PSI - Issue 68

V. Milani et al. / Procedia Structural Integrity 68 (2025) 1181–1187 V. Milani, G. Angella, G. Timelli/ Structural Integrity Procedia 00 (2025) 000–000

1182

2

compromise the mechanical properties of the alloy. The most harmful inclusions arise from the oxidation of aluminium, which results in the formation of double oxide films entrained in the melt during turbulence (Ghanaatian & Raiszadeh, 2022). Also known as bifilms, they act as cracks and contribute to porosity formation (Gyarmati et al., 2018), leading to early initiation of fracture, and causing a substantial scattering of the tensile properties of casting alloys (Liu & Samuel, 1998). Various methods are available to improve the molten metal cleanliness, including the injection of inert or reactive gases, filtration, and fluxing. Rotary degassing combined with flux injection is also common, improving molten metal quality. Salt fluxes assist in the removal of inclusions by enveloping them and floating on top of the melt into the dross (Milani & Timelli, 2023). The effect of oxide films reduces the reliability of castings. The literature widely employs the two-parameter Weibull distribution to characterise fracture-related mechanical properties. Greater values of Weibull modulus and scale parameter indicate that the samples have fewer defects, thus suggesting higher reproducibility (El-Sayed et al., 2022). On the other hand, the three-parameter Weibull distribution can also be useful when two populations of defects, e.g., new and old bifilms, are expected in the casting (Tiryakioglu & Campbell, 2010). The cumulative probability function of specimen failures using the Weibull statistics is expressed as (Timelli, 2018) ! = 1 − (− ) "# % $ * & + (1) Where x is the variable being measured, λ is the threshold parameter, η is the scale parameter and β is the shape parameter, also referred to as the Weibull modulus. This work aims to understand the effects of different salt fluxes and processing temperatures on the scattering of the tensile properties of a gravity cast AlSi9Cu3(Fe) alloy by using the Weibull statistics. In particular, the scattering of the elongation to fracture was investigated using both two and three-parameter Weibull distributions. 2. Experimental Procedure The experimental procedure involved the production of separately poured tensile specimens by preliminary treating the molten metal with two different salt fluxes, namely Flux A and Flux B, in the form of salt granules. The liquid metal was also untreated for comparison. Furthermore, two different processing temperatures were studied, i.e., 720 and 760 ± 5 °C. Flux A and Flux B are both composed of NaCl and KCl, but while Flux A contains NaF and K 2 CO 3 , Flux B contains MgCl 2 . About 5 kg of AlSi9Cu3(Fe) alloy were melted in a SiC crucible using an electrical resistance furnace. The charge comprised 20 wt.% ingots and 80 wt.% scrap originated from a high pressure diecasting foundry returns in the form of gating and runner systems, and overflows of the same alloy. The chemical composition of the alloy, measured before and after casting the tensile specimens, remained constant throughout the entire experimental campaign (see Table 1).

Table 1. Chemical composition of the experimental alloy (wt.%). Si Fe Cu Mn Mg Zn

Cr

Ni

Ti

Al

8.53

0.91

2.26

0.22

0.27

1.01

0.04

0.06

0.04

bal.

After melting, the alloy was manually stirred for approximately 30 seconds, then stirred at 650 rpm for 3 minutes using a coated impeller with a diameter of 90 mm. When flux was used, it was manually introduced in the vortex generated by stirring. After the fluxing process was completed, the impeller was withdrawn, and the melt was held at the processing temperature for 20 minutes. The generated dross was skimmed and the metal was then poured into a pre-heated (250 ± 5 °C) permanent steel mould that was designed according to the CEN/TR 16748:2014 document (16748:2014, CEN/TR, 2014) to produce cylinder-shaped casting bars with a total length and mean diameter of 190 and 15 mm, respectively. The melting, fluxing and casting of the tensile samples process was repeated twice for each experimental condition.