Analysis of directly cast metallurgical slags

Importance of the topic

Directly cast metallurgical slags are fundamental by-products of iron and steel production processes. Rapid and precise assessment of their composition enables optimized furnace operation, improved impurity removal, and enhanced refractory lifespan. Implementing a streamlined analytical workflow supports real-time quality control and process efficiency in the steel industry.

Objectives and overview of the study

This work evaluates the capability of the Thermo Scientific ARL OPTIM’X WDXRF sequential spectrometer to analyze hot, directly cast steelmaking slags without additional sample milling or pressing. Key goals include demonstrating calibration performance, analytical precision, and practical applicability for routine slag monitoring.

Methodology and used instrumentation

Sample Preparation

• Molten slag poured directly into a steel ring using a dedicated ceramic mould, enabling immediate XRF measurement upon cooling.

Instrumentation

• ARL OPTIM’X WDXRF spectrometer (50 W Rh anode tube, no water cooling, SmartGonio goniometer covering F to U).
• OXSAS software for instrument control and multivariable regression calibration.
• Optional 200 W version or additional fixed channels available for enhanced sensitivity.

Calibration

• Ten secondary slag standards used to cover typical concentration ranges for Al₂O₃ (5–12 %), MgO (2.4–9 %), SiO₂ (34–38 %), CaO (32–47 %), and Cr₂O₃ (1.6–7.8 %).
• Multi-variable regression curves established within OXSAS.

Main results and discussion

Calibration accuracy, expressed as Standard Error of Estimate (SEE), was 0.074 % for Al₂O₃, 0.14 % for MgO, 0.23 % for SiO₂, 0.24 % for CaO, and 0.16 % for Cr₂O₃. Short-term repeatability tests (10 consecutive runs) with 100 s total counting time at 50 W yielded relative standard deviations below 0.05 % for all oxides. Extending counting to 200 s or using the 200 W instrument further reduces variance marginally. Comparative analyses confirm results equivalent to those from conventional ground and pressed slag pellets.

Benefits and practical applications of the method

By eliminating pelletizing steps, the direct-cast approach accelerates throughput and reduces sample handling errors. The low-power ARL OPTIM’X offers stable, high-resolution measurements with minimal maintenance and no external cooling. It supports on-site process control, reactive adjustment of slag chemistry, and consistent quality assurance in steel mills.

Future trends and possibilities for use

Advancements may include automated mould handling for continuous monitoring, expanded calibration libraries for diverse slag types, integration with process analytics platforms, and fusion bead preparation protocols to mitigate mineralogical effects. Incorporating machine learning algorithms could further refine predictive quality control and real-time process optimization.

Conclusion

The ARL OPTIM’X low-power WDXRF spectrometer successfully quantifies key slag oxides directly from cast samples with high accuracy and repeatability. Its streamlined workflow and flexible configuration make it a valuable tool for industrial slag analysis. For maximum accuracy across varied slag compositions, fused bead preparation remains recommended.