US 3863478 A
There is described a system for controlling the tensile force acting on a rolling material between individual sets of rollers of a continuous rolling mill. In order to carry out the continuous rolling operation with accuracy, the constants of transfer function of the tensionfree control loop is automatically corrected in accordance with variations in the rolling conditions such as dimension and shape of the steel material being worked, the rolling speed, the distance between individual sets of rollers and the like. Furthermore, the control includes a learning function.
Claims available in
Description (OCR text may contain errors)
United States Patent [1 3,863,478
Harada et al. Feb. 4, 1975 SYSTEM FOR CONTROLLING ROLLING MILLS  References Cited  Inventors: Toshio Harada; Shinichi Nakarnata, UNITED STATES PATENTS both of Kisarazu; Shiro Araki; Koei 3,188,841 6/1965 Wallace 72/9 Nflkashima, bgth of Kitakyushu; 3,677,045 7/1972 Arimura 61 a]. u 72/8 Hironori Kawasaki, Kimitsu; Kazuo g C erson 23:12:: gfii g fi i t ggfg 3,7s2,|51 1/1974 Peterson 72/9 Kobe of Japan Primary ExaminerMilton S. Mehr [731 Assignees: Nippon Steel Corporation;
Mitsubishi Electric Corporation,  ABSTRACT both of Tokyo, Japan There is described a system for controlling the tensile 22 Ffled; Sept 6, 1973 force acting on a rolling material between individual sets of rollers of a continuous rolling mill. In order to  PP N05 394,844 carry out the continuous rolling operation with accuracy. the constants of transfer function of the tension- 0 Foreign Application p Data free control loop is automatically corrected in accor- Se 1 6 I972 J3 an 4188686 dance with variations in the rolling conditions such as 1972 Japan 4.' 88687 dimension and shape of the steel material being p p worked, the rolling speed, the distance between indi-  U S Cl 72/6 72/9 vidual sets of rollers and the like. Furthermore, the [511 it C. 'II"'I""f"'"'"""""'IIIIIIIIIIIIII m; m nnnnn' nnnnnnn n nnnnng fnnnnnn-  Field of Search 72/6-12, i9 15 Claims, 4 Drawing Figures lA lB lC /ROLLERS IROLLERS )ROLLERS LOAD LOAD 7 DETECTOR 7B DETECTORS 70 l 1 l DRIVE n DRIVE l 3A MOTOR 3B MOTOR 3c, ivicii s REVOLUTION 2A H 2B REV3LUTl0N 2c DETECTOR A "gggggtgg' 4 B ozrzcron I If I L SPEED SPEED TENSION rgnsion CONTROL CONTROL I CONTROL I j l i r COMPUT ER PAIEHTEU sum 2 0F 5 ROLLERs ROLLERS ROLLERS IOA IOB IOC G 5 (5 DRIvE DRIvE 0 MM I MOTOR I MQTOR I HA B l IC i 2M 'rRAusPonIIIER 5 mansromzn 12": CURRENT CURRENT OIRRENT sOuRcE sOuRcE sOuRcE' A POWER POwER RHEOST T CONTROL CONTRQL FILTER 9/ FILTER SPEED E RI-IEOsTAT 2 MA SPEED CONTROL 'T'LKSWITCH IGC MEMORY '24C 23B 23C "I SWITCH LS SPEED '78 385$8OL CONTROL SYSTEM FOR CONTROLLING ROLLING MILLS BACKGROUND OF THE INVENTION Field of the Invention The instant invention relates generally to a control system for a rolling mill, and more particularly to a system for controlling the tensile force acting on a rolling material between individual sets of rollers of continuous rolling equipment.
In continuous rolling equipment where a rolling material extends over and is squeezed simultaneously by a number of rollers, it is necessary to control the tensile force acting on the rolling material between the individual rollers. In some cases, for example in continuous strip rolling, a tensile force of a predetermined value is maintained in the rolling direction in order to facilitate the compressive working of the metal. However, it is the usual practice in continuous rolling of steel sections to maintain the rolling material in tension free conditions to avoid defects or irregularities in the shape of the rolled material. For this purpose, conventional rolling mills are provided with a tension preventing control means of one form or another. With the conventional control devices of such nature, however, the same con stant of transfer function is used for all the rolling conditions such as the dimension of the steel material to be rolled, the rolling speed, and the distance between the rollers, regardless of the particular shapes of the steel sections. In actuality, however, the rolling conditions do vary depending upon the particular schedule employed in the rolling operation. In order to ensure an accurate rolling operation, it is therefore desirable to automatically correct the constant of transfer function in response to variations in the actual rolling conditions.
With the foregoing in view, since the tension occurring between the rollers during the rolling operation depends on the roller speeds which in turn are influenced by the degree of complexity in the shape of the steel material, rolling conditions of the steel material and variations in the rolling loads, it has been found that, in order to control or eliminate the tension which would be imposed on the rolling material between the individual rollers, the constant for a rolling mill driving device, and more particularly, the constant for a tension control circuit should be determined as a function of the various rolling conditions including the dimension, shape, sectional area, rolling speed, reduction ratio, forward slip ratio and backward slip ratio of the rolling material, thus eliminating the tension in the rolling di rection.
It is therefore the primary object of the instant invention to provide a system for automatically correcting the constant of the transfer function in response to variations in rolling conditions to ensure accurate control of the rolling mills.
The above and other objects, features and advantages of the instant invention will be apparent from the following descriptions taken in conjunction with the accompanying drawings, wherein:
FIG. 1 is a block diagram of the control system according to the instant invention;
FIG. 2 is a block diagram showing a tension control transfer function loop;
FIG. 3 is a block diagram showing a system arrangement for rolling steel sections in accordance with the present invention; and
FIG. 4 is a block diagram of the system according to the instant invention provided with a learning function.
DETAILED DESCRIPTION OF THE INVENTION Before proceeding with a description of the preferred embodiments illustrated in the accompanying drawings, it is in order to discuss the principles for control' ling and determining the constant of transfer function for steel rolling under tension free conditions between the respective rollers of a continuous steel section rolling mill.
The rolling schedule is generally determined by the reduction, temperature and load distribution characteristics of the particular rolling mill, and the rolling speed is determined in accordance with the rolling schedule which meets a particular rolling mass flow. However, the tension is not always controlled appropriately due to difficulty in speed control. It is therefore the usual practice to provide a tension free control system having a function transfer loop as shown in FIG. 2. However, there is herein described that the open loop transfer function is varied, according to variations in the rolling conditions and to the dimension of the rolling material.
The mathematical relation of the tensile force or compressive force T with the velocity V between two consecutive rollers or between a first and second roller (I) wherein A is the cross-sectional area of the material, E is Youngs modulus, L is the distance between rollers, V, represents the velocity of the rolling material emerging from the first roller and V represents the velocity of the rolling material entering the second roller. Hereinafter, the subscripts I and 2 indicate the first and second rollers, respectively. The relation between the velocity of the rolling material and the angular ve locities N and N of the rollers is VR2= 2) NR2 wherein F. represents the forward slip rate of the rolling material emerging from the first roller while 5 represents the backward slip rate of the material entering the second roller. Experiments have shown that the forward and backward slip rates of the rolling material vary linearly with the unit tension of the material. This is expressed as wherein a and B are constants, F represents a threshold or initial value of the forward slip ratio F represents a threshold or initial value of the backward slip rate A! ATM, and A represents the cross-sectional area of the steel material. If, in this instance, the angu lar roll velocities are varied by AN, and AN respectively, producing a tension T, the following equation is derived from the foregoing equations, (I) to (4).
AT is the tension of the steel section between the two stands;
A is the sectional area of the steel section;
L is the distance between the two stands,
E is Young's modules a and [3 are constants N is the circumferential velocity of the first stand roll;
N is the circumferential velocity of the second stand roll;
5 is the Laplace operator;
F is the initial value of the forward slip ratio of the first stand;
( is the initial value of the backward slip ratio of the second stand;
F, is the forward slip ratio of the first stand;
5 is the backward slip ratio of the second stand; On the other hand, in the tension control transfer loop shown in FIG. 2, the total transfer function (1(3) including transfer function G ls) of the rolling mill and the transfer function G (s) of the speed control device, if the constant of the high degree is disregarded, becomes wherein K represents a tension gain, T a tension time constant. and T a loss time. Here, the tension gain is m U ANR2 m) Ril/( NR] "8' R2) A (W and the tension time constant T is TM B N wherein L represents the distance between rollers and E represents Youngs modulus Thus, it will be understood that the tension gain and the tension time constant of the total transfer function GU) vary in accordance with the parameters 6, f, AM AN 0', ,3, L, E, A, N N Furthermore, if the transfer constant F(s) of the filter circuit is considered as being fixed, optimum control of the tension control loop is obtained by setting the compensating transfer constant K(s) of the tension control loop with regard to the abovementioned variables, as follows Km p+ l/ where K 7 X m/ m n) (lUl K, 5 X l/Km n (Ill wherein y and 6 are numerical constants, K is the proportional constant and K, is the integrated constant. Further, the tension time constant and the tension gain T and K respectively are expressed as functions wherein i represents an intermediate roller tension control number, A represents a sectional area of the steel material in the rolling mill, represents the constant. V represents a angular velocity of the roller. V represents the angular velocity of the first rollers in question, V represents the angular velocity of the second roller in question, K, representsfl L.E.A l, and K represents the ratio of the effective coefficient of backward slip rate to that of forward slip rate.
Thus, K, of Equation (9) is derived from (ll) and l2) as follows lt /Tm l/lCm fAMU fVU ml On the other hand, Kp of Equation (9) is derived from Equations (l0), (ll) and ([2) as follows KP! 7 X n X m 'Y X n fl nunzv mm) (l5) From the foregoing, it is possible to calculate the setpoints K and K, of the compensating transfer function K(s) for forming an optimum tension control loop.
A description will now be presented of a system of the instant invention which operates on the principles discussed hereinabove Referring first to FIG. 1, the reference characters 1A, [B and 1C designate paired rollers of respective roller positions. Indicated at 2A. 2B and 2C are the drive means for the rollers 1A, 1B and 1C, and at 3A, 3B and 3C are detectors connected to the drive means 2A, 2B and 2C for detecting the number of revolutions of the rollers. There are shown at 4A, 4B and 4C speed control means for the aforementioned drive means 2A, 2B and 2C. Shown at 5A and 5B are tension preventing control means. Shown at 6 is a tension computing control means such as a digital computer, and shown at 7A to 7C are rolling load detecting means.
FIG. 2 shows a tension control transfer function loop, wherein the loop is formed by a transfer function F(.r) of the filter circuit, an optimum compensating transfer constant K(s) of the tension control loop, a transfer function G,(s) of the speed control means 4A to 4C. and a transfer function (12(5') of the rolling mill. The loop produces current variation 8 resulting from the tension between two rollers. The current variation 8 produced by the tension between two rollers is fed to a summing point 9 at which a positive signal of the current in tensionfree rolling and a negative signal of the current variation 8 caused by tension are combined together so that tension control is carried out in accordance with the result.
ln the instant invention, the setpoints K, and K for the respective tension preventing control means are operated by the tension computing means 6 of FIG. I in accordance with Equations l2) and (I3) and utilizing other constants such as -y, 8, and T to produce a control output, before the rolling steel material enters the group of rollers of the continuous rolling mill. The respective tension preventing control means 5A and 5B are operated in accordance with the control output which is produced by the computer 6. When the steel material is passed through a first roller IA and then introduced into the next roller 18, it is detected by the detector 78 whereupon the tension preventing control means 5A carries out the optimum control after a time lag corresponding to the decreasing time of the impact speed. In a similar manner, when the steel material enters the roller 1C, the tension preventing control means 58 is actuated. According to the speed correcting values which are issued from the control means 5A and 5B, the rolling speed is adjusted through the speed control means 48 and 4C. The tension computing control unit 6 reads out the corrected speed to make the speed control in a subsequent operation more accurate.
It will be understood from the foregoing that the tension preventing control unit can provide optimum control by way of the theoretical computation which determines the tension control constant between respective rollers, taking into consideration the rolling conditions and the dimension of the steel material, instead of setting the control by experience. The following description presented by way of example, applies the instant invention to the continuous rolling of double-T section steel.
In the continuous rolling of the double-T section steel, the rolling material is maintained in a tension-free condition with no tensile nor compressive force acting thereon. In order to carry out the tension free rolling operation, it is the usual practice to employ the socalled current memory system wherein as the rolling material is compressed at a reference roller position, the current flowing through a drive motor is memorized for comparison with the current flowing through the roller driving motor when the rolling material is introduced into a succeeding roller. The speed of the succeeding roller is controlled in accordance with the difference therebetween. This conventional current memory system has difficulties in that, since it has a fixed control gain and employs sampling or proportional control, a long operation time is required for controlling the respective rollers and therefore it becomes difficult to obtain stabilized operation, particularly when the transfer time between the individual rollers is relatively short and/or there are relatively large variations in the rolling speed or in the size of the rolling material, due to the influences imposed when the succeeding rollers grip the rolling material.
As a result of various experiments and theoretical analysis carried out to elucidate the mechanism of producing tensile or compressive force which gives rise to an unbalance in speed between two rollers, it has been found that the relation between an unbalance in velocity and the force between the two rollers is expressed by a time lag system of the first order.
In this connection, if the speed of the reference roller is constant, that is to say, with AN, =0, we obtain from Equation (5) and from Equations (8) and (I6).
n L/E /l mi B ita/ ail] KM I zul az/ mi B NIH/NIH A If the rolling schedule is fixed, N /N a and ,6 become constant, so that That is to say, the time constant T is substantially inversely proportional to the angular speed N, ofthe reference roller. Similarly, the tension gain K, of Equation I8) is from which it will be understood that ATs becomes a stationary value which is proportional to a product of a percentage change in speed or the amount of unbalancing in the speed and the sectional area of the rolling material.
With (19) and (20), Equations (l4) and 15) are rewritten as K I/Ng K a I/A [Ill Thus, in order to provide control in the tension having a time lag of the first order, the basic control system should preferably include a control element having a transfer constant including a proportional constant and an integrated constant, these constants varying in accordance with the sectional area of the rolling material and the angular speed of the reference rollers, whereby the time constant of the control will be suitable for the particular mill characteristics, and will be controlled by the percentage in speed thereby maintaining the tension gain as a constant.
FIG. 3 shows a system for performing the control of the type discussed hereinabove, in the operation of continuous rolling mill for double-T steel sections. In FIG. 3, 10A to 10C are rollers corresponding to the rollers 1A to IC of FIG. 1. The rollers are driven by DC motors 11A to 11C, which in turn are powered by power sources 12A to I2C. The power sources are controlled by control units 13A to 13C. Current transform ers 21A to ZIC are connected between the motors and the power source. Connected to the transformers are filters 15A to 15C and, memories 16A to NBC. Indicated at 14A to 14C are speed rheostats, and 17A to 17C indicated constant setting units. Indicated at 22A to 22C are 23A to 23C are switches, and indicated at 24A to 24C are adders.
This system controls the tension in the following manner. The armature current of the drive motor 11A which controls the first roller 10A in an arbitrary section of all the rollers, is detected by the current transformer 21A and fed to the memory 16A through the filter ISA and a normally closed contact of the switch 22A. After the following material is gripped by the first rollers 10A and immediately before it is introduced into the second rollers 105, the switch 22A is changed over to its other contact. The memory 16A memorizes and stores the drive motor current of the first rollers (or the reference rollers) flowing immediately before the rolling material is gripped by the second rollers. The value of the current stored in the memory 16A and the value of the actual current through the other contact of the switch 22A is fed to the adder 24A, and the difference between the two currents AI is fed to the constant setting unit 17A through the switch 23A which is closed when the rolling material is passing through the second rollers )8. In response to a preset switch or a command signal from a computer, the constant setting unit 17A deterines the proportional contact Kp and the integrated constant K, in accordance with Equations (2i and (22) or Equations (l4) and (15 t, and produces a percentage change in speed signal AS, as an output. The signal AS serves as a speed change command signal for the drive motor [1B of the second rollers, to adjust the second roller speed in accordance with the current variation Al until the tension AT becomes 0.
Immediately before the rolling material is gripped by the third rollers, the switch 228 is changed over to its other position and the armature current of the driving motor 118 for the second rollers which in this instance is the reference roller, is memorized by the memory I158. The adder 243 gives a current variation Al to the constant setting unit l7B which produces a percentage change in speed signal As which in turn serves as a speed change command signal for controlling the driven motor 12C of the third rollers, until the current variation Al, and thus the tension AT becomes 0. in this manner, the tension between the successive rollers is controlled and maintained in a state.
With the above system, the speed change command signals AS, and A are also applied to all of the successive rollers. The speed of each roller should normally be in proportional relation when gripping the rolling material. However, if the speeds of all the succeeding rollers are corrected at the same time as the second rol ler, the amount of speed adjustment of each of the succeeding rollers may be reduced. In this way as the rolling material reaches each of the succeeding rollers, the operation time required for completing the control of that roller may be shortened to a significant amount.
FIG. 4 shows another embodiment of the instant invention which includes a learning function as an aid in controlling the tension. The control unit containing the learning function is indicated by the block enclosed by line 26. This embodiment is essentially similar in operation as the one shown in FIG. 3 except for the control unit containing the learning function, and thus description of parts similar to that of FIG. 3 is not given here to avoid repetition. In FIG. 4, a computer 27 is employed for setting the speed of the working rollers as well as the reduction value. Designated at 28 is a servomotor which is controlled by the computer 27 by way of reversing switches 30 and 31. The servomotor 28 operates a slide arm of the speed setting unit 14B, There is shown at 29 a shaft encoder which indicates the position of the slide arm of the speed setting device 14B and produces a positional signal in response thereto and applies this latter signal to the computer 27.
The operation of the learning control device 26 is as follows. The percentage change in speed signal AS, is fed to the computer 27, so that. when the current devintion Al becomes 0 upon completion of the tension control, a signal S is supplied and the computer 27 memorizes the change speed command signal A8,. A signal S is supplied when the tail end of the rolling material passes through the second rollers B, and the computer 27 then corrects the speed setting unit 145 according to the memorized value through the servo mechanisms 28 and 29. Thus, assuming the present position is expressed as P01 and the position after learning as Pnl, we can obtain the position of the slide arm of the speed setting unit 148 as ll PM: 1 010 +2 ASi/IOO) Of course, the learning term may be processed by means of the exponential average method or other suitable methods. With the control according to the invention, the roller speed setting for a subsequent rolling material is corrected by means of the actual value obtained from the preceding rolling material, so that the computerized rolling mill control using a model may be carried out far more accurately.
The percentage change in speed command signal AS, obtained from the second rollers IOB is used as a speed command signal for the third rollers and at the same time, controlling the succeeding rollers in the same manner. The computer 27 is capable of providing learning control also for the speed setting unit of the third rollers, by using a similar servo mechanism.
The learning tension control method has been herein described in connection with a current memory system, however, it will be appreciated that this method can be applied to systems other than the current memory system. In such a case, a speed deviation signal obtained from an output of a reference speed generator may be used in place of the control input ASi.
With the system according to the invention, a sufficiently large control gain is obtained in accordance with variations in the sectional area and the speed of the rolling material in a continuous rolling operation for the production of steel sections, so that tension control may be carried out optimumly at all times with the result that it is possible to obtain steel sections of high precision. The instant invention can also be applied to continuous rolling of bar stocks, billet mills, rails, sheet piles and the like, in the same manner as in the continuous rolling of steel sections as discussed hereinbefore.
Furthermore, the rolling mill control system of the present invention can be particularly advantageously applied to steel materials of complicated shapes. Also, it may be used to control rolling mills without the need of tension damping devices between the rollers.
A further advantage of the invention is that a signifi' cantly large gain is obtained through the learning control. The learning control may be attained by directly adjusting the speed setting unit 148 in accordance with the speed change command signals A8,, A5 and so on. However, this results in a far lower gain, as compared with the present invention employing a computer for the control of rolling mills.
What is claimed is:
l. A method for providing a zero tension in a steel section between at least two stands of a continuous rolling mill, comprising the steps of:
a. detecting the speed of the steel section when it enters into the first stand and when it emerges from the second stand;
b. determining a plurality of parameters of rolling conditions including at least the sectional area of the steel section between said first and second stands;
c. calculating a control constant based on said detected speeds and said plurality of parameters, so that both the forward slip ratio and backward slip ratio of the steel section may be linearly varied; and
d. modifying the speed of the steel section between said first and second stands in accordance with said control constant so as to maintain the tension of the steel section therebetween substantially zero.
2. The method of claim 1 wherein said step of determining the plurality of perameters includes determining the forward slip ratio and the backward slip ratio of the steel section.
3. The method of claim 2 wherein said step of determining the plurality of parameters further includes determining the dimension, shape, rolling speed and reduction ratio of the steel section and distance between individual stands.
4. The method of claim 3 wherein said step of calculating includes calculating the proportional gain and the integrated gain of a transfer function, one of said gains being related to the sectional area of the steel sec tion and the circumferential speed of a reference roll stand, to thereby select a time constant suitable for particular mill characteristics.
5. The method of claim 4 wherein said step of calculating further includes the step of obtaining a percentage change of speed signal.
6. The method of claim 1 wherein said rolling mill comprises a plurality of stands and wherein said step of modifying the speed includes simultaneously applying a speed signal to all the stands subsequent to said first stand, for controlling the respective speeds thereof.
7. The method of claim I, further comprising the step of correcting the stand roll speed for a subsequent steel section by means of the actual value obtained from the preceeding steel section, so that the computerized rolling mill control using a model may be carried out far more accurately.
8. The method of claim 1, further comprising the steps of:
e. memorizing the percentage speed change command signal when the tail end of the steel section is rolled in the second stand; and
f. correcting a speed setting unit for the second stand according to the memorized value through a servo mechanism.
9. A method for providing a zero tension in a steel section between at least two stands ofa continuous rolling mill, comprising the steps of:
a. detecting the speed of the steel section when it enters into the first stand and when it emerges from the second stand;
b. using a plurality of parameters of rolling conditions including at least the sectional area of the steel section between said first and second stands;
c. calculating a transfer function K(s) of the form Kl l (Ki/ where K,. is the proportional gain and K, is the integrated gain, and wherein:
where Cm, y and 8 are constants, V is the angular velocity of a roller, K represents f(L.E.A.), K represents the ratio of the effective coefficient of backward slip rate to that of forward slip rate and A is the sec tional area of the steel section. and
d. modifying the speed of the steel section between said first and second stands in accordance with said transfer function so as to maintain the tension of the steel section therebetween substantially zero.
10. The method of claim 9 and wherein for a particular rolling schedule K as l/N X HA and K, l/A wherein N, is the circumferential velocity of the first roll stand.
11. Apparatus for providing a zero tension in a steel section between at least two stands of a continuous rolling mill, comprising:
a. means for detecting the speed of the steel section when it enters into the first stand and when it emerges from the second stand;
b. means for determining a plurality of parameters of rolling conditions including at least the sectional area of the steel section between said first and second stands;
c. means for calculating a control constant based upon said detected speeds and said plurality of parameters, so that both the forward slip ratio and backward slip ratio of the steel section may be linearly varied;
d. means for controlling the speed of the steel section between said first and second stands, and
e. means for applying said control constant to said controlling means in the form of a percentage change of speed, so as to maintain the tension of the steel section between the roll stands at substantially zero.
12. The apparatus of claim 11 and wherein said means for detecting further comprises drive means for rotating each of said sets of rollers, source means providing a current to said drive means, the speed at which each of said sets of rollers rotate being controlled by the amount of current to said drive means, memory means for receiving the amount current controlling a given set of rollers at a time when the material is passing through said given set of rollers, comparison means for receiving the current controlling said given set of rollers at a time when the material has already entered the next successive set of rollers and comparing this last mentioned current with the current in said memory means and providing a current deviation signal.
13. The apparatus as in claim ll and wherein said control constants include the values K w l/N X 1M and K; m wherein N, is the angular velocity of the first roll stand, and A is the sectional area of the steel section.
M. The system as in claim 11 and further comprising learning control means for altering the operation of the sets of rollers in response to said speed correction signal.
15. The system as in claim 14 and wherein said learning control means comprises computing means receiving said percentage change in speed after the speed of the sets of rollers is modified, speed setting means for setting the speed of each of said sets of rollers, and servo mechanical means coupled to said speed setting means and under control of said computing means, said computing means causing said servomechanism means to control said speed controlling means to set the speed of each set of rollers when the rolling material has passed therethrough.
* it: n: a a