SE464465B - PROCEDURES AND DEVICES FOR SEATING AND REGISTRATION OF A ROLLING SPACE AND GEOMETRY OF A STRING MOLDING MACHINE - Google Patents

Description

464 465 A disadvantage of the known devices is that these do not allow a continuous measurement while the device is moved along the runway. Another disadvantage is that these devices do not allow the measurement of casting thickness other than in the case of non-eccentric, i.e. shown, opposite each other at the casting gap arranged along a line perpendicular to the length of the runway direction.

DESCRIPTION OF THE INVENTION The object of the present invention is to provide come a method and a device for measurement and registration of thickness and geometry on a casting gap between rollers in two parallel rollers in an extruded machine, having, for example, a vertical zone, a bending zone and a horizontal zone, which does not exhibit the disadvantages of known devices have.

The purpose has been achieved with a procedure according to present invention wherein one or more measuring blocks brought into contact with two support rollers in the first roller banana. The method is characterized in that a first means are caused to measure the distance between a key plane to the support rollers, or a parallel level, and the other rollers in the first roll the path within the length of the measuring block perpendicular to this plane, that the corresponding distance is measured to the rollers in it the second runway and that the measured values are registered and carried with corresponding values based on the runways geometry in an ideal casting gap. The procedure is characterized further by transforming the measured values into one for all the rollers common coordinate system and that an arbitrary Û function is adapted to the positions of the rollers in the joint the coordinate system.

The device according to the invention comprises a measuring block, intended to be brought into contact with two support rollers in the first runway. The device is characterized in that it comprises at least one first measuring means which is 464 465 arranged to measure the distance between a key plane to the support rollers, or a parallel level, and the other rollers in the first roll the path within the length of the measuring block perpendicular to this plane and by at least a second measuring means arranged to measure the distance between the key plane and the rollers in it second runway. The device is further characterized by that the measuring block has two ends, which are intended to support against each its support roller in the first runway, that it the first measuring device supports a in relation to the the intermediate roller and that the second measuring device net has two or more measuring points, eg measuring rollers or the like, which are arranged to measure the distance between the reference plane and the rollers in the second runway.

The measuring means are connected to each of the sensors which emit signals to a multiplexer controlled by a computer to sequentially connect one sensor at a time to an analog / digital converter which is also controlled by the computer and in which analog voltage signals are converted to digital information. The sensor includes, for example, an optical sensor whose signal is converted in a converter to tric voltage or current signal whose magnitude is one function against the measuring distance.

The invention has provided a and a device for measuring and recording thickness and geometry of a casting gap that is better than hitherto known methods and devices. Especially the invention allows continuous and rapid measurement of the roller tracks in a casting slot in which the rollers can have one completely independent placement in the first lane in relation to the other of the two webs at the casting gap.

DESCRIPTION OF FIGURES The invention is described below in the form of an embodiment examples in connection with the accompanying figures.

Figure 1 schematically shows a runway to one continuous casting machine in a vertical longitudinal section. 464 465 Figure 2 shows a detail of the runway according to figure 1 with entered measured values.

Figure 3 shows the principle for a recalculation of it to a common coordinate system.

Figure 4 shows a coordinate system embedded on the runway loose side.

Figure 5 shows a coordinate system embedded in the fixed side and loose side of the track.

Figure 6 shows the straight part of the runway adapted to a straight line line.

Figure 7 shows the bending zone of the runway adapted to one circle.

Figure 8 shows a graphical presentation of measurement results taten.

Figures 9 and 10 show block diagrams of the measuring system according to the invention.

Figure 11 shows a measuring block for a measuring car which carried along the runway.

Figure 12 shows a calibrator for the measuring block according to Figure 11.

The runway is defined as the envelope of the runway the surfaces on the inside of the casting gap. The upper part of the runway called loose side. The lower part is called the fixed side. For casting the parts of the column mathematically describe the single envelope a straight line, a circular arc with a fixed radius, or a continuous transition where the curvature changes from a circular arc to a straight line. Figure 1 shows the runway with envelopes. When the casting string passes through the casting gap (from point A to B in the figure), comes the casting string to form after the inside of the casting gap. When rolling in the roller conveyor which deviates from the ideal position, the shell to deform, which can lead to cracking formation in the casting shell. The setting of the rollers has therefore crucial for the quality of the casting string.

The casting gap is denoted by C in the figures. 464 465 LN To measure and check the thickness of the casting gap and the relative positions of the rollers, i.e. the roller tracks geometry, a measuring block is placed over three adjacent ones rolls. The block supports the outer rollers on the so-called fixed the side of the continuous casting machine, i.e. the lower one roller conveyor, and moved along the casting gap and the position from a straight line through the support points of the measuring block to the intermediate roller. The distance is also measured to the rollers on the so-called loose side, i.e. the upper roller side banana. Figure 2 shows the measurement principle for an arbitrary group of reels. The measured values are called raw coordinates. The are coordinates in local coordinate systems as described of the key to the outer rollers (abscissa) and the normal to this through the center of one of the outer roll (ordi- nata). The raw coordinates are denoted by Hf (j), for side and Hl (j) for the loose side. Both Hf and Hl are measured as the distance from the straight line through the con- tact points against the rollers denoted (j) and (j + 2) to roll (j + 1). If the total number of rolls is N, it will to be N-2 raw coordinates of the type described above in the string.

Each measured value, ie raw coordinate, is specified in one local coordinate system (š, Q) for food block.

The direction of the coordinate system is determined by the direction that the block takes over when it rests on two of the rollers. Coordinate the direction of the network system can thus vary from a value to another, eg in the bending zone, as shown of Figure 2 where the two local coordinate systems (š / nzi) and (š?: Q ”) are displayed. One can thus not directly compare the different measured values.

To compare the position of several rollers, one must transform the raw coordinates into one for the rollers common coordinate system. The X-axis of the joint the coordinate system may coincide with the key to two arbitrary, rolls. See figure 3. For practical reasons has the direction of the measuring block, which it assumes when it supports 464 465 on the first and third roll, selected as the coordinate the direction of the system. If you know the direction of the to the earth coordinates (horizontal and vertical direction) the geometries of the rollers can be expressed in these coordinates through a simple coordinate transformation.

However, this is not necessary if one is only to set the relative positions of the rollers and calculate the the thickness. As an example, a group is selected if M the following rolls. The first roll begins in an arbitrary roll which, for the sake of simplicity, we have chosen to designate with O second roll l o s v to the last roll denoted nas M-1. The X-axis of the common coordinate system runs through the contact points of the measuring block against rollers 0 and 2.

The Y-axis passes through the center to roller O.

One can mathematically derive a connection between The H coordinates and the XY coordinates of the rollers tangent points on the fixed side of the runway. Samban- it is difficult to express with analytical formulas though it can be recursively calculated in a computer. The calculation reproduces the XY coordinates exactly about Hf, Hl and Lf, Ll according to Figure 2 are precisely indicated.

The loose side and fixed side rollers are measured with block supporting on the same rollers on the fixed side. Normally you get a measured value for each support position as the measuring block intagit. However, there may be borderline cases where nothing or two measured values are obtained in the same measuring mode. This is because fixed and loose side rollers are not always placed opposite each other along a line perpendicular to the casting column. Figure 4 shows the measured values in relation to the block and the local coordinate system (šš ,?) If one imagines a line parallel to the medkš axis with the local coordinate system then constitutes the loose side coordinates the roll's tangent point to the line. The measured rolling distance Hl constitutes the 2 value and the rolling distance 464 465 distance (Ll) from the first roll of the fixed side its šï- value. The distance can be determined, for example, from a drawing substrate or measured. After XY coordinates have calculated for a group of rollers on the fixed side is for each of the local coordinate systems known in relation to the coordinate system that passes through the first and third roll on the fixed side of the group. Rikt- is denoted by an ip and is determined on such manner previously indicated. Consequently, reshape the loose side coordinates to the same XY system as the fixed side and thereby determine the fixed and loose side rollers relative to each other. Figure 5 shows this.

Assuming that the rolling distance in the length of the direction at each measurement occasion corresponds to the drawing the basis, a certain simplification may be allowed in the raw coordinates. You then do not need to measure rolling distance, which can be cumbersome, unless the towing speed of the measuring block is not constant. One such assumption of course introduces a certain measurement error in the specified rolling distances that affect the accuracy of the calculated XY coordinates. It's hard to give one analytically accurate measure of the error. This is because it depends on several parameters, including how many rollers are used when analyzing the deviations. In the bending zone, the error is added both in the X and Y directions. In a horizontal straight section it is added mainly in the Y-direction.

Both gap thickness and deviations between the rollers on the fixed and loose side can be determined from a series XY coordinates determined as previously specified. The the measured values must then be compared with it original model for gap width and roll deviations must be able to be determined.

Mathematical curve fitting is the most applied the method of adapting measured coordinates to a model. However, it requires that the model can be expressed as a mathematical function. The method is called linear regression 464 465 Zion. A mathematical function is then adapted over M rolls. The first reel can be selected arbitrarily in the string because one only seeks to find deviations from it assumed curve.

In straight parts of the casting gap, the adjustment takes place suitably to a straight line. See Figure 6.

Y1 = q1 X1 + qo f _ P1 xf + Po (loose side) F < IN (fixed side) The adaptation takes place, for example, with the least squares method, ie so that the square sum of the deviation between right and the left-hand side of the equations is minimum. When calculating of gap width and roll deviations in the straight section should be divide the shaving part into a suitable number of groups, e.g. 5-10 rolls in each group. There may be a discrepancy i.e. an error also in these points. The custom straight the line will still go through these. To decide the error in the contact points must be the coordinate system rotated so that it coincides with the custom one line. The curve adjustment should therefore take place in steps according to following: 1. Determination of the constants pO and p1 for regression line to the coordinate points of the original the coordinate system XY (which passes through the first and third the roll for an arbitrary group of rolls). Transformation of the original coordinates for both the fixed and loose side so that they coincide with the regression line obtained in step l. 3. Calculation of gap width and roll deviations.

The mathematical solution for curve fitting to a straight line is a standard solution in statistics.

The solution is therefore not described. The coordinate system XY constitutes the local coordinate system of the group scrolling which is analyzed. X'Y 'is a coordinate system whose X 'axis coincides with the regression line plX + po. 464 465 Curve adjustment for the loose side can then be done in X'Y ' coordinates whereby the distance from the matched line to the X 'axis constitutes the mean gap width.

The curve adjustment in the transition zone can take place to one polynomial of appropriate degree. Polynomial derivative i the transitions between a straight part resp. circle part must match the derivatives of the straight line respectively circle bows. The curvature must also be consistent in these points. After the degree number of the polynomial is determined, it takes place the curve fit in the same way as for the straight line.

The difference is that the number of constants to be are met by the minimum conditions are several. Even polynomials adaptation is a mathematical routine and does not need to be described vase. However, it should be borne in mind that any errors in the rollers shift the coordinate system so that the error becomes zero there. Therefore, the resulting curve must be trans- reshaped in the same way as indicated in the three steps above, before the deviation can be determined.

In the bending zone, the XY coordinates are adjusted to a circle.

From a given arbitrary roll in the arc of the circle, M is selected subsequent rollers to be included in the curve fitting.

The first subsequent roll is designated 1. The last roll denoted M. The radius of the circle is R. Its origin in relation to to the local coordinate system by first and the third roll is (a, b). The equation for the circle can written in local coordinates on the form: 2 2 2+ (Xi -a) It is now necessary to determine a and b so that the error in the adapted circle becomes the minimum in the least square todens meaning. In the general case, this leads to a very complicated nonlinear system of equations but by assuming that a and b are the catheters of a tri- angel with the hypotenuse R, the problem can be simplified. This namely leads to a linear system of equations from which a and b can be solved iteratively. The coordinate system 464 465 10 will then touch the circle closer to the new the origin of the coordinate system. When the coordinate transformation done, a and b are calculated again (iteratively) according to what as previously stated. The iteration is repeated until a = o where b = r. Normally this happens after 2-3 iterations.

Finally, the deviation oak from the obtained is calculated the circle. See Figure 7. e = R- xk2 + (Yk-R) 2 When adjusting the coordinate points to a circle is from a mathematical point of view, it is beneficial to have many points. This gives a more secure value of the obtained the circle. For practical reasons, however, one should not take everything for many as the error in the rolling distances otherwise becomes noticeable bart. About 5 - 10 rolls should be suitable ..

A graphical representation of the position of the rollers is displayed on figure 8. The graphic image constitutes the center lines for the fixed and loose sides of the string. The center lines have been straightened out to a straight line for the sake of simplicity and rolling relative positions relative to the center line (the custom curve) has been plotted. If the measurement takes place in several axial planes, each plane can be drawn on the same way.

As a complement to the graphic image can a list over the displacements of the rollers in relation to other or in relation to the expected the width is obtained. The offsets can be listed for all rolls or only for some of them, e.g. those that protrude too far into the mold gap or lots that have too large or too small a column width.

A complete measuring system consists of one or more measuring sensor 1, a measuring instrument 2 and devices 3 for result presentation can refer to a drawing board, a minal, a printer or a central computer. See Figure 9.

The function is that the measuring sensor 1 during the measuring process continuously provides information about the measuring distance to the measuring instrument 2. In this the information is processed 464 465 11 and the number of measurement data is reduced to the appropriate number.

Through analysis of measurement data, the number of rollers can also be determined. are reconciled and a relative distance value is obtained for each roll. Depending on the scope of software in the measuring instrument, further calculations can be made. Ex a curve adjustment can be made and the deviation of each roll from the ideal mode is shown in tabular form. Also limit values can be entered.

The measurement signals from the different sensors 1 are connected to one multiplexer 4. This is controlled by a computer 5 that sequences connect one sensor at a time to an analog / - digital converter 6 which is also controlled by the computer 5. I the converter 6 converts the analog voltage signals to digital (digital) information.

During the measurement, approximately 300 measurement values are generated per second. Most of these values are uninteresting and shall be sorted out. Only the values that represent the minimum distance between the roller and the measuring device shall saved. Therefore, all measurement data is continuously processed by the computer after which they come in for the calculation of the minimum distance. When this is determined, the measured value for is stored each roll partly in the electronic memory (RAM) in computer, partly on a diskette in a diskette drive 7. Di- the sketch unit has two functions: 1. Reading from diskette of instructions to the instrument for determining its function. 2. Storage of measurement data as an additional security measure the electronic memory would be erased, for example due to voltage failure or incorrect operation.

To operate the instrument (tell when measurement starts and ends, loads programs, etc.) there is one display 8 (a screen for characters) and a keyboard 9.

The display shows, among other things, instructions to the operator as well as acknowledgment of given commands. In the instrument there is also a date and time generator 10 (clock) so that each measurement is automatically marked with the correct date and time. The measuring instrument is also battery-powered and 464 465 12 maintenance is charged between the measurement periods via an external charger connected between the mains and the instrument.

The instrument is also equipped with two outputs for connection to external devices. It can be a central computer that can retrieve and provide information to the meant. In that case, the instrument is connected to it existing data network and some simple commands are given from the keyboard of the instrument after which the transfer controlled by the program in the central computer. ' Another option is to connect a printer to the instrument. It is conveniently done via a parallel output. Printing can be done of the measured values for each roll. Only deviations greater than a given limit value or what the instrument has been programmed to calculate is printed. The printout is requested via the keyboard.

The measuring device according to Figure 11 consists of a measuring carriage comprising one or more measuring blocks II in the form of one elongated chassis 12 at each end of which is a support plate 13 is arranged so that resp. support plate 13 is located in one and the same plan and which are intended to support against each of its roller conveyor roll 14. The length of the support plates ll in the longitudinal direction of the roller conveyor is such that they allow distance between the roller tracks 14,17 in the first the roller coaster. A first measuring boom 15, is rotatably mounted on one side of the chassis 12 about a shaft which is para- with the axles of the support rollers 14. The first measuring boom 15.1 carries at its free end a measuring roller 16, which is located between the two support plates 13 and which abuts a roller conveyor 17.1 which is located between then the two support rollers 14, in the first roller conveyor, which are connected to the support plates 13. A second measuring boom 15.2 is rotatably arranged on the other side of the chassis 12 about an axis parallel to that of the rollers 14 axles. The second measuring boom 15.2 carries at its free end of a measuring roller 16, which abuts against a roller conveyor roller 17.2 located in the second roller conveyor. 464 465 By giving the support plate 13 a certain minimum length in the longitudinal direction of the runway, the measuring block ll can be displaced in the longitudinal direction of the runway with unchanged reference plane sà that the measuring roller 16 on the second measuring boom 15.2 can bring to pass the current roller conveyor 17.2 in con- pace with this in particular on this roller conveyor 17.2 is eccentric, i.e. next to the midpoint the normal from the tangent between the support rollers 14, i.e. to the connection line between the contact block 11 of the measuring block points against the support rollers 14.

In alternative embodiments of the invention, the other is the measuring boom 15.2 provided with three measuring rollers 16, which are arranged in the longitudinal direction of the measuring block 11, two of which is shown dashed-dotted in figure ll, with the help of which man can measure distance from the reference plane to a roll 17.2 in the second runway which is strongly eccentric in relation to a corresponding roll 17.1 it first runway.

Through the first-mentioned embodiment, one achieves that one some eccentric placement of the rollers in the other the roller conveyor can be tolerated by the displacement of the measuring block II on the support rollers 14. Through the alternative embodiment in combination with the former, an additional greater eccentricity is tolerated, i.e. an arbitrary placement of the rollers in the second roller conveyor in spirit to the rollers in the first runway is accepted.

A sensor 18 is arranged on the chassis 12 next to resp measuring boom 15 on which a corresponding measuring plate 19 is arranged. The sensor 18 partly comprises a measuring body (sensor) and an adaptive electronics (converter). The output signal from the sensor consists of a voltage or current whose size is a function of the measuring distance. The output signals from multiple sensors can be connected to the multiplexer 4 (channel selector) to be connected sequentially 464 465 14 analog / digital converter 6 which provides a numerical value for further treatment.

Measurement of the distance between two roller conveyor rollers 17 takes place by connecting the measuring trolley to a so-called start chain to a continuous casting machine, not shown in the figures, and displaced by this chain along the machine roller coaster. Measured values are then registered on the measuring rollers 16 passes each roller in the roller conveyor and is converted according to the description above.

The measuring system is regularly calibrated using one calibrator according to figure 12. The calibrator consists of a base plate 20 which is screwed onto the measuring block. ets ll support plates 13. A calibrated measuring sensor 21 (electronic micrometer) is attached in the middle of one of the the roller 16 so that it can be moved within its measuring range by turning the micrometer. The output signal from the micrometer tern 21 is connected to the instrument via a 7th channel.

As the micrometer 21 is rotated, both as setpoints which are read into the instrument. Measuring sensor even 21 reference distances are denoted by D in Figure 12.

After a measuring roller 16 has been brought in this way to move from min. to the max. mode one or more times is given a command that automatically sets the length coefficient in the instrument for this particular sensor. Micro- tern 21 is then moved to the second measuring roller 16 and the process is repeated. The calibrator is moved then to others resp. third measuring block 11. Measuring the system is now fully calibrated. f)

Claims (9)

464 465 15 PATENT CLAIMS A method for measuring and recording the thickness and geometry of a casting gap between rollers in two parallel roller tracks in a continuous casting machine having, for example, a vertical zone, a bending zone and a horizontal zone, wherein one or more measuring blocks (11) are brought into contact with two support rollers (14) in the first roller track, characterized in that a first measuring means (15.1) is caused to measure the distance between a key plane to the support rollers (1U), or an arbitrary plane parallel thereto, and the other the rollers (17.1) in the first roller conveyor within the length of the measuring block perpendicular to this plane, that the corresponding distance to the rollers (17.2) in the second roller conveyor is measured and that the measured values are registered and compared with corresponding values based on the roller conveyor geometry in an ideal casting gap. Method according to Claim 1, characterized in that the measured values are transformed into a coordinate system (x, y) common to all the rollers, and that an arbitrary function is adapted to the positions of the rollers in the common coordinate system. 3. A method according to claim 2, characterized in that the x-axis of the x, y coordinate system is caused to coincide with the tangent of two arbitrary rollers (14). U. A method according to claim 3, characterized in that the positions of the rollers in the two roller tracks are transformed into the x, y coordinate system. 5. A method according to claim 4, characterized in that the coordinates of the rollers in an earth coordinate system are calculated by a coordinate transformation. Method according to Claim 5, characterized in that the x, y coordinates are adapted to a straight line in the straight parts of the runway. Method according to Claim 5, characterized in that the x, y coordinates are adapted to a circle in the bending zone of the roller conveyors. 464 465 16 Device for measuring and recording the thickness and geometry of a casting gap between rollers in two parallel roller tracks in a continuous casting machine having, for example, a vertical zone, a bending zone and a horizontal zone, wherein at least one measuring block (11) is brought into contact with two support rollers (14) in the first roller track, characterized in that it comprises at least one first measuring means (15.1), which is arranged to measure the distance between a key plane to the support rollers (1ü), or one with this parallel arbitrary plane, and the other rollers (17.1) in the first roller track within the length of the measuring block (11) perpendicular to this plane, and by at least one second measuring means (15.2), which is arranged to measure the distance between the reference plane and the rollers (17.2 ) in the second runway. Device according to claim 8, characterized in that the measuring block (11) has two ends, which are intended to support each support roller (1ü) in the first roller track, that the first measuring means (15.1) abuts one in relative to the support rollers (14) intermediate roller (17.1), and that the second measuring means (15.2) has two or more measuring points, for example measuring rollers (16) or the like, which are arranged to measure the distance between the reference plane and the rollers. (17.2) in the second roller conveyor both in the case of rollers (17.1, 17.2) arranged in pairs in the two roller conveyors, and in the case of rollers which are offset in relation to each other in the longitudinal direction of the roller conveyors. 11- '