Official Luthiers Forum!

I'm currently wading through Somogyi's books - the Responsive Guitar and Building the......
What fantastic resources. I'm getting an amazing amount from them, and he's a great writer, so they're both really good reads.
In one of them he presents some results from a study he did on deflection of brace stock relative to grain orientation. What I got from his analysis was that braces are actually stiffer parallel to the grain than perpendicular to the grain, which, at least to me, is counterintuitive.
So .......what's your take on this? Does this data imply that one should orient the brace stock with grain parallel to the top, rather than perpendicular to the top? Are their other impications to doing so? (Wonder if over time the wood might split along the grain lines, especially on a domed top?)
Thanks.

I don't know about the mechanics involved that would make a brace stronger if it was flat-grained, but I'd hazard a guess that it has something to do with the late-growth wood acting like the flanges on an I-beam or H-beam, though in my limited testing the difference was not large in the sizes we use for bracing.

Using vertical grain for bracing makes them less susceptible to splitting from impact at the tapered ends where there is end grain exposed. Also makes for an easier time when carving braces.

Pat

_________________
formerly known around here as burbank
_________________

http://www.patfosterguitars.com

The data does not indicate that this is true for all spruce, only spruce of a certain density.

A rather telling observation is that Somogyi still places his braces in a vertical grain configuration.

In one of his recent e-news letters, Siminoff showed golden era Martins that use rift sawn bracestock and he felt that that orietation was the best of both worlds. I know that my customers still demand that I take exceptional care to saw my brace stock bang on quarter!

Shane

_________________
Canada

If memory serves, the original article about has a bit more info and is on his site.

I'm betting that, as well as corresponding to density, results may also relate to brace height. I bet the shorter the brace, the less favorable the flat sawn would be in comparison. That's just a guess though, based on playing with a few samples.

I asked Ervin about flat-sawn braces being split-prone and he didn't answer. I then did some less than exhaustive tests with Engleman and found that flat performed as well or better than quartered. I believe it was Kevin Gallagher that reported the opposite with Sitka.

The bottom line for me is that there are a lot of variables and no clear benefit to one or the other. So why buck perception. If one person looks inside one of my guitars and and has a fit because the braces aren't quartered then it's not worth any marginal, questionable gain. If it was a clear advantage, I'd be willing to buck the tide.

_________________
http://www.chassonguitars.com

So here is an excerpt from Roger's e-news letter. The picture quality is poor cause I couldn't seem to grab it from the letter so I took a picture of my computer screen.

Shane


Luthier's Tip: Selecting wood for braces and tone bars

The strength of a brace or tone bars is highly dependent on the direction of the grain through the wood. There are three possibilities: vertical grain, grain on the quarter*, or flat grain.

Vertical grain (grain going up and down) is strongest, but is also much too stiff for braces and tone bars. However, one good feature of vertical grain is that braces and bars can be very easily and consistently shaped without slicing or cracking through the grain.

Flat grain (grain going side to side) is the stiffest, but it is also very brittle and prone to checking or cracking through the grain (something you don't want to happen to a brace or tone bar inside the instrument). (Note that it is stiffer than vertical grain, but not as strong from a reliability standpoint.) Also, shaping braces with flat grain can be problematic in that the wood often does not cut with similar ease in both directions.



Grain on the quarter is ideal and is the best compromise between flat and vertical. As shown in this photo of the intersection of X-braces in a Martin guitar, the grain on these braces is on the quarter.

*Note that wood cut "on the quarter" is different from "quarter sawn." Cut on the quarter is a grain direction. Quarter sawn is a log cutting method.

_________________
Canada

So on the quarter is basically 45 degrees? I always thought you were supposed to brace with the grain vertical.

I beliebe that, with most pieces of spruce, you'll find the optimum orientation would be quarter sawn
with the grain lines running perpendicular to the top surface for strength and rigidity. It will also
provide the best integrity at the ends of the bracing where they are tapered down to meet the
perimeter of the guitar at the sides.

Flat or rift sawn pieces can be compromised with a blow or pressure and the grain that is parallel
or close to parallel to the top surface can split and fail while flexed.

You certainly can't argue with the expertise of Ervin on these things, but I do believe that his data
on this topic was intended for Spruce brace stock of specific density.

Regards,
Kevin Gallagher/Omega Guitars

I read quite a bit of published Modulus of Elasticity info talking about grain direction vs stiffness...

The published info I was able to dig up basically says that Pure flat sawn and Pure quartersawn are the stiffest orientations... and the 45* grain orientation is basically the most flexible orientation. It is however much more split resistant.

Consider that those "Golden Era" instruments made just prior to and during WWII had to contend with their Top wood species basically going commercially extinct -- Both Red and White spruce in the Eastern USA were logged into commercial extinction by the early 40's...

Imagine being a Shop Supervisor at Gibson or Martin during those days:
The Spruce you were used to getting and had relied upon was basically gone...
The stuff you could get was very low quality... because you had to contend with the War Effort... and the extinction of the Old Growth American spruce forests in the East....
And.. most of your workforce was drafted to go fight.

You had to make due with what you could get... and so we see Guitars coming out of the factories with funny off-quarter tops, wide grain tops, mismatched tops, Bracewood that went every which way....

They may have been cutting that brace wood off quarter because it had really bad runout and they were trying to prevent splits... Who knows.

And yet, they still made some outstanding guitars.

Thanks

John

Somogyi had his results of testing brace stock published years ago in one of the Big Red Books. That being said I don't think you will find that he advocates using any thing but vertical grain. We should not try to "unlearn" anything from this data,also in my mind this is a very small sample when you consider the HUGE variability within wood stock.We are still on track when we use brace stock in the normally accepted fashion,at least in my mind.

_________________
A person who has never made a mistake has never made anything!!!

Yeah, I remember talking to Ervin about his, in short he said he still prefered using quartersawn stock....if I remember correctly, part of it was because carving/shaping the braces was much easier with quartered stock.

_________________
Gwaltney Guitars

Shane, Am I reading your quote from Rodger Siminoff correctly?
He seems to be saying both that vertical grain timber is much too stiff for braces and that flatsawn timber is the stiffest.???
That on top of the use of the term "on the quarter" for what most people call riftsawn are very confusing

Flat cut wood is stiffer than quarter sawn wood. This is why tops are quarter sawn. braces should also be quarter sawn for the same reason.

As far as I understand, the reason tops are quartersawn is that radial shrinkage and expansion is much less than tangential movement. A flatsawn top would be unstable and crack prone.

Flat cut tops would be more prone to warping. However,a test in FWW years ago proved that flat cut wood is stiffer than quartered. I did a test myself,and verified it using Sitka spruce. I planed 2 thin square cross section spruce rods about 3' long each. When held in identical ways at 1 end,with a weight hanging off the unsupported end,the piece when the grain was horizontal deflected less than when it was turned vertical,with the same weight added. Both test pieces acted the same.

So,it seems serendipitous that we can enjoy the much nicer appearance of quartered spruce,with its beneficial properties,and enjoy better tone from it also. Wouldn't tops look terrible if they had to be made of flat cut wood?

I really think Tradition also has a big contribution here...

I did some reading on and found out that historically "Figured" and flat sawn wood was generally rejected for "Good" furniture as being too unstable, while plain straight grained Quartersawn stock was considered "Premium."

This was long before the days of "Climate control" so most furniture lived in an environment of constant giant humidity and temperature changes... so this probably made perfect sense -- you didn't want your dining room table splitting in the winter when you are heating with Coal...

Thanks

John

Hi Corky,

I did some bracing stock deflection tests on four different species of wood a few years ago: Western red cedar, koa, Douglas fir and redwood. All told, I measured about 35-40 different samples. In each case, I measured the thickness in both the quartersawn and flatsawn directions, measured the deflections on a standard apparatus with the same weight and so forth. Then I calculated the parallel to grain Young's modulus for each grain orientation. So here's the data with quartersawn Young's moduli vs. flatsawn Young's moduli, and a 1:1 solid line drawn through the data for effect:



I interpret these data to show that for these samples, if the dimensions are the same in both flatsawn and quartersawn directions, then the stiffness is essentially the same as well.

I also suggest that experiments testing stiffness in various grain orientations with only a limited number of samples can often distort a significant trend because of statistical noise.

aloha,

Dave Hurd
http://www.ukuleles.com

_________________
David C. Hurd
http://www.ukuleles.com

Alice laughed: "There's no use trying," she said; "one can't believe impossible things." "I daresay you haven't had much practice," said the Queen. Through the Looking Glass by L. Carroll

"Since my house has burned down, I have a much better view of the moon"

Hi Filippo,

For the tests I ran, all the deflection tests were done along the grain of the wood - two sets of them on square cross-section pieces - with the deflection test done on one piece, the piece then rotated 90 degrees along the long axis and the test repeated. So the orientation of the grain relative to deflection never changed.

In a cross-grain flexing test that you describe, the flexing occurs perpendicular to the grain. I like to think of trying the same thing with a piece of corrugated material that doesn't have the surfaces covered with surface sheets. That is, corrugated cardboard without the smooth surface paper on either surface. The Young's modulus across the grain in many woods is many times less than that parallel to grain, but interestingly enough, in koa and I believe mahogany the moduli are much closer to one another.

Perhaps warmer sounding instruments are made from woods which have less of a stiffness contrast (all koa or all mahogany) and brighter sounding instruments are made from woods with greater stiffness contrasts (hardwood back/sides and a softwood such as spruce for the top). This might be another topic for an interesting discussion...

aloha,

Dave Hurd

_________________
David C. Hurd
http://www.ukuleles.com

Alice laughed: "There's no use trying," she said; "one can't believe impossible things." "I daresay you haven't had much practice," said the Queen. Through the Looking Glass by L. Carroll

"Since my house has burned down, I have a much better view of the moon"

Another semantic jungle!

If you take a spruce top that looks like this on the end:
_____________________
||||||||||||||||||||||||
------------------------

and test the stiffness (Young's modulus) for bending along and across the grain, the ratio will be fairly low: the long grain stiffness will be something like 8-10 times as high as the cross grain. For wood that looks like this on the end:
__________________________
__________________________
__________________________
__________________________

the cross grain stiffness will usually be a little lower, with the ratio being something like 12:1

Wood like this:
_________________________
/////////////////////////////////
__________________________

with the ring lines at 45 degrees, may be only 1/100th as stiff across the grain as it is along the grain. Cross grain stiffness in a top depends more strongly on grain orientation than anything else, by far.

Hardwoods have lower stiffness ratios in general, and some, like Osage orange, seem as stiff across the grain when skew cut as they are when well quartered.

None of this has any direct bearing on stiffness along the grain, as far as I can tell. It's possible, as has been said, that on softwood wood that grew very fast, and that has thick latewood lines, that those could act as the flanges on an I-beam if they fell near the upper and lower surfaces of the brace, and you'd end up with significantly higher long grain stiffness measurements _for that brace_ than you would for a brace of the same stock cut with the ring lines perpendicular to the gluing surface. It's also just as possible that you'd end up cutting the brace with the soft earlywood at the upper and lower surfaces, and a significantly _less_ stiff brace would result. The perpendicular orientation may not take the best advantage of the hardness of the latewood grain, but it probably gives more consistent braces, particularly for fast-grown wood.

I'll note that in my tests, the Young's modulus (E) along the grain for all softwoods that we use for tops can be predicted pretty closely from the density. If the density is around 300 kg/cubic meter, the lengthwise Young's modulus will most likely be within 10% of 6000 megaPascals, and if the density is 500 kg.m^3, the lengthwise Young's modulus will be within 10% of 17,000 mPa. That's from about 50 samples of seven different species: Euro, Red, Engelmann, and Sitka spruce, Western Red Cedar, Redwood, and Western Hemlock. There is somewhat more of a tendancy for samples to fall 'below the line', to be denser than they 'should' be for a given stiffness, than 'above the line'. A plot of Elong vs Ecross is almost a perfect scatter: cross grain stiffness is all over the place. I'll note that in my tests Sitka does tend to have lower cross grain stiffness than the other spruces, but I don't have a lot of samples. Larry Stamm, who has tested a lot more wood than I have, has said agrees with this in general.

Having had my hands in a fair amount of Martins and having looked in the archives , Braces in Martins older vintage instruments was not always quartered , and can be seen in any configuration of grain orientation. I heard great and bad guitars with braces both ways. I personally feel the wood is more stable on the quarter .
Most of the spruce was depleted for the war effort but there are pockets of red in NY that are doing well but won't be harvestable in our time . Thanks again Alan for a good explanation and the data .

_________________
John Hall
blues creek guitars
Authorized CF Martin Repair
Co President of ASIA
You Don't know what you don't know until you know it