Official Luthiers Forum!

I didnt read all the posts, so someone may have made this point, but I didnt see any numbers.

Dims:
Width (b) = 0.50
Height (h) = 0.625
Cutout height (h1) = 0.345
Web thickness (tw) = 0.160

Rectangle
Ix=0.01017
A=0.3125

I-Beam
Ix=0.00942
A=0.1952

Triangle .5x.625
Ix=0.00339
A=0.15625

Triangle to match the area of the I-Beam version. Needs to be about 0.78 high.
Ix=0.00659
A=0.195

Ix is the second moment of inertia, A is area.


I had run the numbers in the following post when I was messing around and made a I-Beam brace. Grumpy claimed it didn't improve the strength to weight, so I ran the numbers along with a triangle. In short, a triangular brace with the same length and width will be about 1/3 as stiff as the rectangle at 1/2 the weight. Numbers dont lie.

viewtopic.php?f=10101&t=15775&hilit=beam

I'd have to run the numbers, but I believe that a tall skinny rectangle of the same area as a triangle, the rectangle would win. The main goal in a beam is to get as much of the material away from the centerline as possible. I-Beam is best, and a triangle isnt that great as you have one end that has very little material out there.

However, there is a practical lower limit to the minimum width of the brace. If you start with any width, the triangle will always win when its height is adjusted to get the same I.

BTW, I've been known to I-brace my Braces

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

What a great discussion. I'd love to respond immediately, but frankly, I can't. Too much to process.

Even if you were to figure out the ideal profile (Alan, I meant the profile across the width like you correctly conjectured), then you have to decided if it's something you can do, and if it's something that positively impacts the responsiveness of the guitar.

Starting from a rectangular brace and carving it to a FULLY triangular shape will reduce it's bending stiffness to 1/3 of it's original stiffness.
Whether that is a bad thing is another question.
With an X braced top the area I want really rigid is the X brace intersection. So I leave it rectangular and cap the joint.
Carving the crossection and scalloping are ways of providing increased flexibility in areas where it is needed to allow the top to vibrate.
Designing and building a top that is really stiff is easy, that is your cheap factory guitar.

you can throw a rock and hit 12 engineers around here

_________________
--Kyle Burner--

Ahhh, something just hit me it the back of the head.

Todd had me thinking and wanting to make sure I wasn't wrong. I wasn't wrong...and neither was he. We were comparing apples to oranges. First off it looks like my numbers I ran way back then were in relation to the center of rotation, not the bottom axis. If you take it from the bottom (max width) axis the triangle is 1/4 the stiffness. The numbers differ depending what axis you are referencing.

Todd is correct. Take that same width and made the triangle taller until it is just as stiff, and the weight is about 70% as the original rectangle. About .90" tall in my example...as Todd said. (ref center of rotation)

To further complicate it, take that 0.9" tall and make it a rectangle. Reduce the width until it is as stiff as the other two. The new tall skinny rectangular brace just as stiff as the previous two is less than 70% the weight of the tall triangular brace and less than 50% the weight of the short rectangular. (ref center of rotation again)

We could run variations for a long time.

This does have me thinking about what axis reference should be used for braces. Also when doing theoretical stiffness of braces, these numbers are really a component of the entire top calculation. For this reason, I've always thought peoples idea that of cutting a hole in the upper transverse brace to access the truss rod is a weakness, its not nearly the weakness that they think. I cut my hole close to the top, so when it is glued on, its close to the center of the top+brace thickness. It that point, there shear stresses at the area where the hole is are small. Think about that one for a while.

Well, an I beam is going to give you better stiffness to weight than a rectangle, parabola, triangle,etc. And you can carve them out into trusses like Kevin Ryan and do even better than an I beam. So if your goal is to get the maximum stiffness to weight,

http://t0.gstatic.com/images?q=tbn:ANd9 ... cKWeeIKeg=

But the question whether the guitar will sound better is left unanswered.

_________________
Howard Klepper
http://www.klepperguitars.com

When all else fails, clean the shop.

If you guys will permit:

What would happen if we cut circle holes or ovals in the brace? With a laser it would be very easy to purposely cut meat away from a brace by adding holes correctly spaced and sized without weaking the brace. I don't think that it would change the load carrying dynamics a tremendous amount but it would furhter lighten the brace and the top.

Further why not design a brace that is similar in design to a truss? (Several companies now doing this.)

Thanks,

Kevin Waldron

Just like a Nebraskan....shows up to a potato cannon fight with a rock.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

A couple of points.

First, the bottom of the brace is glued to a top/back with much more mass and stiffness. If one assumes the glue doesn't fail, I imagine we can ignore the loads on that side of the beam when considering stiffness in this direction. Going back to Zach's thought, seems this might make the case for using the distance from the bottom of the brace to the top in calculations instead of the the neutral axis........where is the neutral axis when the brace is permanently attached to the top/back?

Second, I don't think the loads in this direction are the only thing to consider. Loads in this direction are very important when considering how stiff the soundboard/braces should be to resist the pull of the strings. However, when it comes to affecting tone, it seems the braces also twist. How much difference is there in a rectangle or triangle cross section when it comes to resisting a twist? Seems all the dipole modes have a twisting action to the soundboard (I hope I am using the term dipole correctly here).

Here is a link to an iteresting video by John Platko using FEA software that shows the movement of the soundboard in several diggerent vibration modes. The movement of the soundboard is exaggerated so it can be easily seen. Note how much the sounboard and braces twist. Surely chaging the twisting stiffness would also affect tone.

http://www.youtube.com/watch?v=TvxxqViwu9Q

_________________
Formerly known as Adaboy.......

IRRC, Al Carruth noted one time in a thread about top weights that in his experience there is a range of weights for tops, something like 150 - 190 grams, and a range for braces, 25 - 40 grams.

I understand that top + bracing = system, but removing 3 grams from the bracing is about a 10% reduction, isn't it? Not trivial, with respect to braces, if you ask me.

The message that I take away from Al's ranges is that the real weight is in the top itself. Don't forget to graduate that, as well.

L

Another approach.


I do agree with Howard. Theory is great and all, but who says lighter sounds better all the time. I thought Kent Everett said something in his voicing video to like he wants to get as much wood moving as much as possible. I interoperated it that lighter isn't always better, but getting it flexible in the right spots to allow the string energy to move as much wood as possible.

I'm sure the FEA guys could give more accurate info on close to real life numbers. I agree that torsion is also a variable, and the different cross sections will act very differently in torsion. Especially with wood.

Well that didnt work. Go look at the Batson bracing.

Come on guys, this is getting too theoretical and complicated for a simple question! Who fully triangulate their braces? I certainly don't, I shape them a bit and round them on the top. I stand by my earlier assertion that for short bracing, it really doesn't change anything. Again, for those using thin(ner) tops and tall bracing, it is a different story and I wouldn't know.
In the style of building I use, the biggest difference one can make besides the height is how much and where the braces are tapered (or scalloped). That's valid for all braces: X, tonebars and finger braces. In any case I assume everybody leaves the x-braces square at their apex. I consider the UTB as a purely structural brace, and not a "tone shaping" one.

_________________
Laurent Brondel
West Paris, Maine - USA
http://www.laurentbrondel.com/

I do, but then mine have rosewood running through the middle.

It really helps to understand the basic principals behind how things work to more efficiently play with the variables. If you dont have a real understanding about the affect a change will make, your just throwing darts into the dark. It takes a long time to try out one little change, so why no make the most educated guess on what to change.

I guess the simple answer after understanding what an efficient shape to a brace is, how else are you going to tinker with the top. Kent Everetts method would be the alternative.

Isn't that what we're here for?

BTW, on my first acoustic 33 years ago I carved the X to a central point (like the intersection of two gothic arches). No structural purpose, but it looks cool.

As Cicero said, "Nothing so absurd but some luthier has tried it."

_________________
Howard Klepper
http://www.klepperguitars.com

When all else fails, clean the shop.

Hehe....but Cicero also said, "Natural ability without education has more often attained to glory and virtue than education without natural ability." Was he being anti-theoretical?

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

Andy and Todd, good points on the width issue. Still, there is a pretty common piece of advice out there (including in this forum) that goes ~"you can trim rectangular braces into triangles to lose weight without losing stiffness." I do see a few people using tall triangles--maybe more will now.

Andy, Nice work on the I-braces!

Zach and Daryl pointed out the neutral axis issue for moment of inertia calculations. I made a spreadsheet to find the neutral axis and overall inertia for the T-section formed by the top+brace. For a 1/4" x 1/2" rectangular-section brace glued to a typical top, the neutral axis is roughly 1/64" - 1/32" below the glue joint (1st pic below, row 22). The neutral axis ("y-bar") hangs around the glue joint pretty closely for other sizes and cross-sections, due to the comparatively massive top area.

The spreadsheet is available at http://www.malicky.com/davidm/Guitar.T.Inertia.xls (I'd welcome error checking if someone wants to.)

The spreadsheet also compares 3 cross-sections: rectangular, triangular, and parabolic (3 tabs at bottom). I tried the assumption of constant width braces (row 8, b, 0.25"), and a constant inertia for the whole T-section formed by the top+brace (row 23, I-total, 0.0114). These are the required heights of the brace (row 9, h), and the resulting mass of a 16" long brace that has a uniform cross-section for its entire length (row 20, column M):
Rectangular: height = 0.500" (baseline) mass = 14.7g (baseline)
Triangular: height = 0.788" (58% taller) mass = 11.6g (21% lighter)
Parabolic: height = 0.647" (29% taller) mass = 12.7g (14% lighter)

So, if brace width and overall stiffness are held constant, the triangle has the least weight, by about 3g, or 21%. That triangular brace needed to be over 3/4" high to have the same stiffness as the 1/2" rectangular brace. A true triangle seems to me to be a risky cross-section, since the sharp point puts solo wood fibers in the highest stress. Parabolic is less risky and saves most of the triangle's weight.

If brace width and height are held constant (1/4"x 5/8"), here are the comparative overall stiffnesses for the combined T-sections (top+brace):
Rectangular: baseline
Triangle: 68% drop in stiffness, saves 50% in mass for the brace itself, and saves 16% in mass for (brace + 3" width of top)
Parabolic: 48% drop in stiffness, saves 33% in mass for the brace itself, and saves 11% in mass for (brace + 3" width of top)

So Todd has it right: the triangle (and parabola) only make sense if you make them sufficiently taller than the rectangle. Basically, height trumps cross-section. Shaving a rectangle "to save mass while keeping stiffness" just doesn't pan out.

As Howard said, we can do better than these cross-sections we wanted to. It would be easy to make an I-section with a table saw -- that would be lightest of all, and it maintains glue width. And as LPMc and Alan have said, most of the mass is in the top. A lighter density spruce for a top would save many grams for the same deflection test.

_________________
David Malicky

Additional question about bracing...... several guys are now steam bending backs and tops to molds and using a greater compound arch radius. They are also using only a single tone bar style brace on the top and no bracing at all on the backs similar to cello or double bass. Taylor in some of the less expensive models uses a rear back laminate with a slightly bit more arc without bracing.

Whats' the thinking from the engineering guys, sound guys, the old timers or anyone else who thinks themselves qualified to answer?

Kevin

Seems to me that I recall Stu Mossman of Mossman Guitars was working with I-beam type bracing . Not sure if it was experimental or standard fare for his guitars. Any owners or ex employees out there...???
Tom

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

Certainly I-beam braces should be lighter for a given stiffness. That assumes you know exactly what stiffness you need at each point along the brace so that you can design it correctly. If you just make a 'generic' I-beam, you've lost a lot of the ability to 'tune' the stiffness of the brace that a normal solid one gives you. I have to wonder if the small saving is weight would make up for the lack of 'tunability'.

I suspect you're right that there's more than one direction to consider.

John

On I-beams, triangles, etc., I'd agree that the weight differences among any of the brace cross-sections is pretty small. If it were important, one could tune the I-beam shape by selectively narrowing the exposed-flange width, and if needed, shortening the web. Still, it's less linear than shaving a regular brace. With some inertia calcs, one could design the lengthwise-profile of the exposed-flange width, to match a known tapered or scalloped profile of a standard brace. Then fine tune from there.

The torsion properties ('J') of the braces are interesting, too. I'm not sure if one would want more J or less... more 'I' (for a given mass) is good for responsiveness, as we need a certain 'I' to resist the string torque, among other things. But it might be helpful to free up the higher modes, which would mean less J. John P, it would be interesting to see the change in higher mode freqs and amplitudes from micro-slitting the braces lengthwise (that would keep the same I of each brace, but drop its J by more than half).

I checked the J (polar moment of inertia) of the cross-sections posted earlier (which give equal 'I' for the top+brace combo):
Rectangle: 0.0033 in4
Triangle: 0.0037 in4
Parabola: 0.0034 in4
Some differences, but those numbers are all dwarfed by a 3" wide by 0.110" thick piece of top: J = 0.2478 in4
So locally, it appears that J is dominated by the top (near 100%), while 'I' is dominated by the brace (~80%).
Across the whole top, though, the joined X-brace will give torsion resistance simply from its X geometry and the 'I' of each brace... but that may be more a lower mode issue.

_________________
David Malicky

Actually, I think it would be linear. Shaving a regular brace is non-linear.

(not that it really matters in the big scheme but just saying....plus I had never thought of that before and you're giving me ideas)

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