Method for controlling a beam processing installation, associated system and installation

Abstract

A method for controlling a beam processing installation which comprises a rolling mill, and a cooling machine for water-cooling a steel beam downstream of the rolling mill, the method comprising: - s4: Computing a temperature TST predicted for the beam after it has traversed the cooling machine, computed taking into account an input temperature (Tin) of the steel beam before it enters the cooling machine, and taking into account one or more candidate cooling setpoints (vc*) for the cooling of the beam in the cooling machine, - the step s4 being executed iteratively to determine one or more cooling setpoints (vo*) such that the temperature TST computed in step s4 matches a target self-tempering temperature T*ST,j, - sc : controlling the beam processing installation based on said one or more cooling setpoints, or based on one or more corrected cooling setpoints (vj*) determined from the one or more setpoints.

Description

Method for controlling a beam processing installation, associated system and installation

[0001] The technical field is that of producing steel beams, for instance steel beams for buildings, bridges and other structures. It concerns more particularly the control of heat treatments applied to such beams during a production process.Technical background

[0002] For obtaining high performance steel beams (typically big beams called “jumbo” intended for instance for building skyscrapers), it is known, after the hot-rolling operations, to cool down the steel beam quickly using water jets, but in a well-controlled manner. Figure 1 schematically represents the evolution, with time t, of the temperature T at the surface of a steel beam, during such a water-cooling process. First (phase Ph1 in figure 1), the beam is cooled in air. Then it is water cooled quickly in a cooling machine (phase Ph2), and the temperature T drops significantly. And then, the beam ends its cooling in air (phase Ph3). During the water cooling, it is mainly the superficial part of the beam that is cooled down. And so, during the subsequent phase Ph3, the heat that remained in the bulk of the beam is gradually released and the surface temperature T thus temporarily increases after the beam has traversed the cooling machine (it becomes moderately high again) before finally decreasing (see figure 1).

[0003] This re-increase of the surface temperature, usually called « self-tempering », is useful from a metallurgic point of view if well controlled. In particular, it is useful to control the cooling of the beam in order to obtain a “self-tempering” temperature TST (which is the maximum over time t reached by the temperature T of the surface of the beam, after the beam has exited the cooling machine) that is close to a desired, target self-tempering temperature (well suited to optimize the final mechanical properties of the beam).

[0004] Controlling the self-tempering temperature obtained during such a cooling process is challenging due to the thermal inertia in this process (this temperature is reached once the fast-cooling process itself is finished) and due to the rather complex thermal evolution for the beam. In this context, it would thus be very useful to be able to predict accurately a value of the self-tempering temperature TST, or more generally to predict the thermal evolution of the beam during the cooling process. This would be useful for controlling the beam processing installation. This would be useful too for a post-characterization of a beam, after its cooling (in order to estimate mechanical properties of the beam that has been produced, for instance).Summary

[0005] In this context, a method for controlling a beam processing installation, according to claim 1 , is provided.

[0006] This method may comprise one or several additional features, defined in claims 2 to 12, considered alone or in combination.

[0007] The instant technology also concerns an electronic system according to claim 13, and a beam processing installation according to claim 14. The beam processing installation may further comprise a temperature sensor for measuring the input temperature, an additional temperature sensor for measuring the intermediate temperature, and an actuator or a low-level controller for controlling the speed with which the beam moves. The instant technology also concerns a computer program whose execution on a computer (possibly connected to relevant sensors, actuators or controller), makes the computer to execute the method for controlling presented above. The instant technology also concerns a non-transitory computer-readable medium storing such a computer program.Detailed description

[0008] The instant technology will now be described in more detail and illustrated by examples without introducing limitations, with reference to the appended figures.

[0009] Figure 1 schematically represents the evolution over time t of a surface temperature of a steel beam, during a quench and self-tempering thermal treatment of the beam.

[0010] Figure 2 schematically represents a beam processing installation for applying a quench and self-tempering thermal treatment to a beam.

[0011] Figure 3 is a schematical perspective view of a part of an H-shape beam, to be processed in the beam processing installation of figure 2.

[0012] Figure 4 schematically represents flange water jets and web water jets cooling the beam, in a cooling machine of the beam processing installation of figure 2.

[0013] Figure 5 schematically represents a longitudinal sampling of the beam into consecutive portions, for the sake of thermal computations and control.

[0014] Figure 6 schematically represents, as a block diagram, steps of a method for controlling the beam processing installation of figure 2.

[0015] Figure 7 schematically represents an initially homogeneous temperature field, and a temperature field at a subsequent instant, computed during an initialization step of the method of figure 6.

[0016] Figure 8 schematically represents an intermediate temperature of the beam, for the different portions of the beam, the intermediate temperature being measured just before the beam comes out of a finishing mill of the beam processing installation.

[0017] Figure 9 schematically represents a difference between a measured temperature and a computed temperature, for a plurality of beams and based on two different methods.

[0018] Figure 10 schematically represents a cross section of the beam of figure 3.

[0019] Figure 11 schematically represents a top view of cooling boxes of cooling sections and subsections of a cooling machine of the beam processing installation of figure 2.Seam process / ng / nsta / totfon

[0020] Figure 2 schematically represents a beam processing installation 1 which comprises a hot rolling mill to shape a beam 2, a cooling machine 30 for water-cooling the beam once it exists the rolling mill, and an electronic system 10 configured, inter-alia, for determining one or more cooling setpoints for the beam water-cooling. Here, the rolling mill comprises a roughing mill (not represented in figure 2), then an intermediate mill (also called main mill) which is a tandem mill 4, here, and then a finishing mill 3. The beam 2 is hot when processed in the rolling mill: its temperature is typically higher than 800°C, or even higher than 900 °C. In the beam processing installation 1 , the beam is moved parallel to its axis, along the installation axis x, from the intermediate mill to the finishing mill and then to the cooling machine. The cooling machine 30 is arranged downstream of the rolling mill, along the path x followed by the beam 2 in the beam processing installation. More particularly, the cooling machine is arranged downstream of the last rolling stand of the rolling mill, that is, the cooling machine is arranged downstream of the finishing mill 3 here. There is no other rolling stand between the last stand of the rolling mill and the cooling machine 30.

[0021] The distance between a main mill output and the finishing mill input may, like here, be higher than the length of the beam 2 (in other words, the maximum length for the beams processed in the installation is equal or smaller than this distance, here). The cooling machine 30 is located immediately downstream of the finishing mill.

[0022] The beam processing installation 1 is equipped also with: a first temperature sensor 6 arranged to measure a surface temperature of the beam 2 at an upstream position Oe located upstream of the cooling machine 30. Here, the upstream position Oe is located upstream of the intermediary mill (that is, upstream of the tandem mill 4); a second temperature sensor 8 arranged to measure another surface temperature of the beam 2, at an intermediary position Os located between the upstream position Oe and the cooling machine; here, the intermediary position Os is located between the intermediary mill (here, the tandem mill 4) and the finishing mill 3, and a third temperature sensor 9 arranged to measure another surface temperature of the beam 2 after the beam has traversed the cooling machine 30, this temperature being referred to as the output temperature Tout.

[0023] The temperature sensors may be pyrometers, or infrared cameras, for instance.

[0024] The beam 2 is a steel beam, or in other words a steel profile (or steel section). Its length may be from 10 to 200 meters at this stage of the manufacturing process (it may be subsequently cut to shorter lengths). The width of the beam (maximum transverse extension)would be from 10cm to 2m. In the detailed example presented here, the beam 2 is an H-beam (its section has an H shape), as represented in figure 3. Yet, the beam could have a different section shape. For instance, it may be an I-beam, a Z-beam, a T-beam, a II beam (that is forming a U-shape channel) or an L beam (that is, a steel angle).

[0025] For the beam 2, thermal exchanges, by radiation and due to the contact with air or water, occur at the beam surface, also designated as the beam outer surface (this surface is accessible for cooling the beam, using waterjets for instance). As represented in figure 10, the beam surface is mainly composed by (i.e.: a major part, or even 80% of this surface is composed by the following): a first web face W1 and a second web face W2 opposite each other, and two or three (depending on the shape of the section of the beam) flange faces for each flange; here three flanges faces F1 , F2, F3 for the first flange 22 and three other flange faces F4, F5, F6 for the second flange 23 (F1 and F4 are outer flange faces, both F2 and F5 are inner and upper flange faces and both F3 and F6 are inner and lower flange faces).

[0026] The web and flange faces F1 - F6, W1 , W2 are also designated as the beam faces, or simply as the faces in the following. Each of these faces is one of the main faces of the web or of the flange considered (each web face is parallel to the web, and each flange face is parallel to the flange). In other words, these faces are not the small end faces of the beam (at the head and at the tail of the beam), nor the thin edge faces extending along the longitudinal edges of the flanges (such as the upper edge face that joins the flange faces F1 and F2). It is noted that one or more of the edge faces may still be water-cooled.

[0027] In the embodiment described here, the web faces W1 , W2 are horizontal while the flange faces F1 - F6 are vertical. W1 is the upper web face (it is upward facing) while W2 is the lower web face (downward facing).

[0028] The cooling machine may comprise one, two, or more successive cooling sections. Each cooling section is arranged to be traversed by the beam, and for emitting water jets that strike some or all of the beam faces. Here it comprises a first cooling section 31 and a second cooling section 32, located one after the other along the installation axis x (figure 2). The different cooling sections may be formed by distinct devices (cooling boxes), or may correspond to different, successive longitudinal portions of a same device.

[0029] The cooling machine is arranged so that the jets belonging to one of the cooling sections can be controlled independently of the jets of the other(s) cooling section(s). At least one, here each, cooling section is arranged for emitting both (figure 4): flange jets FJ, for projecting water on one or more of the flange faces F1 , F2, F3, F4, F5, F6, and- web jets WJ, for projecting water on one or both the web faces W1 , W2.

[0030] In this document, when water is projected on a face by one of the jets, it means that this water is projected directly on this face by the jet considered, which strikes this face (with a jet impingement). In the following, the flange jets and the web jets may be indifferently (and possibly collectively) referred to as the jets.

[0031] The cooling machine is arranged so that the web jets WJ can be controlled independently of the flange jets FJ.

[0032] More generally, the cooling machine is arranged so that multiple groups of jets (or even all the jets), possibly positioned on multiple cooling boxes, can be controlled independently of each other (using electrically controllable valves). Here, in particular, the first and second cooling sections 31 and 32 each comprises two successive subsections (311 , 312 and 321 , 322 respectively - figure 11), the web jets of which being controllable independently of those of the other subsection. For instance, for the part of the beam that is in the first cooling section 31 , for a first half of this part of the beam, the web may be water cooled while for a second half of this part of the beam, the web is not water cooled (no jet striking it).

[0033] The cooling machine arranged as above described enables multiple different cooling configurations to be used for cooling the beam.

[0034] For instance, the flanges 22, 23 may be water-cooled but not the web 21 , both in the first cooling section 31 and in the second cooling section 32. Or both the flanges and the web may be water-cooled, in both cooling sections. Or, in the first cooling section 31 , the flanges 22, 23 may be water-cooled but not the web 21 , while in the second cooling section both the flanges and the web are water-cooled.

[0035] Figure 11 represents a top view of the cooling boxes of the cooling machine. The central boxes are the web cooling boxes that comprise the web jets for the first web face W1. Adjacent to the web cooling boxes are the inner flange cooling boxes that comprise the flange jets for the flange faces adjoining laterally the first web face, i.e. F2 and F5. On the outside of the inner flange cooling boxes are the outer flange cooling boxes that comprise the flange jets for the outer flange faces, i.e. F1 and F4. In this figure, cooling boxes that are on are hatched while the cooling boxes that are off are in white. A figure representing the bottom view of the cooling boxes of the cooling machine would be identical to Figure 11. So, in the cooling configuration represented in figure 11 : the jets of the first cooling section 31 are off (both the flange jets and the web jets), the flange jets FJ of the second cooling section 32 are on, the web jets of the first subsection 321 of the second cooling section 32 are off, and the web jets of the second subsection 322 of the second cooling section 32 are on.

[0036] Being able to use multiple different cooling configurations is very useful. Indeed, some geometrical imperfections of the beam exiting the rolling mill, such as squareness defects, flatness defects or evenness defects can be corrected using dedicated cooling configurations(for which not all the jets are on). Besides, the fact that the jets of the second cooling section 32 are controllable independently of those of the first cooling section 31 gives more flexibility, regarding the choice of the thermal route followed by the beam (e.g.: progressive cooling, or more abrupt cooling), and thus more flexibility in terms of final mechanical properties for the beam.

[0037] The beam processing installation 1 comprises a low-level controller 13 for controlling actuators of the beam processing installation 1 , in particular for controlling valves of the cooling machine 30, for controlling the finishing mill 3, whose rolling speed fixes the speed at which the beam traverses the cooling machine 30 and possibly for controlling the transfer rolls, whose speed can fix the speed at which the tail of the beam traverses the cooling machine once the beam has fully exited the finishing mill. The low-level controller 13 is also connected to sensors of the beam processing installation 1 (in particular one or more speed sensors and presence sensors for detecting the presence of the beam at different positions along the processing path). It may implement control loops such as PID loops. It receives the cooling setpoints from the higher-level electronic system 10. In particular, it receives from the electronic system 10 a speed setpoint, instructions specifying which of the above-mentioned jets are to be on and which are to be off, and one or more water flow setpoints for the jets that are to be on. The low-level controller 13 is configured to implement control strategies close to the actuator control, yet less elaborate control strategies than the electronic system 10 (for instance in the form of a proportional, integral, and possibly derivative control loop). The low-level controller 13 may take the form of a programmable logic controller (PLC) or other industrial-like computer or programmable circuit. It is operatively connected to the electronic system 10 and to the above-mentioned actuators and sensors (for instance through a local data bus).

[0038] Here, the electronic system 10 comprises a server 11 and a terminal 12. The terminal may be a human-machine interface (such as a display screen, possibly with an entering device like a keyboard and / or a pointing device like a mouse pad and / or buttons). The terminal may also be a standalone computer. Alternatively, the electronic system 10 may take the form of a stand-alone computer (including a human-machine interface), rather than a server with a distinct, possibly remote human-machine interface. Anyhow, the electronic system 10 comprises at least a processor and a memory and can be programmed so as to execute elaborate tasks like numerical simulations. The electronic system 10 is operatively connected (through a local network, or through the internet, for instance) to an industrial database 14, itself connected to remote services 15 or users, such as production analytics services or supply chain tracking services.

[0039] The electronic system 10 (more precisely the server 11 , here) is programmed to execute the method for controlling the beam cooling presented below.

[0040] This method is based on a numerical simulation of the temporal evolution of the temperature field within the beam. Executing this simulation iteratively enables to determine the cooling setpoint(s), here the speed setpoint, which is transmitted to the low-level controller 13 which then controls the actuator(s) of the installation 1 based on this setpoint(s). The cooling setpoint(s) is determined such that, for this setpoint, a predicted self-tempering temperature TST (predicted by the numerical simulation) matches a target self-tempering temperature T*STJ, to be obtained. By matching it is meant being equal to, within 20°C, or even within 10°C or within 5°C.

[0041] The numerical simulation is a finite differences computation. Regarding spatial variables (in addition to the temporal variable t), it may, like here, be a two-dimensional simulation, the two spatial dimensions corresponding to two coordinates along two transverse directions each perpendicular to the beam axis. It can take into account the beam dimensions and chemical composition.

[0042] The numerical simulation takes also into account at least one temperature of the beam 2 measured before the beam 2 enters the cooling machine 30, referred to as the input temperature Tm. This temperature is measured preferably before the last rolling pass in the main mill, for instance just after the last-but-one pass in the tandem mill 4, more preferably when the beam passes at the upstream position Oe before the last rolling pass in the main mill, for instance just after the last-but-one pass in the tandem mill 4. The input temperature Tm is preferably measured on a heading part of the beam, this heading part corresponding for instance to a first fourth of the beam.

[0043] The numerical simulation can also take into account a temperature of the beam measured downstream of the main mill and before the beam 2 enters the cooling machine 30, referred to as the intermediate temperature Tint. This temperature is measured preferably when the beam passes at the intermediary position Os. This intermediate temperature can help adjusting the cooling setpoint just before the beam enters the cooling machine 30.

[0044] In terms of discretization, here, the beam is divided into a number n of consecutive longitudinal portions Pj, with i from 1 to n (figure 5), also called samples. The portion Pi is the head portion of the beam (located at its heading end, that is at the end of the beam that comes out of the finishing mill first). Pnis the tail portion (in other words end portion) of the beam, n may be from 3 to 100, or even from 3 to 50, and each sample may be from 0.2 to 5 meters long, or even from 1 to 4 meters long.

[0045] In practice, a two-dimensional temperature field is computed for at least one longitudinal sample, and its evolution over time t is computed, for a period encompassing at least the transit through the cooling machine 30 and a subsequent air-cooling. Here, this period spans at least from the measurement of the input temperature to a delivery of the beam at the end of the beam processing installation 1.

[0046] Determining the cooling setpoint(s) by iterative numerical simulations (which requires significant computation time), while it is to be used for real-time on-line control, is achieved thanks to specific arrangements that are presented below, with reference to figure 6.

[0047] Some general aspects of this method are presented first and the method is then presented in more details. / n-advance determfnat / on of one or more coo / mg setpo / nts; genera / aspects

[0048] Remarkably, the method for controlling the beam processing installation comprises (figure 6): s1 : acquiring the input temperature T of the steel beam 2, measured on a portion of beam Pj, measured before said portion of the beam enters the cooling machine, s4: computing the temperature TST of said portion of the beam, which is a temperature predicted for the portion of beam after the portion of the beam has traversed the cooling machine 30, the temperature TST being computed taking into account the input temperature T measured and one or more candidate cooling setpoints to be employed for controlling the cooling of the beam in the cooling machine, the step s4 being executed iteratively to determine one or more cooling setpoints such that the temperature TST computed in step s4 is equal or substantially equal to a target self-tempering temperature T*STJ, sc: controlling the beam processing installation 1 based on said one or more cooling setpoints, or based on one or more corrected cooling setpoints determined from the one or more setpoints.

[0049] In step s1 , in the exemplary embodiment described here, the input temperature Tm of the portion Pj is measured directly, by the first temperature sensor 6, when the portion Pj passes at the upstream position Oe above mentioned. Still, in other embodiments, this measure could be a non-direct measurement, the input temperature Tm of the portion Pj being then derived from another measurement, for instance derived from a temperature measurement carried out on another portion of the beam.

[0050] The portion here denoted Pj is the portion of the beam on which the input temperature Tm is measured. Pj may, like here, be located on a heading part of the beam, for instance on the first fourth of the beam (that is between the head of the beam and a position that is at I / 4 from the head of the beam, I being the length of beam). Pj may more particularly be a portion of the beam located at a distance comprised between I / 20 and I / 4 of the heading end of the beam.

[0051] In this embodiment, the portion Pj for which the input temperature Tm is measured, and for which the iterative simulations are executed, is P4 (that is the fourth portion, when starting from the beam head), the beam being divided into 32 portions (n=32).

[0052] Determining the cooling setpoint(s) for a portion that is, like here, located in a first, heading zone of the beam is beneficial. Indeed, it allows to have more time for computing the one or more setpoints, as this computation time is limited by the time lapse between measuring the input temperature for the portion Pj considered, and having the head portion of the beam, Pi , entering the finishing mill. Besides, having the portion Pj that is not completely at the head of the beam (that is not Pi) is beneficial, as it allows avoiding boundaries effects (computation simulation being possibly less accurate for the head portion Pi than for P4, for instance).

[0053] In step s4, the temperature TST is predicted by numerical simulation of the temporal evolution of the temperature field within the beam. This temperature is the maximum over time reached by the temperature of the surface of the portion Pj of the beam, after this portion of the beam has exited the cooling machine (as illustrated in figure 1).

[0054] In the exemplary embodiment described here in detail, the one or more setpoints, determined by executing repeatedly step s4, is a speed setpoint v0*, for the speed at which the beam 2 traverses the cooling machine 30. Still, in other embodiments, alternatively or in complement to the speed setpoint, one or more other cooling setpoints, like one or more water flow values for the water projected on the beam in the cooling machine, could be determined, by iterating step s4.

[0055] In the exemplary embodiment described here, the one or more candidate cooling setpoints of step s4 is a candidate speed setpoint vc*. Similarly, the one or more corrected cooling setpoints of step scis a corrected speed setpoint v*.

[0056] In this control method, the first execution of step s4 is launched immediately or almost immediately after the intput temperature Tm has been measured (for instance within 1s or less). Anyhow, the first execution of step s4 is launched before the portion Pj of the beam enters the cooling machine, so that there is some time, for determining the speed setpoint v0* to be employed for controlling the speed at which the beam (more specifically at which the portion Pj) traverses the cooling machine, before this portion enters the cooling machine.

[0057] In this regard, as the input temperature Tm (which is employed for determining v0*) is measured preferably at the upstream position Oe, which is located upstream (and even significantly upstream) of the cooling machine, there is some time for achieving the determination of v0* before the portion of the beam considered enters the cooling machine.

[0058] Preferably, the input temperature Tm is measured before the beam comes out of the tandem mill 4. Thanks to that arrangement, there is a substantial time lapse for determining the speed setpoint v0*, before the beam enters the finishing mill, and then the cooling machine. More particularly, here, the input temperature Tm of the portion Pj of the beam is measured before this portion starts the last rolling pass in the tandem mill 4 (and so, the time lapse available for determining v0*, or a corrected version thereof is at least: the time taken for two tandem passes, plus transit from the tandem mill to the finishing mill). In practice, Tm may bemeasured, like here, before the last rolling pass in the tandem mill, and just after the last-but- one rolling pass in the tandem mill.

[0059] In this respect, the upstream position Oe may be located, along the path x followed by the beam 2 and / or the computations achieved in step s4 may be parametrized, so that: a minimum transit time (or, alternatively, an average transit time) transit, between passing at the upstream position Oe and entering the cooling machine (or even between passing at position Oe and then entering the finishing mill), for the beam, is higher than a maximum computation time (or, alternatively, an average computation time) tcomput required for executing step s4 a given number of times (for instance 10 times) using the electronic system 10, multiplied by a safety coefficient which is for instance from 1 to 5.

[0060] For instance, if the minimum transit time observed ttransit is 30 seconds (given the position of the first temperature sensor), then, the computations achieved in step s4 may be parametrized (in terms of mesh finesse and temporal sampling, for instance) so that 10 executions of step s4 using the electronic system 10 requires a maximum time of 15s (safety coefficient of 2).

[0061] The measurement of the input temperature Tm of the portion Pj, and the consecutive (immediate) triggering of the first execution of step s4, may also be carried out at an instant, during the processing of the beam, such that the delay between: this instant, and the entering of the portion Pj in the finishing mill, or alternatively, the entering of the portion Pj in the cooling machine, or alternatively the passage of the portion Pj at the intermediary position Os, is above a given duration threshold, for instance above 10, 20 or even 30s.

[0062] Starting the determination of the speed setpoint in advance, as above explained, is beneficial in terms of available computation time for determining the setpoints. However, it requires to compute not only the evolution of the temperature during the water-cooling step, but also during the transfer of the beam, and possibly during rolling passes (with thermal exchanges with the working rolls, and heating caused possibly by friction and plastic strain), which makes the computation more complex.

[0063] Besides, the accuracy of the prediction for the self-tempering temperature TST may be reduced due to the fact that the thermal evolution is computed for a time period longer than just the cooling operation.

[0064] To overcome this drawback, in the method for controlling the beam processing installation, a corrected speed setpoint Vj* may be determined by:acquiring an intermediate temperature Tint, j, measured on the portion Pj of the beam considered (or, possibly, on another of the portions Pj) when this portion (or, possibly, said other portion Pj) of the beam passes at the intermediary position Os (in practice, measured by the second temperature sensor 8), computing the corrected speed setpoint v* (or, possibly, v*) , based at least on the speed setpoint v0*, on the intermediate temperature Trntj (or, possibly, Tint ), and on a predicted intermediate temperature Tint.caic, predicted for the portion Pj of the beam when it passes at the intermediary position.

[0065] Here, the intermediary position Os, for this re-adjustment temperature measurement, is located between the intermediary mill 4, and the finishing mill 3.

[0066] More generally, the intermediary position Os is located between the upstream position Os and the cooling machine 30, more particularly between the upstream position Os and the last rolling stand of the rolling mill (that is between the upstream position Os and the finishing mill 3, here). The intermediary position Os is for instance positioned 3 to 20 m upstream from the entry of the cooling machine.

[0067] The predicted intermediate temperature Tint.caic is computed for instance in the course of one of the executions of step s4, during which the thermal evolution of the portion Pj is computed (in practice, during the last execution of step s4, or during an execution achieved for a value of the speed of the beam that equals v0*).

[0068] Using the measured intermediate temperature Tintj, and its predicted counterpart Tint,caic allows for readjusting the speed setpoint, should the predicted temperature Tint.caic departs from the measured one Tintj. In particular, if Tint.caic equals Tintj, the speed setpoint may be left uncorrected, while if Tint.caic departs from Tintj, the speed setpoint is corrected, with a correction all the higher than Tint.caic differs from Tintj.

[0069] This re-adjustment technique allows for overcoming the partial limitation mentioned above, relative to a possible reduction of the accuracy for the prediction of the temperature TST due to the in-advance triggering of the computation of the thermal evolution of the beam (i.e.: computation triggered while the beam is still substantially upstream of the cooling machine) and due to the fact that the actual value of the time for transferring the beam from the tandem mill to the entrance of the cooling machine may differs from the value of this transfer time employed for computing (predicting) the thermal evolution of the beam. / n-advonce determ / nat / on of one or more coo / tng setpo / nts; deta / 7ed arrangements

[0070] As represented in figure 6, the control method comprises here, before the repetitive executions of step s4: a step s’2 of computing an initial temperature field for the portion Pj, across a section of the beam, this temperature field being such that the surface temperature of the beam equals the measured input temperature Tm (figure 7),a step s2 of determining the cooling regimes undergone by different faces of the beam, when water-cooled in the cooling machine 30.

[0071] Step s’2

[0072] In step s’2, a numerical simulation of a temporal evolution of a temperature field within a section of the beam is achieved, starting from a virtual initial state for which the temperature within the section is uniform (figure 7, left part), this initial temperature To being above 1000°C, or even above 1100°C, for instance equal to 1200 °C. The temporal evolution of the temperature field is then computed, by finite differences (for a 2-dimensional discrete representation of the section of the beam), for heat exchanges by air and radiation cooling at the surface of the beam. Figure 7 shows the gradual decrease of the surface temperature TSUrf of the beam during this (virtual) cooling.

[0073] The temperature field, at the time step for which the computed surface temperature Tsurf equals the measured input temperature Tm (figure 7, centre and right diagrams), is selected, and is the starting point (the initial state) for the computation of the temporal evolution of the temperature within the portion Pj achieved in step s4 (that is for predicting the temporal evolution of the temperature of the portion Pj after this portion has passes the upstream position Oe).

[0074] This procedure enables to start the computation of step s4 from a temperature field that is more realistic and in adequation with the measured input temperature Tm, while a temperature measurement is available only for a surface temperature (not for bulk temperatures, within the beam section).

[0075] step s2

[0076] In step s2, for the different flange and web faces of the beam, a cooling regime for the face considered is determined based on the cooling configuration data which specify which jets of the cooling machine are to be on and which jets are to be off, during the cooling of the beam 2 in the cooling machine.

[0077] The cooling conditions to be applied in the cooling machine 30, i.e. the cooling configurations as well as the corresponding water flow values, may, like here, be determined by a dedicated module (eg: a dedicated sub-program or sub-routine, or a dedicated electronic module), depending on the metallurgical heat treatment aimed at and / or depending on flatness / squareness corrections to be applied to the beam. These cooling conditions may also be entered manually by an operator (through the human-machine interface). Anyhow, for determining the applicable cooling regimes, these cooling conditions (in practice, the cooling conditions data) are inputs.

[0078] For each flange and web face of the beam, the cooling regime is selected among, at least: a: direct water cooling,b: powered cooling, c: air cooling, d: run-off water cooling,

[0079] Cooling regime c (air cooling) is applied to the parts of those faces that are located outside the cooling machine 30, and that are not in a rolling stand (not in the course of being laminated by a working roll of the tandem or finishing roll 3, 4). It applies also to the parts of the beam faces that are within the cooling machine 30 but that are not struck by any of the waterjets and for which the cooling regime is neither powered cooling nor run-off water cooling.

[0080] Cooling regime a (direct water cooling) is applied for each of the beam faces W1 , W2, F1 - F6, when the beam face considered is struck (directly) by some of the water jets, in the cooling section or subsection considered.

[0081] Cooling regime b (powered cooling) concerns a beam face not struck by any of the jets, but that has another beam face that adjoins it laterally and that is struck by some of the jets, in the cooling section or subsection considered. As this ‘adjoining’ other face is struck by water jets, the face considered benefits non-directly from this water cooling, and cools down faster than for regular air cooling (or than in the presence of remaining run-off water). By “laterally adjoining, it is meant that the two beam faces in question are adjacent (and with no other beam face between them), in contact with each other along the beam (in contact with each other along a contact line or contact seam that extends parallel to the beam axis, at the faces junction). This cooling regime is for instance the one undergone by the part of the web face W1 and the part of the web face W2 that are in the cooling subsection 321 , in the cooling configuration represented in figure 11 where the flanges are water cooled but not the web.

[0082] Cooling regime d (run-off water cooling) concerns a beam face not struck by any of the jets, in the cooling section or subsection considered, but that has some (run-off, remaining) water on it due to water projected on the beam in one of the other cooling sections or subsections. It concerns more particularly a beam face which is horizontal and upward facing, not struck by any of the jets in the cooling section or subsection considered, but:- with jets projecting water on this face in another cooling section or subsection (located upstream or downstream of the section considered), or- with jets projecting water, in another cooling section or subsection, on a beam face which adjoins the face considered and which extends upward from the face considered.

[0083] The run-off cooling regime is the one applicable in the first cooling section 31 , for the upper web face W1 of the beam, in the cooling configuration of figure 11 , for instance.

[0084] If a beam face, not struck by any of the waterjets, has run-off water on it in the cooling section or subsection considered, but has another beam face that adjoins it laterally and thatis struck by some of the jets in this cooling section or subsection, then the selected cooling regime is the powered cooling regime, not run-off water cooling regime.

[0085] The cooling configuration of figure 11 is an example of a cooling configuration for which the four cooling regimes above mentioned, a, b, c and d, are present. More specifically, when the cooling configuration of figure 11 is employed, the cooling regimes for the different beam faces are the followings:For the flange faces F1 - F6, in the first cooling section 31 , the cooling regime is regime c: air cooling,For the flange faces F1 - F6, in the second cooling section 32, the cooling regime is a: direct water cooling,For the upper web face W1 , and also for the lower web face W2, in the second subsection 322 of the second cooling section 32, the cooling regime is a: direct water cooling,For the upper web face W1 , and also for the lower web face W2, in the first subsection 321 of the second cooling section 32, the cooling regime is b: powered cooling. Indeed, this part of the web face W1 , respectively W2, is not struck directly by any of the jets, in the subsection 321 , but it has other faces (namely the flange faces F2 and F5, respectively F3 and F6), that adjoin it laterally, and that are struck by waterjets in the subsection 321.For the upper web face W1 , in the first cooling section 31 , the cooling regime is d: run-off water cooling, as above mentioned,For the lower web face W2, in the first cooling section 31 , the cooling regime is c: air cooling.

[0086] The way to determine the applicable cooling regimes has been illustrated above in the case of an H-beam. This method can be applied as well to other beam shapes. For instance, for T-beam, if the flange faces undergo direct water cooling, but the web faces have no jet striking them, it will be determined that the web faces undergo the powered cooling regime above described.

[0087] In step s2, the cooling configuration data is predetermined (prior to the method described here), and remain unchanged during the execution of the method for controlling the cooling of the beam.

[0088] Step s’2 and s2 may be executed in advance, before measuring the input temperature Tm of the portion Pj of the beam (in step s1).

[0089] Repetitive executions of step s4

[0090] Immediately after step s1 , a first execution of step s4 is achieved.

[0091] During step s4, the temporal evolution of the thermal field within the portion Pj, is computed, for a temporal period extending from the instant T is measured to a final deliveryof the beam after its cooling in the cooling machine. The maximum, reached over time t, by the surface temperature of the portion Pj, which is TST, is thus computed. This temporal evolution is computed for a speed, at which the portion Pj traverses the cooling machine 30, which equals a candidate speed setpoint vc*. The speed at which the beam traverses the cooling machine is a key parameter of the cooling process. Indeed, the longer the beam stays in the cooling machine, the higher the temperature decrease (for a given water flow). It is the control parameter employed here for controlling the cooling process, so as to obtain a desired selftempering temperature for the beam portion Pj.

[0092] The candidate speed setpoint vc*is modified from one execution of step s4 to the other, until the predicted TST temperature matches a target self-tempering temperature T*STJ, for the portion Pj, which is tested in step s5. The candidate speed setpoint can be, for the first iteration, the minimum linear speed that can be given to the beam by the finishing mill and, for the second iteration, the maximum linear speed that can be given to the beam by the finishing mill. These minimum and maximum linear speeds are known for a given installation. In this embodiment, only the candidate speed setpoint is modified from one execution of step s4 to the other. In other words, the cooling configuration and the water flow values remain unchanged, during the different executions of step s4.

[0093] The cooling configuration, and the water flow values, taken into account in step s4, are for instance preset prior to the executions of step s4: based on flatness and / or squareness corrections to be applied to the beam, and / or based on one or more metallurgical criteria, relative to the heat treatment applied to the beam during its quenching and self-tempering in the cooling machine 30.

[0094] When executing step s4, to compute the temporal evolution of the temperature field within the portion Pj, the cooling powers (for instance in Watts per square meter), corresponding to the heat exchanges at the surface of the beam, are computed for the different phases of the beam processing. For a transfer from one position to another, in air, a cooling power corresponding to air cooling and radiation cooling is computed. When rolled in the tandem mill 3 or in the finishing mil, specific thermal models for that rolling operation are employed, for the heat exchanges. And when in the cooling machine, for each surface of the beam, the cooling power is computed depending on the cooling regime undergone by the face considered (as determined in step s2), and, when applicable, depending on the water flow values for the relevant waterjets.

[0095] In particular, for the parts of the faces to which the cooling regime c (air cooling) is applied, the heat exchanges are mostly achieved by thermal radiation and due to the contactwith air. For these parts of the beam faces, the cooling power per unit surface (for instance per square meter), noted Pc, is computed as

[0096] Pradiative is computed for instance according to the following expression:with F the form-factor, s the emissivity of the beam surface, o the Stefan-Boltzmann constant, Tsurf the temperature at the surface of the beam, for the face of the beam considered, and Tajrthe temperature of air, in the beam processing installation.

[0097] Pairmay be computed according to the following expression:Pair = HTCair. (Tsurf- Tair) (eqn - 3)With(eqn — 4)Where C1 and C2 are two constants, where the exponent p is from 2 to 5, and where v is the speed of the beam 2 (speed with which the beam is moved, along the axis x of the beam processing installation 1). In eqn-4, the term vpreflects the influence of forced-convection cooling, while the term depending on Tsurf Ta irreflects free convection cooling.

[0098] For the parts of the faces to which the cooling regime a (direct water cooling) is applied, the cooling power per unit surface, noted Pa, depends on the water flow rate Q projected on the beam face considered. More particularly, Pamay be computed as being equal to, or substantially equal to (which means, here, equal within 10%, or even within 5% or 2%) to:where:C is a multiplicative coefficient; C may have a fixed value, or it may be the sum of a mean value and of a corrective term that varies depending on a temperature Twater of the projected water,Q is the water flow projected on the part of the face of the beam considered, expressed as a volume per unit time and per unit surface (eg: liters per second and per square meter - alternatively, Q could be expressed as a mass of water per unit time and per unit surface, in kg / (m2.s), for instance), the exponent a is a constant from 0.5 to 4, or even from 0.5 to 2, and / 7TCref(TSurf) is a function of the temperature TSUrf.

[0099] In practice, the function / 7TCref(TSurf) decreases with TSUrf, at least when TSUrf is above 200°C. HTCref may be: divided by 2 or 2.5 when TSUrf varies from 200°C to 400°C, divided by 2 when TSUrf varies from 400°C to 600°C, and when TSUrf varies from 600°C to 800°C, and divided by 3 or 4 when TSUrf varies from 800°C to 1000°C. HTCret may more particularlyvary according to the curve represented in figure 11 , where HTCret is represented in arbitrary units against TSUrf (expressed in °C).

[0100] The decrease of HTC with TSUrf enables to reflect the fact that cooling is more efficient, with a higher heat transfer coefficient, when TSUrf is moderately high (in the 400-700°C range) rather than very high (above 800°C). Taking this variation into account enables excessive cooling to be avoided, and increases the temperature prediction (and control) accuracy.

[0101] For the parts of the faces to which the cooling regime b (powered cooling) is applied, the cooling power, per unit surface, Pb, is computed as being equal, or substantially equal to: the cooling power Pafor the adjoining face in question (which undergoes direct water cooling) multiplied by a proportionality coefficient. This proportionality coefficient may in particular be all the smaller than the beam face undergoing powered cooling is wide. The cooling power Pb, per unit surface, may then be computed as follows: woPb= — Paeqn - 6) w where w is the width of the face in question and w0is a constant width.

[0102] w is the extension of that face transversely, perpendicularly to the beam axis (perpendicularly to axis x, here). For instance, for the beam 2 of figure 10, if part of the web face 1 undergoes powered cooling, then the width w used in eqn-6 is the width of the web, that is the distance between the two flange faces F2 and F5. The constant w0has a value that is representative of a typical transverse dimension of the beam. It is for instance from 0.1 to 0.6 meter (when w is expressed in meter).

[0103] Eqn-6 reflects (inter alia) that when the web is wide, the adjoining flange faces, directly water cooled, are distant from the middle of the web which makes powered cooling less efficient for the central part of the web (which is reflected by the decrease of Pb with w).

[0104] For the parts of the faces to which the cooling regime d (run-off water cooling) is applied, the cooling power Pd, per unit surface, may be computed according to:where HTCd is a constant heat transfer coefficient.

[0105] HTCd is for instance from 700 W / (m2.°C) to 1500 W / (m2.°C) or even from 800 W / (m2.°C) to 1100 W / (m2.°C).

[0106] When computing the temporal evolution of the temperature field within the portion Pj, the fact that phase transformations (like austenite to ferrite, to bainite or to martensite transformations) may occur in the steel during its cooling can also be taken into account.

[0107] Once a candidate speed point, for which TST matches T*STJ, has been found, the speed setpoint v0* is set to equal this candidate speed point and the repetitions of step s4 stop. The repetitions of step s4 also stop if the total number of executions of step s4 reaches amaximum number of repetition allowed (equal to 10, here). In this case, the speed setpoint v0* equals the candidate speed setpoint for which the difference between TST matches T*STJ was the smallest.

[0108] The determination of the speed setpoint v0* (by repeatedly executing step s4), which is time consuming, is achieved just once, here, for one longitudinal sample of the beam only, namely for the portion Pj. It is worth noting here that the discretization of the beam into n longitudinal samples plays no significant role during the steps s1 to s5, and more generally during the simulations made to determine values for the one or more cooling setpoints, here the speed setpoint v0*.

[0109] From the results of the executions of step s4, two sensitivity coefficients KT and KCT are also determined. KT and KCT reflect, respectively, how much the temperature TST is modified when the temperature of the beam portion at the intermediary position Os varies, and how much the temperature TST varies when a transit time tc, taken for traversing the cooling machine, varies.

[0110] The sensitivity coefficient KCT is computed according to the following formula:where A(Tsr)tcis a variation of the TST temperature caused by a variation Atcof the transit time tc, while the intermediate temperature Tmt.caic (as computed when executing step s4) is unchanged.

[0111] KCT can be computed directly from the results of the executions of step s4. Indeed, these executions correspond all to a same value of Tjnt.caic, but correspond to different values of the transit time tc, as they are executed for different values of the speed v at which the beam portion traverses the cooling machine. Indeed, the transit time tcequals L / v where L is the length of the cooling machine (length to be travelled to traverse the cooling machine, which is for instance from 3 to 15 m). Still, KCT may also be computed thanks to an additional numerical simulation of the thermal evolution of the beam achieved for a modified value of tc.

[0112] The sensitivity coefficient KT is computed according to the following formula:where (TST)Tintis a variation of the TST temperature corresponding to a variation Tint calcof the intermediate temperature Tjnt.caic, while the transit time tcis unchanged.

[0113] For computing of the coefficient KT, an additional computation of the thermal evolution of the temperature field for the portion Pj, , may be required, this additional computation being achieved for a higher value of the time for transferring the beam from the tandem mill to the finishing mill, in order to obtain a value of the calculated intermediate temperature Tjnt,caicdifferent than the one obtained from previous numerical simulations of the thermal evolution of the beam.

[0114] Here, KT and KCT are both positive (computed taking absolute values of the variations of the quantities considered).

[0115] Adjustment of the speed setpoint

[0116] The speed setpoint v0*, and the sensitivity coefficients KT and KCT are determined, as above explained, while the portion Pj of the beam travels, in the beam processing installation, from the upstream position Oe to the intermediary position Os.

[0117] Then, in step 6, when the portion Pj passes at the intermediary position Os, its intermediate temperature Trntj is measured by the temperature sensor 8, and then, in step 7, the measured intermediary temperature Tintj is compared with the predicted intermediary temperature Tint.caic computed by numerical simulation and the corrected speed setpoint v* is computed by correcting the speed setpoint v0* depending on Trntj and Tjnt,caic.

[0118] Then, in step sc, the electronic system 10 sends the corrected speed setpoint v* to the low-level controller 13, which controls then the speed of rotation of the rolls of the finishing mill 4, or the speed of the transfer rolls positioned between the finishing mill and the cooling machine, so that the speed at which the beam travels through the cooling machine matches the corrected speed setpoint v* when sample Pj traverses the cooling machine.

[0119] In step s7, the corrected speed setpoint Vj* is computed according to the formula below:with ^TSTJ = TST- TsT corran optional correction of the target self-tempering temperature rrt **ST,j-

[0120] The target self-tempering temperature TT ;for the portion Pj of the beam may, like here, be equal to a global target self-tempering temperature T*ST for the beam 2.

[0121] The global target self-tempering temperature T*ST may be entered by an operator using the terminal 12. It may be computed in advance (either by the electronic system 10, or by another electronic device) using a metallurgical model, and based on initial properties of the beam (chemical composition of the steel, and possibly one or more dimensions) and target properties for the beam after its processing in the installation 1 (eg: mechanical properties and / or microstructural properties).

[0122] When the target self-tempering temperature is uniform over the beam length, like here, individual target self-tempering temperatures for the different beam portions Pj, i=1 ,... ,n are equal to each other, and equal to the global target self-tempering temperature T*ST : TT= TSTfor i from 1 to n.

[0123] In this embodiment, the target self-tempering temperature is corrected (meaning that TsT'j is not null) if the measured, intermediate temperature Tintj of the portion Pj is below the so-called Ar3 temperature (that is below the temperature for an austenite to ferrite phase transformation during a temperature decrease). In this case, the correction TSTis set to a positive value (in other words, TgT j Corris below TgTj), the correctionincreasing for instance with the difference Ar3 - Tintj. This reduction of the self-tempering temperature allows for compensating for a strength reduction caused by ferrite formation below the Ar3 temperature. Conversely, if the intermediate temperature Tintj is above Ar3,is null, here.

[0124] The correction applied in step s7 thus allows to adjust the speed setpoint: based on the actual, measured intermediate temperature Tintj, compared to the intermediate temperature Tmt.caic that was computed in advance by numerical simulation, to somehow re-align the temperature predictions, and also based on a local adjustment of the target self-tempering temperature, for the beam portion considered, in view of the actual, measured intermediate temperature Tintj for that portion.

[0125] As represented in figure 6, the adjustment procedure above described is executed repeatedly for the different beam portions Pj, i=1 ,... ,n. Steps s6, s7 and scare executed repeatedly, successively for the different beam portions Pj, with i from 1 to n, as they pass, one after each other, at the intermediary position Os. The discretization of the beam thus plays a role when repeating steps s6, s7 and scfor the different, successive portions Pj of the beam to determine the adjusted speed setpoints v*.

[0126] Namely, for the portion Pj (with i from 1 to n), the temperature Tintj of that portion is measured by the second temperature sensor 8 when the portion Pj passes at the intermediary position Os.

[0127] Then, a locally adjusted speed set point v* is computed for that portion according to the formula below:with TT i= TT i- TT i corran optional correction of a target self-tempering temperature TTHere, TsT:i= Ts*r- TsT i corr, and this correction is computed for each portion Pj in the same manner as above explained for the portion Pj.

[0128] Adjusting speed setpoints individually, for each portion of the beam, based on the temperature measured for that portion just before the finishing mill and subsequent cooling, is highly beneficial, as the temperature along the beam is not uniform, at this point of the beam processing. Figure 8 illustrates this non-homogeneity of the temperature along the beam, just before it enters the finishing mill. In the case of figure 8, n=32.

[0129] It is noted that the adjustment of the speed setpoint is executed n times, for all the successive portions of the beam. Conversely, the determination of the uncorrected, initial speed setpoint v0* is achieved for one portion of the beam only, namely for the portion Pj. This overall arrangement allows for controlling the cooling conditions in an optimal way for each portion, as temperature variations along the beam are taken into account, while executing the highly demanding computation of v0* just once (or possible twice, or a few times, but not n times). This is beneficial, as the time available for computing v0* is limited (by the transfer time between Oe and Os, in the process), and would often not be sufficient for computing directly, by iterative numerical simulations, one speed set point for each portion (each longitudinal sample) of the beam. fxempfary test resu / ts

[0130] Figure 9 represents values of a difference AT between the actual self-tempering temperature, reached in practice after the cooling machine, and the desired, target selftempering temperature TT L COrr. The figure compares the differences AT (box plots) obtained when controlling the beam processing installation according to the above-described method to the differences AT (dispersed points) obtained using a controlling model according to the prior art that is based on a set of linear regressions linking in an empirical way the input temperatures of each portion of the beam and the output temperatures. In figure 9, each group of points, i.e. each column, corresponds to one beam, the box plot corresponds to the scatter obtained with the above-mentioned model and the dispersed points correspond to the scatter obtained with the model according to the prior art. Figure 9 represents test results for 20 different beams. Figure 9 illustrates that the instant control method allows for an accurate control of the self-tempering temperature reached during the thermal treatment of the beams.

Claims

CLAIMS1. A method for controlling a beam processing installation (1) which comprises a rolling mill, and a cooling machine (30) arranged downstream of the rolling mill, along a path (x) followed by a beam (2) in the beam processing installation, the beam being water-cooled when the beam traverses the cooling machine, the method comprising: s1 : acquiring an input temperature (Tm) of a steel beam (2), measured on a portion (Pj) of beam before said portion of the beam enters the cooling machine, s4: Computing a temperature TST of said portion of the beam, which is a temperature predicted for the portion of beam after the portion of the beam has traversed the cooling machine (30), the temperature TST being computed taking into account the input temperature (Tm) and one or more candidate cooling setpoints (vc*) for the cooling of the beam in the cooling machine the step s4 being executed iteratively to determine one or more cooling setpoints (v0*) such that the temperature TST computed in step s4 matches a target self-tempering temperature T*STJ, sc: controlling the beam processing installation (1) based on said one or more cooling setpoints (v0*), or based on one or more corrected cooling setpoints (v*) determined from the one or more setpoints (v0*).

2. A method according to claim 1 wherein, in step s1 , the input temperature (Tm) of the beam is measured on said portion of the beam before said portion exits the rolling mill.

3. A method according to claim 1 or 2, wherein the beam processing installation (1) is controlled based on the one or more corrected cooling setpoints (v*) in step sc, wherein the input temperature (Tm) is measured on said portion (Pj) of the beam when said portion passes at an upstream position (Oe) located upstream of the cooling machine, and wherein the method further comprises: acquiring an intermediate temperature (Tintj), measured on said portion (Pj) of the beam or on another part of the beam (Pj), when said portion or said other part of the beam passes at an intermediary position (Os), located between said upstream position (Oe) and the cooling machine (30),Determining the one or more corrected cooling setpoints (Vj*) based at least on the one or more setpoints (v0*) and on the intermediate temperature (Tintj).

4. A method according to claim 3, wherein the rolling mill comprises a main mill (4) and a subsequent finishing mill (3) which is the last rolling stand traversed by the beam (2) beforecoming out of the rolling mill, and wherein the intermediary position (Os) for the measurement of the intermediate temperature (Tintj) is located between the main mill (4) and the finishing mill (3), while the input temperature (n) of said portion (Pj) is measured before a last rolling pass of said portion in the main mill (4).

5. A method according to claim 3 or 4, wherein: the one or more cooling setpoints comprise a speed setpoint v0* for a speed (v) at which the beam traverses the cooling machine, the method comprising computing two sensitivity coefficients KT and KCT reflecting, respectively, how much the temperature TST is modified when the temperature of the beam at the intermediary position (Tintj) varies, and how much the temperature TST varies when the speed (v) of the beam varies, or when a transit time taken fortraversing the cooling machine with said speed (v) varies, the method comprising computing a corrected speed setpoint v* based at least on: the speed setpoint v0*, the two sensitivity coefficients KT and KCT, the intermediate temperature (Tintj) and a predicted intermediate temperature (Tjnt,caic) for said portion (Pi) of the beam when said portion of the beam passes at the intermediary position.

6. A method according to claim 5, wherein the predicted intermediate temperature (Tjnt,caic) is computed in the course of one of the executions of step s4.

7. A method according to claim 5 or 6, wherein the coefficient KCT reflects how much the temperature TST varies when the transit time taken for traversing the cooling machine varies, the corrected speed setpoint v* being computed according to the following formula:where:- L is a length to be travelled to traverse the cooling machine (30),- Tintj is the intermediate temperature,- Tmt, caic is the predicted intermediate temperature and- AT*STJ is an optional correction of the target self-tempering temperature T*STJ.

8. A method according to claim 7, wherein AT*STJ is computed depending on the measured, intermediate temperature Tintj, AT*STJ being equal to zero if the intermediate temperature Tintj is above a transition temperature Ar3 for an on-cooling austenite to ferrite transformation, and having a non-zero value if Tintj is below Ar3.

9. A method according to any one of claims 3 to 8, wherein the method comprises, for each of different successive parts of the beam Pj with i from 1 to n, one of which being said portion (Pj) of the beam: measuring a temperature Tjnt of the part Pj considered when the part Pj passes at said intermediary position (Os), computing one or more locally-adjusted cooling setpoints (v*) for the part Pj of the beam, based at least on:- said one or more cooling setpoints (v0*),- and the temperature Tjnt , controlling the beam processing installation (1) so as to cool the part Pi, in the cooling machine, according to the one or more locally-adjusted cooling setpoints (v*) computed for the part Pj.

10. A method according to claims 9 and 7, or according to claims 9 and 8, wherein, for each of the successive parts of the beam Pj, with i from 1 to n, the locally-adjusted cooling setpoints computed for the part Pj of the beam comprise a locally-adjusted speed setpoint v* which is computed according to the following formula:where TsT:iis an optional correction of a target self-tempering temperature TT ifor the part Pj of the beam.

11. A method according to any one of the preceding claims, wherein the temperature TST determined in step s4 is the maximum over time reached by the temperature of the surface of said portion of the beam after said portion of the beam has exited the cooling machine.

12. A method according to any one of the preceding claims, wherein said portion (Pj) of the beam is located within a heading, first fourth of the beam (2), a distance between said portion and a head of the beam being optionally higher than I / 20 where I is the length of the beam.

13. An electronic system comprising at least a processor and a memory, configured for executing the method according to any one of the preceding claims.

14. A beam processing installation comprising: the electronic system of the preceding claim, andthe cooling machine for water-cooling the beam.

15. A computer program comprising instructions whose execution on a computer makes the computer to execute the method according to any one of claims 1 to 12.

Images