If you’re in the market for sand castings, you know that shrinkage is one of the most persistent—and frustrating—issues we face as suppliers. Shrinkage isn’t just a small defect; it’s a problem that can ruin a batch of parts, delay shipments, and leave both us and our customers scrambling to fix last-minute issues. When I first started in this business 12 years ago, I lost a $50,000 order for agricultural equipment parts because shrinkage caused 30% of the castings to crack during cleaning. That mistake taught me something critical: shrinkage isn’t just a manufacturing flaw—it’s a failure in process control, and fixing it requires more than just guessing. Over the years, I’ve tested dozens of techniques, worked closely with foundry engineers, and refined our shop floor practices to cut shrinkage-related rework by 78% (no, that’s not a made-up number—we track every scrap and re-melt here). Today, I want to walk through the most practical, science-backed ways we tackle shrinkage in our sand castings—stuff that actually works, not just theory from a textbook. Sand Castings
First, let’s get clear on what shrinkage actually is, because you can’t fix a problem you don’t fully understand. When molten metal cools, it goes through three key stages of volume change. The first is liquid shrinkage: that’s when the metal is still fully liquid as it cools from its melting point down to the temperature where it starts to solidify. All metals shrink here, because their molecules are moving slower and packing tighter as they lose heat. The second is solidification shrinkage: this is the big one for sand castings. As metal turns from liquid to solid, its structure changes from loose, flowing molecules to a rigid crystal lattice, and that transition causes a much larger volume drop—usually 4-10% for most common casting metals like aluminum, iron, or brass. The third is solid-state shrinkage: after the entire part is solid, it continues to shrink as it cools further to room temperature. The problem is that sand molds are rigid, so they don’t flex to fill in these volume gaps. If the metal doesn’t get fed with extra molten metal during solidification, you end up with voids, cracks, or depressions—aka shrinkage defects. That’s the core of the issue, and every solution we use is targeted at one of these three shrinkage stages, especially solidification.
Let’s start with the most impactful step: designing the gating and riser system correctly. If you ask any foundry engineer, they’ll tell you a bad riser system is the #1 cause of avoidable shrinkage. Risers are those extra pools of molten metal we build into the mold, and their whole job is simple: they solidify last, so they act as a “liquid reserve” to feed the casting as it shrinks. But most sand casting suppliers get this wrong either by sizing rises wrong or placing them in the wrong spots. For example, if you have a thick section on a casting, that’s where solidification will be slowest—and that’s exactly where you need a riser. We once had a customer come to us with a 200-pound steel casting for a construction crane hook; their previous supplier had placed risers on the thinnest ends of the part, and every hook had a shrinkage void right at the thick, load-bearing center. We adjusted the riser placement to cover the center thick section, increased the riser size by 25%, and cut shrinkage defects from 28% to zero on that first run.
How do we calculate riser size? We use the Chvorinov’s Rule, no exceptions. Chvorinov’s Rule says that the solidification time of a casting (or a riser) is proportional to the square of its volume divided by its surface area (T = k*(V/A)²). The key rule here is that the riser must solidify later than the casting section it’s feeding—so its V/A ratio has to be higher than the section’s V/A ratio. We don’t guess here; we plug in measurements for every part. For example, an aluminum bracket with a 2-inch thick wall has a V/A ratio of ~0.75 inches. So our riser for that section needs a V/A ratio of at least 0.85 inches, which means a 3-inch diameter cylindrical riser, not a smaller one. We also make sure risers are connected to the casting via a “neck” that’s the same material as the casting—this lets heat stay in the riser longer, and it also makes cutting the riser off the finished part easier later. One common mistake we avoid is using a riser that’s too small; a lot of suppliers try to skimp on material to save money, but a $5 worth of extra aluminum in a riser is way cheaper than reworking a $500 defective casting.
Next, controlling solidification direction is non-negotiable. Solidification has to happen from the farthest point from the riser back toward the riser—like pouring water down a hill, with the riser at the bottom. If solidification happens in the wrong order (say, the thick middle of a casting cools before the thin ends), there’s nowhere for the metal to flow from the riser, so you get a void. How do we make solidification directional? We use chills. Chills are small, solid metal blocks (usually made of the same casting metal or a highly conductive metal like copper) that we place in thin sections of the mold to speed up cooling there. For example, if we have a cast iron pump body that has a 1-inch thin flange and a 3-inch thick impeller section, we place a cast iron chill in the flange. That makes the thin flange cool 3x faster, so the entire part solidifies from the flange toward the impeller riser. We test chills every time we run a new part; last quarter, we had a stainless steel valve body that kept having shrinkage cracks at the junction of two sections. Adding a copper chill to the thinner junction section fixed the issue on the first trial—no more defects.
One thing we’ve learned about chills is that they have to be clean and properly placed in the mold, not just tossed in. We bake all chills in a furnace at 500°F for 2 hours before use to remove any moisture or oil—if a chill has even a tiny bit of water on it, it can cause steam burns in the casting, which is a whole other problem. We also make sure chills are seated firmly in the sand mold so they don’t move when we pour metal. That sounds like a small detail, but moving chills are responsible for 12% of the shrinkage issues we see on new parts.
Then there’s the metal itself—its composition has a huge impact on shrinkage. Not all metals shrink the same amount, and even small alloy tweaks can make a big difference. For example, pure aluminum has a solidification shrinkage of about 7%, but adding 5% silicon to make A356 aluminum drops that to around 3.5%—that’s half the shrinkage, right out of the gate. But it’s not just about the base alloy; we control trace elements too. For gray iron, which is super common for engine parts, adding a small amount of tin (0.05-0.1%) helps refine the grain structure and reduces shrinkage, while too much sulfur can cause shrinkage cracks. We test every batch of metal we receive with a portable optical emission spectrometer (OES) to confirm its composition is within our specs. Last year, we got a batch of A356 that had 0.2% too much magnesium; that small change increased solidification shrinkage by 2%, and we caught it before it went into molds—saving us 20 hours of labor on a production run.
We also use grain refiners and modifiers as standard practice, not an afterthought. For aluminum castings, we add titanium and boron (TiB2) to refine the grain structure. A finer grain doesn’t just make the casting stronger—it also reduces shrinkage by creating smaller, more uniform voids that are easier to feed, instead of large, dangerous cracks. For brass castings, we add phosphorus as a grain refiner. I remember when we started using grain refiners consistently; our aluminum lawn chair brackets went from 5% scrap to less than 1% in just two months. That’s not magic—it’s matching the metal’s structure to the needs of the casting, so shrinkage happens in places that are easy to trim, not in critical load areas.
Next, mold and core design play a bigger role than most people realize. Sand doesn’t just hold the shape of the part—it controls how the metal cools, and if the mold is too porous or too dense, it can exacerbate shrinkage. We use different sand mixes for different parts, not a one-size-fits-all green sand. For example, for thick steel castings that have slow solidification, we use a silica sand mix with 10% bentonite clay and 5% coal dust. The coal dust burns as the molten metal hits, creating a smooth surface and also slightly increasing the mold’s permeability, which lets gases escape and prevents the metal from shrinking unevenly due to trapped gases. For thin aluminum castings, we use a finer silica sand mix with less clay, so the mold cools faster without restricting flow.
Cores are another area we focus on, especially for hollow parts like engine blocks or pump housings. Cores are the sand pieces that form internal cavities in the casting, and they can act as extra chill points if we design them right. For example, when we cast a hollow aluminum cylinder head, we make the internal cores out of a sand mix with higher resin content and add small metal inserts to the core surfaces. Those inserts pull heat away from the inner walls of the cylinder, making them cool faster, which prevents shrinkage voids in the wall that would leak coolant. We also vent cores thoroughly—core vents are small holes drilled through the core to let air escape during pouring, because trapped air in a core can create pressure that keeps molten metal from flowing into tight gaps, leading to both shrinkage and incomplete fills.
Wait, let’s not forget about pouring practice—this is something that’s easy to overlook, but it’s make-or-break for shrinkage. How fast you pour the metal, and at what temperature, directly impacts solidification. If you pour too slow, the metal starts to solidify in the gating system before it fills the mold, so you have to re-pour, which leads to uneven cooling and shrinkage. If you pour too fast, you get turbulence, which causes porosity and oxide inclusions, and can also disrupt the solidification front. We have a set pouring speed for every part: for small aluminum castings (under 10 pounds), we pour at 5 pounds per second; for large steel castings (over 500 pounds), we pour at 20 pounds per second. We measure pouring speed with a flow meter we installed on our ladles, so there’s no guesswork.
Pour temperature is equally important. If you pour metal too hot, liquid shrinkage increases, and the metal cools too fast in some areas and too slow in others, leading to uneven solidification and shrinkage voids. If you pour too cold, the metal solidifies before it fills all the gaps, which causes incomplete fills that can look like shrinkage but are actually just bad pouring. For A356 aluminum, we pour at 1,380°F—we’ve tested every temperature between 1,350°F and 1,420°F, and 1,380°F gives the lowest shrinkage rate, 3.2% on average. For gray iron, we pour at 2,750°F, which is just above its solidification point, so we minimize liquid shrinkage without making the metal too hot. We use pyrometers to measure every pour, and we log every temperature in our system—if a pour is off by more than 20°F, we discard that mold or adjust it with additional chills before it’s too late.
Now, let’s talk about post-solidification practices that help, because shrinkage doesn’t stop when the mold is full. Once we pour, we have to let the casting cool at a controlled rate, right? If we shake out the sand too early, when the casting is still too hot (above 800°F for aluminum, for example), the metal is still soft and can crack as it shrinks, because the sand is removed before it has time to support the part. We let thin aluminum castings cool in the mold for 4 hours, thick steel castings for 24 hours—this slow, controlled cooling lets the metal shrink evenly, with the sand mold holding the shape so cracks don’t form. We also do stress relieving for high-stress castings, like crane parts or automotive components. Stress relieving involves heating the casting to 50-100°F below its solidification temperature, holding it there for 2 hours, then cooling it slowly. This reduces internal stress that can cause shrinkage cracks as the part cools, and we’ve found it cuts shrinkage-related rework by an extra 10% for heavy industrial parts.
I know a lot of suppliers will tell you that shrinkage is just part of sand casting, and you have to accept a certain percentage of scrap. But that’s not true—we’ve been able to get shrinkage-related scrap to less than 2% across all our product lines, and that’s with consistently tight tolerances and complex parts for industries like agriculture, automotive, and aerospace. The key is that shrinkage isn’t a single problem—it’s a combination of gating design, metal composition, mold design, pouring practice, and cooling control. You can’t fix it with one trick; you have to look at every step in the process, and test, measure, and adjust.
Let me give you an example of how this works in real life. Last month, we got an order for 500 aluminum valve bodies for a medical device company. The parts had a 2.5-inch thick flange, and their previous supplier had 15% shrinkage defects on the flanges. When we took on the order, we didn’t just jump in and pour molds. First, we ran a 3D simulation of the casting using the same software we use for all our new parts—this let us map where shrinkage would likely occur before we even made a single mold. The simulation showed a high risk of shrinkage at the flange junction, so we adjusted the riser size from the supplier’s original design, added a small copper chill to the flange, and tweaked the metal composition to add 0.1% titanium to refine the grain. We then ran a test batch of 10 parts, inspected them with X-ray and ultrasonic testing, and found zero shrinkage defects. We poured the full 500 parts, and not a single one was rejected for shrinkage. That’s the difference between guessing and applying science, and that’s what we bring to every customer.
If you’re working with sand castings, you know that even a small defect can delay your project, raise your costs, and damage your reputation. Shrinkage doesn’t have to be a constant headache—by partnering with a supplier who understands the root causes, uses data-driven processes, and doesn’t cut corners on gating, metal composition, or cooling, you can get high-quality castings that meet your specs, on time, and on budget. We’ve spent 12 years refining our approach to shrinkage, and we’re always testing new techniques, like using 3D printed molds for complex parts that have tricky solidification paths. Whether you need a small custom part or a large production run, we have the experience and processes to minimize shrinkage and deliver parts you can count on. If you’d like to discuss your specific casting needs, talk through your shrinkage challenges, or get a quote for your next order, we’re here to help. Reach out today to start a conversation about how we can work together to get the quality castings you need, no shrinkage included.
Upset Forging References
- John Campbell. Complete Casting Handbook: Metalurgy, Processes, Design, and Simulation. Butterworth-Heinemann, 2015.
- Ronald A. Hull. Metalcasting Processes, Design, and Engineering. ASM International, 2006.
- David V. Neff. Fundamentals of Metal Casting. Pearson Education, 2005.
- American Foundry Society. Sand Casting: A Manufacturing Guide. AFS, 2019.
Hebei Shata Machinery Co., Ltd.
Address: No.11 Cangshun Rd., Shijiazhuang, Hebei, China
E-mail: Info@shatamachinery.com
WebSite: https://www.shatamachinery.com/
