Heavy Timber Beam Testing

[noxer]

I am going to have some Heavy Timber beams sampled for testing and wanted to see if anyone had any thoughts on where the samples should be taken. The samples will need to be about 4"x4"x36" Long. The beams are 11 1/2" x 11 1/2" and span about 12 ft.

Obviously, I am concerned about permanently removing a portion on the beam. The samples I would imagine need to be taken from either the top corner of the beam or the bottom corner based on how the strips are going to be removed.

My initial thought is to locate the samples near the end of the span and just make sure that the reduced shear capacity would be acceptable. Any other thoughts?

I'd appreciate any thoughts or suggestions especially from anyone that has dealt with analyzing heavy timber beams based on tests on samples.

Thanks!

 

[dhengr]

Some of these OP’s are real head scratchers..... I’d make the samples 4.5"x4.5"x22", because 36" long won’t fit in most fireplaces or wood stoves. What on earth are you testing for, what are you trying to prove and what kinda damage do you do to the timber taking your samples? Why not have someone knowledgeable in timber grading, grade the beams for you? Alternatively, instead of ruining the timbers; test each one to 1.5 or 2 (you pick a number) times the load they will see in action and use those that pass the test. That is, they don’t fail, crack, etc. under load, and they return to their original shape, elastically, when the load is removed.

 

[Brad805]

I too am wondering why? If it is a decay issue, there are non-destructive test methods that will help you decide if testing is needed.

Brad

 

[msquared48]

Unless you want to push the beams to the limit here, most jurisdictions will accept sight grading, lowering the grade determined by two grade, and working with those allowable stresses.

Did you get these from an old tin storage warehouse, and are they clear grain Doug-Fir from the turn of the century?

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[BAretired]

noxer,

I too, question the approach you have outlined. Perhaps it would be prudent to review it one more time.

BA

 

[noxer]

dhengr: I don't think testing the beams in the field would work since if it fails, I will have completely ruined the beam. I don't have much experience with visual grading, but I would expect that the results would be very conservative and of not much use.

Brad085: What non destructive test methods are out there? I would love to see if it would be applicable to my project. By the way, the beams are in perfect condition...the issue is that my client would like to double the live load capacity of the floor. My initial calcs suggest that this is possible if I can justify a higher bending stress.

Thanks!

 

[a2mfk]

I wonder if it would not be more economical to design a composite section maybe using steel plate or something rather than spend time doing testing? Or some other method of reinforcing or strengthening the beam. I am guessing keeping the exposed aesthetics of the beams is paramount, maybe you can hide the steel plate at the bottom with some furred out wood casing...

 

[Brad805]

We have used stress wave testing on existing glulams that were exposed for years.

The problem I see with the testing, is you will get results for one specific beam and then you need to apply that to the others somehow.

I would use either steel plating or carbon fibers to reinforce the beams. We have reinforced beams to increse the bending capacity by as much as 250%.

Brad

 

[MiketheEngineer]

You could do a simple deflection test. Load a couple of beams say with a 1,000 lbs in the center. Accurately measure the deflection.

You know the Moment of Inertia - now you can find the E. Knowing what species of timber you have - you should now be able to select some Fb that makes sense.

 

[dhengr]

Noxer:
You didn’t say the beams were already part of an existing structure. Obviously that makes testing more difficult, but that also makes sampling more difficult. This isn’t exactly a problem that you will find the answer to on page 367, in three short sentences, in your wood design text book from school. This is a problem which involves a bunch of engineering judgement and experience, and some agreement from your client to back off if problems start to show themselves, and to hold you harmless, for what you can’t control. You are offering your best judgement as to how to stretch an existing structure to its limit without hurting anyone in the process.

But, from the engineering standpoint, you’re asking a difficult question and then leaving out about a hundred important pieces of info. The only thing you have really told us is that the beams are 11.5" x 11.5" x 12' span length. What kind of timbers are they, what spacing, what floor construction details, any plans, specs. and structural notes or calcs. which show lumber grade used, etc., what D.L & L.L. right now, what new L.L. and type of loading, duration of loadings (long or short term), tolerance for deflection and vibration or creep and permanent set, are beams adequately supported for new loads, bearing area on cols., C perpendicular, horiz. shear, do beams have any knots in the tension face, do they have any significant checking, and on and on and on. Again, get a timber grader involved to help you set the grade and allowable stresses.

You still might test, in place, to start to get a handle on the problem. We know how E (ksi) & I (“^4) fit in moment, stress, and deflection equations. You can calc. I for the beams, and will finally get a handle on some sort of a composite I for the floor system. Now, load a beam at its center with 200#, 3, 4, 500#, measure D.L. deflection and the defection under each new concentrated load. Do this on several different beams. Work the deflection calc. back wards, and see what this says about E and I composite. You will have to make some judgement about actual E & I comp. contributions to the deflection, and may not know the exact value of either. This is a crude form of machine stress grading, where they do know I for the piece of lumber, and measure deflection to determine E. Read the grading rules for the timber you are working with and also the NDS and its commentary on allowable stresses vs. ultimate stresses. You are still kinda flying by the seat of your pants, but you are honing in on the floor system and beam strength.

 

[shobroco]

I can't imagine that cutting a piece out of an existing beam & testing it is going to give you much useful information, since the strength of the beam depends on the grade and condition of the whole thing. Is your sample split or checked? Does it have cross grain? large knots? Are they on the tension face? etc. etc. A 12x12 spanning 12' is going to have pretty decent capacity anyway; are you sure that's where you want to start checking? What about the joists or the decking? I think Mike's & dhengr's suggestion of testing for deflection makes the most sense if you're not comfortable simply looking at the condition, grade, & species and making a judgement call. There's nothing wrong with being conservative if you're worried about what you see.

 

[BAretired]

I think a2mfk had the best idea. Reinforce the existing beam to carry the additional weight. You can use a steel plate each side or a steel plate on the bottom, fastened to the beam using glulam rivets.

BA

 

[HSIII]

I agree with a2mfk, the better, faster, safer bet would be to reinforce the existing wood beam.

And in case you're worrying about the wood strength, E, and I to use for the composite, I would believe that being conservative would be better than the notching and cutting which wouldn't really accurately tell much as far as bending strength, E, and I.

I only saw a wood beam notched out to be tested in a laboratory once, and to be honest the results I observed were far less than stellar.

Hope this helps.

 

[RHTPE]

The North American forestry industry conducted a massive testing program conducted over an eight year period, involved the destructive testing of 70,000 pieces of lumber from 33 species groups. This resulted in:
- A new test method standard, ASTM D4761, developed to cover the mechanical test methods used in the program.
- A new standard practice, ASTM D1990, was developed to codify procedures for establishing design values for visually graded dimension lumber from test results obtained from in-grade test programs.

The 1991 NDS design values were based on this testing program. You should check out:

http://www.fpl.fs.fed.us/documnts/fplgtr/fplgtr113/ch06.pdf

The problem with cutting out a test piece or pieces from your existing beam or beams is that you will have no way of insuring that the sample(s) are truly representative of all of them. Wood is nature's product, and as such it will be variable over the length and cross-section of any particular member.

I think that employing the services of a professional grader would be of equivalent cost to a serious testing prgram. I harvested a lot of timber from my land here in the Northeast and contact NELMA about having the resulting timber graded to satisfy the local building code and building inspector. They were willing to send a grader to my site and either grade stamp each piece or provide a certificate of grade for the batch. I would have to pay the grader and his travel expenses to have this done.

By having the existing beams graded by a professional, you have the advantage of knowing the grade of each piece (based on the current grading rules) AND you can easily apply the design critera of the 2005 (or later) NDS. Plus you have the bonus of not having compromised any of the existing beams by cutting out samples.

I think it's worth considering.


Ralph
Structures Consulting
Northeast USA

 

[RHTPE]

PS Noxer: Perhaps you should contact the closest mill to find out how they grade their structural lumber. They may actually have a grader on-site who might be able to examine what you have.



Ralph
Structures Consulting
Northeast USA

 

[RHTPE]

PPS[\b] IBC 2302.1.1 permits "... a certificate of inspection as to species and grade issued by a lumber-grading or inspection agency meeting the requirements of this section is permitted to be accepted for precut, remanufactured or rough-sawn lumber, and for sizes larger than 3 inches nominal thickness."

I would submit that your situation would be covered by this.


Ralph
Structures Consulting
Northeast USA

 

[druminman]

The ASTM tests must be conducted on specimens completely free of all defects, with straight grain. There is a correction for slope of grain. This requirement makes getting a representative sample somewhat more realistic, as one only needs to assure themselves that all of the beams in the structure are of the same species and grade.

I did this once on a project and removed one beam and one column completely in what would become a mechanical room, replacing them with a modern materials. I then sent them to a local cabinet shop with instructions for preparing the specimens, which I then sent to a testing lab. The whole process for two species (I tested bending and compression parallel and perp. to the grain) only cost about $2500.

Once you have graded the timber, you can calculate allowable stresses based on your testing results. The factor of safety for bending is about 5, even for Select structural grades. Even after this, and using conservative grading, I got allowable bending stresses around 4000psi. I wasn't comfortable using values that high, so I reduced them 50% further for my own comfort.

 

[a2mfk]

I hope Noxor has already designed some steel reinforcement and is now working on his next project... Its two beams for crying out loud.

BUT, all this is useful info if we are ever asked to analyze a large timber structure with many, many members. The it would seem economical to consider grading or testing.