Cost effective casting design: what every foundryman and designer should know.

Viewing these six key factors as a system - while sketching geometries - provides a workable methodology for consistently good casting designs.

Structural design engineers who work successfully with castings commonly design in a narrow group of casting types poured from similar alloys (like the family of irons or the 300 series of aluminum) and molded from similar foundry processes (like green sand or nobake). Rules of thumb have been developed over the years for common design situations.

Close inspection of these rules reveals that they sometimes recommend conflicting geometry. For example, the use of gusseting instead of mass for stiffness might be labeled "recommended" in one set of design rules and "poor" in another.

Further, when a design engineer leaves a familiar casting design realm for an unfamiliar one, unexpected trouble may result. For example, let's say we are moving from ductile iron to aluminum bronze while staying in a familiar foundry process, nobake molding. No alarms are sounded among the "rules of thumb," but there's likely trouble in the usual "ductile iron-style" geometry. Good aluminum bronze geometry is different than typical ductile iron geometry, and the molding process may need to supplement the different geometry with heat transfer techniques. Not suspecting this, the design engineer's new casting design may suffer from "no-quotes," higher-than-expected prices and foundry requests for design changes.

How are design engineers supposed to know that successfully casting geometry for aluminium bronze should somehow be different? And if a design engineer did know that, what would be the proper course of design action?

The answer lies in a better understanding of the relationship among geometry, various foundry alloys and structure. As shown in Fig. 1, there are six parameters (based on physics) that underlie cost-effective casting design. Part I of this two-part series examines the four parameters that drive alloy castability. Part II will review the two parameters of mechanics for structure.

All six, applied as a system, drive the geometry of casting design. Geometry is not only the result of casting design but is also the most powerful weapon in creating successful casting design.

This six-faceted system is capable of optimizing geometry for castability, structure, downstream processing (machining and assembly) and process geometry (risering, gating, venting and heat transfer patterns) in the mold. The process geometry forms the casting geometry.

Quickly sorting through possible casting and process geometries by marking up blueprints or by making engineering sketches is the way to find optimal "system" geometry. An elegant result of good sketched brainstorming can be a solid model of the casting and its process geometry, the basis of rapid prototyping and/or computerized testing.

Applying the System

Optimizing casting geometry using the six-parameter system is not difficult. The six physical and mechanical characteristics in Fig. 1 influence important variables in designing, producing and using metal castings. These variables include:

* casting method;

* design of casting sections;

* design of junctions between casting sections;

* surface integrity;

* internal integrity;

* dimensional capability;

* cosmetic appearance.

Both the designer and metalcaster possess a vital ally to streamline any casting design. Casting geometry is the most powerful tool available to improve castability of the alloy and mechanical stiffness of the casting.

Carefully planned geometry can offset alloy problems in fluid life, solidification shrinkage, pouring temperature and slag/dross forming tendency. Section modulus, an attribute of structural geometry, has the capability to increase stiffness and/or reduce stress - a capability that can be very important when applied to alloys with lower strength and stiffness. Modulus of elasticity, an alloy's inherent stiffness, can be combined with section modulus and section length to limit or allow: deflection in a casting design.

To preview geometry's ability to influence the four physical characteristics of "castability," consider the simple steel fabrication in Fig. 2a that was converted into carbon steel and gray iron casting designs, Figs. 2b and 2c, respectively.

The fabrication is a guide block to constrain low velocity/low toad sliding motion, and it was welded from rectangular bar stock, subsequently milled, drilled and tapped. The geometries in 2b and 2c are considerably different as a consequence of differences among fluid life, solidification shrinkage type and amount, pouring temperature and tendency to form nonmetallic inclusions (See Junctions, [ILLUSTRATION FOR FIGURE 6 OMITTED]).

1. Fluid Life

Fluid life more accurately defines the alloy's liquid characteristics than does the traditional term "fluidity." Molten metal's fluidity is a dynamic property, changing as the alloy is delivered from a pouring ladle, die casting chamber, etc. into a gating system and finally into the mold or die cavity. Heat transfer reduces the metal's temperature, and oxide films form on the metal front as this occurs. Fluidity decreases most rapidly with temperature loss, and it can decrease significantly from the surface tension of oxide films.

The absolute value of temperature is not the test of fluidity at a given moment. For example, some aluminum alloys at 1200-1400F (650-750c) have excellent fluid life. However, some molten steels at 3000F (1650c) have much shorter fluid life. In other words, a molten alloy's fluid life also depends on chemical, metallurgical and surface tension factors.

Fluid life affects the design characteristics of a casting, such as the minimum section thickness that can be cast reliably, the maximum length of a thin section, the fineness of cosmetic detail (like lettering and logos) and the accuracy with which the alloy fills the mold extremities.

It is essential to understand that moderate or even poor fluid life does not limit the cost-effectiveness of design. Knowing that an alloy has limited fluid life tells the designer that the part should feature:

* softer shapes and larger lettering;

* finer detail in the bottom portion of the mold, where metal arrives first, fastest and generally hottest;

* coarser detail in the upper portions of the mold where the metal is slower to arrive and more affected by oxide films and solidification "skin" formation. Even an alloy with good fluidity, when overexposed to oxygen, may form a high surface tension oxide film that makes the fluidity die, "rounding off" of the leading edges as it flows.

* more taper toward thin sections.

Some alloys, like 356 aluminum, have been specifically designed metallurgically to enhance fluid life. In the case of 356, the high heat capacity of silicon atoms "revive" aluminum atoms as their fluid life begins to wane.

[TABULAR DATA FOR TABLE 1 OMITTED]

2. Solidification Shrinkage

There are three distinct stages of shrinkage as molten metals solidify: liquid shrinkage, liquid-to-solid shrinkage and patternmaker's contraction.

1. Liquid shrinkage is the contraction of the liquid before solidification begins. It is not an important design consideration.

2. Liquid-to-solid shrinkage is the shrinkage of the metal mass as it transforms from the liquid's disconnected atoms and molecules into the structured building blocks of solid metal. The amount of solidification shrinkage varies greatly from alloy to alloy. Table 1 provides a guide to the liquid-to-solid shrinkage of common alloys. As shown, shrinkage can vary from low to high shrinkage volumes.

Alloys are further classified based on their solidification type: directional, eutectic-type and equiaxed (see definitions in Table 1). The type of solidification shrinkage in a casting is just as important as the amount of shrinkage. Specific types of geometry can be chosen to control internal integrity when solidification amount or types are a problem.

Figures 3-5 illustrate what is implied by the three solidification shrinkage types defined in Table 1. In each case, a simple plate casting is shown with attached risering (a "riser" is a reservoir of liquid metal attached to a casting section to feed solidification shrinkage). Cross sections of the plate and riser(s) show conceptually how solidification takes place; metallurgical reality is similar, but microscopic.

Figure 3 shows solidification on and perpendicular to the casting surfaces, known as "progressive" solidification. At the same time, solidification moves at a faster rate from the ends of the section(s) toward the source of feed metal (risers) - this is known as directional solidification. Directional solidification moves faster from the ends of the sections because of the greater amount of surface area through which the solidifying metal can lose its heat. The objective is for directional solidification to beat out progressive solidification before it can "close the door" to the source of the feed metal. As shown, directionally solidifying alloys require extensive risering and tapering, but they also have the capability for excellent internal soundness when solidification patterns are designed properly.

Figure 4 illustrates the eutectic-type alloy, the most forgiving of the three. Such alloys typically have less solidification shrinkage volume. Risers are much smaller, and in special cases can be eliminated by strategically placed gates. The key feature with these alloys is the extended time that the metal feed avenue stays open. The plate solidifies more uniformly all over and all at once, similar to eutectic solidification. Eutectic-type alloys are less sensitive to shrinkage problems from abrupt geometry changes.

Alloys that exhibit equiaxed solidification respond the most dramatically to differences in geometry [ILLUSTRATION FOR FIGURE 5 OMITTED]. Shrinkage in these alloys tends to be widely distributed as micropores, typically along the center plane of a casting section. The reason is that solidification occurs not only progressively from casting surfaces inward and directionally from high surface area extremities toward lower surface area sections, but also equiaxially via "islands" in the middle of the liquid. These islands of solidification interrupt the liquid pathway of directional solidification. Gradually, the pathways freeze off, leaving micropores of shrinkage around and behind the islands that grew in the middle of the pathway. Larger risers, thicker sections and tapering (shown at center of [ILLUSTRATION FOR FIGURE 5 OMITTED]) are counterproductive, causing micropores to coalesce into larger pores across more of the casting cross section. As illustrated at the bottom of Fig. 5, microporosity is kept small and confined to a narrow mid-plane in the casting section by more "thermally neutral" geometry with smaller, further-spaced risers.

As illustrated in Fig. 3-5, there is a significant bilateral and reciprocal relationship between solidification shrinkage and geometry. Most simply, eutectic-type solidification is tolerant of a wide variety of geometries; the least reciprocity is required. Most complex, equiaxed solidification requires the most engineering foresight in the choice of geometry and may require supplemental heat transfer techniques in the mold process. In the middle lies directional solidification, while capable of the worst shrinkage cavities, it is the most capable of very high internal integrity when the geometry is properly designed. Well-planned geometry in a directionally solidifying alloy can eliminate not only shrinkage but the need for any supplemental heat transfer techniques in the mold.

In fact, the real mechanism behind the bilateral and reciprocal relationship between solidification shrinkage and geometry is heat transfer. All three modes of heat transfer, radiation, conduction and convection are involved in solidification of castings, and all three depend on geometry for transfer efficiency. Convection and conduction, are very important in casting solidification, and transfer rates are highly affected by geometry.

3. Patternmaker's Contraction is the contraction that occurs after the metal has completely solidified and is cooling to ambient temperature. This contraction changes the dimensions of the casting from those of liquid in the mold to those dictated by the alloy's rate of contraction. So, as the solid casting shrinks away from the mold walls, it assumes final dimensions that must be predicted by the pattern- or diemaker. This variability of contraction is another important casting design consideration, and it is critical to dimensional accuracy. Tooling design and construction must compensate for it.

Achieving dimensions that are "just like the blueprint" require the foundry's pattern- and/or diemaker to be included. The unpredictable nature of patternmaker's contraction makes tooling adjustments inevitable. For example, a highly recommended practice for critical dimensions and tolerances is to build the patterns/dies/coreboxes with extra material on critical surfaces so that the dimensions can be fine-tuned by removing small amounts of tooling stock after capability castings have been made and measured.

3. Slag/Dross Formation

Among foundrymen, the terms slag and dross have slightly different meanings. Slag typically refers to high-temperature fluxing of refractory linings of furnaces/ladies and oxidation products from alloying. Dross typically refers to oxidation or reoxidation products in liquid metal from reaction with air during melting or pouring, and can be associated with either high or low pouring temperature alloys.

Some molten metal alloys generate more slag/dross than others and are more prone to contain small, round-shaped nonmetallic inclusions trapped in the casting. Unless a specific application is exceedingly critical, a few small rounded inclusions will not affect casting structure significantly. In most commercial applications, nonmetallic inclusions are only a problem if they are encountered during machining or appear in a functional as-cast cosmetic surface.

The best defense against nonmetallic inclusions is to inhibit their formation through good melting, ladling, pouring and gating practices. Ceramic filters, which can be used with alloys that have good fluid life, have advanced the foundry's ability to eliminate nonmetallics. Vacuum melting and pouring are applied in extremely drossprone alloys, like titanium.

4. Pouring Temperature

Even though molds must withstand extremely high temperatures of liquid metals, interestingly, there are not many choices of materials with refractory characteristics. When pouring temperature approaches a mold material refractory limit, the heat transfer patterns of the casting geometry become important.

Sand and ceramic materials with refractory limits of 3000-3300F (1650-1820c) are the most common mold materials. Metal molds, such as those used in diecasting and permanent molding, have temperature limitations. Except for special thin designs, all alloys that have pouring temperatures above 2150F (1180c) are beyond the refractory capability of metal molds.

It's also important to recognize the difference between heat and temperature; temperature is the measure of heat concentration. Lower temperature alloys also can pose problems if heat is too concentrated in a small area - better geometry choices allow heat to disperse into the mold.

Design of Junctions

A junction is a region in which different section shapes come together within an overall casting geometry. Simply stated, junctions are the intersection of two or more casting sections. Figure 6 illustrates both "L" and "T" junctions among the four junction types, which also include "X" and "Y" designs.

Designing junctions is the first step to finding castable geometry via the six-faceted system for casting design. Figure 6 illustrates that there are major differences in allowable junction geometry, depending on alloy shrinkage amount and pouring temperature. Alloy 1 allows abrupt section changes and tight geometry, while alloy 3 requires considerable adjustment of junction geometry, such as radiusing, spacing, dimpling and feeding. Figure 7 illustrates a very high form of the foregoing principles in a critical automotive application.

Considerations of Secondary Operations in Design

System-wide thinking also must include the secondary operations, such as machining, welding and joining, heat treating, painting and plating.

One aspect that affects geometry is the use of fixturing to hold the casting during machining. Frequently, the engineers who design machining fixtures for castings are not consulted by either the design engineer or the foundry engineer as a new casting geometry is being developed. Failure to do so can be a significant oversight that adds machining costs. If the casting geometry has been based on the four physical characteristics of the alloy, then the designer knows the likely surfaces for riser contacts and may have some idea of likely parting lines and core match lines. These surfaces and lines will be irregularities on the casting geometry and will cause problems if they contact fixturing targets.

It is best to define the casting dimensional datums as the significant installation surfaces, in order of function priority, based on how the casting is actually used. Targets for machining fixtures should be consistent with these datum principles.

There is nothing more significant in successful CNC and transfer line machining of castings than the religious application of these datum fixture and targeting principles.

Drawings and Dimensions

The tool that has had the most dramatic positive impact on the manufacture of parts that reliably fit together is geometric dimensioning and tolerancing (GD&T), as defined by ANSI Y14.5M - 1994. When compared to traditional (coordinate) methods, GD&T:

* considers tolerances, feature-by-feature;

* minimizes the use of the "title block" tolerances and maximizes the application of tolerances specific to the requirement of the feature and its function;

* is a contract for inspection, rather than a recipe for manufacture. In other words, GD&T specifies the tolerances required feature-by-feature in a way that does not specify or suggest how the feature should be manufactured. This allows casting processes to be applied more creatively, often reducing costs compared to other modes of manufacture, as well as finish machining costs.

GD&T encourages the manufacturer to be creative in complying with the drawing's dimensional specifications because the issue is compliance with tolerance, not, necessarily compliance with a manufacturing method. By forcing the designer to consider tolerances feature-by-feature, GD&T often results in broader tolerances in some features, which opens up consideration of lower cost manufacturing methods, like castings. Figure 8 illustrates GD&T principles applied to a design made as a casting. Note the use of installation surfaces as datums and the use of geometric zones of tolerance.

Factors that Control Casting Tolerances

How a cast feature is formed in a mold has a significant effect on the feature's tolerance capability. The following six parameters control the tolerance capability of castings. In order of preference, they are:

Molding Process - The type of molding process (such as green sand, shell, investment, etc.) has the greatest single influence on tolerance capability. How a given molding process is mechanized and the sophistication of its pattern or die equipment can refine or coarsen its base tolerance capability.

Casting Weight and Longest Dimension - Logically, heavier castings with longer overall dimensions require more tolerance. These two parameters have been defined statistically in tolerance tables for some alloy families.

Mold Degrees of Freedom - This parameter is least understood. Just as some molding processes have more mold components (mold halves, cores, loose pieces, chills, etc.) than others, some casting designs require more mold components. Each mold component has its own tolerances, and tolerances are stacked as the mold is assembled. More mold components means more degrees of freedom; hence more tolerance. Good design for tolerance capability minimizes degrees of freedom in the mold for features with critical dimensions.

Draft - It is common for casting designs to ignore the certainty of draft, including mold draft, draft on wax and/or styrofoam patterns made from dies, and core draft. Since 1 [degree] of draft angle generates 0.017 in. of offset per in. of draw (about 0.5 mm/30 mm), draft can quickly use up all of a tolerance zone and more.

Patternmaker's Contraction - The uncertainty of patternmaker's contraction is why foundrymen normally recommend producing first article and production process verification castings (sometimes called "sample" or "capability" castings) to establish what the dimensions really will be before going into production. A common consequence of patternmaker's contraction uncertainty is a casting dimension that is out of tolerance, not because it varies too much, but because its average value is too far from nominal. In other words, the dimension contracted more or less than was expected.

Cleaning and Heat Treating - Many casting dimensions are touched by downstream processing. At the least, most castings are touched by abrasive cutting wheels and grinding - even precision castings. Many castings are heat-treated, which can affect straightness and flatness.

When considering the breadth and depth of geometry's importance in casting design, from its influence on castability, the geometry of gating/risering, structural form, cosmetic appearances and downstream fixturing, extensive brainstorming of geometry is highly recommended. The standard for "optimal" casting geometry is high, but the possibilities for geometry are limitless. Find ways of exploring geometry quickly, such as engineering sketching, before committing to a print or solid model.