Method for estimating the temperature and the oxide thickness of a steel strip

Abstract

A method for estimating the oxide thickness and the temperature of a heated steel strip, undergoing a heat treatment performed at a temperature from 100°C to 1100°C, comprising the steps of: 1. Measuring at least two radiation intensities at different wavelengths, in a range from 1 to 5 µm, emitted by said heated steel strip, 2. Estimating the temperature of said heated steel strip, TESTIMATED, based on - said at least two measured radiation intensities and - a reference radiation intensity for at least a reference wavelength, emitted by a reference steel strip having a determined oxide layer thickness, 3. Estimating the emissivity coefficient of said heated steel strip, eESTIMATED, using at least one of said measured radiation intensities and the estimated temperature, TESTIMATED, 4. Estimating the oxide thickness, OxESTIMATED, of said heated steel strip using said estimated emissivity, eESTIMATED.

Description

The present invention relates to a method permitting to estimate the temperature and theoxide layer thickness of a steel strip.Steel strips undergo several thermal treatments in order to enhance their properties. Inmost of those treatments, the steel strip is heated above a determined temperature and then cooledmore or less rapidly.One of the most common thermal treatment is the annealing which permits to increase theductility of the steel strip and reduce its hardness. In this process, the strip is heated and maintainedabove its recrystallization temperature and then cooled. During the annealing, the strip surface isgradually oxidised and a layer of oxide is generally formed on its surface. ITowever, depending onseveral factors (such as the annealing condition : e.g. temperature, dew point, atmosphere, and thesteel grade), the oxides layer thickness varies from 0 to 200 nm. Generally, the oxide layer isessentially composed of FeO due to the thermo-dynamical conditions.Controlling the strip temperature and the oxide layer thickness is key to ensure a goodquality of the strip, control the process and adapt the subsequent process steps. In an anncalingfurnace, this control is usually done by means of pyrometers using the sirip radiation to measurethe temperature.Yet, the thickness variation of the oxide layer impacts the temperature measurement doneby the pyrometers. Indeed, it is admitted that thicker is the oxide layer, greater is the emissivity andso greater is the intensity of the detected signal by the pyrometers. Ilowever, an increase of thesteel temperature also leads to a greater detected signal. Consequently, a pyrometer cannot reliablydetect the presence of an oxide layer, let alone its thickness. When the detected signal intensityincreases, it is not possible to determine if it is due to an increase of the temperature, of the oxidelayer thickness or of both.Thus, the measured temperature by a pyrometer is not reliable because it does not take intoconsideration the emissivity variation due to the oxide layer thickness variation. Therefore, acocfhicient, depending on the emissivity of the measured layer, is applied to the temperature givenby the pyrometers. Several methods have been developed to estimate the temperature and theemissivity of a steel strip during an annealing.JP 09 033 464 discloses a method to measure online the scale thickness. It claims a six-steps process comprising the steps of :- detecting in an annealing furnace an infrared emission light,- determining a first radiance temperature S1 at a wavelength L1 between 12 and 20 μmwherein the emissivity is supposed to be independent of the scale thickness,- determine a second radiance temperature S2 at a wavelength 12 between 2.5 and 4 μmwherein the emissivity depends on the scale thickness,- determine a steel plate temperature based on an emissivity at L1 and S1,- calculate emissivity e2 at L2 based on the determined steel plate temperature and theradiance temperature S2,- determine the oxide thickness based on the emissivity e2.The reliability of this measure is limited because even though the emissivity is nearlyconstant in the 12 to 20 μm domain, its variation in percentage is not negligible and can lead totemperature measurement error of more than 50°C. Moreover, the emissivity in this wavelengthdomain is particularly influenced by parasite flow in industrial condition which lower thetemperature reliability.JP 11 324 839 discloses a method to precisely measure the thickness of an oxide film formed on asteel plate. The method comprises two steps :- assume that the steel temperature is equal to the soaking temperature of the steel,- measuring the radiance from the surface of a steel sheet ata plurality of wavelength between2.5 and 10μm,- determining the oxide film thickness based on the relationship between the radiance, theoxide film thickness and the emissivity.The reliability of this measure is mited because industrially, aimed soaking temperaturecan be different soaking temperature in the furnace. Moreover, there might be a temperaturediscrepancy between the soaking temperature and the one of the steel during the radiancemeasurement.Conscquently, it is necessary to develop a method permitting to accurately and reliablydetermine the temperature of the steel strip in order to increase the reliability and the precision ofthe measurement of its oxide layer thickness.This object is achieved by providing a method according to claim 1. The method can alsocomprise any characteristics of claims 2 to 8. Claims 9 to 11 relate to thermal treatment methodsusing the measure done in claims 1 to 8.Other characteristics and advantages of the vention will become apparent from thefollowing detailed description of the invention.To illustrate the invention, various embodiment and trials of non-limiting example will bedescribed, particularly with reference to the following figures:Figure 1 exhibits process flow diagrams of a measuring method as known in the prior artand as claimed in the present invention.Figure 2 illustrates the steps of an embodiment of the present invention.Figure 3 is a plot representing a relative luminance in function of the wavelength for steelstrip having various oxide layer thickness.Figure 4 exhibits two temperature measurements, one according to the prior art and theother according to a method of the present invention.The invention relates to a method for estimating the oxide thickness and the temperatureof a heated steel strip, undergoing a heat treatment performed at a temperature from 100°C to1100°C, comprising the steps of:1) Measuring at least two radiation intensities at different wavelengths, in a range from1 to 5 μm, emitted by said heated steel strip,2) Estimating the temperature of said heated steel strip, TESTIMATED, based on- said at least two measured radiation intensities and- a reference radiation intensity for at least a reference wavelength, emittedby a reference steel strip having a determined oxide layer thickness,3) Estimating the emissivity coefficient of said heated steel strip, eESTIMATED, using atleast one of said measured radiation intensities and the estimated temperature,TESTIMATED,4) Estimating the oxide thickness, OxESTIMATED, of said heated steel strip using saidestimated emissivity, eESTIMATED.The heat treatment performed at a temperature from 100°C to 1100°C can be an annealingtreatment comprising a heating step and a soaking step. Moreover, after said heat treatment, thesteel strip can be cooled and coated.The steps of the claimed method are illustrated in Figure 1.In the first step of the process, the intensity of at least two radiation, emitted by the heatedsteel strip, at different wavelengths of the 1-5 μm domain, are measured by any suitable measuringmeans. For example, a first radiation intensity at a wavelength of 2 μm is measured and a secondradiation intensity at a wavelength of 4 μm is measured. The measuring means can be twospectrometers or a hyperspectral camera. This first step is represented, in figure 1, by a plotrepresenting the radiation intensity in function of the wavelength which can be produced by saidsuitable measuring means.The wavelength, of the measure intensity, is preferably not more than 5 μm because the 5-8 μm range lies in the absorption domain of air and also because greater is the wavelength, greateris the estimation error on the temperature difference as it can be deduced from the followingequation :ΔT = Δε / εxλxTK2 / 14388.The radiation intensity of each wavelength detected by the recording means depends mainlyon two factors : the radiance and the emissivity of the heated steel strip. In the following terms, λrefers to a wavelength, T refers to a temperature of the steel strip and Oxra refers to the thicknessof the oxide layer.A steel strip radiance, Radiance (λ, T), depends only on the steel strip temperature and themeasured wavelength as explained by the Planck Law.The steel strip emissivity of a steel grade, Emissivity (λ, OxTH), depends on the oxide layerthickness and the wavelength. Consequently, the recorded intensity can be defined by the Equation(1):(1) Intensity (T, λ, OxTH) = Emissivity (λ, OxTH) * Radiance (λ, T)In the second step, the goal is to estimate precisely the temperature of the heated steel stripusing said measured at least two radiation intensities and at least two reference radiation intensityat different wavelength, emitted by a reference steel strip having a known temperature for at leastN oxide layer thickness from 0 to 200 nm. Saud N oxide layer thickness are noted as OxTHn.N is an integer. Preferably, Nis greater than 10. Fyen more preferably, Nis greater than25. Preferably, the step between each reference oxide layer thickness is of 5 nm.One way to achieve that will be presented hereunder. The terms of the Equation (1) canbe divided by the radiation intensity of a reference steel strip leading to Equation (2).Equation (2) From the Equation (2) can be easily deduced the Equation (3).Equation (3)The term, In [Radiance(λ,T) / Radiance (λREF,TREF)], is equal to C2 / λ(TREF).T-TREF / T which is rewritten as CT(T) / λwherein CT(T) is equal to C2x(T-TREF) / TREFxT wherein T is the temperature of the heated steel strip, TREFis the temperature of the reference steel strip and C2 is a constant from the Planck's formula andequals to he / k, wherein h is the Planck's constant and k is the Boltzmann constant.A hncarived emissivity being equal to : λ x In [Emissivity (λ, OxN) / Emissivity(λREF, OxThn)], can be defined. By combining said linearized emissivity and said at least two reference emissivity at differentwavelength of a reference steel strip having a known temperature for atteast N oxide layer thicknessfrom 0 to 200 nm, it is possible to approximate the linearized emissivity with an affine function.For example, said affine function can have a slope "a" and a y-intercept "b" wherein 'a'and 'b' are approximated using a polynomial function. For example, 'a' = a1 x OxN2 + a2 x OxN +a3, and 'b' = by x OxN2 + b2x OxN + b3.In an analogue manner, a linearized intensity can be defined as being equal to:λxIn[Intensity(T,λ, OxTH) / Intensity(TREF,λREF, OxTHn).] By combining said linearized intensity and said at least tworeference radiation intensity at different wavelength, emitted by a reference steel strip having aknown temperature for said at least N oxide layer thickness from 0 to 200 nm, it is possible toapproximate the linearized intensity with an affine function.For example, said affine function can have a slope "a" and a y-intercept "b". 'a' and 'b' canbe approximated using a polynomial function.For example, 'a' = a1 x OxN2 + a2 x OxN + a3, and b= b1 x OxN2 + b2 x OxN + b3 + CT(T).By combining the equation 3 with the linearized intensity and emissivity, the followingequation can be established :Equation (4)Then CT(T) can be found be resolving the equations systems. Solving the equation systemsleads to two pairs of a value of oxide thickness associated with a CT(T), i.e. the temperature of theheated steel. The skilled m the art can easily rule out the pair presenting an incoherent value, bysetting acceptable domain for the values. For examples, oxide thickness value being negative orexceeding a threshold value (such as 500 nm) or a steel temperature higher than the steel meltingtemperature can be considered as not possible.It permits to find the estimated temperature of the heated steel sheet, TESTIMATED.Greater is the number of measured radiation intensity and of reference radiation intensityand of reference emissivity, the more accurate is the coefficients of the polynome and thus greateris the accuracy of the estimated temperature.Preferably, the reference steel strip and the heated steel strip have a similar composition orbelong to the same steel grade. Even more preferably, said reference steel strip has the samecomposition as the heated steel strip.As vastly known, based on the Planck's Law, the emissivity of a body can be calculatedwhen its temperature is known. Consequently, in the third step, the emissivity of the heated steelstrip can be estimated using the Planck's Law and the estimated temperature, TESTIMATED. Forexample, the Equation (5) wherein L is the luminance of the Plank's law can be used to estimatethe emissivity. This is illustrated in Figure 1. The estimated emissivity is noted egsrivatep.Equation (5)More than one emissivity of the heated steel strip can be estimated by using more than oneof the at least two measured radiation intensities.In the fourth step, the iron oxide thickness can be estimated using abacus wherein the ironoxide thickness is plotted in function of the emissivity of a steel strip for a determined wavelength.Such a curve is plotted in Figure 1, wherein the oxide layer thickness is plotted in function of theemissivity of the FeO oxide for a determined wavelength.More than oxide thickness of the heated steel strip can be estimated by using more thanone of the estimated emissivity.In the present invention, the temperature of the steel strip is estimated using measurementsand reference values. On the contrary, in the prior art, the temperature was estimated usingforecasted process temperature or two radiance temperature, as illustrated in Figure 2. Moreover,the assumption that the emissivity is independent of the scale thickness for a wavelength between12 and 20 μm is not correct as illustrated in Figure 3 wherein the relative luminance is plotted infunction of the wavelength for oxide thicknesses from 0 to 500 nm.Thus, the estimated temperature of the present invention is more precisely and reliablydetermined because it takes into account the surface state (e.g. true emissivity) of the heated steelstrip. Consequently, it also permits to improve the estimation of the oxide layer thickness.Preferably, said heated steel strip is running.Preferably, in step 1), at least ten radiation intensities, emitted by said heated steel strip, atdifferent wavelengths of the 1-5 μm domain, are measured and in step 2), TESTIMATED is estimatedusing said at least ten radiation intensities. Even more preferably, in step 1), at least twenty radiationintensities, emitted by the steel strip, at different wavelengths of the 1-5 μm domain, are measured.and in step 2), TESTIMATED is estimated using said at least twenty radiation intensities. The moreradiation intensities are used, the more reliable are the estimations.Preferably, the at least two radiation intensities have a wavelength difference of at least 0.1μm, more preferably of at least 0.5 μm and even more preferably of at least 1 μm. Apparently,greater is the wavelength difference, the more precise will be the temperature estimation.Preferably, the heated steel strip and the reference steel strip have similar composition.Preferably, the composition of the heated steel strip and the reference steel have for each element,a mass proportion difference of maximum 10%, more preferably of maximum 5% and even morepreferably of maximum 2%. For example, for a mass proportion difference of maximum 10%, ifthe heated steel strip comprises 5% of silicon, the reference steel strip comprises from 4.5% to5.5% of silicon.Preferably, the steel strip is set at a temperature from 500°C to 1100°C. Such a temperaturerange permits to increase the radiance of the strip in the 1-5 μm range thus improving themeasurement precision. Setting the steel strip temperature in this range is preferably done duringan annealing process.Even more preferably said heat treatment is performed at a temperature from 500°C to1100°C, and in said step 1) the wavelength range of the measured at least two radiation intensitiesis from 1 to 1.7 μm. This wavelength domain is advantageous, for this temperature range, becausea variation of the oxidised thickness layer impacts strongly the emissivity compared to otherwavelength ranges of the 1 to 5 μm range. Secondly, this range exhibits the smallest impact of themeasurement uncertainty of the estimated temperature on the estimated emissivity because withinthe 1 to 5 μm range.Preferably, said heat treatment is performed at a temperature from 100°C to 500°C, and insaid step 1) the wavelength range of the measured at least two radiation intensities is from 3 to 5μm. Such a range is advantageous because the radiation intensity variation is, for this temperaturerange, greater in this domain than in the 1 to 3 μm range.Preferably, said steps 1) to 4) are repeated for several points of said heated steel stripsurface. Even more preferably, said steps 1) to 4) are done for several points along the heated steelstrip width and along the heated steel strip length. Doing the steps 1) to 4) at several points of theheated steel strip surface permit to map the oxide layer thickness and the heated steel striptemperature at different locations of the heated steel strip. Advantageously, measurements are doneclose to the strip edges and close to the middle of the strip width.Preferably, the method comprises a step of mapping the oxide thickness and thetemperature of said steel strip using the estimated oxide thicknesses and estimated temperatures ofsaid several points of the steel strip surface.The invention also relates to a method of a thermal treatment of a heated steel stripperformed in a furnace, wherein the previously described method is performed and said TESTIMATEDis used to control the furnace temperature.Preferably, said furnace comprises a heating section and a soaking section the previouslydescribed method is performed in said heating section and said TESTIMATED is used to control saidfurnace temperature during said heating step.During the heating and the soaking steps, target temperatures for the heated steel strip areset in order to achieve the desired properties. Thanks to the previously explained method, theheated steel strip temperature can be monitored more precisely and reliably. Consequently, thefurnace temperature and the heat quantity brought to the heated steel sheet can be varied to matchthe TESTIMATED with the target temperature.The invention also relates to a method of thermal treatment of a steel strip, comprising aheating step and a soaking step, being performed in a furnace comprising burners with adjustablepower along the width of said heated steel strip, wherein the previously explained method isperformed during said heating step and said estimated oxide thickness, OxESTIMATED, is used to varythe power of said burners along said heated steel strip and to homogenize the oxide thickness alongthe width of said heated steel strip width.Because several oxide thicknesses are estimated along the strip width, the oxide thicknessvariation along the strip width can be estimated. Then, the intensity of the fumers can be variedin order to homogenize the oxide thickness along said heated steel strip width.LXPLRIMENTAL RESULTSA comparative experiment has been conducted to evaluate the reliability of the claimedmethod. In this experiment, the temperature of the heated steel strip has been measured by threedifferent techniques : pyrometers, wedge measure and the method according to the presentinvention. The results are plotted in Figure 4.It is known that the wedge measure is highly reliable for stable conditions, when thetemperature is more or less constant, but is not reliable for unstable condition, when thetemperature of the steel strip varies.It can clearly be seen in Figure 4, where the temperature is stable for circa 12 minutes, thatthe estimated temperature with the claimed method is closer to the temperature measured by thewedge measure than the temperature measured by pyrometers.Consequently, the claimed method provides a more accurate method.

Claims

1. A method for estimating the oxide thickness and the temperature of a heated steel strip, undergoing a heat treatment performed at a temperature from 100°C to 1100°C, comprising the steps of: 1) Measuring at least two radiation intensities at different wavelengths, in a range from 1 to 5 μm, emitted by said heated steel strip, 2) Estimating the temperature of said heated steel strip, TESTIMATED, based on - said at least two measured radiation intensities and - at least two reference radiation intensity and at least two reference emissivity, at different wavelength, of a reference steel strip having a known temperature for at least N oxide layer thickness from 0 to 200 nm, 3) Estimating the emissivity coethicient of said heated steel strip, εESTIMATED, using at least one of said measured radiation intensities and the estimated temperature, TESTIMATED, 4) Estimating the oxide thickness, OxESTIMATED, of said heated steel strip using said estimated emissivity, εESTIMATED.

2. A method according to claim 1, wherein said heated steel strip is running.

3. A method according to claim 1 or 2, wherein in step 1), at least ten radiation intensities, emitted by said heated steel strip, at different wavelengths of the 1-5 μm domain, are measured and in step 2), TESTIMATED 18 estimated using said at least ten radiation intensities.

4. A method according to claim 3, wherein in step 1), at least twenty radiation intensities, emitted by the steel strip, at different wavelengths of the 1-5 μm domain, are measured. and in step 2), TEsrpparsn is estimated using said at least twenty radiation intensities.

5. A method according to any one of the claims 1 to 4, wherein said at least two radiation intensities are measured at a wavelength difference of at least 0.1 μm, more preferably of at least 0.5 μm and even more preferably of at least 1 μm.

6. A method according to any one of the claims 1 to 5, wherein said heat treatment is performed at a temperature from 500°C to 1100°C, and in said step 1) the wavelength range of the measured at least two radiation intensities is from 1 to 1.7 μm.

7. A method according to any one of the claims 1 to 5, wherein said heat treatment is performed at a temperature from 100°C to 500°C, and in said step 1) the wavelength range of the measured at least two radiation intensities is from 3 to 5 μm.

8. A method according to any one of the claims 1 to 7, wherein said steps 1) to 4) are repeated for several points of said heated steel strip surface.

9. A method of thermal treatment of a heated steel strip being performed in a furnace, wherein the method according to claims 1 to 8 is petformed and said TESTIMATED is used to control the furnace temperature.

10. A method according to claim 9, wherein said furnace comprises a heating section and a soaking section and wherein the method according to claims 1 to 8 is performed in said heating section and said TESTIMATED is used to control said furnace temperature during said heating step.

11. A method of thermal treatment of a steel strip, comprising a heating step and a soaking step, being performed in a furnace comprising burners with adjustable power along the width of said heated steel strip, wherein the method according to claims 1 to 8 is performed during said heating step and said estimated oxide thickness, OxESTIMATED, is used to regulate the power of said burners along said heated steel strip and to homogenize the oxide thickness along the width of said heated steel strip.