The present invention relates to a regulating method of furnace atmosphere during the heat treatment of a steel strip. More particularly, the method can be used during an annealing of a steel strip.

During the annealing of a steel strip, the latter is heated and then maintained above its recrystallisation temperature in order to increase its ductility and reduce its hardness. This is done in a furnace wherein the temperature and the dew point of its atmosphere are regulated. Controlling the atmosphere is key to ensure the strip quality.

The tuning of the temperature is ensured by heating devices (radiant tubes, inductors, . . .) providing a desired quantity of heat.

The tuning of the dew point is ensured by providing water, e.g. in the form of vapour, to the atmosphere.

Each steel grade and format have preferred process conditions, including a preferred furnace atmosphere, i.e. ranges of temperature and dew point. So, for continuous production line manufacturing different grades and formats of steel strip, the furnace atmosphere needs to be adapted for each product. During a product transition, shorter is the time from one furnace atmosphere to another, better is the achievable quality of the product.

In order to regulate the dewpoint, the existing system uses a Proportional Integral Derivative controller (PID controller) which is a part of the control loop mechanism coupled with a dew point measuring device and a system to inject water.

In this system, as illustrated in Figure 1, the dew point of the atmosphere is measured (DP MEASURED) and a targeted dew point (DPTARGET) is defined. Then an error value, e.g. a difference between the measured dew point and the targeted dew point, is calculated. After that, a flow of H ₂ O to inject (H ₂ OINJECT) is determined based on proportional, integral and derivative terms calculated by the PID controller. Ultimately, the determined flow of H ₂ O is injected into the furnace. This sequence is regularly repeated throughout the heat treatment.

However, it is key to remember that an annealing furnace has usually an important volume compared to the flow of vapour that can be injected in the furnace. So, when two steels grades requiring different dew points in the furnace are sent consecutively in a short period of time, an optimal dew point cannot be achieved on time over the full coil length. It may lead to a yield decrease or a poor strip quality, especially for the tail of the first strip and head of the second one.

The goal of the invention is to improve the regulation of the H ₂ O injected in a section of an annealing furnace.

More particularly, the goal of the invention is to improve the transition management from a preferred annealing atmosphere A _TAR1 of a steel strip S ₁ to a preferred annealing atmosphere A _TAR1 of a steel strip S ₂ , in a section of a furnace leading to an optimised quality of the final products. The preferred annealing atmosphere can comprise a range of values.

This object is achieved by providing a method according to claim 1. The method can also comprise any characteristics of claims 2 to 7.

Other characteristics and advantages of the invention will become apparent from the following detailed description of the invention.

To illustrate the invention, various embodiment and trials of non-limiting example will be described, particularly with reference to the following figures:

Figure 1 illustrates a regulation method as known in the prior art.

Figure 2 illustrates an embodiment of the regulation system as per the present invention.

Figure 3A illustrates the dew point regulation as known in the prior art.

Figure 3B illustrates the steam injection regulation as known in the prior art

Figure 4A illustrates the dew point regulation according to the invention.

Figure 4B illustrates the steam injection regulation according to the invention

The invention, as illustrated in Figure 2, relates to a method for regulating an atmosphere A, comprising H ₂ and Nz, inside a furnace, wherein a steel strip S ₁ having a composition C ₁ and an exposed surface area A _SURF1 is heat treated from a time T ₀ to a time TSIEND and a steel strip S ₂ , having a composition C ₂ and an exposed surface area A _SURF2 is heat treated from a time T _S2START to a time TN, comprising the following steps:

A. A data acquisition step wherein : i. at a time T ₀ , the dew point; DPo, of said atmosphere A is measured, ii. a target atmosphere, A _TAR1 , for said steel strip SI is retrieved, wherein said target atmosphere A _TAR1 comprises at least a dew point value, iii. a target atmosphere, A _TAR2 , for said steel strip S2 is retrieved, wherein said target atmosphere A _TAR2 comprises at least a dew point value,

B. A tuning step comprising the steps of : i. defining N projected atmospheres, A _PRO-1 to A _PRO-N , corresponding at N times, T1 to T _N , based on: a) said dew point, DP ₀ , of said atmosphere measured at the time T0, b) the volume of said furnace, c) the compositions and the exposed surface areas of said steel strips SI and S2 inside said furnace for each of said times T1 to T _N , ii. estimating an amount of H ₂ O to be injected, Q H ₂ O, at T ₁ inside said furnace using a model predictive control controller using the following data : a) said projected atmospheres A _PRO-1 to A _PRO-N b) said target atmospheres A _TAR1 and A _TAR2 c) said volume and the renewal flow of N ₂ and H ₂ of said furnace, d) the composition and the exposed surface area of said steel strip SI at T ₁ ,

C. Injecting at T ₁ , said estimated amount of H ₂ O.

The furnace comprises heating means. The heating means can be for example inductors and/ or radiant elements such as tubes.

The furnace can comprise several sections such as a pre-heating section, a heating section, a soaking section and a cooling section. The method can be applied to each and every section of a furnace. Preferably, the method is applied in the heating and/ or in the soaking section.

The heat treatment can be any type of heat treatment. The heat treatment is preferably an annealing treatment, and more particularly a recrystallisation annealing.

The steel strip S1 and the steel strip S2 can have the same composition or similar compositions corresponding to the same family of products. The steel strip SI and the steel strip S2 can have different compositions.

Preferably, the steel strip comprises, in weight percent, from 0.0001% to 0.50% of C, from 0.01% to 5.0% of Mn, from 0.001% to 5.0% of Si. The exposed surface area of a steel strip represents the surface of a steel strip inside the furnace and exposed to the furnace atmosphere. The exposed surface can be expressed per unit or per volume area.

TSIEND is the first step at which the strip S ₁ is not in the furnace. In other words, it is the first time at which A _SURF1 is equal to zero. T _S2START is the first step at which the strip S2 is in the furnace, at least partly. In other words, it is the first time at which A _SURF2 is not equal to zero.

In the first step, the data acquisition step, the dew point of the atmosphere is measured and the targeted dew point for each steel strips undergoing a heat treatment in any of the times T ₁ to T _N are retrieved. The targeted dew point can also be a targeted range of dew point so the target atmospheres, A _TAR1 and/ or A _TAR2 , can comprise a range of dew point values.

A target atmosphere of a steel strip permits to perform the continuous heat treatment ensuring a satisfying product quality.

The dew point values or ranges depends on the steel composition. They can, for example, be retrieved from a database.

In the second step, the tuning step, the goal is to estimate the atmosphere of the furnace at N times (T ₁ to T _N ), i.e. at N steps, which correspond to N projected atmospheres ( A _PRO-1 to A _PRO-N ) if no water is added inside the furnace (e.g. the atmosphere) while determining the required water injection flow rate at each step.

Alternatively, the estimation of the atmospheres of the furnace at N times (T ₁ to T _N ) can be performed for different amount of H ₂ O to be injected.

Preferably, N is a positive integer from 2 to 100. Even more preferably, N is a positive integer from 5 to 50. Even more preferably, N is a positive integer from 10 to 30.

Preferably, each of said times T ₀ to T _N are spaced apart by 5 seconds to 1 minute. Even more preferably, each of the times T ₀ to T _N are spaced apart by 15 seconds to 45 seconds.

Preferably, each of the times T ₀ to T _N are spaced from the same period of time. For example, if N=50 and that each of the times are spaced of 10 seconds : T ₀ is at a time t=0 second, T ₁ is at a time t= 10 seconds, T ₂ is at a time t=20 seconds and TN=T _{50 i} s at a time t=500 seconds. The definition of the projected atmospheres, A _PRO-1 to A _PRO-N , is done using the dew point measured in step A.i., the volume of the furnace, the composition and the exposed surface area of the steel strip inside said furnace for each of said times T ₁ to T _N . Such estimation of projected atmospheres are well known by the skilled in the art.

Alternatively, the definition of the projected atmospheres, A _PRO-1 to A _PRO-N , is done using the dew point measured in step A.i., the volume of the furnace, the composition and the exposed surface area of the steel strip inside said furnace for each of said times T ₁ to T _N and an amount of H ₂ O to be injected.

It is possible that at one or several steps, the strip S ₁ and the strip S ₂ are partly in the furnace at the same time, e.g; this is notably the case when two strips are welded one to another. In that case, T _S2START happens before T _S1END.

In that case, the definition of the projected atmosphere is done using the compositions and the surface areas of each of the strip in the furnace.

It is also possible that at one or several steps, no strip is in the furnace, e.g; this is notably the case when T _S2START happens after T _S1END.

The definition of the projected atmospheres is preferably done assuming that the temperature and the pressure of said atmosphere are constant for the steps T ₁ to T _N .

Preferably, said target atmospheres, A _TAR1 and A _TAR2 , comprises at least a minimum and a maximum for the H ₂ concentration, a maximum for the CO concentration.

Even more preferably, said target atmospheres, comprises from 0.1 to 50% by volume of H ₂ . Even more preferably, said target atmospheres, comprises from 1 to 50% by volume of H ₂ .

Then the amount of H ₂ O to inject at T ₁ in the furnace, QH ₂ O, is estimated. This is done by using a model predictive control (MPC).

Generally, a model predictive control is based on iterative, finite-horizon optimization of a plant model. At time t the current plant state is sampled and a cost minimizing control strategy is calculated for a relatively short time horizon in the future [t, t+T], For example, in this invention, the definition of an amount of H ₂ O to be injected is performed by defining a cost function to be minimised wherein the cost function represents a difference between at least one target atmosphere, ATAR, and the N projected atmospheres, A _pRon.

For example, the cost function can have the following structure : wherein

- ApRo-n is the projected atmosphere at the step n

- ATAR is the targeted atmosphere

- Q H ₂ On is the amount of H ₂ O injected at the step n

- Matl and Mat2 are tuning parameters to weigh the importance between amount of H ₂ O injected and the dew point control error, i.e. the difference between the targeted dew point and the projected dew point.

The volume and the renewal flow of N ₂ and H ₂ of said furnace and the composition and the exposed surface of the steel strip S ₁ at T ₁ allow to model the dynamic between the amount of H ₂ O injected and the evolution of the dew point.

Alternatively, in the step B. ii, a flow of H ₂ O, or steam, to be injected atT ₁ inside said furnace can be estimated using a model predictive control controller using the following data : a) said projected atmospheres A _PRO-1 to A _PRO-N b) said target atmospheres A _TAR1 and A _TAR2 c) said volume and the renewal flow of N ₂ and H ₂ of said furnace, d) the composition and the exposed surface area of the steel strip SI at T ₁ .

In the fifth step, the amount, or the flow, of H ₂ O defined in the step B.ii. is injected into the furnace. This injection can be done by directly injecting water into the furnace or by means of a porter gas.

The present invention permits to have an atmosphere inside a furnace closer to the setpoint value. The present invention is particularly advantageous when there is a strip composition and/ or set-point change in the furnace. The invention also relates to a method for regulating an atmosphere A, comprising H ₂ and N ₂ , inside a furnace, wherein a steel strip S ₁ having a composition C ₁ and an exposed surface area A ₁ is heat treated from a time T ₀ to a time T _N , comprising the following steps:

A. A data acquisition step wherein : i. at a time T ₀ , the dew point, DP ₀ , of said atmosphere A is measured, ii. a target atmosphere, A _TAR1 , for said steel strip SI is retrieved, wherein said target atmosphere A _TAR1 comprises at least a dew point value,

B. A tuning step comprising the steps of : i. defining N projected atmospheres, A _PRO-1 to A _PRO-N , for n times, T ₁ to T _N , based on: a) said dew point, DP ₀ , of said atmosphere measured at the time T ₀ , b) the volume of said furnace, c) the composition and the exposed surface area of said steel strip S ₁ inside said furnace for each of said times T ₁ to T _N , ii. estimating an amount of H ₂ O to be injected, QH ₂ O, at T ₁ inside said furnace using a model predictive control controller using the following data : a) said projected atmospheres A _PRO-1 to A _PRO-N b) said target atmosphere A _TAR1 c) said volume and the renewal flow of H2 and N2 of said furnace, d) the composition and the exposed surface area of said steel strip S ₁ inside said furnace at T ₁ ,

C. Injecting at T ₁ , said estimated amount of H ₂ O.

EXPERIMENTAL RESULTS

In order to assess the impact of the present method on a furnace atmosphere regulation, an atmosphere is modelled for a transition between an Interstitial Free steel, of the type sold by ArcelorMittal, and a Dual Phase 780 steel, of the type sold by ArcelorMittal, for two regulation methods.

In a first simulation, the atmosphere is a regulated using a PID controller. In a second simulation, the atmosphere is regulated using the claimed method. In both simulations, the Interstitial Free steel strip requires a dew point of -30°C and is in the furnace from the minute 0 to the minute 9.5 whereas the Dual Phase 780 steel strip requires a dew point of -15°C and is in the furnace from the minute 9.5 to the minute 20.

For each target dewpoint, there is a tolerance of 5.1 O ' of the molar fraction of H ₂ O. So, for the target dewpoint of 30°C, the tolerance is of ±8°C and for the target dewpoint of -15°C the tolerance is of ± 3°C.

For both simulations, the measured dewpoint (continuous line), the targeted dewpoint (thick dash line) and the tolerances (thin dash lines) are plotted in Figure 3A, for the first simulation, and in Figure 4A, for the second simulation. Moreover, the steam injection is also plotted in Figure 3B, for the first simulation, and in Figure 4B, for the second simulation.

In the first simulation, the time spent outside of the tolerance ranges is of 14% whereas in the second simulation, according to the claimed method, the time spent outside of the tolerance ranges is of only 4%. So, with the claimed method, the time spend outside of the tolerances ranges is divided by 3.5 in this example.