WO1992000817A1

WO1992000817A1 - Interstand tension control

Info

Publication number: WO1992000817A1
Application number: PCT/AU1991/000302
Authority: WO
Inventors: Duncan Campbell Mcfarlane; Peter Mitchell Stone
Original assignee: The Broken Hill Proprietary Company Limited
Priority date: 1990-07-06
Filing date: 1991-07-08
Publication date: 1992-01-23

Abstract

The interstand tension of a multi-stand rolling mill is controlled by: (a) estimating interstand tension at each stand from measured values of roll speed, (b) measuring variables such as motor torque and/or roll separation force at each stand, (c) using the measured values as a basis for comparison with the estimations from roll speed, such comparison being carried out by a Kalman filter, (d) adjusting the estimates of interstand tension on the basis of the comparison, and (e) varying the roll speed at each stand on the basis of the adjusted estimates of interstand tension.

Description

INTERSTAND TENSION CONTROL

The present invention relates to a method and a system for predicting and controlling interstand tension in a multi-stand rolling mill.

The term "interstand tension" is herein understood to mean the longitudinal stress state in a bar being rolled. It is noted that interstand tension can denote that the bar being rolled is in tension or compression.

The term "bar" is herein understood to cover what are conventionally termed "bar", "rod", "strip", and "elongate sections" such as rails and structural members.

Interstand tension in a rolling mill occurs when individual elements of the bar undergo changing rates of volume flow relative to those which existed prior to entry into the rolling mill. Changes in volume flow rate can be caused by a mismatch between rotational speeds of adjacent rolling stands, temperature

differentials along the bar, and/or incorrect roll partings, leading to an accumulation or deficit of material between the stands at a given time.

Fluctuations in interstand tension can result in

variations in bar cross sectional area which lead to dimensional variations in the final product. In this regard, increasing demands for high quality products with high dimensional accuracy and consistency make the control of interstand tension a critical task.

Additionally, interstand tension control reduces the probability of interstand stock buckling and cobbling caused by compression (negative tension) developing between rolling stands.

In order to control interstand tension it is necessary first to have an indication of the interstand tension level at any given time and then to adjust a process variable to maintain an optimum interstand tension level.

Interstand tension can either be measured (direct control) or estimated (indirect control) from measurements of other process variables using a process model. Direct control methods are more accurate but are often relatively costly and impractical and in some cases applicable only to products of small enough dimension .

Known indirect control methods fall into three categories which are listed below in order of increasing sophistication:

(i) Current or Torque Comparison methods (also referred to as "Head End Methods") which deal primarily with establishing correct speed relationships in terms of motor torques while threading the head end of the bar and then maintaining the head end or motor torque values through the bar length.

(ii) Ratio or Quotient Control methods which are widely used and often incorporate features of (i) above and additionally incorporate roll force variables.

(iii) Forward Slip methods which model the bar tension as a direct integral function of bar speed.

An object of the present invention is to provide an improved indirect control method and system for controlling interstand tension in a rolling mill.

According to the present invention there is provided a method of controlling interstand tension in a rolling mill comprising:

(a) dynamically predicting the interstand tension between successive stands in the rolling mill from detected values of controllable input at each stand,

(b) predicting a measurable output at each stand from the predicted values of interstand tension; (c) measuring actual values of the measurable output,

(d) comparing the predicted and measured values of the measurable output,

(e) adjusting the dynamically predicted values of interstand tension in response to any mismatch between predicted and measured values of the measurable output, and

(f) adjusting the controllable input at each stand to control interstand tension on the basis of the adjusted predicted values of the interstand tension.

The term "measurable output" as used herein is understood to mean process variables which can be measured directly or statically derived from directly measured process variables by simple algebraic

relationships.

It is preferred that the controllable input is roll speed and the measurable output is motor torque and/or roll separation force.

It is preferred that the step (a) of predicting the interstand tension between successive stands in the rolling mill comprises Inputting detected values of the roll speed into a dynamic model which comprises a speed control model, a bar speed model and an interstand tension model.

It is also preferred that the dynamic model further comprises a temperature model for estimating the bar temperature at each stand. It is preferred that the step (b) of predicting the motor torque and/or the roll separation force at each stand comprises inputting the predicted values of interstand tension and estimates of bar temperature at each stand into a measurement model which comprises a motor torque model and/or a roll separation force model.

It is preferred that the roll speed adjustment step (f ) is carried out automatically by a controller and comprises:

(a) comparing the adjusted predicted values of interstand tension with predetermined tension set points that are known to produce sections of rolled

product within particular dimensional ranges having prescribed tolerances, and

(b) adjusting the roll speed at each stand to produce interstand tension values within the limits of the tension set points.

In an alternative arrangement it is preferred that the roll speed adjustment step (f) is carried out manually by an operator.

It is preferred that the steps (d) and (e) of comparing the p: dieted and measured values of the motor torque and/or rcil separation force and adjusting the predicted values of interstand tension in response to mismatch between predicted and measured values of motor torque and/or roll separation force are carried out by a Kalman Filter.

According to the present invention there is also provided a method of rolling metal or similar such materials in a rolling mill comprising controlling interstand tension in accordance with the method

described above.

According to the present invention there is also provided a system for minimising interstand tension in a rolling mill comprising:

(a) a means for dynamically predicting the interstand tension between successive stands in the rolling mill from detected values of a controllable input at each stand,

(b) a means for predicting a measurable output at each stand from the predicted values of interstand tension,

(c) a means for measuring actual values of the measurable output, and

(d) a means for comparing the predicted and measured values of the measurable output and for

adjusting the predicted values of interstand tension in response to any mismatch between the predicted and measured values of the measurable output so that the interstand tension between each stand can be controlled by adjusting the controllable input at each stand on the basis of the adjusted predicted values of interstand tension.

It is preferred that the system further comprises a means to automatically adjust the roll speed of each stand to control interstand tension on the basis of the adjusted predicted values of interstand tension.

According to the present invention there is also provided a method for controlling interstand tension in a rolling mill comprising:

(a) dynamically predicting the interstand tension between successive stands in the rolling mill from detected values of a controllable input,

(b) predicting the interstand tension between successive stands in the rolling mill from detected values of a measurable output,

(c) comparing the predicted values of the interstand tension from steps (a) and (b),

(d) adjusting the predicted values of

interstand tension from step (a) in response to any mismatch between predicted values from steps (a) and (b), and

(e) adjusting the controllable Input at each stand to control interstand tension on the basis of the adjusted predicted values of the interstand tension.

It is preferred that the controllable input is roll speed and the measurable output is at least one of bar temperature and motor torque and/or roll separation force.

It is preferred that the step (a) of predicting the interstand tension between successive stands in the rolling mill comprises inputting the detected values of the roll speed into a dynamic model which comprises a speed control model, a bar speed model and an interstand tension model.

It is preferred that the step (b) of predicting the interstand tension comprises inputting the detected values of bar temperature and motor torque and/or roll separation force into an extended head end tension prediction model.

It is preferred that the roll speed adjustment step (e) is carried out automatically by a controller and comprises:

product within particular dimensional ranges having prescribed tolerances, and

In an alternative arrangement it is preferred that the roll speed adjustment step (e) is carried out manually by an operator.

It is preferred the comparison and adjustment steps (d) and (e) are carried out by a Kalman Filter.

described above.

(b) a means for predicting the interstand tension between successive stands in the rolling mill from detected values of a measurable output at each stand, and

(c) a means for comparing the predicted values of interstand tension and for adjusting the predicted values of interstand tension in item (a) in response to any mismatch between the predicted values in items (a) and (b) so that the interstand tension between each stand can be controlled by adjusting the controllable input at each stand on the basis of the adjusted

predicted values of interstand tension.

It is preferred that the controllable input is roll speed and the measurable output is at least one o bar temperature and motor torque and/or roll separation force.

The present invention is described further by way of example with reference to the accompanying drawings in which: Figure 1 is a block diagram of a preferred embodiment of a method and system for controlling

Interstand tension in a merchant bar rolling mill In accordance with the present invention,

Figure 2 is a block diagram of the main components of the computer model illustrated in Figure 1; and

Figure 3 is a block diagram of another preferred embodiment of a method and system for controlling interstand tension in a merchant bar rolling mill in accordance with the present invention.

The preferred embodiments of the present invention are described in relation to a multi-stand merchant bar rolling mill, although it is noted that the present invention is not limited to this application and extends to rolling mills generally and to any process where elastic or semi-elastic material passes between adjacent stages whose relative tractive effort can impart tension in the material.

The preferred embodiments are concerned with controlling interstand tension to an acceptably low level which is constant between adjacent stands in the merchant bar rolling mill.

The method and system of the preferred

embodiment shown in Figures 1 and 2 combines together:

(a) a computer model which comprises:

(i) a dynamic model for predicting

interstand tension from measured values of roll speed (ie a controllable input) at each stand and for predicting temperature at each stand,

(ii) a measurement model for predicting values of roll separation force and motor torque (ie a measurable output) from inter alia the predicted values of interstand tension and bar

temperature, and

(b) a Kalman Filter for comparing measured and predicted values of roll separation force and motor torque and adjusting the predicted values of interstand tension in response to any mismatch between the measured and predicted values of roll separation force and motor torque so that the adjusted predicted values of

interstand tension can then be used as a basis for assessing the extent to which the roll speed at each stand should be varied to effect control of the

interstand tension as required.

Such assessment can be carried out and appropriate adjustments made to the roll speed at each stand by experienced operators.

Alternatively, the assessment and adjustment can be made automatically. In this regard, the method and system of the preferred embodiment shown in Figures 1 and 2 includes a controller operable to compare the values of adjusted predicted interstand tension with a dimensional model which prescribes the maximum and minimum values of interstand tension that can be safely accommodated to produce bar of given dimensions and tolerances while ensuring that product buckling does not occur, and to adjust the roll speed at each stand in a stable manner to produce interstand tension values within the range of these maximum and minimum values.

An important aspect of the method and system of the preferred embodiment shown in Figures 1 and 2 is the selection of the dynamic and measurement models.

With reference to Figure 2, the dynamic model comprises the following models:

(a) temperature;

(b) speed control;

(c) bar speed; and

(d) interstand tension.

The temperature model essentially is peripheral to the main purpose of the dynamic model but is critical to the measurement model. In accordance with the temperature model, the input comprises measured or inferred values of temperature at each roll stand.

The function of the speed control model, the bar speed model and the interstand tension model is to reflect accurately the relationship between roll speed at each stand and the interstand tension. As can be seen from Figure 2, the input to the speed control model comprises the roll speed reference signal at each stand and the output from this model becomes one input to the bar speed model. The other input to the bar speed model is the predicted values of interstand tension. It is noted that this feedback aspect of the bar speed and interstand tension models reflects the fact that there is an inter-dependence between the models and a change in the value of one quantity will necessarily affect the value of the other quantity, and vice versa. This feedback phenomenon is well known and is owing to the slipping of the bars under tension. The output of the bar speed model becomes one input to the interstand tension model and the other input is the predicted values of interstand tension. The output of the

interstand tension model comprises predicted values of interstand tension.

The measurement model comprises the following models

(a) roll separation force; and

(b) motor torque.

The input to the roll separation force model comprises the predicted values of temperature and interstand tension from the dynamic model. The input to the motor torque model comprises the predicted values of temperature and interstand tension from the dynamic model.

The various models referred to above are discussed hereinafter.

Dynamic Model

(a) Bar Temperature Model - the bar temperature in the rolling stands of the intermediate mill is nominally assumed to be constant with time. However, the

inclusion of a pyrometer upstream from the intermediate mill allows for the detection of temperature fluctuations, such as those caused by uneven bar heating in the furnace (e.g. ''skid' effects).

where T - stand bar temperature (K)

δT - stand temperature disturbance (K)

A delay is introduced into the pyrometer signal to align the temperature variations correctly with each stand.

(b) Speed Control Module - roll speed, at any time step k, is determined dynamically from a roll reference speed (set by the mill operator) via a speed control loop. This relationship can be approximated by the first order difference equation:

where t_r - speed control system constant,

ω - roll speed (rad/sec),

ω(ref) ~ roll reference speed (rad/sec),

ΔT - sampling period (sec),

k - time step number,

i - 1=1/ . . . ,4 rolling stand number.

(c) Bar Speed Model - this model determines the speed of the bar at the entry to each stand. This bar speed differs from the tangential speed of the rolls because of roll slip: relative motion between the bar and the roll. Roll slip is directly related to bar tension, and an equation has been developed relating bar velocity to roll speed and tension.

(3) where v - bar speed at stand entry (m/s)

σ - roll speed (rad/s)

It is very difficult to determine an accurate theoretical bar speed model for on-line use, and instead a simplified emperical structure has been selected.

(d) Bar Tension Model - tension between stands i and i+l is dynamically related to bar speed mismatch across the relevant roll gap. The relationship is derived by considering the quantity of bar accumulating between adjacent rolling stands at any given time. The

speed/tension relationship is most accurately defined by a continuous differential equation but is approximatec here by a discrete difference equation with sampling period The tension/bar speed relationship can neatly be written in terms of bar velocity mismatches and tension

differences across a rolling stand:

(4) where L - distance between rolling stands (m)

E - elastic modulus (kN/m )

Ar - stand cross sectional area ratio

(entry/exit)

It is note that this model can be further generalized to include the effects of upstream bar

deformations. Measurement Model

(a) Roll Separation Force Model - the separating force applied by the bar to the rolling stands is affected by two factors: (i) yield strength of the heated bar ('hot strength') and (ii) the tension in the bar. The hot strength equation is complicated and highly non linear, but well known, and we simply note it here as km, - h_liT_i,ω_i)

(5) where k denotes hot strength. The roll separation force is then given by

^Fi ⁼ a₂a₃(km_i - β_1b i - β_{1f i+i})

(6) where a2 - bar/roll contact area (m ),

a3 - geometric term,

F - roll separation force (kN)

β_{1f ,} β_1b - force/tension gradients

It is noted that equations (5) and (6) were developed for flat products and have been modified in order to provide predictions for non-flat products.

(b) Motor Torque Model - motor torque can be

calculated from tension and roll separation force using a lever arm equation which can be derived from inspection of Figure 2. The equation is:

(7) where A_entry - bar cross sectional area at stand entry (m ), l_d - bar/roll contact length (m),

G

- torque (kNm). The dynamic and measurement models can be combined into state space form = xk+1 = f(xk, u_k, d_k, k)

y_k = h(x_k, u_k, k) where x, defines the "process state', u_k, the "process input', d_k, "process disturbances' and y_k 'process

outputs'.

In the tension model, the roll speed reference signals represent process inputs, rollspeed, temperature, and tension are process states, and roll separation force and motor torque are process outputs. In terms of the variables previously defined: u_k = {ω(ref)₁,ω (ref)_2,ω (ref))₃,ω(ref)4}'k x_k = { ω₁,ω₂,ω₃, ω₄T₁,T₂,T₃,T₄,σ₂,σ_3,σ₄}'k

y_k = {F_1'F_2' F_3' F4' G_1' G_{2 '} G_3' G₄}k d_k = {δT₁,δT₂,δT₃, δT₄}'_k where the feedforward temperature measurements δT₁- δT₄, represent disturbance inputs to the model.

The preferred embodiment of the method and system of the present invention shown in Figure 3 combines together:

(a) a dynamic model for predicting interstand tension from measured values of roll speed at each stand;

(b) an extended head end model for predicting interstand tension from measured values of bar temperature and motor torque at each stand; and (c) a Kalman Filter for comparing the predicted values of interstand tension from the dynamic model and the extended head end system and adjusting the predicted values of interstand tension from the dynamic model in response to any mismatch between the predicted values of interstand tension from the dynamic model and the extended head end system so that the adjusted predicted values of interstand tension can then be used as. a basis for

assessing the extent to which the roll speed at each stand should be varied to effect control of the interstand tension required.

As with the preferred embodiment shown in Figures 1 and 2, such assessment can be carried out and appropriate adjustments made to the roll speed at each stand by experienced operators or alternatively such adjustment can be made automatically. In this regard, the automatic adjustment is carried out in the preferred embodiment shown in Figure 3 in the same way as in the preferred embodiment shown in Figures 1 and 2.

An important aspect of the method and system of the preferred embodiment shown in Figure 3 is the selection of the dynamic model and the extended head end model.

The dynamic model is identical to that described in detail in relation to Figure 2. In this regard, it is noted that the outputs of this model are dynamically predicted values of interstand tension.

In the extended Head End Model interstand tension is predicted from the "measured" values of torque and measurements of bar temperature in accordance with the following equation. ^{ύ (} * >

(8) where

G_i - rolling torque (torque at the rolls) (Nm), G_ti, - tensionless rolling torque (Nm),

ld, - rolling contact length projected along the bar (m),

A_i - cross sectional area of bar entering stand

(m),

R_i - effective rolling radius (m),

α_2, - rolling load equation geometric parameter, α₃, - rolling load equation geometric parameter, βl, - gain factor of downstream tension on rolling load,

βb, - gain factor of upstream tension on rolling load,

βi - torque arm length multiplier,

σl, - downstream (forward) tension (N/m ),

σb, - upstream (backward) tension (N/m ),

Many modifications may be made to the preferred embodiments without departing from the spirit and scope of the present invention.

In this regard, whilst the preferred embodiment in Figures 1 and 2 includes the prediction and measurement of both motor torque and roll separation force, this is not an essential requirement of the present invention and it is sufficient that only motor torque is predicted and measured.

In addition, whilst the preferred embodiment in Figure 3 includes the measurements of motor torque, this is not an essential feature of the present invention, and it is within the scope of the present invention to measure other measurable outputs, such as roll separation force, which can be used as the basis to predict interstand tension.

In addition, whilst the dynamic model described in relation to the preferred embodiment shown in Figures 1 to 3 includes a bar speed model, it is noted that the

inclusion of the bar speed model in the dynamic model is not an essential requirement.

In addition, whilst the preferred embodiment described in relation to Figure 3 includes measurement of motor torque, this is not an essential requirement of the present invention and direct motor power measurements could also be used.

In addition, whilst the preferred embodiments include measurement of bar temperature, this is not an essential requirement of the present invention provided roll separation force is measured.

Furthermore, it is noted that one consequence of the method and system of the preferred embodiment in

Figures 1 and 2 is that the accuracy of the relationship between the roll speed and interstand tension, as

reflected by the dynamic model, is checked by comparing the predicted and actual values of the measurable

quantities of roll separation force and motor torque. It is within the scope of the present invention that the dynamic and measurement models are then adjusted in response to the results of the comparison so that the adjusted predicted values of interstand tension from the dynamic model are more accurately related to the roll speed.

Claims

THE CLAIMS DEFINING THE INVENTION ARE AS FOLLOWS:

1. A method of controlling interstand tension in a rolling mill comprising:

(a) dynamically predicting the interstand tension between successive stands in the rolling mill from detected values of a controllable- input at each stand,

(b) predicting a measurable output at each stand from the predicted values of interstand tension;

(c) measuring actual values of the measurable output,

(d) comparing the predicted and measured values of the measurable output,

(e) adjusting the predicted values of

interstand tension in response to any mismatch between predicted and measured values of the measurable output, and

2. The method defined in claim 1, wherein the controllable input is roll speed and the measurable output is motor torque and/or roll separation force.

3. The method defined in claim 2, wherein the step of predicting the interstand tension between successive stands in the rolling mill in step (a) comprises inputting detected values of the roll speed into a

4. The method defined in claim 2, wherein tehe dynamic model comprises a speed control.

5. The method defined in claim 3 or claim 4, wherein the dynamic model further comprises a

temperature model for estimating the bar temperature at each stand.

6. The method defined in claim 5, wherein step (b) of predicting the motor torque and/or the roll

separation force at each stand comprises inputting the predicted values of interstand tension and the estimates of bar temperature at each stand into a measurement model which comprises a motor power model and/or a roll separation force model.

7. The method defined in any one of claims 2 to 6, wherein the roll speed adjustment step (f) is carried out automatically by a controller and comprises:

product within particular dimensional ranges having prescribed tolerances, and

8. The method defined in any one of claims 2 to 6, wherein the roll speed adjustment step (f) is carried out manually by an operator.

9. The method defined in any one of claims 2 to 8, wherein the steps (d) and (e) of comparing the predicted and measured values of the motor torque and/or roll separation force and adjusting the predicted values of interstand tension in response to mismatch between predicted and measured values of motor torque and/or roll separation force are carried out by a Kalman

Filter.

10. A system for minimising interstand tension in a rolling mill comprising:

(c) a means for measuring actual values of the measurable output, and

11. The system defined in claim 10, wherein the controllable input is roll speed and the measurable output is motor torque and/or roll separation force.

12. The system defined in claim 11, further

comprises a means to automatically adjust the roll speed of each stand to control interstand tension on the basis of the adjusted predicted values of interstand tension.

13. A method for controlling interstand tension in a rolling mill comprising:

(a) dynamically predicting the interstand tension between successive stands in the rolling mill from detected values of a controllabel input,

(d) adjusting the predicted values of

14. The method defined in claim 13, wherein the controllable input is roll speed and the measurable output is bar temperature and motor torque and/or roll separation force.

15. The method defined in claim 14, wherein the step (a) of predicting the interstand tension between

successive stands in the rolling mill comprises

inputting the detected values of the roll speed into a dynamic model which comprises a speed control model, a bar speed model and an interstand tension model.

16. The method defined in claim 14 or 15, wherein the step (b) of predicting the interstand tension comprises inputting the detected values of bar

temperature and motor torque and/or roll separation force into an extended head end model.

17. The method defined in any one of claims 14 to 16, wherein the roll speed adjustment step (e) is carried out automatically by a controller and comprises:

product within particular dimensional ranges having prescribed tolerances, and

18. The method defined in any one of claims 14 to 16, wherein the roll speed adjustment step (e) is carried out manually by an operator.

19. The method defined in any one of claims 13 to 18, wherein the comparison and adjustment steps (d) and (e) are carried out by a Kalman Filter.

20. A system for minimising interstand tension in a rolling mill comprising: (a) a means for dynamically predicting the interstand tension between successive stands in the rolling mill from detected values of a controllable input at each stand,

(b) a means for predicting the interstand tension between successive stands in the rolling mill from detected values of a measurable output at each stand,

(c) a means for comparing the predicted values of interstand tension and for adjusting the predicted values of interstand tension from item (a) in response to any mismatch between the predicted values from items (a) and (b) so that the interstand tension between each stand can be controlled by adjusting the controllable input at each stand on the basis of the adjusted

predicted values of interstand tension.

21. The system defined in claim 20, wherein the controllable input is roll speed and the measurable output is bar temperature and motor torque and/or roll separation force.

22. The system defined in claim 21, further

23. A method of rolling metal or similar such materials in a rolling mill comprising the method of controlling interstand tension in accordance with any one of claims 1 to 9 and 13 to 19.

24. A system for rolling metal or similar such materials in a rolling mill comprising the system for controlling interstand tension in accordance with any one of claims 10 to 12 and 20 to 22.