In new casting development projects, the casting purchaser and the casting manufacturer come to some form of agreement that outlines the technical requirements for qualification to production. These arrangements may range from more comprehensive specific dimensional and metallurgical requirements associated with aerospace fixed processes to a lone AS9102 dimensional requirement, which may apply to an industrial part application. 

A gap in mutual understanding exists where the casting designer/purchaser may assume metallurgical requirements, including the notion that machined-from-casting (MFC) test bars are intended to and must meet ASTM requirements. Often, the designer/purchaser requires MFC, expecting ASTM properties. And many times, foundries agree to the requirement, because they think there is no other choice, and the designer must be correct. This can cost the casting producer in terms of excessive mechanical testing and lost production. 

The metallurgical differences of cast shapes are responsible for the inimitability (or uniqueness) of castings with different geometries. This includes test bars, because fundamentally, test bars are castings. Certain factors contribute to this inconsistency, but it may be possible to close the gap between test bar properties and MFC based on recent work. For now, a better approach for these discussions could be that separately-cast then subsequently-machined test bars are the preferred way to material assessment. Key to this discussion is that all cast shapes are inimitable, including test bars. Therefore, an appropriate mechanical testing plan should be agreed on at the beginning of each new casting project.

Background Basics

The first critical factor in making a sound casting is the essential solidification ratio. This is expressed in Figure 1 where the width “W” should not exceed B/2. This shows the need to gate wider at the top, for example. The part geometry, volume, and the surface area factor into how well this ratio can be met. The keel block on the right meets the essential solidification ratio. ASTM mechanical properties are developed using this ideal casting shape that provides good solidification, good grain, and no isolated shrink.

Figure 2 shows a keel block compared with a small casting. These two may not have the same metallurgical characteristics. This is not a sound comparison of the small casting, which does not meet the essential solidification ratio, and the keel block that does. Testing to ASTM standards is intended to verify the quality of the steel, not to establish actual casting properties. Each casting shape has different solidification ratios, different gating, different geometry size effect, different 
metallurgical size effect, and different heat effects. These factors lead to differing amounts of microporosity, different locations of shrinkage, dendrite arm spacing, grain size, and grain directionality that impact mechanical properties. The importance of interpreting tensile results is to review test bar data determining the quality of the steel from which the castings have been poured.

Castings have been affected by the rate of cooling post casting and during subsequent heat treatment, which are in turn influenced by casting thickness, size, and shape. Grain size and heat-treating impact mechanical properties in steel, for instance. As section size increases, grain size increases. Properties reduce. Grain size also affects fracture strength and fatigue. Heat treatment alters the as-cast microstructure. Thick sections and edge conditions provide different cooling, microstructure, and properties providing variation—surface to center. The single most important and least avoidable effect of section size is the coarseness of the microstructure. Cooling internally may not be rapid.

ASTM specifications A781 and A703 recognize that castings and test bars exhibit different properties. The mechanical test requirements are intended to verify the quality of the steel. They are not intended to establish actual casting properties that are impacted by solidification conditions.

Even with the essential solidification ratio being met, precise test bar machining and no “lab effect” impacting mechanical results, ASTM A370 observes intrinsic variation in an alloy’s tensile testing results. Even though every operation was done in the same laboratory and by the same technicians, the results were not the same. ASTM observed that foundries could see 95% assurance of being within 1,000 psi for ultimate tensile strength, within 95% assurance to be within 1,600 psi in yield strength, +/- 3% variation in elongation and +/- 5% in reduction of area. This level of variation can alone disqualify metal.

The non-equilibrium factors at work during solidification help explain why there is variation in tensile properties within the same casting shape. For example, if a casting freezes more quickly, grain size, grain directionality, and chemical segregation differences occur that might influence casting outcomes. Microporosity and dendrite arm spacing can be impacted. Metallurgical differences occur within controlled foundry conditions, too. With intrinsic variation occurring in property testing, mechanical test results can become less meaningful. The ASTM A370 specification goes so far as to describe basic riser conditions to establish consistency in addressing this problem.

Metallurgical effects ensure that each cast shape solidifies differently. In Figure 3, the casting, on the left, will experience different solidification than the thin-walled part on the right. Comparatively, the middle drawing is the shape that is used to cast then subsequently excise ASTM test bars from. Each of the three unique shapes has different solidification, different grain, different shrink, and unique-to-itself mechanical properties. Cast shapes are impacted by: (1) Geometrical size effect-dimensions; (2) Metallurgical size effect; 3) Mass effect-solidification and heat transfer in casting solidification and in heat treatment; and 4) Bi-film effect-Bi-films occur especially in castings with turbulent metal flow.
Figure 4 shows an ASTM E8 sub-sized test bar required by various customers. Compared with standard specimens that are several times larger, the keel block test bar is bigger than some customer-required test bars. The keel block test bar was used by ASTM to establish the mechanical properties we are working toward in specifications. Some customers require sub-sized MFC bars from a thin-walled casting,  that may be only capable of netting a sub-sized test bar; for example, with an 0.125-in. diameter. There is now the problem of not having established properties to meet or compare to. Different sized test bars going for the same mechanical result may not be realistic in an accept/reject scenario.

Experimental Work

Figure 5 and Table 1 show the intrinsic variation even in a cluster of parts. In this cluster, the top casting (A) exhibited lower properties than the bottom casting (B). The properties in the bottom casting were expected to be better, and they were. Different locations on a cluster can be likened to different ice cubes in an ice cube tray, freezing differently. Following the intrinsic variation test, additional functions were performed, including fracture morphology examination, metallography, texture corrosion mounts, X-ray, MPI, surface visual, and straightness.

A look into the testing effects brings out the importance of test bar condition. ASTM mechanical property specifications are established from a near-perfect casting shape and an ideally-gated keel block. The ASTM E8 test bar should come out of the soundest section near the lower portion of the keel block. The test bars should be precisely machined with no knicks, dents, or notches. The bars should also be non-destructive-tested for straightness. The metallurgical laboratory calls out precise and consistent settings with specific work instructions. Note, too, that mechanical testing machines generally have an as-manufactured margin of error of approximately +/- .5% accuracy.

Test Bar Variants

An array of test bar variations exists between casting purchasers and casting manufacturers. The impact of this may not be fully understood. Each variant can produce different results. When flowed down to the supplier, these requirements are often accepted. It’s essential to have mechanical property test plans fully understood before starting a new project. Separately-cast then subsequently-machined test bars for material assessment most closely align with the original ASTM procedure for making the bars. Following are examples of different test bar variants customers may use or casting producers may allow, which can lead to an array of different mechanical property results causing potentially unresolvable differences.

Test bar variant examples:
• 0.250 -in. diameter, cast to size (CTS) bars on a separately-cast mold, with 10 test bars per mold, top and bottom gated.
• 0.250-in. diameter, separately cast mold, with 10 test bars per mold, top and bottom gated, then subsequently machined.
• 0.250-in. diameter CTS, cast within a tree of parts, side gated, attached to the side of the sprue.
• 0.250-in. diameter, cast within a tree of parts, side gated, attached to side of the sprue, then subsequently machined.
• 0.125-in. diameter, CTS test bars on a separately cast mold with 10 test bars per mold, top and bottom gated.
• 0.125-in. diameter, separately cast mold with 10 test bars per mold, top and bottom gated, then subsequently machined.
• 0.125-in. diameter, CTS cast on tree of parts, side gated, attached to the side of the sprue.
• 0.125-in. diameter, cast within a tree of parts, side gated, attached to the side of the sprue, then subsequently machined.
• 0.125-in. diameter, excised from the centerline of a thick casting section (MFC).
• 0.125-in. diameter, excised from a thin section, at (1/4T) of a casting (MFC).
• Flat test bar excised from the centerline of a thick section of a casting (MFC).
• Flat test bar excised from a thin section at (1/4T) of a casting (MFC).

Shrink and Microporosity 

Centerline shrink is a casting certainty and impacts ultimate tensile properties. Differing casting shapes result in unique-to-that-shape centerline shrink and ultimate tensile properties. The surface of the casting impacts yield strength and the strongest part of the casting is at 1/4T, midway between the surface and the centerline. 

Customers use the terms shrink and shrinkage. These expressions have multiple meanings. Shrinkage is the decrease in volume that occurs during the transition from liquidus to solidus. The reason it occurs is because the specific volume of alloys is highest in the liquid state. All metals, except bismuth, have a higher density in a solid state. Therefore, castings undergo contraction as they cool. As castings cool from the outside inward, all castings will have microporosity, much of which is concentrated in the centerline region. As mentioned, different cast shapes, including test bars, will have differing microporosity impacting ultimate tensile.

Interdendritic shrink is shrinkage between the feather-like features (dendrite arms) that form when castings solidify. Dendrite arm spacing is important in this discussion because of its impact on casting properties. Another measure is secondary dendrite arm spacing (SDAS). As metal solidifies, the liquid reforms into “tree-like” shapes called dendrites. As the arms close, metal may not be filled in. There is usually about 0.004-in. dendrite arm spacing in a normal cooling profile. Temperature gradient and cooling rate are important in addition to some of the more common variables to control. These effects manifest differently in various cast geometries.

Figures 6, 7, and 8 are examples of microporosity typical in center sections. These examples are of automotive castings with microporosity shown at 100-times magnification. This testing was part of a study where a customer originally thought that 0.78% microporosity was a decent specification. That number changed as the project progressed and we showed them how to produce and properly assess excellent cast structures without intolerable specification call outs.

An approval that is meaningful to the product and economically attainable at the foundry, is needed for cast components. It’s typical to have an AS9102 dimensional document and a Production Part Approval Process (PPAP) for quality. Often, creating a metallurgical and material assessment plan is needed, as well. This can occur in advance of a new part kick-off. The material assessment is critical in this discussion.

Being well informed on the critical factors discussed here will help provide a path to successful material assessment.

Click here to view the column in the May/June Casting Source Digital Edition.