Calming the turbulence in your runner system.
Aluminum oxide contamination in aluminum casting alloys has long been an area of major concern to aluminum casting facilities. The presence of oxides as films or discrete folded particles can reduce liquid metal fluidity, resulting in casting filling issues, such as misruns or short pours. Of even greater concern is the detrimental effect aluminum oxides have on the resultant casting's mechanical properties, such as yield strength, tensile strength and elongation, and the contribution toward a reduction in casting pressure tightness.
Because of the high reactivity of liquid aluminum with oxygen, no matter how clean the metal just prior to pouring, significant damage to the metal still can occur. Most aluminum metalcasting facilities must constantly fight these oxides. The increasing acceptance of in-mold filtration has added another weapon to this battle; however, most firms have found that filtration alone is not sufficient to solve their lingering oxide contamination problems.
Unfortunately, what most aluminum plants miss is the total relationship that exists in the runner system. While most aluminum casters agree that an unpressurized runner system is preferable, they typically only look at the relationship of cross-sectional areas between the sprue choke, the runner and the ingate. Two major mistakes are often made.
First, if a foam filter is used, there is inadequate compensation (enlargement of the area into and out of the filter) to ensure that the filter will not be a choke in the system. Second, little, if any, consideration is given to runner geometry.
Going with the Flow
Regardless of orientation, runner and gating systems usually have a pouring cup or basin, tapered sprue (with or without a well), choke, runner (with or without a runner extension and typically in the drag) and one or more ingates (typically in the cope) (Fig. 1).
[FIGURE 1 OMITTED]
All of the calculations developed over the years to determine the filling speed of the mold have concentrated primarily on the relationship and ratios of the cross-sectional area of the system. For aluminum castings, the preferred relationship is one where the metal velocity is continuously and progressively decreasing so as to minimize turbulence and metal splashing and therefore reoxidation. The key areas would be at the bottom of the sprue, the runner and the ingates.
A typical proportion of cross sectional areas would be something like 1:4:4.1--the runner cross sectional area is four times the area of the choke, and the total area of the gates is slightly larger than the area of the runner. In pouring configurations where the smallest cross sectional area cannot be at the bottom of the sprue, a choke can be placed immediately after the transition between the sprue well and the start of the runner.
In addition to the issue of metal velocity in mold filling, air entrainment often is not given as much thought as it should, especially early in the filling process before a steady state is established. Most often, it is assumed that after the choke, metal flowing into the runner completely fills the runner as it progresses toward the first gate. In actuality, most runner systems allow for significant air bubble entrapment (Figs. 2-4). The air bubbles are gradually released into the mold cavity later in the pour than what would be assumed or expected, typically when the mold is close to half full or more. These air bubbles contain the same ingredients for continuing aluminum oxidation ([O.sub.2] and [H.sub.2]O) in the mold before solidification and, like a balloon with a trailing string, the bubbles continuously slough off oxide strings and films as they move through the mold cavity (Fig. 5).
[FIGURES 2-5 OMITTED]
All air bubbles leave behind some sort of oxide trail. Some exit the metal before solidification. Some become trapped just below the solidifying skin where they can be confused with hydrogen gas, subsurface shrinkage or other general porosity related issues. Others can become trapped within the body of the casting or in and around a core.
In any case, the double impact of the potential hole or void, coupled with the oxide trail, leads to a decrease in potential casting quality. Increasingly, castings that may have been accepted in the past with such small internal defects are no longer acceptable to the casting customer. This has created a situation where closer examination of the root cause of these kinds of defects and possible solutions has generated renewed interest in runner and gating system design.
Runner geometry is typically dictated more by available pattern or mold real estate or by past running programs than on the actual influence the geometry has on mold filling behavior. For automotive castings converted from iron, the tendency for many aluminum casting plants had been to keep the same general runner and gating configuration as the iron casting with only minor adjustments.
Conventional wisdom for iron casting has been to use tall, narrow runner systems to allow iron slags to float to the top of the runner and stick to the upper runner surface. However, aluminum oxides are unlikely to easily float to the top of the runner. Still, most firms tend to run as close to the limits on pattern or mold space as possible, which has reinforced the use of tall, narrow runners out of convenience.
Since these iron-based runner systems were the most common and familiar systems during the time when many automotive castings were converted to aluminum, their basic configurations and geometries were looked upon as normal and became the basis for the design of future runner systems. A tall, narrow runner system minimizes the room on the pattern plate or in the permanent mold to locate the runner, potentially reducing equipment and molding costs. However, the effect of these runner system designs on the behavior of flowing liquid metal, and thus on the ultimate quality of the produced casting, is missed.
Modeling Metal Flow
A casting process modeling program was initiated to investigate the effect of runner design on metal flow turbulence and air entrainment in a horizontal sand runner and gating design. The "casting" was a simple plate casting with a cope ingate. A flow-modifying foam filter was used at the base of the sprue to act as a barrier against any initial air being carried over into the runner during early filling of the sprue. This provided a situation in the model where any air subsequently trapped in the advancing metal would have to come from the runner itself.
Because many of the tall, narrow runner systems have an aspect ratio of 2:1 or greater, the first orientation investigated was that of a runner with an almost 3:1 aspect ratio (Fig. 6). The modeling showed that metal preferentially runs along the bottom of the runner at relatively high velocity (150 cm/second) and that there is a substantial air gap above the flow of metal toward the gate. At the end of the runner and immediately below the ingate, a significant fold of metal, which would result in aluminum oxide entrapment, was seen (Fig. 7). This swirl of oxide-laden metal was the first metal into the ingate.
[FIGURES 6-7 OMITTED]
The swirl of metal at the end of the runner then began to push metal back toward the sprue, as well as into the ingate, at a high velocity. The trapped air at the top of the runner was compressed back toward the sprue (Fig. 8).
[FIGURE 8 OMITTED]
The velocity of the metal moving back along the top of the runner towards the sprue began to increase, further compressing the entrapped air. As the runner system became completely full, the multi-directional flow equalized and became unidirectional toward the ingate. By this time, the filling was 20% full and the air that was trapped in the runner was carried into the mold cavity with the advancing metal.
Calming the Waves
A series of aspect ratio reductions was then investigated in relation to the effect on liquid flow behavior. The cross sectional area was kept the same, so the overall metal velocities and mold filling speeds were equal.
Only minor improvement in flow behavior resulted from reducing the aspect ratio of the runner to 2:1. The metal flow initially was still along the bottom of the runner at relatively high velocity. A significant amount of metal still was swirling at the ingate area, causing the metal to fold over itself and entrap oxide skins. Likewise, the potential for a large amount of entrapped air in the top portion of the runner still remained.
A runner with a 1:1 aspect ratio showed some reduction in metal velocity, but the tendency still existed for the metal to fold over itself at the area of the ingate as well as for a small pocket of air to be entrapped. A lag in time before the runner was completely full, combined with the metal folding and trapped air, could provide the opportunity for reoxidation and the transport of air bubbles into the mold cavity early in the pour rather than later.
When the aspect ratio was reduced further to 1:2, a marked improvement in metal flow could be seen. The overall metal velocity was reduced to less than 100 cm/second, and the runner was completely full side-to-side and top-to-bottom with unidirectional flow only toward the ingate. When the metal met the end of the runner and began to move up into the gate, the remaining air was moved into the gate area before the metal rather than mixed in with it (Fig. 9). When the metal completely filled the gate and just began to enter the mold cavity, there was no sign of turbulence or entrapped air.
[FIGURE 9 OMITTED]
Inside This Story
* The presence of aluminum oxide as films or discrete folded particles can reduce liquid metal fluidity, resulting in casting filling issues, such as misruns or short pours.
* Filtration alone cannot sufficiently resolve oxide contamination problems, and most metalcasting facilities miss the importance of the total relationship that exists in the runner and gating system.
* Metal flow simulation has shown that shorter and wider runners in aluminum casting production can help decrease turbulent metal flow and reduce the regeneration of oxides post filtration.
For More Information
"Observations on the Effect of Gating Design on Metal Flow and Defect Formation in Aluminum Lost Foam Castings: Part 1," S. Bennet, M. Tschopp, A. Vrieze, E. Zelkovich, C.W. Ramsay, and D.R. Askeland, 2001 AFS Transactions (01-056).
Robert Pischel is the aluminum filtration applications manager for Foseco Metallurgical Inc., Cleveland.