Shrinkage is the casting defect whose root cause is the most fundamental and the least avoidable in principle — it arises directly from the physical fact that aluminium contracts when it solidifies, and the metal that was liquid and filled the cavity completely is no longer sufficient in volume to fill that same cavity in its solid state. Every aluminium alloy shrinks during solidification; the question is not whether shrinkage occurs but where in the casting it occurs, whether that location is inside the casting boundary where it creates an internal void, or outside the casting boundary in a feeder where it is acceptable and expected. Feeding system design is the engineering discipline whose objective is to ensure the former rather than the latter — to direct the shrinkage to locations that do not compromise the casting's structural integrity or dimensional accuracy.
The Physics of Solidification Shrinkage — Why It Happens Where It Does
Aluminium alloys shrink approximately 3.5 to 7 percent by volume during solidification, depending on the alloy composition and the solidification range. This volumetric contraction is accommodated during solidification by liquid metal feeding from adjacent regions that are still molten — as one region solidifies and contracts, it draws liquid metal from the surrounding melt to fill the contracting volume. This feeding process continues until the feeding path — the channel of liquid metal connecting the shrinking region to a liquid metal reservoir — solidifies and feeding ceases. At the moment feeding ceases, any remaining volume deficit in the already-solidified regions can no longer be compensated, and a void forms.
The location of shrinkage voids within a casting is therefore determined by the solidification sequence — the order in which different regions of the casting transition from liquid to solid. The region that solidifies last — the thermal centre of the casting, typically the thickest cross-section or the region furthest from the die wall — is the region where the feeding path solidifies last and where the shrinkage void forms if feeding has been inadequate. This principle — that the last region to solidify is the region most vulnerable to shrinkage porosity — is the foundation of feeding system design, because it identifies where feeders must be located and why the feeding path must remain open until the regions it feeds have fully solidified.
The solidification sequence is controlled by the temperature gradient within the casting — the rate at which temperature falls with distance from the die wall, which determines which regions cool fastest and therefore solidify earliest. A steep temperature gradient — produced by a cold die, effective die cooling, and thin casting sections — drives rapid solidification from the die wall inward, producing a directional solidification front that feeds shrinkage efficiently toward the die wall and away from internal thermal centres. A shallow gradient — from a hot die, inadequate cooling, or thick sections — produces slow, poorly directed solidification that creates multiple isolated liquid pools that each require feeding independently and that frequently receive inadequate feeding because the feeding paths between them solidify prematurely.
Identifying Shrinkage Risk — Before the Die Is Built
The most cost-effective point at which to address shrinkage risk in a gravity die casting is before the die is built — during the casting design and die design phase where changes can be made to section geometry, feeder location, and cooling channel design without incurring tooling rework costs. Identifying shrinkage risk at the design stage requires either casting simulation — computer modelling of the solidification sequence and temperature field within the casting geometry — or the application of systematic geometric analysis to identify the thermal centres that will solidify last and therefore carry the highest shrinkage risk.
Geometric shrinkage analysis uses the inscribed sphere method — finding the largest sphere that can be inscribed within the casting geometry at each location, which represents the thermal mass at that location and therefore indicates its solidification rate relative to its surroundings. The location where the largest inscribed sphere is found is the location with the highest thermal mass and therefore the last to solidify — the location of highest shrinkage risk. Connecting these maximum inscribed sphere locations across the casting geometry traces the path of progressive solidification and identifies the feeding requirements that the casting geometry imposes.
Casting simulation software performs this analysis computationally, modelling the heat transfer from the liquid aluminium through the solidifying casting shell to the die and cooling channels, and predicting the solidification sequence and the location of remaining liquid metal as a function of time during solidification. The simulation output — typically visualised as a colour map showing solidification time or remaining liquid fraction at each point in the casting — directly identifies the regions where the last liquid remains and where feeding is required. Modern casting simulation is accurate enough to predict shrinkage void location within a few millimetres of its actual position in trial castings, and its use at the die design stage consistently reduces the number of tooling rework iterations required to achieve acceptable casting soundness.
Feeder Design — Reservoir, Location, and Modulus
A feeder is a reservoir of liquid metal, connected to the casting by a feeding channel, whose function is to supply liquid metal to the shrinking casting during solidification. For the feeder to fulfil this function, it must satisfy three requirements simultaneously: it must contain sufficient volume of liquid metal to compensate for the casting's solidification shrinkage; it must remain liquid until after the region it feeds has solidified, so that feeding can continue throughout the casting's solidification period; and it must be connected to the region it feeds by a feeding channel that also remains liquid throughout the solidification period.
The first requirement — sufficient volume — is addressed by feeder sizing. The feeder must contain at least the volume of metal that the casting requires for shrinkage compensation, plus the metal that the feeder itself requires to compensate for its own shrinkage. Feeder volume calculation methods — ranging from simple percentage-of-casting-volume rules to the more rigorous modulus method based on solidification time theory — provide the minimum feeder volume for a given casting section. In practice, feeders are typically sized somewhat above the theoretical minimum to provide a safety margin against variability in pouring temperature, alloy composition, and die temperature that affect the actual shrinkage volume in production.
The second requirement — the feeder must remain liquid longer than the casting section it feeds — is addressed by feeder modulus. The modulus of a casting section or feeder is defined as the ratio of its volume to its surface area. Sections with higher modulus solidify more slowly — they have more thermal mass relative to their heat-losing surface — and sections with lower modulus solidify faster. For a feeder to remain liquid longer than the casting section it feeds, the feeder's modulus must be greater than the casting section's modulus — typically by a factor of 1.2 to 1.5 in standard feeder design practice. This modulus relationship ensures that the feeder is always the last part of the metal system to solidify, drawing its shrinkage void safely outside the casting boundary.
The third requirement — an open feeding channel — is addressed by the feeding channel modulus and geometry. The feeding channel connecting the feeder to the casting section must have a modulus intermediate between the feeder's and the section's, ensuring progressive solidification from section through channel to feeder rather than premature channel solidification that isolates the feeder from the section it is intended to feed. Feeding channel design — its cross-section, length, and connection geometry — is as important as feeder sizing in achieving effective shrinkage feeding, and it is frequently the element of feeding system design that receives the least systematic attention in practice.
Practical Feeding System Design — The Decisions That Determine Outcome
Translating the theoretical principles of feeding system design into a practical die design requires decisions about feeder shape, feeder placement relative to the casting, and the integration of feeders with the gating system that also fills the cavity with metal. Each decision affects the feeding effectiveness, the die cost, the casting yield — the ratio of casting weight to total metal weight poured — and the ease of feeder removal from the finished casting.
Feeder shape affects the ratio of feeder volume to feeder surface area — the modulus — for a given volume of metal. A spherical feeder has the highest modulus for a given volume but is difficult to achieve in a gravity die. A cylindrical feeder with height-to-diameter ratio of approximately 1.5 is the practical optimum for most gravity die casting applications, providing adequate modulus with a geometry that can be formed by a simple cylindrical pocket in the die. Taller, narrower cylinders have lower modulus because their higher surface area relative to volume causes faster cooling; shorter, wider cylinders also have lower modulus because their flat ends cool rapidly. The height-to-diameter ratio of approximately 1.5 represents the optimum balance for a gravity die casting feeder geometry.
Feeder placement determines the feeding distance — the maximum distance from the feeder at which adequate feeding can be supplied. Feeding distance in aluminium gravity die castings is limited by the freezing range of the alloy and the temperature gradient in the casting during solidification. LM6 — the near-eutectic alloy — has a short freezing range and high fluidity, allowing relatively long feeding distances. LM25 — with its wider solidification range — has a shorter effective feeding distance that requires feeders to be placed closer to the thermal centres they serve. Exceeding the alloy's effective feeding distance — by placing a feeder too far from a thick section — produces shrinkage porosity at the location beyond the feeding range regardless of feeder sizing, because liquid metal cannot be drawn from the feeder across a solidified or semi-solid region.
The integration of feeders with the gating system — using the gate as an in-gate feeder that provides both the metal fill path and the feeding reservoir — is a design approach that reduces total metal weight and die complexity when it can be applied. An in-gate feeder must be sized to satisfy both its gating function — providing adequate flow area for cavity fill — and its feeding function — providing adequate modulus and volume for shrinkage compensation. When these two requirements are compatible, the in-gate feeder is an elegant solution. When they are not compatible — when the gate size required for fill is too small for adequate feeding modulus, or vice versa — separate feeders and gates are required, and the design must balance their competing requirements without compromising either fill quality or feeding effectiveness.