MXPA97000530A - Method and apparatus for detecting the condition of liquid metal liquid in and from a filling concrete - Google Patents

Abstract

The present invention relates to an apparatus for detecting the condition of the liquid metal flow within or from a filling vessel, comprising: a sensor for detecting vibration caused by a liquid metal flow in and from the filling vessel and for outputting a sensor signal corresponding to a quantity of vibration detected by the sensor, a signal processor for receiving the sensor signal and for comparing the sensor signal with a reference signal, and outputting a comparison signal; and a logic unit for receiving the comparison signal and for outputting a status signal indicative of the liquid metal flow condition inside or from the filled container

Description

METHOD AND APPARATUS FOR DETECTING THE CONDITION OF THE LIQUID METAL FLOW IN AND FROM A FILLING CONTAINER BACKGROUND OF THE INVENTION FIELD OF THE INVENTION The present invention relates to a method and apparatus for detecting the condition of the flow of liquid metal in and from a filling container, and more particularly to a method and apparatus for detecting the presence of an undesirable condition in the liquid metal flow. inside and from a filling container.

DESCRIPTION OF THE PREVIOUS TECHNIQUE The liquid metal, and in particular the liquid steel, is drained from a drainage or filling vessel, usually a cauldron, into one or more molds usually through an intermediate or receiving vessel, usually a funnel. In this procedure, a controlled flow of liquid metal passes from the cauldron, usually through a nozzle and valve at the bottom of the cauldron, to a ceramic tube and then to a receiving vessel, usually a funnel. A funnel is a container of refractory lining with one or more outlets, through which metal flows into the mold (s). As the filling cauldron empties, the slag and oxidation products, which float on top of the liquid steel in the cauldron, can enter the filling flow and transfer to the funnel. Usually, as the filling cavity is emptied, the surface of the liquid steel in the cauldron is observed visually and when the slag is seen to enter the receiving vessel, the valve in the cauldron is closed in order to reduce the contamination of the metal in the cauldron or mold with the slag and oxidation products. Alternatively, an electromagnetic coil can be used to assist in the detection of the presence of slag and non-metal in the fill flow and automatically signal the closing of the valve. Typically, this coil surrounds the filler cavity nozzle and senses variations in the electromagnetic field produced by the excitation of the coil relative to changes in the non-metallic content of the flow. It is well known that the flow of a filling cauldron induces the vibration of the same cauldron, the ceramic tube, which is attached to the cauldron, and the funnel, in particular, the vibration of the tube can be substantial. Attempts have been made to perceive this vibration manually. The prior art does not address the following problems: Visual Scum Detection The visibility of the slag entry inside the funnel is poor. Therefore, the capacity of the cauldron filling operator to see the slag is difficult, and the consistency of the cauldron flow closure is poor. The premature closure of the cauldron results in a loss of metal production, and the late closing of the cauldron results in the contamination of slag from the liquid metal in the funnel. Since multiple cauldrons are emptied into a funnel, slag development occurs and the problem of visibility is combined. A significant problem associated with the visual detection of slag is that the slag is not seen until it is already present in the funnel.

Electromagnetic Detection of Slag The sensing coil is located in the cauldron and, therefore, is highly susceptible to physical and term damage. The cauldron must be specifically adapted to accept the coil and as each cauldron arrives in the filling position, a cable connection must be made. The penetration of steel into the nozzle block can damage the coil or impede its operation. A filling cauldron must be removed from the operational cycle to replace a damaged or non-functioning coil. In this situation, the slag is not detected until it is present within the nozzle block of the fill cauldron and is ready to flow into the funnel.

Perception of Vibration of the Prior Art Manual perception of vibration is inconsistent and dependent on the operator. The threshold of the human being to perceive and discriminate the change in vibration is limited. As with the two previous methodologies, the slag is detected when it is present and flows through the tube attached to the cauldron.

COMPENDIUM OF THE INVENTION Accordingly, it is an object of the present invention to provide a method and apparatus, which uses vibra-acoustic signals to characterize and detect changes in the condition or behavior of the liquid metal flow, which presage the onset, or are characteristic of, the entrance of slag into the metal flow that passes from a drainage vessel. Another object is to provide a logic that allows means for flow discrimination and provides alarms indicating a deviation between a desired flow condition of a cauldron and unwanted flow conditions such as eddy formation, irregularity in flow rate, surface crushing, flow obturation, slag entry, and gas suction. It has been found that the above objects and other objects of the present invention are obtained in an apparatus for detecting the condition of liquid metal flow in or from a filling container, including a sensor for detecting vibration caused by a liquid metal flow. in or from the filling container and to output a sensor signal corresponding to a quantity of vibration detected by the sensor. A signal processor receives the sensor signal and compares the sensor signal with a reference signal and outputs a comparison signal. A logic unit receives the comparison signal and outputs a status signal indicative of the liquid metal flow condition in and from the filling vessel. A method for detecting a condition of a liquid metal flow in or from a filling container includes detecting an amount of vibration caused by the liquid metal flowing in and from the filling container. The detected amount of the vibration is converted to a sensor signal. The sensor signal is compared with a reference signal to output a comparison signal. A state signal is produced in response to the comparison signal, the status signal is indicative of a condition of the liquid metal flow within and from the filling vessel. Other aspects and advantages of the present invention will be apparent from the following description of the invention, which refers to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS For the purpose of illustrating the invention, a mode that is presently preferred is shown in the drawings; however, it is understood that the invention is not limited to the precise arrangements and instruments, shown.

Figure 1 is a typical arrangement in a steel making process showing the apparatus used to empty the liquid metal from a filling vessel or cauldron and finally into a mold. Figure 2 is a schematic diagram showing the elements used to detect an undesired flow condition in the liquid metal flow in and from the filling container or cauldron of Figure 1. Figure 3 is a graph representing a change in the signal strength at 40 Hz associated with an undesirable flow condition, such as the entry or transportation of slag. Figure 4 is a graph representing a change in signal strength at 50 Hz associated with an undesirable flow condition, such as the entry or transportation of slag.

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method and apparatus that discriminates the vibration associated with the flow of liquid metal, the slag or the flow of liquid metal contaminated with slag, and alterations in the behavior or condition of the flow, which presage the beginning of the entrance of slag into the metal flow passing from a filling or draining vessel, such as a cauldron. Referring now to the drawings, in which similar numbers indicate similar elements, a general arrangement of an apparatus typically used in a method for making constant casting steel is shown in Figure 1., where you want to detect the presence of contaminants in the liquid metal flow. In the continuous casting process, a draining vessel or cauldron 10 is filled with a liquid metal 12 and transfers the liquid metal 12 to one or more molds 14 through an intermediate or receiving vessel called a funnel 16. A controlled flow of the liquid metal 12 passes from the cauld 10 to the funnel 16 through a nozzle 18 located at the bottom of the cauld 10, and through a tube 20, preferably a ceramic tube. The nozzle 18 includes a valve 22 for controlling the flow rate of the liquid metal 12 out of the cauldron 10. The funnel 16 is equipped with one or more outlets 24, through which the liquid metal 12 flows towards a corresponding number say, one or more) of molds 14. When the cauldron 10 contains very little or no liquid metal 12, the cauldron 10 is replaced by another cauldron, not shown, filled with liquid metal 12 to ensure that the liquid metal 12 flows continuously towards the mold (s) 14. When the second cauldron, not shown, similarly runs out of liquid metal 12, it is also replaced with another cauldron, not shown, filled with liquid metal 12. This is a continuous process. A problem with the continuous casting of steel arises when the cauldron 10 is emptied. The presence of an impurity such as slag 26, which typically forms a layer on liquid metal 12, arises when it enters liquid metal 12 passing through valve 22 and tube 20. This contaminates liquid metal 12 which it passes into the funnel 16, and finally into the mold (s) 14. This is undesirable. When the kettle 10 is filled with the liquid metal 12, the presence of the slag 26 or other impurities floating on the top of the liquid metal 12 is far enough away from the nozzle 18 so that the slag 26 does not enter with the metal liquid 12 passing from the cauldron 10 to the funnel 16. The flow of liquid metal 12 from the cauldron 10 and through the nozzle 18 at that point is, therefore, an uncontaminated flow, or a substantially uncontaminated flow. The method and apparatus of the present invention utilizes vibration perception, analysis and alarm logic for the flow discrimination of the liquid metal 12 of the cauldron 10. This procedure provides alarms, which indicate a deviation between the desired condition of the liquid metal flow from cauldron 10, and unwanted flow conditions such as slag entry as cauldron 10 is emptied. It is within the scope of the present invention to detect other undesirable flow conditions such as vortex formation, flow rate irregularity, surface crush, flow obturation, and gaseous aspiration. Alarms are provided to indicate undesirable changes in the flow condition. Alarms associated with vortex formation, and / or irregularity of flow velocity, and / or surface crush may presage the start of slag flow or entry. The present invention can be used to assist the operator in deciding to stop the flow by closing the valve 22. Alternatively, an alarm logic can provide a signal to automatically or manually initiate the closing of the valve 22. Referring now to Figure 2 , a general arrangement of the method and apparatus of the present invention is shown at 28 for detecting the condition of liquid metal flow 12 inside and / or from the cauldron 10. A vibration perception device or sensor 30, such as a microphone , or in a preferred embodiment a delta shear stress accelerometer, is used to sense the vibration induced by the flow of liquid metal 12 inside or from the cauldron 10. The vibration sensor 30 outputs an analog electrical signal or signal of sensor 12, which allows to measure the vibration. The accelerometer can be obtained from any of the known suppliers, including Brüel &; Kjaer of Denmark or of Hewlett Packard. The sensor 30 may be coupled to the vibration source by any of the known methods, for example, bolts or magnetic means. The sensor 30 is not required to be in direct contact with the liquid metal flow channel. For example, the sensor 30 can be directly coupled to the ceramic tube 20, if the sensor 30 is able to withstand temperatures associated with the ceramic tube 20. In a preferred embodiment, the sensor 30 is coupled to a lifting device, not shown, which moves the tube 20 in and out of alignment with the cauldron 10 and the funnel 16. The sensor 30 can be coupled to the tube 20 or to the lifting device, not shown, by means of a magnet. Alternatively, the sensor 30 may be placed directly to the ceramic tube 20 or to the lifting device, or to any other elastic solid, i.e., a solid capable of passing the vibration, which is in direct contact with the source of vibration. Once the vibration sensor 30 generates the sensor signal 32, it passes to a differential load amplifier 34 consisting of two high-gain, low noise operational amplifiers. A filter system around the input amplifier provides a damping in a response below 10 Hz. This eliminates the influence of low frequency noise from the sensor signal 32, which may be caused by the effects of temperature fluctuation. Preferably, the load amplifier 34 has a balanced low impedance output for pulling, i.e., passing signals through long wires. The charge amplifier can be obtained from any of the known suppliers, including the suppliers described above. The sensor signal 32 of the load amplifier 34 is passed to an analysis electronics or to an analysis unit 36 such as a signal processor for continuous analysis. The analysis or signal processing of the sensor signal 32 allows the discrimination of the signal 32. In a preferred embodiment, the analysis / signal processing is performed by a real-time frequency analyzer, which allows a rapid frequency analysis and the simultaneous spectral comparison, so that no signal data is lost. The signal processor may be obtained from any of the known suppliers, including the suppliers described above. Within the signal processor 36, the analog sensor signal 32 is first converted to a digital data signal by an analog-to-digital converter 38. In a preferred embodiment, the analog sensor signal 32 is converted to a digital data signal. by means of a signal analyzer using a nine-pole low-pass elliptical filter, which provides at least 84 dB of high-frequency signal attenuation. During and / or after the conversion, the digital data signal is processed using broadband filters 40 of constant percentage to divide the digital data signal into those portions associated with several frequency bands on the frequency scale of interest, by example, a frequency scale of 0.1 Hz to 20 kHz. This is referred to as a frequency analysis. The output of the frequency analysis is a data spectrum. The data spectrum is rapidly and continuously generated and compared to a calibration spectrum by means of a comparator 42. A calibration spectrum is generated for the desired flow condition (ie, free of eddies, slag, contaminants, etc.). . In comparator 42, the data spectrum is continuously on the calibration spectrum to determine when the intensity level of the data spectrum is outside a preprogrammed standard deviation, which is an integral part of the calibration spectrum. This is termed as a spectral comparison. The differences between the data spectrum and the calibration spectrum are calculated by the comparator 42 and produce spectral comparison data 44. The magnitudes of the spectral comparison data 44 are processed within a logical unit or a central processing unit ( "CPU") 46. In a preferred mode, the CPU 46 is a compatible IBM Compaq 486DX266 programmable computer, although other types and brands of CPUs that are within the scope of the present invention may be used, eg, Apple computers, RISC-based computers, workstations of Silicon Graphics, or similar. The CPU 46 generates a status signal 54 such as a pollution warning or change of condition signal in the flow and / or a closing signal of the cauldron, using logic based on the magnitudes of the spectral comparison data. This logic is based on the magnitude of the variation of the spectral comparison data 44 outside the confidence intervals of about 88% to 95%. In other words, the difference between the spectrum of data generated in real time and the calibration spectrum is calculated at frequent intervals (typically of the order of 500-1000 milliseconds) and logic assumes that the variation in this difference greater than a range of specific confidence placed around the calibration spectrum, is indicative of alterations in the flow condition, which can be characterized at different frequencies or spectral bands such as swirl, surface crushing, flow velocity, gaseous aspiration and slag input. The CPU 46 contains adjustable limits of the confidence interval and the magnitude of the differences between the adjusted confidence interval and the measured spectral data, to allow the adjustment of the sensitivity of the system to the previously defined flow condition changes. In addition, the logic allows the independent adjustment of the confidence interval associated with the individual frequencies or bands in order to allow adjustment of the sensitivity of the system to the different flow conditions that are experienced. The logic may simply involve the presence of a specific magnitude of the spectral comparison data 44 or may include a time history, and / or cauldron weight factors to improve the alarm sensitivity. The CPU 46 may have as input, signals from the process equipment for making steel such as the weight of the kettle determined from a kettle weight sensor 48, the level of the funnel or the determined weight of a level / weight sensor 50. of funnel, and / or the position of gate or valve of the cauld determined from the position sensor 52 of gate or valve of the cauldron. This allows the sensitivity of the logic to be varied as a function of the weight of the cauldron, and for the system to automatically control and close the gate or valve 22 of the cauldron. For example, as the cauldron 10 empties more and more, the probability of slag entry or an undesirable similar flow condition is increased, and, consequently, the sensitivity of the logic increases. The output of the CPU 46 is used to pass the status signal 54 to the operator signal panel 56, which indicates the status, eg, an alarm state, of the flow condition. The status signal 54 can be used to assist the operator of the signal panel 56 to decide whether to stop the flow of liquid metal by closing the valve 22. The status signals 54 can also be used to signal or automatically initiate the closure of the boiler valve 22. During use, the method and apparatus 28 of the present invention is first used to determine the normal deviation of the calibration spectrum for the flow of liquid metal in the particular process for making steel, for which it seeks detection of the liquid metal flow condition, such as the procedure shown and described in Figure 1. The normal deviation is then preprogrammed to the signal processor 36 and applied to the calibration spectrum. This is subsequently compared to the data spectrum. The data spectrum is continuously generated and updated to measure most of the updated liquid metal flow conditions. To determine the normal deviation that will be applied to the calibration spectrum, the flow of the liquid metal is analyzed. The vibration sensor 30 is attached to the steelmaking equipment used in the process for making steel to generate the sensor signal 44 in the manner described above. The signal 32 is then divided into different frequency bands by means of the filter 40 of the signal processor 36 at a scale of between about 0.1 to 20 kHz. The intensity of each of these frequency bands is measured and analyzed to determine which are the frequency bands that respond to the flow conditions that are to be detected, for example, the slag input. For conditions or flow cases such as slag input that generate a low final noise, the lower end of the frequency band, between approximately 10 Hz-1 kHz, only needs to be considered. Referring now to Figures 3, 4, two representative frequency bands are shown that have been generated to determine an illustrative normal deviation, which can be applied to a calibration spectrum. Figure 3 illustrates a 40 Hz signal, and Figure 4 illustrates a 50 Hz signal. The y-axis unit is dB and the x-axis unit is time. Figures 3, 4 were generated during a typical operation of the method and apparatus of the present invention 28, in conjunction with a given process for making steel, such as that shown and described in Figure 1. For the particular process for manufacturing steel from which Figures 3 and 4 were generated, the sharp drop observed at the end of the graphs is associated with the slag 26 passing through the ceramic tube 20 between the cauldron 10 and the funnel 16. The intensity of the Signal levels for both figures are absolutely constant until the change in the liquid metal flow 12 associated with the slag input. The small tines in Figures 3, 4 just before the great fall, may be associated with the swirling formation within the tube 20 just before the entrance or transportation of the slag. From Figures 3, 4, the frequency bands are determined to deviate in intensity of approximately +5, -15 dB. In other words, in Figures 3, 4, the intensity level of an uncontaminated flow of liquid metal at 40 and 50 Hz, respectively, is about 70 dB. Once the liquid metal flow is contaminated with the slag or when some other undesirable flow conditions occur, the intensity of the frequency bands generated by the liquid metal flow deviates +5, -15 dB. It should be noted by those skilled in the art that although only the analysis of two frequency bands as described in Figures 3, 4 is described, all frequency bands, within the range of .1 Hz-20kHz, should be analyzed. , although the lower end of the frequency band, approximately between 10 Hz-1kHz, only needs to be analyzed in the steelmaking process, which typically generates low frequency noise, that is, less than 1 kHz. Once the normal deviation to be applied to the calibration spectrum is determined, it is programmed to the signal processor 36. The normal deviation can be increased or reduced to decrease or increase, respectively, the sensitivity of the method and apparatus. The apparatus 28 of the present invention is now ready to detect the condition of the flow of metal liquid in the desired process for making steel, such as that described and shown in Figure 1. In a preferred embodiment, the method and apparatus 28 of the present invention is turned on when the amount of liquid steel 12 in the cauldron 10 approaches approximately 20 tons, to generally first the calibration spectrum to which the standard deviation applies. . A typical cauldron used in the steelmaking process holds anywhere from 100-300 tons of liquid metal. At 15-20 tons of liquid metal 12 remaining in the cauldron 10, the flow of liquid metal 12 out of the cauld 10 is a substantially uncontaminated flow, although it should be observed by those skilled in the art that the apparatus 28 of the present invention can be ignited at any level of the liquid metal 12, wherein the liquid metal flow is not contaminated or in an undesirable condition to generate the calibration spectrum.

To generate the calibration spectrum, the vibration of the flow of the liquid metal 12 passing through the ceramic tube 20 is analyzed by the signal processor 36 in the manner described above, for about 3 to 15 seconds. This information is then fed to the comparator 42. After the calibration spectrum is generated and fed to the comparator 42, the vibration of the liquid metal flow 12 passing through the ceramic tube 20 is analyzed by the signal processor 36 in the form described above, to generate the data spectrum. The comparator 42 of the signal processor 36 then compares the intensity of the vibration frequency of the liquid metal flow 12 traveling through the ceramic tube 20, i.e., the data spectrum, with the normal deviation applied to the calibration spectrum. to generate the spectral comparison data 44, which is fed to the CPU 46. The data spectrum is constantly updated and compared with the normal deviation applied to the calibration spectrum to reflect the latest flow conditions. The CPU 46 analyzes the spectral comparison data 44 to determine if the data spectrum is within the normal deviation applied to the calibration spectrum. If the spectral comparison data 44 determines that the data spectrum deviates outside the normal deviation applied to the calibration spectrum, the CPU 46 is programmed to determine the degree of deviation, for example, that the data spectrum of the calibration spectrum, and for how long. If the deviation is determined by the CPU 46 as acceptable, a positive state signal is passed to the operator signal panel 56. If the deviation is determined by the CPU 46 as unacceptable, the status signal 54 declares an alarm state . This information is passed to the operator signal panel 56 to inform the operator of the signal panel 56 of the alarm status, so that he can initiate the closing of the valve 22 if appropriate. Alternatively, the status signal 54 in the alarm state can initiate the automatic valve shut-off of the valve 22. Once the valve 22 is closed, the cauldron 10 is replaced by another cauldron filled with liquid metal, not shown. When the amount of liquid metal contained in the second cauldron approaches 15-20 tons, the method and apparatus 28 is again turned on and operated in the manner described above. A new calibration spectrum is generated and applied to the preprogrammed standard deviation and compared to a data spectrum in the manner described above. This is a continuous procedure. Although the present invention has been described in relation to its particular embodiments, many other variations and modifications and other uses will be apparent to those skilled in the art. It is preferred, therefore, that the present invention be limited not by the specific description herein, but only by the appended claims.

Claims (1)

1 - . 1 - An apparatus for detecting the condition of the flow of liquid metal in and from a filling container, comprising: a sensor for detecting vibration caused by a flow of liquid metal in and from a filling vessel and for outputting a sensor signal corresponding to a quantity of vibration detected by the sensor; a signal processor for receiving the sensor signal and for comparing the sensor signal with a reference signal, and outputting a comparison signal; and a logic unit for receiving the comparison signal and for outputting a status signal indicative of the liquid metal flow condition in or from the filling vessel. 2. The apparatus for detecting the liquid metal flow condition of claim 1, wherein the sensor signal is received by a differential load amplifier to eliminate the low frequency noise before it is received by the signal processor. . 3. The apparatus for detecting the liquid metal flow condition of claim 1, wherein the filling container includes a valve for regulating the flow of liquid metal out of the filling vessel. 4. The apparatus for detecting the liquid metal flow condition of claim 3, wherein the valve is in fluid communication with a passage for flowing the liquid metal from the filling vessel to a receiving vessel. 5. The apparatus for detecting the liquid metal flow condition of claim 4, wherein the passage is in operative coupling with a lifting device capable of moving the passage in and out of the fluid communication with the receiving vessel. . 6 - The apparatus for detecting the liquid metal flow condition of claim 5, wherein it includes a control panel for receiving the status signal. The apparatus for detecting the liquid metal flow condition of claim 6, wherein the control panel includes means for closing the valve to stop the flow of liquid metal out of the fill container in response to a status signal. indicative of an undesired flow condition of the liquid metal in or from the filling vessel. 8. The apparatus for detecting the liquid metal flow condition of claim 6, wherein the control panel includes means for displaying the liquid metal flow condition in and from the fill container in response to the status signal. . 9. The apparatus for detecting the liquid metal flow condition of claim 1, wherein the sensor is a microphone. 10. The apparatus for detecting the liquid metal flow condition of claim 1, wherein the sensor is an accelerometer. 11. The apparatus for detecting the condition of the liquid metal flow of claim 1, wherein the signal processor divides the sensor signal into frequency bands. 12. The apparatus for detecting the liquid metal flow condition of claim 1, wherein the signal processor divides the sensor signal into a frequency scale of about 0.1 Hz to 20 kHz. 13. The apparatus for detecting the liquid metal flow condition of claim 11, wherein the signal processor outputs the frequency bands as a data spectrum. 14. The apparatus for detecting the liquid metal flow condition of claim 1, wherein the logic unit is a central processing unit. 15. The apparatus for detecting the liquid metal flow condition of claim 1, wherein wherein the status signal is indicative of an undesirable flow condition of the liquid metal in or from the filling vessel. 16. The apparatus for detecting the liquid metal flow condition of claim 1, wherein the status signal is indicative of a desirable flow condition of the liquid metal in or from the filling vessel. 17. The apparatus for detecting the liquid metal flow condition of claim 15, wherein the undesirable flow condition is the slag input. 18. The apparatus for detecting the condition of the liquid metal flow of claim 17, wherein the undesirable flow condition includes vortex formation, flow rate irregularity, surface crush, flow obturation or gaseous aspiration. 19. A method for detecting the condition of the liquid metal flow in or from a filling vessel, the method comprising the steps of: detecting an amount of vibration caused by the liquid metal flowing in or from the filling vessel; convert the detected amount of vibration to a sensor signal; comparing the sensor signal to a reference signal to output a comparison signal; and outputting a state signal in response to the comparison signal, the status signal being indicative of a liquid metal flow condition in or from the filling vessel. 20. The method for detecting the liquid metal flow condition of claim 19, wherein the sensor signal is amplified to eliminate the low frequency noise before the sensor signal is compared with the reference signal.