Deflection testing on Adirondack braces... slab vs quartered

[j. Kinnaird]

I just read Robert Lundberg’s book on Historical Lute Construction. One of the surprising (at least to me) sections dealt with brace construction. Lundberg specified that the braces would be cut from stock that was slab sawn when viewed from above so that the load would be applied tangentially to the direction of the grain lines. I did some internet research regarding which orientation would make the strongest brace and found a lack of consensus so i thought that I would do my own test on a sample of red spruce. The results of this small sample size i realize cannot be generalized to all spruces nor even to all samples of Adirondack spruce but here are the results none-the-less

I cut a piece of spruce stock that was square in cross section and put it in my deflect-o-meter.



The end of the spar had a groove cut around it to hold the wire in the same location when the spar was rotated.


The other end of the spar was placed against a stop glued to the support board so that the exposed length would be constant.

A weight was hung from the end of the spar to cause it to deflect.


Since my intent was to test deflection in all four planes I labeled each surface as either slab or vertical, side one or side two.

I checked the unloaded starting point for each orientation and since the spar was quite straight the starting points for all surfaces was within about a 64th of an inch.

The first surface i tested placed the load on the slab surface. For photographic clarity i used masking tape to mark deflection depth. On other tests i marked the deflection with a pencil on the board.


With this slab deflection i cut the tape flush with the top of the spar.



Then i rotated the spar so that the load was applied to the quartered face and you can see for yourself the extra deflection that caused.
The top of the blue tape marks the deflection of the slab surface and the new location of the spar marks the extra deflection the quarter brace allowed.


I tested sides 2 vertical and 2 slab with very similar results.

What to make of this? I think I can imagine a mechanism for these results and it is something mechanically similar to an I beam with the hard summer grain forming the cross bars of the I beam and the spring wood forming the webbing.

How these two grain orientations would respond to brace tapering and brace scalloping is another variable that needs to be tested. It would be nice to be able to increase strength just by changing grain orientation of the braces. They could be Made lighter With the same strength and presumably Would be more resonant.

[Mark Hatcher]

Thanks for showing us your test John. I have gotten similar results of my own.
I’ve found much less stiffness when the grain is running at a 45 degree angle.

One area of concern when setting up brace stock is grain runout. It is possible to have runout that looks like it is well quartered which is why it is important to work from stock that is split out on both sides, both top and side.
Another large advantage of splitting out both side is it is much easier to carve the braces without getting any unexpected splits or chisel dives.

You end up with more waste wood by the time you split out the stock, but the question should be: do we want braces made from waste wood?

Mark

[j. Kinnaird]

Originally Posted by Mark Hatcher

Thanks for showing us your test John. I have gotten similar results of my own.
I’ve found much less stiffness when the grain is running at a 45 degree angle.

One area of concern when setting up brace stock is grain runout. It is possible to have runout that looks like it is well quartered which is why it is important to work from stock that is split out on both sides, both top and side.
Another large advantage of splitting out both side is it is much easier to carve the braces without getting any unexpected splits or chisel dives.

You end up with more waste wood by the time you split out the stock, but the question should be: do we want braces made from waste wood?

Mark

Thanks for that input Mark. It’s good to know that your results are similar to mine. I think you must be right regarding the runout thing and the benefits of splitting braces and it makes sense that a brace cut on a 45 degree bias would have less stiffness than either quartered or slab cut braces. I wonder if factory produced guitars concern themselves overly much with those kind of details. Probably depends on the factory.

John

[Guitars44me]

I am glad John does the research and I get to play his (our) guitars!

An excellent division of labor...

Interesting info abounds hereabouts

SALUD

Paul

[Tim36]

I have no idea how this could be tested, but I do wonder if braces have more resistance to warping due to environmental changes when glued on the quarter sawn side (grain going vertically) or the flat sawn side (grain going flat). I would have always assumed the vertical grain would be best, but this test ruins my assumptions. Thanks for sharing.

Come to think of it, I once received a bowed fingerboard that was perfectly quartersawn, and it was bowed along the vertical grain.

[j. Kinnaird]

Originally Posted by Tim36

I have no idea how this could be tested, but I do wonder if braces have more resistance to warping due to environmental changes when glued on the quarter sawn side (grain going vertically) or the flat sawn side (grain going flat). I would have always assumed the vertical grain would be best, but this test ruins my assumptions. Thanks for sharing.

Come to think of it, I once received a bowed fingerboard that was perfectly quartersawn, and it was bowed along the vertical grain.

Good question. Right now I’m leaving the brace under stress and will do so for a few days to test it’s resistance to creep over time

[TomB'sox]

If I am understanding things correctly, all the tests were done on the one piece of stock. Did you repeat the deflection test again with the same orientation with the same side and weight etc. to see if it deflected to the same place again. In other words, testing the same piece in different orientations could weaken the brace on different planes? I would wonder if after testing if you did it again, if you would find the same results with the same piece?

[Portland Guitar]

Excellent test!

[Alan Carruth]

Good test as far as it went.

I, too, suspect the hard latewood would add stiffness on a flat cut piece if it ended up at the upper or lower surface of the brace; it's like those CF-balsa 'I' beams that Smallman makes on his tops. If the softer earlywood ended up at the surface the result would be the other way. I suspect, then, that you'd see more variation in flat cut than quartered brace stock, but similar averages, but that's just a hunch. There's no substitute for sample size.

Classical guitar makers used to split the wood for the fan braces off the outside edge of the top blank. If the blank started out at, say, 5-6 mm thick, and you split them off before you thicknessed it, you'd have plenty for those low braces.

Lutes use a lot of tall, narrow braces, none of them so wide that you could not get them by splitting them from the edges of top blanks. I note that lute braces tend to be about five times as tall as they are wide; I suspect that's related to the strength of the glue line; much more than that and they tend to peel up.

Often,when thinking about the way they used to do things, I find it makes sense to keep in mind that good wood was expensive and hard to get, but labor was relatively cheap. They'd put in extra time to save wood.

[j. Kinnaird]

Originally Posted by TomB'sox

If I am understanding things correctly, all the tests were done on the one piece of stock. Did you repeat the deflection test again with the same orientation with the same side and weight etc. to see if it deflected to the same place again. In other words, testing the same piece in different orientations could weaken the brace on different planes? I would wonder if after testing if you did it again, if you would find the same results with the same piece?

There are 4 different ways to deflect that beam. I tried all four ways,rotating it so that first the quartered surface was up and then the slab surface, then quartered again and slab again. Each test involved rotating the beam a quarter turn. The results were consistent.

[j. Kinnaird]

Originally Posted by Portland Guitar

Excellent test!

Thank you sir. I might try to improve on the test by doing a larger sample size and different woods

[j. Kinnaird]

Originally Posted by Alan Carruth

Good test as far as it went.

I, too, suspect the hard latewood would add stiffness on a flat cut piece if it ended up at the upper or lower surface of the brace; it's like those CF-balsa 'I' beams that Smallman makes on his tops. If the softer earlywood ended up at the surface the result would be the other way. I suspect, then, that you'd see more variation in flat cut than quartered brace stock, but similar averages, but that's just a hunch. There's no substitute for sample size.

Classical guitar makers used to split the wood for the fan braces off the outside edge of the top blank. If the blank started out at, say, 5-6 mm thick, and you split them off before you thicknessed it, you'd have plenty for those low braces.


Lutes use a lot of tall, narrow braces, none of them so wide that you could not get them by splitting them from the edges of top blanks. I note that lute braces tend to be about five times as tall as they are wide; I suspect that's related to the strength of the glue line; much more than that and they tend to peel up.

Often,when thinking about the way they used to do things, I find it makes sense to keep in mind that good wood was expensive and hard to get, but labor was relatively cheap. They'd put in extra time to save wood.

You are right about sample size. Increasing it might factor out the location of the spring and summer wood. (And give more validity to the results)

[Portland Guitar]

How heavy was the weight that you used? It looked like a pretty big piece of wood. Given that the stiffness is linear it makes sense you'd want something heavy to test with to see the difference. I'm wondering how much stiffer the flat sawn beam is under realistic loads.

[j. Kinnaird]

Originally Posted by Portland Guitar

How heavy was the weight that you used? It looked like a pretty big piece of wood. Given that the stiffness is linear it makes sense you'd want something heavy to test with to see the difference. I'm wondering how much stiffer the flat sawn beam is under realistic loads.

Those are good questions to which I have no answer. I need to quantify the test. I’ll do that and get back to you.

[Steve Kinnaird]

Most interesting, brother. I'd like to see you suspend your weight from the center of each brace, while you support the ends. That should come closer to simulating the brace under string load. And to be really curious about the deflection, you could mount a dial indicator over the brace and get your results in thousandths.

I know...this is feeling like a can of worms you just opened, right?