The Acoustic Guitar Forum

Go Back   The Acoustic Guitar Forum > General Acoustic Guitar and Amplification Discussion > Build and Repair

Reply
 
Thread Tools
  #31  
Old 10-09-2017, 01:03 PM
Alan Carruth Alan Carruth is offline
Registered User
 
Join Date: Feb 2011
Posts: 4,180
Default

Wood starts to 'cold creep' as soon as you put a load on it. The longer the load is on the greater the deflection. The folks I know who get the most consistent measurements this way put the load on, zero the gauge, and then remove the load, noting the change immediately. On advantage of the vibration testing method is that cold creep doesn't enter in directly: it probably has some bearing on damping but doesn't alter the Young's modulus measurement.

If you know some things, like the spacing of the supports, the weight, and the dimensions of the part being tested you can calculate the Young's modulus of the wood. That allows you to compare your results directly with those of other people, and effectively increases the size of your data base.
Reply With Quote
  #32  
Old 10-10-2017, 06:50 AM
redir redir is offline
Registered User
 
Join Date: Aug 2011
Location: Mountains of Virginia
Posts: 7,657
Default

Quote:
Originally Posted by Alan Carruth View Post
Wood starts to 'cold creep' as soon as you put a load on it. The longer the load is on the greater the deflection. The folks I know who get the most consistent measurements this way put the load on, zero the gauge, and then remove the load, noting the change immediately. On advantage of the vibration testing method is that cold creep doesn't enter in directly: it probably has some bearing on damping but doesn't alter the Young's modulus measurement.

If you know some things, like the spacing of the supports, the weight, and the dimensions of the part being tested you can calculate the Young's modulus of the wood. That allows you to compare your results directly with those of other people, and effectively increases the size of your data base.
That's exactly the way I do it too. Is there a tutorial or even just a formula that you could share or point me to online perhaps for figuring that out. I can see that being useful.

And if, "just Google it..." is the answer then I can do that too

Last edited by redir; 10-10-2017 at 08:48 AM.
Reply With Quote
  #33  
Old 10-10-2017, 10:19 AM
Rodger Knox Rodger Knox is offline
Registered User
 
Join Date: Jul 2013
Location: Baltimore, Md.
Posts: 2,431
Default

Quote:
Originally Posted by redir View Post
Is there a tutorial or even just a formula that you could share
Formula for simple beam with concentrated load at center, deflection at center:

deflection = (Load x ((distance between supports)squared)) / 48 x (Young's modulus)x(Moment of inertia of beam)
Moment of inertia of a rectangular beam = (width x ((thickness)cubed))/12

A little algebra gives:

Young's modulus = (Load x ((distance between supports)squared)) / (48 x (width x ((thickness)cubed))/12)
__________________
Rodger Knox, PE
1917 Martin 0-28
1956 Gibson J-50
et al

Last edited by Rodger Knox; 10-10-2017 at 10:25 AM.
Reply With Quote
  #34  
Old 10-10-2017, 10:33 AM
redir redir is offline
Registered User
 
Join Date: Aug 2011
Location: Mountains of Virginia
Posts: 7,657
Default

Cool thanks for that. I wish I had known that when I started. I presume the type of support is pertinent. In my case as you can see the supports are a 'point' so I imagine that is better than a square though maybe it doesn't matter. Also does the weight have to be dead center? I imagine that the dial gauge would have to be center too but of course that is not possible with my set up. They could be very close though.
Reply With Quote
  #35  
Old 10-10-2017, 11:17 AM
Rodger Knox Rodger Knox is offline
Registered User
 
Join Date: Jul 2013
Location: Baltimore, Md.
Posts: 2,431
Default

The equation is probably close enough for your setup. I'd guess that the effect of the brick not being a point load and the deflection measurement not being exactly centered would be less than 5%. Don't forget that most of these numbers have some type of units, and the units need to be consistent.

edit: If you're really interested in the physics and math, I'd recommend the Gore/Gilet books.
http://www.goreguitars.com.au/main/p..._the_book.html
__________________
Rodger Knox, PE
1917 Martin 0-28
1956 Gibson J-50
et al

Last edited by Rodger Knox; 10-10-2017 at 11:35 AM.
Reply With Quote
  #36  
Old 10-10-2017, 07:25 PM
Monsoon1 Monsoon1 is offline
Registered User
 
Join Date: Jul 2007
Location: N.E. Ohio
Posts: 1,810
Default

Alan Carruth, redir, thanks for the well thought out posts.
__________________
Something something, beer is good, and people are crazy.
Reply With Quote
  #37  
Old 10-10-2017, 07:29 PM
Monsoon1 Monsoon1 is offline
Registered User
 
Join Date: Jul 2007
Location: N.E. Ohio
Posts: 1,810
Default

Quote:
Originally Posted by redir View Post
Cool thanks for that. I wish I had known that when I started. I presume the type of support is pertinent. In my case as you can see the supports are a 'point' so I imagine that is better than a square though maybe it doesn't matter. Also does the weight have to be dead center? I imagine that the dial gauge would have to be center too but of course that is not possible with my set up. They could be very close though.
Take the brick and epoxy a thin flat piece of wood or metal to the side.
Then lay the brick dead center on the wood and take your measurement off of the thin flat piece you've epoxied to it.
Now you'll be able to get a dead center measurement.
__________________
Something something, beer is good, and people are crazy.
Reply With Quote
  #38  
Old 10-10-2017, 08:29 PM
printer2 printer2 is offline
Registered User
 
Join Date: Dec 2012
Location: Middle of Canada
Posts: 5,098
Default

Use a piece of steel tubing or rod as the end supports. It would be best if the load was applied the same way but even a rectangular rod would be fine and a lot easier than making the weight balance on it. Say a half inch width glued onto the bottom of the brick. You could also make the same measurement lengthwise on the plate to get the crosswise value although you might need some more weight to get adequate deflection.
__________________
Fred
Reply With Quote
  #39  
Old 10-11-2017, 11:03 AM
Rodger Knox Rodger Knox is offline
Registered User
 
Join Date: Jul 2013
Location: Baltimore, Md.
Posts: 2,431
Default

While these are all good suggestions to make your loading closer to the assumptions of the equation, they probably will not make your readings more accurate. The precision of measurement for the thickness and deflection will probably be the limiting factors in accuracy of results, and your setup would be better served by maximizing the precision of these measurements.
__________________
Rodger Knox, PE
1917 Martin 0-28
1956 Gibson J-50
et al
Reply With Quote
  #40  
Old 10-11-2017, 05:54 PM
printer2 printer2 is offline
Registered User
 
Join Date: Dec 2012
Location: Middle of Canada
Posts: 5,098
Default

Dial gauge and digital micrometer works fine and while we are at it a digital kitchen scale will take you a far way. I have used the same precision micrometers in measuring test sample dimensions in the aerospace manufacturer test lab I worked in. To get better precision with the dial gauge use double the weight (within reason).
__________________
Fred
Reply With Quote
  #41  
Old 10-12-2017, 11:08 AM
Rodger Knox Rodger Knox is offline
Registered User
 
Join Date: Jul 2013
Location: Baltimore, Md.
Posts: 2,431
Default

Quote:
Originally Posted by printer2 View Post
Dial gauge and digital micrometer works fine and while we are at it a digital kitchen scale will take you a far way. I have used the same precision micrometers in measuring test sample dimensions in the aerospace manufacturer test lab I worked in. To get better precision with the dial gauge use double the weight (within reason).
Those are certainly the right tools, and the most difficult variable to measure is the thickness of the plate. If it's uniform thickness (which it's not), then it's easy, one measurement does it. It's not uniform thickness, it varies by some amount, probably less than our ability to measure (~.0001") or so. If that's the case, edge measurements would be adequate to get a valid average thickness. Looking at the equation, the thickness is cubed, so any measurement error is also cubed, so I'm concerned more about the thickness than anything else. Incidentally, in engineer speak, error is the variation between the measured value and the actual value, and is a function the precision of the measurement instrument, ie a micrometer is more precise than a tape measure.

Bottom line is the uniformity of the thickness of the plate is critical for the equations to be valid.
__________________
Rodger Knox, PE
1917 Martin 0-28
1956 Gibson J-50
et al
Reply With Quote
  #42  
Old 10-12-2017, 12:48 PM
redir redir is offline
Registered User
 
Join Date: Aug 2011
Location: Mountains of Virginia
Posts: 7,657
Default

With my process the uniformity in thickness could be a problem depending on what tolerances is acceptable. I run the plates through a drum sander till it's just about where I want it. Then I use a hand plane to smooth off one side, the side that the bracing gets glued too. The reason I do that is, without getting too far into it - long story short, a planed surface is better than a sanded one. IMHO opinion of course. I don't plane the outside face because I always use sand paper to finish the top anyway. I can't ever seem to finish soft wood with a scraper like I can with backs and sides.

But anyway then that allows me to take my first deflection measurement and run it through the drum sander till I get what I want. So the hand planed face might be a bit thicker in the middle or to one side but it would be very small.

I suppose I could take 3 measurements with my dial calipers along each side and take an average.

Oh and also I have been doing this with the rosette installed so I imagine I'd have to change that process as well since that might throw the weight off.
Reply With Quote
  #43  
Old 10-12-2017, 02:10 PM
Rodger Knox Rodger Knox is offline
Registered User
 
Join Date: Jul 2013
Location: Baltimore, Md.
Posts: 2,431
Default

It's really much to-do about nothing. Your data will be good for your use, even if the Modulus you calculate is a bit off. (You're shooting for a target deflection, so you're not using the Modulus in your design.) It's a bit like tuning to standard, it only make a difference if you're playing with others.
__________________
Rodger Knox, PE
1917 Martin 0-28
1956 Gibson J-50
et al
Reply With Quote
  #44  
Old 10-12-2017, 07:01 PM
Alan Carruth Alan Carruth is offline
Registered User
 
Join Date: Feb 2011
Posts: 4,180
Default

A couple of things:

As has been said, you can't get an answer that's more exact than your least accurate measurement. It's worth putting some effort into improving that first. For most of us, that's going to be thickness.

I saw a paper once that claimed there was a systematic difference in measurements of stiffness depending on the way the piece was surfaced. It makes some sense: the stress is highest on the surface, and if that's fuzzy, say, from sanding with coarse paper, you might in effect be sacrificing some thickness.

I use vibration measurements to find wood properties. It has been pointed out that the usual equations that calculate the Young's modulus (E) based on the size and mass of the piece and the resonant frequencies, are simplifications. For example, they may assume that the shear modulus is zero, or infinite, either of which would give a somewhat different result when solving for E. There are more complex models that can give answers that are closer to 'correct', but they require more measurements of other sorts of resonances and more calculation.

In the end, the 'simple' measurement and 'basic' equation are supposed to get you to within 10% or so of the 'real' value. That implies a thickness change of about 3% to maintain the same stiffness. That's about +/- .004" on a top that's .125" thick. Do any of us actually measure thickness that accurately and maintain it? Do you know how much wood you remove on final sanding?

And then there's the variation within each piece. How do you get that without actually chopping the wood up small? If you do how do you make a guitar from it?

And so it goes.
Reply With Quote
Reply

  The Acoustic Guitar Forum > General Acoustic Guitar and Amplification Discussion > Build and Repair

Thread Tools





All times are GMT -6. The time now is 06:49 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.
Copyright ©2000 - 2022, The Acoustic Guitar Forum
vB Ad Management by =RedTyger=