EP3271092B1 - Method for producing metal strips - Google Patents

Description

The invention relates to a method for producing metal strips in a rolling mill with a desired profile contour according to the preamble of patent claim 1 or 3.

Background of the present invention is the fact that the requirements for the setting accuracy of a profile of a metal strip increase at least at individual predetermined bandwidth positions, so-called reference positions, as well as the dimensional accuracy of the profile contour of the metal strip. Depending on the planned application of a metal strip z. B. parabolic hot strip profile contours expected with a predetermined profile height at a certain reference position to facilitate further processing in a downstream cold rolling mill (tandem mill). Alternatively, box sections may be required, d. H. Metal bands with a flat cross-section in the middle, which decreases more towards the band edges; This requirement is made for example on metal bands, which are to be divided later in the longitudinal direction. Conversely, concave band profiles, i. H. Band profiles, which have thicker or raised edges compared to their middle region, and metal bands with Kantenwulsten usually not desired.

In order to produce the desired band profiles as precisely as possible, various approaches have already been proposed in the prior art.

Thus, the international patent application discloses WO 1995/034388 a detection system for detecting the profile of a metal strip at the exit of a finishing train. The band profile K detected there is compared with a predetermined target profile at this position, and the use of profile actuators is proposed in order to minimize the deviation of the measured profile from the target profile in subsequent bands. Furthermore, a Deciding if a measured band profile shape is acceptable or not and proposing measures, e.g. B. Change of the thermal crown shape of the work rolls to improve the profile shape, if necessary.

Also the EP 0 618 020 B1 , on which the preamble of claims 1 and 3 is based, aims to adapt the profile of a metal strip at the exit of a hot strip mill to a predetermined target contour. For this purpose, mechanical actuators are used so that a possibly determined deviation between a calculated, ie predicted band shape and the predetermined target contour is minimized. Also, a measured band profile C40 (at the position 40 mm from the band edge) is used for correction or adjustment of control systems.

Furthermore, the procedure according to the preamble of claim 1 or / and 3 is known in the art. Accordingly, a prediction value for the band profile and setting values for profile actuators when rolling an nth metal band at a predetermined reference position are simulated and calculated with the aid of a mathematical-physical process model. If necessary, the simulation takes into account restrictions and the use of different profile actuators. After a rolling of the nth metal strip, an adaptation value is calculated on the basis of the difference between said prediction value and a measured actual value for the strip profile of the nth metal strip at said reference position. The reference position is a predetermined bandwidth position measured from the natural edge of the metal strip, for example 25 or 40 mm. According to the prior art, said prediction value and the said adaptation value are determined or predefined only at a single reference position in order to define individual target values for the band profile of the metal band on this basis.

Based on this prior art, the invention has the object, a known method for producing metal strips in one To further develop rolling plant to the effect that - in the future production of metal strips - a more accurate forecast of the profile contour of the metal strip across the width and a more accurate adjustment of profile actuators of the rolling mill is possible.

This object is achieved by the method claimed in claims 1 and 3.

In the method according to claim 1, the prognosis value for the profile contour is calculated in the context of the simulation of the rolling process before the rolling of the metal strip. In contrast, the prediction value according to the method of claim 3 is not calculated in the simulation before rolling, but by a recalculation after the rolling of the metal strip.

In other words, alternatively, the adaptation value calculation depending on the adaptation philosophy in the prognosis value according to claim 1 to the calculated in the simulation of the rolling process value of the profile with the use of preset values (expected rolling force, etc.) or according to claim 3 to the result of recalculation Act with actual conditions (measured rolling forces, etc.).

In principle, the aim in both methods is that the calculated forecast values coincide with the predetermined target values; However, due to process- or analgen-specific peculiarities, it may happen that the forecast values do not exactly match, but only approximately with the target values.

The calculation of the prediction values for the band profiles at the different reference positions bi is carried out with the same setting of the profile actuators. This applies in both claimed methods.

The term "metal strip" also includes metal sheet.

The term "rolling mill" includes both single scaffolding, such as heavy plate scaffolding, Steckel or Twin Steckel scaffolding, etc., but also whole finishing mills with a.

The term "reference position bi" preferably designates a subjacent of the general positions m in the width direction of the metal strip. While normal bandwidth positions are defined by their respective distance from the center of the metal strip in the width direction, reference positions are defined by a respective predetermined distance from the belt edge or natural edge of the metal strip. For standardized reference positions, eg. B. 25 mm, 40 mm or another reference position, z. B. 100 mm from the natural edge of the metal strip are typically given values for the profile contour, z. As C25, C40 or C100 values. The reference positions are preferably the same for different bandwidths or for all metal bands. Whether the C ... values are target values, forecast values or adaptation values is determined from the context.

The term "process model" means a mathematical / physical model for simulating a rolling process. In particular, it is suitable to calculate prognosis values and profile contours for the metal strip as well as the setting values of profile actuators. The process model is also referred to as "Profile Contour and Flatness Control" PCFC.

The term "calculated value" means "forecast value". Similarly, "calculated contour" means "predicted contour".

The term "later production" or "future production" means a manufacturing or rolling time after the determination of the new adaptation value for the at least n'te metal strip. Later manufacturing can be further Obtain longitudinal sections of the same nth metal strip or a completely new metal strip n + x.

bezeichnet ein zukünftig nach dem n'ten Metallband hergestelltes bzw. herzustellendes Metallband. So bezeichnet beispielsweise n+2 das zweite nach dem n'ten Metallband herzustellende, insbesondere zu walzende Metallband.The term "n + x" with x = 1, 2, 3, ... etc x∈

refers to a future produced after the n'ten metal strip or metal strip. For example, n + 2 denotes the second metal strip to be produced after the nth metal strip, in particular to be rolled.

The respective future to be rolled band is thus generally designated for the corresponding preset calculation each n + x. The previously calculated adaptation values are used here.

The terms "profile contour" and "band profile", each seen in the width direction of the metal strip, are used synonymously.

The core idea of the claimed claimed invention is that an adaptation value as the difference between a measured actual value and a calculated, ie predicted value for the profile contour of the metal strip not only, as usual in the prior art, at only one (number) determined Reference position, but at a plurality of reference positions is determined. This advantageously a Bandkonturadaption is possible. This plurality of determined adaptation values over the bandwidth can be taken into account in the calculation and adjustment of the profile actuators and in the calculation of the profile contour or in the calculation of the prognosis values for the metal strips to be rolled in the future. By providing the plurality of adaptation values and due to more accurate knowledge of the profile contour, the profile actuators can advantageously be more accurate with regard to the desired target values for a long longitudinal section of the nth metal strip or for the profile contour of the n + x'ten metal strip or the profile contour in the future to be rolled metal bands. Also the calculation of Forecast values for the profile contour is therefore more accurate for the n + x'th metal strip, ie for metal strips to be rolled in the future.

According to an advantageous embodiment, a distinction is made in the determination of the adaptation values at the reference points bi between short-term adaptation values and long-term adaptation values. This advantageously makes it possible to use the learned at at least one band n n for a later to be rolled band n + x, because the same profile contour deviations between measured and predicted profile contour values occur again and again quite frequently in a follower band or at a later rolled under similar conditions on.

• mit K: Kurzzeitadaption und
• ΔC_K (n-x)bi : Alter Kurzzeitadaptionswert
• C_Ist (n)bi: Gemessener Ist-Wert für die Profilkontur des n'ten Bandes
• C_P (n)bi: Errechneter Prognosewert bzw. errechnetes Bandprofil
• x=1, 2, 3 ...
• n: betreffendes Metallband

The calculation of the short-term adaptation value takes place according to the formula: $$\mathrm{\Delta }C\left(n\right)\mathit{bi}=\mathrm{\Delta }{C}_K\left(n\right)\mathit{bi}=\mathrm{\Delta }{C}_K\left(n-x\right)\mathit{bi}+\left[C_{\mathit{is}}\left(n\right)\mathit{bi}-C_P\left(n\right)\mathit{bi}\right]$$

• with K: short-term adaptation and
• Δ C _K ( nx ) bi : age short term adaptation value
• C _Is ( n ) bi : Measured actual value for the profile contour of the nth band
• C _P ( n ) bi: Calculated forecast value or calculated band profile
• x = 1, 2, 3 ...
• n: concerned metal band

Using this formula for the short term adaptation value, the summand ΔC _K (nx) bi at restart of a rolling process, e.g. B. after a work roll change, with z. 0 or another typical initial value. The short-term adaptation value is then calculated as the sum of the initial value and the difference between the actual value C _actual (n) bi for the profile contour and the prediction value C _P (n) bi of the nth metal band at the reference position bi.

• Ermitteln der Adaptionswerte durch Wiederholen der Schritte a) bis f) nach Anspruch 1 oder 3 an der Mehrzahl I von Referenzpositionen bi für eine Mehrzahl von vordem n+x'ten Metallband gewalzten Metallbändern einer Adaptionsgruppe; und
• Berechnen der Langzeitadaptionswerte ΔC_Lbi durch Bildung der Mittelwerte der Adaptionswerte oder Bildung der Mittelwerte der Differenzen zwischen Ist-Werten und Prognosewerten für die Profilkontur für die Mehrzahl von Metallbändern jeweils an einer der Referenzpositionen bi.

The long-term adaptation value ΔC _L bi at a reference position bi is obtained by performing the following steps:

• Determining the adaptation values by repeating steps a) to f) according to claim 1 or 3 at the plurality I of reference positions bi for a plurality of metal bands of an adaptation group rolled prior to n + x th metallic strip; and
• Calculating the long-term adaptation values ΔC _L bi by forming the average values of the adaptation values or forming the mean values of the differences between actual values and prognosis values for the profile contour for the plurality of metal bands, each at one of the reference positions bi.

For the determination of the prognosis value C _P (n + x) bi of the metal strip n + x according to claim 1 or 3, the long-term adaptation value ΔC _L bi is optionally taken from the corresponding adaptation group to which the metal band n + x belongs.

In other words, the long-term adaptation value may also result from an averaging of the total adaptation values (long-term and short-term adaptation value) of j bands that have been rolled in the same adaptation group in the past.

The maximum number used in the past rolled bands j can e.g. 100 or 50 and is freely definable. The difference in a band thus affects the long-term adaptation value only for a jth part. The determined long-term adaptation value can be used in the PCFC preset calculation to 100% or only partially, depending on freely definable boundary conditions.

The definition and the calculation of the long-term adaptation value ΔC _L (n) bi may presuppose the knowledge of the short-term adaptation value ΔC _K (n) bi. In contrast, in exceptional cases, the Kurzzeitadaptionswert can also be used alone.

As an alternative to the long-term and / or short-term adaptation value, it is also possible to determine a total adaptation value for determining the setting values of the profile actuators and for determining the contour of the band at the reference points bi according to claim 6. This total adaptation value is then calculated as the sum of the short-term adaptation value and the long-term adaptation value in each case at a reference position bi.

How the adaptation values, calculated profile values and measured values etc. can behave at a reference position from band to band for 4 bands of the same long-term adaptation group is illustrated in the following example:

According to a further embodiment, the determined short-term adaptation value, the determined long-term adaptation value or the determined sum adaptation value can be used in the calculation for presetting the profile actuators either to 100% or only to a desired part. The desired proportion can be selected depending on freely definable boundary conditions. Depending on the selected weighting, z. B. 33 or 50%, the adaptation effect is attenuated or smoothed. The change of short-term adaptation values from band to band may be limited by a maximum value, e.g. B. 10 microns, are limited to not weight any individual measurement errors too high. Also, the short term adaptation value may be furnace dependent or dependent on other process variables. The Kurzzeitadaptionswert usually refers to the profile differences of the last band n. In exceptional cases, z. B. the Profile difference be related to the penultimate band. Then n corresponds to the band n-1 or generally nx.

The adaptation values calculated according to the invention at the individual width positions bi of the metal strip can advantageously also be used to determine the adaptation contour of the metal strip by connecting the individual existing adaptation values with one another to at least one suitable attachment function. The adaptation contour can be guided by the adaptation values Δ C ( n + x ) bi determined for the metal band n + x, or the adaptation contour runs close to the adaptation values depending on the approach function or smoothing function (approximation). An approach function is thus used to connect adaptation values, interpolation, smoothing, extrapolation or approximation and is for example so designated. As a rule, adaptation values are present at at least two reference positions bi, and preferably at least one further adaptation contour value is present at a further bandwidth position m, which is not a reference position. Other bandwidth positions are typically dictated by the process model. Depending on which bandwidth positions the adaptation values are known for, the adaptation contour can be determined either only over a limited section or area or over the entire width of the metal strip. The density of the known adaptation values may be different in individual regions over the width of the metal strip. Preferably, the density of the known adaptation values in the edge region of the metal strip, there preferably at the reference positions, greater than in the central region, also called body region. This is due to the fact that the requirements for the accuracy of the profile contour in the edge region are often higher than in the middle region. If, for an extreme special case, every smoothed measuring point supplied by a profilometer has an adaptation point bi, then the adaptation contour can also be determined without further determination by an interpolation function; In this case, the adaptation contour simply exists in the adjacent sequence the plurality of adaptation values. As a rule, the maximum number I of bandwidth positions, in particular reference positions, is less than 10.

According to an advantageous exemplary embodiment of the invention, the said and determined adaptation contour for the n + x'th metal band is added to a non-adapted calculated profile contour predicted by the process model in order to obtain an adapted profile contour for the n + x'th metal band.

The determination of the approach functions or interpolation functions of the adaptation contour or of the adapted profile contour can be carried out differently for different width sections of the metal strip. A first width section may be, for example, in the middle width area and second width section or further width sections may be, for example, in the edge area, also called edge area of the metal strip.

In the case of two width sections adjoining in the width direction, the attachment functions or the adaptation contour or the adapted profile contour over the two width sections are preferably selected such that the contour progressions are continuously differentiable at the border from one band section to another, in particular have the same pitch. This condition avoids that the contours at the boundary between the two band sections have a kink; instead, they go smoothly together.

The adaptation contour or the adapted profile contour over a width section of the metal strip can be extrapolated into an adjacent width section for determining an extrapolated adapted adaptation contour or an extrapolated adapted profile contour over the adjacent width region, in particular if no adaptation values or measured profile contour values are known there.

The said at least one starting function or approximation function or interpolation function for connecting individual adaptation or profile contour values or the said extrapolation function can be formed from a linear function, a polynomial function of any order, an exponential function, a trigonometric function, a spline function or a combination of different functions. The starting functions or interpolation functions can also be different for different width sections of the metal strip.

Instead of the measured actual value of the profile contour of the metal strip at the reference position bi, it is also possible to use an average of measured actual values at the mirror-image reference positions bi on the right and left half of the metal strip, as viewed in the rolling direction. The imaginary plane, also called the width plane, acts as a mirror plane at half the width or width of the metal strip, which extends in the longitudinal direction of the metal strip.

The adapted profile contour values or the adapted profile contour can initially only for a band half, z. B. the band half can be determined on the operating side and below for the other band half, z. B. mirrored for the band half on the drive side.

The measured actual value of the profile contour can be used as a direct measured value at the reference position bi or as a profile measured value smoothed by a compensation function across the width, for example a measured value interpolation function.

The measured actual values C _ist (n) bi in the profile contour can be determined at a defined tape length position or averaged over a tape segment length or averaged over an entire tape length.

Advantageously, the adapted profile contour determined according to the invention is determined with regard to profile anomalies, such as, for example, band bulges, d. H. unwanted thickening in the band edge region, or steep band profile waste, especially in the edge region of the metal strip analyzed. The analysis is preferably carried out online or in a real-time mode. Then, the profile actuators can be suitably adjusted to actively combat or reduce said profile anomalies in subsequently rolled sections in the longitudinal direction of the same metal strip or subsequently rolled metal strips.

Without the use of the adaptation contour according to the invention, it may happen that metal bands are calculated with normal profile contours, and that, however, in practice, band bulges form on the edges. The inventively made possible determination of the adaptation contour and the thus made possible determination of a more precise adapted profile contour opens up new possibilities of improved determination of the profile contour. If z. B. calculated for a metal band an edge bead height which is higher than an allowable threshold, the band profile level 40 mm away from the natural edge of the metal strip automatically by the process model within allowable predetermined profile level limits between, for example, C40 target _min and C40 target _max set to a value, usually raised, so that the maximum allowable bead height is not exceeded or reduced or / and there is a targeted use of profile actuators (eg roller displacement, etc.) to reduce the bead height.

Further advantageous embodiments of the method according to the invention are the subject matter of the dependent claims, in particular of claims 21 to 23.

Taking advantage of the material transverse flow behavior, the body band profile, ie the profile contour in the middle region of the metal band, and the edge band profile using the contour adaptation can be adjusted more precisely in two steps. First, the profile actuators in the front Area of the rolling mill or in the first stitches of a reversing mill used so that adjusts the body profile. In the second step, the profile actuators for the rear scaffolds or last stitches are set so that the nominal profile is also set at the edge of the strip or so that an overall contour is shaped.

Thus, several target profile values for different width positions can be specified, all of which are set or / and which are kept or monitored within certain limits. For example, by an extended process model, a target profile value C25 = 30 μm can be set in the edge region or the deviation can be minimized and at the same time the limit C100> 15 μm can be maintained for a target profile value in the bodyband region.

In the setting strategy, the profile value in the band edge region may be e.g. C25 or alternatively the bodyband profile value e.g. C100 as the primary target variable and given differently from band to band. Appropriately, the band contour values or the band contours are adapted (as described) at these reference points.

The adapted profile contour function, consisting of m _max profile contour values C (n + x) m, is advantageously analyzed with respect to band profile anomalies, and by means of the process model the information of the analyzed finished band contour errors is transmitted to the calculation of the interstitial or intermediate stitch contours by means of transfer functions or weighting factors not described in detail. Alternatively or additionally, the determined adaptation values at the positions bi are transmitted to the calculation of the interstand or intersection contours by means of transfer functions or weighting factors not described in greater detail.

The exact quantitative knowledge of the location of the band contour anomalies (bead height, bead width, edge drop between two defined profile points (eg C25-C100) as well as profile deviations in the middle band range (or at C100, C125, C150 or C200) thus allow a targeted analysis of whether band contour errors occur at the edge, in the middle range or in both ranges. With this knowledge, in a profile and flatness calculation, the profile actuators of the different frameworks are iteratively used in a more targeted manner in order to avoid or reduce band profile anomalies.

This allows profile actuators, e.g. variable work roll cooling systems, zone cooling or local roll heating for influencing the thermal crown, a work roll displacement in conjunction with roll grinding ("anti-bead roller" or "tapered roll", CVC roller coiling) Rollers, higher order polynomial or trigonometric functions), band edge heaters, band zone cooling, work roll bends, and / or scaffolds with pair-cross function. In addition to the mechanical and thermal profile actuators, the rolling force redistribution for influencing the contour may also be used selectively.

Figur 1

die Profilkontur eines Metallbandes mit zum Verständnis der Erfindung wesentlichen Begriffsdefinition;

Figuren 2.1, 2.2 und 2.3

eine Veranschaulichung des erfindungsgemäßen Verfahrens;

Figur 3

eine erste Möglichkeit zur Reduzierung eines unerwünschten Wulstes am Rand des Metallprofils auf Basis des erfindungsgemäßen Verfahrens;

Figuren 4.1 und 4.2

eine zweite Möglichkeit zur Reduzierung von unerwünschten Wulsten am Rand des Metallbandes; und

Figur 5

die Einstellung der Profilkontur des Metallbandes durch Vorgabe von Zielwerten an mehreren Referenzpositionen

veranschaulicht.The description is a total of 5 figures attached, where

FIG. 1

the profile contour of a metal strip with essential for the understanding of the invention definition of terms;

Figures 2.1, 2.2 and 2.3

an illustration of the method according to the invention;

FIG. 3

a first possibility for reducing an undesirable bead on the edge of the metal profile based on the method according to the invention;

Figures 4.1 and 4.2

a second possibility for reducing unwanted beads on the edge of the metal strip; and

FIG. 5

the setting of the profile contour of the metal strip by specifying target values at several reference positions

illustrated.

The invention will be described in detail below with reference to the said figures in the form of embodiments.

FIG. 1 shows a cross section, ie the profile contour of a metal strip registered in a coordinate system, wherein the abscissa the band width position m and bi and the ordinate a profile value for the profile contour is applied. The coordinate system is designed to the curved profile contour so that it is placed in the middle of the width of the curved profile contour. Positive values for the bandwidth position extend in FIG. 1 to the right and negative values for the bandwidth position extend in FIG. 1 to the left, respectively in the width direction of the metal band. Individual profile values, which are each assigned to specific positions in the width direction of the metal strip, denote the deviation of the profile contour from a rectangular profile contour, as represented by the horizontal abscissa m / bi. The profile values are accordingly ablated perpendicularly from the abscissa and indicated with positive signs. In other words, the profile values describe, in particular, the curvature of the metal strip at a specific bandwidth position in relation to the center of the metal strip. The profile value CL is in FIG. 1 specified with CL = 0, because this profile value forms the origin of the coordinate system.

In FIG. 1 are initially two profile contours to recognize, namely on the one hand a measured profile contour, in FIG. 1 shown as a dashed line. In addition, as a solid line z. B. Predictive profile contour without adaptation, which was calculated using a process model. The predicted profile contour, as in FIG. 1 is not yet adapted in the context of the invention, as will be described below.

The core idea of the present invention is an adaptation of the predicted profile contour or an adaptation of the profile contour values, also called prognostic values C _P (n) bi, of the nth metal band, each at a plurality of bandwidth positions bi with i = 1,2,3 etc. , in FIG. 1 at the positions bi = b1 to b4. The predicted profile contour corresponds to a juxtaposition of calculated profile contour values or the profile contour or prognosis values connected to one another via an approach or interpolation function. Essential for the adaptation according to the invention is the determination of a corresponding adaptation value ΔC (n) bi, which determines the profile deviation, ie the difference between the actual value C _actual (n) bi and the associated prognostic value C _P (n) bi at the plurality of bandwidth positions b1 to b4 describes.

Basically, the bandwidth positions bi are arbitrary positions in the width direction of the metal strip; Usually, latitude positions are defined by their positive or negative distance from the mid-band. In some standardized cases, however, these bandwidth positions can advantageously also be defined by their distance from the respective natural edge of the metal strip on the drive side or / and on the operating side of the metal strip, then respectively in the direction of the strip center. The bandwidth positions thus defined are typically referred to as reference positions. These normalized reference positions are then typically associated with specific profile values, which are then referred to as C40 or C100, for example become. The figure behind the C then corresponds to the distance of the bandwidth position of the respective natural edge of the metal strip.

In FIG. 1 the profile contour is shown over the entire width of the metal strip from the drive side to the operating side. In the following Figures 2 and 5 For reasons of simplification, only the right half of the profile contour of the metal strip is shown in each case. In this half determined adaptation values or differences between predicted and measured profile contour can be assumed at least approximately by mirroring for the left half of the profile contour.

Alternatively, the measured and calculated values for the profile contour can also be obtained by averaging the contour values at the mirror-image positions i = 1, i = -1; i = 2, i = -2; i = 3, i = -3 and / or i = 4, i = -4 are formed on the drive and operator side. Negative index values are only intended to make it clear that this is an opposite page. In this case, a smoothing function is preferably applied by the entire measured band contour in order to suppress any noise of the band contour signals. The calculation of the profile contour and the corresponding adaptation according to the invention can be symmetrical only for one band half or asymmetrically over the entire width.

FIG. 2 illustrates the inventive method for producing a metal strip or in particular for adapting the profile contour of the metal strip.

The Figures 2.1 - 2.3 illustrate the situation using a simplified example. Only a short-term adaptation was used. The aim of the figures is to illustrate the effect of the contour adaptation and the profile adaptation at several reference points 2 bi.

Figure 2.1 first describes the determination according to the invention of the adaptation values on an n-th metal band, shown in simplified form only for the right-hand band half and on the example of only two adaptation points. For the description of Figure 2.1 can on the previous description of the FIG. 1 to get expelled; this applies to the Figure 2.1 alike. Merely as a supplement, it should be mentioned again that the bandwidth positions or the points in the width direction where a calculation of a profile value takes place are generally numbered consecutively with the parameter m, in particular if they are counted from the center of the band CL. The reference positions bi are equally bandwidth positions, which are not defined by the band center but by their distance from the natural edge of the metal band.

Not only in Fig. 2.1 but also in the following figures, the parameter m is also used as an indication of the entire contour or total number of contour calculation points in contrast to the parameter bi, which is to be understood regularly only as an indication of discrete values (reference positions).

The distances of these reference positions bi from the band edge are in Fig. 2.1 and 2.2 as well as 2.3 for the different bandwidths n and n + 1 alike.

Fig. 2.1 illustrates the determination according to the invention of individual adaptation values ΔC (n) b1 and ΔC (n) b2 as the difference between individual prognostic values C _P (n) bi with i = 1 and i = 2 and the actual values C _Ist (n) bi for the profile contour of the nth metal band.

Fig. 2.2 illustrates the determination of an adaptation contour according to the invention. The adaptation contour is determined for the following band n + x. On the band n can z. For example, the width may be different than for band n + x. Only the adaptation values bi at the band n or / and long-term daptation are used Averaging is determined for a number of bands j and used for a following band n + x. The adaptation contour and the point sequence ΔC (n + x) m (with the index m) is always used only in connection with the band n + x.

In Fig. 2.2 and Fig. 2.3 are the in Fig. 2.1 entered adaptation values Δ C (n) b 1 and Δ C ( n ) b 2 entered. They are used there in the simplified example for the following band n + x (with x = 1) for the adaptation contour determination. Therefore, the above adaptation values (x n +) are designated 2 (x = 1) b 1 and Δ C b with Δ C (n + x). In addition to these two adaptation values at the reference positions b1 and b2, another trivial value, in this case the value in the middle of the band, is also used to determine the adaptation contour Figure 2.2 with m = 1, taken into account. The value .DELTA.CL in the center of the band is .DELTA.CL = 0 because the coordinate system has been arranged as passing through this point. The adaptation values were determined at points b1 and b2 on band n and used for band n + 1 (x is = 1).

The adaptation contour ΔC (n + 1) m for the n + 1'th metal band then results, as in Figure 2.2 shown as at least piecemeal approach or interpolation function through the band center CL = 0 and the two mentioned adaptation values and at the reference points C100 and C25, the latter two being measured as the distance from the natural edge of the metal strip.

The formation of an approach or interpolation function and the interpolation between the band center and the reference point b1 and the corresponding formation and interpolation between the reference point b1 and the reference point b2 can basically be done separately and independently of each other in the respective bandwidth sections. To make a kink at a transition point of two interpolation functions, in Figure 2.2 For example, at position b1, the additional condition that the two adjacent partial interpolation functions are continuous at the transition point is filled in with the formulation of the two partial interpolation functions must be differentiable, ie in particular that the respective functions must have the same slopes there. This procedure is basically carried out for all adaptation areas in the width direction of the metal strip. In this example (symmetrically), the adaptation contour starts at the center of the tape CL with a horizontal tangent.

From the last adaptation value, in Figure 2.2 at the reference position i = 2, up to the edge point m _{max of} the metal strip, where no profile value is specified, the adaptation contour can be determined by extrapolation. The interpolation or extrapolation is used to interpolate or extrapolate on the profile values at other bandwidth positions m based on the given profile values at the reference positions.

Figure 2.3 illustrates how the previously according to Figure 2.2 For the n + 1'te metal strip determined adaptation contour can now be considered in the forecast and subsequent production to be rolled n + 1'ten metal strip.

Figure 2.3 shows, inter alia, the calculated adapted profile contour C _p (n + 1) m and the calculated adapted predicted values C _P (n + 1) b1 and C _P (n + 1) b2 and a related calculated predicted profile contour C _P (n + 1) m _OA , with oA: without adaptation, here by way of example for the n + 1'th metal strip, ie here as an example for the next metal strip to be rolled.

The previously according to Figure 2.1 Adaptation values ΔC (n) b1 and ΔC (n) b2 determined for the nth metal band can be added to the prediction values at the corresponding reference positions in order to obtain improved adaptive prognosis values for the predicted adapted profile values or profile contour there.

Alternatively or additionally, the previously according to Fig. 2.2 for the n + 1'te metal band determined adaptation contour .DELTA.C (n + 1) m to the determined for the n + 1'te metal strip predicted profile contour C _p (n + 1) m _OA are added in order in this way to obtain a correspondingly improved or adapted profile contour C _p (n + 1) m; see also claim 9.

The new adapted prognosis values or the new profile contour obtained in this way can advantageously be used to set the profile actuators even more precisely with respect to desired target values and / or target contours in the production of the n + 1'th, in general the n + x'ten metal band to be able to.

mit

C_P(n+1)m

korrigierte bzw. adaptierte Profilkontur des n+1'ten Metallbandes über der Bandbreite m;

C_P(n+1)m_OA

eine berechnete bzw. prognostizierte Profilkontur des n+1'ten Metallbandes über der Bandbreite m ohne Adaption;

ΔC(n+1)m

Adaptionskontur: Werte der Adaptionskontur an der Position m für das Metallband n+1

$$\mathrm{m}=1\dots {\mathrm{m}}_{\max }.$$

Expressed mathematically, the adapted band contour values or the adapted band contour for the n + 1 metal band to be rolled, for example, are calculated according to the following formula: $${\mathrm{C}}_{\mathrm{P}}\left(\mathrm{n}+1\right){\mathrm{m}}_{\mathrm{OA}}+\mathrm{\Delta }\mathrm{C}\left(\mathrm{n}+1\right)\mathrm{m}={\mathrm{C}}_{\mathrm{P}}\left(\mathrm{n}+1\right)\mathrm{m}$$

With

C _P (n + 1) m

corrected or adapted profile contour of the n + 1'th metal band over the bandwidth m;

C _P (n + 1) m _OA

a calculated or predicted profile contour of the n + 1'th metal band over the bandwidth m without adaptation;

.DELTA.C (n + 1) m

Adaptation contour: values of the adaptation contour at the position m for the metal strip n + 1

$$\mathrm{m}=1...{\mathrm{m}}_{\mathrm{Max}},$$

The width position m may also be reference positions bi.

The difference or adaptation .DELTA.C (n) m between measured and calculated correction is at the in Figure 2.2 shown example for ease of description / illustration shown only for a metal band. As a rule, this difference is formed on the last rolled metal strip and / or on the penultimate rolled metal strip and / or on a plurality of metal strips of the same type, optionally with different weighting, and in this way a sum adaptation value is determined.

FIG. 3 shows an application example for the use of the contour adaptation according to the invention for reducing or avoiding unwanted beads in the edge region of a metal strip. At this first in FIG. 3 Shown embodiment shown, the reduction of the beads by a targeted increase in a value for the profile contour at a reference position, in FIG. 3 Position C40, ie 40 mm away from the natural edge of the metal strip.

Without the use of contour adaptation, it may happen that bands with supposedly normal profile contours are calculated or predicted; see the dashed output contour after the first calculation step without contour adaptation in FIG. 3 , After carrying out the invention and previously with particular reference to Figure 2.3 described contour adaptation by addition of the predicted for the band n + x profile contour and a determined for a previous band adaptation contour can according to the invention in FIG. 3 shown adapted profile contour C _P (n + x) m are determined for the n + x'te metal strip. The advantage of the profile contour C _P (n + x) m adapted according to the invention over the non-adapted predicted profile contour C _P (n + x) m _OA is in FIG. 3 clearly recognizable, because the adapted profile contour makes the unwanted bead with the bead height W1 in the edge region of the metal strip detect in the first place; the unadapted predicted profile contour (dashed line) did not show the bead so clearly. In this respect, the profile adaptation according to the invention provides an improved calculation result for determining a more accurate profile contour and opens up new possibilities for improving the profile contour, here in particular for reducing the bead height. For example, for the metal strip according to FIG. 3 calculates an edge bead height W1, which is higher than a threshold value for an allowable bead height, then the process model within given allowable limits z. B. C40 target _min and C40 target _max the profile value at the corresponding edge position, here 40 mm from the natural edge of the metal strip away, automatically set to a new value, raised here, so that the maximum allowable bead height is not exceeded or reduced. By said increase of the predetermined profile value by the amount Δ P is reduced in the in FIG. 3 shown example, the bead height of W1 to W2.

Alternatively or in addition, for the same conditions and the same profile contour as in FIG. 3 With the use of the adapted profile contour for controlling the bead height, a raised force level within the limits of the process and equipment limits in the rear stands of a finishing train or in a reversing stand in the later back stitches can be used. This can be achieved by a rolling force redistribution, ie a relief of the front scaffolds or earlier stitches and a greater load on the rear scaffolding or later stitches and / or by driving up one or more scaffolds (last scaffold or last stitch or scaffolding within the finishing train or middle stitch) happen. Figure 4.1 shows examples of advantageous Walktkraftumverteilungen to the bead height W1 (see Figure 4.2 ) to reduce. An iteratively determined higher load in the rear scaffold increases the work roll flattening. As a result, the bead W2 reduces or disappears after the rolling force redistribution, see the dashed line in FIG Figure 4.2 (2nd billing step). The mechanical profile actuators are adapted in the iterative calculation process this new boundary conditions and the z. B. C40 destination profile set.

The knowledge of the expected profile contour due to the physical modeling of the relationships and the said adapted profile contour a plurality of width positions bi across the width of the metal strip is further actively utilized to assist in setting a nominal strip profile at the strip edge, e.g. B. at position C25, in addition, the band profile in the band center area - expressed by CBody or C100 - in allowable minimum and maximum limits C100 _min , C100 _max to hold, as is an example in FIG. 5 is shown. In an advanced profile presetting advantageously additional process limits are introduced and the minimum and maximum band profile limits for multiple band contour points, eg. C25 and C100. The improved result (2nd calculation section) represents the band contour with the solid line.

Claims (24)

1. Method for producing metal strips in a rolling installation with a desired profile contour, comprising the following steps:

   a) presetting a target value for the profile contour at at least one reference position bi in width direction for at least an nth metal strip;

   b) simulating a rolling process at the rolling installation for producing the metal strip with the aid of a process model, wherein setting values for profile setting elements and a prognosis value C_P(n)bi for the profile contour of the nth metal strip at the reference position bi are so calculated that the target value is achieved as far as possible - sofar as present - with consideration of old adaptation values at the reference position bi and possible restrictions;

   c) setting the profile setting elements by the calculated setting values;

   d) rolling the nth metal strip;

   e) measuring the actual value C_ist(n)bi of the profile contour of the rolled nth metal strip at the reference position bi; and

   f) determining a new adaptation value ΔC(n)bi on the basis of the difference between the actual value C_ist(n)bi and the prognosis value C_P(n)bi for the profile contour of the nth metal strip at the reference position bi;
   characterised in that
   the steps a), b) and c) are carried out before the rolling of the at least nth metal strip for a plurality I, wherein I ≥ 2, of reference positions bi, wherein 1 ≤ i ≤ I in at least one width section of the at least nth metal strip and that the steps e) and f) are carried out after rolling of the at least nth metal strip for the plurality I of reference positions bi in order to determine the new adaptation values ΔC(n)bi at the plurality I of the reference positions bi in the at least one width section of the at least nth metal strip; and

   g) during later production of a further longitudinal section of the nth metal strip or of an n + xth metal strip, wherein x = 1, 2, etc., at least the steps a) to d) are repeated, wherein n = n + x, wherein the new adaptation values ΔC(n)bi determined previously according to step f) at least for the nth metal strip are taken into account for the plurality I of the reference positions bi in the calculation of the settings of the profile setting elements and the calculation of the prognosis values according to step b) for the n + xth metal strip as old adaptation values.
2. Method according to claim 1, characterised by determination of the new adaptation values ΔC(n)bi according to step f) at the reference positions bi of the nth metal strip at least partly in the form of a short-term adaptation value ΔC_K(n)bi according to the following formula: $$\mathit{\Delta C}\left(n\right)\mathit{bi}={\mathit{\Delta C}}_K\left(n\right)\mathit{bi}={\mathit{\Delta C}}_K\left(n-x\right)\mathit{bi}+\left[C_{\mathit{Ist}}\left(n\right)\mathit{bi}-C_P\left(n\right)\mathit{bi}\right]$$

   

   wherein

   K: short-term adaptation;

   x=1,2,3...;

   ΔC_K(n-x)bi: old short-term adaptation value;

   C_ist(n)bi: measured actual value for the profile contour of the nth metal strip at the reference position bi; and

   C_P(n)bi: calculated prognosis value or calculated strip profile.
3. Method for producing metal strips in a rolling installation with a desired profile contour, comprising the following steps:

   a) presetting a target value for the profile contour at at least one reference position bi in width direction for at least an nth metal strip;

   b) simulating a rolling process at the rolling installation for producing the metal strip with the aid of a process model, wherein setting values for profile setting elements - sofar as present with configuration of old adaptation values at the reference position bi and possible restrictions - are so calculated that the target value is achieved as far as possible;

   c) setting the profile setting elements by the calculated adjustment values;

   d) rolling the nth metal strip;

   e) measuring the actual value C_ist(n)bi of the profile contour of the rolled nth metal strip at the reference position bi;

   e') calculating a recalculated prognosis value C'_P(n)bi for the profile contour of the nth metal strip at the reference position bi on the basis of the rolling installation conditions and current processing positions, such as they were present during rolling of the nth metal strip according to step d); and

   f) determining a new adaptation value ΔC(n)bi on the basis of the difference between the actual value C_ist(n)bi and the recalculated prognosis value C'_P(n)bi for the profile contour of the nth metal strip at the reference position bi;
   characterised in that
   the steps a), b) and c) are carried out before the rolling of the at least nth metal strip for a plurality I, wherein I ≥ 2, of reference positions bi, wherein 1 ≤ i ≤ I, in at least one width section of the at least nth metal strip and that the steps e), e') and f) are carried out after rolling of the at least nth metal strip for the plurality I of reference positions bi in order to determine the new adaptation values ΔC(n)bi at the plurality I of the reference positions bi in the at least one width section of the at least nth metal strip; and

   g) during later production of a further length section of the nth metal strip or of an n + xth metal strip, wherein x = 1, 2, etc., at least the steps a) to d) are repeated, wherein n = n + x, wherein the new adaptation values ΔC(n)bi determined previously according to step f) at least for the nth metal strip are taken into account for the plurality I of the reference positions bi in the calculation of the settings of the profile setting elements and in the calculation of the prognosis values according to step b) for the n + xth metal strip as old adaptation values.
4. Method according to claim 3,
   characterised by
   determination of the new adaptation values ΔC(n)bi according to step f) at the reference positions bi of the nth metal strip at least partly in the form of a short-term adaptation value ΔC_K(n)bi according to the following formula: $$\mathit{\Delta C}\left(n\right)\mathit{bi}={\mathit{\Delta C}}_K\left(n\right)\mathit{bi}={\mathit{\Delta C}}_K\left(n-x\right)\mathit{bi}+\left[C_{\mathit{Ist}}\left(n\right)\mathit{bi}-C{\prime }_P\left(n\right)\mathit{bi}\right]$$

   

   wherein

   K: short-term adaptation,

   x=1, 2, 3 ...;

   ΔC_K(n-x)bi: old short-term adaptation value;

   C_ist(n)bi: measured actual value for the profile contour of the nth metal strip at the reference position bi; and

   C_P(n)bi: recalculated prognosis value or recalculated strip profile.
5. Method according to any one of the preceding claims,
   characterised by
   determination of the new adaptation values ΔC(n)bi according to step f) in claim 1) or 3) at the reference positions bi at least partly in the form of long-term adaptation values ΔC_Lbi by carrying out the following steps:

   determining the adaptation values by repeating the steps a) to f) according to claim 1 or 3 at the plurality I of reference positions bi for a plurality of metal strips, which are rolled before the n + xth metal strip, of an adaptation group; and

   calculating the long-term adaptation values ΔC_Lbi by formation of the mean values of the adaptation values or formation of the mean values of the differences between actual values and prognosis values for the profile contour for the plurality of metal strips in each instance at one of the reference positions bi.
6. Method according to claim 2, 4 and 5,
   characterised by
   determination of the adaptation values ΔC(n)bi according to step f) each time in the form of the sum adaptation value ΔC_s(n)bi as a sum of the short-term adaptation value ΔC_K(n)bi and the long-term adaptation value ΔC_Lbi for use for the metal strip n + x.
7. Method according to any one of claims 2, 4, 5 and 6,
   characterised by
   determination of the adaptation value ΔC(n)bi according to step f) and/or use of the adaptation value ΔC(n)bi in the form of a short-term adaptation value, long-term adaptation value or sum adaptation value, the value being weighted by a weighting factor g, wherein 0 ≤ g ≤ 1, or by a weighting function.
8. Method according to any one of the preceding claims,
   characterised by
   determination of an adaptation contour ΔC(n+ x)m for the n + xth metal strip in the form of a set-up function, which is preferably conducted by the adaptation values, which are determined at the at least nth metal strip, at at least two of the reference positions bi and preferably additionally by at least one further calculation point - calculated/predetermined by the process model - at at least one further strip width position m.
9. Method according to claim 8,
   characterised by
   determination of an adapted profile contour C_P(n + x)m for the n + xth metal strip by the addition of a non-adapted calculated profile contour C_P(n+x)m_oA - forecast by the process model - for the n + xth metal strip and the calculated adaptation contour ΔC(n+x)m for the n + xth metal strip.
10. Method according to claim 8 or 9,
    characterised in that
    determination of the adaptation contour or of the adapted profile contour for ≥ 2 width sections of the metal strip is carried out, wherein the first width section of the metal strip lies in, for example, the central region and the second width section or further width sections lies or lie in, for example, the edge region of the metal strip.
11. Method according to claim 34,
    characterised in that
    in the case of two sections adjoining one another in width direction the adaptation contour or the adapted profile contour over the two width sections is preferably selected so that the contour courses can be continuously differentiated at the boundary of one strip section to the other strip section, in particular so that they have the same gradients.
12. Method according to one of claims 10 and 11,
    characterised in that
    the set-up function is formed over at least one of the width sections from a linear function, a polynomial function, an exponential function, a trigonometric function, a spline function or a combination of different functions.
13. Method according to claim 12, characterised in that the set-up functions are different for the different adjacent width sections.
14. Method according to claim 8 or 9,
    characterised in that
    the adaptation contour or the adapted profile contour over one width section of the metal strip is extrapolated to an adjacent width section for determination of an extrapolated adaptation contour or an extrapolated adapted profile contour over the adjacent width region.
15. Method according to any one of the preceding claims,
    characterised in that
    instead of the measured actual value C_ist(n)bi of the profile contour of the metal strip at the reference position bi a mean value from the measured actual values at the mirror-image reference positions bi on the righthand half and lefthand half of the metal strip seen in rolling direction is used.
16. Method according to one of claims 1 and 9,
    characterised in that
    the prognosis values C_P(n + x)bi or/and the adapted profile contour C_P(n + x)m is or are initially determined for only one strip half, for example the strip half on the control side, and subsequently mirrored for the other strip half, for example the strip half on the drive side, at the strip centre plane extending in longitudinal direction of the metal strip.
17. Method according to any one of the preceding claims,
    characterised in that
    the measured actual value C_ist(n)bi of the profile contour is used as a direct measured value at the reference position bi or as a profile measurement value smoothed by an equalising function.
18. Method according to any one of claims 9 to 17,
    characterised in that
    the adapted profile contour C_P(n+x)m is analysed with respect to profile anomalies, such as strip beads or steep edge drops, particularly in the edge region of the metal strip.
19. Method according to claim 18,
    characterised in that
    when calculated strip beads are present the adapted profile contour C_P(n+x)m is iteratively improved by means of the process model by successively increasing a value of the profile contour at at least one of the reference positions bi within the scope of the allowable profile setting limits and by appropriate resetting of the profile setting elements in order to reduce the height of the strip bead.
20. Method according to claim 18,
    characterised in that
    calculated strip beads are reduced or avoided by increasing the load in the last roll stand (run-out stand) or the last roll stands of a rolling train or in the last rolling passes of a stand of the rolling installation by redistributing the load from the front to the rear or by deselecting at least one roll stand or rolling pass within the scope of the processing and installation limits.
21. Method according to any one of the preceding claims,
    characterised in that for production of the n + xth metal strip:

    the profile setting elements are so set in step b) that the target values predetermined for a plurality of reference positions bi or calculated prognosis values C_P(n + x)bi for the profile contour are achieved within allowable minimum or maximum profile limits; or

    the profile setting elements are so set in step b) that the target value predetermined for a reference position bi is achieved or the deviation from the target value is minimal and at the same time the strip profile is maintained at at least one further strip width position within allowable minimum or maximum profile limits.
22. Method according to any one of the preceding claims,
    characterised in that
    the determined adaptation values at the positions bi and/or the adapted profile contour and/or the adaptation contour in the process model is or are taken into account - particularly transferred to the preceding rolling passes or stands with weighting factors or transfer functions - for calculation of the intermediate stand contours or intermediate pass contours of the front stands or the preceding passes and for optimised adjustment of the profile setting elements.
23. Method according to any one of the preceding claims,
    characterised in that
    the reference position bi is defined by way of its spacing from the edge of the metal strip.
24. Method according to any one of the preceding claims,
    characterised in that
    for the setting of the target contour, with use of the strip contour adaptation, the following profile setting elements are employed: variable working roll cooling systems or zonal cooling means or local roll heating means for influencing the thermal crown and/or working roll displacements in conjunction with roll grinding (special roll grinds for combatting strip beads or strip edge drops, "tapered roll", CVC rolls, CVC rolls with a grind of higher order or polynomial nth order or trigonometric functions), strip edge heating means, strip zone cooling means, working roll bending means and/or stands with a roll pair cross function.

Images