Why Did A Clean Part Rust Overnight?

The call came in the first week of winter. A plant running cast iron housings through a tunnel washer had found rust on the whole batch the next morning, and the same machine had been running aluminium parts all week without a trace. The batch was due to ship the following day.

I have taken a lot of these calls, and most of them end up in drying or storage. But this one took two wrong turns before it got there, and both of them were mine. That is the reason it is worth writing down.

What We Saw

I went in the next morning and looked at the rejected parts first. The rust was a thin, even bloom, yellowish, spread across the two large faces and the inside of the bearing bores. The flange faces were clean.

That distribution was the first real information. The flanges are the last feature to be finished and have the best surface finish; the faces and bores are either as-cast or semi-finished and are rougher. Rust was growing on the rough surfaces and not on the smooth ones, which says the water was sitting in microscopic pits and that there was not much of it. A lot of water would have produced a continuous film rather than isolated patches.

The line itself looked fine. Tunnel washer, hot air drying at the exit, parts coming out dry. The operator ran one through while I watched and handed it to me; there was no visible water on the surface.

The First Answer Was Wrong

My first read was insufficient drying, and the logic was comfortable: rough cast surface, water trapped in pits, hot air only reaching the outside. We extended the drying zone and raised the air temperature.

They ran it. Parts came out hotter and visibly drier, and the next morning the rust was back, but the patches had moved from the middle of the faces towards the edges.

The movement was the clue I should have picked up faster. If the cause were drying alone, the pattern would have stayed where it was. Rust that relocates is following something other than the dryer, and the obvious candidate is how the part was sitting.

So Was The Second

The second suspect was the bath. Cast iron in alkaline chemistry carries a real corrosion risk, and concentration, temperature and rinse quality were all plausible contributors. We checked concentration, pH and bath temperature, and all three were inside the specified range. Rinse water conductivity was slightly high but not alarming.

To settle it, we took ten parts from the same batch, rinsed them with clean water, wiped them dry by hand and left them on a bench on their own. The next morning all ten were clean.

That largely cleared the bath. Parts washed in the same liquid and then handled individually did not rust, so the surface was not inherently reactive when it left the machine. Something after the machine was doing it.

What It Actually Was

It was the box.

They were packing into steel-framed crates with a layer of corrugated board on the bottom. The board was recycled paper, which absorbs moisture. Parts went in dry, a layer of board between each layer of parts, lid on, shipped the next morning.

That crate was fine in summer. The shop was warm and ventilated, and whatever moisture was inside escaped. In winter the shop dropped to the low teens, the crate was closed, and the small amount of water still on the parts could not leave. Inside the crate the air reached close to saturation overnight, and recycled board carries enough acidity to help. One night was enough for a surface bloom.

We tested it the cheapest way available. Half the batch went into the original crates with the paper, half into dry plastic totes with no paper between parts, both stored in the same place overnight. The crates rusted. The totes did not.

That closed it. The parts were not badly washed and not badly dried; the packing step was undoing a result that had already been achieved. The aluminium parts never showed the problem because aluminium does not bloom like this in the same conditions, which is also why the crate had gone unquestioned for years.

What Changed

Three things, none of them expensive:

• Packing moved to plastic totes with internal dividers, no paper between parts
• A ten-minute forced cooling step before packing, so parts enter the tote close to shop temperature and do not condense inside it
• A small desiccant pack for any batch held longer than three days

The first two cost almost nothing. The third is a few hundred yuan a year at their volume.

The Result

No rust complaints in the first month. The following rainy season produced one recurrence, with a different cause: that batch had sat in the shop for two days before packing, during a stretch of wet weather. That is a holding-time problem rather than a packing problem, and the desiccant covered it.

The plant wrote the fix into their work instruction as one line: cool before packing, no paper inside the tote. One line was worth more than any parameter we discussed.

What I Still Cannot Explain

One thing never fully resolved. The same crate, the same process, and some batches rusted while others did not. The earlier records from the Shimada Big Bird Industrial period show the same inconsistency, which suggests it is not new.

My working explanation is holding time in the shop before packing. Two hours and half a day leave different amounts of residual water, and the starting humidity inside the crate differs accordingly. But if you ask me how long is safe, I do not have data I would defend. I can only say shorter is better, and that the cooling step buys more than any time limit.

If your parts rust overnight after washing, look at the container before you look at the machine. That order saves a lot of time.