Detection error control of portable X-ray fluorescence spectrometer

Portable X-ray fluorescence spectrum analyzer has the advantages of simple operation, rapid analysis and no damage to samples. It is widely used in metal production, processing and industrial pipeline installation engineering quality control. However, when testing some alloy materials, the results of elements such as copper, manganese, titanium, vanadium and aluminum show a deviation of more than - 15%. When the content of vanadium and titanium in the material is low, there is an undetectable phenomenon, which seriously interferes with the judgment of the tester. In order to solve these problems, after long-term detection accumulation and test, two methods to control the detection error are proposed: ① optimize the sample and detection environment. ② Taking the test results of several characteristic elements as the judgment index, the material type and brand are directly determined.

IMG_ 20211021_ 133647.jpg

1. Carbon steel and alloy steel testing error control (1) carbon steel testing error control. The commonly used carbon steels are mainly recommended grades of GB / t699, GB / T700 and GB / T701, among which Q235 series, 20#, 35#, 45# etc. are the most common. The composition is composed of iron base, five conventional elements such as carbon, silicon, manganese, phosphorus and sulfur and a small amount of impurities. There are two types of spectrometer for detecting carbon steel. One is an all element detector, which can quickly detect the components of metal and non-metal elements in the sample, and the other is only to detect the components of metal elements in the sample. When tested by the all element spectrometer, the elements in the sample are excited by high-energy rays. After the non-metallic elements are excited, the characteristic spectral line is very unstable and easy to be absorbed by the air. In addition, the amount is small, the spectral line intensity is weak, and the fluorescence absorbed by the detector is small, so the result is low. In order to reduce the error, the surface of the sample is deeply cleaned and shows metallic luster. At the same time, 99.99% argon is equipped to protect the excitation elements and fully absorb them. In order to make the analysis accurate, there are strict requirements for the test sample and environment, it is necessary to carry argon, which makes the operation inconvenient and the use of all element spectrometer limited. The metal element detector is more widely used. When testing carbon steel, the results show the composition of iron, manganese and impurity alloy, and the brand can not be identified. However, compared with the standard value, the error is large. Select 20 #, Q235B and 45 # steel, use kemais k-1688 portable X-ray fluorescence alloy spectrum analyzer, test at 25 ℃ room temperature for 25s, and compare with the test results in the laboratory. The results are shown in Table 1.

1.png

In order to reduce the detection error of carbon steel, it is controlled by improving the sample surface roughness and optimizing the detection environment. The sample is polished to 5cm with 400 mesh metallographic sandpaper × 5cm detection window, metallic luster, surface roughness Ra ≥ 0.25 μ m. Optimized detection conditions: no strong electromagnetic interference, temperature 25 ℃, detection for 25s. Thirdly, the spectral detection of 20#, Q235B and 45# samples after treatment showed that the manganese content was 0.40, 0.38 and 0.50 respectively, and the error decreased significantly. The standard stipulates that the manganese content of carbon steel is between 0.25% ~ 1.00%. For the alloy spectrum detector, the manganese content detected in this range, no matter what method is used to control, only carbon steel is displayed, and the specific brand can not be identified. However, when carbon steel and alloy steel are mixed, it is a good method to classify with spectrometer. The low manganese content is mainly caused by three reasons: ① the detection environment and sample preparation are not standardized. ② The instrument itself has large ray energy. After the sample is excited, the characteristic spectral line is easily absorbed by the outside world. ③ Manganese and iron in steel interfere with each other after being excited. It can be seen from the periodic table that the atomic numbers of manganese and iron are 25 and 26 respectively, and the mutation values of characteristic spectral lines are 25MN respectively γ k＝8.07，Jk＝0.876，26Fe γ K = 7.87, JK = 0.873, the absorption difference of manganese and iron JK lines is 0.003. After the sample is excited, the iron JK line and manganese JK line interfere with each other. In addition, the iron content of carbon steel is ≥ 98%, while the manganese content is ≤ 1.00%. After being excited, the manganese fluorescence is less than that of iron, and the absorbed manganese spectrum is less, so the manganese component value is low.

(2) Alloy steel detection error control. According to the alloy content, alloy steel is divided into high and low alloy steel. In low alloy steel, except for iron base, the alloy composition is ≤ 2.00%. Typical examples are 40Cr, Q345, 15CrMo and 12Cr1MoVG. A small amount of alloy elements are added on the basis of carbon steel to greatly improve the material properties. High alloy steels form a variety of systems and grades according to the types and amounts of alloys added, such as 18-8 austenitic stainless steel, which is widely used in production and life. When the alloy spectrometer detects it, it can detect the components with an alloy content of more than 0.05%. After comparison with the instrument, the material composition and brand will be displayed. When the content of copper, manganese, titanium, vanadium and aluminum in the sample is low, there may be no display value, or the display value is seriously low, which can not reach the lower limit of alloy composition specified in the standard. In particular, when there is welding within 5m of the detection point, or the sample surface temperature exceeds 55 ℃, it is difficult to detect the V (vanadium) component when the V (vanadium) content is ≤ 2.00%. Taking boiler tube 12Cr1MoVG as an example, no matter how to treat the sample surface and increase the detection time, the display of V (vanadium) content is ≤ 0.06%, while the display of Cr and Mo is normal.

The first is to control the detection error of low alloy steel by improving the detection environment: clean the sample, the temperature is between 5 ~ 35 ℃, stop the welding operation within 5m of the detection point, detect for 25s, repeat the detection, and the data has good reproducibility and accuracy. Second, determine the material brand according to the Cr and Mo composition display, that is, when the Cr and Mo contents are stable for many times, more than 0.05% v (vanadium) display (vanadium standard value 0.15% ~ 0.25%), although the deviation is large, it can also be determined that the V (vanadium) composition of the material meets the requirements. However, it is better to retest the surplus materials purchased in the same period and of the same specification and model to corroborate the test conclusion.

The alloy detector cannot detect non-metallic elements, so it cannot identify the materials with the example number determined by the carbon content, such as low alloy steel 30crmoa and 35CrMo. The content of Cr (0.80% ~ 1.10%) and Mo (0.15% ~ 0.25%) is the same, and the carbon content is slightly different. High alloy austenitic stainless steels such as s30403 / s30408 and s31603 / s31608 have the same alloy composition. The grades are distinguished only by the carbon content, and the specific grades cannot be accurately determined only by the spectral alloy composition.

This kind of error control is carried out from several aspects: ① conduct full quantitative detection on the samples that are allowed to be damaged, and find the corresponding brand according to the quantitative analysis results. ② In the petroleum and petrochemical industry, according to the petrochemical standard sh3501-2013 and Sinopec Jian [2015] No. 495 document, there is no mandatory requirement to detect the non-metallic content. As long as the alloy composition is qualified, the material is judged to be qualified. The basis for this regulation in the petroleum and petrochemical industries: ① establish a dynamic qualified supplier system and strictly monitor the supply quality. ② From the development of smelting technology, smelting on-line analysis is the standard configuration of large steel plants, and the composition control is good. In addition, in recent years, the steel production capacity has been reduced, small smelters and strip mills have been shut down, and the low-quality steel has been greatly reduced. Based on this, only metal elements can be detected, and there is no need to tangle whether there is non-metallic composition data. When the composition of chromium, nickel and molybdenum in alloy steel is qualified, the material is directly determined to be qualified.

With the rise of manufacturing costs and the improvement of smelting level, manufacturers generally control the components such as manganese, chromium, nickel, molybdenum and titanium at the lower limit of the standard allowable value. Taking the acetylene furnace tube of a chemical plant as an example, it is designed as American Standard 321 (the national standard corresponds to 08cr19ni11ti). During one overhaul, the spectra of three furnace tubes are tested, and the results are shown in Table 2.

2.png

It can be seen from the table that the content of chromium and nickel is low and does not meet the requirements of GB / t14976-2013. However, combined with the data error processing in GB / t222-2012, when the content of chromium and nickel is ≥ 10.00%, the allowable deviation is ± 0.3%, and the content of chromium and nickel meets the requirements. The content of titanium is 6 × C% ～ 0.80%. When the carbon content cannot be determined, as long as the chromium and nickel components are qualified and there is a small amount of titanium in the tested components, the material is directly determined to be qualified.

2. Influence of detection conditions on data error

(1) The influence of testing time on the results is reflected in two aspects: ① the attenuation degree of X-ray tube, the cumulative use of equipment is less than 1000 h, the sensitivity is high, and accurate results can be obtained within 6 s. After using for 2500 h, the attenuation of the ray tube increases. When the detection time is less than 15s, the data error increases obviously. The detection time is adjusted to more than 25s, and the accuracy of the results is improved. ② The sample itself requires the detection time. When detecting materials with low alloy content and light metal elements, the detection time is set at more than 25s, which can obtain more accurate results. With the long service life of the equipment, the testing time should be increased.

(2) Detect the impact of the environment. The data error is large in three environments: ① strong magnetic environment. During detection, there are welding operations within 5m of the measuring point, the component display is abnormal, or even disordered code, because the welding high-frequency arc interferes with the fluorescence absorption, and the results will be abnormal when detecting strong magnetic equipment. For such errors, the welding operation should be stopped and the strong magnetic equipment should be closed. ② Temperature. According to the equipment manual, the applicable temperature is - 10 ~ 50 ℃. In fact, the ideal detection temperature is 5 ~ 35 ℃. When the ambient temperature is 38 ~ 45 ℃, the detection times should be reduced. Every 50 times of detection, shut down for 15min. When the sample surface is higher than 60 ℃, it cannot be detected. ③ Humidity. The optical and electrical components in the equipment have high requirements for dry humidity. They shall be tested in rain and fog, and protective measures shall be taken. When the equipment is not in use, it shall be placed in the desiccant environment and maintained at ordinary times to reduce the error caused by equipment instability.

IMG_ 20211025_ 174120.jpg

3. Conclusion

At present, there are many models of portable X-ray fluorescence spectrometer, but the basic principle is the same. For complex material varieties, inspectors can quickly and accurately detect the results by applying the methods mentioned in the paper on the basis of certain metal material knowledge. Of course, detection is a technology integrating professional knowledge and experience in materials science, physics and chemistry, which can not be fully understood in one article. It is hoped that through the description and analysis of this paper, the detection error can be reduced and the detection level can be improved.

one thousand six hundred and forty trillion and two hundred and forty-one billion eight hundred and seventy-two million nine hundred and fifty-one thousand six hundred and thirty-one jpg

one thousand six hundred and forty trillion and two hundred and forty-one billion eight hundred and seventy-eight million three hundred and fifty-two thousand four hundred and seventy-eight jpg