Brian Burns early tonewood testing

[kencierp]

His conclusions regarding Bearclaw and super tight grain most likely rattled a few builders:

Notes on Stiffness, Density and Q Testing of Tonewoods ~ Brian Burns

Healdsburg Guitar Festival ~August 1997
The reason for doing these tests is to gather some actual measurable data on guitar making woods that might help in judging their quality, and help in deciding how to thin and brace particular tops and backs. At this point I have tested 110 tops, 18 backs, 17 bridges, and miscellaneous billets of rosewood, redwood and spruce. I am about halfway through building a batch of 8 instruments (6 flamencos 2 classics} using wood selected with the
data from these tests as a guide. What properties make the "best" wood is still very much an open question. I have chosen the tops and backs with the highest overall stiffness to density ratios, and highest QE/D ratings. I build Flamenco and classical guitars "in the Spanish style", and I'm looking for woods that will produce that particularly bright tone.

How well this wood testing predicts the tone of finished guitars will certainly not be decided by these eight instruments, but it's a start.
In good journalistic style, conclusions first~

1. Stiffness in guitar soundboard materials (modulus of elasticity) varies much more widely than I thought it would. In a batch of 70 ungraded European Spruce tops, some were almost twice as stiff parallel to the grain as others. Perpendicular to the grain, some were
three times as stiff as others!

2. There is no correlation between parallel and perpendicular stiffness. A top can be stiff with the grain and flexible across the grain, or vice versa. Spruce is always much stiffer parallel to the grain than perpendicular. This batch averaged 17 times as stiff, with one top 9 times stiffer parallel to the grain, and one 29 times stiffer parallel to the grain.

3. Hazeling, from the German, Hazelfichte, or "bear claws", a cross grain figure in some tops,
tends to decrease stiffness parallel to the grain, and increase it perpendicular to the grain. In
heavily figured tops, parallel stiffness decreased up to 17%, and perpendicular stiffness
increased up to 7%. The usual effect is some loss of stiffness overall.

4. Finer grained tops tended to be stiffer parallel to the grain (23% at max.), but extremely fine grained ones tend to fall back to a lower value of parallel stiffness. Grain spacing had no effect on stiffness perpendicular to the grain.

5. In top wood there is some correlation between increased density (specific gravity) and increased stiffness parallel to the grain, but no effect on perpendicular stiffness.

6. In about half of my spruce and redwood tops, the stiffness to density ratio went up markedly in the denser boards. Since the finished weight of a soundboard is dependent on this ratio, roughly half of these denser tops would produce lighter weight finished soundboards than the less dense ones. As they increased in density, they put on "muscle", not fat.

[deadedith]

Very interesting. Goes against some of the 'accepted' wisdom.

[Herman]

It does not surprise me. But nice point of view.

Don't know who it was (I think Erwin Somogyi) that tested braces glued quatersawn versus flatsawn. His conclusion was that their strenghth was equal.

Ken, can you confirm this?

Herman

[kencierp]

Actually Somogyi is a real "sticklier" about using and gluing braces on the vertical, grain lines at perfect perpendiculars. His belief is that the braces can be much thinner and therefore add less weight to the sound-board -- a good thing. However, Roger Siminoff tested and published information that seemed to comfirm the point that flatsawn "necks" were stronger then 1/4 sawn necks. It would seem the same holds true for braces?

[johnnparchem]

A lot of lutherie practices are evolutionary, that is the ideas and practices that make the best instrument for the times tend to get copied. I often suspect studies comparing one technical attribute of a wood species vs. another and arriving at a conclusion that counters an accepted practice. It is not that I am a traditionalist, rather I am not sure the studies are focusing on all of the relevant attributes. The studies have value for further investigation, testing them on bunches of guitars. As an engineer I often times find that people are blinded by numbers, even accurate numbers, missing the errors in the original assumptions.

[David L]

I still don't like the inconsiderations for the variances when dealing with organic materials no matter how controlled the testing!

David L

[deadedith]

But does that mean abandoning all testing? Or just ensuring that we do not draw conclusions more specific than the data is able to support? Brian is pretty careful to say that here is the data, what it really means is not as clear.

A careful observer, with calibrated instruments and solid methodology, will provide good data. Interpretation of data is something different; but I like having data.

The more that is at stake, the less our certainty, or at least, the more sceptical we become. But this is not rocket surgery :-) so I am not too troubled about trusting some data over no data when it comes to using a piece of wood.

[David L]

I like data too Dave, I just try to be cautious as to how much stock I put in it. I don't think that comparetive testing should be abandoned, it's just that I'm a doubting Thomas and I tend to have a sensitive nose when it comes to "smelling a rat"

David L

[deadedith]

I'm with you, DavidL! Well put.

[johnnparchem]

But does that mean abandoning all testing? Or just ensuring that we do not draw conclusions more specific than the data is able to support? Brian is pretty careful to say that here is the data, what it really means is not as clear.

Testing is good and I have no issue with Brian's information. More like you say not drawing conclusions more specific than the data will report. I personally feel (not know) if one is working with the same type of wood personal testing can lead to some consistency. I also think a good/experienced luthier will get better consistencies relying on their feel for the wood and adjustments based on that feel than a scientific set of density/deflection tests.