Abstract
The present study was performed on A356 and B319 alloys in mechanically stirred or degassed condition. Melts were Sr-modified and grain-refined. Hydrogen content was varied from less than 0.1 ml/100 g Al to ∼0.4 ml/100 g Al; Fe was increased to 0.8% in B319 alloy. Lanthanum and cerium were added as 99.5% pure metals. Two main techniques were used to investigate porosity formation: fracture surface of tensile or fatigue test bars, or reduced pressure test (RPT) method. Porosity type and shape were examined. The results show that pore size is more influential than small scattered ones from a mechanical point of view. Tensile testing is affected by porosity located at the center of the testing bar, whereas edge porosity is responsible for crack initiation in case of fatigue testing. Intermetallics precipitate in the form of intercepted platelets which restricts the flow of the molten metal, leading to formation of shrinkage cavities. Precipitation of clusters of compounds from the liquid state such as Al2Si2Sr, Mg2Sn, Al3Ti, or added Al2O3particles would as well act as nucleation sites for porosity formation. Most oxides were observed in the form of long branched strings. In some cases, bifilms were also reported in addition to SrO and MgO.
1. Introduction
Macroscopic cavities occurring in a casting are normally a shrinkage type due to poor feeding. Volumetric shrinkage in metals transforming from liquid to solid state may range from 3 to 10%, with 5–8% being typical of most cast alloys and taking place in different forms. The other major source of porosity is caused by the dissolution of the hydrogen gas in the solidifying solid as a result of the reduction of hydrogen solubility [1–10]. Majority of porosity observed in castings is caused by a combination of gas and shrinkage. These pores take place mainly in the interdendritic regions, since these are the last parts to solidify [11–13]. It is suggested that the pore nucleation and growth can be expressed as [13–15]where = equilibrium pressure of dissolved gases in the melt; = pressure drop due to shrinkage; = pressure of the atmosphere over the system; = pressure due to the metallostatic head; and = pressure due to surface tension at the pore/liquid interface.
In order for a pore to grow, its internal pressure ( = + ) > ( = + + ) to avoid pore collapse. The dissolved gas pressures and the pressure drop due to shrinkage are the main parameters controlling the formation of porosity. Iron and rare earth metal-based intermetallics are considered as favorable sites for porosity formation mainly due to intersection of the intermetallic platelets preventing/retarding the motion of the liquid metal during solidification [16–18].
A new model to predict location-specific volumes of entrained defects as well as oxide induced defects in aluminum castings was developed by Ridgeway et al. [19]. The authors claim that their model (Oxide Entrainment Number) can accurately determine the gaseous pore number in a selected position in the casting. Effect of defect size on tensile elongation in aluminum castings was investigated by Koprowski et al. [20]. The study showed that there is a good correlation to total defect size and the reduction of % elongation of the samples. Prediction of quantifiable damage that oxide films may cause to the quality of aluminum castings during the filling process remains theoretical, due to lack of supporting data. Hoefert [21] carried out a set of experiments to turn on and turn off turbulent filling conditions with film entrapment in order to obtain the needed data required to predict the actual porosity and tensile damage.
The present study was undertaken to highlight some important metallographic aspects of porosity in Al-%Si castings that were observed by our research group during the past two decades in the course of our studies and laboratory experiments, to be shared with other scholars. These castings were produced using different methods, namely, metallic mold, sand mold as well as the application of the RPT and ultrasonic techniques.
2. Experimental Procedure
The chemical composition of the base alloy is shown in Table 1. The ingots were melted in a 40 kg capacity SiC crucible using an electrical resistance furnace. The melting temperature was kept at about 750°C. Addition of alloying elements was made using master alloys depending on the required composition: Al-25%Fe, Al-10%Sr, Al-5%Ti-1%B, Al-10%Ti, Al-10%Ti, Al-15%Zr, pure La, and pure Ce.
The molten metal was either mechanically stirred or subjected to degassing for a sufficient period of time (about 25 minutes) applying high purity, dry argon gas that was introduced into the molten bath at a constant rate of 20 m3/h, using an impeller made of chemically treated graphite (rotating at ∼135 rpm), to reduce the hydrogen concentration in the molten alloys as well as to ensure the uniform mixing of the additions. Strontium was added towards the end of degassing, followed by another 3 minutes of degassing to ensure complete dissolution of the strontium.
Hydrogen content was varied using different amounts of small potato pieces that were introduced into the molten bath using a perforated graphite plunger. In each case, the hydrogen content was analyzed by casting a sample of the melt in a Ransley mold. This is a mold specially designed to prepare samples for hydrogen analysis using the Leco subfusion technique [22]. Four samples were used for each melt prepared and the average of the four was taken as the hydrogen content of the specific melt. Complete details are provided in [1]. The hydrogen was also monitored using an AlScan apparatus [23].
Three samples for chemical analysis were also taken at the time of the casting; this was done at the beginning, in the middle, and at the end of the casting process to ascertain the exact chemical composition of each alloy. Degassed melt was poured into a metallic permanent mold (ASTM B-108, that was preheated at about 460°C, DAS ∼ 22–25 μm) to prepare test bars, the dimensions of which are shown in Figure 1. Samples from mechanically stirred melts were examined using the RPT technique. In certain cases, a sand mold preheated at 150°C was used (DAS ∼ 10–120 μm) (Figures 2and 3). In addition, porosity in fractured fatigue tested bars (DAS ∼ 15–20 μm, Figure 4) was also investigated [24].
(a)
(b)
For mechanically stirred melts, for each condition, RPT samples were taken. Samples for metallographic examination were prepared from the casting. The study was carried out using a Clemex image analyzer in conjunction with an Olympus PMG3 optical microscope. Characterization of the phases observed in both alloys was carried out employing a field emission scanning electron microscope (FESEM) equipped with an EDS system.
3. Results and Discussion
3.1. Gas and Shrinkage Porosity
The solubility of hydrogen in pure aluminum, and 319 and 356 aluminum alloys, is shown in Figure 5[25], whereas Figure 6illustrates the porosity development during solidification [26]. The results of Boudreault et al. [27] on the formation of gas porosity in 356 and 319 alloys solidified under different rates (using different molds) are controlled by the amount of absorbed hydrogen and solidification time, applying the same techniques described in the experimental section. The results are presented in Figures 7(a)and 7(b). Figure 8demonstrates the presence of a mixture of gas and shrinkage cavities at the edge of a tensile bar of non-modified A356 alloy in the T6 condition (samples were supplied), whereas long shrinkage cavities can be seen at the center of the sample. In both cases, pores are seen as disconnected (i.e., separated by a few dendrites). However, some of these pores may join the neighboring ones under the applied stress (blue arrow in Figure 8(b)). Figure 8(c)is a high magnification image of the dendrites in the white circle in (b) revealing two rows of dendrites with a narrow crack passing through whereas Figure 8(d)is an optical micrograph exhibiting the possibility of the growing of gas pores through joining each other (arrowed). Considering the topography of the fracture surface, it is difficult to measure the actual pore size that leads to crack initiation or propagation.
(a)
(b)
(a)
(b)
(c)
(d)
Figure 9depicts the shape and number of pores in Sr-treated A356 alloy with two levels of hydrogen (0.25 ml/100 g and 0.4 ml/100 g Al). Figure 9(a) shows several round pores (∼400 μm in diameter) distributed throughout the fracture surface of a tensile bar in T6 temper condition (H2 = ∼0.25 ml/100 g Al). These pores were characterized by their large size (∼400 μm in diameter) and relatively shallow depth. Porosity in high hydrogen-containing alloy (0.4 ml/100 g Al) is shown in Figure 9(b) where pores are much deeper than those displayed in Figure 9(a). As illustrated in Figure 9(b), their growth was limited by the surrounding dimple structure. In some areas, the bottom of the pore could not be reached as demonstrated in Figure 9(c), long white arrow.
(a)
(b)
(c)
(d)
(e)
(f)
In all cases, the pore surfaces are covered by fine eutectic Si particles, arrowed. Some of these particles could be SrO oxides as shown in Figure 9(d) [26, 27]. Figure 9(e) is a backscattered electron image produced from a similar pore showing the form of the SrO as confirmed from the associated x-ray diagram displayed in Figure 9(f). This part will be discussed in more detail in a separate section. It should be borne in mind that the Si Kα line (1.739 keV) is very close to Lα Sr (1.806 keV). Note that in both Figures 8and 9the porosity is caused by excess of H2.
Nevertheless, increasing hydrogen content and hence porosity will lead to reduction in the amount of shrinkage cavities, but not to eliminate them. Figure 10(a)shows a shrinkage cavity where the crack is passing through in 0.4 ml/100 g Al treated A356 alloy (T6 temper) under uniaxial loading. On the other hand, repeated compression/tension testing may lead to impingement of the dendrites (white circles) narrowing down the spacing between two adjacent sets of dendrites as seen in Figure 10(b). Figure 10(c)represents the microstructure corresponding to a polished sample that was sectioned from a casting poured in a sand mold preheated at 150°C showing a severe case of shrinkage formation. The molten material (non-modified A356 alloy) was degassed for about 30 minutes using high purity Ar gas prior to casting (humidity was about 11–13%).
(a)
(b)
(c)
The results of Magnusson and Arnberg [28] and Leitner et al. [29] show that the density of liquid aluminum-silicon alloy increases with increasing silicon content, while silicon reduces the density in the solid state. Silicon content reduces the solidification shrinkage from 6.6 pct for pure aluminum to 4.4 pct for Al-11.6 pct Si. Deev et al. [30] studied the influence of the melt cooling rate on shrinkage behavior during solidification of aluminum alloys. At low cooling rate, the rate of the shrinkage process is much less than the filtration rate and capillary feed, and shrinkage defects are effectively healed. High rates of heat removal from the surface of the casting lead to the frontal nature of crystallization, the mushy zone is practically wedged out, and the rate of healing of hot tears due to filtration again exceeds the rate of their opening due to the shrinkage process.
Ammar et al. [31–33] highlighted the importance of (i) pore position in the casting, size and pore distribution and (iii) inter-pore distance. Figure 11depicts the difference in pore position; pore at the center has less effect on the fatigue life than that on the surface due to stress distribution. In fact, pores at the casting center have a great effect on the premature failure of the castings under uniaxial loading. Examining Figure 12, in the case of separate small pores, the propagation of the crack from one small pore to the next will take a longer time, resulting in increasing the resistance to crack propagation and, hence, longer fatigue life. The high stress concentration associated with a large single pore will enhance the driving force for crack propagation and lead to decreasing the sample fatigue life. Thus, it is the pore size that has a more influential role in determining the fatigue life than disconnected small ones.
(a)
(b)
The effect of surface porosity on the tensile properties of the used A356 alloy in the T6 condition is illustrated in Figure 13(tensile and fatigue testing at room temperature). As can be seen, regardless of the type of testing, the increase in total surface porosity is accompanied by a marked reduction in the alloy strength parameters, i.e., ultimate tensile strength (UTS), yield strength (YS), and fatigue life. The decrease in properties in Figure 14may be interpreted in terms of the high stress concentration at the pore surfaces resulting in a stress gradient between the sample surface and the specimen center as displayed in Figure 11for fatigue-tested samples [33].
3.2. Intermetallics
Figure 15shows the solidification curve of B319 + 0.8%Fe revealing the temperatures of the precipitation of the three types of Fe-based intermetallics. Obviously, it is expected that as the solidification temperature decreases, the size and distribution of the precipitated Fe-phases will decrease. The backscattered electron image in Figure 16(a)illustrates the effectiveness of β-platelets in restricting the flow of the liquid metal leading to formation of a coarse shrinkage cavity. Although precipitation of α-Al12(Fe, Mn)3Si2phase takes place at a temperature lower than that of β-Al5FeSi phase (Figure 16(b)), it does not seem to have an effective role in porosity formation due to branching which would facilitate the motion of the liquid metal between the script arms as displayed in Figure 16(c).
(a)
(b)
(c)
(d)
(e)
(f)
At 540°C, a part of β-Al5FeSi transforms into π-Al8Mg3FeSi6in the form of a group of short platelets as shown in Figure 16(d), as was confirmed from the associated EDS spectrum displayed in Figure 16(e). Due to narrow distances between the π-platelets (about 10 μm in length and 1 μm in thickness) compared to the original β-phase platelets (∼100 μm in length and 5 μm in thickness) coupled with the reduction in the liquid metal fluidity, pockets of small voids may form within the π-phase platelets as illustrated in Figure 16(d), white circles. It should be mentioned here that not only are β-platelets an effective site for causing shrinkage cavities, but also other intermetallics such as rare earth-based intermetallics were found to be capable of restricting the flow of the liquid metal as depicted in Figure 16.
In addition to platelet-shaped intermetallics, high temperature melting compounds were found to nucleate gas and shrinkage cavities. In this section, examples obtained from well-degassed molten alloys will be reviewed. Figure 17(a) shows the precipitation of Mg2Sn compound in branched star-like form within the α-Al dendrites. Considering the melting point of this compound is about 770°C [34] and the α-Al network precipitates at about 615°C (Figure 15), thus the α-Al dendritic structure was precipitated on the pre-existing phase Mg2Sn phase particles leading to formation of a large shrinkage cavity (H2 < 0.1 ml/100 g Al).
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Another feature to be considered is the precipitation of Al2Si2Sr phase in the form of tetrahedral pyramids at the bottom of a cavity (Figures 17(b) and 17(c)). Since the melting point of Al4Sr phase is 1034°C, it is suggested that the reaction between Al4Sr and the Si resulted in formation of the present compound. However, the exact precipitation temperature is not yet determined. In this case, the size of the pore depends on the number of the Al2Si2Sr particles and their occupied size as exhibited in Figure 17(d).
Mahmoud et al. [35, 36] reported that the shape of rare earth (RE) based intermetallics depends on the Ti/RE ratio. In the case this ratio is <1, RE-intermetallics precipitate in the form of thin platelets as shown in Figure 17(e). However, when Ti > RE, the RE-intermetallics precipitate in the form of gray sludge (star-like form) as demonstrated in Figure 17(e) where Ti is about 7 at. % and Li is about 4 at %. In this case, high Ti-containing RE phase particles are joined together at the bottom of a large pore. In contrast, particles outside this pore show no sign of micro/macro-porosity associated with their presence, indicating that the precipitation temperature of this phase is high enough to render the molten metal good fluidity to fill up the spacing between the branches.
Another aspect to be considered is the excess use of grain refiner, in particular Al-10%Ti that leads to formation of Al3Ti phase in the form of clusters of short platelets as shown in Figure 17(f). As a result, several micro-pores may occur between these particles.
In the case that the alloy was grain refined using a master alloy containing Ti and B, it is possible that these two elements may react with Sr in modified alloys forming a complex compound AlTiBSr which apparently has a tendency to form cavities as seen in Figure 17(g), similar to those shown in Figure 17(f). Figure 17(h) offers another possibility for both AlTi and AlSr to be potential sources for porosity formation.
According to Sato et al. [37] and Vakhobov et al. [38], the aluminum corner of Al-Si-Sr ternary alloy system was investigated by means of inverse rate thermal analysis, with identification by x-ray diffraction, EPMA analysis, and microscopic observation. A ternary compound SrAl2Si2, named “T,” exists in this ternary system and the Al-T line is considered as a quasibinary system with a eutectic point at 1.1%Si, 2.4%Sr, 645°C. The aluminum corner is divided into two regions by the Al-T quasibinary system. A ternary eutectic point was determined to locate at 575°C, 13.1%Si, 0.03%Sr, which is very close to the Al-Si binary eutectic point. The modification effect of small amount of strontium addition on Al-Si alloys can be regarded to be due to the formation of a fine ternary eutectic structure. In the Al-SrAl4-T region, a ternary eutectic point was found to locate at 643°C with 1.7%Si, 2.4%Sr. The solubility of strontium and silicon in aluminum was examined along three ray sections. The greatest solubility is shown to occur along the SrSi2-Al (T phase) quasibinary section at Sr : Si = 1 : 2. It is found that the SrAl2Si compound melts at 1010 ± 10°C. This compound forms constitutional diagrams of the eutectic type with aluminum, silicon, strontium, and also SrSi2and SrAl4compounds as shown in Figure 18.
(a)
(b)
3.3. Oxide Films and Dispersoids
Atakav et al. [39] investigated the effect of Sr addition on the melt cleanliness of A356 alloy. Reduced pressure test (RPT) was used to assess melt quality change and it was found that cleanliness was increased due to the fading of Sr. Characterization of the effect of melt treatments on melt quality in Al-7wt %Si-Mg alloys was investigated by Uludag et al. [40] who found that, in both degassed and non-degassed melts, Sr additions degraded the melt quality significantly. Another source to be considered is oxide films, mainly Al2O3and SrO [40–44]. Dispinar and Campbell [45] proposed the concept of bifilms defined as folding/unfolding and furling/unfurling mechanism. According to the authors, the bifilm index range is defined as follows [46]: 0–25 mm: best quality, 25–50 mm: good quality, 50–100 mm: average quality, 100–150 mm: bad quality > 150 mm: do not cast. An example of bifilm-porosity formation mechanism is exhibited in Figure 19. Figure 20shows an example of RPT test samples obtained from A356 alloy melts in the present study after (a) degassing, and (b) degassing plus addition of ∼200 ppm Sr, where the development of increased porosity may be observed due to the presence of SrO.
(a)
(b)
In this section, B319 alloy was used without degassing. The melt was held at 780°C for 3-4 h. In order to avoid particle sedimentation, the melt was mechanically stirred from time to time. Porosity formation was examined using RPT technique. Figure 21(a)shows a good example of so-called bifilms as described by Dispinar and Campbell [45] and Campbell [46] in their investigations. In this case, pores are massive with irregular shape. However, this type of oxide was not observed in all the examined samples. It was seen following vigorous mechanical stirring of the melt. Oxide films in the form of long strings passing randomly throughout the matrix (solid arrows) were more commonly observed as shown in Figure 21(b)creating several micro-porosities <10 μm in length (broken arrows).
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
Strontium based oxides were often seen in the form of thin platelets within the shrinkage cavities as displayed in Figure 21(c). Although the exact composition could not be confirmed, x-ray imaging revealed the presence of Al, Sr, and O in this plate. As shown in Table 1, the present B319 alloy contains about 0.3%Mg. As a result, Mg could be oxidized forming MgO as displayed in Figure 21(d)tangled with other oxide films. Since the atomic number of MgO is close to Al, an optical microstructure was used to demonstrate such a complex situation. Figure 21(e)illustrates the O2distribution in a similar area.
Grain refining using Al-5%Ti-1%B is commonly employed in the aluminum industry (approximately 0.1–0.15%Ti). The master alloy decomposes into Al3Ti + TiB2phases. The TiB2particles are ultra-fine (∼300–500 nm) which make them very suitable for grain nucleation, i.e., refining as shown in Figure 21(f). Exceeding the proper amount would lead to formation of high density of the TiB2particles which would restrict the fluid motion and hence create spongy porosity (see blue arrows showing a large pore surrounded by tiny ones). Another example is depicted in Figure 21(g)where Al2O3dispersoid particles were added to harden the alloy. During mechanical stirring, the aluminide particles tangled with the oxide films. If the Al2O3particles are well dispersed, they only form micro-porosity at the particle/matrix interface. Porosity formation only happens when the Al2O3particles are clustered/segregated together preventing the flow of the liquid metal as illustrated in Figure 21(h).
4. Conclusions
In this article, the different sources of the formation of porosity in Al-Si castings were discussed, mainly due to hydrogen level, high stress precipitation in the form of intercepted platelets or compounds, and the type of oxide films. The following conclusions have been drawn:(1)The β-Fe, π-Fe, and low Ti-rare earth intermetallics platelets are very effective in restricting the flow of molten metal leading to formation of shrinkage cavities, whereas α-Fe is comparably less active.(2)Precipitation of high temperature compounds such as Al2Si2Sr, MgSn, Al3Ti, or TiB2could also be sites for porosity formation in the form of pockets or spongy-like. The size of pores in this case depends on the number of segregated or clustered compound particles.(3)Oxide films (Al2O3, MgO, SrO) in the form of branched strings or bifilms are strong candidates for creating micro-porosity compared to those (macro-porosity) in 1 and 2.(4)Strontium oxides appear in the form of thin platelets, dispersed particles, or thin films.(5)Increasing the gas porosity balances a part of the volumetric change from liquid to solid state and thus the magnitude of shrinkage cavities, in particular in sand castings.(6)Alloy mechanical properties depend on the location of the pores in the casting and their sizes. Some of the fine pores tend to join together under the applied load to form coarser ones.
Data Availability
Data will be available upon request to the corresponding author.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
The authors would like to thank Amal Samuel for enhancing the quality of the artwork used in the present article.