Cooling of Strip on Run out Table in a Hot Strip Mill

Cooling of Strip on Run out Table in a Hot Strip Mill

Hot strip mill is used to produce hot rolled coils of prescribed strip thickness and properties from a slab of known dimensions and chemical composition. Typically, a present day hot strip mill is located after continuous casting facility and consists of reheating furnaces, descaling units, roughing mill stands, shear, finishing mill stands, run out table and down coilers. For conventionally controlled hot rolling, controlled deformation is normally applied in the austenitic temperature range accompanied by austenite recovery and / or recrystallization. Controlled run out table cooling is central to the control of mechanical properties of hot rolled strip.

In a hot strip mill, the hot rolled coils are produced from slabs through hot deformation and subsequent cooling with water on a run out table. In conventional rolling, the starting slab thickness is around 200 mm to 250 mm. On the other hand, in thin slab casting and rolling, the starting slab thickness is around 50 mm to 90 mm. In both cases, the slab is then subjected to a sequence of passes to reduce its thickness down to around 1 mm to 18 mm. After all the rolling deformations are complete, the strip is cooled with water, and the huge length of strip (1.5 km to 1.7 km) is coiled in a down coiler to produce the final product (hot rolled coil).

A run out rolling table normally consists of (i) a bed of rollers on which the hot strip moves, and (ii) an arrangement of cooling nozzles which cools down the hot strip. Ultra-fast cooling is done by using a high flow of water and air. Cooling methods include curtains of jets, spray, and circular jet etc.

In a high capacity continuous hot strip mill, the length of a run out table can be as long as 150 m. The run out table is equipped with a water cooling system normally arranged in groups or banks for convenience of control between 75 m to 105 m. The strip is conveyed through the run out table by an arrangement of motorized rolls. Typically, the rolls used on the run out tables are around 300 mm in diameter and are hollow, cast iron or low carbon steel units on forged steel hubs located at 450 mm centres. They are individually driven by DC motors of 2 kilowatts (kW) to 4 kW under adjustable voltage control and are skewed and canted to keep the strip centered along the table.

Run out table and strip cooling zone is between the last finishing mill stand and down coiler in the hot strip mill. This area is used to cool the hot strip to reach the desired temperature and expected quality level and before it is coiled by down coiler. After leaving the finishing mill, the strip is carried down by a large number of individually driven rollers through 4 to 12 banks of low-pressure, high-volume water sprays which cool the red hot strip to a specified coiling temperature which can be ranging between 400 deg C and 900 deg C before the strip is led into down coilers. Metallurgically critical to the properties of hot rolled steel is the coiling temperature, as the coil cools from this temperature to ambient over the course of 3 days. Essentially a heat treatment comparable to annealing, the stresses imparted to the steel during reduction from the slab thickness down to the thickness of the hot strip are given the opportunity as the coil cools to relieve them.

For a given steel grade, the most important processing parameters in the run out table operation are (i) finishing temperature, (ii) coiling temperature, (iii) strip temperature and moving velocity, (iv) cooling water temperature, (v) layout of the active cooling jets, (vi) water flux, and (vii) strip gauge. Fig 1 displays the hot strip mill run out roller table showing top cooling headers.

 Fig 1 Hot strip mill run out roller table showing top cooling headers

In cooling process, as the strip moves from the finishing mill to the down coiler, water distribution is to be adjusted with variable rolling velocity to meet target coil temperature. Water cooling process is a complex process affected by several factors, consisting mainly of strip temperature, water flow, strip thickness, chemical composition, and strip velocity etc. Once the strip enters cooling zone, with little difference in strip thickness, width, and composition in strip length direction, start cooling temperature and velocity are the main factors influencing temperature control. Presently, to increase the productivity and decrease heat loss on delay of transfer bar, finish rolling process in the hot strip mill adopts the variable-velocity rolling strategy, which leads strip velocity to vary in a large scale.

Meanwhile, the adjustment of rolling velocity is also used to control finishing mill delivery temperature, (FDT), which is also the start cooling temperature. The rolling acceleration / deceleration and velocity are dynamically changed to meet the aim FDT and considerably influence coiling temperature (CT) control accuracy. Under variable rolling velocity, water distribution expressed as open / closed cooling headers and header flow is the only adjustment method to ensure target coil temperature. Hence, precise water distribution calculation based on accurate velocity prediction is crucial for cooling path and temperature distribution control in the cooling process.

Several run-out table pumps feed cooling water to the header tank, which supplies the laminar flow cooling tubes on a hot strip mill. Constant water level in the header tank produces a constant pressure at the flow control valves in the laminar flow lines and allows good flow control. If too much flow enters the header tank, the surplus water overflows into the scale water pit, and wastes pumping energy. Fig 2 shows the strip cooling control system.

Fig 2 Strip cooling control system

In order to improve this situation, a water supply control system has been developed which uses existing control information and predictive information from the strip mill models. This control sets the speed of variable frequency drives (VFD) connected to the pump motors, and operates as (i) the coiling temperature control (CTC) computer calculates the required flow of the cooling water in advance, (ii) the water supply controller receives the required cooling water flow data from CTC computer, and the mill setup and piece tracking models, (iii) the water supply controller predicts the water volume using this information and determines the speed of the run out table feed pumps, and (iv) during a roll change, the run out table feed pumps are stopped or slowed. Using this system, pump energy savings of 5,010 MWh/y (megawatt hour per year) can be realized

In production of hot strips, control cooling process on run out table is important for the steel product pursuing excellent and stable performance. After rolling in the finishing stands, the strip is cooled by water from a start cooling temperature to target temperature with a given cooling rate before its coiling. This promotes achieving more fine grain size and improves strengthening and toughness. In control cooling process, high precise temperature distribution prediction and cooling path control are key factors for ensuring the mechanical property of the steel. The final properties of the steel are also dependent on the possible grain growth after the run out table cooling.

Run out table temperature control is one of the most critical processes to ensure mechanical properties of the steel strip. Accelerated cooling has been implemented in hot strip mills to reduce the heat loss on the delay table, to achieve the needed coiling temperature and to achieve the desirable grain structure. Appropriate cooling application with better understanding of strip cooling characteristics ensures the precise temperature and cooling rate control.

For achieving high precise temperature control of entire strip length, the strip is divided into number of segments, and each segment is separately controlled. Similarly, cooling zone is also divided into several micro cooling zone, same length with segment. In cooling process, each segment passes through all micro cooling zone in sequence from FDT pyrometer to CT pyrometer, with temperature drop increasing, strip temperature finally meets the target value.

Controlled cooling of the hot rolled strip on the run out table has become a common practice in the production of hot rolled coils. Controlled water cooling systems are used along the run out table to control the phase transformation process and flatness while precipitation strengthening mainly occurs in the process of coiling and cooling after coiling. Over the past 30 years or so, a considerable effort has been made to quantify the thermo-mechanical and metallurgical phenomena during controlled hot rolling with the aid of more fundamentally based mathematical models.

The introduction of water jet cooling by BISRA (British Iron and Steel Research Association) in 1957, opened a new technology to produce high strength low alloy (HSLA) steels. Hence, fine ferrite grain size, combined with precipitation strengthening, can be manipulated to achieve high strength in steels with a reduced carbon content. As a result of reduced carbon content, a smaller quantity of pearlite is present in the microstructure which improves the weldability and toughness of the steel. Accelerated cooling refines ferrite grain size because of the lower transformation temperature and the resulting undercooling of the austenite.

The purpose of using water jet cooling is to achieve a very high heat extraction rates to produce a fine ferrite grain size (around 10 micrometer). A uniform temperature field in the thickness direction is desirable. However, the thermal resistance of steel is an impediment. Inhomogeneity in mechanical properties and distortion of the strip results from localize cooling. Hence, optimization of the cooling pattern is needed to avoid such problems.

The control of water cooling on the run out table is aimed for achieving a good microstructure and resultant mechanical properties of the steel strips. Different steel grades need different thermal treatments, or cooling path, to optimize their mechanical properties. Tab 1 gives typical cooling strategies together with the products for which they are used. From this point of view, flexible set up of run out table facility and flexible control of cooling water play a key role in fulfilling the cooling strategies enumerated in Tab 1. Hence, the flexible run out table control for water cooling is normally considered as a potential and powerful approach to further improve mechanical properties for meeting the requirements of the steel strip users.

Tab 1 Cooling strategies
Early coolingFor tensile grades, i.e. hard materials
For ‘interstitial free’ (IF) steels
Late coolingFor soft grades for further processing
For cold rolled products
Interrupted coolingGap temperature control
Gap interrupt time control
Initial cooling rate control
For dual phase steels
For carbon-manganese steels
For TRIP (transformation induced plasticity) steels
For galvanizing products

For conventional controlled rolling in the hot strip mill, the finishing temperature lies in the range of 850 deg C to 950 deg C and the coiling temperature is in the range of 510 deg C to 750 deg C. For reducing the temperature of the strip from its value on emergence from last stand of the finishing mill to a value appropriate for coiling, the strip is transported along a run out table, such as shown in Fig 1, and cooled mainly with the application of water. Fig 3 shows schematic setup of run out table.

Fig 3 Schematic setup of run out table

The rolls are placed at appropriate distances to avoid excessive bending of the strip. The water cooling systems are mounted at the top and bottom of the strip. The cooling banks consists of headers which supply water to the arrangement of jet lines from which the water impinges on the steel through rectangular nozzles creating a water curtain, or through circular nozzles which produce water bars. The bottom cooling headers are normally installed less than 100 mm below the moving strip. The top cooling headers are normally at least, 1,200 mm above the table to avoid damage by cobbles of the moving strip.

The water is supplied from water towers to large storage tanks located below or above the run out roller table. When the water tanks are below the run out table, the water is pumped to the headers and then to the jet nozzles. Pressure is needed to ensure that the quantity of water is sufficient and to maintain the stability of the water flow. When the tanks are above the run out table, water flow by gravity. The level of water is maintained constant in the storage tanks to control the quantity and stability of the water flow. The cooling water is normally supplied at room temperature through the top and bottom water jets, and the water is frequently filtered and recycled to the cooling towers after the application. Originally, spray cooling or the application of water by jets operating at 1.4 MPa to 1.8 MPa was standard for a run out table. However, during last 30 years or so, laminar cooling has gained increasing acceptance with the use of low pressure streams or water curtains. Fig 2 shows schematics of spray cooling, laminar flow cooling, and water curtain systems. Fig 4 shows three typical run out table cooling systems.

Fig 4 Three typical run out table cooling systems

Water jet configurations can be classified into five categories namely (i) free surface jets, (ii) plunging jets, (iii) submerged jets, (iv) confined jets, and (v) wall jets (Fig 5a). The free surface jet is injected into an immiscible environment (liquid into gas), and the liquid travels relatively unrestricted to the impingement surface. Laminar flow cooling and water curtain cooling belong to this category. The first four configurations induce flow fields on the impingement surface which are quite similar.

Fig 5b depicts representative conditions for planar, free surface jet. The inviscid pressure and stream wise velocity distribution for a uniform jet velocity profile are also shown. The pressure is maximum at the stagnation point because of the dynamic contribution of the impingement jet. With increasing stream wise distance, the pressure declines monotonically to the ambient value. Conversely, the stream wise velocity is zero at the stagnation point and increases to the velocity of the jet with increasing distance along the surface.

Fig 5 Schematics of cooling water jets used or hot strip cooling

For clarifying the boiling phenomenon at different locations on the impingement surface, the flow has been defined into (i) stagnation, (ii) acceleration, and (iii) parallel flow regions (Fig 5b). The combination of stagnation and acceleration regions is also named as impingement region. The stagnation region coincides with that of the impingement jet, in both size and location, and contains a nearly linear increase in the stream wise velocity. Within the acceleration region, the fluid continues to accelerate and approaches the jet velocity to within a few percent. For the parallel flow region, the stream wise velocity is essentially that of the jet and the hydrodynamic effects of impingement are no longer realized within the flow.

As shown in Fig. 6, the cooling strip is made up of four sections. These sections include (i) impingement region, (ii) parallel flow region, (iii) spray water at the strip bottom, and (iv) surrounded air. In Fig 5b more details are shown on the impingement zone.

Fig 6 Cooling region on run out roller table

The earliest water jet system consisted of arrangements of bars impinging on the top surface of the strip. Water jet cooling is sometimes called ‘laminar cooling’ because of the streamline flow and transparent glassy appearance of the jet, though the jet is not necessary in a laminar flow. Spray cooling, or water bar cooling at the bottom surface has been used, in an attempt to assure a more uniform cooling. Intrinsic inhomogeneity of the water flow in the width direction was expected to cause non-homogenous properties, but this effect was greatly reduced by alternating the pattern of the arrangement.

In order to improve heat extraction with water, planar water jets (sometimes called water curtains) issuing from slot type nozzles mounted in both upper and lower low pressure headers were developed.  The advantages of planar jets include (i) improved temperature control, (ii) elimination of clogged nozzles and nozzles, (iii) nozzle erosion, and (iv) simple valve arrangement which is compatible with the computerized control.  More uniform temperature profiles are expected from these systems jets.

The success of water jet cooling results from the direct liquid water contact with the high temperature surface of the strip. The jet momentum assures solid-liquid contact and hence enhances the heat transfer as compared to film boiling which is produced by spray cooling.

The results of experimental studies at Krupp Stahl AG as shown in Fig 7 clearly explains why the spray cooling was replaced by laminar flow cooling and water curtain. Laminar flow cooling has intermediate cooling capacity and water curtain has the highest cooling capacity among the three cooling systems. Besides the cooling capacity, uniform water distribution and resulting heat removal are other advantages of the laminar flow cooling and the water curtain.

Fig 7 Cooling rate of hot strip as a function of strip thickness

Immediately after the strip rolls out from the last stand of the finisher, it enters the run out table cooling zone. Depending on the finishing temperature, mill speed, and the needed coiling temperature, the hot strip is cooled by arrangements of top and bottom cooling headers in a distance of around 70 % of the run out table length, and continuously cooled by air convection and radiation until it is wound in the down coiler.

Under heat flux control, to maintain a higher heat flux than the critical heat flux, a large increase in the surface temperature is needed because of the very high thermal potential needed to sustain the heat flow by pure film boiling. In contrast, with the smaller temperature gradients, the liquid contacts the surface, and the cooling is very effective during that contact, generating a very high heat flow. The transition boiling mechanism is combination of both film boiling and nucleate boiling, which is also intermediate in effectiveness of cooling.

The typical shape of temperature controlled and heat flux controlled boiling curves are shown in Fig 8. The basic difference between this, and the heat flux controlled boiling curve is that in the latter the transition boiling regime does not exist.

Fig 8 Cooling of hot rolled strip on run out roller table

Jet cooling in the run out table is employed in a wide range of surface strip temperatures, typically from 900 deg C down to 300 deg C, and hence the cooling is carried out by different boiling mechanisms. More specifically, nucleate boiling, the critical heat flux, transition boiling, and film boiling are important for strip cooling on the run out table.

In the zone adjacent to the jet centerline, the free stream flow changes the direction and is required to develop in a finite length (impingement zone) in a direction parallel to the strip motion (co-current and counter-current). This involves a change in the pressure energy (pressure gradient) which is also parallel to the strip motion.

The symmetric cooling effect of the last dry cooling section allows the hot strip to re-distribute temperature by internal conduction. It is particularly important for mill operation with inevitable non-symmetric cooling in the water cooling zone. This cooling process cools the strip from around 870 deg C to below 730 deg C. It includes internal and external conduction, stagnation forced convection, forced boiling convection, air convection and radiation, and heat generation from material phase transformation. Although table roller conduction was recently found to be also a cooling element to cause a large local temperature drop, external conduction as well as air convection are normally treated together to achieve an empirical heat transfer coefficient equation for the mill application. Radiation heat transfer follows Stefan-Boltzmann rule with temperature dependent emissivity. Heat generation because of phase transformation depends on the ratio of transformed austenite. Transformation detectors have been applied to the run out table temperature control. However using temperature dependent thermal properties reduces the problem complexity and is widely adopted and accepted.

Forced convection by the cooling medium which counts for more than 90 % of the entire heat transfer is still not fully understood since heat transfer mechanism involves a complex mixing phenomenon of water impingement and boiling on a moving surface. For a production mill, analyzing the operating data can hardly achieve a concrete conclusion because of inherent data error from the mill harsh environment and the confounding effects between rolling parameters. Studies using a small scale equipment can provide more information, however these results do not reflect the true mill operation and call for modifications before application. In order to apply these theoretical knowledge to mill practice, it is useful to review the various studies of stationary, steady state moving, and mill application cases in the following.

Stationary cases – For a steady state condition of water impingement on a stationary plate, heat transfer adjacent to a jet impinging has been roughly classified into five regimes. The single phase forced convection zone is located right beneath the jet. Followed by the wet zones of nucleate / transition boiling, forced convection film boiling, and agglomerated pools, heat transfer gradually reduces to the dry zone by air convection and radiation. The impingement area has been estimated around 2 times to 4 times of the nozzle diameter or the curtain width. At the vicinity of the impact zone, an observable darken line shows that this zone can have the highest heat transfer. One of the studies has shown that heat transfer coefficient at this stagnation area held constant (23.26 kw/sqm.K) regardless of the impinging speed and the surface temperature if the surface temperature exceeds 400 deg C.

Another study demonstrated that the cooling surface temperature near the stagnation point was cooled from an initial temperature of 1,000 deg C to 180 deg C in 2.5 sec, an average cooling rate of 328 deg C/sec. From the boiling curve, the transition and nucleate boiling temperature was found around 675 deg C and the maximum heat transfer occurs at 300 deg C. Using various impingement angle from a spray nozzle, one of the studies has concluded that heat transfer sharply increases once the surface temperature passes the Leidenfrost point, which is around 670 deg C although varies from test to test. It has also been reported that heat flux of the nucleate boiling is not a function of the impinging speed and the plate surface temperature.

Summarizing from these results, the stagnation zone does have highest heat transfer effect which depends on the surface temperature and the boiling condition. Heat transfer is relatively small at the onset of water impingement on the plate because film boiling prevails and no audible boiling noise (no active boiling). After a certain time delay which depends on the experiment, heat transfer rapidly raises as the surface temperature reduces to change film boiling to transition and nucleate boiling. Away from the impingement zone, an ‘effective cooling zone’ can be visualized around 20 times to 48 times of the nozzle diameter depending on the jet Reynolds number. Since the wet zone has higher heat transfer because of the cooling medium, it is also called the effective cooling zone. The local and average heat transfer of the effective cooling zone is rreported as a function of Reynolds number and the distance from the stagnation point. For a constant jet speed, heat transfer coefficients in the effective cooling zone are nearly linear to the distance even the regression equations which has been used in the study have higher order terms.

Moving cases – Several studies have also been carried out for the steady state condition with a moving plate. The impact zone length remains the same as the stationary case, but the wet cooling zone length stretches downstream to 19 times of the nozzle diameter and only 3 times upstream, which makes a non-symmetric effective cooling zone. The cooling mechanism appears similar to the stationary case.

The highest heat transfer occurs at the regime of nucleate boiling and the surface temperature and the plate speed has a negligible effect at that state. The heat transfer coefficient is around twice of film boiling regime. However, the nucleate boiling occurs earlier than the plate reaches the stagnation point and causes the peak heat transfer to shift to the direction opposite to the plate motion. Besides, the heat transfer coefficient seems symmetric to the maximum plane and shows a similar trend as observed in the stationary case.

Obviously, the maximum heat transfer location appears to be in the vicinity of the onset point of nucleate boiling instead of the stagnation point. However, both the Lindenfrost temperature and its corresponding location are difficult to formulate theoretically. In fact, the Lindenfrost point and the heat transfer coefficient are highly correlated each other. Normally, they depend on cooling capacity of water flow as well as the initial temperature. In one of the early studies, it has been pointed out that the surface temperature of 538 deg C and 649 deg C represented the nucleate and transition boiling temperature using different initial temperature.

One of the studies has selected an initial temperature of 644 deg C and 108 deg C, and found that nucleate boiling cannot occur in the low initial temperature. Using slightly higher initial temperature of 240 deg C, one of the studies has found the nucleate boiling regime is in the range of 120 deg C to 170 deg C. Since water saturated temperature is 100 deg C, the low initial surface temperature in the impingement zone can be too cool for boiling. A higher initial temperature causes nucleate boiling with a time delay which depends on the water cooling effect.

Operating cases – Controlling coiling temperature in the production mill has to take additional operating conditions into consideration. Since strip thickness, finishing temperature, mill speed, and cooling headers vary during rolling, the mill hardly reaches a steady state condition. Low pressure, high volume cooling headers are activated or deactivated every second with a large response time. The laminar jet speed and its flow rate are in the neighborhood of 1 m/sec (metres per second) and 22 litre/min (lpm) respectively. Since the mill speed is in a range of 2.5 m/sec to 15.3 m/sec, the plate to jet speed ratio of 2.5 to 15.3 is much larger than the experimental cases of less than 0.5. Header cooling effect is considerably reduced because of cooling water interaction if two adjacent headers are open simultaneously.

Because of high volume, water stays on the strip and causes wash back and wash down effect. Water wash back effect can be realized by the reported problems of the X-ray gauge meter error due to water wash back from the first cooling header, although the distance from the first header to the last finishing stand is as far as 15 metres. Similarly, water wash down effect is observable that the residual water is normally blown away by high pressure water (or air) wipers mounted at both sides of the run out table.

Unlike the bottom header, water drops when it lost its momentum to the opposing gravity force, the large quantity of cooling water from the top header cannot escape from the strip surface in a very short time. Except a portion of water evaporates, most of water washes downstream or upstream and gets off the strip from the strip edge. Particularly for a wider strip, the water in the central portion of the strip has more difficulty to escape from the strip edge. Hence, it is very difficult to apply the experimental result to mill operation directly. Alternative approaches adopted in the steel industry is to utilize statistical methods to get empirical equations to calculate the needed headers for cooling temperature control. Because of the feed forward and adaptive control loops, several satisfactory results have been reported. Simple and low maintenance are the advantages of this approach. The disadvantages are difficulty of fine tuning the controller and the frequencies of re-tuning the controller when changing rolling schedules.

Run out cable cooling and strip mechanical properties

For a given steel grade, the most important operating parameters in the run out table operations are (i) finishing mill exit temperature, (ii) coiling temperature, (iii) layout of the active cooling jets, (iv) water flow rate, (v) water temperature, (vi) strip velocity, and (vii) strip thickness. A number of studies have been carried out to improve this critical stage of the strip production.

Controlled run out table cooling has the most important effect on the final mechanical properties of the steel strip during hot rolling in the hot strip mill. The mechanical properties of steel strip are adjustable by effective controlling microstructural evolution such as austenite decomposition and precipitation. Further, the microstructural evolution processes are basically thermal driven, and directly associated with applied cooling water intensity and distribution. Following this line of reasoning, products with prescribed properties can be made if the appropriate water flux and distribution are achieved.

Obviously, a   fundamentally based correlation bridging prescribed mechanical properties and processing parameters is a prerequisite. Basic knowledge of water jet impingement boiling heat transfer and phase transformation is needed to create the above correlations. Over the past half century, extensive theoretical and experimental work on boiling heat transfer and austenite decomposition has been conducted and hence extensive data is available.

Basically, the issue at the run out table is the definition of an expression for the local heat flux during the water jet cooling as a function of the most important operating parameters such as water temperature and flow rate, velocity of the strip, and the local surface temperatures. Other parameters of importance are jet arrangement along the roller table, nozzle shape, dimensions, height, and angle.

In the jet cooling, there are two main fluid zones. These are (i) parallel flow, with and against the direction of moving strip, and (ii) pressure gradient flow, in both strip and counter current directions. Additionally, depending on the jet arrangement, there can be a stagnant zone where two opposite parallel flows meet. Further, as is expected, a function for the heat transfer coefficient for the run out table has to be reliable. However, because of the nature of the two phase heat transfer mechanism and the different flow regimes, it is unlikely that a simple expression can include all important variables, and remain general.

In the production of steel strip for the automotive sector, the mechanical properties are largely defined by the process conditions in the hot strip mill. In particular the phase transformation, from austenite to ferrite, which takes place during the forced cooling stages on the run out table after the finishing mill, governs the microstructure and hence the mechanical properties.

In combination with the chemical composition, the cooling rate in different sections of hot strip mill in the temperature trajectory from 1,200 deg C to coiling temperature, as well as the coiling temperature itself, largely determines the microstructure of the steel strip. An important step in this process is the phase transformation from austenite to ferrite which takes place in the dynamic hot rolling process on the run our table in the temperature range of 600 deg C to 800 deg C. The exact transformation temperature (range) is determined by the cooling rate and chemical composition.

The post rolling cooling serves four main purposes namely (i) controls the microstructural phase formations, (ii) controls the final cross sectional ferrite grain size / distribution, (iii) controls the interphase precipitation strengthening mechanisms, especially important in higher strength titanium carbide precipitation strengthening grades, and (iv) controls the final texture formations which affects toughness.

In regard to microstructural control, the Ar3 temperature needs to be known and then an understanding of the CCT (continuous cooling transformation) kinetics for the given alloy is helpful, so that the correct cooling rate / cooling stop temperature can be chosen. In addition, cooling rate affects final ferrite grain size / distribution, texture formation, and interphase precipitation formation. As with all processing within the steel plant, water temperature / yearly seasonal conditions can play role in the cooling rate and hence the final metallurgy / mechanical properties that can be achieved.

Cooling temperature has an important influence on the microstructure and mechanical properties of rolled strip on run-out table. The factors which affect the strip cooling process on run-out table include strip chemical constituents, thickness, velocity, flatness, cooling water flow, cooling water pressure, and finishing temperature etc., among which the effect of velocity disturbance is the most significant one. As one of the main means, adjusting mill velocity is normally applied to ensure precision control of the finishing temperature for hot strip mill, so the forecast of the speed progression is frequently not realized as a result of the current production conditions. Due to the mass flow, all the strip segments located between the pyrometer of the last finishing stand and that of the coiler are disturbed simultaneously in the event of strip velocity change. The effect of the velocity disturbance for each strip segments differs, depending on temperature / microstructure state, convection, and radiation. It is difficult to control this effect within certain limits for classical run out table cooling control system.

The knowledge of geometrically location-dependent temperature progression within run-out table cooling section has a great value to hot strip cooling process control. But the progression of the strip temperature can only be measured at a few positions within the cooling section due to the difficult ambient conditions. Taking values easily measured, such as finishing temperatures, strip velocity, acceleration rate, and value actual on / off status, etc. as auxiliary variables, online monitor function of cooling temperature on run-out table of hot strip mill can be developed using soft-sensing technique. The control method can be developed to eliminate the effect of velocity disturbance on strip temperature on run-out table.

One of the studies investigated the effect of the transfer phase on the prediction of the accuracy of the strip temperature at the end of roller table. The results were predicted by the model for different steel grades, and the results had shown that high carbon steel had a higher error because of the lack of attention to the latent heat of the transition phase. Another study considered the effect of latent heat and achieved the heat transfer coefficients empirically. One more study considered and achieved sheet temperature profiles for three low-carbon, medium-carbon and carbon rich grades and calculated the effect of latent heat transfer using empirical heat transfer coefficients. This study explored the relationship between metallurgical properties, heat transfer and control of the cooling system in the rolling of steel. The results had shown that the surface movement effect indicates the increase in the cooling efficiency of the sheet and rollers.

One of the studies investigated three high-strength steels at an ultimate temperature of 820 deg C and observed that the ultimate tensile strength was around 50 MPa to 80 MPa at a rapid cooling rate increasing relatively to conventional cooling. Another study has shown that the heat transfer coefficient was constant in the stillness region (23 kW/sqm.K). Regardless of the collision velocity and surface temperature, if the surface temperature exceeds 400 deg C, using a collision angle from a spray nozzle, one study concluded that heat transfer increases rapidly when the surface temperature of the Leidenfrost point is close to 670 deg C.

In this study, the run-out table length is divided into three different zones including inlet zone between the latest finishing stand and the first curtain water jets, curtain water jet cooling zone and outlet zone between the latest curtain water jets and down coiler. At inlet and outlet zones, the strip is cooled in air, while main strip cooling occurs in the curtain water jet cooling zone. The length of the run out cooling table from the outlet of the last finishing mill stand to the first down coiler is around 146.5 m. The hot strip is cooled down by the air in the first and final parts of the area, and in the middle of the laminar cooling zone, the hot strip is cooled down with curtain water on the sheet and spraying water from below the sheet. In the upper part of this area, tanks for reserve cooling water are embedded, which are opened and closed in accordance with the cooling process of the desired strip.

The tanks (reservoirs) above the laminar cooling zone are 15, and each tank has headers for water guidance on the surface of the strip. So there are 13 tanks with four headers with a distance of 1.8 m from each other. The nominal diameter of each header is 200 mm. These heads direct the water toward the water showers, which are kind of curtains, and flow of each header is about 225 cum/h. The distance from the strip to headers of water curtain showers is around 1,695 mm,  the distance between water curtain showers is around 1,800 mm, the distance between the spray nozzles to the surface of the strip is around 266 mm, and the distance between the water spray nozzles is around 900 mm.

The total number of water curtain showers is 52, and bottom water sprays are 104. According to the existing procedure, 26 of 52 water curtain showers and 26 of 104 bottom water sprays are enough for cooling strip from 870 deg C to 630 deg C. The average strip speed is around 3.48 m/s, and it takes around 13.55 s for the strip to reach the end of the run out table from the beginning of the showers to the end of the effective shower curtain (twenty-sixth showers). So, the cooling rate is 17.85 deg C/s.

After rolling in the finishing mill, the strip is cooled by water from a start cooling temperature to target temperature with a given cooling rate before coiled, which promotes obtaining more fine grain size and improving strengthening and toughness. In control cooling process, high precise temperature distribution prediction and cooling path control are key factor for ensuring the mechanical property.

In cooling process, as the strip moves from finishing mill to down coiler, water distribution is to be adjusted with variable rolling velocity to meet target coil temperature. Water cooling process is complexly affected by several factors, mainly including strip temperature, water flow, strip thickness, chemical composition and strip velocity, etc. Once the strip enters cooling zone, strip thickness, width, and composition have little difference in strip length direction, start cooling temperature and velocity are the main factors influencing temperature control. Presently, to increase the productivity and decrease heat loss on delay of transfer bar, finishing rolling process always adopts the variable-velocity rolling strategy in hot rolling strips making, which leads strips velocity varying in a large scale.

Meanwhile, the adjustment of rolling velocity is also used to control finishing mill delivery temperature (FDT, herein also is start cooling temperature), by which the rolling acceleration / deceleration and velocity are dynamically changed to meet the aim FDT and considerably influence coil temperature control accuracy. Under variable velocity rolling, water distribution expressed as open / closed cooling headers and header flow is the only adjustment method to ensure target coil temperature. Hence, precise water distribution calculation based on accuracy velocity perdition is crucial for cooling path and temperature distribution control in cooling process.

To build a high-accuracy temperature distribution prediction and control model, people have been paying attention to related studies. So far, three main kinds of models have been developed and applied. Index model is an important scheme to deal with convective heat transfer, considering convective between cooling water and strip surface as the main heat exchange in cooling process. However, because of simplifications and ignoring temperature gradient in strip thickness, the application effect of index model is not good, especially for heavy thickness strips.

Another kind of model is the statistical model, which has been built relying on statistical regularity. In statistical modelling, by fully considering physical mechanism of cooling process, the main influence parameters of cooling process are considered and determined by statistical regressive methods. Further, to improve the control precision, the influence parameters are classified to categories based on steel grades and thicknesses of strip or other boundary conditions. Then, each category of strips has a series of model parameters. However, the temperature control accuracy of this method gets limited by the number of categories. Besides, some advanced methods are used to improve the control accuracy, such as fuzzy control and model prediction control. The advanced methods in control cooling greatly rely on the statistical model or index model.

The third model is built by finite difference method (FDM) and heat transfer theory, which divides strip and time into nodes and finite number of steps to solve differential equations by approximate derivatives. Then, the spatial and sequential temperature distribution of strip can be obtained. During modeling, for surface node, the heat transfer form is radiation and convection heat transfer, while the heat transfer form is heat conduction for inner nodes. In control cooling process, the temperature gradient in thickness direction is fully considered in FDM, consequently, control accuracy is obviously improved, especially for heavy thickness strip. At present, the most used method is combining with FDM and statistical method, the former is used for temperature prediction model building, and latter is used for seeking optimal model parameters.

Leave a Comment