| Summary: | Growth in the aluminium casting industry has lead to greater competitiveness among manufacturers. It has been suggested that improving metal quality may provide a greater advantage for ION. Inclusions are one aspect of metal quality to which ION Ltd has found difficulty in the past to identify and quantify. The Prefil method has recently claimed to be a method that could provide both quantitative and qualitative analysis of inclusions in molten metal. It operates based on filtration and was ideally suited to ION because one form of testing molten aluminium (Prefil curves) was claimed to be simplistic enough for plant operators to use. The other testing method (Prefil metal residue analysis) is claimed to be more complex yet able to provide identification of inclusions in the metal. However, the exact mechanisms filtration occurring in the Prefil was not clear. This prevented ION from using this machine effectively. Thus the aim of this research was to examine the mechanisms of filtration occurring in the Prefil, to identify inclusions trapped by the Prefil and to quantify oxides trapped in the plant. This was done for four locations in the plant; two before master alloy addition Al-Ti-B and two after Al-Ti-B. The two before Al-Ti-B additions were where the metal was melted (CL) and then first degassed (D-EL). The other two measurements occurred before (LBD) and after a second degassing process (LAD). Identification of inclusions was obtained through SEM EDX analysis. These inclusions were then characterised through image analysis on a number of regions in the metal residue to identify the mechanism of filtration and the quantities of inclusions. Metallographic examinations have shown that particles were trapped above the filter during a Prefil test. These particles were identified to be mainly oxides at CL and D-EL locations. It was found that the flow rate determined in Prefil curves correlate with cake thickness of oxide inclusions in and above the filter. Al-Ti-B additions caused small TiB2 particles to be trapped above the Prefil filter. A combination of oxides and TiB2 particles were thus characterised for the LBD and LAD locations, with the LAD location having the lowest flow rate due to flow resistance dominated by TiB2 particles. Thus, it is suggested that Prefil curves may not be an effective means to evaluate inclusions in the plant when TiB2 particles are present.
|
|---|
