Official Luthiers Forum!

I know it is a very difficult question to answer to: If I understood properly, poeple working to stiffness instead of thickness made up their own standars depending of their model.

Up to now I've always worked to thickness and then adjusting until it felt "right" to my teachers (as a beginner I couldn't do it) But is their some kind of average to respect using this technique (like between this and this) or does it REALLy differ that much depending n the builder and his model?

I don't have the answer to your question. I haven't found that info 'all in one place' anywhere.
'Targets' for thickness/stiffness are quite different for nylon and steel-string guitars, and the body size makes a difference as well.
A search on 'Deflection testing' will turn up some information.
The Yahoo group 'Left Brain Luthiers' has participants interested in this sort of thing:
http://launch.groups.yahoo.com/group/leftbrainluthiers/

Testing wood deflection before the start of construction is fairly common; some builders also test the deflection of the top of finished guitars.

Cheers
John

Kent Everett demoed his method in his voicing DVD, and I found it particularly helpful because it lets you start to get a sense for stiffness with a concrete cue in toe. He sends it through the sander, holds the top at the sides, and wiggles it back and forth. If it makes a fluttery saw blade type sound, he stops. If it doesn't, he sends it through again and again until it does.

Doing something like working the top to a certain stiffness takes a lot of experience and "reps" before anything actually gets out of your short-term memory. If it's something you do frequently, great, but if not, Kent's method accounts for the variability in each piece of wood but also gives us a certain threshold to find.

I've been testing the wood I get in to find the Young's modulus along and across the grain. This allows me to more or less predict the stiffness of the top at a given thickness, and, conversely, to determine the 'right' thickness for a given top. I use a simple 'index number' system.

First, as always, you need to have some sort of model of how you think things work, so you can figure out what might be important. To my mind, the thing that limits how thin you can make the top is the static stiffness needed to resist bridge torque over time. This will obviously vary somewhat depending on things like how big the box is, and how high you like the strings to be off the top. I also simplify a lot by assuming that, over time, the cross grain stiffness will not be much help, since cold creep of the top will allow it to distort into a permanent 'set'. A less floppy top will distort less over the short term, but eventually they all seem to get to the same shape. At any rate, this simplification allows me to concentrate on the lengthwise Young's modulus.

The other thing you need to make this work is a successful guitar where you know both the Young's modulus and thickness of the top. The best is to use one of your own where you made all the relevantmeasurements as you built it. However, you could probably get away with using 'average' values from a bunch of production guitars, under the assumption that, if you look at enough of them, the material properties will approach some mean value.

Let's suppose that, by whatever method, you've decided that for a given model of guitar, top wood with a Young's modulus of 10,000 MegaPascals (metric measurement) along the grain will be stiff enough if it's .110"thick (2.8mm). Using a deflection or resonance test you figure out that the wood you have has a lengthwise Young's modulus of 8000 MPa. How thick should that top be?

Remember that the stiffness of a beam goes as the cube of the depth. I cube the thickness of the 'good' top and multiply that by the Young's modulus to get an 'index number'. since I work in metric, that would be (2.8)^3 x 10000= 219,520. Now, divide that index number by the Young's modulus of the new wood, 8000=27.44. That's the cube of the thickness, in mm, that the new top should be. That's just a little more than 3mm, or, say, .120".

I realize that there's a LOT of room in this for the introduction of higher math. Those of you who are good at that sort of thing are welcome to do so. I beleive that this system is reasonably conservative: you'll probably err on the side of too thick rather than too thin, and that's probably better in the long run. It's a place to start, anyway, and probably a bit better than simple guesswork, for those of us who lack 'the feel'.

That sounds a bit more complicated than I thought: LOVE IT

Now here is another tricky question: I heard some poeple "tune" their soundboard and back (I believe it is done quite precisely in the violon's familly too)

Now: lets say you want to tune your top or back, for a specific model. If you use the same deflection, at each time, for the very same model, meaning the same dimensions too, does that mean you will get the same "frequencies" out of it, or if not the same, something close (prior to bracing)?
Or is it more complicated than this?

Does any one know a formula to calculate the frenquency (something general for any material given, not necessary top/back)?

If building to get the same deflection, the result will probably have different dimensions and even different freqencies, because each chunk of wood is different.

_________________
Dave
Milton, ON

Al,

I've read about your method many times and I definitely need to start using it for my banduras. Question for you though, do you have any idea what level of accuracy and precision you're getting on this?

I guess what I'm asking is, let's say you think you want to make a top 10% stiffer than your last one for whatever reason. Are these methods good enough that you can actually make a 10% change or is your variance big enough that 10% falls in the noise?

_________________
http://www.birkonium.com CNC Products for Luthiers
http://banduramaker.blogspot.com