, Japanese Patent Publication No. 57-4407 provides cooling for the entire lower inner surface of H-shaped steel and the outer surface of the flange, and Japanese Utility Model Publication No. 1977 provides a device for cooling only the outer surface of H-shaped steel flanges.
-18679 etc. are well known. By the way, recently,
Attempts have been made to produce thin web H-section steel by rolling. Thin-web H-beam steel is a high-performance product that has improved cross-sectional performance relative to unit weight by making the H-beam steel web thinner, or by making the web thinner and thickening the flange, in response to the rise in the height of buildings in recent years. It is H-shaped steel. Now, since this thin-walled web H-section steel has a large wall thickness ratio between the flange and the web, the thermal stress becomes larger.In addition, the rigidity of the thinned web is small, so the web easily seats due to thermal stress. The cooling web waves are generated. For this reason, in order to manufacture thin-walled web H-section steel by rolling, forced cooling of the flange is essential, and the degree of cooling is much greater than in the case of conventional sizes. The degree of uniform cooling also increases, and the degree of bending also increases.

By the way, the forced cooling state of the outer surface of the flange of the H-section steel in the rolling position is such that cooling water is injected from the horizontal direction to the vertical plate. For this reason, the lower part of the flange is affected by the cooling water flowing down the flange surface, and the cooling rate is different between the upper and lower parts of the flange, causing the problem that the temperature distribution in the width direction (up and down direction) of the flange surface of the rolled material is not uniform. There is.

Now, in general, methods such as spray cooling, laminar flow cooling, jet cooling, and mist jet cooling are adopted as forced cooling methods for steel materials, but the heat transfer coefficient for estimating the cooling rate and temperature is based on the cooling water volume density, Empirical formulas are used as a function of surface temperature, etc., such as "forced cooling of steel" (
Various experimental formulas are shown in November 1973, published by the Iron and Steel Institute of Japan). However, most of the cooling systems that are the premise for deriving these experimental formulas are configured to cool slabs, plates, etc. whose surface to be cooled is in a horizontal position from above or below.

Therefore, in forced cooling of the outer surface of the H-section steel flange in the H position, it has been found that it is not appropriate to evaluate the cooling capacity using such a general experimental formula.

[Problem to be solved by the invention]

The present invention provides a flange forced cooling method for suppressing vertical bending in the H-rolling posture when manufacturing H-shape, ■-shape, etc. shaped steel sections having flanges by rolling. provides a method for setting an optimal cooling water volume density distribution in order to obtain a uniform or targeted effective water volume density distribution in the width direction (height direction) of a vertical surface of a section steel.

[Means and actions to solve the problem]

FIG. 9 shows an example of an H-shaped rolling mill train for carrying out the present invention, in which the rolled material is heated in a heating furnace 3, then roughly rolled in a breakdown mill 4, and then rolled in an intermediate universal mill 5 and tandem thereto. It is rolled to a size and shape close to that of the final product using an etching mill 6 that reduces the flange width, and is further shaped using a finishing universal mill. The flange cooling equipment 8a, 8b is for water cooling during reverse rolling by the intermediate universal mill 5 and the edging mill 6, and is provided on both sides of the rolling line. Similarly, the flange water cooling equipment 8c is installed to perform water cooling after finish rolling in the finishing mill 7.

Figure 2 is viewed from the front in the rolling direction. Cooling water injection nozzle (
This is an example of the arrangement of cooling water (hereinafter referred to as nozzles), and cooling water is injected onto the outer surface of the flange f of the H-beam 1 by a plurality of nozzles 9. FIG. 2(a) shows an example in which the nozzle 9 is arranged right sideways with respect to the flange width direction, and FIG. 2(b) shows an example in which the nozzle 9 is arranged so as to spray downward from diagonally above. In the H-type tI41 shown in Figure 3, the width f of the flange 1a and the height h of the web 1b differ depending on the product size, so in the nozzle arrangement as shown in Figures 2(a) and (b), the outer surface of the flange and the nozzle are horizontal. It is necessary to select the distance @d, the height direction position of the nozzle, and the nozzle stage. Figure 4 shows the cooling water injection form, and (a) shows the flange Ia.
(b) shows a cross-sectional view of the jet water as viewed from the front in the rolling direction.

First, when actually setting the position of the nozzle 9, the following conditions are considered. That is, the inclination angle α between the center of the nozzle jet direction and the horizontal plane in the cross section perpendicular to the rolling direction, the twist angle β of the nozzle with respect to the horizontal plane in the rolling direction, the nozzle interval UV in the flange width direction, the longitudinal nozzle interval fLh,
Items include horizontal pressure @d from the outer surface of the flange to the nozzle. In addition, these conditions are determined by considering the impact width Wh and Wv of the cooling water in the flange width direction and length direction, and the overlap and gap between the positions where the cooling water injected from adjacent nozzles collides with the flange, such as Sv and Sh. etc. should also be considered.

FIG. 5 schematically shows the form in which the cooling water collides with the flange 1a and spreads on the flange surface, and the shaded area A is the area where the cooling water is directly sprayed and forcibly cooled.
Range B, indicated by the horizontal line, is not directly injected, but is forcibly cooled by the cooling water rising upward along the flange surface, and range C, indicated by the vertical line, is not directly injected, but is cooled. The area that is cooled by flowing water, the white area is the area that is cooled by flowing water and rising water, and the dotted area E is an area that is not forcibly cooled. By the way, in the case of the above-mentioned example of the rolling apparatus, the nozzles are installed vertically on both sides of the roller table in the conveying direction, and are arranged so that the direction of jetting cooling water is constant on the flange surface of the H-section steel, so that the H-section steel can be rolled by the rollers. The flange surface is cooled while being transferred on a table. Therefore, it is easy to uniformly cool the flange surface in the product length direction, but the flange width direction (
Uniform cooling in the height direction) is not easy. That is, within the range of the flange width f in FIG. 5, the width W1 of the upper end is the part that is not subjected to forced cooling, the width W2 is the part that is forcibly cooled by the swelling of water jetted from the top nozzle, and the width W3 is the part that is not forcedly cooled. Although there is a swell of water jetted from the nozzle, the portion that is forcibly cooled mainly by the direct jet of cooling water and the flowing water, and the width W4 is the portion that is forcibly cooled only by the flowing water. The existence of the portion W1 that is not forcedly cooled makes uniform cooling in the flange width direction and symmetrical cooling of the upper and lower flanges impossible, so it is preferable to make it as small as possible. If the injection position of the nozzle is too close to the top of the flange, the cooling water will enter the upper surface of the web, causing serious problems such as the generation of web waves due to thermal stress in thin-walled web H-beam steel, etc., where it is necessary to avoid rapid cooling of the web. will occur. Furthermore, since the portion W2, which is cooled only by the rising water, has a smaller cooling capacity than the portion W3, it is desirable to make it smaller as well as Wl. It is clear from FIG. 5 that setting the twist angle β is effective in expanding the contact area between the cooling water and the material to be cooled.

Based on the above-described nozzle arrangement, the present invention provides means for setting the jet water volume density of each stage nozzle for the water-cooled area of the flange (W3 in Fig. 5), which can be cooled and controlled by adjusting the cooling water volume. In order to cool the entire surface of the flange uniformly or in a cooling pattern that is vertically symmetrical with respect to the center axis in the width direction of the flange, the jet of nozzle at each stage is determined based on the cooling rate measurement results obtained through experiments in advance. The effective water density (described later) is determined for each region, and the ratio of this effective water density to the effective water density of the top nozzle where the flowing water is zero (effective water density ratio) is calculated as the density of water flowing downward from the top and the actual water density. The purpose of this paper is to find an appropriate amount of water to be ejected from the nozzles at each stage by expressing it as a function of the cooling water amount density. Hereinafter, a specific method of the present invention will be described in more detail.

FIG. 6 shows an example of a model experiment method for investigating the influence of cooling water flow density, and (a) and (b) are a front view and a side view, respectively. The cooled test piece 10 has a thickness tfJ
(It is a thick plate with a width (height) of 19mm and a width (height) of 30θmm, and the nozzle 9 used is a flat spray type and weighs 3kg.
f/crr? The injection angle is 110' and the injection water amount is 10 fL/win.

The cooling water injection direction has an inclination angle of 30° and a twist angle β of 8°.
and a horizontal path 11 from the water-cooled surface of the test piece to the nozzle.
1d is 100mm. With this arrangement of the nozzle 9, the pressure is 2 kgf/err? The spray width Wv of the cooling water from the nozzle in the width direction (height direction) of the specimen is approximately 50 m.
The injection width wh in the mb length direction is approximately 350 mm. The nozzle arrangement interval is 50 mm in the specimen width direction (height direction).
The length was set to 350 mm in the length direction, and the overlapping of the injection areas of the cooling water injected from each nozzle was considered to be small. Therefore, the width of the cooling water sprayed from all nozzles in the width direction of the specimen is 250 mm, and the position of the nozzle in the height direction is the width W2° of the upper end of the specimen and the width W4 of the lower end (both 25+n
m) is not directly sprayed with cooling water, that is, the upper end 25 mm is a part that is forcibly cooled by a swell of cooling water except for the extreme end, and the lower end 25 mm is a part that is forcibly cooled only by flowing water. . The portion W3 excluding these ends is cooled by a mixture of injection cooling and downstream cooling.

The density of the amount of cooling water injected from the nozzles arranged in this way (the amount of water ejected per unit time by each nozzle is expressed as the area of the water cooling surface directly injected from that nozzle wh)
Water cooling was performed with the same value (value divided by Xwv) for all stages. The temperature measurement position 11 is carried out by attaching a thermocouple to the position indicated by . Therefore, the part W3 that receives the cooling water directly, the upper end part W2 that is cooled only by the rising water, and the part W2 that is cooled only by the flowing water. The temperature of the lower end W4 is also measured. To evaluate the cooling capacity, use heat transfer calculations to determine the relationship between the cooling water volume density and cooling rate at the same plate thickness as the test piece used in the experiment and at the same distance from the water cooling surface as the temperature measurement position. The cooling water density at which the cooling rate in the experiment satisfies this calculated relationship is determined by back calculation, and this value is used. In the present invention, this cooling water density is referred to as the effective water density.

The water-cooling heat transfer coefficient used for heat transfer calculations is based on the above technical document “
``Forced cooling of steel materials.'' Note that the average cooling rate in the same temperature range was used for both calculations and actual measurements.

Regarding the amount of water flowing down, in the area that receives direct injection from the nozzles of each stage, we ignore the amount of water jetted by the nozzles of that stage themselves, and focus on the flow of water jetted from the nozzles of the upper stage. That is, the amount of flowing water in the injection area of the top nozzle can be expressed as τ, and the density of the amount of flowing water in the area that is directly injected from the first-stage nozzle, with the top layer as the first stage, can be expressed by equation (1). That is, in the present invention, the flowing water density refers to a value obtained by dividing the amount of water injected from a nozzle above the nozzle by the area of the injection region of the nozzle.

Qr+−Ws X (i −1) ・・・”・・・(
1) However, the flowing water volume density (It/m'm1n) of the area that receives the direct cooling water injection from the nozzle on the Qr11 step surface Ws: Cooling water volume density per nozzle (117 m''
lll1n) From equation (1), the cooling water amount density Ws of the first stage nozzle is Ws (in this example, since the injection conditions are uniform across all stages).

The ratio of the flow rate density QP to the flow rate density QP (equal to ) (hereinafter referred to as flow rate density ratio) can be expressed as (i-1) when the jet water rate density of the nozzles at each stage is the same as in the present experimental conditions.

Regarding the influence of running water, the ratio of the effective water volume density in the direct injection area of each stage to the effective water volume density of the top stage where running water is assumed to be zero (hereinafter referred to as the running water influence coefficient), that is, (2)
It can be evaluated with an expression.

ni,=W! ,/W! ：、・・・・・・・・・・・・ （
2) However, KI: Effective water flow density at the top of the effluent water influence coefficient W (fL/rr? win) wE,: Effective water flow density at the top row (fL/rr? m1n) Figure 7 shows the effective water flow density at the top row (fL/rr? m1n) In the example, the cooling water density is 40 in all stages.
0IL/nfmin are the same.

The area at the top ■ of the sample (range W2'' in Figure 6) is a region that is forcibly cooled by the swelling of cooling water, and the area at the bottom ■ (range W4 in Figure 6) is forced only by flowing water. The area to be cooled, the area marked with ■, is the area of W3 shown in Figure 6 that is forcibly cooled by direct injection from the top nozzle, and the area marked with ■ is forcedly cooled by direct injection of water flowing from the top. This is the area to be cooled. * indicates the value at each temperature measurement point, which is the average value of the values at two points in the length direction. A stepped solid line 12 is an area (■) that is not directly sprayed with cooling water.

For (2), the values at the measurement points are representative values for each area, and for the basins (2) and (2) that receive direct injection, the values are averaged for each spray area of the nozzle at each stage.

From this figure, we can see that the effective water flow density of areas ■ and ■ that do not receive direct cooling water injection is smaller than that of areas ■ and ■ that receive direct injection, and that the effective water flow density of area ■ that is cooled by rising water is that of flowing water. Although it is smaller than that in the basin (2), which is cooled only by water, it can be seen that there is no major difference.

In addition, in the areas (2) and (2) that receive direct cooling water injection, the effective water flow density is large in the lower nozzle injection area, indicating that the flowing water promotes cooling. In this way, if the water density distribution is made uniform for each film, the cooling will be asymmetrical with respect to the width center of the specimen to be cooled, which is the same in the case of flange cooling of H-beam steel, and this will not cause bending. it is obvious.

Areas ■ and ■ that receive direct cooling water injection can be controlled by adjusting the cooling water amount density for each nozzle in each stage, and areas ■ and ■ that do not receive direct cooling water injection are the top nozzles, respectively. This is automatically determined by the jet water density and the total water volume of all stages, and cooling control is not possible.

However, since the effective water density in the regions ■ and ■ is small and the width of the regions is small, the effect of the difference in cooling in these two regions on the bending is so small that it can be ignored. Therefore, if cooling control is performed such that only the area receiving direct cooling water injection is symmetrical, bending can be significantly suppressed.

FIG. 1 shows the relationship between the density ratio of the amount of flowing water and the influence coefficient of flowing water in the area directly receiving the injection of cooling water. in this case,
The cooling water amount density is changed by changing the injection pressure.

The solid line (O mark) for the cooling water flow density is 21042/rrl”m
in, the broken line (x mark) is 280 fL/m'min.

Δ,・.

mouths are 300, 400, and 500 ft/ni” respectively.
However, these three have almost the same tendency as shown by the dashed line.

As shown in this figure, the relationship between the effluent volume density ratio and the effluent influence coefficient varies greatly depending on the cooling water volume density.

Is the cooling water density 300-400 Il/rr? In the case of min, the flowing sewage promotes cooling, but the cooling water density is 200j! /rn"min, the flowing sewage actually suppresses cooling.

Although it is not clear why the relationship between the flowing water ratio and the flowing water influence coefficient differs greatly depending on the cooling water flow density, it is presumed that the cooling water flow density is adjusted by pressure rather than nozzle replacement.

In other words, when the flow rate is large, the injection pressure is high, so the injection water destroys the film of the flowing water, and the sum of the cooling capacities of both the injection water and the flowing water becomes greater than the cooling capacity of the injection water alone, and the amount of injection water increases. When is small, the injection pressure is low, so the force that destroys the film of the flowing sewage is weak, and the sum of the cooling capacities of both the injection water and the flowing sewage is smaller than the cooling capacity of only the injection water. It is interpreted that

The relationship between the effluent volume density ratio and the effluent influence coefficient in FIG. 1 can be roughly expressed by a simple equation such as equation (3).

However, W□: Cooling water flow density of the first stage nozzle (fL /
m'min) Here, if the coefficient a is limited to the range of the water flow density of this experiment, we will now discuss setting the appropriate cooling water flow density for actual equipment.

In equation (2), the effective water flow density WtI in the injection region from the top nozzle with no flowing water is almost equal to the actual cooling water flow density if the empirical formula for the heat transfer coefficient is highly accurate, and since there is no flowing water, It may be assumed that it is equal to the target effective water flow density (target effective water flow density) of the first stage nozzle.

Therefore, equation (2) can be expressed as equation (5).

To +mW! +/We Otori” W Ol / W 61 = W 6 I/ W g 1 ・・・・・・・・・
(5) Here, Wo+: Target effective water flow density of the i-th stage (
fl 7m”m1n) The amount of cooling water for each stage is Equation (3), Equation (4), and Equation (5).
It can be obtained from the formula. The calculations can be performed sequentially from the top stage assuming that the cooling water amount density Ws1 in the first stage is equal to the WOI.

When Wl≦300, when WS+≧300, Wg+-We+-0, 115XΣW,...
...(6°) ul Therefore, in order to perform vertically symmetrical cooling with respect to the center of the flange width to suppress bending, it is sufficient to make the target effective water density at the vertically symmetrical positions the same, and when aiming for uniform cooling, The amount of water to be injected may be calculated by keeping the target effective water amount density of each stage the same.

In G&D, non-uniform cooling also occurs within the injection area of the nozzles at each stage, but this is an unavoidable problem because it is difficult to control the distribution of the amount of water injected from the nozzles.

However, this problem can be almost solved by increasing the number of stages of nozzles installed and controlling the density distribution of the amount of cooling water from the nozzles at each stage. There isn't.

If the number of nozzle stages is one or two, and the width of the spray area from the nozzles in each stage is large, the influence of the flow of water jetted from the nozzle itself is large within the spray area of each nozzle, making cooling control difficult. becomes impossible.

Among the cooling conditions in the experiment, the nozzle placement, inclination angle, and twist angle settings were almost the same as in the actual equipment, and the nozzle specifications were also the same as in the actual equipment to obtain highly reliable data.

Instead of using the value obtained by dividing the amount of water injected from a nozzle above the nozzle by the injection area of the nozzle as the flowing water density as in this example, we use the collision width of cooling water in the longitudinal direction (Wv in Fig. 4). ), but if the collision width of the cooling water in the width direction (height direction) of each stage (wh in Figure 4) is not -like, it is better to use this value. Good.

In addition, the reason why the cooling water flow density is the same for all stages is that even if there is variation in data due to the influence of the surface condition of the water-cooled test specimen (for example, scale thickness, etc.), the injection area of the top stage nozzle with no flowing water This is because the effect on the cooling capacity of flowing water can be evaluated using equation (2) using the effective water density as an evaluation criterion.

(Example) Flange water cooling of H-section steel was carried out using actual equipment in which nozzles of the same type as above were arranged under the same conditions.

The conditions are shown in Table 1. Condition 1 is a cooling water flow density distribution that is uniform throughout all stages, and Condition 2 is a water cooling water flow density distribution that is symmetrical and uniform with respect to the center of the flange width determined from the above formula, and the target effective water flow. Density is 400 (J2/ln"1in) for all stages of nozzle
)be.

The size was expressed as web height x straw 29 width x web thickness/flange thickness.

The number of tests is 10 each.

The bending in the vertical direction of the rolling posture (δ in Fig. 8) was measured for a length of 5 m in the rolled material 1. Table 2 shows the bending amount immediately after water cooling and at room temperature.
+ indicates an upwardly convex curve, and - indicates a downwardly convex curve.

From this result, it is clear that the cooling water amount distribution throughout the stages results in non-uniform cooling, and it was confirmed that the setting of the cooling water amount density in the method of the present invention is highly effective.

When the material control technology that performs forced cooling after controlled rolling comes into use for shaped steel, similar to thick steel plates, stricter cooling conditions will be required, but by applying the method of the present invention, the problem of bending can be solved It is possible to solve the problem, and it is also possible to set an appropriate cooling water amount density distribution even when a target effective water amount density is required depending on the temperature distribution.

Table 1 Table 2 (Effects of the Invention) When forcibly cooling a vertical flange surface of a section steel, the method of the present invention requires only setting the cooling water density of each of a plurality of vertically installed cooling nozzles in a fixed procedure. The temperature distribution on the flange surface can be made uniform or the temperature distribution on the flange surface can be controlled to a desired pattern. Therefore, when the cooling method of the present invention is applied to the rolling process of H-section steel, vertical bending is eliminated, and transportation or straightening work becomes easier. Furthermore, in the production of steel sections with a large difference in thickness between the flange and the web, such as thin-web H-section steel, the effect of improving the shape or product purchase is extremely large.

[Brief explanation of drawings]

'M1 diagram is a graph showing the relationship between the effluent volume density ratio and the effluent influence coefficient related to the mechanism of the present invention, Figure 2 (a)
, (b) is a schematic front view showing the state of cooling water injection to the flange surface of the cooled material, Fig. 3 is a cross-sectional explanatory view of the H-section steel,
Figure s4 (a) is a schematic side view showing the cooling water injection pattern to the cooled surface, Figure 4 (b) is a schematic front view showing the cooling water injection situation of Figure 4 (a), and Figure 5 is the cooled surface. Figure 6 (a) and (b) are diagrams explaining the types of water-cooled parts shown above, Figure 6 (a) and (b) are diagrams explaining the water cooling model of the cooling test piece, and Figure 7 shows the distribution of effective water density by position of the cooling test piece. The graph shown in FIG. 8 is a perspective view illustrating the bending state of H-section steel, and FIG. 9 is a schematic diagram showing an example of a section steel rolling apparatus row. 1... H type rlIla... Flange lb-... Web 2... Roller table 3... Heating furnace
4... Breakdown mill 5... Intermediate universal mill 6... Etching mill 7... Finishing universal mill 8a, 8b, 8C... Flange water cooling equipment 9... Cooling water injection nozzle 10... Covered Cooling test piece 11...Temperature measurement position and others 4
Figure 1 (a) (b) 9: Cooling water injection nozzle Figure 3 Figure 4 (b) Figure 5 10: 1ffl cooling test piece 11: Temperature measurement position

Claims (1)

[Claims] 1 In a method of forcibly cooling a section steel with at least one surface to be cooled placed substantially perpendicular to a horizontal surface using a plurality of cooling water injection nozzles installed vertically facing the section steel, The effective water flow density when a constant cooling water flow density is injected from each of the cooling injection nozzles to the specimen to be cooled is determined, and the effective cooling water density corresponding to the cooling width of each nozzle in the height direction of the surface to be cooled is determined. The flowing water influence coefficient is determined from the water flow density and the effective cooling water flow density of the uppermost nozzle of the cooling water injection nozzles, and then the above form is calculated from the relationship between the ratio of the flowing water flow density and the cooling water flow density and the flow flow influence coefficient. A forced cooling method for shaped steel, characterized in that cooling is performed by successively setting a cooling water flow density that satisfies a target effective water flow density for each cooling width of the steel from an upper cooling water injection nozzle to a lower nozzle.