Official Luthiers Forum!

I didn't want to hijack the thread on Adi tops, but the question of deflection testing comes up fairly often and as a newbie, I'm not certain of the technique(s).

It seems like one should attempt to clamp one end (or side?) then apply a known weight a known distance from the clamp and measure the deflection.

It also seems that holding some of these measurements, most notably the length between the clamp and weight constant would be difficult.

I guess that I am wondering how the measurements can be accomplished?

Thanks in advance--Jim

_________________
Jim Howell
Charlotte, NC

I sort of explain my (Ervin's?) jig in the other thread, but essentially the only stiffness you really care about with respect to the raw plate is longitudinal. You can more easily influence the cross grain stiffness with the bracing.

The jig I use essentially suspects the plate between 2 dowels centered 18" apart. The deflection is measured from a dial indicator that can be zero'd from above.

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

How much deflection should we be aiming for over that 18" distance? How much weight? Numbers Brock, we need numbers

I'd really like to take Ervin's class, but I'd have to add a couple grand for airfares. Try getting that past the missus. PaulB38939.9188310185

Numbers are relative. The more your weight weighs the more the deflection... right? Everyone will have different numbers depending on what they are using to deflect the top.

If I use a mouse to deflect my tops and you use an elephant, we can't aim for a given deflection.

I came home from ES's class with 2 tops that were at target deflection and I use them as a reference point.

Sorry... I wish I could be more helpful ....

You will have to follow John How's advice and take notes, but I will tell you that you can sand far beyond what you think is safe.

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

[QUOTE=Brock Poling]
Numbers are relative. The more your weight weighs the more the deflection... right? Everyone will have different numbers depending on what they are using to deflect the top.

If I use a mouse to deflect my tops and you use an elephant, we can't aim for a given deflection.

[/QUOTE]

Which is why I asked how much weight you were using and how much deflection you were aiming for.

Consistency of method seems to be the most important factor in obtaining meaningful results. David Hurd offers a very specific and interesting approach to this process. Check out:

http://www.ukuleles.com/Technology/compliance.html

_________________
D..

[QUOTE=PaulB] [QUOTE=Brock Poling]
Numbers are relative. The more your weight weighs the more the deflection... right? Everyone will have different numbers depending on what they are using to deflect the top.

If I use a mouse to deflect my tops and you use an elephant, we can't aim for a given deflection.

[/QUOTE]

Which is why I asked how much weight you were using and how much deflection you were aiming for.


[/QUOTE]

Brock's carefully guarded secret

_________________
Jim Kirby
kirby@udel.edu

[QUOTE=jtkirby] [QUOTE=PaulB] [QUOTE=Brock Poling]
Numbers are relative. The more your weight weighs the more the deflection... right? Everyone will have different numbers depending on what they are using to deflect the top.

If I use a mouse to deflect my tops and you use an elephant, we can't aim for a given deflection.

[/QUOTE]

Which is why I asked how much weight you were using and how much deflection you were aiming for.


[/QUOTE]

Brock's carefully guarded secret [/QUOTE]

I honestly don't know how much it weighs. It is just a chunk of metal. I can try to find a postal scale and weigh it.

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

[QUOTE=Brock Poling] [QUOTE=jtkirby] [QUOTE=PaulB] [QUOTE=Brock Poling]
Numbers are relative. The more your weight weighs the more the deflection... right? Everyone will have different numbers depending on what they are using to deflect the top.

If I use a mouse to deflect my tops and you use an elephant, we can't aim for a given deflection.

[/QUOTE]

Which is why I asked how much weight you were using and how much deflection you were aiming for.


[/QUOTE]

Brock's carefully guarded secret [/QUOTE]

I honestly don't know how much it weighs. It is just a chunk of metal. I can try to find a postal scale and weigh it.

[/QUOTE]

I find Ervins ideas to be very aluring, what little I know of them. Seems to me he's onto something and that we're all over-building, otherwise we'd hear of a lot more failures. Guitars collapsing under the pull of the strings etc. But in the few years that I've been into this, I've NEVER heard of one catastrophic failure, not one, from a newbie building too light. Seems to me that if we were all close to the line, we'd have heard tales of woe. But we haven't, well I haven't at any rate, maybe I should get out more. So how light can we build?

When I made kayaks, we'd all try to build as light as possible. With boat building everything is a trade off. You sacrifice strength for a lower weight (especialy kayaks when you sometimes have to carry the bloody things). We'd often hear of failures because the builder had crossed that line. So where are the failures in guitarmaking? Where exactly is that line?

Now I know Brock has paid a lot of money for Ervin's school, something I don't see me ever being able to justify especially when trans-pacific airfares are tacked on to the overall cost. If I were Ervin I'd make all students take a vow of silence, this is his bread and butter after all. But I still wanna know! PaulB38940.3602546296

it seems as if some are trying to expect perfect results to arise because they are using an "objective" measuring process.

for what it's worth, to me it seems that what this technique gives is a basis for approaching the problem of getting consistent results from different pieces of wood.

one has to do the measurements, take notes, build, ask whether you like the sound? if so then that degree of deflection with that type of wood may continue to give good results if you also have all the other aspects of your building technique under consistent control, e.g. bracing size and stiffness comes to mind.

is there any magic significance in having the measurement span 18"? why not 20", or 17". perhaps it is because that span will support small tops without their defecting so much under the weigh used that they fracture or fall through?

what weight? enough to make it measureably bend, but not enough to cause damage.

the technique is just one way to try to objectify what one does with their hands, or should do at least, as they thin the top, namely, flex the wood and feel how it is changing as it is worked, and compare the results with what is known from experience to work for you.

CMM, there is not much scientific about this approach. For sure it has a process and it is well thought out, but beyond measuring the density of the plate and measuring the deflection there really isn't much more "science" going on.

I am certainly not saying my (i.e. Ervin's) way is the ONLY way to approach the issue of building the plates, BUT I know it works, the results it produces are stellar, and (perhaps most importantly) I understand it.

The reason that it is 18" is because "it is". You are right, the distance can be anything you like (within reason). It just needs to be the SAME for every plate you make so you have consistant readings. The deflection test IS taking into account the properties of each individual piece of wood. That is the point of it.

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

i agree with each point you make. but what i suspect is that some are viewing the technique as a panacea, i.e. do this and you will get perfect results.

rather, it seems to me that it is merely an aid, if used correctly, to help achieve consistent results.

the degree of perfection achieved will depend on many things besides the technique used to measure how stiff the plate is.

Yes CMM, you are right. It is only a method of quantifying what you see and feel with your hands. It's part of training your senses. You need to do all of the things you mentioned measure the deflection (and your numbers will be only your numbers), flex the wood, tap the wood, just be consistent and take notes. You will begin to get a feel for the wood but it's not magic and it's realy not science as Brock pointed out.

_________________
Tickle your guitar daily, and it'll tickle you back.

[QUOTE=Dennis E.] Consistency of method seems to be the most important factor in obtaining meaningful results. David Hurd offers a very specific and interesting approach to this process. Check out:

http://www.ukuleles.com/Technology/compliance.html
[/QUOTE]

Dennis, thanks for that link. The topic is now much clearer to me. It seems that correlating the amount of deflection to some tonal quality is still going to be a trial and error process. It's not clear to me how to measure the frequency response of the result and "dial it in".

The test is probably of more use as a way of becoming more consistent. If I apply the test as I build a series of guitars, find one that sounds really good, then I have a set of deflection numbers I can use on future instruments.

_________________
53% of all statistics are made up on the spot
http://driftwoodguitars.blogspot.com/

Doh! Just like CMM said.

_________________
53% of all statistics are made up on the spot
http://driftwoodguitars.blogspot.com/

[QUOTE=Hesh1956]Brock can't be expected to do a data dump of Ervin's methods and Brock has shared a very fair amount of information in this tread and some others recently. I have been hanging on every word…… From what I have read we do over build our tops and this is probably an understatement too.

[snip]

In fact I WANT to take Ervin’s class……..
[/QUOTE]


CMM I agree with you that this method won't help you build the "perfect" guitar. But... I would challenge anyone to define exactly what a perfect guitar is. What is great to one person is only "ok" to another. Sound quality is somewhat subjective to the player and the style of music s/he is trying to acheive.

However, what this method does produce is a METHOD of allowing you to control the sound with a high degree of predictability. I am 100% sure that I could not build a guitar that sounds like Ervin's. I will never be able to out Ervin Ervin. I have no desire to do that. Applying his method simply allows me to build and develop my own sound.

I know I sound like I "have religion" ... and I do.    But I openly acknowledge that there may be multiple roads to the same goal. I am simply pushing the merits of this one.


And, Hesh... your comments are good, and I really think you should try to take the class if you can. It totally changed my thinking....   However, I have shared a few nuggets about the approach, but I caution you that there is NO WAY I could even articulate this method to you live and face to face.... and certainly not in the format of a forum.. To do so would be a disservice to you and Ervin. It took him a week of VERY full days to impart this to us...

But there are a few main themes worthy of consideration, but I wouldn't go off and start trying to apply this without seeing the whole show.

Brock Poling38940.4593518519

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

Consistent thickness in test comparison is essential also. Deflection "test" three pieces, one .250, one .125 and one .140 and your results are almost meaningless for comparison points. Keep in mind that species variances are plus or minus 20% making any two same species pieces of wood on the workbench up to 40% different in properties. Deflection testing will give a builder a leg up at least in picking. It seems well experienced builders intuitively can hand flex and just "know".

One thing that Hesh comments on, and I tell my students, is that my feel for the flexibility of a top is to be used in conjunction with a certain bracing style and sizing. change the top deflection and you now have to change the bracings, or you will over/under build that guitar. Mario once stated that a great sounding guitar he built had a top that was 140 thou. Maybe it also had the lightest braing he has ever used - he never said. For those of you that came to Fort Erie ... my Dragonfly guitar has a top that was .135, the Guiding Light was more like .110 or less. Both made pretty good guitars. They have identical bracing sizings and pattern (both are multiscales). i refus eto go below .100 on a top for fear of having the bracing telegraph thru badly - you will likely see some with thin tops of say 100-105, but thinner will make it way worse IMO. The guitar may sound livelier, but i think over time it will deform much more.

Brock - you mention that you are testing for longitudinal strength because cross grain can be more easily controlled with bracing - the X brace does both due to its orientation, so cant you also be controlling the other at the same time ??? The soundhole completely negates the longitudinal strength in a top in fornt aof a bridge, the X brace is there to divert the pull around the hole - so are you really controlling the pull from behind the bridge then with this test ??? Its all very interesting.

_________________
Tony Karol
www.karol-guitars.com
"let my passion .. fulfill yours"

[QUOTE=Larry Davis] Consistent thickness in test comparison is essential also. Deflection "test" three pieces, one .250, one .125 and one .140 and your results are almost meaningless for comparison points. Keep in mind that species variances are plus or minus 20% making any two same species pieces of wood on the workbench up to 40% different in properties. Deflection testing will give a builder a leg up at least in picking. It seems well experienced builders intuitively can hand flex and just "know".[/QUOTE]

Of course... the idea is to test deflection and thin the plate until you hit a target deflection. The thickness of the plate is not the main concern.

Or, did I miss your point?

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

[QUOTE=TonyKarol]    Brock - you mention that you are testing for longitudinal strength because cross grain can be more easily controlled with bracing - the X brace does both due to its orientation, so cant you also be controlling the other at the same time ??? The soundhole completely negates the longitudinal strength in a top in fornt aof a bridge, the X brace is there to divert the pull around the hole - so are you really controlling the pull from behind the bridge then with this test ??? Its all very interesting.[/QUOTE]

I need to think about how to respond to this..... This could lead us right down the rabbit hole.   

_________________
Brock Poling
Columbus, Ohio
http://www.polingguitars.com

I hesitate to reply to these threads because these threads often degenerated into "art" versus "science" and that gets tiresome for everyone.

I build to a target deflection for consistency among instruments. Folks have been doing that forever, but gauging flex with their hands. If you build a lot of instruments, then I think you can get the touch for that in a few years. Because I only build a handful a year, I figured measuring deflection would get me consistent faster.

Deflection, in part, is a function of thickness, mass of the material, the span used, and the weight used. So use a constant span and constant weight and you don't have to think about those anymore.

If you always measure deflection for materials of the same thickness, you will find that heavier materials almost always deflect less than lighter materials, which is why we can usually thin Sitka more than Engelmann to reach a target deflection.

If you measure materials of different thicknesses, as I do for rough plates, measurements need not be meaningless as Larry suggests, but they will need to be adjusted to a common scale. A useful scale is Young's modulus (MOE) which incorporates mass, span, thickness, and deflection.

So I measure and weight my plates, measure deflection across a measured span with a known weight, and stick it in a spreadsheet to calculate MOE. This way I know where that plate stands before I actually start to build with it.

If you are only after a target deflection (and that value is up to you), you have no need for MOE.

David Hurd's work on compliance maps has extended the idea of consistency in deflection of fixed plates to consistency of plates that are installed on the instrument. I have not tried it yet, but it makes sense to me.

[QUOTE=crazymanmichael] i agree with each point you make. but what i suspect is that some are viewing the technique as a panacea, i.e. do this and you will get perfect results.

rather, it seems to me that it is merely an aid, if used correctly, to help achieve consistent results.

the degree of perfection achieved will depend on many things besides the technique used to measure how stiff the plate is. [/QUOTE]

It's not a question of finding a panacea. It's just that if you use an 18" span and weight X, and I use an 18" span and weight X, then we begin to have facts that we can discuss with each other. That doesn't happen until there is consistency, both self-consistancy but also consistency between the group of people performing tests.
(Unless you want to convert your weights and deflections to elastic moduli, and report those.)

_________________
Jim Kirby
kirby@udel.edu