This article sets forth principles for casting design geometry improvements that enable robust solutions for variation in solidification integrity or microstructure quality sometimes found in first article preparation, pre-production capability assessment, or later in-production sampling. In other words, after a metalcasting producer has been chosen, well after the casting design geometry has been finalized. 

These design geometry improvements are casting producer initiated “tweaks” to the design geometry in very specific local regions, not wholesale redesigns. They are relatively easy to accomplish in design, relatively minor in tooling changes, and not likely to upset existing machined or assembly surfaces. 

There are two scenarios in which a casting producer would seek approval for a design geometry tweak from the customer’s design authority:

1.    Improved solidification integrity. These changes typically involve casting interior thicknesses that don’t solidify to specification without the use of chills, chill sands, or correction with in-process weld rework. Such process additions are less certain for consistent solidification integrity than geometry improvement. The customer design authority should welcome these geometry tweak suggestions for mutual solidification integrity benefit. 

2.    Variation in microstructure quality. In some critical service structural casting applications, the alloy’s required mechanical properties are highly dependent on defined microstructure attributes. These casting alloys are the ones whose mechanical properties are dependent on complex melting, melting dynamics, holding, and/or pouring metallurgical adjustments. Specifics can be various forms of stirring, de-gassing, addition and distribution of ceramic particles, micro-alloying, treatments, modification, and innoculation. These are typical, in all or in part, of the high-strength aluminums and ductile iron.

Typical Specification Requirements

Among high-specification aluminum alloys, requirements involve sampling during the melting and pouring processes, as well as in coupons in the mold with the castings. The samples are evaluated with metallography, looking at microstructure attributes such as gas pore size among the dendrite arms, secondary dendrite arm spacing, grain size, the compounds in intermetallic phases, and the size and shape of the primary alloy phases. These are metallurgical quality variables that underlie the specified mechanical properties, especially elongation percent for aluminum alloys.

Among high specification ductile irons, microstructure requirements involve the proportion of ferrite and pearlite phases, the nodularity percentage, and the nodule count. These microstructure variables drive mechanical properties, especially the allowable transformed shear stress1 for yield and fatigue limit. Sampling is done at the melting furnace, holding furnace, and in coupons in the mold with the castings.

The quality assurance function monitors the sampling results, and variation on the low side of microstructure requirements results in melted alloy poured into pig molds instead of casting molds and/or rejected castings.

Consequences of Challenging Microstructure Specifications

Since the microstructure quality required in such high specification castings is difficult to achieve, there is drifting toward or below the specification minimums. The concern is a corresponding downward drift in mechanical properties, which could shorten the strain life of aluminum alloy castings or shorten the fatigue life of ductile iron castings. 

A More Robust Alternative

There is a more robust way to mitigate the consequences of microstructure quality variation. This alternative is 100% reliable without sampling, and it reduces the casting producer’s cost rather than adding to it. 

The idea is to reduce the transformed shear stress (von Mises stress) at the location of the highest load case forces and bending moment on the casting. It is an idea that needs to be initiated by the casting producer because the customer design authority, from the outside looking in, has no driving force to want to change the casting’s geometry.

Obviously, a casting design geometry tweak has to be a collaboration between the casting producer and the casting’s design authority, but as we will see, the tweak is relatively easy to conceive and implement in the casting design. 

Usually, the existing production tooling can be modified to incorporate the geometry tweak at reasonable cost and not involve building new tooling. With today’s powerful stress and strain simulation software, the design authority can be assured that the geometry tweak will be successful in lowering the Von Mises stress or total strain amplitude sufficiently to sustain the intended, and previously tested, cyclic life. Strain life or fatigue life can also be simulated with high confidence levels.

Examples of How It Works

The details of a successful casting geometry tweak lie in the author’s methodology for overlaying casting structural geometry on a foundation of “castability geometry.” In a sound bite, good castability geometry is a mold cavity shape that the chosen casting alloy “likes” to flow into and solidify with high integrity all by itself. The spectrum of metalcasting alloys vary widely in what they “like,” but knowing that and designing for that is readily possible. The structural geometry tweak we are talking about has to be compatible with the temperature distribution of the alloy as it arrives in the mold, and the temperature and mass gradients that are set up in gating and risering to solidify with high integrity.

There is an attribute of structural cross-sections of any kind that is especially powerful in casting structural design. It is a cross-section that has high stiffness against Von Mises stress and total strain amplitude. That cross-section also has a large perimeter that accommodates the heat transfer necessary for casting solidification integrity. Its name is Area Moment of Inertia, and common versions are the I-Beam (against bending) and the Torque Tube (against twist.)  One especially powerful in-casting structural design is the Omega because it is stiff against both bending and twist, and the Omega has natural draft, handy for mold cavity design. The Omega is illustrated in Figures 1 and 2.

Our first example is a A356-T6 cast aluminum gear case. It is fairly large with a bounding box of 34 x 34 x 30 in. (864 x 864 x 762 mm). The mechanical properties specification comes from AMS-A-21180 for tensile test coupons excised from designated areas in the casting. The designated areas are in the center deck of the gear case where the load case from torque transmission is highest.
The design intent of the specification is to provide strength and toughness in the center deck, but aluminum gear cases need resistance to deflection more than stress. It is deflection in the center deck that will cause premature wear in shafts and bearings. 
Regardless of mechanical properties, aluminum itself has poor stiffness. Its Modulus of Elasticity is only one-third of the steels and only 40% of ductile iron. However, Area Moment of Inertia is also in the equations for deflection, and Area Moment of Inertia can easily provide stiffness from cross-section geometry that aluminum can’t do by itself.

Shown in Figure 4’s sketch is the conversion of the center deck geometry for improved castability and, most importantly, stiffness from Omega cross-sections where deflection should be resisted the most.

Affirming the earlier point about the cost-effectiveness of geometry tweaks, these geometry changes can be readily accomplished with redesign of the coreboxes and/or printed cores. No other tooling changes would be necessary.

Our concluding example is a ductile iron brake caliper for the front wheels of a heavy duty pickup truck. The design is generic, but it is similar in the way it works with this Meritor brake caliper, made in two halves. The piston side of the two halves is our design. The ductile iron alloy is Grade 80-55-06, and the microstructure specification is ferritic/pearlitic with 90% nodularity minimum and a nodule count of 100 nodules per square millimeter at 100X magnification. As Figure 6 shows, the nodularity percent directly affects the cyclic life of a ductile iron casting design.

From a casting producer standpoint, the microstructure requirements for nodularity and nodule count are challenging because the control of nodule shape and nodule count fades continually with oxidation in the holding furnace. 

Nodularity and nodule count can be touched up as the holding furnace fades by pouring molds with in-mold innoculation. Those molds contain magnesium in the gating system, and they have to be set up in the pouring line, in advance, anticipating when they might be needed to touch up fading nodularity in the holding furnace. So, sustaining 90% nodularity and 100 nodule count is complex metallurgically and logistically. Reality is that a significant amount of the holding furnace iron will have to be poured into pig molds rather than casting molds. That is a costly consequence for the producer.

The original design geometry is shown in Figure 7 with the Von Mises stress at the highest bending moment from brake fluid pressure. The Von Mises stress is below the allowable for a fatigue life of 1 million cycles as shown by the “all blue” fatigue life simulation.

Figure 8 is a cutaway view of the geometry tweak to lower the Von Mises stress at that high-bending moment location. The geometry tweak allows the nodularity percent and nodule count specification to be lowered without compromising the intended fatigue life of the casting.

The radius between the two piston bore features has been increased with no interference with surrounding components. The cavity inside the brake caliper is a reservoir that shares brake fluid from the brake line and the two piston bores. That cavity is a place where stiffening geometry can be added with no effect on surrounding components or the hydraulic performance. 

The top of the upside-down Omega and the thicker radius section adds more “Y2” stiffness to the Area Moment of Inertia equation, which, in stiffness jargon is called “depth of section.”  So, the result is a new, stiff box section that can be formed with a small change to the pattern for the radius change and a printed core for the piston bores and center cavity feature. 

The Von Mises stress reduction at the highest bending moment region has been reduced by 20%, far more than the 3% allowable Von Mises stress reduction from a drop in nodularity from 90% to 85%. A nodularity drop from 90% to 80% would only be a 5% reduction in allowable Von Mises stress for the same 1 million cycle fatigue life.

As shown in Figure 9 above, if this were an aluminum strain-driven cyclic life, the stiffness increase against strain from the geometry tweak was even more at 30%.

Conclusion

It is the casting producer who can and should initiate casting design geometry tweaks for the mutual benefit of the producer and the customer design authority. There are two scenarios for producer-initiated design geometry tweaks. 

]First, tweaks for improved solidification integrity: These are common requests and not always considered by the customer design authority. But they should be. Improved solidification integrity is a benefit to the casting’s durability. 

Second, tweaks for relaxing tight microstructure quality specifications: High confidence comes from reduction of transformed shear stress or total strain amplitude at the highest load case regions of the casting. Either reduction is the consequence of improved structural geometry that is 100% reliable casting after casting. As illustrated by examples in aluminum and ductile iron castings, a collaboration between the customer design authority and the geometry change suggestion from the producer is relatively easy to accomplish, involves minimal modification of tooling, is mutually cost effective, and reliable.