Slag analysis by laser-induced breakdown spectroscopy

Citation
G. Doujak et al., Slag analysis by laser-induced breakdown spectroscopy, STAHL EISEN, 121(3), 2001, pp. 53-58
Citations number
16
Categorie Soggetti
Metallurgy
Journal title
STAHL UND EISEN
ISSN journal
03404803 → ACNP
Volume
121
Issue
3
Year of publication
2001
Pages
53 - 58
Database
ISI
SICI code
0340-4803(20010315)121:3<53:SABLBS>2.0.ZU;2-T
Abstract
Within the context of optimizing blast furnace and liquid steel processes t he timeliness of the slag analysis deserves special attention. In addition to analytical precision, the cost and time required for the analysis play c entral roles. The laser-induced breakdown spectroscopy provides all require ments necessary for quick analysis of both metallic and oxidic materials as an important element of production control. To enable the direct analysis of slags the sampling method of liquid slag had first to be modified in ord er to meet the requirements of this new analysis technique. Special probes were used for taking the slag samples. The optimization of the filling and cooling conditions led to samples with smooth surfaces and only very few st ress cracks. The samples were suitable for immediate analysis by laser-indu ced breakdown spectroscopy (LIBS). This method requires considerably less t ime for the analysis than XRF spectroscopy. Due to the lack of reference ma terials, the new analytical equipment is calibrated based on the results ob tained from parallel XRF analyses of solidified vitreous slag samples. The comparison of the LIBS results with the results obtained by XRF analysi s has shown good agreement. However, precision is poorer. But as long as th e achievable precision is within the required tolerance limits the LIES is a suitable analytical means for process control purposes. The different sla g amounts and viscosities in liquid steel processes as well as specific on- site situations (converter, ladle) pose additional problems for successful slag sampling. Currently, yield rates of 80 % are achieved. The target is > 98%.