End grain and oak

[NormandieStill]

Even by my normal standards, this is going to be long... very long so:

TLDR (of the TLDR): In my opinion, try to minimise the end grain in your staves

TLDR: The proportion of end grain to side grain in a stave in glass ageing has a significant effect on extraction of compounds from the oak. It stands to reason that it will also have a significant effect on the woodiness of the resulting spirit. It might be that fast extraction of compounds will have a beneficial effect on speeding up ageing overall but will probably decrease the early drinkability of the spirit and require longer time (or eventually, heat) to encourage reaction and ester formation.

Background

We'd all love to be ageing large quantities of spirits in large barrels, but realistically this is not an option. Badmotivator's design allows for a decent compromise, but most of us have to content ourselves with ageing in glass containers with wood sticks. There's been a lot of research (sometimes anecdotal) into techniques for "rapid" ageing, sometimes using wood chips rather than sticks, but as a woodworker, I'm very much aware of the varying physical properties of wood when you work with, or against the grain. An example would be cutting a tenon perpendicular to the grain. To the newcomer this might not seem heretical, but the resulting tenon would have zero strength because the strength of wood is in the fibres and you'd have cut them. This is also why a round branch can be many times stronger than a larger sectioned length of cut wood. When barrels are made, the staves are split from logs. Not cut. Originally this would have been the only real technique for making staves, but it would have soon been found that cut staves tend to break when bent.

This process originating from a mechanical need results in a peculiarity of barrels. The spirit "sees" little to no end-grain. Staves are obviously shaped with cutting tools so some grain is cut, but the amount of exposed end-grain is minimal. The Badmotivator barrels replicate this (again, perhaps not intentionally, but as a natural consequence of the build process). Users of barrels report better results, than those ageing in glass. This is possibly in part due to the bias caused by the financial investment, but it's a near universal feeling. And the single biggest difference between ageing in a barrel and ageing in glass is the presence of end-grain.

So I decided to investigate the effect of end-grain on ageing... or at least, start exploring. Over a period of some weeks I developed an experiment that would allow me to compare end-grain and side-grain.

The experiment

End-grain cannot be eliminated in oak sticks, but it can be exaggerated. The idea was to compare samples with the same overall surface area, changing the proportion of end-grain. Some back of envelope maths and a bit of tweaking led to the following dimensions for the sticks.

4cm² section (2cm x 2cm)
Sample 1: 1 x 5cm length
Sample 2: 3 x 1cm length

This produces samples with a total of 48cm² of surface area, with 8cm² of sample 1 being end-grain compared to 24cm² of sample 2. So in sample 1 the end-grain represents 16.6% of the surface area, where as in sample 2 is makes up 50%.

Badmotivator calculated approximate values for surface area:volume for standard oak barrels, arriving at roughly 98cm² / litre. So my 48cm² would be appropriate for 500ml of liquid. In order to better observe the results I decided to put each sample into 1L of water.

The samples were made by splitting some weathered French oak with a froe (which doesn't work nearly as well on dry wood as it does on green wood!). I then selected a straight section and cleaned it up with my planer / thicknesser to produce a 2cmx2cm length which I then cut into the sticks for the experiment.

To see if char had an effect, I doubled up my samples and lightly charred half of them.

To record the change, I weighed the samples on my jeweller's scale before putting them into the water and then every time I took a "reading" (after patting them dry). This would monitor the absorption of water. I also built a light box using a flash and using my DSLR I photographed each sample (in the jars) setting the exposure so as not to burn out the highlights. These photos would be processed minimally and identically. The captured "white" section being used to adjust the exposure so that an average RGB value could be taken for the sample. This RGB value was converted to HSV on the assumption that the saturation and hue data would allow "objective" comparison.

Faults

- The most obvious "defect" in the design is the use of water rather than alcohol. I would predict that the results will be the same, albeit it at a potentially different rate.
- I did my best to limit fibre cutting, but obviously some of the side-grain will present cut fibres. To rely on split wood would make the surface area calculations somewhat complicated.
- I have no idea what compounds are being extracted. Detailed comparison of the hue values might provide some insight, but I'm not convinced enough of the reliability of the data processing to provide that level of fine information.
- I'm technically removing small amounts of water each time I take a sample, but these amounts are minimal compared to the 1L volume.
- I can't talk to impact on taste. This just provides data on extraction rates. Whether it is good to extract quickly or slowly remains unknown.

Results (so far)

Jar 1: Low end-grain, raw oak
Jar 2: High end-grain, raw oak
Jar 3: Low end-grain, charred oak
Jar 4: High end-grain, charred oak

Mass-percent.png

Saturation-HSV.png

So the first graph shows the weight of the samples over time as a percentage of their dry weight. The second shows the saturation of the samples. There is a consistent change in the hue and a difference between the hue from the low end-grain and high end-grain samples, but I'm still trying to work out how to represent that data in a meaningful and useful way.

IMG_1717.jpg.modified.jpeg

The first sample taken of jar 1

IMG_1759-strip.jpg

Jar 1 after 16 days

IMG_1718-strip.jpg

The first sample taken of jar 2

IMG_1760-strip.jpg

Jar 2 after 16 days

Conclusion

The experiment is ongoing, and I hope to continue it for some time, although at some point I may need to reclaim the jars for ageing actual spirits in! My current conclusions are in my TLDR(s) at the start.