Timber Joists

[Sasas2]

Hey,
Im undertaking some research at the moment which aims to determine how load is spread between timber floor joists. In order to achieve this I have made a floor joist system which consists of five joists. I have loaded the central joist in the system with two point loads at the 1/3 points along it's lenght. When testing the joist system I recorded deflections on each joist and also strain. It turns out that the strain gauges will not bond to the joists properly and this led to mis- reading of strain on each joist. All I have left is deflection readings to work with and im not sure how to accuratly determine the degree of load spread using just deflection. Also I cannot use Young's Modulus in any way as the system was only partially composite (joist- timber flooring)........any ideas on how to determine load spread using deflection? Thanks!

 

[hokie66]

The load and the deflection would be directly proportional. So twice the deflection, twice the load.

 

[BAretired]

Hey, Sasas2, Hokie66 came up with a sensible answer. Why would you attempt to attach strain gauges to wood? And why would you ask such a question?

BA

 

[msquared48]

I agree with Hokie here.

For the percentage of the total load taken by any particular joist, divide the deflection of that particular joist at the centerline of its span by the sum of the midspan deflections of all five joists. It's just that simple. No need for all the fancy stuff...

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

 

[frv]

Not trying to play devils advocate here, but I don't think it's that simple.

E varies significantly between two pieces of wood (heck, it even varies along the length of a single piece of wood). We use a relatively conservative value of E for design but, at best, it could be considered an "effective" modulus. That's perfectly OK for design, but not for research if one is attempting to understand something a bit more fundamental.

 

[BAretired]

I agree it is not simple, but the variability of E will always be present so an approximation is really all that can be expected.

BA

 

[csd72]

To allow for the variability of E apply a single point load along various points along the joists and take the deflections.

You can then do a statistical analysis to compare the average comparitive deflections. Remember that you have two sample for each load as the distribution should theoretically be a mirror image on both sides.

Also it is not usual practice to treat the joists as a composite T beam even if they are glued to the floor ply. This is usually done to minimise vibration rather than for strength.

 

[rscassar]

What is the flooring you are using to distribute the load to the outer joists? (19mm plywood??). I believe for the sake of the exercise it would be reasonable to use deflections to gauge how the load is distributed. If you really want to get exact you could load test each joist individually to determine an E-modulus for each. Also, I don't use the flooring and joist as a composite system for joist design but it would be interesting to see the results and how much stiffness increases with the addition of flooring.

 

[BAretired]

Another important variable is load duration. Permanent loading can be expected to change the results of a short term test due to creep of the various materials.

BA

 

[Sasas2]

Thanks...ya the flooring is 19 mm ply and I have already determined Young's Modulus of the individual joists.

 

[a2mfk]

Along with what BA said I'd also measure the moisture content of the joists at the time of testing, at least to have the data if you need it at a later date. I read recently that elevated moisture contents can affect creep due to some microscopic fiber action and the moisture acts as a lubricant, to put simply...

 

[apsix]

With hindsight; The individual joists should have been deflection tested before the load spread test to ensure reasonably uniform stiffness.

As it is the deflection results give as good an indication of relative applied load as the strain gauges would, if not better.
The deflection measurement will average out any stiffness variability along the joist length, which the strain gauge won't.