Official Luthiers Forum!
http://mowrystrings.luthiersforum.com/forum/

Something I’ve been up to
http://mowrystrings.luthiersforum.com/forum/viewtopic.php?f=10102&t=14419		 Page 1 of 2
Author:  Jim Watts [ Wed Nov 07, 2007 2:30 pm ]
Post subject: 
I've been building some nomex cored double tops over the last few years and have decided I really want to focus on that type of structure. I've used pretty standard X bracing on them but have decided I want to enter into more of a development phase with them and see where I can take it. I think there's good fruit here but will know more later.

The postal service was kind enough to remove one of my necks for me, so I thought I'd turn that guitar into a test bed of sorts and make up a variety of tops to place on it.

So all of this lead me to the following; I thought I needed to know more about how a standard guitar top looked like under a load. So I built this FEA model based on my standard design and applied some material properties to it that I've measured over the years. I was pretty pleased when I put a dial indicator on one of my instruments and measured the deflection. I found that the model matched up pretty good, not perfect but pretty good, within about 10%.
I feel that this gives me a platform to spring board off of, checking my stress distributions and deflections on my new designs using the FEA tool comparing that data to my standard design. I will then build the candidate tops and place them on my test bed and see if I can correlate the sound to the model, who knows
Eventually I'll try to move this model from a static model to a dynamic one.

I thought there may be some out there how are interested in type of thing and also thought that many of you might be interested in what a top looks like under a string load (Daddario light). All the deflections are exaggerated for easy visualization. The red areas represent .029 inches, the actual guitar deflected .026. Right near the sound hole the actual guitar deflected .009 and the model predicted .007

Here are a couple of screen shots for you guys.





Author:  Hesh [ Wed Nov 07, 2007 2:46 pm ]
Post subject: 
Interested?  You bet my friend!!!

Your model came out very close to your measurements and that makes this very valuable in visualizing what in fact happens to a guitar top under load.

Is your objective to model the difference bracing configurations as well?

Jim when I met you at HGF I was most impressed with your use of carbon fiber and the engineering that went into your guitars.  It was clear to me that your guitars are outstanding.  Now you are taking it to the next level which should be very interesting indeed.

Thanks for sharing and updates would be greatly appreciated too with what you are comfortable sharing here.

Also, I'll be in touch to continue our conversation and I want to start incorporating some CF into my guitars too.

Great post Jim!!!
Author:  erikbojerik [ Wed Nov 07, 2007 2:53 pm ]
Post subject: 
OK James...as if the CF discussion wasn't enough, now you REALLY have
tweeked my scientist's antenna!!  It is always nice to get
correspondence between model and data.



So I must ask, what FEA software are you using to model this?
Author:  Parser [ Wed Nov 07, 2007 3:05 pm ]
Post subject: 
Looks like SolidWorks and maybe Cosmos to me...(?)

Did you model these parts as orthotropic materials? Or are they all effectively part of the same chunk of an isotropic material? I've been pretty curious about how accurate an FEA analysis would be if you just modeled everything to be isotropic (same stiffness in every direction).

Do you have the capability to do Modal Analysis? I used to do vibration analysis at my old job....I don't have those tools anymore at this new job. Modal will show you the Chladni patterns for each of the vibration modes and will also predict the resonant frequency for each mode....I would expect you'd have to model the material as orthotropic in order to get decent results from a modal analysis(?).



Author:  Jim Watts [ Wed Nov 07, 2007 3:06 pm ]
Post subject: 
Hesh - Yes, my idea is to look at different bracing patterns and top thickness. Lattice bracing particularly.

Erik, I'm using Solidworks and Cosmos.
Author:  KenH [ Wed Nov 07, 2007 3:17 pm ]
Post subject: 

WOW this is something I can use!


I have been overpowering my brain to figure out why the tone bars are placed in the position they are in, and this explains alot more than talk alone can produce.


Thanks so much for posting this!
Author:  Jim Watts [ Wed Nov 07, 2007 3:24 pm ]
Post subject: 
Ken, your welcome.
I think this is of limited use, so try not read more into it than is there (I don't think you are). This does not show how a top vibrates (next step for me), but helps give a clue about structural integrity.
Author:  Ben Furman [ Wed Nov 07, 2007 3:26 pm ]
Post subject: 
That's awesome, Jim.  I'm rather surprised by the distribution.  Way cool.

-Ben
Author:  Jim Watts [ Wed Nov 07, 2007 3:35 pm ]
Post subject: 
Parser, missed your post, sorry about that.
Each part is modeled as an individual part. I can change angles and placement, etc..
Right now I'm using isotropic properties which is why I'm surprised at how close it comes out. In the static load case I can kinda see how the isotropic properties work as the load is along the fibers and the braces have such a small cross section in that direction, I think.
I'm moving to a vibrational model, but I,m sure I'll need the orthotropic properties for that analysis. I'm having trouble finding the shear modulus in each plane for spruce for the orthotropic properties. I think maybe I'll contact the FPL and see if they can help with properties.
Author:  Jim Watts [ Wed Nov 07, 2007 3:37 pm ]
Post subject: 
[QUOTE=Ben Furman] That's awesome, Jim. I'm rather surprised by the distribution. Way cool.-Ben
[/QUOTE]
Me too, at least around the sound hole which is why I took an indicator to verify it. It was a really nice surprise
Author:  James Ringelspaugh [ Wed Nov 07, 2007 4:07 pm ]
Post subject: 
Does your model have an arch built into it?
Author:  Hesh [ Wed Nov 07, 2007 4:10 pm ]
Post subject: 
Jim sorry to deluge you with questions my friend but your post is VERY thought provoking.

The braces are not scalloped, I wonder what it would look like with scalloped braces?
Author:  Jim Watts [ Wed Nov 07, 2007 4:13 pm ]
Post subject: 
[QUOTE=lex_luthier] Does your model have an arch built into it?[/QUOTE]
Yes, It has a 25 ft arch in it.
Author:  Jim Watts [ Wed Nov 07, 2007 4:15 pm ]
Post subject: 
[QUOTE=Hesh] Jim sorry to deluge you with questions my friend but your post is VERY thought provoking.The braces are not scalloped, I wonder what it would look like with scalloped braces?
[/QUOTE]
No problem, I thought I'd scallop them also. I'll post an updated screen shot when I do.
Author:  erikbojerik [ Wed Nov 07, 2007 11:36 pm ]
Post subject: 
Sweet, thanks Jim!  I've been involved with several fluid dynamic modelling projects in my day job, so this appeals.
Author:  JJ Donohue [ Wed Nov 07, 2007 11:59 pm ]
Post subject: 
Very cool, Jim! Thanks for sharing. I'll be interested also to hear Alan's comments on this.

This is the kind of information that could be the catalyst for some new directions.
Author:  Parser [ Thu Nov 08, 2007 1:34 am ]
Post subject: 
Hey Jim W, here's a good journal article:
Sitka Spruce Shear Modulus

Just skimming throught it real quickly, it looks like the shear modulus measured along an axis parrallel to the grain for Sitka Spruce appears to be about 900 MPa (130.5 kPsi) while the shear modulus along the two axes perpindicular to the grain would measure about 1.75 times that (1600 MPa ~ 232 kPsi).

Author:  Parser [ Thu Nov 08, 2007 1:41 am ]
Post subject: 
Now let's see some Modal Analysis! That's where the action is. Of course you realize once everyone sees what you can do with this you're going to have to spend your time analyzing the effect of soundboard thickness, brace geometry, bridge geometry, bridge plate geometry, the effect of the sides, guitar shape, the effect of the back, as well as alternative kerfing, etc, etc, etc....

I can't wait to see what you get!
Author:  grumpy [ Thu Nov 08, 2007 2:58 am ]
Post subject: 
Nice, but... the model doesn't jibe completely with the levels and areas of distortion we see on older guitars.
Author:  Jim Watts [ Thu Nov 08, 2007 3:17 am ]
Post subject: 
Parser, Thanks for article, I appreciate it.

[QUOTE=grumpy] Nice, but... the model doesn't jibe completely with the levels and areas of distortion we see on older guitars.[/QUOTE]
Agreed, the model does not take into account wood creep so you can't apply the results to an aged guitar. You also can't apply the results to any random guitar. I modeled it after one of mine and used the material properties I measured from the wood I built it with.
Author:  Bruce Dickey [ Thu Nov 08, 2007 3:51 am ]
Post subject: 
Way to go Jim, thanks for showing it and sharing your thoughts.
Author:  erikbojerik [ Thu Nov 08, 2007 5:01 am ]
Post subject: 

[QUOTE=Jim_W]I ...used the material properties I measured from the wood I built it with.[/QUOTE]



Just curious as to what properties...density?  MoE?  Any others?
Author:  Jim Watts [ Thu Nov 08, 2007 6:56 am ]
Post subject: 
Erik,
As far as the properties I measure it was just the MoE and Density of the wood.
Simple deflection testing and a gram scale. I measure the tops in 2 axes but I only check my braces along the grain.
Author:  grumpy [ Thu Nov 08, 2007 10:05 am ]
Post subject: 
So the modeling doesn';t take into consideration that the top wood is much stiffer in one direction over the other?

And why are the strings pulling down into the top? Is that just part of the drawing, or is that where the actual string pull direction is modeled?


Interesting stuff, but before weighing-in too seriously, we need to establish some basics as to how it was modeled.
Page 1 of 2 All times are UTC - 5 hours
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/