Cracks in a Steel Casting. Where They Come From. What We Actually Do About Them.

Honestly, cast steel after sand inclusions, cracks are the thing that keeps me up. You don't always spot one until the part's machined, or it's been running for a while, and then you're not fixing a casting anymore. You're on the phone with a customer whose line just stopped, trying to explain something you can't fully see.

Cracks break down into hot and cold, and honestly in most cases it's not one thing. It's a stack of small things adding up.

Hot cracks first.

These happen while the metal is still solidifying. And that's the worst possible moment, because right at the end of solidification the steel is at its weakest. Low strength, low plasticity. The crystal skeleton's just formed and it wants to contract, and if the mold won't give—or the wall thickness jumps around too much—the stress climbs fast. Push past what the steel can take at that temperature and it just tears.

You can tell a hot crack because it's wobbly. Uneven width. Runs along the grain boundaries. Wider at the surface, narrowing as it goes in. And that dull, oxidized look—it broke while it was hot.

What causes it? Uneven walls. Tight internal corners. Too much sulfur and phosphorus hanging around—they drift to the grain boundaries and weaken everything. We try to keep the total under 0.025, call it 0.016 phosphorus and 0.006 sulfur, just to have a safe margin. Pour too hot and the grain goes coarse. Ram the sand too tight and it won't yield when the metal needs to shrink.

Then cold cracks.

These wait. They show up later, once the casting's cooled down to where it behaves like elastic. So it's usually during knockout, or cleaning, or machining when you find out. The classic setup: a thick section chills and hardens fast on the outside, and the inside is still shrinking, and the two pull against each other until the thing snaps.

Cold cracks read different. Thin, straight lines. Cut clean across the section. Bright metallic fracture, no oxidation, because it broke cold.

They come from cooling too fast, pulling the mold open too early, knocking out while the part's still hot, heat treatment done sloppily, hammering on risers, and design choices like sharp corners or abrupt changes in thickness. High carbon, or phosphorus much past 0.05%, and the steel gets brittle and makes all of it worse.

So here's what we watch. Four things. We don't skip any.

Because cracks stack up from several causes at once, you only really lower them by tightening all four areas together.

One, the melt. Sulfur and phosphorus settle at the grain boundaries and that's the direct driver of hot cracks. We run an electric arc furnace plus LF refining, and we hit every heat three times with a spectrometer. Total sulfur and phosphorus under 0.025. Then argon purge to pull the inclusions and gas out.

Two, the geometry. Keep wall thickness as even as you can, and round the corners where things transition, so stress doesn't pile up at a point. On wheels and parts like that, we've curved the spokes to give them somewhere to move. When you've got thick and thin sections you can't avoid, use the ingate position and the gating and riser design to slow down the thick parts. Get everything cooling at roughly the same rate. That's the "simultaneous solidification" idea, and it helps both kinds of cracks.

Three, the cooling and the timing. The sand mold's actually a decent insulator—it evens out the temperature between thick and thin areas. So leave the casting in the mold longer. Don't pull it early, because opening too soon locks in a lot of internal stress. And be gentle at knockout, cleaning, moving it. No dropping, no banging. We pour around 1520 to 1600 degrees. Too hot and the grain gets coarse and crack-prone. Too cold and the mold doesn't fill cleanly.

Four, the stress relief. Castings carrying a lot of residual stress should get a stress-relief or aging treatment early, before that stress turns into a cold crack. And if you've cut off risers or done a weld repair, run another cycle afterward. Normalizing refines the grain. Tempering pulls the stress out.

And the ones you can't see

Surface cracks, your eye can catch. The ones buried inside? Those need NDT. Magnetic particle for surface and near-surface. Ultrasonic for the internal ones. On critical load-bearing parts we run both. That's how we keep cracked parts from ever leaving the floor.

Someone usually asks how to tell hot from cold. Hot forms during solidification—wavy, along the grain boundaries, oxidized face. Cold forms after cooling—thin, straight, cutting clean across, bright metallic break.

Sulfur and phosphorus matter more than people think. They gather at the grain boundaries and weaken the steel. Keep the total under 0.025; past about 0.05% phosphorus and it gets noticeably brittle.

Pouring temperature? We hold it around 1520-1600. Hotter coarsens the grain. Cooler leaves short fills and cold shuts.

Can a cracked casting be saved? Surface defects can be repaired—clean the defect out, weld with the right filler at low heat, then re-inspect. But the real fix is upstream: clean chemistry, gentler geometry, simultaneous solidification, controlled knockout timing.

About us, quick. Thirty years of large steel castings, casting to heat treatment to machining to inspection, all in-house. On big girth gears we hold sulfur and phosphorus tight, use simultaneous solidification,  Send the drawing and we'll have a process plan in 24 hours. sales@xtfcasting.com. WhatsApp +86 19837330051.