Oilcloth Making

Recreating old techniques for boiled linseed oil and oilcloth


Boiled Linseed Oil

"Boiled" linseed oil can mean many things. Some sources describe a low-temperature heating that causes the oil to polymerise. Modern oils are barely heated at all, but have a chemical drier added to them. The traditional "boiling" process though consisted of long heating at high temperatures, in combination with a metallic compound. Such oils are now almost unobtainable.

The traditional driers for linseed oil were either lead or manganese based. They were added to the oil in solid form, then heated to high temperatures for long periods. Modern driers are no longer lead-based (for safety reasons) and are usually cobalt-based. They're prepared as liquids (you can buy these as "driers") and added to cold or slightly-warmed oil without further heating.

Following the recipe in Bill Knight's book, I produced small batches of both manganese and lead-dried linseed oils. I describe some of the process here, but I'm not intending to reproduce or replace their text. If you're interested, I really do recommend getting hold of a copy.

Boiling the oil

Equipment for boiling oil

Equipment used for the first tests was extremely rudimentary. Obviously we worked outdoors. Heat was supplied by a butane camping stove. The oil was heated in 500ml batches in a cast-iron pot. Although bare iron has been advised against, this is what was available and seems in keeping with the period. The pot was simply suspended over the gas burner from a stand, and a small windshield of scrap steel was hung around the burner. Temperature monitoring was by means of a modern K-type thermocouple.

Boiling was carried out according to the book. Driers were mixed into the cold oil as a powder, and stirred thoroughly. The oil was then heated, and marble chips added when hot. Temperature was maintained at around 270°C for two hours for the lead / manganese oil, or four hours for the plain lead. Fortunately the burner size was about right for holding a steady temperature and no further control was required.

The feedstock for both was raw linseed oil (obtained from Screwfix)

The lead used was lead dross from a shooter's bullet-casting pot. This is mainly lead oxide, although there may be other nasties and critters in there too.

The manganese was a commercial burnt umber pigment (from Liberon).

The boiling process produces an acid product. For good long-term stability, this acid should be neutralised. As described

My first oil batches produced around 300ml om lead/manganese oil and 500ml of lead oil. It's unclear why there was so much variation.

Waiting for the oil boiling process

It's a slow process. Boiling a lead-based oil takes around 4 hours.

Oilcloth

The initial use for this oil was to reproduce some traditional oilcloth, according to the recipes in Bill Knight's book.

Oilcloth, coated and stretched on a tenter frame to dry

This oilcloth is for use covering a small portable chest.

The oilcloth process is much simpler than the oil-making. The cloth is a fine-woven cotton canvas, stretched out and stapled onto a simple wooden frame. To improve adhesion, the cloth was washed hot beforehand.

The first step is to size the fabric, using kitchen gelatine.

As can be seen in the picture, the surface soon showed uneven colouring. As far as I can tell, this was caused by the darker patches being less well sized, thus allowing the oil to penetrate more deeply. Even after many coats, this unevenness was still present.

A later problem became evident when fixing the oilcloth to the chest. Although the oil layer had good cohesion, it lacked adhesion. Whenever I folded the oilcloth tightly over an edge, it tended to flake off - particularly after any impact. This was easy enough to re-touch, but still worrying. Some of this may have been due to the oil being poorly dried at that time - 4 month old offcuts are much less prone to it.

It's a thirsty process for oil. I'd estimate about 3 sq. yards / litre of finished oil. In the future, I think I'd skip the sizing process and use fewer coats (about 6 each side, for a double sided cloth. 6 and 4 for cloth used to cover a box).

A later experiment with the remainder of the cloth used some commercial bitumen paint (BlackJack, sold as a primer for roofing felt). This gave a rapidly drying black "bitumencloth" and was generally easier to work with. The surface finish wasn't as good, although that may have been due to carelessness for an experiment and using a thinner coating. It certainly didn't body as rapidly or as well.

Other Resources