Guides / Process

Season Cracking: Why Brass Parts Split Weeks After Assembly

September 21, 2026

There is a failure mode in brass that behaves unlike anything else a buyer deals with. The parts pass inspection. They assemble fine. Then somewhere between two weeks and six months later, a batch of them splits.

It is called stress corrosion cracking, and in brass it has an older name: season cracking.

The three ingredients

Stress corrosion cracking needs all three of these at once. Remove any one and it does not happen.

Residual tensile stress inside the part. Not the load it carries in service. Stress locked into the material by how it was made.

A susceptible alloy. High zinc alpha-beta brasses are the classic case. Low zinc and single phase alloys are far more tolerant.

A specific chemical environment. Ammonia is the textbook agent. Amines, which appear in many cleaning products and some corrosion inhibitors, do the same job. Moisture is usually needed too.

That combination is why the failure feels random. Two identical parts from the same batch, one in a dry electrical box and one in a plant room next to a cleaning store, behave completely differently.

Where the residual stress comes from

This is the part buyers can actually control, so it is worth listing.

• Cold working. Drawing, bending, staking, crimping, thread rolling, swaging over an insert. Any operation that deforms the metal at room temperature leaves stress behind.
• Press fits and interference assemblies. A brass sleeve pressed over a steel pin is under sustained tensile hoop stress for its whole life.
• Over-torqued threads, particularly tapered threads, which wedge and keep wedging.
• Heavy machining, especially aggressive roughing on thin walls, where the removed material was holding the rest in balance.
• Solidification. Bismuth brasses in particular can carry high residual stress from bismuth particle growth as the bar solidifies.

Note what is absent from that list: normal service load. A part can be nowhere near its yield strength in use and still crack, because the stress that drives the crack was already inside it before it left the factory.

Which parts are at risk

Work down this list for your own part.

1. Is the alloy a yellow brass with high zinc, C360, C385 or similar?
2. Was it cold formed, bent, crimped, staked or thread rolled after machining?
3. Is it pressed, shrunk or otherwise assembled under interference?
4. Does it live anywhere near ammonia, amines, urea or cleaning chemicals?
5. Is it in a humid environment?

Two or more yes answers and this is worth a conversation with your supplier before the first production run, not after.

Common real cases: cable glands in agricultural buildings, fittings in refrigeration plant rooms, crimped hose ends in food processing washdown areas, and brass inserts moulded into plastic and then loaded.

The fix: stress relief annealing

The usual remedy is simple and cheap compared to a field failure. Heat the part to around 330 to 350 degrees Celsius for two to four hours and let it cool.

That is low enough not to recrystallise the material, so you keep the mechanical properties you paid for, and hot enough to let the locked-in stress relax out.

Three practical notes.

Order matters. Stress relief goes after the last cold working operation. Relieving a part and then crimping it puts the stress straight back in.

It has to be on the drawing. No shop adds a heat treatment operation nobody asked for. If it is not specified, it does not happen.

It is not free, but it is cheap. A batch through an oven costs far less than a recall, and usually less than the argument about who pays for one.

Proving it worked

Two standard tests, both accelerated and both destructive on the samples used.

ASTM B154, mercurous nitrate test. Parts are immersed in mercurous nitrate solution for a defined time, then examined for cracks. Fast and severe. Note that the reagent is mercury based, so fewer labs run it than used to.

ISO 6957, ammonia vapour test. Parts are held over an ammonia solution in a sealed chamber, typically for 24 hours, then examined. This is now the more common choice.

Both are sampling tests. You test a few parts to destruction to gain confidence in the batch, not every part.

What to write on the drawing

Stress relief anneal after all cold working.
330 to 350 C, 2 hours minimum, air cool.
Verify per ISO 6957, 24 h ammonia vapour,
sample 3 parts per lot, no cracking permitted.

If you are not sure whether your part needs it, describe the assembly and the environment in your RFQ and let the shops that quote say so. A shop that has seen this failure before will raise it unprompted, and that in itself tells you something useful about the shop.

Not the same as dezincification

These two get confused because both are corrosion and both affect brass in service. They are different.

Dezincification is zinc leaching out into water, leaving porous copper. It needs water chemistry, not stress. Stress corrosion cracking is intergranular cracking driven by residual stress plus a specific agent, usually ammonia.

A grade that resists one does not automatically resist the other, and the remedies are different. See dezincification and the grades that resist it.

Have a part that is cold formed or lives in a chemically active environment? Send the drawing and say so in the description. We match it to shops that have dealt with this before.

Frequently asked questions

Keep reading

• Brass CNC Machining: Why It Is Cheap, and Where the Cost Hides
• Swiss Turning Brass Parts and When It Is the Cheaper Route