Combating Al-Si porosity: the strontium hydrogen myth.
It is common practice for aluminum foundrymen to modify aluminum-silicon (Al-Si) alloys with strontium (Sr) to change the shape of the silicon and enhance the metal's mechanical properties. Unfortunately, Sr also tends to increase porosity in castings, sometimes to the point where the porosity cancels out the Sr's contributions to mechanical properties.
Porosity in Al-Si castings occurs because of the negative pressures generated by solidification contraction, as well as the pressure from the evolution of dissolving hydrogen from the growing solid into the adjacent liquid. It has been thought that Sr introduces more hydrogen to the melt, thus increasing porosity. However, there is no quantitative data to verify that assumption.
Besides hydrogen content, factors such as cooling rate, modification, alloying elements, grain refinement and inclusion content play a role in forming porosity. This article details an extensive study of several of these factors, with an emphasis on the effect of cooling rate on porosity formation.
The research's objective was to clearly establish quantitative correlations between the above-mentioned variables, the pore volume fraction and pore size in Al-Si alloys to comprehensively explain the effect of modification on porosity. In the process, the data proved that while Sr does increase the severity of porosity, it adds no significant amount of hydrogen to the metal, and must therefore affect it in other ways.
Developing a Test
For the research, three Al-Si alloys with different Silicon levels were used. In addition to the standard A356 alloy (7.0 wt% Si), pure silicon and commercial purity aluminum were added to create alloys with 5.7 and 8.1 wt% Si, respectively.
For each experiment, a 22-lb (9-kg) melt was prepared in a silicon carbide crucible using an electric resistance furnace. Melt temperature was held constant at 1346F (730C). The tests were conducted on both Sr-modified and unmodified melts at various hydrogen levels and without grain refiners.
To study the effects of cooling rate, directional solidification of the test castings was necessary. The alloys were poured into preheated molds made from inorganic refractory insulating board coated with iron oxide wash [ILLUSTRATION FOR FIGURE 1 OMITTED]. At the base of each mold was a copper plate sprayed with a jet of cooling water to quickly establish a solid layer at that end of the casting.
Once that solid layer was established, the copper plate was removed and the water jet was allowed to impinge directly onto the casting's solid surface. The cooling rate was monitored by eight thermocouples placed at various intervals along the length of the mold.
After solidification, the castings were machined into nine pieces for image analysis. Each sample was evaluated for area percentage porosity, average pore diameter and pore density (number of pores per square inch).
Porosity is measured against the theoretical density of an alloy - the ideal density at which there is no porosity at all. For this study, the theoretical densities of the three Al-Si alloys were determined from the bottom portion of degassed specimens that had been poured into a water-cooled copper mold that, due to the relatively high solidification rate, would produce a density very close to theoretical. The theoretical density of the Sr-modified alloy was assumed to be the same as the unmodified alloy. The theoretical densities of A356, Al-5.7 wt% Si and Al-8.1 wt% Si were determined to be 2.679, 2.685 and 2.674 grams per cu centimeter, respectively. A linear relationship exists between the theoretical density and the silicon content in these alloys.
The effect of Sr modification on melt hydrogen pickup was studied by measuring the hydrogen content before and after the Sr addition. Contrary to previous reports, Sr did not raise the hydrogen level of the alloy.
However, for the same preaddition hydrogen content and casting cooling rate, Sr significantly increases the pore volume fraction, pore size and pore number density. For example, at a cooling rate of 33.8 [degrees] F (1 [degrees] C)/sec and 0.26 ml/100g hydrogen, average pore diameter increases from 30 to 50 microns upon modification.
Sr addition also introduces a much wider variation in the size of pores. In unmodified alloys the maximum pore size is about 270 microns. Sr modified alloys exhibit a wider size range, with a maximum pore size of 390 microns. The results also suggest that the total pore number density is about twice as high in modified alloys.
From this data, it is clear that some other attribute of besides hydrogen aggravates porosity.
A decrease in the cooling rate increases both the total amount of porosity and the average pore size in both modified and unmodified Al-Si alloy castings. With increasing solidification rates, however, less time is available for hydrogen to diffuse into the interdendritic spaces of the partially solidified metal, resulting in smaller pore sizes.
In addition, the higher temperature gradients present in rapid solidification tend to limit the length of the mushy zone to make feeding easier and retard porosity formation. Castings cooled at higher rates tend to contain less and more finely dispersed porosity.
Test data indicates that in both modified and unmodified alloys with the same given hydrogen content, the pore volume fraction decreases as the cooling rate increases [ILLUSTRATION FOR FIGURE 2 OMITTED]. Conversely, at a given cooling rate the pore volume fraction increases with increased hydrogen content. At lower cooling rates (typically less than 33.8 [degrees] F/sec) the pore volume fraction becomes much more sensitive to the hydrogen level.
Results also show that a critical cooling rate of about 33.8 [degrees] F per second exists, below which the difference in pore volume between modified and unmodified alloys becomes very significant. That rate seems to be slightly elevated for the 7.0 and 8.1 wt% Si. From these curves, it is clear that to avoid modification-induced porosity, the melt hydrogen content should be lowered to at least 0.1 ml/100g by degassing the melt, while the local cooling rate should be greater than 33.8 [degrees] F/sec.
The cooling rate also affects the individual pore size as measured by the average pore diameter. An increase cooling rate allows less time for the hydrogen to diffuse before complete solidification, limiting pore size.
To gauge the significance of silicon content (independent of Sr or cooling rate) on porosity formation, the 5.7, 7.0 and 8.1 wt% alloys were poured in their unmodified state and with a hydrogen content of about 0.2 ml/100g. Tested over all ranges of cooling rates, the pore volume fraction of the 7.0 alloy was slightly higher than that of the 5.7 or 8.1. Thus, the 7.0 may be the Al-Si alloy most prone to porosity among those within the silicon range of 5.7-8.1 wt%.
Although no measurement of melt inclusion content was made in this study, the increase in pore number density in the Sr-modified alloy may be due to higher inclusion content or a decrease in surface tension associated with modification. Inclusions can act as a catalyst for the heterogeneous nucleation of pores, and lower surface tension can also increase the nucleation rate and lead to a larger number of pores. In either case, a pore that forms at an early stage of solidification can grow for a longer period of time, increasing pore size.
The results of this study indicate that Sr modification does not automatically increase melt hydrogen in Al-Si alloys. The observed increase in porosity does not seem to be caused by a single factor, but by a complex interaction of different parameters, which may include surface tension, inclusion content and freezing in modified alloys.
The information in this article was originally presented at the 1994 AFS Casting Congress, and is available in full through AFS Transactions.
|Printer friendly Cite/link Email Feedback|
|Author:||Gruzleski, John E.|
|Date:||Mar 1, 1995|
|Previous Article:||Channel furnace start-up demands careful attention.|
|Next Article:||Forecast remains bullish for '95.|