Laser radiation has maintained entry in his capacity as a contactless working tool into many industrial applications in the field of laser material processing like drilling or cutting of material. In the field of chemical analysis by laser-induced breakdown spectroscopy (LIBS) laser setups are used routinely e.g. for testing for mixed-up components of pipes. This work bridges the two fields of defined material ablation and chemical analysis by LIBS. Time and spatially modulated Nd:YAG laser radiation is used to ablate effectively scaled low-alloy steel surfaces and to analyse simultaneously. By using triple pulse bursts the depth penetration can be enhanced by a factor of up to 6 in comparison to single laser pulses taking the same total laser burst energy. For several elements the necessary ablation depths are estimated for a representative analysis at the scale side of the production control samples. Some elements can be analysed in the upper surface layer whereas e.g. for the element carbon a depth of at least 200 µm has to be removed. By using double pulse bursts the value of 55.2 rel.-% for the relative standard deviation of the procedure of carbon calibration at the scale side without prior material ablation can be reduced to a value of 2.8 rel.-% after removing a surface layer with a thickness of approx. 300 µm. The first time in the field of LIBS tailored crater geometries were generated by a 3D laser scanner and optimised scan strategies which lead to smaller crater debris in comparison to locally fixed laser radiation. Interactions between plasma and crater walls and with it falsified plasma compositions by scale material are reduced analysing in the middle of the produced crater. For an adequate analysis in a laser generated crater aspect ratios of smaller than 1 are to be aspired. Calibration measurements showed that the LOD values for many elements are lower than 10 µg/g up to an analysis depth of about 300 µm. With increasing number of evaluated laser pulses of up to 3000 on each sample position the LOD values are significantly improved, e.g. for Ti to 1 µg/g. Using the curve-of-growth theory conclusions can be made concerning the self-absorption of emission lines. By means of a double logarithmic plot of calibration functions the linear section can be separated from the non-linear one showing self-absorption. With calibrations in sections the analytical performance can be improved by a factor of up to 3. For the analysis of scaled steel samples a calibration method is developed using in a first step double pulse bursts. Double pulse bursts lead to an effective scale ablation and simultaneously to an analysis of elements which are already contained representatively in the upper scale layer. In the second step single laser pulses are focused onto the same sample position leading to smaller plasmas and therefore smaller self-absorption. In this way the use of several analyte lines for covering bigger concentration ranges of one element can be disclaimed. On-site measurements in a steel mill and in a hot strip rolling mill were carried out for testing the results in the industrial practice. For the test in the steel mill a laser system was built up for the combination of ablation and analysis of scaled low alloy steel samples analysing automatically the samples at different sample positions. Two methods are used for calibration. In a first step double pulse bursts and in the second step single laser pulses are taken for ablation and analysis. During the test phase the optical elements in the laser trace were not cleaned. By means of referencing and periodical recalibrations a relative standard deviation of the procedure of 2.3 rel.-% is achieved for the element Mn. The drift behaviour of the intensities can be reduced by using reference lines which lie in a similar wavelength range as the analyte line. Thereby the ambient gas emission line Ar 415.85 nm is proved to be most suitable for the analyte line Al 394.4 nm. For the on-site measurements in the hot strip rolling mill LIBS analysis at up to 520°C warm steel coils were carried out concerning the testing for mixed-up components. Seven different coils were measured by a water-cooled measuring head constructed especially for the analysis of warm coils. The setup including a 5 m long fibre optics was optimised for a high transmission of UV light in the wavelength range of about 190 nm. With this setup the accuracies of warm coils lie in a comparable range as for chilled scaled samples.
展开▼
