Design of Structural Aluminum Castings
Excellence in casting design and manufacturing engineering will create a high value structural component at reasonable cost. Castings offer the designer incredible flexibility and many avenues to use that geometric flexibility to produce stronger and stiffer components. However, mechanical properties can vary throughout a casting based upon the temperature distribution and solidification rates throughout the mold cavity. The resulting microstructure differences can affect the structural performance of casting, especially in cyclic life, depending on the location within the casting where the microstructural differences occur.
Solidification integrity, part cost, volume of parts and part size govern the casting process selection. The most common casting processes are sand, permanent mold, and die casting. Tooling cost is very high for high-pressure die casting (HPDC), with permanent molds approximately 50% less. Sand casting is the preferred process for low quantity and very large castings. A very fine grain structure and, hence, relatively high mechanical properties can be expected in permanent mold and die casting. However, those processes are more prone to feeding defects that may outweigh this advantage in some instances. The production and use of castings, particularly those needing to perform under high loads, is a shared responsibility of the casting designer and the foundry. The implementation of the design by the foundry is critical to the part success, and certain attributes of the design can be critical to the success of the foundry.
The manufacturing engineering phase of a metalcasting design is the link between component shape, physical and mechanical properties, and the physics of converting liquid metal into solid metal. Success with aluminum structural castings engages the interrelation between the aluminum alloy’s castability characteristics, the design geometry’s influence on the alloy’s castability, the choice of mold cavity-making process, and the design of the liquid metal delivery system…all combined so that the temperature and mass gradients result in specified mechanical properties in the right locations to enable robust structural component performance.
It is the original equipment manufacturer (OEM) design engineer who defines the structural shape, the required mechanical properties in designated locations of the aluminum casting geometry, important functional physical properties (i.e., corrosion resistance), and the dimensional tolerances required. It is the integration of all the previous design and manufacturing engineering, executed in forethought, that enables the spectacularly successful aluminum structural castings currently in the most challenging applications.
More specifically, the integration of design and manufacturing engineering of aluminum structural castings has many facets that include:
- The aluminum alloy’s preference for a geometry type that enhances its ability to solidify within the specified microstructural integrity;
- Reconciliation, via iterations of design geometry and the resulting solidification integrity, of the specific aluminum alloy chosen and the structural geometry required; and
- Development of casting-friendly approaches to meet the structural geometry specifications.
- Casting design is optimized by examining the geometry of what is to be cast through four different geometric concepts, and two key material mechanics concepts:
- Structural geometry (engineering mechanics of the stress-resisting stiffness, coupled with the alloy’s stiffness from modulus of elasticity).
- Castability geometry (shape that the chosen alloy “likes” for enabling solidification integrity, especially the shape of junctions inside the mold cavity space, away from the mold cavity walls).
- Process interface geometry (gating, risering, chilling, and venting connections to the mold cavity surfaces).
- Downstream processing geometry (features of the casting’s surface geometry specifically intended to interface with casting finishing functional gauges, machining fixtures, and/or assembly fixtures, any process step after the casting is poured including heat treatment, machining and assembly).
- Crack initiation.
- Crack growth.
This article explores the four geometric concepts.
Most casting designs begin by considering the structural geometry of the part. The aim of structural design is to meet the strength requirements of the part and to prevent fracture by eliminating (as much as possible) structural details that act as stress raisers and can be potential fracture initiation sites.
Specific to aluminum casting structural design, most component requirements involve low mass and long cyclic life. The capability of aluminum’s low density must be supplemented with area moment of inertia to develop a geometry that is stiff, adds the least mass, and simultaneously protects the aluminum from plastic strain.
Since all aluminum alloys have a modulus of elasticity = ~10 x 106 lb/in2 [68,900 MPa], compared to steel = 30 x 106 lb/in2 [206,800 MPa] or ductile iron = 24 x 106 lb/in2 [165,400 MPa], geometric provisions are often made using ribbing or increased section thickness where greater stiffness is required. Further, aluminum alloys have an elastic elongation = 5 to 5% compared to common low alloy steels = 20%, and common ductile iron = 12%. Area moment inertia is the enabler for cast aluminum structural geometry that can locally protect the structure from plastic strain. Only castings can enable that local strain protection so easily.
There will be multiple considerations depending on how the part will be used in service. A casting design may begin by understanding the typical stresses the casting will be subjected to and compared with the yield strength of the alloys that are under consideration. Some additional safety factors may be applied. Typical casting alloys have yield strengths between 100 and 350 MPa,1 with those strengths varying minimally with casting quality. The casting must possess enough ductility to yield locally and redistribute the load in the vicinity of the stress raisers. Alloy selection is based on many factors that may be out of the designer’s control including cost and supplier capabilities. Due to the characteristics of aluminum casting alloys, some design changes may be required to accommodate castability of the alloy chosen. This design loop is iterative, with castability dependent on design and design dependent on castability.
If stresses in a part are cyclical, fatigue strength may be the critical design criteria. Aluminum castings typically do not have a true fatigue limit, but there is good data available for most casting alloys for an endurance limit at 500 million cycles determined by reverse bending tests on R.R. Moore rotating beam machines. Fatigue performance is primarily influenced by defects in the casting larger than the microstructural features of the alloy. As an example, in an Al-Si alloy such as 356, if produced to perfection, the fatigue life will depend on the silicon structure of the alloy. Various and often contradictory rules of thumb are used to estimate the fatigue limit of aluminum casting alloys, such as 0.4 of the tensile strength, or elongations of greater than 5% ensure better than a 70MPa fatigue limit. While these “rules of thumb” may indicate some broad trends, they do not translate well across alloy families or casting processes. The best way to maximize the fatigue life of a structure is to iterate between structural and castability geometry to enable the production of castings with minimum size defects.
Appropriate alloy selection and good casting practice such as grain refinement, modification, use of chills (to promote directional solidification) and filters (for inclusion removal, etc.) are important considerations to minimize casting defects. According to Gundalch, fatigue strength at 10 million cycles can be as low as 28 MPa (axial fatigue testing) in A356 and A319 alloys due to crack nucleation at microshrinkage voids. However, fatigue strength can be increased to 97 MPa at 10 million cycles by using chills to promote directional solidification, which reduces porosity to 0.01%.
Damage-tolerant design assumes that castings have defects at some level and fracture mechanics can be used to determine tolerable crack size and crack growth rates and better estimates design stress levels and fatigue performance. Make sure that the applied stress intensity factor, KI, is always less than the critical stress intensity factor, KC.
Castability geometry can be defined as how the shape of the product interacts with the physics of metal flow and the metallurgy of the alloy poured. The physics of the flow will include fluid characteristics of the alloy, interactions of the alloy and the atmosphere in which it is poured, and temperature at which the metal can be poured. Equally important is how the flow will interact with the molding medium (interface effect, thermal capacity, thermal conductivity, etc.) and how thermal gradients in the mold develop over time.
The castability geometry defines casting success. It controls whether the casting can fill within a reasonable amount of superheat and the type and amount of solidification shrinkage that forms. The geometry controls heat transfer within the casting as well as heat transfer to the molding medium. The castability geometry is the major factor that controls shrinkage porosity in most aluminum castings. Successful castability geometry is significantly influenced by the junctions of the walls and ribs. Often solidification software is used to define areas that are difficult to feed and are prone to porosity. A natural solidification (a simulation without a feeding system) can quickly pinpoint areas of concerns. Figure 1 shows one such example, where a heavy section remote from any feed area looked to be of concern. The customer redesigned this area to create a feed path, eliminating the problem.
Geometric features of a casting come together in junctions. Those junctions have been defined by the geometric form that they take, i.e., the T, Y, and L junctions. These types of junctions are shown in Figure 2 view (a) and (b). The junctions influence the thermal gradients that will occur in the casting and those thermal gradients have an influence on the performance of the part.
The casting soundness is heavily influenced by the thermal gradient that can be established during solidification. One way to think about this is to consider the Niyama criterion as a conceptual framework to design and process castings.
In general, the larger the Niyama criteria the more sound the casting will be. This is partially influenced by the design of the gating system through the strategic use of chills and risers. It is heavily influenced by the design of the casting itself.
Most junctions cause inherent hot spots that can lead to porosity. In the T junction shown in Fig. 2, the hot spot can be fed directly with a riser. The casting can easily be made sound in that area but the strength will be reduced somewhat because of slower solidification at the riser connection. The Y junction does not offer an attachment point for a riser that can be subsequently removed. In this case the junction should be made small enough to make it possible to chill the intersection. The solidification of the L junction is primarily controlled by both the inside and outside radii of the intersection.
The X junction in Fig. 3 (a) demonstrates how solidification will naturally occur. The intersection area, possessing more mass will solidify more slowly, leaving an unfed pool of liquid metal after the arms have solidified. The problem and solution in Figure 1 are essentially a junction design exercise. Solidification patterns for other junctions corresponding to Fig. 2 (a) and (b) are shown in Fig. 3 (b) and (c). These show fraction solid corresponding to later solidification stage.
The castability factors include (i) liquid delivery temperature and (ii) solidification shrinkage and amount. For aluminum alloys, the liquid delivery temperature is usually in the range 705 to 788C (1300 to 1450 F). The volumetric contraction from liquid to solid state for aluminum alloys ranges from 3 to 8.5%. This is influenced by the freezing range of the alloy. Extensive microporosity can be expected in long freezing range alloys (i.e., 206, 390, and 520) which need effective gating and risering design and the use of chills to promote directional solidification. Distributed shrinkage porosity is much less in short freezing range alloys (i.e., 356 and 413).
One of the most effective design tools is the proper use of radii and fillets These often have advantages for both the structural geometry through redistribution of stresses and the prevention of stress raisers as well as the castability geometry through better temperature distribution. This is illustrated in Figure 4. Note the improvement in heat distribution with the matched internal and external radii.
How an alloy reacts with the casting geometry is influenced by the castability characteristics of the alloy. The better the fluid life, the more intricate the detail that can be achieved in the casting. Alloys with shorter solidification ranges feed better and are less likely to hot tear. Alloys with high levels of magnesium (i.e., 535 [7% Mg]) are more prone to gas porosity. Alloys such as 356 and 357 are generally considered the best alloys for casting. The alloys containing copper (i.e., 319, 355 and 206) generally have a longer solidification range and are more subject to shrinkage porosity. However, they have many desirable properties such as good strength at elevated temperatures. Apparent alloy castability can be improved by insightful castability geometry and good gating systems.
Process Interface Geometry (Gating, Risering & Venting)
One important detail that is sometimes overlooked by those not in the metalcasting industry is that the gating system must allow the metal to enter and flow into the casting. These systems have been studied in depth and there are many software packages that can assist in the development of gating systems for castings. Figure 5 shows a typical gating system on an aluminum casting. While this gating system produces perfectly sound castings (the goal of a gating system!), it is not a good gating system. The gating connections extend up the side of the casting necessitating an additional machining operation. The risers on top of the casting slow down solidification and reduce mechanical properties in the most highly-stressed area of the castings. Resolving these difficulties is at the heart of the design process for castings.
Delivering high mechanical properties in sections requiring risers is a common issue. Often, section size is increased to compensate for this loss in properties, which may require a larger riser to feed it, which reduces properties further, and so on. A better solution may be to reduce the section size and use a chill to solidify it faster. There are many solutions to such problems but compromise for the sake of the application is required.
DOWNSTREAM PROCESSING GEOMETRY
A customer may never experience seeing a casting with all its gates, risers and chills attached before processing begins. This can be a teaching moment about the casting manufacturing process. All of the “stuff” that is critical to producing good castings must be removed before the final product is shipped and minor design changes will facilitate better castings delivered to the customer. Figures 6 and 7 show some examples of castings in the early stage of processing.
Figure 8 is an example of a modification done to the flange of a casting so ingates and risers could be attached and removed easily. This modification was made before the design was finalized and helps to emphasize the need for concurrent engineering of the casting and the casting manufacturing process. Figure 9 shows riser connections beneath the flange line requiring expensive manual removal. In this case, this was necessary given the need to feed the heavy sections in the interior of the casting. Perhaps earlier consultation with the customer to tie the heavy sections to the exterior of the part may have eliminated or reduced this problem.
Preparing for Machining—Finish Stock and Datums
An excellent guide to finish stock and basic information on datum structures is given in “Standards for Aluminum Sand and Permanent Mold Castings,” published by the Aluminum Association. Given the advancements in high-speed machining, having more machining stock than the minimum requirements can be used to an advantage by the foundry. For example, excess machine stock can increase the ability of a section to feed other areas of the casting and make it possible to increase the size of fillets and radii. Extra stock can also help prevent handling damage and make a casting less sensitive to grinding or sawing errors.
Design Considerations to Prevent Machining Issues
A well designed and executed casting should be able to be machined without great difficulty. However, most foundries and machine shops would agree that many issues can arise that would have been better handled during the design phase. Some of these issues can be summarized as follows:
- Cast locators are positioned in areas that are difficult to mold cleanly or are subject to distortion or damage during casting finishing operations. Locators that do not protrude from casting edges and are integrated into the casting geometry work best. Since peripheral locators are often used, it is important to consider the gating required for the casting to avoid the risers and gates from interfering with the integrity of the locators. Early consultation between the foundry and machine shop usually results in mutually beneficial locator positioning.
- A casting may be too complicated to achieve good machining alignment from a standard XYZ datum structure. Complicated aerospace or military casting may be produced using 10 to 50 cores, and alignment of those cores to the datum structures and to themselves presents huge challenges. In cases such as this, the castings are often “targeted”. Targeting a casting uses a special fixture to tie core positions or features to a datum structure that is machined in by either the foundry or the machine shop. Assuming that the core setting is done with standard foundry practice, this is a very effective technique to ensure accurately machined and tolerance castings. An example of a complicated transmission housing being targeted is shown in Figure 10. Recent progress in additive manufacturing of sand cores has also made it possible to combine cores in one-piece printed assemblies. This can reduce the total number of cores and improve positional accuracy.
- Distortion after machining can occur, often discovered in round bores with tight tolerances. This post-machining distortion is usually caused by residual stress induced in the casting during heat-treatment quench or, in non-heat-treated castings during solidification of the casting in a rigid mold. Distortion can usually be controlled with stress relief treatments or modifications to the quench parameters. For example, a warm water quench, spray or air quench can be used to reduce quench-related stresses. Modification to standard heat treatment practices may affect the mechanical properties. Reducing the severity of the quench usually reduces the tensile strength of the material. This should be reviewed and verified and may require changes to the part design. CS