Rust migration in a steel case: eight years of the same damage
· job 189 · a folio, 41 × 28 cm, one board with a steel plate laid inside it
The rust did not spread outward from the plate. It travelled along the grain of the board, in long pale streaks, and the only thing that slowed it was giving up on neutralising the plate and instead putting a barrier between the plate and the board.
What the owner believed
The book came from a estate sale in 2018 and the owner had kept it in an unheated outbuilding. Their account was that the board had "gone brown from damp". That was true but incomplete: the browning started at a line 60 mm from the joint, which is the edge of the hidden plate.
Opening the board
I opened the board at the tail where the endpaper had already lifted. Inside: a steel plate about 1.5 mm thick, roughly 22 × 14 cm, fixed with four flat-headed screws and what appeared to be a paper interleaving that had gone nearly black at the edges.
| Rust extent | Heaviest within 30 mm of the plate edge; visible streaks to 140 mm from it |
|---|---|
| Direction of streaks | Parallel to the grain of the board, not radial from the plate |
| Board condition at streaks | Still strong. Slightly softer than surrounding areas, no loss of stiffness |
| Text block | Sound. No rust found anywhere in the sewing or the folds |
The wrong thing I tried first
On a corner offcut from the same board I tried the tannin approach I had read about, then a chelating approach. Both removed colour. Both left the corner measurably weaker, and the chelated corner darkened again within three weeks in a normal room.
I stopped. My conclusion, which may be wrong and is written here as a conclusion anyway: the iron was not sitting on the fibre waiting to be removed. It was already bonded into it, and removing the colour did not remove the pathway.
What I did instead
- Removed the four screws and lifted the plate out of the board.
- Brushed the loose rust off the plate and coated it with a thin microcrystalline wax, twice, warm. Not to seal it forever. To stop it trading moisture with the board.
- Cut a new barrier: a sheet of archival board, 2 mm, between plate and original board.
- Re-fixed the plate with four new brass screws at the same holes, no adhesive anywhere near the original streaks.
Six-month check (not yet done)
Due December 2026. What I am watching for is whether the streaks advance past 140 mm. If they do, I have failed, and I will say so in this entry rather than deleting it.
Job number 189. Written the day the board closed.