Official Luthiers Forum!

Being pretty green still, I wanted to ask you all about testing tops for
stiffness and what you're looking for as far as flex.

The reason I ask is I have a #2 grade sitka to that I have recently joined and
thicknessed to .011. When the halves were seperate they seemed pretty stiff
but after joining and thickness sanding there seems to be quite a bit of flex
to my untrained eyes.

Is there something I should look for and what are the affects of a top that
has too much flex? What is too much flex?

Thanks

_________________
Christopher C. Cordle
North Muskegon, MI

http://www.yardsaleunderwear.com
http://www.facebook.com/christopher.cordle

Oops, I think that should read .110 or one-hundred and ten thousandths.
I always question myself on the proper way to write that as a decimal.....
it's been just a few years since I took that math class

_________________
Christopher C. Cordle
North Muskegon, MI

http://www.yardsaleunderwear.com
http://www.facebook.com/christopher.cordle

The stiffness of each plate has not change but the distance from edge to edge has doubled therefore the amount of deflection for the same force has increased near proportionally.

Chris,
You should be fine. My plates get very loose after thickness sanding generally thinner than your dimension.
Be sure to keep good notes and develop a way to measure stiffness/deflection for a reference point.

_________________
"An adventure is only an inconvenience rightly considered. An inconvenience is an adventure wrongly considered." G. K. Chesterton.

Hey Chris.  I don't really think that there is "too much flex", within reason of course.  I think that the masters basically make up for it with bracing, maybe making it a little beefier than on a really stiff top.  I think Steve's suggestion is excellent because the only way to tell about flexibility and stiffness is through experience.  The more tops you flex the better you will be at judging them.

It is possible to aquire a reliable and accurate feel for this, but it takes time and practice. As with tuning strings, there are technical aids you can use until you develop more natural skill, or if you feel you want more accuracy or even just some numbers to talk about. There are many levels of complexity.

Many people are using one or another form of deflection testing. You support the top off the bench with a pair of parallel tubes or bars at some known distance, and place a weight in the center. Measure the deflection with a dial gauge. If you always use the same setup, the same weight, and the same thickness of wood, you can just compare results directly without worrying about exactly what you're measureing. That is, if you always use the same brick, and thickness the tops to .140" for testing, the one with the greater deflection will be less stiff, and you can thickness them accordingly. A modification of this would be to thickness all of your tops to give the same deflection, assuming you have some idea what that should be.

If you can calibrate everything, find out how much weight you're using, the thickness of the top, and so on, there are formulas that will allow you to calculate the 'Young's modulus' of the material, a measure of how hard it is to stretch it. That, and the density, which you get from the volume and weight of the piece, will tell you a lot about whther it's a good piece of wood or not. It's also helpful to get the stiffness in both the lengthwise and crosswise directions, as this can effect the way you brace the top.

If you have the facilities to do 'glitter testing', a signal generator, frequency counter, and amplifier, with some other stuff, you can find not only the lengthwise and crosswise Young's modulus, but also the 'damping factor'; a measure of how much energy the wood absorbs as it vibrates.

As far as anybody knows at the moment, these are the most important material properties that influence the tone of the guitar. They vary quite a lot from piece to piece, even within a given species, and knowing more or less exactly what they are can help you fine tune the instruments you make for better results.    

many builders include in their building journal objective measurements of the top's stiffness as defined by measurement of its deflection when suspended over a given span under a given weight, along with notes on the tone produced by each. some maintain that this is the only way to maintain consistency in the tone of their guitars even made with the same materials since there is so much variation in the stiffness of wood.

others depend on their muscle memory after flexing 100 or so tops.

oops, al beat me to it!

[QUOTE=uncleshish]....after joining and thickness sanding there seems to be quite a bit of flex....[/QUOTE]

This freaks me out too until I recall that stiffness changes as the cube of the change in thickness.

_________________
now known around here as Pat Foster
_________________
http://www.patfosterguitars.com

I do the deflection tests, like Al says, with a known weight and check the deflection with a dial gauge. I weigh everthing, test the braces for flex. Tap tune the top as I shave my "parabolic" braces. Record everything.

However, I still don't finish tuning the top, by selective sanding it, until the bridge is on and it's strung up to pitch. In fact I've just done it to an OM. Every part of the system affects all of the others, to tune the whole system ideally it should be complete. Only when I'm happy with the final result do I take the strings off, unbolt the neck again and French polish the top.

Colin

_________________
I don't believe in anything, I simply make use of a set of reasonable working hypotheses.

_________________
Dave
Milton, ON

Alan - or anybody else,

This is probably a stupid question, but deflection testing as I am reading about it here and have seen pictures of set-ups elsewhere, involves measuring the deflection in just one place on the top that has been thicknessed to a constant thickness over it's surface. Does this imply that when you have a measure in one place, the relative stifness in every other point on the top will be in a constant relationship with the stiffness in the chosen measuring point as you go from one top to another top? If not wouldn't you want to measure in a lot of different places and see the relationship on each piece of wood?

Also, I've read that the same weight is used each time. Wouldn't you also want to see gradients - ie how the deflection changes as the weight changes in the same place - or is this a constant relationship in all top woods?

_________________
Dave White
De Faoite Stringed Instruments
". . . the one thing a machine just can't do is give you character and personalities and sometimes that comes with flaws, but it always comes with humanity" Monty Don talking about hand weaving, "Mastercrafts", Weaving, BBC March 2010

[QUOTE=Dave White]........wouldn't you want to measure in a lot of different places and see the relationship on each piece of wood?......[/QUOTE]Dave,
I suspect that you are like me in that when you make a selection for top wood, you flex it and check that it is nice and stiff across the grain, without any noticeable spots that are either a concern visually or that feel odd when flexing. Without any uneven grain issues, a nice looking top wood should be very consistent in deflection both with and across the grain. I don't think it is necessary to make these measurements.
If someone has some spare tops to send me, I would be happy to cut them into 1" strips and take deflection measurements. Perhaps a few of our suppliers might be interested in seeing this kind of data.

[QUOTE=Dave White]
Also, I've read that the same weight is used each time. Wouldn't you also want to see gradients - ie how the deflection changes as the weight changes in the same place - or is this a constant relationship in all top woods?[/QUOTE] If you stay well under the yield strength of the material, it is a linear relationship.

_________________
"An adventure is only an inconvenience rightly considered. An inconvenience is an adventure wrongly considered." G. K. Chesterton.

Steve,

Thanks for the explanations.

I'm happy doing 4 or even 6 piece tops but 14-16 piece tops would be a real challenge

_________________
Dave White
De Faoite Stringed Instruments
". . . the one thing a machine just can't do is give you character and personalities and sometimes that comes with flaws, but it always comes with humanity" Monty Don talking about hand weaving, "Mastercrafts", Weaving, BBC March 2010

[QUOTE=SteveS] a nice looking top wood should be very consistent in deflection both with and across the grain. I don't think it is necessary to make these measurements.
[/QUOTE]

Steve,
You might take 5 or 10 joined tops, cut them into identical length and width dimensions, thickness them all to the same dimensions and then measure the deflection. You might change your mind. Looks can be deceiving. Don't loose sight that every piece of wood is unique even though two tops came from the same billet they will not be the same stiffness.

_________________
tim...
http://www.mcknightguitars.com

I just spent 20 minutes writing a review on Roger Siminoff's new DVD/Book set on "The art of tap tunning". Lost the whole thing with an improper placement of my cursor. Anyway, The book/dvd spends a good amount of time on stiffness and tuning of various parts (obviously related to stiffness). IT in NO WAY can compare to what you would learn from Master Luthier ERVIN SOMOGYI's class...BUT...it is a start. As Tim said above, you could "take 5 or 10 tops..." They WILL ALL BE DIFFERENT. If you really have a deep desire to learn more...Start with Siminoff's DVD/Book set, But definitely Go see the Master, Somogyi! He will change your luthiery world!

p.s. Stew Mac is selling the Siminoff set for around $35 or there abouts.

_________________
Dave Bland

remember...

"If it doesn't play in tune...it's just pretty wood"

I guess I will have to check out the book/dvd as this has all been very
confusing and much more in depth than I had thought it might be.

_________________
Christopher C. Cordle
North Muskegon, MI

http://www.yardsaleunderwear.com
http://www.facebook.com/christopher.cordle

[QUOTE=Tim McKnight] [QUOTE=SteveS] a nice looking top wood should be very consistent in deflection both with and across the grain. I don't think it is necessary to make these measurements.
[/QUOTE]

Steve,
You might take 5 or 10 joined tops, cut them into identical length and width dimensions, thickness them all to the same dimensions and then measure the deflection. You might change your mind. Looks can be deceiving. Don't loose sight that every piece of wood is unique even though two tops came from the same billet they will not be the same stiffness.[/QUOTE]

Tim,

I think Steve was referring to my question and saying that he didn't think it necessary to do deflection measurements on more than one place on any one top.

I suspect that he shares your view on every piece of wood being different, and from his point does deflection measuremnets on his tops.

_________________
Dave White
De Faoite Stringed Instruments
". . . the one thing a machine just can't do is give you character and personalities and sometimes that comes with flaws, but it always comes with humanity" Monty Don talking about hand weaving, "Mastercrafts", Weaving, BBC March 2010

[QUOTE=Dave White]Tim,

I think Steve was referring to my question and saying that he didn't think it necessary to do deflection measurements on more than one place on any one top.

I suspect that he shares your view on every piece of wood being different, and from his point does deflection measuremnets on his tops.[/QUOTE] Exactly.
Thank you Dave. I was going to write something similar, but you did it better than I could have.

_________________
"An adventure is only an inconvenience rightly considered. An inconvenience is an adventure wrongly considered." G. K. Chesterton.

Dave Stewarts asked:
" So, it sounds as if getting an accurate density of the chunk of wood you're dealing with would be a good initial predictor of it's relative stiffness and thus how thin you can take it. Have I got this right?"

Yes, more or less; there is some scatter in the data. I just went through my list of tested tops, and picked out all the ones with lengthwise Young's modulus (E) values between 9000-11000 MegaPascals(MPa). The densities varied from a low of 350kg/m^3 to 474 kg/m^3. That's a density range of 35% for a range in E values of 1.22. That list, BTW, includes Engelmann, Western Red Cedar, Redwood, and European spruce, and there's a lot of scatrter in any one of those woods as well. You could use density as a rough stand in for stiffness, but would be much better off really knowing the numbers.

When you stand on a trampoline the load is taken up by the whole surface, although most of the deflection is right under your feet. It's the same with a top in a deflection test: you support the edges and put a load in the middle, but the whole top bends, and it's the properties of the whole top you're measuring.

You could object that, just as every piece of wood is different from every other, so different areas within a top can be different. That's certainly true. The wide grain toward the outside is probably different from the narrow grain in the center, and, ideally, you'd like to measure those differecnes too. Similarly, it's often the case that the tree grew with some twist, and the grain line angle varies from one end to the other, and that has to have an effect. The problem is that there's no way I know of to look at the thing that closely without sawing it into forty little pieces, and then you can't make a guitar out of it any more.

In fact, of course, there is no perfect way to test anything, and anyway, the only test that really means anything is what the customer says when they pick the guitar up. All of this other stuff is meant to be a way of getting closer to that desired result.