JPH04203954A - Emission spectral analysis method of acid-soluble aluminum within steel - Google Patents

Abstract

PURPOSE:To enable analysis to be improved by eliminating a tip of a projecting part which is generated after coagulation and then analyzing an emission signal obtained by discharge. CONSTITUTION:An electron beam is emitted onto a surface layer of a sample 1 for using one part of it and then a tip of a projecting part 3 of a fused and coagulated part 2 which is generated by fusion and coagulation operation is removed after this fused part is coagulated. Then, emission is made by discharge, the obtained emission signal is analyzed, and a quantitative value is obtained. Then, an acid-soluble Al within steel is condensed and eliminated previously so that an emission signal when the acid-soluble Al within steel exists can be distinguished from an emission signal when it does not exist clearly, thus enabling analysis accuracy to be improved and obtaining an analysis value in a short time even if no chemical analysis is made.

Description

[Detailed Description of the Invention] [Industrial Application Field] Related to the rapid analysis of components in steel, acid-soluble aluminum, which is particularly susceptible to the formation of inclusions and difficult to obtain sufficient accuracy with instrumental analysis, Regarding emission spectroscopy.

[Prior Art] During steel manufacturing, aluminum is added as a deoxidizing agent, and most of it becomes an oxide and migrates into the slab, or a portion remains in the steel. Some of this remaining aluminum exists in the form of inclusions, while others exist as a solid solution in the steel. These types of aluminum are precisely measured by chemical analysis based on stoichiometry, and the existence form of aluminum is determined by whether it dissolves in acid or not. That is, the former is mostly acid-insoluble aluminum (which is an oxide and does not dissolve in acids).
The latter is acid-soluble aluminum (hereinafter referred to as 5olAA) that dissolves in acid. Moreover, the combination of both is total aluminum (hereinafter referred to as T and Af).

5olAl is an important component that affects the mechanical properties of steel, and rapid feedback of its analytical values is required for steelmaking process control. For this reason, wet chemical analysis takes too much time, so research is continuing to obtain highly accurate information using optical emission spectroscopy, which is a short-time analysis method.

In emission spectroscopic analysis of components in steel, elements in a sample of solidified molten steel are excited by discharge energy, and the type and quantitative information of each element is obtained from the wavelength and intensity of the light emitted when returning to the ground state. And In5olAI! and 5olA I
! When performing a quantitative analysis of each type of light emission, based on the difference in light emission behavior between the two, these can be analyzed by analyzing the frequency distribution of the intensity of the light emission signal obtained for each discharge pulse (hereinafter referred to as intensity distribution). The amount is being determined. There are various analysis methods, including the distortion method, time difference method, and energy alternating conversion method.

However, conventionally, in emission spectrometry, In5oI in the sample
Al! If the content of is high, 5olA t! However, recently, efforts have been made to improve the accuracy of analysis. For example, CAMP-I S I JVo
13 (1990)-601, a method (hereinafter referred to as oxygen analysis method) of improving the accuracy of 5olA 1 analysis by obtaining and analyzing an oxygen emission signal simultaneously with an Af emission signal is reported. That is, by applying an extrapolation method to the relationship between the luminescence signal intensity of oxygen and the luminescence signal intensity of Af, AI! without oxygen luminescence is obtained. The amount of 5olAl is determined by estimating the luminescence signal intensity of .

[Problem to be solved by the invention] However, the wavelength of the measurement line that can be used to measure oxygen is short and absorption is very easy. There still remained the problem that the values were not sufficiently accurate.

This invention was made to solve this problem.
The purpose is to improve the accuracy of 5olA 1 analysis by emission spectrometry by sufficiently removing 1nsolA 1 before emitting light.

[Means for Solving the Problem] The means for achieving this purpose is to irradiate the surface of a sample for component analysis of steel with an electron beam to melt a part of the surface layer of the sample, and to melt this melted part. This is an emission spectroscopic analysis method for soluble aluminum in steel, in which a quantitative value is obtained by removing the tips of the protrusions produced by the melt-solidification operation after solidification, and then emitting light by electric discharge and analyzing the resulting luminescent signal.

[Operation] In the emission spectroscopic analysis of components in steel, discharge energy is generally given as a pulse of several hundred meters. Therefore, a light emission signal is obtained in response to this discharge, but there is no In5O within the discharge point.
Luminescence signal when lA j7 exists (hereinafter referred to as Sl) and emission signal when it does not exist (hereinafter referred to as S)
It is known that the strength is different between S and S, and that the strength of S is greater. This is the difference in light emission behavior. This difference is used to analyze the luminescence signal, but this difference can only be known by comparison. In other words, it cannot be determined that a light emission signal with an intensity above a certain level is 31, and a light emission signal with an intensity lower than that is S.

[n5olA I! The intensity distribution of the Af emission signal of the sample containing almost no ion is the distribution of S, and [n5olA I
! The intensity distribution of the Af emission signal of the sample containing SI and S.

Although the distribution is a mixture of the two, there are differences in the pattern or median value between the former and the latter. From this, Sl
It is estimated that there is a clearer difference between the intensity distribution of 83 and the intensity distribution of 83. However, the problem here is that the amount of In5olA is large and all discharge points are
This is a case where 1nsolA1 exists. In this case, the light emission signal is only S, and there is no S to be compared, making it impossible to distinguish between the two from the signal strength. Samples with a large amount of In5olAf approach the above state, and this is the cause of the decrease in the accuracy of 5olAf analysis.

Figure 6 shows 5olA! for four sample types analyzed using the strain method. ! This is a comparison of the emission spectroscopic analysis values and the chemical analysis values. The vertical axis is the emission spectroscopic analysis value, and the horizontal axis is the chemical analysis value. The circle indicates 0.01-0.02wt% In5olA I! in converter molten steel with a large amount of In5olA I! Contains. All ○ marks are above the straight line, indicating that there is a correct error.

In order to solve this problem, it is necessary to obtain In5olA in the sample so that S8 can be obtained with a frequency comparable to Sl.
I! All you have to do is reduce it.

For the target steel, the upper limit of 5olA I content is 0.08%.
, the upper limit of the amount of In5olA is about 002%, [
n5olA! ! The above-mentioned accuracy problem occurs with samples where the amount exceeds 0.005%.

Therefore, it is sufficient to reduce the amount of In5olAl to 0005% or less.

The surface to be analyzed is irradiated with a Nirecroton beam to melt part of the surface layer of the sample, and after solidification, the protrusions created by melting and solidification are removed to obtain 1nsolA of the target area.
It is possible to reduce I. When a part of the sample is melted,
In5olA 1. It has a high melting point, does not melt, and has only half the specific gravity of steel, so it floats to the surface. Furthermore, when the material is melted by irradiating it with an electron beam, it melts in a circular shape around the beam and solidifies in a convex shape at the center. This situation is shown in FIG. In the figure, 1 is a sample, 2 is a melted solidified part, and 3 is a convex part.

The floating In5olAβ is concentrated and collected at the tip of the convex portion 3. When the tip of this convex part 3 is removed, most of the 1nsolA that was present in the melted solidified part 2 is removed! ! or will be removed. The removal may be carried out by polishing with alumina-free sandpaper, which is commonly used for sample surface preparation, or by pre-discharge, since discharge is likely to occur at the top. Since 5olA1 must not change during the melting and solidification process, melting must be performed in a non-oxidizing atmosphere, and electron beam irradiation in a vacuum is also suitable for this purpose. Furthermore, if excessively high-density melting energy is input, there is a risk that the sample may change due to evaporation, bumping, etc., but electron beam melting is also suitable in this respect.

In this way, the surface layer part that has been melted and solidified and the tips of the convex parts have been removed is the subject of analysis, but [n5olA f is 0.02
Even in samples containing approximately 0.0%, the amount decreased to 0.002 or less in the analysis target area. In this way, since the amount of 1nsolAf is drastically reduced, analysis can be performed under the most convenient conditions, and the precision and accuracy of analysis is greatly improved. Naturally, conventional research has been carried out for this analysis [n5olA I! and 5
A method that utilizes the difference in the luminescence behavior of olA f, for example,
A distortion method, a time difference method, an energy alternating conversion method, etc. can be used.

[Example] Samples were taken from molten steel extracted from a converter, ladle refined molten steel, RH degassed molten steel, and continuous cast molten steel, solidified into a cylindrical shape of 30 strokes in diameter, cut and polished in cross section, and used as analysis samples. .

The sample types containing the most 1nsolAl are 0.01 to 0.02 wt% of molten steel extracted from a converter, and 0.0% of molten steel refined in a ladle.
05 to 0.010 wt%, and 0.05 to 0.010 wt% for RH degassed molten copper.
002-0.005wt%, O, OO for continuous casting molten steel
1~0. 002 wt% each.

Using these samples, 5olA I! We analyzed the reproducibility and accuracy of the analytical values. As examples of the present invention, two cases were investigated: one in which the tip of the convex part was removed mechanically with a sander, and one in which it was removed by preliminary discharge. At the same time, as a comparative example, a sample in which the analysis part was not melted and solidified was also compared.

(Example 1) An electron beam was irradiated perpendicularly to the surface of the analysis sample. The acceleration voltage of electrons was 10 kV, and the analysis area was melted by irradiation at 30 W+A for 5 seconds, followed by irradiation at 80 mA for 5 seconds. The degree of vacuum in the irradiation chamber was set to 10-'Tar, and after irradiation, the material was allowed to cool and solidify. This process results in a diameter of 10 mm.
An analysis part was obtained which was approximately circular in shape and had a protrusion in the center with a sharp tip and a height of approximately Im+. Approximately half the height of this convex part,
It was removed by light sanding with a 5i-C belt sander.

For light emission, an emission spectrometer of DC spark light emission DCLVS-bW according to JIS-G-1253 was used. That is,
The secondary voltage was 400 V, the inductance was 10 μH, the capacitance was 10llF, the resistance was 2Ω, and the discharge frequency was 400±. The analytical value was the average value for two analytical parts in one sample, as in normal emission spectroscopic analysis.

As the intensity signal of AI, the intensity ratio with Fe serving as an internal standard was used, and the signal was analyzed using the time difference method. That is,
Let T be the median value of the intensity distribution of the light emission signals in the discharge orders 4001 to 5000, and let ΔT be the difference between T and the median of the intensity distribution of the light emission signals in the discharge orders 501 to 1500, 5olA I! The amount was calculated using the following formula.

(5olAf:l=, T-, ΔT-(11 However, +kl and 2 were determined by regression calculation using a large number of standard samples.

(Example 2) The melting and solidification of the sample analysis section was performed as in Example 1. The same is true for . The tip of the convex portion could be removed by preliminary discharge for 15 seconds immediately after the start of discharge.

As in Example 1, the ratio of the Af emission signal intensity to the Fe intensity was used, but an energy alternating conversion method was used for analysis. This method alternates between high and low energy level discharges and adds this information to the emission intensity signal for analysis. The light emission conditions at the high energy level were the same as in Example 1, and at the low energy level, the charge amount for discharging was half that of the high energy level. The amount of 5olAf was determined using the following formula.

(5olA!ri=k (X, -α (βX Ac
-X, c) 1... (2) where, : Average value of the high intensity side 5% of the cumulative frequency distribution of the low energy level. k, α, and β were determined by regression calculation using a large number of standard samples.

In the comparative examples, Comparative Example 1 was analyzed using the time difference method, and Comparative Example 2 was analyzed using the energy alternating conversion method.

These results are compared with the chemical analysis values and shown in Figures 2 to 5, respectively.
The results are shown in the figure, and a comparison of their analysis accuracy is shown in Table 1. Figure 2 is Example 1, Figure 3 is Example 2, and Figure 4 is Comparative Example 1.
, FIG. 5 shows the emission spectroscopic analysis values of each of Comparative Example 2 on the vertical axis, and the chemical analysis values on the vertical axis. In the figure, the ○ mark represents molten steel from a converter, the mouth represents ladle-refined molten steel, the △ mark represents RH degassing refining molten steel, and the * mark represents continuous casting molten steel. The correct errors in the molten steel tapped from the converter shown in FIGS. 4 and 5 are not seen in FIGS. 1 and 2.

Table 1 σM: Reproducibility accuracy when analyzing two points of the same sample at different measurement points.

σ: Accuracy of the average value of two analyzes of the same sample.

It can be seen that both the reproducibility precision and accuracy are significantly improved in the example.

Note that the conventional oxygen analysis method requires the use of a special analyzer, making it impossible to compare using the same sample; however, the reported accuracy is taken as a conventional example.
I compared the accuracy of the analysis value with that of the example with the same 5olA f amount range, that is, 5olA It amount 0.
This is a comparison in a small range of about 0.02 wt%. In the conventional example, the reproducibility is 0.0007wt% and the accuracy is 0.00.
It is 18wt%. On the other hand, in this range, in Example 1, the reproducibility is 0.0005wt% and the accuracy is 0.000%.
9we%, and the reproducibility accuracy in Example 2 is 0.0004w
t%, and the accuracy is 0.0009wt%.

It is clear that both reproducibility and accuracy are improved compared to the conventional example.

[Effects of the Invention] As described above, according to the present invention, 1nsolA 1 is concentrated and removed in advance in the emission spectroscopic analysis of 5OIA f. For this reason, it becomes possible to more clearly distinguish between the luminescent signal when [n5olA I exists within the discharge point and the luminescent signal when it does not exist, and as a result,
The analytical accuracy was greatly improved even in the range where the amount of lAf was large. In steelmaking process management, which conventionally had to rely on time-consuming chemical analysis, the present invention has a great effect in that accurate analysis values can be obtained in a short time.

[Brief explanation of the drawing]

FIG. 1 is a cross-sectional view of a sample after melting and solidification for explaining the present invention in detail, FIG. 2 is a diagram showing the relationship between emission spectroscopic analysis values and chemical analysis, which is an embodiment of the present invention, and FIG. The figure shows the relationship between the emission spectroscopic analysis value and chemical analysis as another example. Figure 4 shows the relationship between the emission spectroscopic analysis value and chemical analysis as a comparative example. FIG. 6 is a diagram showing the relationship between the emission spectrometry value and chemical analysis as a comparative example, and FIG. 6 is a diagram showing the relationship between the emission spectrometry value and chemical analysis as a conventional example for explaining the operation. l...sample, 2...melt solidification part, 3...convex part.

Claims (1)

[Claims] The surface of the sample for steel component analysis is irradiated with an electron beam to melt a part of the surface layer of the sample, and after solidification, the tips of the convexities produced by melting and solidification are removed, and then the light is emitted by electric discharge. An emission spectroscopic analysis method for acid-soluble aluminum in steel, characterized in that analysis is performed by analyzing signals.