JPH10175007A - Method for controlling roll gap in rolling mill - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain the control method of a roll gap taking backlash which varies every rolling chance into consideration using separated components by separating the change of the roll gap which is generated at the time of rolling into a linear component which continuously varies by rolling load and a nonlinear component which intermittently varies by the mechanical backlash of a rolling mill. SOLUTION: This method is constituted so that a screw-down location detecting sensor 13, load cell 16 and backlash detecting sensor 17 are connected to a controller 18 and the prescribed process is executed to each measured value in the controller 18. The basis of the control system is the formula ΔS=(GO+G)+F(ΔP, MO, CT, CH, CG). (Where, GO: the standard value of backlash amount, G: the detected value of backlash amount, ΔP: the deviation of the tetected value of the rolling load from the calculated value, MO: the standard value of a mill constant, CT: the constant related to a rolling material, CH: the constant related to the rolling mill, CG: the constant related time series change, F: the function for determining the controlled variable S of the roll gap regarding the deviation of the rolling load). In this way, the change in the roll gap which is generated at the time of rolling a sheet is controlled in the optimum.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method for controlling a roll gap in a rolling mill.

[0002]

2. Description of the Related Art A conventional roll gap control method is based on input information (specifically, input side thickness, sheet width, and sheet temperature) of a finishing mill so as to satisfy a predetermined product specification. The roll gap of the stand is set and calculated.
The method of this setting calculation is to determine the exit side target thickness of each stand and the exit side passing speed of the final stand, and then the sheet temperature, advance rate, roll peripheral speed, when rolled material bites into each stand,
The deformation resistance and the rolling load are sequentially calculated, and finally the rolling position is calculated using a gauge meter formula.

The general form of the gauge meter type is expressed as h = S + P / M + α. Here, h: outlet plate thickness, S: roll gap at no load, P: rolling load, M: mill stiffness coefficient, α:
Correction value. In such a roll gap control method, a method of calculating a rolling position from an actually measured rolling load during sheet rolling using a relational expression between a rolling position and a load obtained in advance, and controlling a deviation from a target, For example, Tokuhei 2
-42001.

As disclosed in Japanese Patent Publication No. Sho 63-29606, there is a method of calculating a roll gap by a gauge meter method for each of the work side and the drive side of a rolling mill.

[0005]

By the way, a rolling mill is an assembly of many mechanical parts, and although the mechanical accuracy is limited, the conventional gauge meter type requires a rolling machine for every rolling chance. Fluctuating mechanical gaps (hereinafter referred to as "play") were not considered.

That is, in a work roll shift type rolling mill in which a work roll is shifted in the axial direction, the position of the backup roll in the axial direction does not change, but the position of the work roll in the axial direction changes every rolling opportunity. Change.
Therefore, when there is play in the rolling mill, the play amount also changes for each rolling chance. However, in the related art, no consideration is given to the amount of backlash that changes for each rolling chance.

That is, in the method disclosed in Japanese Patent Publication No. 42001/1990, the roll gap position is set to true, and the rolling load error is calculated as the sheet deformation resistance difference. There is no consideration for the play. In Japanese Patent Publication No. 42001/1990, S = h-MS ₁ -MS ₂ + OF-GC is shown as a gauge meter formula, and “gauge meter constant GC” is introduced in the formula. The gauge meter constant GC can be learned from the results of the wear and shrinkage of the roll which changes in time series, and the influence of the mechanical play which changes for each rolling chance cannot be completely absorbed. Was something.

In the above-mentioned Japanese Patent Publication No. Sho 63-29606, the gauge meter type disclosed as h = S + f (F, B,...) + Α is of the same type as the above, and the "learning term" “α” is also a term that is effective only for those that change in a simple time series, and does not take into account the play that changes for each rolling chance.

[0009] In addition, "Third Edition" Steel Handbook "Volume III (2)
The Japan Iron and Steel Institute, November 20, 1980, page 1308, left column, describes a technique for actually measuring backlash. However, this backlash measurement is for a change in diameter, and is measured for each rolling chance. It did not take into account the changing play. Therefore, in the conventional method using the gauge meter type, the play due to the diameter change is taken into account, but the error of the roll gap due to the play which changes every rolling chance has not been properly evaluated.

Accordingly, an object of the present invention is to provide a method of controlling a roll gap in consideration of a play that changes for each rolling chance.

[0011]

In order to achieve the above object, the present invention takes the following measures. That is, the feature of the present invention is that the change in the roll gap that occurs during rolling is separated into a linear component that continuously changes by the rolling load and a nonlinear component that changes intermittently due to the mechanical play of the rolling mill. In controlling the roll gap.

That is, control is performed as ΔS = G + F.
Here, ΔS: roll gap control amount, G: a non-linear component that changes intermittently due to mechanical play of the rolling mill, and F: a linear component that changes continuously due to the rolling load. More specifically, the rolling load during rolling, the rolling position, the detection value of the play detection sensor is measured, and the relational expression between the rolling position and the load determined in advance, the relational expression of the detection value of the play detection sensor and the load is obtained. Is used to determine and control the play during the rolling.

[0013]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows the configuration of a rolling mill for implementing the method of the present invention. The housing 1 of the rolling mill is provided with a pair of upper and lower work rolls 2, 3, and upper and lower backup rolls 4, 5 for supporting the work rolls 2, 3. The left and right ends of the upper and lower backup rolls 4, 5 are supported by upper and lower chocks 6, 7, respectively. The chocks 6, 7 are supported by the housing 1 so as to be vertically movable. The vertical movement of the upper chock 6 is controlled by a rolling-down device 10 including a hydraulic jack 8 and a screw 9. Lower chock 7
Is supported on the housing 1 by a multi-stage liner 11.

The upper and lower work rolls 2 and 3 are shift rolls which can be moved in the axial direction, and a bender device (not shown) is provided at both ends thereof. Shape control is possible. The rolling-down device 10 is provided with a rolling-down position detecting sensor 13 for detecting a rolling-down position. The sensor 13 includes a top hat sensor 14 for detecting a screw position and a sensor 15 for detecting an oil column value of the hydraulic jack 8. further,
The rolling device 10 includes a load cell 16 for measuring a rolling load.
Is provided.

[0015] Between the lower chock 7 and the housing 1,
A play detection sensor 17 for measuring the vertical displacement of the lower chock 7 is provided. FIG. 2 shows details of the play detection sensor 17. That is, a non-contact range finder is fixed to the housing 1, and is configured to detect a vertical displacement of the lower chock 7. The play detection sensor 17 and the rolling position detection sensor 13 are provided on both the work side (WS) and the drive side (DS) (both sides in the axial direction of the roll).

The rolling position detecting sensor 13 and the load cell 1
6. The play detection sensor 17 is connected to the control device 18 so that each measured value is subjected to predetermined processing in the control device 18. FIG. 3 shows actual measurement data of the backlash detection sensor 17 during plate rolling in the rolling mill. From this figure, it can be seen that the lower chock 7 is displaced downward during plate rolling.

FIG. 4 is a plot of the displacement of the lower chock 7 shown in FIG. 3 based on the rolling load. From this figure, it was found that the displacement of the lower chock 7 varied for each rolling chance, even with the same rolling load. By the way, even if there is mechanical play in the rolling mill, the displacement amount should not vary if the rolling load is the same. However, in reality, as shown in FIG. 4, there was a variation of about ± 50 μm at the same rolling load. As a result of examining the cause, the shift position of work rolls 2 and 3 differs for each rolling chance, the load acting on the work side and drive side of upper and lower chocks 6, 7 changes, and the backlash becomes apparent and the variation amount It is presumed that it became.

[0018] Such a chock 6,7 for each rolling chance.
Unless the roll gap is set in consideration of the variation in the displacement of the roller, high-precision roll gap control cannot be performed. However, conventionally, even if the play is considered in the roll gap control, the play is considered as a constant value, and the change in the diameter is sometimes actually measured (refer to the above “Steel Handbook”). The control in consideration of the backlash in each case has not been performed.

The present invention has been made on the basis of the above findings, and provides a roll gap control method that takes into account play that changes with each rolling chance. The method includes the following equation (1). As shown in (1), a formula combining a play component and a conventional formula is used. That is, the change in the roll gap generated during rolling is separated into a linear component that changes continuously due to the rolling load and a non-linear component that changes intermittently due to the mechanical play of the rolling mill, and the roll gap is controlled. is there.

ΔS = (G ₀ −G) + F (ΔP, M ₀ , C _T , C _H , C _G ) (1) where ΔS: roll gap control amount G ₀ : play amount standard value G: Backlash detection value ΔP: Deviation from calculated rolling load detection value M ₀ : Mill constant standard value C _T : Constant for rolling material C _H : Constant for rolling mill C _G : Constant for time series change F: Rolling load deviation The following shows a function to obtain the roll gap control amount for.

The play back standard value G ₀ is determined by contacting the work rolls 2 and 3 (with tightening down) in a state where there is no plate and separating the play back into the amount of mechanical play and the amount of elastic deformation. FIG. 5 shows the relationship between the rolling position detected by the rolling position detection sensor 13 and the rolling load. FIG. 6 shows the relationship between the displacement of the lower chock 7 detected by the play detection sensor 17 and the rolling load.

The backlash standard value G ₀ is obtained as a total value from the ratio of the backlash shown in FIGS. 5 and 6 obtained by the rolling position detection sensor 13 and the backlash detection sensor 17. Here, the rolling position detection sensor 13 and the backlash detection sensor 17
Will be described. The rolling-down position detecting sensor 13 is located at the upper part of the rolling mill, and includes a top hat sensor 14 for detecting a roll gap change due to rotation of the rolling-down screw 9 and a sensor 15 for detecting a roll gap change by a hydraulic rolling-down device. In other words, the rolling position detection sensor 13
Is a sensor that detects where the upper roll has been set. The play detection sensor 17 is a sensor that detects the position of the lower roll. The lower roll has no moving mechanism,
It only changes due to load or play.

Now, taking a spring model as an example where the roll is tightened without a plate, when the screw 9 is tightened, it is displaced without generating a load (reaction force) by the amount of play first. When the screw 9 is tightened, the spring between the rolls 2 and 3 and the spring of the liner 11 are compressed to generate a reaction force. The spring of the housing 1 is extended by the reaction force, and the forces are balanced.

The rolling position detecting sensor 13 and the play detecting sensor 17
The relationship between the vertical position play sensor 13 and the displacement of the spring, which is the displacement of the whole upper and lower, and the play detection sensor that the displacement of only the lower part
Although the output level is different as 17 but the function is the same. Next, in the case of plate rolling, the upper roll 2 is set at a predetermined position, and the backlash component becomes zero at the same time as the plate is engaged, and the spring component is deformed by the rolling load. At this time, the displacement of the rolling position cannot be detected by the rolling position detection sensor 13.

That is, the displacement position sensor 13 has a displacement of 0 if the screw 9 is not rotated and if the hydraulic cylinder 8 is not moved. On the other hand, the play detection sensor 17
Since the position of the lower roll 3 is directly measured, the displacement due to the biting of the plate can be detected. Thus, the rolling position obtained from the rolling position detection sensor 13 shown in FIG.
In the relationship of the rolling load (measured by contacting a roll without a plate), the horizontal portion is G ₀ (standard value of play). On the other hand, FIG. 6 shows the relationship between the displacement amount obtained from the play detection sensor 17 and the rolling load field at this time. This horizontal portion is defined as GS0 (detection play standard value). Because of the structure of the rolling mill, there is also a portion where play occurs at the upper part, so that G ₀ > GS ₀ . Also,
In the present invention, the total play is estimated by providing the play detection sensor 17 in the lower part of the rolling mill, assuming that G ₀ / GS ₀ = γ (constant).

Therefore, the ratio of G ₀ / GS ₀ is obtained by the measurement without the plate, and the amount of play during the actual rolling is estimated as follows.
The above-mentioned ratio is multiplied by the above-mentioned ratio to the amount of play at the lower part of the rolling mill obtained from the detection value G of the play detection sensor 17 to obtain a total play. That is, the play value G is a displacement amount of the lower chock 7 obtained by the play detection sensor 17 during rolling.
This is as shown by white circles in FIG.

As described above, during the rolling, the play detection sensor 17 is used.
The play amount GS is obtained from the displacement amount (white circle in FIG. 6) obtained by the above. The displacement amount detected by the play detection sensor 17 includes a play component and a linear component. Therefore,
The amount of subtraction of the linear component from the amount of displacement detected during rolling is the play amount GS. This linear component is obtained using the inclination of the straight line when the detection play standard value GS0 is obtained. Therefore, the play amount G can be represented by G = GS × γ.

The play amount standard value G ₀ can be described by the following equation. That is, since the relationship between the rolling load P and the rolling position S is in the relationship of the above-described formula (1), the following formula (2) is satisfied in a state without a plate. ΔS = G ₀ + F (ΔP, M ₀ , C _H ) (2) ∴G ₀ = ΔS−F (ΔP, M ₀ , C _H ) (3) Now, as a simple example of the function F, Assuming a linear linear equation,
It can be expressed as the following equation (4).

F (ΔP, M ₀ , C _T , C _H , C _G ) = ΔP / M ₀ + C _H (4) ∴G ₀ = ΔS−ΔP / M ₀ −C _H (5) The position of the play detection sensor 17 is set at the lower part of the lower chock 7 due to the configuration of the rolling mill (there are many mechanical components between the lower backup roll and the housing). It is preferable to install it at the place where the cause of occurrence is the largest.

The roll gap control may be performed individually on the work side and the drive side, or may be performed as an average of both.

[0031]

【Example】

(Embodiment 1) Next, as a more specific method, a method of roll gap control in a rolling mill setting calculation before rolling a plate will be described with reference to a flowchart of FIG. Load cell 16, rolling position detection sensor during rolling of Nth plate
13. The play data (rolling load,
The rolling position and the amount of play are sampled at a plurality of points, and the influence coefficient A (corresponding to the slope M of the straight line in FIG. 5) on the rolling load at the rolling position is determined.

Next, a relational expression between the rolling position and the rolling load determined in advance (the formula in the absence of the plate corrected by the plate width and the rolling load level. Details are omitted because they are not related to the present invention).
Is obtained from the influence coefficient B, and the mill constant M in the above equation (1) is learned. As learning, time series learning is performed by the following equation (6).

M = M _a × α + (1−α) × A (6) where M _{a is} the mill constant after the (N−1) -th rolling (M _a = M ₀ at the time of the first rolling) A: Mill constant α at the time of N-th rolling α: Smoothing coefficient At the same time, a difference between a plurality of actual play amounts G and a play standard value G ₀ is obtained as an error C by a statistical method.

That is, assuming that the play amount measured during the N-th rolling is G ₁ , G ₂ ,... G _k , the error C is obtained as follows. _{C = G 0 -G = G 0} - (G 1 + G 2 + ... + G k) / k ...... (7) Then, in the rolling mill setting calculation before rolling the N + 1 th plate mill learned by N th Constant M and backlash error C
Is used to set the roll gap.

That is, the roll gap is set by the following equation (8) using the mill constant M and the play error C, and the control to be performed on the nonlinear component (play error C) and the load This is performed in combination with the control performed on the linear component related to the change. S = S ₀ + ΔS = S ₀ + C + F (ΔP, M, C _T , C _H , C _G ) (8) Here, S ₀ is a roll gap standard value obtained from the (N + 1) th product target dimension.

As the rolling load in the above calculation, a value obtained by correcting the actual values of the work side and the drive side, respectively, by the shift (off-center amount) between the center of the plate and the center of the rolling mill is used. In the above calculations, the work side and the drive side are not separately distinguished, but the roll gap control is basically performed independently for each side, but may be treated as an average. (Embodiment 2) Next, an example of roll gap control of a rolling mill during rolling of a sheet will be described.

This control is basically the same procedure as the method described in Japanese Patent Publication No. 42001/1990. When a plurality of rolling mills are continuous, the next stand is corrected based on the results of the first stand. The improvements of the present invention are described in
The S term in the formula GC = h-MS ₁ -MS ₂ + OF-S of the gauge meter constant described on page 6 of 2001 is to be S = S ₀ + C.

Here, C is the error C of the first embodiment.

[0039]

According to the present invention, it is possible to optimally control the change in the roll gap that occurs during sheet rolling.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a rolling mill for carrying out the method of the present invention.

FIG. 2 is an enlarged view of a lower chock in FIG. 1;

FIG. 3 is a graph showing measured values of a play detection sensor.

FIG. 4 is a graph showing a relationship between a rolling load and a displacement of a lower chock.

FIG. 5 is a graph showing a relationship between a rolling position and a rolling load by a rolling position detection sensor in a state where no plate is provided.

FIG. 6 is a graph showing a relationship between a displacement of a lower chock and a rolling load by a play detection sensor in a state without a plate.

FIG. 7 is a flowchart illustrating the method of the present invention.

Claims (2)

[Claims] The roll gap is controlled by separating a change in a roll gap generated during rolling into a linear component that continuously changes according to a rolling load and a non-linear component that changes intermittently due to mechanical play of a rolling mill. A method of controlling a roll gap in a rolling mill. 2. The method according to claim 1, wherein a rolling load during rolling, a rolling position, and a detection value of a play detection sensor are measured, and a relational expression between the rolling position and the load determined in advance, A roll gap control method in a rolling mill, wherein a play amount during the rolling is determined and controlled using a relational expression between a detected value and a load.