A man boards a train, and finds himself sitting in a compartment opposite another passenger who is reading The Times. Every time the other finishes a page, he tears it from the paper, rolls it into a ball and throws it from the train. Perplexed, the man asks what he is doing.
“Ah,” says the man with the newspaper. “A trick I learned in Africa. Keeps the elephants away, don’t you know.”
“But there aren’t any elephants around here!”
“Yes. Works well, doesn’t it!”
Nearly twenty years ago when I was an Artist Blacksmith I found myself in Paul Zimmerman’s forge in Germany, I was impressed as to how well seated his hammer heads were compared to mine, and was told that every Christmas they would soak all their hammers for three days in linseed oil, this would swell the wood and hold the heads on. This made a deep impression on me, not in the least that Christmas was the only time they took a break long enough to complete this process.
When I returned from this trip I set up shop more professionally, soaked all my hammers and have not suffered loose heads to the extent that I used to; forging hammers have a tough life, they often get hot from contact with hot steel, often enough to expel uncured linseed from the heads. But in my new shop the process seemed to work and I have used the same process on Axe and Adze heads with similar success. Or have I ? my new shop is much less damp than my old one, maybe the heads would not have come loose without a soak. In a similar vein I have heard and repeated the idea that soaking a handle in water is a bad idea as the fibres will try to expand greatly to the extent that they will become crushed and on drying the handle will be a worse fit than before.
Is this true though? can the fibres expand and become crushed at the same time? Yet another version I have heard is that in Japanese woodworking if an irregularity is found in the socket of an axe head the corresponding area in the handle is peened down and then carved to make as good a fit as possible, the handle is then fitted and water poured down the handle/socket junction. This soaks and expands the compressed wood which then pops back into place locking the handle on. Yet would soaking the handle not make it swell, compress the fibres and then contract on drying and come loose? It is said that if something is repeated eleven times on the internet it becomes irrefutable fact, but it is hard to know which facts to believe sometimes.
Heads coming loose on axes and adzes, especially where the contact area is relatively small because the eye socket is quite short- such as the Chineland pattern axes is something that comes up fairly frequently, even in my workshop I have had students bring along such axes for attention.
A slight window in my schedule gave me the few hours I needed to finally put in the time to continue a series of experiments to address some of these questions. I had made a tentative start over a year ago and then as ever work took over. But I am finally on track again. The full cycle of these including wetting and drying cycles will take around a month to complete and will be quite long winded. I will document results as they progress partly because I am uncharacteristically excited by this and want to share the results, also there is quite a lot to get across and it seems to make sense to break it down a bit.
To start off with though I had, last year, rough turned some ash blanks in the workshop, these were now skimmed in my metal lathe to be an accurate fit in some 25mm tube I have.
These blanks were then cut into sections as were the tubes. They were then weighed and measured ( unfortunately my lathe was cutting a very slight taper) then driven into the tubes with a few very light taps of the hammer. This replicates an axe/adze/hammer handle as it would be fitted in my workshop. These tubes were then left over a radiator in my home and tested again. Unsurprisingly they had come loose to varying degrees, again replicating a common problem.
The next stage was to soak these test pieces in various solutions to try and fix them, I used the following:
Water- will it get even looser after drying?
PEG – polyethylene glycol- this was used quite commonly in the woodturning community a few years back but seems to have fallen out of favour somewhat now, it was quite hard to source. The idea is that the PEG will saturate the wood like water but will leave a waxy residue in the wood, reducing shrinkage, and in the case of woodturning it will lubricate the tool cutting it leaving a finer finish, check out Wiki for its other common uses.
Veritas Chair doctor glue- ‘ If a chair has a loose rung an injection of Chair Doctor Glue will first swell the rung and then bond it in position. The secret is the low viscosity. It soaks into the end grain, swells the wood then ‘freezes’ the wood in its swollen state as it cures. A film of dry glue is left on the walls of the wood cells preventing contraction. ‘ – sounded like it was worth a go, so a bottle was purchased.
Boiled linseed oil- My usual choice, rather than test different types of BLO I also tested a sample cut with 20% white spirit to thin it, hopefully allowing increased penetration. Accepted idea is that the oil penetrates and wets the fibres which then swell, and stay swollen as the oil will not evaporate out of the wood.
I used two samples in each solution- trying to match one loose and one tight fitting in each pair. They were left to soak for three days, replicating a German blacksmith’s Christmas break.
I didn’t soak the Chair doctor glue pair- just added some to the end grain at the end of the three days- these seemed to glue solid into the tubes.
After the soak weights and measurements were again taken, but the most obvious thing was that the water and PEG soaked pairs were now rock solid in the tubes and the protruding ends were noticeable oval. However all the BLO samples were as loose as when they had gone in, the oil actually made them slide in and out of the tubes easier. At this point I did wonder if all the time I had spent soaking tools in BLO had been a complete waste of time, maybe the only effect was to reduce the rate at which moisture could pass in and out of the wood as atmospheric conditions changed? Still I wanted to complete the experiment and see how the water and PEG samples dried so have been completing the weights and measurements on all the samples over the last couple of weeks as they dry or cure, I should have final results completed by next weekend, so far they have surprised me.