Method for automatically analyzing and scanning inclusions in pure steel

Abstract

The invention creates a brand new analysis method for automatically scanning the inclusions in the clean steel, and by optimizing the detection setting of the scanning method and the setting of the processing method, the time for scanning the large-area inclusions on the sample is shortened, and the efficiency for scanning the large-area inclusions on the sample is greatly improved. According to the method, the time required for scanning a single sample is reduced to be less than 4 hours by 30 hours or even more, the inclusion scanning efficiency is controlled to be 70% or above, and the scanning efficiency is improved by 40% or above compared with that of a traditional method. According to the method, the requirement of scientific research personnel for automatically analyzing and scanning batch inclusions of a large batch of samples in a plurality of steelmaking procedures is well met, the working efficiency is greatly improved, and the method has important value for defect analysis and quality improvement of materials.

Description

Technical Field

[0001] The invention relates to a method for automatic analysis and scanning of inclusions in pure steel, in particular to a method for scanning and analyzing using an INCA-Feature automatic inclusion analysis system. Background Art

[0002] In 1962, Kiessling [1] first proposed the term clean steel in a special report drafted for the British Iron and Steel Institute, which generally refers to steel with low content of impurity elements such as O, S, P, H, N, and Pb, As, Cu, and Zn. Thirty years after the first oil crisis in 1973, the mass production of commercial clean steel gradually matured [2,3]. Finally, clean steel has turned from a scientific research term to mass production, and a clean steel production platform has been formed in some famous steel mills in Europe, the United States, and Japan. Since then, the production of clean steel has expanded to a wide range of fields from ultra-low carbon steel to high carbon steel, such as deep-drawn thin plates, IC lead pins, tire radials, ball bearing steel, rails, hot working molds, etc.

[0003] In order to meet market requirements, the production and quality improvement of clean steel is one of the focuses of steel material research, and the control and analysis technology of inclusions in steel is the focus of clean steel quality assessment and improvement. In order to implement full-process monitoring of clean steel quality, there is a need for automatic batch inclusion analysis and scanning of large batches of samples from multiple steelmaking processes. Only by correctly and comprehensively analyzing the inclusions in each process of clean steel production can we more clearly identify the source and generation mechanism of inclusions, and thus clarify the corresponding methods to reduce and control inclusions in steel.

[0004] After the clean steel is polished, there are still inevitably contamination or pits on the surface that interfere with the automatic scanning process, making the scanning time far longer than expected, greatly affecting the efficiency of automatic scanning of inclusions in clean steel. The present invention has created a new analysis method for automatic scanning of inclusions in clean steel, which greatly improves the scanning efficiency. The present invention can control the single scanning time to within 4 hours, control the inclusion scanning efficiency to more than 70%, and improve the scanning efficiency by more than 40% compared with the traditional method.

[0005] The control and analysis technology of inclusions in steel is the focus of steel cleanliness research. In order to implement full-process monitoring of the quality of pure steel, there is a need for automatic analysis and scanning of inclusions in large batches of samples from multiple steelmaking processes. The present invention successfully establishes a new method for automatic analysis and scanning of inclusions in pure steel, which well meets the needs of scientific researchers for automatic analysis and scanning of inclusions in large batches of samples from multiple steelmaking processes, greatly improves work efficiency, and lays a solid foundation for scientific researchers to improve the quality of materials. Summary of the invention

[0006] The purpose of the present invention is to provide a method for automatically analyzing and scanning inclusions in pure steel in view of the above problems.

[0007] The object of the present invention is achieved as follows: a method for automatic analysis and scanning of inclusions in pure steel, comprising the following steps: step 1: according to the analysis requirements, select the required analysis area, select a rectangle as the analysis area for the steel sample, WD, that is, the working distance = 8.5-9.0 mm, and refocus each target point by adjusting the Z-axis distance; step 2: processing method setting, the total feature quantity limit number of a single field of view in the analysis area is 18-22, the total time spent in the steel sample is limited to 230-250 minutes, and the spectrum and chart of each feature are saved, and the rest are default values, (0); step 3: detection setting, field setting Set to 1024×766, BSE signal, the first pass image is 1.5~2.0ms, the second pass image is 9~11ms, the feature value setting, the magnification is 950~1050, the features smaller than the area to be ignored are 0.45~0.50mmECD, and the protection area is 16.11~16.31μm; Step 4: Spectrum setting, the number of passes is 1; the first pass live time is 1.9-2.1s; the processing time is 4.5~5.5s, the spectrum range is 0-20keV, the number of channels is 900-1100, and all features are collected; Step 5: Start running.

[0008] Step 5: After the operation starts, the feature point data of all fields of view in the area are automatically scanned and analyzed.

[0009] The beneficial effects of the present invention are as follows: the present invention creates a new analysis method for automatically scanning inclusions in clean steel, which reduces the time for scanning large-area inclusions on samples by optimizing the detection settings and processing method settings of the scanning method, and greatly improves the efficiency of scanning large-area inclusions on samples. This method reduces the time required for scanning a single sample from 30 hours or even more to less than 4 hours, controls the efficiency of inclusion scanning to more than 70%, and improves the efficiency of scanning by more than 40% compared with the traditional method.

[0010] This invention meets the needs of scientific researchers for automatic batch inclusion analysis and scanning of large batches of samples from multiple steelmaking processes, greatly improves work efficiency, and is of great value to material defect analysis and quality improvement. DETAILED DESCRIPTION

[0011] The present invention has created a new INCA-Feature self-inclusion automatic scanning method, which can control the single scanning time within 4 hours and the inclusion scanning efficiency to more than 70%. The specific experimental method is as follows: 1. According to the analysis requirements, select the required analysis area. The steel sample is generally rectangular, WD=8.5~9.0mm, and each point is refocused by adjusting the Z-axis distance.

[0012] 2. Processing method settings: the total feature quantity for a single field of view is limited to 18 to 22; the total time spent in the sample is limited to 230 to 250 minutes; the spectrum and chart of each feature are saved, and the rest are default values.

[0013] 3. Detection settings, field of view set to 1024×766, BSE signal, first pass image 1.5~2.0ms, second pass image 9~11ms; feature value setting, magnification 950~1050, ignore features smaller than the area is 0.45~0.50mmECD, protection area is 16.11~16.31μm.

[0014] 4. Spectrum settings, number of passes is 1; first pass live time is 2s; processing time is 4.5~5.5; spectrum range is 0-20keV; number of channels is 1k; collect all features.

[0015] 5. Start running and automatically scan and analyze the feature point data of multiple fields of view in the scanned area.

[0016] The specific implementation of the method is described in detail below in conjunction with examples, but the specific implementation of the present invention is not limited to the following examples. Embodiment 1

[0017] 1. Sample preparation: Take a clean steel electroslag process sample with a size of (15-18) mm × (18-22) mm × (13-17) mm. Polish the surface of the test sample, perform rough polishing, medium polishing, and fine polishing to achieve a mirror effect, clean and dry it to meet the analysis and detection requirements of inclusions. Then mark the analysis area, which is approximately a (0.9-1.1) mm × (0.9-1.1) mm rectangle.

[0018] 2. Analysis and testing: The polished sample is loaded into the EVO18 scanning electron microscope of the German Zeiss company, and vacuum is evacuated and high voltage is applied (the voltage parameter is set to 15KV) in sequence, and then the brightness, contrast, focus, astigmatism and other parameters are adjusted to make the sample surface image clear, which is convenient for subsequent inclusion scanning.

[0019] 3. Observe in the marked analysis area, WD = 8.5 ~ 9.0mm, and refocus each point by adjusting the Z-axis distance.

[0020] 4. Processing method settings: the total feature quantity for a single field of view is limited to 18 to 22; the total time spent in the sample is limited to 230 to 250 minutes; the spectrum and chart of each feature are saved, and the rest are default values.

[0021] 5. Detection settings, field of view is set to 1024×766, BSE signal, the first pass image is 1.5~2.0ms, the second pass image is 9~11ms; feature value setting, magnification is 950~1050, features smaller than the area to be ignored are 0.45~0.50mmECD, and the protection area is 16.11~16.31μm.

[0022] 6. Spectrum settings, number of passes is 1; first pass live time is 2s; processing time is 4.5~5.5; spectrum range is 0-20keV; number of channels is 1k; collect all features.

[0023] 7. Start running, the final acquisition time is 239.6 minutes, the number of fields of view is 418, the number of features is 4436, the number of inclusions after screening is 3441, and the scanning efficiency is 77.57%.

[0024] 8. The traditional method takes 994.2 minutes to collect data, with 1735 fields of view, 18458 features, 4527 inclusions after screening, and a scanning efficiency of 24.53%. Compared with the traditional method, this method increases the scanning efficiency by 53.04%. Embodiment 2

[0025] 1. Sample preparation: Take a clean steel mold ingot sample with a size of (15-18) mm × (18-22) mm × (13-17) mm. Polish the surface of the sample to be tested, perform rough polishing, medium polishing, and fine polishing to achieve a mirror effect, clean and dry it to meet the analysis and detection requirements of inclusions. Then mark the analysis area, which is approximately a (0.9-1.1) mm × (0.9-1.1) mm rectangle.

[0026] 2. Analysis and testing: The polished sample is loaded into the EVO18 scanning electron microscope of the German Zeiss company, and vacuum is evacuated and high voltage is applied (the voltage parameter is set to 15KV) in sequence, and then the brightness, contrast, focus, astigmatism and other parameters are adjusted to make the sample surface image clear, which is convenient for subsequent inclusion scanning.

[0027] 3. Observe in the marked analysis area, WD = 8.5 ~ 9.0mm, and refocus each point by adjusting the Z-axis distance.

[0028] 4. Processing method settings: the total feature quantity for a single field of view is limited to 18 to 22; the total time spent in the sample is limited to 230 to 250 minutes; the spectrum and chart of each feature are saved, and the rest are default values.

[0029] 5. Detection settings, field of view is set to 1024×766, BSE signal, the first pass image is 1.5~2.0ms, the second pass image is 9~11ms; feature value setting, magnification is 950~1050, features smaller than the area to be ignored are 0.45~0.50mmECD, and the protection area is 16.11~16.31μm.

[0030] 6. Spectrum settings, number of passes is 1; first pass live time is 2s; processing time is 4.5~5.5; spectrum range is 0-20keV; number of channels is 1k; collect all features.

[0031] 7. Start running, the final acquisition time is 213.7 minutes, the number of fields of view is 807, the number of features is 3199, the number of inclusions after screening is 3199, and the scanning efficiency is 73.30%.

[0032] 8. The traditional method takes 834.8 minutes to collect data, with 3158 fields of view, 13063 features, 3992 inclusions after screening, and a scanning efficiency of 30.56%. Compared with the traditional method, this method increases the scanning efficiency by 42.74%.

[0033] The above description is only a specific embodiment of the present invention, but the structural features of the protection scope of the present invention are not limited thereto. Any changes or modifications made by any technician in the field of the present invention are all covered by the patent scope of the present invention.

Claims

1. A method for automatic analysis and scanning of inclusions in pure steel, characterized in that: The following steps are involved: Step 1: Select the required analysis area according to the analysis requirements. For steel samples, select a rectangle as the analysis area. WD, or working distance, is 8.5-9.0 mm. Each target point is refocused by adjusting the Z-axis distance. Step 2: Set the processing method, limit the total number of features in a single field of view in the analysis area to 18-22, limit the total time spent in the steel sample to 230-250 minutes, save the spectrum and chart of each feature, and the rest are the default values ​​(0); Step 3: Detection settings, field of view set to 1024×766, BSE signal, first pass image 1.5~2.0ms, second pass image 9~11ms, feature value setting, magnification 950~1050, ignore features smaller than area 0.45~0.50mmECD, protection area 16.11~16.31μm; Step 4: Spectrum settings, number of passes is 1; first pass live time is 1.9-2.1s; processing time is 4.5-5.5s, spectrum range is 0-20keV, number of channels is 900-1100, and all features are collected; Step 5: Start running.

2. The method for automatic analysis and scanning of inclusions in pure steel according to claim 1, characterized in that: Step 5: After the operation starts, the feature point data of all fields in the area are automatically scanned and analyzed.