Deflection vs. Frequency for measuring flexural modulus

[Michael Minton]

I agree that the regression basically shows no fit at all. What are the units you are using on the X and Y axis? They don’t line up will values I would expect to see. E.g., what are the deflection units (are 9/5 to 13 in thousandths?), and the frequency units (10-13 doesn’t make any sense to me for frequency units).

[Neil Miller]

The units are gigapascals for flexural modulus on both axes. One calculated from frequency and the other from deflection. They’re within a reasonable range for spruce, just no correlation between the two methods.

• This reply was modified 3 months ago by  Neil Miller.

[Giuliano Nicoletti]

Hello Neil,

as you rightly pointed out, they are all in reasonable range for the species, but do not perfectly correlate (at least, not all of them). In my experience, the deflection method is is not very reliable, while the opposite is true for a dynamic test.

Did you try to repeat both test on the same sample to check dispersion?

Cheers,

Giuliano

[Neil Miller]

I didn’t repeat the test. When you say you find deflection is less “reliable” what standard are you using? Do you mean less “repeatable” or is there another standard against which to measure them? These same tops were tested by Pacific Rim Tonewoods, and I found a similar (weak) correlation between theirs and my deflection-based E, but absolutely no correlation between theirs and my frequency-based E (r2 = 0.087).

[Paul M]

I’m not super smart and all of this is on the outer edge of my ability to understand things, but for me, Guiliano’s methods are the ones that are most approachable to me and additionally he’s been exceptionally helpful in talking to luthiers here and elsewhere.

For that reason I’m sticking with his methods and not worrying about anything else. I’m grading my own wood and just going through the process I think I’m learning things. Importantly I think it’s a big thing to build guitars as you go, you will figure out what you want to do having some experience (at least that’s the case for me, the guitars I built last year led me to questions that provided ideas I’m trying to implement this year).

Anyway, just my thoughts. If you feel confident in your own methods, it sort of doesn’t matter what how someone else measured something.

I’m not getting super freaky about bracewood, mostly focusing on the density at this point.

[Neil Miller]

I’m not trying to push any one method, just trying to understand the options. The T. Gore books talk about both, so it seems reasonable that they should produce similar results. I’m curious to know if others have produced results similar to mine, or if perhaps I’m doing something wrong…

I will differ with you on the issue of brace selection. Brace wood contributes significantly to both the mass and stiffness of a top. Why should we give it any less attention than the top plate? Too often suppliers throw the left overs in the brace wood pile, and I’ve found HUGE variation in both density and stiffness from the same source.

[Paul M]

I guess I don’t think that the minute differences in bracing stock are going to make measurable differences, ultimately I’m tuning the braces towards target frequencies and very importantly, target deflection which I think will likely account for the vast majority of these miniscule differences in tiny pieces of wood.

Ultimately I’m looking for things that a) will allow me to improve my guitars significantly without b) taking tons of time as I’m of the opinion that the knowledge you get from finishing multiple guitars is far greater than you get from splitting hairs on one guitar and I have experience that seems to bare that out. I like Guiliano’s methods because it’s a small ecosystem of data selection and he’s made it accessible to someone like me without any significant math knowledge but most importantly it’s starting to become quick and seamless in my process. Ultimately I have a goal of what I want a guitar to sound like and act like structurally and I can’t really understand that vs the numbers unless I finish a bunch of them.

long story short: I’m measuring the density of my brace wood and comparing it to the density of nice Sitka spruce that I’m using for tops but I’m not getting nuts about it past that, I think that’s pretty reasonable.