DE19912796A1 - Flatness control to achieve flat cold strip - Google Patents

Abstract

The invention relates to a method for rolling a strip in a strip rolling mill which has at least two roll stands or an individual stand with upper and lower, optionally adjustable working rolls respectively, said rolls being optionally supported indirectly on support rolls or with intermediate rolls. At least one pass is rolled in said rolls in order to modify the state of the rolled strip. A target distribution of stresses or any target form of unevenness is predetermined for the rolled strip and compared with the actual distribution of stresses and the mechanical or physical control elements are employed in such a way as to minimise the difference between the predetermined and actual distribution of stresses as far as possible. In addition to the actual distribution of stresses in the hot strip, the distribution of temperature over the width of the strip is determined and used to calculate the distribution of stresses which would be present after the strip has cooled if the strip were free of stresses in the hot state with the distribution of temperature that exists behind the roll gap and this temperature-induced distribution of stresses is used to correct the actual stress before this is compared with the target form of unevenness.

Description

The invention relates to a method for rolling a rolled strip in a Belt mill with at least two roll stands or a single stand with each upper and lower, if necessary, adjustable work rolls, which may be immediate or Support on intermediate rollers on support rollers and in which to change the state of the rolled strip is rolled at least one pass, one for the rolled strip Target stress distribution or any target unevenness form specified and with the actually achieved voltage distribution is compared and available standing mechanically or physically effective actuators used in the way be that the difference between predetermined and factually achieved Stress distribution is minimized as much as possible.

The flatness measurement of cold rolled strip is usually done using Flatness measuring rollers, which show the current tension distribution over the bandwidth measure up. The measurement is usually carried out on the outlet side of the roll stand, where average strip temperatures between 80 and 160 ° C occur. This applies analogously also for other metals such as aluminum or copper.

Experience has shown that temperature gradients of the order of 10 ° -30 ° C occur between the strip center and the strip edge. For this purpose, two strip temperature distributions measured during the reversing rolling of stainless steel are plotted in FIG. 1 as an example. Assuming the thermal expansion coefficient of steel of about 1.2 10 ^-5 / degrees and an elastic modulus of 210 000 N / mm ² , a temperature difference of 10 degrees already induces a tension difference of about 25 N / mm ² in the cold-rolled steel strip; However, the aim is usually to keep the voltage differences across the bandwidth below 20 N / mm ² . Without flatness control, this temperature-induced tension would have no influence on the flatness of the end product, the cooled strip, at least not if it remains in the elastic range.

With the flatness control, the contour of the roll gap is to be regulated in such a way that the strip flatness or a specific non-flatness shape is maintained or achieved. One problem, however, is that the measured variable is the stress distribution in the hot band, which is associated with the temperature-induced stress described (in the example above, more than 25 N / mm ² ). This means that the known flatness control with a target flatness or target function for flat strip will also regulate this temperature-induced tension in the warm strip. After the temperature difference has cooled and subsided, a voltage component corresponding to the temperature-induced voltage is additionally superimposed in the cooled band of the target unevenness form, but with the opposite sign.

The target stress distributions or target non-flatness forms could be for one given product together with a given rolling or special In principle, cooling strategy can also be specified in such a way that an adequate one Flatness of the cooled product is preserved. However, the search for the required objective functions or unplanned forms an elaborate empirical Process that involves unwinding the federal government with sampling for each product makes necessary.

It would be more technologically satisfactory, no longer between the measurable Flatness in the warm state immediately behind the roll gap and for different products and rolling strategies different forms of cold flatness to distinguish cooled bundles.

In 1973 and 1975, in the absence of one not yet developed at the time direct measurement of the belt tension using tension measuring rollers and subsequent regulation or measurement of the local strip elongation proposed been to measure the strip temperature distribution itself (DE-A-23 49 611 and DE-A-25 54 246). Differences in the strip temperature across the width become more direct there Way interpreted as a difference in the degree of deformation across the width and this Difference in the degree of deformation in turn as the cause of flatness errors. By now the  Differences in temperature in the strip directly in differences in Coolant exposure to the roller should be converted, the roller contour and thus the local degrees of deformation are changed.

This process is at least incomplete because of temperature differences occur not only because of different degrees of deformation, but also others There are different causes, like thermal roller profiles Coolant flows and the larger surface / volume ratio at the Band edge especially in the wound state, which causes the temperature to Band edge usually falls off. In contrast, the degree of deformation when rolling is thin steel strips usually because of the embedding of the rolling stock in the roller higher at the belt edge, and yet the temperature at the edge is because of the the effects just described are lower overall.

That described in the patent applications DE-A-23 49 611 and DE-A-25 54 246 Accordingly, the procedure did not prevail, but was later implemented the described direct measurement of the tension distribution with the following Regulation replaced. One of the actuators of this regulation is still the local one Control the amount of coolant on the work roll, which is usually the residual error should minimize that of the other actuators such as roll bending, displacement or the inflatable roller cannot be covered.

In the company brochure "ABB Automation and Drives (document 3BSE005258) dated February 17, 1994, page 3", the technical status of the flatness control at this time is summarized as follows:

• 1. Definition of target flatness
• 2. Measure the flatness of the strip
• 3. Calculation of the flatness error
• 4. Description of the function of each mechanical roll gap actuator by means of Influence functions
• 5. Best possible overlay of the influencing functions to avoid the flatness error eliminate
• 6. Use the roll gap actuators to adjust the roll gap  
• 7. Use the locally adjustable roller cooling for the selective adjustment of the Roll gap through thermally induced work roll Changes in diameter.

This state of the art can also be found in the European patent application EP 0850704 A1 supplemented by the note that to achieve a cold condition plan the tape with the warm tape different over the tape lengths Unplanarity forms can be specified. No statements can be found at the state of the Technology about what the exact cause of a ribbon unplanned when cold is, as described above, the state of the art Control concept itself, in interaction with inhomogeneous temperature distributions about the bandwidth. Likewise, no statements can be found about what is next to against it simply empiricism ("different forms of unplanarity in the warm band") can.

The aim of the present invention is to find a method for flatness control, with that no longer directly between the measurable flatness in the warm state behind the roll gap and for different products and rolling strategies Different cold unevenness forms of the cooled coils can be distinguished must and with the same flatness or non-flatness in the warm state can be specified.

To solve the problem it is provided that in addition to the actual Tension distribution in the hot band also the band temperature distribution over the Bandwidth is recorded and the voltage distribution is calculated from this, which would occur after the tape had cooled if the tape was in the warm condition with the temperature distribution behind the roll gap would be stress-free, and that this temperature-induced voltage distribution for Correction of the voltage actually reached is used before this with the Target flatness is compared.

The advantage of this proposal according to the invention is that the Formulation of the target function for flatness can be done in such a way that on the best possible flatness of the cooled bundle is achieved directly. in the  Contrary to the intent of DE-A-23 49 611 and DE-A-25 54 246 uses the Invention the temperature measurement not as a directly linked to the flatness Measured variable (this task is performed by the tension measuring roller), but according to the technological knowledge of the causes of unplanarity as one Supplementary correction variable for the tensile stress in the warm, which can be measured directly Strip, so that the control system specifically cooled cold strip in each individual case can achieve defined flatness.

Embodiments of the method according to the invention are in the subclaims specified.

The invention is explained below with reference to the drawing figure 2. The strip is computed into strips in the direction of travel, conveniently, but not necessarily with a strip width as specified by the tension measuring roller used. In many technically executed cases this is approx. 52 mm and usually also corresponds to the division of the nozzles on the chilled beam. For each of these strips i, an average temperature T _{i is} determined from the temperature signals. If the temperature signals are not given in a - in the given example - 52 mm division, an interpolation to this strip division must take place beforehand. In addition, an average temperature T _{m is} calculated over the entire bandwidth. On the outlet side, the difference between the strip temperatures T _{Ai and} the average temperature T _Am is denoted by ΔT _Ai

ΔT _Ai = T _Ai -T _Am (1).

By multiplying the values ΔT _Ai by the thermal expansion coefficient of the rolling stock α on the outlet side, an imaginary relative change in length ΔL _Ai / L _{Am is obtained}

ΔL _Ai / L _Am = α.ΔT _Ai (2)

this strip compared to the average length L _{Am of} all these strips. However, since these strips are connected, they cannot take up this change in length, but are kept under tension (approximately) on the length L _Am , which specifies the mean strip temperature. This strip- _related tension σ _TAi is then

σ _TAi = -E.ΔL _Ai / L _Am (3),

with E as the modulus of elasticity of the rolling stock.  

The sign arises from the convention that tensile stresses with positive and compressive stresses have a negative sign.

If any tension gradients are corrected by the actuators in the warm band, the tension σ _a remains in the cooled band.

For temperature-independent material values α and E, this is the negative expression of (3):

σ _ai = E.ΔL _Ai / L _Am = E α ΔT _Ai (4).

The formulas (2) to (4) give a first approximation for the conversion from the temperature distribution to the voltage distribution, because:

• - The material values α and E are not completely independent of temperature and
• - The individual strips are not completely drawn back to the length L _m given by the mean temperature, but in principle remain at slightly different lengths L _i ≠ L _m , which leads to somewhat reduced stresses than calculated according to (4).
  In order to calculate this length contour L _i , a method similar to that described in the chapter "Thermoelastic coupling" on page 20 of the article Thermal camber model for hot and cold rolling for a cylindrical roller can be selected. The resulting formula (26) for the cylindrical case shows that the radial expansion contour R (z) disappears when R becomes very large. Similarly, the longitudinal band length contour L _i disappears in our case when the band length L _m becomes very large, which makes the calculation according to (1) - (4) practically sufficient.

In comparison to empirically determined target functions, which can implicitly contain a thermally induced voltage component, the method according to the invention has further advantages:
By eliminating the inspection of the cooled coil, the determination of target unevenness for the hot strip is considerably simplified. Scattered cold unevenness due to temperature fluctuations are avoided. Differences in target unevenness for different products are reduced. When using the new cooling strategies for the belt, the target unevenness need not be adjusted. Bundles that were previously rolled in cold operation and consequently cool down on the order of a day after each pass can be rolled further after a shorter period of time, provided that the permissible total temperature is not exceeded. It also applies here that the new procedure makes the goal of unplannedness independent of the waiting time between stitches.

Claims (7)

1.Method for rolling a rolled strip in a strip mill with at least two roll stands or a single stand with upper and lower optionally adjustable work rolls, which may be supported directly or via intermediate rolls on support rolls and in which at least one pass is rolled to change the condition of the rolled strip , whereby a target tension distribution or any target unevenness form is specified for the rolled strip and compared with the actually achieved tension distribution, and available mechanically or physically effective actuators are used in such a way that the difference between the specified and actually achieved tension distribution is minimized as far as possible , characterized in that, in addition to the actual tension distribution in the warm strip, the strip temperature distribution over the strip width is also recorded, and from this the tension distribution is calculated, which is obtained after cooling of the strip would stop if the strip were tension-free in the warm state with the temperature distribution behind the roll gap, and that this temperature-induced tension distribution is used to correct the tension actually reached before comparing it with the target flatness shape. 2. The method according to claim 1, characterized, that the temperature-induced voltage distribution from the factual Subtracted stress distribution, the resulting stress distribution with the shape of the target is compared, the difference is calculated from it and the Available mechanically or physically effective actuators of the type be used so that the difference is minimized as much as possible.   3. The method according to claim 1, characterized, that the temperature-induced stress distribution to the target flatness shape is added, the factually determined flatness distribution with this Sum distribution is compared, the difference is calculated from it and the Available mechanically or physically effective actuators of the type be used so that the difference is minimized as much as possible. 4. The method according to any one of claims 1 to 3, characterized, that in the sense of a technological compromise of several aspects the Flatness control the temperature-induced stress distribution before Further processing is multiplied by a gain factor. 5. The method according to claim 4, characterized, that in terms of a technological profile weighting, this gain factor for each strip has a different value. 6. The method according to any one of claims 1-5, characterized, that the strip temperature distribution behind the roll gap is measured, preferably detected without contact. 7. The method according to any one of claims 1-5, characterized, that the strip temperature distribution behind the roll gap from a calculation model is derived.