Official Luthiers Forum!

Sorry to do it to you all again but...

It's often posted how spruce is the desired wood for the top and bracing due to it's stiffness to weight ratio. I'd like to understand that a bit more please, especially in regards to bracing.

I think I agree that the bracing is not only there for strength but for an effect on the top vibration. From a newbie point of view, if I was going to make a marimba, one of the last woods I'd choose is spruce ( this is sure to bring out someone who has built a spruce marimba! ), so the question that brings to mind is... Does the bracing really contribute to the sound in any way other than as a damping mechanism that can help to shape the vibration patterns of the top?

Next, what is the physical numerical property of wood that is looked at when determining the best species to use? I think it is stiffness rather than bending strength (any other value?). From the structural properties of wood, what other values make sense to look at when comparing wood types for braces?

Here's where I am leading to... if woods have very similar stiffness values, and one has a greater hardness value might it be more desirable to use the harder wood? My though train here is that a harder wood may have better acoustical properties than a softer wood.

I took a quick look at some properties of wood as listed on: http://www.thewoodexplorer.com/woodsearch.html
Initially I looked at red oak, white ash and Engelman spruce. I also looked at the bending strength, which I now think was wrong, and then looked at stiffness. Bending strength of oak and ash left spruce in the dust, but stiffness was won by spruce so that must be the factor people look at. I then looked poplar. It had a slightly greater stiffness and only a slight increase in weight.

Yellow Poplar: 1506psi / 29lbs/ft3 = 51.9 (weight to strength)
Engelman: 1438 / 26 = 55
White ash: 1720/42 = 40

So the question is: would the harder wood (poplar) have better tonal properties that would make it a better bracewood?

Please bear with me... I know that I'd be better off following the traditional methods, and I realize that there's reasons that people use the wood they do. I'm just trying to resolve the "why" of it all and if it doesn't come off a clear winner in MY head, I may try it so I DO know why the next time. I plan to do some basic, non-scientific stress testing with poplar and junk spruce (smashing things is always poplar with the kid) since I have them laying about in the barn.

Thanks for any input.
Rob & Robbie

I know people use engleman for bracing, but I believe that sitka and red/adi are much more prevalent - they are also stiffer for a given mass in general I think. Something that is not in the data is velocity of sound - sitka is very fast - 5000m/sec - it also has a good tap/ring tone .. I dont know about poplar ....

I am not so sure that bending strength is that important in bracewood ... seeing as its not really bent, like say a hoop back chair (oak) or a lacrosse stick (hickory)

_________________
Tony Karol
www.karol-guitars.com
"let my passion .. fulfill yours"

Your ratios are actually reversed - the numbers you gave are actually more like strength/weight and so higher is better.

For bracewood you want the same qualities as you want for the top - you want it to hold string tension, but also be free to vibrate.

I am not certain why hardness would be valued in bracewood, as it generally will not be subject to any kind of impact.

_________________
The member formerly known as erikbojerik....

Sound velocity? Is that a "real" measuement? Probably what i was assuming the hardness/density accounted for. Anyone know of a database that includes such a measurement?

Probably not something one normally needs to know when building furniture.

Rob, sound velocity is a real measurement, but not likely one you can make in you shop. I believe a lucci meter measure this. The other way is a scope and some accelerometers.
Sound velocity is related to the stiffness (modulus) and the density of the material,
it's actually the square root of the stiffness/density.
The takeaway information here is that stiff light things transmit sound faster than not so stiff heavy things.

_________________
Jim Watts
http://jameswattsguitars.com

You can calculate the "velocity of sound" in any piece of wood. It is always a function of the stiffness and the mass. This "velocity of sound" measurement is basically the same thing as the "first mode" of vibration also referred to as the first natural frequency. There are an infinite number of potential vibration modes, of which there are a few that dominate the tonal response of something like a guitar top.

The reason that the stiffness to weight ratio is important is because of efficiency. If you had a race care and wanted to go as fast as possible, you'd make it as light as you could, for a given engine size. Using spruce for soundboards and braces allows you to create the lightest structure that can withstand the forces of the strings...this makes it easier for the strings to activate the top and then to have good response in terms of volume and tone.

Trev

_________________
http://www.PeakeGuitars.com

Erik
"Your ratios are actually reversed - the numbers you gave are actually more like strength/weight and so higher is better."
No, they are the stiffness values. It's probably a matter of algebra and which number i put on top. Haven't used algebra since high school so many years ago!


Jim/Parser.

Perfect! My ears like what you posted. Now i need to actually plug in numbers to make sure my other brain agrees.

Thanks!

As a side bar: I think this is where dynamic testing for the modulus has the edge vs static testing. I static test and am happy with that, but if I try to match my results to the actual first mode I'd more than likely be off due to the inconsistencies and anisotropic nature of the wood.

So actually I retarct my statement that it's hard to measure, Al Curruths dynamic system of looking at the Chladni patterns will give it to you, well that some math.

BTW Trevor it's nice to know your name instead of just knowing you as Parser

_________________
Jim Watts
http://jameswattsguitars.com

So Al's the go to guy with Chladni patterns? I'm not too far from him so i may need to sign up for some of his class time in the future.

I hope you're just retracting your side bar statement and not the prior post, else i withdraw my understanding.

The Larson Brothers' seemed to be going for something other than optimizing the stiffness to weight ratio with their laminated braces. You might want to check those out.

Rob, all I'm saying is that it is easier to measure it that I initially stated.

_________________
Jim Watts
http://jameswattsguitars.com

Hey Jim, it's nice to be a bit more public now that I'm switching jobs..! The past few years has been fun and educational..

In my past experiences doing vibration testing, we expected a larger margin of error between predicted natural frequencies and the ones we were able to measure. The culprit was assumed to be the supports for the part being tested. (these were parts for drivetrain components, not guitar parts)

Have you been doing any more FEA work? I'm hoping for a full version of SolidWorks Simulation and a laptop at my new job...

_________________
http://www.PeakeGuitars.com

I have been doing some more, and I'll say my experience mirrors yours. I seem to get the static stuff right but have missed the mode frequencies.

_________________
Jim Watts
http://jameswattsguitars.com

As usual, the reasons for choosing one material over another for bracing involve a balancing of factors.

Braces are mostly there to add stiffness to the top without adding too much weight. If you look at the Chladni patterns of a bare spruce top they will be quite similar to those of a properly braced one, but lower in frequency. You could make the top thicker to get the stiffness and mode frequencies up, but then it would be heavy. Braces add the needed stiffness without as much weight. So it would seem that the first thing you'd like to se in braces is a high stiffness/weight ratio.

Young's modulus along the grain seems to scale pretty closely with density in soft woods, and the stiffness of a brace is proportional to the Young's modulus (E), the width, and the cube of the height. It follows that using a less dense wood will give you lighter braces for a given stiffness, since you can always make them taller to make up for the generally lower E value. So why don't we normally use balsa wood?

One problem with it might be that, while it glues well, it's not very strong. A sharp rap on the top could break the brace at the glue surface, and it would peel up. You could get around that by making the brace wider, so that it had more gluing surface, but that takes away some of the advantage of the low density. Also , balsa has low shear strength, and a low modulus of rupture; its not very tough. So it looks as though toughness is another thing we're looking for.

Spruce comes up pretty well in that respect. It's much tougher than some of the less dense softwoods, like cedar. This allows you to make a fairly narrow and tall brace if you want. Lute braces were typically about five times as tall as they are wide, and I take that as a practical maximum for top bracing. We seldom see braces that tall and narrow on guitars.

There's also the matter of availability. Maybe, if Torres or Strad had had access to balsa they'd have figured out how to use in in instruments. Spruce was the best stuff they had, so they worked out their designs to take advantage of it's properties.

Wood bracing has the disadvantage of cold creep: under even a small sustained load it takes a set. This can be countered by using something like carbon fiber, and people are working out the best methods of doing that.

Given that the total weight of bracing is far exceeded by that of the top itself I don't spend a lot of time worrying about finding the lightest possible brace stock. IMO it's far more important to get the top and bracing working well together than getting the lightest possible structure, although, of course, one does not want to build a tank.