Abstract
The subject of this thesis is the mathematical modeling of a new method for detecting oxidic inclusions, specially of Aluminum, in steel. For detecting inclusions with a spark-induced emission spectroscope, electric strokes of high energy are discharged on the surface of a steel sample. As a result, the surface evaporates and the elements in the vapor emit energy on their characteristic frequencies. The intensity of this energy is measured. The results have to be described by a mathematical model, so that conclusions regarding the amount of impurities can be drawn. Therefore a probability distribution describing the measurements is derived and discussed in detail. The parameters of this distribution are characteristic values of one data record. Since common methods for the estimation of parameters fail in this case, a modified procedure for the identification is developed. Finally the available data sets of different steel samples are statistically analyzed with the previously described method.
| Translated title of the contribution | Modeling and Statistic Analysis of Data Sets Resulting from the Detection of Aluminum Inclusions in Steel by means of Spark-Induced Emission Spectroscopy |
|---|---|
| Original language | German (Austria) |
| Publication status | Published - Jun 2000 |
Fields of science
- 101 Mathematics
- 101029 Mathematical statistics