Abstract
For the analyses of oxide in aluminium, several methods have been proposed, most of which have turned out to be still unsatisfactory having respective merits and demerits. The M. T. method, reported 1949 by M. Tournaire, however, was a good method, standing on the metallurgical view point, to nalise the nonmetallic impurities, especially the total insoluble oxide.
The authors, therefore, made a test analysis of alumina as a preliminary experiment by the M. T. method, and, as a result, found out that the insoluble oxide still remained almost unchanged, provided that the conditions, such as decomposition temperature, reaction time, etc, were appropeiate.
As the results of further experiments were also favorable, the authors made routine analyses by the M. T. method, of the total oxide of the same samples from each stage of rolling of aluminium sheet. The results were as follows:
1. Total oxide in original virgin ingot varies from 0.008% to 0.020%.
2. The total oxide gradually decreases during melt-treatment and shows a constant value in the final slab.
3. Distribution of oxide in the slab doesn't change by rolling.
4. Blister of the rolled sheet seems to have nothing to do with the oxide content, while the parts around scratch, crack, abnormal blow-hole, etc. contain more oxide than the normal and sound parts.
5. After all, it may be concluded that the oxide content of the normal parts of the rolled aluminium sheets show nearly constant value, about 0.010%.
Moreover, by means of the X-ray counter method, one of the physical methods other than the above, the authors observed α- and γ-Al2O3 in the vergin ingot, and α-, β-, and γ-Al2O3in the cast slab.
