US3650135A

US3650135A - Control for rolling means having successine rolling stands

Info

Publication number: US3650135A
Application number: US832571A
Authority: US
Inventors: Charles Roger Skelton; Dalip Tarachand Malkani
Original assignee: British Iron and Steel Research Association BISRA
Current assignee: British Iron and Steel Research Association BISRA
Priority date: 1968-06-14
Filing date: 1969-06-12
Publication date: 1972-03-21
Anticipated expiration: 1989-03-21
Also published as: GB1270246A; BE734579A; JPS5110193B1; DE1930169A1; SE359463B; FR2010900A1

Abstract

This invention is concerned with the automatic rolling of stock other than strip, plate, sheet and the like. The rolling is carried out in at least two passes and the rolling operation is controlled by adjusting at least one of the following four parameters: roll speed and screwdown of the first of the passes; roll speed and screwdown of the second of the passes, in response to at least one of: direct stock measurement; direct mill measurement; indirect stock measurement; indirect mill measurement.

Description

United States atent Skelton et a1.

[ 1 Mar. 21, 1972 CONTROL FoR ROLLING MEANS HAVING SUCCESSIVE RQLIANG STANDS Inventors: Charles Roger Skelton, Dronfield, Sheffield; Dalip Tarachand Malkani, Chesterfield, both of England Assignee: The British Iron and Steel Research Association Filed: June 12, 1969 Appl. No.: 832,571

Foreign Application Priority Data June 14, 1968 Great Britain ..28,562/68 U.S. Cl ..72/8, 72/9, 72/16, 72/19 Int. Cl. ..B2lb 37/02, B21b 37/08 Field of Search ..72/8-12, 16, 19, 72/21 [56] References Cited UNITED STATES PATENTS 3,222,900 12/1965 Helsing ..72/234 X 3,251,207 5/1966 Wilson ..72/12 3,526,113 9/1970 McNaugher ..72/8 3,049,036 8/1962 Wallace et al. ...72/9 3,212,310 10/1965 Brys ..72/12 3,468,145 9/1969 Yeomansm. 72/l2 3,531,961 10/1970 Dunn ..72/8

Primary ExaminerMilton S. Mehr Att0rney-Bacon & Thomas [57] ABSTRACT This invention is concerned with the automatic rolling of stock other than strip, plate, sheet and the like. The rolling is carried out in at least two passes and the rolling operation is controlled by adjusting at least one of the following four parameters: roll speed and screwdown of the first of the passes; roll speed and screwdown of the second of the passes, in response to at least one of: direct stock measurement; direct mill measurement; indirect stock measurement; indirect mill measure- 12 Claims, 14 Drawing Figures C221 26 cm ROD/14575? fwd/73 PATENTEUMARZI I972 3,650,135

snmaure SMND 7(21) STAND 2 (22) HR [Q WIDTH ERROR 24 STAND I (2!) STAND 2/22) l8 I8 12 F --12 Flo. Z4.

INVENTORS CHARLES Roe-ER SKELTON DAL/P THAHC/IAND Mnum/w d w ATTORNEYS Pm-mnmza I972 3,650,135

SHEET 6 [IF 6 A STAND 1(21) 23 STAND 2(22) RH U 2 2 5Q QQ 64" L F/GBA.

STAND H21) 23 STAND 2(22) FIR Q \L L. F: HSMND M21) 23 TAND 2(22) i I F/a8 INVENTORS gm ATTORNEYS CONTROL FOR ROLLING MEANS HAVING SUCCESSINE ROLLING STANDS This invention relates to the rolling of relatively thick metal stock. Such stock may be material in bloom billet, rod, bar or like form which generally has thickness of the same order as width and the invention is concerned with the problem of rolling of such stock to a desired cross section. Such stock is to be distinguished from stock in the form of strip, plate and sheet the thickness of which is relatively small and of an order of magnitude lower than the width.

A specific solution to this problem has been given in our British Pat. No. 1,150,073 (British Pat. application No. 25,437/65). The present application proposes a general solution and a number of rolling schemes which can be derived from the general solution.

According to the present invention in one aspect, there is provided a method of rolling stock other than stock having, in a cross section transverse the length of the stock, one dimension which is small compared with, and of an order of mag nitude lower than, another dimension normal to said one dimension, said method involving subjecting the stock to at least two passes for producing stock on the discharge side of the last of the two passes having a cross section substantially the same as a desired cross section, the method comprising controlling the rolling operation to produce said desired cross section stock by adjusting at least one of the following parameters: roll speed of the penultimate of the two passes; screwdown of the penultimate of the two passes; roll speed of the last of the two passes; screwdown of the last of the two passes; in response to at least one of: direct stock measurement; direct mill measurement; indirect stock measurement; indirect mill measurement.

According to the present invention in another aspect, there is provided a rolling mill for rolling stock other than stock having, in a cross section transverse the length of the stock, one dimension which is small compared with, and of an order of magnitude lower than, another dimension normal to said one dimension, said mill comprising means providing at least two passes for producing stock on the discharge side of the last of the two passes having a cross section substantially the same as a desired cross section, and means for adjusting at least one of the following parameters: roll speed of the penultimate of the two passes; screwdown of the penultimate of the two passes; roll speed of the last of the two passes; screwdown of the last of the two passes, in response to at least one of: direct stock measurement; direct mill measurement; indirect stock measurement; indirect mill measurement to thereby control the rolling operation to produce said desired cross section stock.

The present invention will be more readily understood from the following derivation of the said general solution and of the following description, given by way of example only, of a number of rolling schemes which can be derived therefrom. In the description, reference will be made to FIGS. 1 to 3 which illustrate various rolling schemes.

Where the term constant roll gap system or constant gap system is used herein this term refers to a system like that described in our British Pat. No. 692,267.

Both the solution in the said British Pat. No. 1,150,073 and the solution of the present application require the stock to pass through means providing two rolling mill passes. The passes may be provided by successive rolling mill stands, alternatively, a reversing mill may provide the passes. The second of the two passes may be the last pass of a sequence or, alternatively, it may be followed by a further pass, or passes.

The solution in the said British Pat. No. 1,150,073 requires that the screw setting of the first pass is controlled in response to measured stock error and that the roll gap of the second pass is fixed, or maintained at a substantially fixed value by automatic gauge control. This solution requires there to be zero, or constant, tension in the stock between the two passes.

1n the general solution of the present application, account can be taken of tension in the stock between the two passes (in the case of the passes being provided by successive stands) and such tension does not have to be zero, or constant. Alternatively, in certain cases, it is possible to properly control the rolling operation ignoring the interstand tension.

To arrive at the general solution of the present application, equations are established relating the input and output variables when rolling simultaneously in two stands of a rolling mill with interstand tension. The coefficients of the equations are established by experiment. The notation is as follows:

H Stock height.

W= Stock width.

6 Stock temperature.

S No load roll gap setting.

N Light load roll speed setting.

(I Roll speed under load.

P= Roll separating force.

T= lnterstand tension.

V= Stock velocity.

R A reference signal used to set the required roll gap in a mill stand incorporating a constant roll gap control system as described herein.

q Selected proportion of a signal from a position transducer fitted on a mill stand incorporating a constant roll gap control system.

r Selected proportion of a signal from a pressure transducer fitted on a mill stand incorporating a constant roll gap control system.

X Closed length of hydraulic jack.

1: Extension of hydraulic jack.

Y= Roll gap with no load on the mill stand and the hydraulic jack in its full closed position (i.e., with x=0).

Z =Roll gap under load.

M Stiffness of mill stand.

a, b, c, d, e, f, g, j, k, l, u, denote coefficients.

Where both upper and lower case symbols are given, the lower case symbol signifies a small change in the corresponding upper case symbol. The suffix convention adopted is such that if a symbol has one suffix then that suffix refers to the stand number. If a symbol has two suffixes then the first suffix refers to the stand, while the second denotes whether the quantity occurs before the stand (0) or after it (1 This suffix convention does not apply to the symbols denoting co-efficients.

When rolling simultaneously in two stands of a rolling mill with interstand tension, the following linear equations with u 8 l0 B 10 8 l3 1 8 2 fl3 1 g8 2 Similar relationships may be drawn up for any of the other output variables such as rolling load and torque.

If it is necessary to evaluate the co-efficients in the above equations they may be determined by experiment for particular steady state standard conditions and for a particular mill and rolling condition.

As an example, the co-efficients can be calculated for rolling conditions in which a round rod is rolled in an oval pass and then in a second, round, pass, the rolling being alternately horizontal-vertical, (i.e., the axes of the rolls on stand 2 at to those on the first stand). Experiments are first conducted to establish the steady state standard conditions. Then small perturbations, each in turn, are made in H W 6,, S S N, and N and the effects of these perturbations on H W 0,, .0 V and T are measured in each case; it will be apparent that the co-efiicients in equations (i) to (viii) can then be evaluated. Similarly, the effects on any other output variables may be measured if required. It must be emphasized that the results so obtained apply to the particular mill and the particular initial steady state rolling conditions for which they are obtained.

vii.

viii.

Using equations (i), (ii) and (vi) above with the calculated values of the co-efficients inserted into them, it is possible for various combinations of the four controls to calculate the necessary changes s 3,, n, and n, which.must be made in order to compensate for changes in the input variables. Equation (vi) is used as a means of indicating which combinations produce acceptable tension changes if two controls only are used, and with three controls supplies the third necessary equation to solve for three unknowns if reasonable values of t are assumed.

A check on the reliability'of the equations can be made by rolling oversize input stock such that the ingoing height and width errors are both, for example +5 percent above steady state values. Equations (i) and (ii), with the known values for the co-efficients a, b, c, d, etc., can be used to predict the necessary screw changes on

stands

1 and 2 in order to eliminate the effects of the input errors and produce out-going stock with the same dimensions as the steady state standard.

During actual experiments, it was found that the predicted changes in the two screw settings were successful in eliminating the effects of the input errors on the output dimensions and did not lead to any excessive changes in the tension level.

For this reason a control system which controls errors by means of gap adjustments in the two mill stands without any control of tension has been found to perform satisfactorily.

Reference is made herein to out-going or output stock height and width; this is to be understood as being a reference to the stock height and width on the output side of the second stand, or pass.

Equations (i) AND (ii) may be written as:

Where X and Y represent input variations:

It will be realized that, in theory, the two output dimensions, 11,, and W can only be controlled simultaneously by making simultaneous changes to at least two of the mill settings e.g., two screws, one screw and speed or two speeds.

By way of example two general modes of control are discussed in detail below.

a. Control by Two Screws If the speed settings are not to be altered and control is effected by varying s and s,, then n, =n =0 at all times.

Equations (ix) and (x) may be written as:

From equations (xiii) and (xiv) the input variables X and l during any particular steady state are given by:

Where s, and s are the existing screw setting variations of the particular steady state from the standard steady state.

Under these input conditions, the desired values of s, and s: which will give zero values of h and w may be predicted by solving equations (xiii) and (xiv) for h,, w 0.

Substituting for X and Yfrom equations (xv) and (xvi):

xiii.

xvii

xviii.

xix.

The right hand sides of equations (xxi) and (xxii) give the desired changes in the screw settings nquired to reduce It and w to zero (assuming the speed settings are maintained constant). b. Control by One Screw and One Speed lf control is to take effect by altering stand 2 screw and roll speed settings, and stand 1 screw and speed settings are not to be altered, .r, =n, 0 at all times.

Equations (ix) and (x) may be written as: hm n, +e Sg Xxill. w =Y+ g, n, +e,s, ulv. Equations (xxiii) and (xxiv) are identical to equations (xiii) and (xiv) excepting that g, n, and g, n; have replaced d s, and d, s,.

Therefore, by exactly the same mathematicalprocess used 5 in (a) it can be shown that The co-efficientsj j,, k,, 1,, n 1, and u, of equations (xxi), (xxii), (xxiii) and (xxiv) are all expressible (as indicated) in terms of the basic co-efficients of the control equation and can therefore be determined by experiment in the manner in-{ dicated. In order to calculate the changes in screw setting on:

the two stands, or the changes in screw and roll speed setting on stand 2, to reduce the errors in output stock height and .'right hand sides of. these equations from a knowledge of the;

determined co-efficients and the measured height and width errors.

' A similar process may be used for determining similar relationships if it is desired to use different combinations of two of the four control parameters (roll speeds and screw settings for the two passes). Similar equations to (xxi), (x xii), (xxv) AND (xxvi) may be derived in terms of the stock input errors using equations (i) and (ii).

It may be desired to additionally measure variations in they interstand tension and if these variations are found to be large under same steady states to use them to alter a third mill setting e.g., n for the first example given over leaf and s for the second. In this case, each of equations (xxi), (xxii), (xxv) AND (xxvi) would include an additional term which can be determined by use of equation (vi), in addition to equations (i) and (ii). I

Reference will now be made to FIGS. 1 to 8 in which like reference numerals indicate like parts.

FIG. 1 shows calculating apparatus for evaluating the right hand sides of equations (xxi) and xxii) so as to give the changes in screwdown on

stands

1 and 2 necessary to reduce the output height and width errors to zero.

FIGS. 2 and 3 each show such calculating apparatus controlling two stands of a rolling mill according to the invention.

FIGS. 4 to 8 illustrate schematically rolling schemes which are at present believed to be particularly important embodiments of the present invention.

Turning to FIG. 1, output height and width errors are shown 2592s?! I9. Em ts The ile k s a s in screwdown on

stands

1 and 2 necessary to reduce the output height and width errors to zero are shown emanating from terminals 11. The values of the

resistances

12, 13, 14 and 15 are chosen relative to the values of the resistances 16 and 17 to give the values as calculated by experiment of the coefiicients jh h h, and h The summation of the products h h and k w and the summation of the products j h and k w are achieved by means of the

operational amplifiers

18 and 20 respectively. FIG. 1 accordingly shows how signals representing the desired screw setting changes to eliminate output stock errors h and W can be obtained in a simple manner.

The methods whereby these signals can be used to actuate the screwdown will now be discussed with reference to FIGS. 2 and 3, where the two screw method of control has been applied.

These Figures show the rolling mill comprising a first stand 21 and a second stand 22. The stock, for example rod, is indicated at 23 and travels in the direction of the arrow A. The roll pass of stand 21 is oval in cross section and the roll pass of stand 22 is round in cross-section. As shown in FIGS. 2 and 3, the axes of the rolls of

stands

21 and 22 are all horizontal. However, to achieve an alternately horizontal-vertical rolling sequence, twist guides of known form and located between the

stands

21 and 22 cause the rod 23 to be twisted through 90 between

stands

21 and 22. As an alternative arrangement, to produce an alternately horizontal-vertical rolling sequence, the roll axes of the stand 21 may be horizontal and the roll axes of the stand 22 vertical.

In each of FIGS. 2 and 3, the control system which is shown controlling stand 21 will be duplicated for stand 22 but with the use of coefficientsj and k so that the screwdown of both these stands will be adjusted in accordance with the measured output height and width errors.

FIG. 2 shows one method of using the signals as computed in FIG. 1 to actuate the desired screw changes on

stands

21 and 22. Errors in output stock height and width are measured by a rod, or bar, meter shown generally at 24 and having

channels

25 and 26. The signal carried by channel 25 will represent the output height error and the signal carried by channel 26 will represent the output width error. The signals carried by

channels

25 and 26 will be processed by a circuit which has already been described in relation to FIG. 1 which circuit will emit an error signal representing the right hand side of equation (xxi). This signal represents the desired change in screw setting for stand 1 and may be termed the screw setting error signal. This signal is fed to an amplifier 27 and from thence to servo-valve 28. The servo-valve 28 is fed by a pump 36 and is connected to a drain 31. The servo-valve 28 operates a jack 32 which is used to adjust the screwdown of the rolls of the stand 21. The jack position is representative of the screw setting and with the arrangement shown in FIG. 2 the jack velocity would approximately be proportional to the screw setting error signal. The jack would stop in a new position only when the error signal was zero. The disadvantage with this method is that the jack position is controlled solely by computations made from the rod meter readings and rapid changes in load would cause errors in jack position, which would not be detected until the stock reached the rod meter. In the case of stand 21, this could represent a fairly large time interval and also because of this transport lag, the rate of correction cannot be very fast without causing instability. With this method there is also the problem of maintaining the correct gap when no rod is being rolled. The above two disadvantages may be eliminated by the arrangement shown in FIG. 3. As between FIGS. 2 and 3, like parts have been given the same reference numerals.

The arrangement shown in FIG. 3 may be best understood if the operation of the constant gap control system shown within the dotted lines 37, is described first. This system is shown applied to stand 1 and with no external inputs into the system at terminal 49, the system will operate to maintain a constant roll gap under load, i.e., the system makes adjustments in the mill setting to compensate for the elastic deflection of the various components of the mill stand 21. Although the embodiment shown in this diagram uses a hydraulic jack 32 and a servovalve 23 to actuate movement of the rolling mill chucks, screws and wedges actuated either by electric motors or hydraulically may also be used for this purpose. A positive reference signal R at 38 is used to select the required constant roll gap.

The signal from the position transducer 51 indicates the extension x or the change in X which is the closed length of the jack. The sense of this signal is negative, i.e., as the jack extends the position transducer 51 gives a negative signal whose magnitude is proportional to the extension of the jack. Any proportion q of this signal can be fed into the resistor 42 by suitable adjustment of the potentiometer 40. The pressure transducer 50 indicates the jack pressure (which is proportional to rolling load) and provides a positive signal proportional to the magnitude of the rolling load P. Any proportion r of this signal can be fed into the resistor 43 by adjustment of the potentiometer 45. The operational amplifier 46 sums the signals fed into 38, 43 and 42, i.e., its output is given by This error signal is fed into the servo-valve and if the signal has a positive sense the servo-valve allows flow into the jack so that it extends it and vice-versa. The jack attains a steady position only when the above error is zero, i.e.,

If Y is the no load roll gap of the mill stand when the jack is in its fully closed position (i.e., with PC) at any time under rolling load I, the actual roll gap Z will be given by xxviii. where M is the stiffness of the mill stand as a whole and the term P/M gives the elastic deflection of the stand.

Substituting for x from xxvii) Z=Y-R/qrP/q-+-P/M xxix. if the values of r and q are selected such that qlr=M equation (xxix) simplified is Z=YR/q xxx. It can be seen that Z, the roll gap under load is independent of the load and is constant when reference R is constant. The reference R may be used to initially set the gap to the desired value.

When the feed back signal from the pressure transducer 50 is cut out from the system, the error fed to the valve will be given by Rqx, and the control system endeavors to maintain the jack extension at at a constant value given by FIR/ll xxxi. This is independent of load and from equation (xxviii) the actual roll gap under load is given by xxxii. The term R/q is equivalent to the screw position in a conventional conventional mill and can be used to set the no load roll gap (Y-R/q). This gap however, increases by an amount P/M when the rolling load P is applied.

In FIG. 3 the error signal representing the desired change in screw setting for the stand 21, is shown as being fed to a resistance 33. The screw setting error signal is then fed through a three term controller 34, (a three term controller being a well known piece of equipment), whose output signal may contain three terms, integral, proportional and derivative with respect to the input error. This signal is then presented to the control system shown generally at 37 for the stand 21. In particular, this signal is added to the reference signal of the gap control system 37 which operates with both its .position and pressure feed back loops closed. The signal from the three term controller and the signals from the position and pressure feed back loops are combined by the operational amplifier 46 having resistance 47 in parallel therewith and the resultant signal is passed to amplifier 48; the resultant amplified signal is fed to servo-valve 28. The servo-valve 28 operates the jack 32.

In any steady state, the achieved roll gap will be equal to the set roll gap (proportional to the reference signal) plus an amount proportional to the value of the signal from the three term controller 34. The roll gap will attain a new constant value only when the screw setting error is zero (as in the system of FIG. 2). However, if any rapid change in load occurs causing errors in jack position, the pressure and position transducers will detect these changes and rapid corrective action will be taken by the inner control loops of the gap control system. Because of the transport lags, the rate of correction initiated by the error signal which is computed from rod meter readings (outer control loop) will have to be relatively slow.

Another alternative is the use of the system as described in relation to FIG. 3 but with the pressure loop of the gap control system 37 open. Here, the change in jack position will be proportional to the signal from the three term controller and this will reach a new steady state position only when the screw setting error is zero. However, the jack position will still be maintained when sudden changes in load occur because of the action of the inner position loop. The existing values of the coefficients should be used in the equations in this alternative system. However, if the complete gap control system is used (with the pressure loop closed) new co-efficients will have to be determined. It would be possible to do this experimentally with the gap control system in operation or by deriving the relationship between the two analytically.

Other methods of control are also possible. These include discontinuous systems, self-adaptive systems and semi-predictive systems, etc.

In the discussion above, the screw corrections for the two stands have been predicted with the use of equations (xxi) and (xxii). In relation to the screw correction for the second stand (i.e., stand 22), however, it can be shown that only a constant roll gap control system on its own on this stand can satisfy the system requirements. This assumes that roll wear and expansion are neglected; this should provide the same screw setting change as predicted by equation (xxii). It will be appreciated that the use of a constant roll gap control system on the second stand is the solution put forward in the said British Pat. No. 1,150,073. The reasoning behind putting a constant roll gap system on the second stand is as follows:

The changes in screw setting predicted by equations (xxi) and (xxii) must reduce both h and w,, to zero, but 11,, can only be reduced to zero by reducing the roll gap variation from the steady state to zero, i.e., by keeping the gap constant. In practice, however, roll wear and/or expansion will not be negligible and a rod meter will in all probability be required to measure I1 purely for resetting the reference signal to offset errors arising out of roll wear and/or expansion.

It can be seen from the control equations that stock can be rolled to the desired cross section by taking two stock mea' surements (H and W,,) and adjusting, in response to these measurements, 1 mill parameter (screw setting on stand 1) the roll gap on the second stand being kept constant. In this case the only control equation in issue is equation (xxi) because the stand 2 screw setting is solely controlled by the constant gap system. If the constant gap system on stand 2 does not over come all errors in H then both H and W must be measured.

Instead of obtaining control by the use of equations (xxi) and (xxii) control may be obtained by the use of equations (xxv) and (xxvi). In this latter case, the speed and screw settings of stand 22 would be varied to attain the control over the output height and width of the rod. The screw setting change of stand 22 may be made in the same way as for the method using two screws, namely by using a constant roll gap control system on stand 22. A suitable circuit will adjust the speed setting of stand 22 in response to measured errors in output height and width of the rod.

It will be seen that in the last-mentioned rolling schemes, where the second stand (stand 22) is controlled solely by a constant gap system, then the only control equation in issue is equation (xxi) or equation (xxv) which reduce respectively to:

because due to the constant gap control system on the second stand (or a second stand of high stiffness) 12,, is substantially zero.

It can be seen from the foregoing control equations that a system is possible in which the only stock parameter measured is the output width. In such a system, correction of the output width may be obtained by using errors in output width to adjust screwdown on the first stand only while stand 2 operates with a constant gap control system. Alternatively, the second stand roll gap could be kept at a substantially constant value by using a second stand of high stiffness without any form of gap control.

The errors caused by long term variations (such as roll wear) in arrangements using a constant gap control system or a stiff mill design on the second stand may be overcome by feedback of the output height error to stand 2 screwdown and the output width error to the screwdown or roll speed of stand 1 or the roll speed of stand 2.

When one stock dimension only is measured, that dimension must be the output width namely the width of the stock after leaving the second stand, i.e., the second stand in the direction of travel of the stock through the two stands. This is because h is assumed to be zero. If any other stock dimension were measured, it can be shown from the control equations that an additional measurement (which would not be zero) would be required to compute the necessary mill correction. When one dimension only (namely output stock width) is measured, it is preferable that the rolling planes of the two stands relative to stock be mutually perpendicular. This can be achieved by arranging for the axes of the rolls of one stand to be perpendicular to the axes of the rolls of the other stand; alternatively, the axes of the rolls of one stand may be parallel to the axes of the rolls of the other stand and the stock twisted through between the two stands. Relative to the stock, the rolling sequence may be horizontal-vertical, or vertical-horizontal, relative to the direction of travel of the stock.

It is possible to use the control equations on the basis of stock measurement elsewhere than after the second stand. Thus control may be effected on the basis of stock measurements taken before the first stand, or between the two stands, or both before the first stand and between the two stands. However, in these cases, both stock height and width must be measured; also in these cases it is expected that the stock measurements will have to be used to effect adjustment in at least one of: first stand screwdown, first stand roll speed, second stand screwdown, second stand roll speed.

In application of the invention to a reversing mill, a predictive system must be used. Thus height and width of the stock prior to entry into the first pass must be measured. These mea sured ingoing height and width errors may be used to predict the changes in the screwdown setting of the first and second passes to eliminate output height and width errors. Adjustment of the roll speed of both the passes would have little or no effect. The prediction for the second pass may be stored. Alternatively, the roll gap of the second pass may be kept constant. After this control action, any height and width errors measured on the output side of the second pass may be used to update value of the constants used in determining the screwdown adjustment of the first pass.

Instead of referring to stock height and width, it may be considered more appropriate, particularly in the case of round stock, to refer to the rollway dimension and guideway dimension respectively. The rollway dimension is the stock dimension measured in the direction of the roll gap and guideway dimension is the stock dimension measured perpendicular to the rollway dimension. Prior to the first stand, or pass, the rollway dimension will be measured in the direction of the roll gap of that stand or pass and after the second stand, or pass, the

rollway dimension will be measured in the direction of the roll gap of the second stand, or pass. In between the two stands, or passes, the rollway dimension may be measured relative to either the first stand, or pass, or the second stand, or pass.

Measurements of stock height and width may be made either directly, for example by means of a rod, or bar meter, or indirectly. As an example of indirect stock measurement, the stock height measurement which instead of being measured directly by a meter can be taken to be equal to the roll gap under load which in turn may be determined by adding the known no load roll gap to the mill spring which in turn is a known proportion of the measured rolling load. The stock width can be obtained indirectly by dividing the height into a measured value of mass flow of the stock from the second stand, or pass, or roll gap.

Also, there are means of indirect dimension error measurement that can be drawn from equations (iii) and (iv). At any time variable changes s,, s n and n are known and it should be possible to measure stand speed changes under load, to, and with great accuracy using good quality tachogenerators or other suitable means.

Therefore we can write:

i io a io "2 q l 4 2 f4"1 84"2 assuming that temperature variations are negligible. Consequently,h and w can be determined provided 11 b, 0 17., 0 which generally will be the case. Small temperature variations about the chosen steady state may be assumed negligible. In a similar way, it would also be possible to utilize equations relating load changes on each stand with the input variables or even those that could be drawn up for torque changes.

Thus it is possible, by use of the control equations, to measure roll speed changes under load, or roll load changes, at one or both passes and use variations in these measurements to adjust at least one of: first pass screwdown, first pass roll speed, second pass screwdown, second pass roll speed to obtain stock on the output side of the second pass of substantially the desired cross-section. The roll speed changes may be measured directly, for example by means of a tachogenerator or indirectly for example by means of sound measuring apparatus which would detect pitch changes in the noise of the roll drive on changes of roll speed. Roll load changes may be measured by the use of load cells.

The invention of the present application requires some method for measuring stock height and/or width, or stock rollway and/or guideway dimensions. For this purpose, a TV. bar meter has been developed and is the subject of our copending application No. 700/68. It is such a meter which may comprise the rod, or bar, meter referred to with reference to FIGS. 2 and 3.

The following rolling schemes (see FIGS. 4 to 8) are at present believed to be particularly important embodiments of the present invention:

1 FIG. 4

Stock output width error feedback for the adjustment of stand 1 screwdown with a constant gap control system on stand 2 or a stand 2 of stiff design. No control action would be taken to alter the steady state speed settings of the stands. This scheme can be performed by the rolling mills ofFIGS. 2 and 3, it being understood that the following modifications would be made: (a) channel 25 carrying the output height error and associated circuitry would be removed so that output width errors only from channel 26 would be used to control stand 21 screwdown by the means disclosed; (b) stand 22 would be of stiff design or have a constant roll gap control system applied to it in place of the stand 22 control of FIGS. 2 and 3.

In FIG. 4, reference 24 is a width meter only feeding width error to three term controller 34. Box 60 is a screwdown controller for adjusting the screwdown on stand 1 (21) in response to the output of the controller 34. In the scheme of FIG. 4 and also in other schemes where it is used, the screwdown controller 60 may comprise the screwdown adjustment mechanism of FIG. 2 or may incorporate a constant gap control system and therefore comprise the screwdown adjustment mechanism of FIG. 3. The jack 32 adjusts the roll screwdown of stand 1 (21) in response to the output from controller 60. The reference 61 indicates a constant gap control system for stand 2 (22) feeding jack 32; alternatively, this stand may be of stiff design.

2 FIG. 5

Stock output width error feedback for the adjustment of stand 2 roll speed with constant gap control on stand 2 or a stand 2 of stiff design. N0 control action would be taken to alter the steady state screwdown and roll speed settings of stand 1. This scheme can be performed by the rolling mills of FIGS. 2 and 3 it being understood that the following modifications would be made: (a) channel 25 carrying the output height error and associated circuitry would be removed; (b) output width errors from channel 26-would be used to control stand 22 roll speed; (c) stand 22 would be of stiff design or have a constant roll gap control system applied to it in place of the stand 22 control of FIGS. 2 and 3; (d) the control of FIGS. 2 and 3 would be removed from stand 21.

In FIG. 5, the width meter 24 feeds its signal to three term controller 34, the output of which feeds roll speed controller 62. In the scheme of FIG. 4 and also in other schemes where it is used, the controller 62 may be comprised by any one of a number of well known means for controlling motor speed, e.g., a Ward-Lennar control system. The output of controller 62 is fed to the rolls of stand 2 (22). Reference 61 indicates a constant gap control system for stand 2 (22); alternatively, this stand may be of stiff design.

3 FIG. 6

Stock output wi th error feedback for the adjustment of stand 1 roll speed with a constant gap control system on stand 2 or a stand 2 of stiff design. No control action would be taken to alter the steady state screwdown setting of stand 1 and roll speed setting of stand 2.

In FIG. 6, the width meter 24 feeds its signal to three term controller 34, the output of which feeds roll speed controller 62 for stand 1 (21). Reference 61 indicates a constant gap control system for stand 2 (22) feeding jack 32; alternatively, this stand may be of stiff design.

4 FIGS. 7A-7E A predictive system with input stock width and height error (i.e., width and height error measured immediately upstream of stand 1) fed forward (FIG. 7A) to stand I and stand 2 screwdowns (no control action to alter the steady state stand 1 and stand 2 roll speed settings), or (FIG. 7B) to stand 1 roll speed and stand 2 screwdown (no control action to alter the steady state stand 1 screwdown and stand 2 roll speed settings), or (FIG. 7C) to stand 2 roll speed and screwdown (no control action to alter the steady state stand ll speed and screwdown settings), or (FIG. 7D) to stand 1 screwdown and stand 1 roll speed (no control action to alter the steady state stand 1 roll speed and stand 1 screwdown settings), or (FIG. 7E) to stand 1 screwdown and stand 2 roll speed (no control action to alter the steady state stand 1 roll speed and stand 2 screwdown). The co-efficients used may be updated by measurement of output stock errors.

These schemes are illustrated in FIGS. 7A to 7E and it is believed that no detailed description of these Figures is necessary bearing in mind that the items referenced 23, 32, 34, 60, 62 have already been described. In FIGS. 7A to 7E, reference 24 denotes a rod, or bar, meter measuring both stock height and width. Line 63 carries the rollway dimension (height) error and line 64 the guideway dimension (width) error. The

boxes

65 and 66 of FIGS. 7A to 7E enclose circuits having the same electrical components l2, l3, l6 and 18, and arranged in the same manner, as appear in FIGS. 2 and 3. The values chosen for these components in

boxes

65 and 66 are not necessarily the same as the values of these components in FIGS. 2 and 3 and are determined by the appropriate control equations.

5 FIGS. 8A-8C Stock output width error feedback for adjustment of stand 1 screwdown or stand 1 roll speed or stand 2 roll speed together with stock output height error feedback for adjustment of stand 2 screwdown for correction of the long term variations in the roll gap at stand 2, which embodies a constant roll gap system or is of stiff design. No control action would be taken to alter that mill parameter (steady state stand 1 screwdown and roll speed settings and steady state stand 2 roll speed settings) not adjusted in response to the said stock measurement.

These schemes are illustrated in F IGS. 8A to SC and it is believed that no detailed description of these Figures is necessa ry bearing in mind that the items referenced 23, 24, 32, 34, 60, 62 have already been described. Lines 63 carry the rollway dimension (height) error and lines 64 the guideway dimension (width) error.

It will be realized that in the rolling schemes comprising embodiments of the invention, the mill parameters (screwdown of each pass in the case of rolling with a reversing mill, and screwdown and roll speed of each stand in the case of rolling with a multi-stand mill) will be initially set to steady state values. Then to control the rolling operation to eliminate output height and width errors which would arise as a result of variations in input stock height and width, at least one of the mill parameters is adjusted in response to stock or mill measurement. The term stock measurement is to be construed .as covering measurement of one or more stock dimensions either directly or indirectly as previously indicated. The term mill measurement is to be construed as covering measurement of one or more mill parameters (roll speed or roll load) either directly or indirectly as previously indicated.

The following paragraphs (1) to (9) comprise a summary of the control systems which are intended to fall within the scope of this invention.

1. A control system for ensuring that both the output width and height of the stock emerging from a rolling mill are kept constant. This control is applied while rolling in at least two consecutive passes either on a single stand if the mill is of reversing type or on two stands if the mill is a continuous type. 2. The control system of 1) may operate by direct measurement of outgoing width errors only in order to actuate either a screw change on the first stand or a speed change on either of the two stands. The second stand should be of a stiff design and/or could incorporate a constant roll gap control system to provide the stand with infinite stiffness. The first stand may also be operated to advantage by incorporating a constant roll gap control system.

3. The control system of 1) may operate by direct measurement of both output height and width errors, such that height errors actuate the screw setting of the second stand (to achieve constant roll gap at that stand) and width errors the screw setting on the first stand or the speed setting of either stand. A stiff second stand design would be advantageous or alternatively the second stand could incorporate a constant roll gap control system. A constant roll gap control system may also be usefully incorporated in stand 1.

4. The control system of 1) may operate by measurement of both output height and width errors, which when utilized in accordance with equations as devised herein, may be used to actuate stand one or stand two speed or stand one gap setting. Stand two should be of a stiff design or have a constant roll gap control system.

5. The control system of (1 )may operate by measurement of both output height and output width errors which when utilized in accordance with the equations as derived herein may be used to actuate stand 1 or stand 2 speed or stand 1 gap setting. Stand 2 is controlled by height error feedback and is of stiff design or has a constant gap control system.

6. The control system of I) may operate by direct measurement both of output height and width errors which when utilized in accordance with equations as derived herein may be used to actuate two or more of the following four control variables, namely: the two screw and speed settings of the stands. It may also beadvantageous to have a constant gap control system incorporated in each stand.

7. The control system of 1) may operate by direct measurement of both the input height and width errors which when utilized in accordance with equations derived in the same manner as the equations herein may be used to actuate two or more of the following four control variables, namely the two screw and speed settings of the stands. It may also be advantageous to have a constant gap control system incorporated in each stand.

8. For the predictive case (7) above, this control system may use the measurements of output height and width errors for the specific purpose of updating the co-efficients in the equations that enable screw or speed setting changes to be predicted.

9. The control system of 1) may use indirect measurement of dimension errors by monitoring changes in mill parameters which can be related by equations derived in the same manner as the equations herein.

We claim:

1. A method of rolling stock other than stock having, in a cross section transverse the length of the stock, one dimension which is small compared with, and of an order of magnitude lower than, another dimension normal to said one dimension, said method including rolling the stock in at least two successive rolling stands with variable interstand tension, and controlling the rolling to produce on the discharge side of the last of the two rolling stands desired stock having both its height and width controlled, the control of the rolling being provided by simultaneously adjusting during rolling the roll gap of one of the stands and at least one other of the following independently adjustable parameters (i) roll speed of the penultimate rolling stand, (ii) roll gap setting of the penultimate rolling stand, (iii) roll speed of the last rolling stand, and (iv) roll gap setting of the last rolling stand, provided that when the last rolling stand has a constant roll gap control system at least one of parameters (i) and (iii) is adjusted, both adjustments being made in response to errors in two mutually perpendicular dimensions of the cross section of the stock measured at one location.

2. A method according to claim 1, wherein both adjustments to said parameters are made in response to errors in cross section of the desired stock as rolled by the last rolling stand.

3. A method according to claim 1, wherein the rolling operation is controlled by adjusting the roll speed of the last rolling stand in response to error in the width of the desired stock as rolled by the last rolling stand, the roll gap of the last rolling stand being kept substantially constant.

4. A method according to claim 1, wherein the rolling operation is controlled by adjusting the roll speed of the penultimate rolling stand in response to error in the width of the desired stock as rolled by the last rolling stand, the roll gap of the last rolling stand being kept substantially constant.

5. A method according to claim 1, wherein the rolling operation is controlled by adjusting the roll gap setting of the penultimate and last rolling stands.

6. A method according to claim 1, wherein the rolling operation is controlled by adjusting the roll gap setting of the last rolling stand in response to error in height of the desired stock and another of said parameters in response to error in width of the desired stock.

7. A rolling mill for rolling stock other than stock having, in a cross section transverse the length of the stock, one dimension which is small compared with, and of an order of magnitude lower than, another dimension normal to said one dimension, said mill including at least two successive rolling stands through which the stock is rolled with variable interstand tension; and control means for controlling the rolling to produce on the discharge side of the last of the two rolling stands desired stock having both its height and width controlled, the control means being adapted to provide the control by simultaneously adjusting during rolling the roll gap setting of one of said stands and at least one other of the following independently adjustable parameters (i) roll speed of the penultimate rolling stand (ii) roll gap setting of the penultimate rolling stand, (iii) roll speed of the last rolling stand, and (iv) roll gap setting of the last rolling stand, provided that when the last rolling stand has a constant roll gap control system at least one of parameters (i) and (iii) is adjusted, both adjustments being made in response to errors in two mutually perpendicular dimensions of the cross section of the stock measured at one location.

8. A rolling mill according to claim 7, wherein the control means are adapted such that both adjustments to the said parameters will be made in response to errors in cross section of the desired stock as rolled by the last stand.

9. A rolling mill according to claim 7, including a speed control for adjusting the roll speed of the last rolling stand in response to error in the width of the desired stock as rolled by the last rolling stand, and a constant gap controller for keeping the roll gap of the last rolling stand substantially constant.

10. A rolling mill according to claim 7 including a speed controller for adjusting the roll speed of the penultimate rolling stand in response to error in the width of the desired stock as rolled by the last stand, controller for keeping the roll gap of the last rolling stand substantially constant.

1 l. A rolling mill according to claim 7 including a controller for adjusting the roll gap setting of the penultimate and last rolling stands.

12. A rolling mill according to claim 7 including a controller for adjusting the roll gap setting of the last rolling stand in response to error in height of the desired stock and a controller for adjusting another of said parameters in response to error in the width of the desired stock.

t l I t

Claims

11. A rolling mill according to claim 7 including a controller for adjusting the roll gap setting of the penultimate and last rolling stands.