System and method for estimating both thickness and wear state of refractory materials in metallurgical furnaces - Patents.com

Abstract

The present invention primarily relates to a system for estimating both the thickness and state of wear of refractory material (1) of a metallurgical furnace (12), comprising processing means including at least a database of simulated frequency domain data, referred to as simulated spectra, representing simulated shock waves reflected at simulated refractory material of known condition and thickness, each simulated spectrum being correlated with both the known condition and thickness data of the considered simulated refractory material, the processing means being configured to record the reflected shock waves as time domain signals and convert them into frequency domain data, referred to as experimental spectra, and further configured to compare the experimental spectrum with at least a plurality of simulated spectra from the database, determine a simulated spectrum that best matches the experimental spectrum, and estimate the thickness and state of the refractory material (1) of the furnace (12) using the known condition and thickness data correlated with the best-matching simulated spectrum.

Description

[Technical field]

[0001] The present invention is in the field of metallurgy, and more particularly of refractory materials in metallurgical furnaces.

[0002] The present invention relates to a system and method for estimating both the thickness and the wear state of refractory materials in a metallurgical furnace, such as a blast furnace. [Background technology]

[0003] A blast furnace typically comprises an outer layer of metal sheets, preferentially steel sheets, and an inner layer and an intermediate ramming layer, both made of refractory material, which allows good thermal insulation of the outer layer and forms the refractory wall of the blast furnace.

[0004] However, since blast furnaces can be exposed to very high temperatures, exceeding 1500 degrees Celsius, and the molten metal is produced by chemical reactions involving more than 1000 gases and materials, the refractory material is subjected to high chemical and thermal stresses, resulting in its progressive deterioration and thickness reduction. Thus, reduction of the refractory material to a critical thickness can cause major malfunction of the blast furnace, often resulting in serious accidents.

[0005] A non-destructive testing technique is already known for estimating the thickness of refractory materials in blast furnaces, known as impact echo testing.

[0006] However, these known techniques are not accurate enough and do not estimate the thickness and wear state of multi-layered materials such as refractory materials in blast furnaces. For example, when exposed to high chemical and thermal stresses, weakened layers may appear in refractory materials, especially if they are made of carbon. The impact echo technique is not accurate enough to accurately determine the location and thickness of the weakened layers, often called brittle layers. Also, the impact echo technique cannot detect the ramming gaps of the ramming layers. Summary of the Invention [Problem to be solved by the invention]

[0007] The object of the present invention is therefore to remedy the shortcomings of the prior art by providing a system and a non-destructive method for estimating both the thickness and the wear state of refractory materials in a blast furnace. [Means for solving the problem]

[0008] To this end, a first subject of the invention consists of a system for estimating both the thickness and the wear state of a refractory material (1) of a metallurgical furnace, comprising at least: a shock wave generating means configured to generate at least one shock wave which propagates into the refractory material; a shock wave sensor configured to sense at least one reflected shock wave into the refractory material; a processing means comprising a database of simulated frequency domain data, referred to as simulated spectra, representing simulated shock waves reflected at simulated refractory materials of known condition and thickness, each simulated spectrum being correlated with both known condition and thickness data for the simulated refractory material considered, the processing means comprising: The reflected shock waves are recorded as time-domain signals and converted to frequency-domain data called experimental spectra. The method is configured to compare the experimental spectrum with a plurality of simulated spectra from a database, determine a simulated spectrum that best matches the experimental spectrum, and estimate a thickness and condition of the refractory material of the furnace using the known condition and thickness data that is correlated with the best matching simulated spectrum.

[0009] The system may also have any of the following features, either individually or in combination:

[0010] - a processing means, in sequence: i. determining resonant frequency peak locations in the frequency domain data; ii. filtering the simulated spectra from the database by resonant frequency peaks and selecting a corresponding reduced group of simulated spectra comprising said resonant frequency peaks; iii. determining a unique simulated spectrum from the reduced corresponding group whose resonant frequency peak is closest to a height of the peak of the experimental spectrum, said unique simulated spectrum being the best fitting simulated spectrum.

[0011] the shock wave generating means comprises an instrumented hammer configured to strike the metallurgical furnace wall to generate at least an acoustic shock wave that propagates and reflects into the refractory material;

[0012] the shock wave sensor comprises at least an accelerometer configured to measure the mechanical response of the refractory material caused by the reflection of the generated shock wave;

[0013] A second subject of the invention consists in a method for estimating both the thickness and the wear state of the refractory material of a metallurgical furnace in a system such as that described above, said method comprising at least: i. generating at least a shock wave that propagates into a refractory material; ii. sensing at least the reflected shock wave into the refractory material; iii. recording the reflected shock wave as a time domain signal; iv. converting the time domain signal into frequency domain data called an empirical spectrum; v. Comparing the experimental spectrum with at least a plurality of simulated spectra from a database, each simulated spectrum being correlated with both known state and thickness data of the considered refractory material; vi. determining a simulated spectrum that best matches the experimental spectrum; vii. Estimating the thickness and condition of the furnace refractory material using the known condition and thickness data correlated with the best fitting simulated spectrum.

[0014] The method may also take into account any of the following features, either individually or in combination: - Steps v and vi, in order: a. Determining the resonant frequency peak position in the experimental spectrum; b. filtering the simulated spectra from the database in which the resonant frequency peaks are detected and selecting a corresponding reduced group of simulated spectra comprising said resonant frequency peaks; c. Performed by determining the unique simulated spectrum from the reduced corresponding group whose resonant frequency peak is closest to the height of the peak in the experimental spectrum, said unique spectrum being the best fitting simulated spectrum.

[0015] - step vii is carried out by estimating at least the total thickness of the refractory material and the location and thickness of at least one layer in which the refractory material is weakened by an anomaly, said layer being called the brittle layer.

[0016] A number of shock waves are generated and a number of reflected shock waves are sensed and recorded as time domain signals and made into frequency domain signals.

[0017] - Determination of the resonant frequency peak position in the experimental spectrum: a. arithmetically averaging the experimental spectra and selecting a first set of representative peaks; b. geometrically averaging the experimental spectra and selecting a second set of representative peaks; c. selecting a final set of peaks that are both selected in the first set and in the second set, said final set of peaks being resonant frequency peaks.

[0018] A peak of the first set is selected if its width is greater than a threshold comprised between 10 and 20 Hertz.

[0019] - the simulated spectra from the database are filtered with the detected resonant frequency peaks using at least a numerical dispersion curve model to determine the propagation modes of the simulated shock waves, filter them with the detected resonant frequency peaks, and select a reduced corresponding group of simulated spectra.

[0020] - A unique simulated spectrum is determined from the reduced corresponding group using at least a numerical transient model.

[0021] - the corresponding group of spectra reduced is selected and then: deriving the difference between the simulated spectrum and the experimental spectrum; comparing the overall shape of the simulated spectrum with the experimental spectrum; determining the difference between the maximum height peak positions of the simulated spectrum and the experimental spectrum; Cross-correlating the simulated spectrum with the experimental spectrum. The spectrum is compared to the experimental spectrum by performing at least one of the following:

[0022] Other characteristics and advantages of the present invention are explained in the following description, by way of indication and in no way limiting, and with reference to the accompanying figures: [Brief description of the drawings]

[0023] [Figure 1] FIG. 1 is a cross-sectional view of a blast furnace. [Diagram 2] FIG. 1 is a cross-sectional view of a refractory material of a blast furnace. [Figure 3a] FIG. 2 is a cross-sectional view of a first variant of the refractory material of a blast furnace with a cooling system. [Figure 3b] FIG. 2 is a cross-sectional view of a second variant of the refractory material of a blast furnace with a cooling system. [Figure 3c] FIG. 11 is a cross-sectional view of a third variant of the refractory material of a blast furnace with a cooling system. [Figure 4] 1 is a graph having a first curve representing the arithmetically averaged experimental spectrum and a second curve representing the geometrically averaged experimental spectrum. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0024] The system and method of the present invention finds particular application in estimating both the thickness and the wear state of the refractory material 1 of a metallurgical furnace 12, and more particularly of a particular bottom portion of a blast furnace 12 containing the molten metal with the highest temperature.

[0025] The bottom of the blast furnace 12 includes an outer layer 11 made of metal, typically steel. To thermally insulate this outer layer 11, and as shown in Figures 1 and 2, the blast furnace 12 includes a refractory material 1 whose composition correlates with the temperature to which it is exposed.

[0026] At the bottom of the blast furnace 12, where temperatures exceed 1500 degrees Celsius, the refractory material 1 comprises a thick inner layer 3 and a thin intermediate ramming layer 2 located between the metallic outer layer and the inner layer 3. The ramming layer 2 is about 5 to 10 centimeters thick and preferably made of ceramic. The inner layer 3, made of carbon, is about 80 to 100 centimeters thick.

[0027] As depicted in Figures 3a to 3c, the blast furnace 12 further comprises cooling means 13, 14, 15 configured to cool the blast furnace 12 with a suitable coolant, e.g., water. These cooling means 13-15 are adjacent to the steel outer layer 11 of the blast furnace 12.

[0028] In a first variant shown in Fig. 3a, the cooling means 13 comprise a spraying device configured to spray atomized coolant regularly or continuously against the outer surface of the outer layer 11. Evaporation of the coolant on the outer layer 11 serves to cool the blast furnace 12.

[0029] In a second variant shown in figure 3b, the cooling means comprise cooling channels 14 arranged against the outer surface of the outer layer 11 (only one cooling channel 14 is shown in figure 3b). Coolant flows through these channels 14 and contributes to the cooling of the blast furnace 12 via the outer layer 11.

[0030] In a third variant shown in FIG. 3c, the cooling means comprise cooling staves 15 inserted between the outer layer 11 and the intermediate ramming layer 2 (only one cooling stave 15 is shown in FIG. 3c). The coolant flows through these staves 15 and contributes to the cooling of the blast furnace 12 via the ramming layer 2.

[0031] As indicated above, the objective of the system and method of the present invention is to estimate both the condition and thickness of the inner and intermediate layers 2, 3 in a non-destructive manner.

[0032] By condition of layers 2, 3 is meant the wear state of said layers 2, 3: the system and method of the invention is able to detect defaults and weak points in the refractory material 1. In particular, the invention is designed to detect ramming gaps in the intermediate layer 2 and to detect the location and thickness of any weakened layer 5 in the inner layer 3, said weakened layer being called brittle layer 5.

[0033] As shown in FIG. 2, the system and method of the present invention can also estimate the remaining “good” inner layer 4, i.e., the thickness L1 of the remaining carbon layer between the intermediate layer 2 and the brittle layer 5, and the “damaged” inner layer 6, i.e., the thickness L3 of the carbon layer beyond the brittle layer 5.

[0034] The system and method of the present invention applies to any of the blast furnaces 12 described above, regardless of whether the blast furnace 12 is equipped with cooling means 13-15 or not, and regardless of the variants of the cooling means 13-15.

[0035] The system of the present invention will now be described.

[0036] The system includes at least one electric hammer configured to strike the outer layer of the blast furnace 12. The purpose of each strike by the hammer is to excite the refractory material 1 of the blast furnace 12: with each strike a mechanical shock wave, typically an acoustic shock wave, is generated which propagates through the refractory material 1. Of course, multiple shock waves can be generated by the hammer, since each strike generates a shock wave.

[0037] The system also includes at least one shock wave sensor for detecting an excitation of the refractory material 1 of the blast furnace 12. In other words, said shock wave sensor is adapted to detect at least one reflected shock wave into the refractory material 1. As the shock waves cause successive compressions and expansions in the refractory material 1, each shock wave sensor is preferably an accelerometer installed on the outer surface of the outer layer of the blast furnace 12 so as to be able to detect the mechanical excitation of the refractory material 1.

[0038] The system includes a processing means electrically connected to the electric hammer and the accelerometer and is configured to record the reflected shock waves sensed by the accelerometer. The recorded signal represents the displacement of the refractory material 1 versus time. The recorded signal is therefore a time domain signal.

[0039] The processing means is also configured to convert the time domain signals, for example via a Fast Fourier Transform algorithm, into frequency domain data, hereafter referred to as "experimental spectrum". Thanks to this conversion, the experimental spectrum shows the propagation modes, meaning the respective resonant frequencies, which are correlated with the mechanical excitations of the refractory material 1. The propagation modes of the experimental spectrum are correlated with the structural properties of the refractory material 1, meaning its gap in the ramming layer 2, its total thickness L and, when applied to the brittle layer 5, its position and thickness L2.

[0040] As will be explained in a later section, according to the invention, the system uses the propagation mode to estimate both the thickness L and the wear state of the refractory material 1. More specifically, the processing means is configured to compare the experimental spectrum with a plurality of simulated spectra.

[0041] Thus, the processing means of the system includes a database of simulated spectra, each simulated spectrum being a Fast Fourier Transform of a corresponding simulated time domain signal corresponding to a known refractory material of the blast furnace 12. In other words, each simulated spectrum is correlated to a simulated refractory material whose ramming gap, total thickness and brittle layer location and thickness are known.

[0042] The simulated spectra are generated using known finite element analysis software in which a wide range of refractory material configurations have been studied. The database may contain more than 200.000 simulated spectra corresponding to the same number of refractory material configurations.

[0043] According to the invention, a method for estimating both the thickness L and the state of the refractory material 1 of a blast furnace 12 is carried out by the system described above. This method will now be described.

[0044] In a first step of the method, the processing means controls an electric hammer so that it strikes the surface of the outer layer of the blast furnace. At the same time, an accelerometer is activated to sense any excitation signal of the refractory material 1 of the blast furnace 12. The hammer thus generates a shock wave that propagates through the refractory material 1.

[0045] In a second step, the accelerometer senses the reflected shock wave into the refractory material 1, said reflected shock wave thus being recorded as a time domain signal in the memory of the processing means.

[0046] In a third step, the processing means transforms the time signal domain into frequency domain data to generate an empirical spectrum representative of the structure of the refractory material 1 .

[0047] In a fourth step, the processing means compares the experimental spectrum with at least a plurality of simulated spectra recorded in the database to determine the simulated spectrum that best matches the experimental spectrum. To do so, the processing means looks for representative resonant peaks in the experimental spectrum and compares them with the resonant frequency peaks of the plurality of simulated spectra, where such resonant peaks are representative propagation modes. The following description explains exactly how the best matching simulated spectrum can be selected in said fourth step.

[0048] Preferentially, to increase the signal-to-noise ratio, these three steps are repeated sequentially to obtain multiple experimental spectra, typically between 10 and 20, representative of the structure of the refractory material 1. To reduce noise, these multiple experimental spectra are then summed together using the formula

number

[0049] The processing means then select a first set of representative peaks 9, 10, said selection being based on the height and / or width of the peaks. Preferentially, the processing means select only peaks having a width greater than a threshold comprised between 10 and 20 Hertz.

[0050] The experimental spectrum can also be expressed by the formula

number

[0051] The arithmetic and geometric averaging are performed in parallel by the processing means. The objective is to obtain two independent averages, the arithmetic mean and the geometric mean, on the experimental data based on these two parallel processing steps. The processing means thus selects a final set of peaks that are both selected in the first set and in the second set. The peaks selected in the final set are representative resonant frequency peaks 9, which are directly correlated with the propagation modes representative of the refractory material 1 of the blast furnace 12.

[0052] Once the final set of resonant peaks 9 has been determined, the processing means filters at least a number of the simulated spectra with said selected resonant peaks 9. This filtering is accomplished by using known numerical dispersion curve models to determine a propagation mode of the simulated shock wave, then filter with the detected resonant frequency peaks 9, and finally select a reduced group of simulated spectra that includes said resonant frequency peaks 9.

[0053] The processing means implements at least the known numerical transient model, e.g., a finite element model, on the reduced group of simulated spectra to determine a unique simulated spectrum from the reduced group of simulated spectra. The unique simulated spectrum is the simulated spectrum that best matches the arithmetically averaged experimental spectrum 7. The best matching spectrum is the spectrum that includes resonant peaks whose respective heights are closest to the heights of the resonant peaks 9 of the arithmetically averaged experimental spectrum 7.

[0054] While applying the transient model, the processing means determines a similarity parameter by performing one or more of the following steps to determine the best matching simulated spectrum from the reduced group of simulated spectra. For each performed step a considered similarity parameter is generated: i. deriving the difference between the simulated spectra of the final set and the arithmetically averaged experimental spectrum 7; ii. comparing the overall shape of the final ensemble of simulated spectra with the arithmetically averaged experimental spectrum 7; iii. determining the difference between the decimated simulated spectrum and the decimated arithmetically averaged experimental spectrum 7; the decimated spectrum being, for example, a simplified spectrum obtained by applying a low pass filter to suppress selected features of the original spectrum; iv. determining the difference between the maximum height peak positions of the final set of simulated spectra and the arithmetically averaged experimental spectrum 7, respectively; and / or v. Cross-correlating the final set of simulated spectra with the arithmetically averaged experimental spectrum 7 by measuring the similarity between the final set of simulated spectra and the arithmetically averaged experimental spectrum 7. In other words, the more distant the spectra are from each other in the frequency domain, the smaller the correlation value.

[0055] To select the best-fitting simulated spectrum from the reduced group, the processing means generates a table showing the percentage of similarity between each simulated spectrum and the arithmetically averaged experimental spectrum 7. More specifically, the percentage of similarity is determined taking into account all simulated parameters obtained by performing one or more of the above steps. Furthermore, all similarity parameters are considered with equal weight. In other words, if five similarity parameters are determined, each one contributes 20% in selecting the best-fitting simulated spectrum. In this approach, the more similar the spectra are in shape, maximum position, etc., the higher the percentage of similarity between these spectra. The simulated spectrum with the highest percentage in the table is finally selected by the processing means to be the best-fitting simulated spectrum.

[0056] Finally, in a fifth step of the method, the processing means selects the simulated refractory material in the database that correlates with the best matching simulated spectrum. The processing means then generate a table showing all the characteristics of said simulated refractory material that are highly comparable to the tested refractory material 1. This provides a highly accurate non-destructive estimation of the characteristics of the tested refractory material, in particular the presence or absence of ramming gaps in the intermediate layer 2, the thickness L of the total carbon layer 3, the thickness L2 and location of the brittle layer 5, if applicable, and the thickness L1 of the "certain" carbon layer 4 between the brittle layer 5 and the ramming intermediate layer 2. Also, with the method and system of the present invention, the estimation is performed accurately whatever the configuration of the metallurgical furnace, in particular the location and type of cooling means as shown in Figures 3a to 3c.

Claims

1. A system for estimating both the thickness and the wear state of a refractory material (1) of a metallurgical furnace (12), comprising at least a shock wave generating means configured to generate at least one shock wave which propagates into the refractory material (1); a shock wave sensor configured to sense at least one reflected shock wave into the refractory material (1); processing means comprising a database of simulated frequency domain data, referred to as simulated spectra, representing simulated shock waves reflected at simulated refractory materials of known thickness and wear state, each simulated spectrum being correlated with both the known thickness and wear state data of the simulated refractory material under consideration; The system comprises: The reflected shock waves are recorded as time-domain signals and converted to frequency-domain data called experimental spectra. The system is configured to compare the experimental spectrum with a plurality of simulated spectra from a database, determine a simulated spectrum that best matches the experimental spectrum, and estimate a thickness and wear state of the refractory material (1) of the furnace (12) using the known thickness and wear state data that is correlated with the best matching simulated spectrum.

2. The processing means sequentially i. determining the resonant frequency peak (9) location in the frequency domain data; ii. filtering the simulated spectra from the database at the resonant frequency peak (9) and selecting a corresponding reduced group of simulated spectra comprising said resonant frequency peak (9); iii. The system of claim 1, configured to determine a unique simulated spectrum from the reduced corresponding group whose resonant frequency peak is closest to a height of the peak (9) of the experimental spectrum, the unique simulated spectrum being the best fit simulated spectrum.

3. 3. The system according to claim 1 or 2, wherein the shock wave generating means comprises an instrumented hammer configured to strike the metallurgical furnace wall (11) to generate at least an acoustic shock wave that propagates and reflects into the refractory material (1).

4. 4. The system according to any one of claims 1 to 3, wherein the shock wave sensor comprises at least an accelerometer configured to measure the mechanical response of the refractory material (1) caused by the reflection of the generated shock wave.

5. A method for estimating both the thickness and the wear state of a refractory material (1) of a metallurgical furnace (12) in a system according to any one of claims 1 to 4, said method comprising at least i. generating at least a shock wave that propagates into a refractory material (1); ii. sensing at least a reflected shock wave into the refractory material (1); iii. Recording the reflected shockwave as a time domain signal; iv. converting said time domain signals into frequency domain data called empirical spectra; v. Comparing the experimental spectrum with at least a plurality of simulated spectra from a database, each simulated spectrum being correlated with both known thickness and wear state data of the considered refractory material; vi. determining the simulated spectrum that best matches the experimental spectrum; vii. Estimating the thickness and wear state of the refractory material of the furnace (12) using the known thickness and wear state data correlated with the best fitting simulated spectrum.

6. Steps v and vi, in turn: a. determining the location of the resonant frequency peak (9) in the experimental spectrum; b. filtering the simulated spectra from the database according to the detected resonant frequency peaks (9) and selecting a corresponding reduced group of simulated spectra comprising said resonant frequency peaks (9); c) The method of claim 5, performed by determining the unique simulated spectrum from the reduced corresponding group whose resonant frequency peak is closest to the height of the peak (9) of the experimental spectrum, said unique spectrum being the best fitting simulated spectrum.

7. Step vii determines the total thickness (L) of the refractory material (1) and the location and thickness (L) of at least one layer where the refractory is weakened by an anomaly. ２ 7. The method according to claim 5 or 6, which is carried out by at least estimating:

8. 8. A method according to any one of claims 5 to 7, wherein a plurality of shock waves are generated and a plurality of reflected shock waves are sensed and recorded as time domain signals and converted to frequency domain signals.

9. The determination of the position of the resonant frequency peak (9) in the experimental spectrum can be done by a. arithmetically averaging the experimental spectra and selecting a first set of representative peaks (9, 10); b. geometrically averaging the experimental spectra and selecting a second set of representative peaks (9'); and c. selecting a final set of peaks that are both selected in the first set and in the second set, said final set of peaks being resonant frequency peaks.

10. 10. The method of claim 9, wherein a peak (9,10) of the first set is selected if the width of the peak is greater than a threshold comprised between 10 and 20 Hertz.

11. 11. The method according to claim 9 or 10, wherein the simulated spectra from the database are filtered with the detected resonant frequency peaks (9) using at least a numerical dispersion curve model to determine the propagation modes of the simulated shock waves, filter them with the detected resonant frequency peaks (9) and select a reduced corresponding group of simulated spectra.

12. 12. The method of any one of claims 9 to 11, wherein the characteristic simulated spectrum is determined from the reduced corresponding group using at least a numerical transient model.

13. The corresponding group of spectra of the reduced size is selected and then deriving the difference between the simulated spectrum and the experimental spectrum; comparing the overall shape of the simulated spectrum with the experimental spectrum; determining the difference between the maximum height peak positions of the simulated spectrum and the experimental spectrum; Cross-correlating the simulated spectrum with the experimental spectrum. The method of claim 11 or 12, wherein the experimental spectrum is compared by performing at least one of the following:

Images