Analysis of Steel Micro Inclusions Using OES

Optical emission spectrometry (OES) performs elemental analysis of solid iron and steel samples rapidly and cost effectively, with ease. The Thermo Scientific™ ARL iSpark™ Series metals analyzer (Figure 1) is an OES spectrometer platform delivering superior performance with unprecedented precision and accuracy to analyze iron and steel from trace to alloying element levels.

Figure 1. The Thermo Scientific™ ARL iSpark™ Series metals analyzer.

With Spark-DAT (Spark Data Acquisition and Treatment) methods, the ARL iSpark can perform ultra-fast inclusion analysis in addition to spectrochemical analysis. These methods are gaining traction in the steel industry, especially for their ability to deliver data about the inclusions during the steel elaboration process. This article discusses both the Standard Inclusion Analysis option and Advanced Inclusion Analysis option available with the Thermo Scientific™ ARL iSpark™ Series metals analyzer.

Principles of the Spark-DAT Analysis

With the Spark-DAT methods, the ARL iSpark enables using different treatment principles compared to OES concentration analyses. The light intensity values for all the single sparks are presented to a special mathematical treatment rather than integrating and translating into concentration. The intensity of a single spark signal relies on the sample composition at the position struck by the corresponding single spark. An intensity peak will be the outcome if the concentration of an element in the ablated sample material is considerably greater than the concentration of its soluble form in the matrix.

Advantages of ARL iSpark with Optional Spark-DAT Methods

The ARL iSpark metals analyzer with standard Spark-DAT methods yields the following benefits:

• The ability to perform inclusion analysis as well as elemental concentration analysis significantly reduces investment costs for inclusion analysis
• The ability to provide data on inclusions shortly after sample taking offers key insights for in-process control of the metal elaboration
• Conveniently detects randomly distributed exogeneous inclusions and rapidly analyzes very large surface areas
• Requires very less time for inclusion analysis and associated sample preparation
• Performs inclusion analysis simultaneously with the analysis of elemental concentrations for over 30 samples per hour
• Can perform inclusion analysis on all samples analyzed by OES
• Ensures minimal cost and time for operations as sample preparation, maintenance, and service remain the same as for the standard ARL iSpark spectrometer.

The following are the additional benefits offered by advanced Spark-DAT applications:

• Enables quantitative oxygen analysis in killed steels at levels even lower than 30ppm, minimizing the requirement for costly combustion analysis
• Determines quantitative inclusion size and size distribution, thereby providing comprehensive inclusion data, especially during the steel elaboration.

Practical Aspects and Analysis Time

The Spark-DAT methods comprise software and specialized algorithms and are offered with PMTs only. The single spark intensities obtained with Single Spark Acquisition (SSA) are utilized for inclusion analysis and conventional elemental concentration analysis, thereby performing the two types of analyses simultaneously. Since the Spark-DAT raw dataset is very large and complex, the values pertaining to the data of interest are calculated via fast dedicated algorithms. The resulting values can then be processed like traditional OES results by the analytical software.

For a single measurement, the Spark-DAT analysis alone usually takes 7 seconds that includes 2 seconds of Ar flush. This mode is appropriate only for rapid counting and confirmation of inclusion types, for acquiring raw data for off-line interpretation.

Basic Applications of Spark-DAT Methods on Standard Inclusion Analysis Option

The Standard Inclusion Analysis option provides great benefits for:

• Inclusion control for quality assurance
• Process control through on-line monitoring of inclusions
• Screening hundreds of samples for inclusions in a day
• Evaluation of number and type of inclusions
• Replacement of long or costly analysis techniques.

Soluble/Insoluble Contents of Elements

The soluble or insoluble part concentration of elements like B, Ti, Ca, Al is long-established and widely used indicators of the steel fabrication process. Hence, the algorithm Insoluble is one among the generally used Spark-DAT applications. It depends on estimating the soluble or insoluble part of an element. The algorithm Insoluble is used to calculate the ratio R_insol of the sum of intensity signals attributable to the matrix insoluble part to the sum of all the intensity signals.

Calibration samples are not required in the Insoluble Spark- DAT method for soluble/insoluble concentrations, thus enabling the method to be used for any partly insoluble element such as Ca, Ti, and B, or for cases where no certified reference material is (CRM) available. The accuracy of the Insoluble Spark- DAT method obtained on CKD low alloy steel standards having certified Al_sol concentrations is shown in the following table:

Estimation of Number and Type of Inclusions

Counting intensity peaks on the channel of a specific element using the algorithm peaks is the simplest application of the Spark-DAT methods. Counting intensity peaks allows for estimating the number of inclusions having this element. As delineated in Figure 2, it is possible to easily determine clean and dirty steel samples through comparison of the number of peaks counted on the channels of the inclusion elements.

Figure 2. Comparison of the number of peaks counted on the channels of the inclusion elements helps identifying clean and dirty steel samples.

The algorithm Composition enables counting coincidental peaks appearing on the channels of different elements simultaneously during the same single spark. It is possible to count coincidences of up to four channels with the algorithm Composition, thereby enabling the chemical formulation of complicated inclusions or inclusion clusters (Figure 3). Moreover, the option to check for non-coincidences as well as coincidences eliminates ambiguities on the inclusion type.

Figure 3. Coincidences of up to four channels can be counted with the algorithm Composition.

Qualitative Size and Size Distribution

Since large inclusions normally affect the metal quality, it is essential to understand the size of the inclusions or their size distribution. It is possible to use the two algorithms Peaks and Composition to count signals pertaining to different intensity classes. All the visible peaks can be counted by setting the threshold 3•SD greater than the intensity of the element in the matrix. Evaluating the inclusions between consecutive threshold values yields the count of inclusions in the size class that they delimit.

Figure 4 shows an example of the count of peaks and coincidences between 3 and 9.SD relative to small size inclusions, between 9 and 15.SD to medium size inclusions, and greater than 15 to large size inclusions. Such evaluations enable producing qualitative inclusion size distributions.

Figure 4. Calculating the inclusions between consecutive threshold values provides the number of inclusions in the size class that they delimit.

Additional Benefits of Advanced Spark-DAT Applications

The algorithm QuIC (Quantification of Inclusion Content) yields the insoluble fraction of an element owing to its existence in a specific type of inclusion for numerous peak intensity classes, which then enable estimating the average Equivalent Spherical Diameter (ESD) of inclusions for the relative size classes.

Quantitative Size Determination

Since large inclusions normally affect the metal quality, it is essential to understand the size of the inclusions or their size distribution. As mentioned earlier, the algorithms Peaks and Composition can be utilized for counting signals relative to different intensity classes, but enable constructing only qualitative size distribution diagram. Conversely, the Spark-DAT algorithm QuIC facilitates quantitative analysis of inclusions in terms of size and size distribution. Figure 5 illustrates an example of a size distribution diagram.

Figure 5. An example of a size distribution diagram.

Quantitative Oxygen Analysis at Low Concentration

The Advanced Inclusions Analysis with the Spark-DAT algorithm QuIC enables estimating the total oxygen concentration directly from the data acquired on the oxide inclusions (composition and concentration). It is a quantitative method and calculates the oxygen concentration by OES even lower than 30ppm, as delineated in Figure 6.

Figure 6. The comparison results obtained with a combustion analyzer for low alloy steel samples taken in the continuous casting mold (samples and combustion results with permission of R. Dumarey and F. Medina, from ArcelorMittal, Gent).

Figure 7 illustrates the quantitative oxygen analysis, where the Spark-DAT algorithm QuIC is used to compare total oxygen concentrations obtained by measuring several CRMs with their certified values.

Figure 7. Comparison of total oxygen concentrations obtained by measuring several CRMs with their certified values.

Other Applications of the Algorithm QuIC

The following are the other applications of the Algorithm QuIC:

• Area fraction (or surface fraction Sf) (i.e., the fraction of the surface occupied by the inclusions of the given type)
• Concentration of an inclusion type
• Insoluble concentration (or fraction) of an element as a specific type of inclusion.

On-Line Analysis and Off-Line Investigations

It is possible to simultaneously monitor the Spark-DAT results, count of intensity peaks and coincidental peaks with concentration values. It is possible to process, display, transmit, and store the Spark-DAT results like any standard OES result. Figure 8 shows an OXSAS screen depicting partial results of an analysis including elemental determinations and inclusion related data (peak counts, inclusion counts and size).

Figure 8. An OXSAS screen depicting partial results of an analysis including elemental determinations and inclusion related information.

It is possible to store the Spark-DAT intensity data in standard text (.txt) or comma separated value (.csv) files, which can be utilized off-line for subsequent analysis on inclusions or for research and development of new techniques or algorithms. They can be graphically represented with the Spark-DAT viewer integrated in OXSAS. Thermo Scientific continues to improve existing algorithms or create new algorithms, which can be accessed by existing users together with OXSAS upgrades.

Conclusion

The ARL iSpark metals analyzer becomes a versatile instrument when coupled with the optional Spark-DAT methods. In the steel industry, Spark-DAT methods offer rapid, convenient and economical solutions for inclusion analysis for routine use or research. The standard Inclusion Analysis Spark-DAT methods enable ultra-fast on-line qualitative inclusion analysis like counting of inclusions and determination of their type.

The Advanced Inclusion Analysis Spark-DAT method facilitates quantitative inclusion analysis, including size determination, determination of total oxygen content and other parameters within a few seconds to a couple of minutes. This makes it an effective tool to control inclusions and steel cleanness during production.

This information has been sourced, reviewed and adapted from materials provided by Thermo Fisher Scientific - Elemental Analyzers.

For more information on this source, please visit Thermo Fisher Scientific - Elemental Analyzers.