Static tests of timbers in structural sizes (ASTM D198-84)

[Pyrrhus]

Hey everybody,
My friend Brian and i are trying to understand the procedures for this specific test, has anyone here worked with this before?

Anyway, we are not sure when the wood beam should be considered broken. The test doesn't seem clear.

A simple outline of this test is a simply supported beam with 2 concentrated loads in order to induce the pure bending condition. The idea is to find the rupture bending stress.

We know the max deflection of this beam can be calculated from Mechanics of Material. Most of the variables in the equation are known for us, except for the modulus of elasticity. How should we approach this problem?. Brian told me he had the beams made from a wood supplier, maybe we could get the specs? or use tables according to the tree specimen? and if such which tables?

Thanks in advance.

 

[FredGarvin]

No I haven't worked with that ASTM spec. Heck, I've never had to work with wood in a professional setting, but I can feel your pain in this. Wood has got to be tricky to work with and a statistical mess for data.

Have you looked here?:

Wood as an engineering material. General Technical Report 113 .
http://www.fpl.fs.fed.us/documnts/fplgtr/fplgtr113/fplgtr113.htm

Wood mechanical properties:
http://www.fpl.fs.fed.us/documnts/fplgtr/fplgtr113/ch04.pdf

 

[Pyrrhus]

Hey Fred, thanks for the response i forwarded the links to Brian, we will be looking into that. So far, we have an idea of how to know when the wood beam is considered broken.

 

[FredGarvin]

Have you set any kind of parameters to control moisture content? I would think that if testing were done on samples that were as dry as possible that would help you in your break-no break criteria, but I don't know if it would change the mechanical properties.

If you can find anything, you may want to do some looking into what metallurgists do for very ductile materials in tensile tests. It may shed some light. If I get over to my materials lab anytime soon I'll see if I can ask some of those guys.

 

[Pyrrhus]

Have you set any kind of parameters to control moisture content? I would think that if testing were done on samples that were as dry as possible that would help you in your break-no break criteria, but I don't know if it would change the mechanical properties.

If you can find anything, you may want to do some looking into what metallurgists do for very ductile materials in tensile tests. It may shed some light. If I get over to my materials lab anytime soon I'll see if I can ask some of those guys.

For the moisture content, we relied on the ASTM D198-84 test. That's under control, or at least Brian told me. I'll check again.

Yea, if you could ask somebody who has done this test in the past, it'll really be appreciated.

 

[haynewp]

I have not done this test but I have done the same bending test setup on concrete beams.

I would say when rupture of the bottom wood fibers occur visually, and this should correspond to a maximum on the loading data recorder. You should see from the recorded loading data a maximum load point reached with a sharp rise in recorded deflection after this point without any increase in load required. You can also calculate the modulus of elasticity through testing the beam prior to failure, rearrange the deflection formula and solve for "E" using the recorded deflection.

 

[chief]

I'm a little late here, but I think I can help. The modulus of elasticity can be found in the NDS (national wood design code) tables. You just have to know a few things... the wood species, how the wood was graded (mechanically or visually), that actual grading (no. 2, dense, etc.), and the size of the beam or joist. All of this information should be located on a stamp found on the piece of wood.

A common type of wood is visually graded no. 2 yellow pine. For a 2x4 of this material, the E value is 1,600,000 psi.

The moisture content would only affect the calculations if it is above 19% (wet conditions).

 

[AlephZero]

I once got involved in trying to do 4-point bending tests on large (100 x 4 x 2 inch) beams that were composite structures - a metal or plastic outer skins and metal honeycomb or foam filling.

One big problem trying to measure the deflections was local crushing of the structure at the loading and support points. I would suspect testing wood will have the same problem. (I haven't looked at your test spec so I don't know if it specifies a way to avoid this).

We never found a good (i.e. accurate and repeatable) way to do our tests - which probably isn't what you want to hear!

One way to measure impending failure is just to listen for cracking noises. Making a video recording, with sound from a microphone close to the test piece, is cheap to do and very useful for a post mortem analysis if/when things go wrong.

 

[haynewp]

I still say that you want to measure the modulus of elasticity for the actual piece of wood that you have, and not pull the modulus from NDS tables. This is because you are under experimental circumstances and not design.

Alephzero is likely correct about bearing failure at the supports, but I would rely on the electronic data where the maximum point of allowable load is reached. Although the point of "failure" may also be considered where bearing at the supports crushes to the point that it would be considered unserviceable in real life. You might also find the failure to be in the form of horizontal splits (shear failures) depending on the length of the beam in the test setup.

 

[AlephZero]

To state the obvious, wood grows on trees - so it's mechanical properties like Youngs Modulus are not consistent to the same degree as metal alloys with tightly controlled specifications.

But for a 4-point bending test the structure is statically determinate, so you can calculate the stress from the applied loads without knowing E. And you can calculate E for your test piece from the load-deflection curve, if you want, using the standard beam equations.

Sure, you need estimates for E and the expected failure stress when you are planning the test, so you have some idea if your test piece is going to deflect 0.1 inch or 10 feet before it breaks, and whether you will need a load of half a pound or half a ton to break it.

Data from a wood handbook should be good enough to get you in the right ballpark for your test setup.

 

[haynewp]

I know wood grows on trees, and I never said you shouldn't estimate it for general purposes from the NDS handbook. Anyone (you?) that has done experimental work would know that you also want to verify the properties EXPERIMENTALLY for the specimen you are working with for your data and not just go on published values.

To state the obvious, wood grows on trees - so it's mechanical properties like Youngs Modulus are not consistent to the same degree as metal alloys with tightly controlled specifications.

But for a 4-point bending test the structure is statically determinate, so you can calculate the stress from the applied loads without knowing E. And you can calculate E for your test piece from the load-deflection curve, if you want, using the standard beam equations.

Sure, you need estimates for E and the expected failure stress when you are planning the test, so you have some idea if your test piece is going to deflect 0.1 inch or 10 feet before it breaks, and whether you will need a load of half a pound or half a ton to break it.

Data from a wood handbook should be good enough to get you in the right ballpark for your test setup.

 

[PhanthomJay]

Having worked with wood timbers for a number of years, i have always used, for southern yellow pine, an E modulus of 1.6 x 10^6 psi (as noted above), and a breaking stress of 8000 psi (same for Douglas fir; I think Western red cedar is about 6000 psi). I'm sure experimentally that those numbers will vary somewhat, but as AlephZero noted, the breaking strength is independent of the E modulus. I never had much luck calculating deflections, though; wood seems to have a mind of its own in that regard.

 

[Pyrrhus]

Hey everyone, thanks for the help so far. We figured what was need to be done and already started breaking the specimens. The only problem so far is with the deflection, the wood beams are making "homeruns" with the sensor. We are working on a mechanism to fix the problem.