Official Luthiers Forum!

I hoping some of you out there can give me some tips to try some Chladni testing.
I've been doing some finite element modeling and am hoping to see how the actual results stack up against my model. I suspect they'll be off primarily due things such as damping, but it'll be fun to see.

My questions are very basic at this time such as:
1) What do set the free plate on when testing it? I'm assuming it's not just resting on the bench top.

2) Is there a typical support arrangement that works better than others?

3) How many watts of power do I need.

4) How do you pronounce "Chladni" (did I even spell it right?)

I was planning on down loading the sweepgen software and running that through an old stereo amplifier to a 6 or 8 inch speaker. Will this work?

TIA

_________________
Jim Watts
http://jameswattsguitars.com

I don't think you want to test a free plate. If I were setting up to do this, I'd want to:

(1) build a two sided (sandwich) form to clamp the rim around the edge to about the
depth that the linings would be glued, or;

(2) test the top after it's glued to the sides, using the side mold to support the assembly.

That way you'd be closer to examining the resonances of the finished box.

But Al C. would be the person to listen to, once he responds.

_________________
Jim Kirby
kirby@udel.edu

Take a look at Brian Burns' page on his website; Lessons in Lutherie. He has developed some interesting fixtures to test response and modes with Chaldni patterns and also impulse measuring software. There are some nice photos that are self explanatory.

http://lessonsinlutherie.com/woodtestin ... hirez.html

_________________
Peter

You set the plate on four foam pyramids. Mine are about 2 inches tall. The arrangement that works best is determined by which mode you are checking. When you start to get a resonance, the glitter will start to dance around and move into node lines. These lines will intersect the edge of the top. Move the foam supports to these node line positions. Make sure you choose positions that will still support the top.

I have my signal generator feed into a 10w guitar amp. I use a 5 inch speaker. I hold the speaker over the top and move it around to help excite various part of the plate. Make sure that you use hearing protection since sine waves don't sound as loud as they actually are.

I pronounce it "clod knee". Though clod knee sounds like a medical condition.

Most off the research has been done on braced free plates. Once the plate has been mounted to the rims, you will almost always get the same patterns. Think of the rims as 4 inch tall braces that effectively stops all top vibrations. The working hypothesis is that, if you tune the plate before it is mounted to the rims so that it has a pattern that is close the pattern you get after it is mounted to the rims, than the braces won't be fighting the motion of the top that the rims are forcing it to make. This should make for a more efficient top. We don't really know if that is the actual reason Chladni tuning works but it is an answer that makes sense to me.

Jim,

I agree with JT that Al is the one to talk too. He just gave the NEL a presentation that Mark Blanchard presented and was fascinating. Hopefully he'll chime in here.

-Joe

_________________
Joe Breault
Merrimack, NH
Perpetual novice

I think that the fixturing depends on what your goal is. If you are trying to analyze the top by itself, then work on fixturing the top in a free state (i.e. unrestrained boundary conditions). I would want to analyze the top by itself before jumping into analyzing the whole assembly, or even before looking at a restrained top.

I would try suspending the top from a string somehow (maybe drill a small hole up near the top in an inactive part of the top?). You'll need some type of accellerometer or transducer. I'd mount this at the bridge location. You'll have to hook that up to some sort of oscilloscope (or scope software)...then you can simply "bonk" the top with a mallet. The thing will vibrate and you can use an FFT function to pull out the dominant modes and amplitudes. I think it would be important to mount accellerometer/transducer at the bridge location as this would allow you to read the max amplitude and compare it to your FEA results. It might be worth it to remove some of the wood from the bridge that would be equivalent to the mass of the transducer, depending on how small of a transducer you can find.

Really cool stuff...I'm REALLY looking forward to your posts. If you post some videos of the modal results, I'm sure you will really blow everyone away...!!!

_________________
http://www.PeakeGuitars.com

My experience tells me:

Supporting the top on foam blocks works great. There is no need for elaborate suspension methods. The exact position of the blocks doesn't matter. The top is going to virbrate the way it wants to vibrate regardless of where the supports are.

15 watts is plenty of power.

The simpler the system, the more complex, variable and useful the Chladni data will be. For me, running modes on an unbraced free plate provides the most useful information. Very little useful information can be gained by looking at Chladni patterns with the edge of the top retstrained.

Producing nice clean pattern shapes can help improve the efficiency of an instrument, thereby making it more sensitive and possibly louder. It will not, necessarily, make it sound "better". If your building style tends to create a brash, nasal, tone, then refining the Chladni pattern shapes will probably just result in guitar that does that more efficiently.

Manipulating the tone of an instrument requires changing the relationship between the modes. That is done by changing the structure. Once you figure out the Chladni relationships that are indicative of a tone that you like, They can help you detect, and compensate for, natural variations in the materials that you are building a particular guitar from.

Mark

Don't know when Al's going to get here, but what MikeM and MarkB describe is how I learned to do it from Al and their explanations are very similar to his. The reason the foam pyramids are moved to the nodes is just to give the top a little more freedom to move (since the nodes don't) which makes the patterns a bit sharper.

_________________
Bob Garrish
Former Canonized Purveyor of Fine CNC Luthier Services

Thanks for the info guys, this is what I was looking for.

I had been wondering a lot about the subject Mark mentions in the quote below prior to him mentioning it. Upon playing with my model I notice it's pretty easy to shift these things around and change the shape. I can't really say though that if I do "this" then "this" will happen, but things definitely change. Any way I was wondering if it can be generally expanded on. That is, what's the tonal effect of having the resonances bunched close together vs. far apart?



Right now I'm mostly trying to verify my model to reality (dream on). So thanks for the information. It helps.
J

_________________
Jim Watts
http://jameswattsguitars.com