% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@PHDTHESIS{Middelmann:854354,
author = {Middelmann, Olaf},
othercontributors = {Bührig-Polaczek, Andreas and Senk, Dieter},
title = {{D}etektion von {P}hasen in {A}luminiumlegierungen mittels
optischer {E}missionsspektrometrie mit intensitätsbasierter
{E}inzelfunkenauswertung},
volume = {34},
school = {RWTH Aachen University},
type = {Dissertation},
address = {Aachen},
publisher = {Gießerei-Institut der RWTH Aachen},
reportid = {RWTH-2022-09493},
isbn = {978-3-944601-23-6},
series = {Ergebnisse aus Forschung und Entwicklung},
pages = {1 Online-Ressource : Illustrationen, Diagramme},
year = {2022},
note = {Druckausgabe: 2022. - Auch veröffentlicht auf dem
Publikationsserver der RWTH Aachen University 2023;
Dissertation, RWTH Aachen University, 2021},
abstract = {The motivation for this work is the methodical extension of
spectral analysis, to additionally determine present phases
in the microstructure. In this thesis the three phase types
titanium borides, aluminum oxides and AlFeSi phases are
considered. With regard to the ferrous phases, the
differentiation between the α-Al15(Fe,Mn)3Si2 and
β-Al5FeSi modification is the main focus. Samples for
spectral analysis are produced in sand casting process.
Spectral analysis is performed using the QSG 750-II,
SPECTROLAB LAVM12 and Q8 Magellan at the manufacturers
facilities. The measured intensities of the element channels
were used for method development. Metallographic evaluations
serve to control the cast microstructure and to evaluate the
influence of the analysis process. Using a scanning electron
microscope (REM-EDX) the microstructural constituents are
determined. The analysis process of OES-PDA shows a massive
influence on the sample surface condition and
microstructure. Remelting zones show inhomogeneous
compositions based on the mixture of adjacent phases. With
regard to the evaluation of titanium borides in the
material, the procedure of outlier determination with
matching spark numbers for titanium borides is appropriate.
A comparison of the considered methods for limit value
determination shows a higher sensitivity for the procedures
IQR and MAD in comparison to SD and $SK_IQR.$ While
quantification depends strongly on the choice of method and
parameters. The evaluation of the oxide load requires a
prior determination of the measurement error for oxygen. A
high-purity aluminum sample was used to define a new zero
intensity level. The determination of the percentage
deviation of each spark discharge to this zero intensity
allows for the definition of a new criterion for the
determination of oxides. To distinguish between the two
phase types α- and β-AlFeSi the presented method relies on
the element intensities of iron, manganese silicon and the
ratio of manganese to iron; calculating the minimum squared
deviation to defined phase regions. The quantified ratios of
these phases do not correspond to the real phase fractions
of the microstructure because of a strong influence by the
remelting zone. However, the comparison of different samples
shows the same tendencies as the metallographic evaluation.
The self-learning analysis methods from the field of data
mining show themselves to be methods with great potential,
but in the studied case the necessary boundary conditions
could not be sufficiently fulfilled in order to obtain
reliable results. The presented methods successfully extend
the functional range of OES PDA.},
cin = {526110 / 520000},
ddc = {620},
cid = {$I:(DE-82)526110_20140620$ / $I:(DE-82)520000_20140620$},
typ = {PUB:(DE-HGF)11 / PUB:(DE-HGF)3},
doi = {10.18154/RWTH-2022-09493},
url = {https://publications.rwth-aachen.de/record/854354},
}
guest ::