US3763678A

US3763678A - Apparatus for automatic control of product fill dimension

Info

Publication number: US3763678A
Application number: US00268085A
Authority: US
Inventors: P Dornbusch; T Johnson
Original assignee: General Electric Co
Current assignee: General Electric Co
Priority date: 1972-06-30
Filing date: 1972-06-30
Publication date: 1973-10-09
Anticipated expiration: 1990-10-09

Abstract

Apparatus for automatic control of the fill dimension of a product formed in a rolling mill by sensing the fill dimension of the finished product and comparing this value to a predetermined reference value to form a resultant error signal. This resultant error signal modifies the height of loops between successive stands of the mill to vary the product fill dimension. The apparatus includes provision for increasing the gain if successive errors of fill dimension in the same direction are sensed and for decreasing the gain if successive errors in the opposite directions are sensed.

Description

United States Patent [191 Dornbusch et a].

[4 Oct. 9, 1973 APPARATUS FOR AUTOMATIC CONTROL 3,526,113 9/1970 Mclslaughen 72/8 ()1? P O U FILL DIMENSION 3,650,135 3/1972 Skelton et al 72/8 [75] Inventors: Paul E. Dornbusch; Thomas D. Prima ry Exammer-Mrlton S. Mehr Johnson both of Roanoke Att0rneyArn0ld E. Renner et al. [73] Assignee: General Electric Company, Salem,

' Va. [57] ABSTRACT [22] Filed: June 30, 1972 Apparatus for automatic control of the fill dimension of a product formed in a rolling mill by sensing the fill [2]] Appl' 268085 dimension of the finished product and comparing this value to a predetermined reference value to form a re- [52] U5. Cl. 72/9, 72/16 sultant r r ign l- This re ltan e r r ign l modi- [51] Int. Cl B21b 37/12 ties e igh of lo p w n c sive s nd of [58] Field of Search 72/8, 9, 10, 11, the mill t vary the product fill dimension. The appa- 72/12, 16 ratus includes provision for increasing the gain if successive errors of fill dimension in the same direction 5 6] References Cited are sensed and for decreasing the gain if successive er- UNITED STATES PATENTS rors in the opposite directions are sensed. 3.251.207 5/l966 Wilson 72/12 .7 Claims, 4 Drawing Figures 26 27 LOOP LOOP LOOP LOOP MOTOR SENSO MOTOR SENSOR MOTOR SENSOR MOTOR SENSOR MOTOR 3| I l l L OP SPEED L P P ED P SPEED LOOP SPEED REG. REG. RES. E RES REG. RESv A 37 MAX. 3940 MIN.

| REFERENCE I BUZZER GATE 7 LOOP HEIGHT 7 LIMIT ERROR PATENTEDUET 9% 3,763,678

SHEET 10F 3 iii) FIG. 1

NOISNEHIO UHGTDQHS PATENIED DDT 91873 SHEET 30F 3 LET HEAD HAS CHED SENSOR YES BILLET TAIL HAS YES REACHED STAND 22 UNDER FILL 59 TOGGLE SET T No 6| DECREASE GAIN I SET OVERFILL TOGGLE RESET UNDERFILL TOGGLE YES UNDERFILL YES TOGGLE sET 48 OVERFILL TOGGLE sET DECREASE INCREASE GAIN T /50 MULTIPLY ERROR BY GAIN SET UNDERFILL RESET OVERFILL I sum WITH HEIGHT RE LOOP FERENCE TOGGLE TOGGLE APPARATUS FOR AUTOMATIC CONTROL OF PRODUCT FILL DIMENSION BACKGROUND OF THE INVENTION .changes to obtain the desired tension regulation throughout the mill speed range. This was generally accomplished by providing interstand tension sensing devices, such as a loop height sensor betweenadjacent stands in the mill, and using a signal from the sensor to vary the speed of the rolls which has the effect of varying the tension in the product.

In attempts to keep a constant product dimension, the ratio of the stand speed to mass flow is generally the controlling factor. Although the stand speed is relatively insensitive to variations and easy to control, the mass flow is relatively sensitive to product parameters and therefore is relatively difficult to control. The difficulty of controlling mass flow in the prior art limited the ability to maintain product dimensions.

Many modern, high production mills for forming rods, bars, etc., utilize a plurality of stands in a finishing train and have regulated loops between successive stands. The difficult problem of controlling the mass flow of a product in such mills has been substantially overcome by the practice of this invention. This is accomplished by sensing immediately after the final finishing stand the product fill dimension and comparing this dimension to a predetermined reference value. Any deviation from the desired fill dimension on the finished product forms an error signal and automatically modifies the loop heights as a function of the dimension deviation or error signal. The loop heights are regulated by changing the speed of a plurality of preceding stands.

SUMMARY OF THE INVENTION lt is therefore an object of this invention to provide an improvedautomatic control of the fill dimension of a product.

It is another object of this invention to provide an automatic control of mass flow to control the fill dimension of a product.

It is a further object of this invention to provide an improved automatic control of the fill dimension of a product by regulation of the speed of a plurality of preceding stands.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 2 is a simplified diagram of a portion of a mill train showing a regulated loop between two stands of the finishing train.

FIG. 3 is a simplified diagram of an automatic fill control for controlling the fill dimension in accordance with this invention.

FIG. 4 is a simplified logic flow chart further illustrating the operation of the automatic fill control.

DESCRIPTION OF THE PREFERRED EMBODIMENT Referring now to FIG. 1 of the drawings, there are illustrated three sectional views of a product, specifically in the form of a rod 11, after it leaves the last finishing stand and shaping has been accomplished. Two dimensions are indicated in FIG. 1, a shoulder dimension and a fill dimension. The shoulder dimension is the dimension perpendicular to the rolls of the aforesaid stand. The fill dimension is the dimension parallel with the rolls of the stand.

The shoulder dimension is usually determined primarily by the roll grooves and screw settings of the stands. It is relatively insensitive to variations in incoming product parameters and stand speeds; therefore, it is relatively easy to control. The fill dimension is controlled primarily by the ratio of stand speed to mass flow; it is sensitive to product parameters and relatively difficult to control.

Assuming that the mill is properly set up and running a good product, a satisfactory rod section will develop, as shown at llla in FIG. 1. The rod 11b in FIG. 1 shows the cross section of a finished product in an underfilled state, forming flats in the fill dimension. The rod in FIG. 1 shows the cross section of a finished product in an overfilled state, forming fins in the fill dimension.

Inspection of the geometry of the sections in FIG. 1 reveals that the relationships between the cross section area and the fill dimension is non-linear. This fact, in addition to other non-linearities in the process (such as roll spacing), results in wide variations in the relationships between mass flow and fill dimension.

FIG. 2 is illustrative of a portion of a modern, high production mill having a regulated loop between two stands of the finishing train. FIG. 2 illustrates a product in the form of a rod 12 moving in the direction indicated by arrow 13 and driven by horizontal stand 14 and vertical stand 15. Hold down rolls l6 and 17 keep the product 12 in contact with loop thrower 18 and the height of the formed loop is sensed at the center line 19 of the loop formed by the loop thrower.

FIG. 3 shows the product 12 advancing in the direction of arrow 20 and driven by

vertical stands

21, 23, and 25 and

horizontal stands

22 and 24. Between each horizontal and vertical stand, such as

stands

22 and 23 is a loop, corresponding to the loop shown in FIG. 2, the height of which is sensed by loop sensor 26. The motor 27 of stand 23 is regulated through a speed regulator 28 by a loop regulator 29 which compares an input from the loop sensor 26 with an input from an amplifier 30. The output of the loop regulator 29 is supplied to the speed regulator 28 for controlling the speed of the motor 27, and thereby controlling the loop height. Each of the

stands

22, 24, and 25 is also provided with a corresponding loop sensor, motor, speed regulator, and loop regulator. The loop regulator of each stand is connected to the amplifier 30.

In order to control automatically the plurality of stands forming the finishing train, the control circuit shown at the bottom portion of FIG. 3 is employed. After the product 12 travels through the final stand 25, the fill dimension of the product 12 is sensed by a sensor 31, which may be in the form of an infrared micrometer, located immediately after the final stand. sensor 31 produces a fill dimension error signal which is intergrated in calculator 32 for a fixed time. The calculator 32 also receives a signal from a loop height reference 33. The loop height reference is the loop height value against which the actual loop height sensed by loop sensor 26 is compared in the loop regulator 29. The signal from the loop height reference 33 is summed in the calculator with the error signal. The resultant signal is supplied through a gate 34 to the loop height reference and incrementally varies the loop height reference in proportion to the error signal. This proportion is automatically calculated to compensate for variations (non-linearities) in the relationships between mass flow and fill dimension. The loop height reference, which is varied as indicated, establishes a loop height which, if maintained and with conditions remaining unchanged, would result in a correct fill dimension. The loop height reference signal is supplied to the calculator, it is summed with the loop height reference to establish a new value for the loop height reference.

In addition to the automatic control signal from calculator 32 to gate 34, a manual input is also provided to gate 34. This manual input may be utilized if the sensor 31 is temporarily disabled or if, for any reason, it is desired to override the automatic control.

The output of the loop height reference is connected through digital-to-analog converter 37 to the amplifier 30 to provide a signal to the loop regulator for controlling the speed of stand 23 and thereby the loop height of the product between

stands

22 and 23. It will be noted from FIG. 3 that the signal from the amplifier 30 is connected not only to the loop regulator 29 of stand 23 but also to corresponding loop regulators associated with

stands

22, 24, and 25. Hence, all four loops are simultaneously adjusted, thereby quadrupling the change in fill dimension which can be achieved by adjusting a single loop.

Loop height reference 33 is also connected to a loop height limit device 38 which is equipped with maximum limit light 39, a minimum limit light 40 and a buzzer 41 to indiciate that the maximum or minimum setting of the loop height which can be controlled by the automatic fill control has been reached. Each of the loops may be automatically controlled by the control system of this invention so long as the loop height remains within a predetermined range corresponding to the maximum and minimum limits. If the loop height goes beyond this range, the appropriate light signals the operator to make appropriate adjustment of roll screws to bring the loop height into the range for automatic control.

The operation of the automatic control of this invention can be more clearly understood by reference to the further details of operation shown in the logic flow chart of FIG. 4. When a sensor, represented by logic unit 42, senses the head of the product, at stand 25, operation of the automatic control is initiated provided the tail of the product has not yet reached stand 22, as indicated by logic unit 43. If the above two conditions are met, fill sensor logic unit 44, corresponding to infrared micrometer 31 in FIG. 3, supplies an error signal to the calculator 32 in FIG. 3 where the fill dimension error signal is integrated over a fixed period of time. In the logic flow chart of FIG. 4, the elements included in the dashed block 32a indicate the functions performed in the calculator 32 shown in FIG. 3.

When the sensing function indicated by logic unit 44 is completed, the calculator 32 shown in FIG. 3 will have integrated the error signal over a period of time corresponding to the travel of a given length of product past the sensing point, and the resultant signal will be proportional to the average error in that length of product. Completion of the sensing function by logic unit 44 initiates error sensing logic in the logic unit 45. If no error exceeding the error preset in logic unit 45 is indicated, the error sensing logic unit 45 will restart the total logic system through a logic unit 46 incorporating overfill and underfill toggles (flip flops in an electronic system). In the flow chart, logic unit 46 is identified as resetting overfill and underfill toggles. It will be understood that this language indicates that under this condition the toggles (flip flops) are placed in their original condition, that is, not set.

Anytime an underfill greater than the error preset in logic unit 45 is sensed, the sequence of events is illustrated on the underfill (right side) of the logic flow chart. If this condition occurs, and neither the underfill toggle nor the overfill toggle is set at the time, as determined by

logic units

47 and 48 respectively, the logic unit 49 sets the underfill toggle and resets the overfill toggle, that is, places the overfill toggle in its original (non-set) state.

In the preferred embodiment of this invention, a single toggle or flip flop comprises both the

logic units

48 and 55 identified as overfill toggles in FIG. 4 and a single toggle or flip flop comprises both the

logic units

47 and 56 identified as underfill toggles in FIG. 4. It will, therefore, be understood that the setting or resetting of

logic units

47 and 48 shown on the right, or underfill, side of the flow chart in FIG. 4 results in the setting or resetting of the corresponding overfill and underfill toggles of

logic units

56 and 55 shown on the left, or overfill, side of the flow chart. Thus the setting of the toggle by any logic unit, such as logic unit 47, will perform an identical function in the other corresponding logic unit 56 in logic flow chart of FIG. 4.

Logic unit 49 is connected to a logic unit 50 where the error signal is multiplied by the gain. This multiplied signal is summed with a signal from the loop height reference 33 (FIG. 3) in logic unit 51 (FIG. 4) to produce a new and lower loop height reference. The new loop height reference causes all four loops as depicted in FIG. 3 to decrease in height. This results in an increased fill dimension after a transient time sufficient to permit the loops to reach a steady state condition at the new height plus the time required for the product to travel from the stand 22 to the sensor 31.

After a correction, operation of the sensing logic unit 44 (i.e., sensor 31 in FIG. 3) will not be initiated again until after the aforementioned transient time plus the aforementioned product travel time, as indicated by the logic unit 54.

In a similar manner, anytime an overfill greater than the error preset in logic unit 45 is sensed, the sequence of events is illustrated on the overfill (left side) of the logic flow chart. If this condition occurs, the loop height reference will be increased. This causes all four loops to be increased in height, resulting in decreased fill dimension. This is accomplished by connecting the logic unit 45 to overfill logic unit 55 which in turn is connected to logic unit 50 through the serially connected underfill logic unit 56 andlogic unit 57. Logic unit 57 sets the overfill toggle and resets the underfill toggle, that is, places the underfill toggle in its original (non-set) state. This affects the state of

logic units

55 and 56 and also the state of

logic units

47 and 48.

After a correction, reading by the fill sensor indicated by the logic unit 44 will not be initiated again until after transient time sufficient to permit the loops to reach a steady state condition at the new height, plus the time it takes the product to travel from stand 22 to the sensor.

If two consecutive underfill or two consecutive overfill errors greater than the preset error of logic unit 45 are detected, the gain will automatically change by an amount calculated to minimize the error after the second correction. With two consecutive underfill errors, this is accomplished on the underfill portion of the logic flow chart. Logic unit 47 is set by the first sensing of an underfill error exceeding the preset error; when the second underfill error is sensed, logic unit 47 is then connected to logic unit 49 through increase gain logic unit 58. Logic unit 58 causes the gain multiplied in logic unit 50 to increase by an amount calculated to reduce the error to zero.

Likewise, with two consecutive overfill errors, the gain is changed as illustrated on the overfill portion of the logic flow chart. Logic unit 55, which is set by the first sensing of an overfill error exceeding the preset error, is connected to logic unit 57 through increase gain logic unit 59. Logic unit 59 then causes the gain multiplied in logic unit 50 to increase by an amount calculated to reduce the error to zero.

When an overfill is followed by an underfill, the operation is illustrated as follows. Since there was no preceding sensing ofan underfill error, the underfill logic unit 47 is not set. Since there has been a preceding sensing of overfill error, however, logic unit 48 has been set. Under these conditions logic unit 48 is connected to logic unit 49 through decrease gain logic unit 60. This decreases the gain by an amount calculated to reduce the error to zero.

When an underfill is followed by an overfill, the sequence of operation is illustrated on the overfill portion of the logic flow chart. Since overfill logic unit 55 is not set and underfill logic unit 56 is set, logic unit 56 is connected to logic unit 57 through decrease gain logic unit 61. This decreases the gain by an amount calculated to reduce the error to zero.

To relate the significance of gain as discussed above in connection with the flow chart of FIG. 4 to the physical arrangement of the mill as shown in FIG. 3, it is observed that gain in the automatic control system of FIG. 3 refers to the magnitude of loop height change initiated by a given increment of measured fill error. The effect of a change in loop height on the fill dimension is not constant, differing, for example, dependent upon whether overfill or underfill is involved and further involving a non-linear relationship. The system illustrated in FIG. 4 adapts to varying conditions by modifying the system gain when successive samples show error in the same direction, indicating undercorrection, or in opposite directions, indicating overcorrection.

In operation an operator would provide coarse control of the fill dimension by manually adjusting the speed and/or screws on the preceding stands until the fill dimension is in the range of the automatic fill control.

If the sensor 31 senses an oversized fill dimension, it will apply a higher loop height reference. The resulting higher loops cause higher pull back force (back tension) on the product 12. The higher back tension reduces the mass flow, thus reducing the product fill. The reduced mass flow in the delivery stand would cause the loop height to increase beyond the new loop height reference; however, the loop sensor senses loop height above the new reference and automatically increases the speed of the delivery stand to maintain the loop at the new height. The higher stand speed increases the mass flow to maintain the mass flow at a constant valve. The net effect is as follows:

1. Higher loop height 2. Increased back tension 3. Higher delivery product linear speed 4. Constant mass flow 5. Reduced product fill dimension.

In one embodiment of this invention the variation in one loop results in fill dimension variations of between 5 to 10 mils in diameter. Therefore, in order to automatically control the fill dimension of a product in a range of at least 20 to 40 mils in diameter, an automatic control system involving four loops, as illustrated in FIG. 4 is employed.

While a specific embodiment of this invention has been shown and described, it will be apparent to those skilled in the art that modifications are possible without departing from the inventive concepts herein described. It is intended, therefore, to cover by the appended claims all modifications falling within the spirit and scope of this invention.

What is claimed as new and desired to be secured by Letters Patent of the United States is:

1. Apparatus for the automatic control of the fill dimension of a product formed in a rolling mill, said apparatus comprising: v

a. a plurality of stands for performing a work function on the product;

b. sensing means for sensing the fill dimension of the product after completion of the work function on the product;

c. comparator means for comparing the fill dimension with a predetermined reference value to form a resultant error signal; and

d. modifying means for automatically modifying the speed of the plurality of stands as a function of said error signal to control the product fill dimension.

2. Apparatus as recited in claim 1 wherein the means for modifying the speed of the plurality of stands includes means for automatically developing a signal so proportioned to said error signal as to compensate for non-linearities in the relationship between mass flow of the product and the product fill dimension.

3. Apparatus as in claim 1 wherein said sensing means is an infrared micrometer.

4. Apparatus for the automatic control of the fill dimension of a product formed in a rolling mill, said apparatus comprising:

a. a plurality of stands for performing a work function on the product;

b. each of said stands including a means for forming loops in the product during the performing of a work function on the product;

c. sensing means for sensing the fill dimension of the product after completion of the work function on the product;

d. comparator means for comparing the fill dimension with a predetermined reference value to form a resultant error signal; and

e. modifying means for automatically modifying the height of the plurality of loops as a function of said error signal to control the product fill dimension.

5. Apparatus in claim 4 wherein the means for modifying the height of the loops includes means for automatically developing a signal so proportioned to said error signal as to compensate for non-linearities in the relationship between mass flow of the product and the product fill dimension.

6. Apparatus for automatic control of the fill dimension of a product formed in a rolling mill, said apparatus comprising:

a. a plurality of stands for performing a work function on the product; p b. means for forming a loop in the product between successive stands, each loop having a height within a predetermined range;

d. comparator means for comparing the fill dimension with a predetermined reference value to form a resulting error signal; and

e. means for receiving said error signal and transmitting an output signal to effect a change in the height of said loops dependent upon said error signal;

f. said last-named means including means for modifying said output signal to increase the magnitude thereof when said sensing means indicates two successive fill errors in the same direction and to decrease the magnitude thereof when said sensing means indicates successive fill errors in opposite directions.

7. Apparatus as recited in claim 6 wherein said means for receiving said error signal includes means for multiplying said error signal by a variable gain, and wherein said gain is increased when said sensing means indicates two successive fill errors in the same direction and said gain is decreased when said sensing means indicates successive fill errors in opposite directions.

Claims

1. Apparatus for the automatic control of the fill dimension of a product formed in a rolling mill, said apparatus comprising: a. a plurality of stands for performing a work function on the product; b. sensing means for sensing the fill dimension of the product after completion of the work function on the product; c. comparator means for comparing the fill dimension with a predetermined reference value to form a resultant error signal; and d. modifying means for automatically modifying the speed of the plurality of stands as a function of said error signal to control the product fill dimension.

4. Apparatus for the automatic control of the fill dimension of a product formed in a rolling mill, said apparatus comprising: a. a plurality of stands for performing a work function on the product; b. each of said stands including a means for forming loops in the product during the performing of a work function on the product; c. sensing means for sensing the fill dimension of the product after completion of the work function on the product; d. comparator means for comparing the fill dimension with a predetermined reference value to form a resultant error signal; and e. modifying means for automatically modifying the height of the plurality of loops as a function of said error signal to control the product fill dimension.

6. Apparatus for automatic control of the fill dimension of a product formed in a rolling mill, said apparatus comprising: a. a plurality of stands for performing a work function on the product; b. means for forming a loop in the product between successive stands, each loop having a height within a predetermined range; c. sensing means for sensing the fill dimension of the product after completion of the work function on the product; d. comparator means for comparing the fill dimension with a predetermined reference value to form a resulting error signal; and e. means for receiving said error signal and transmitting an output signal to effect a change in the height of said loops dependent upon said error signal; f. said last-named means including means for modifying said output signal to increase the magnitude thereof when said sensing means indicates two successive fill errors in the same direction and to decrease the magnitude thereof when said sensing means indicates successive fill errors in opposite directions.