Composite Timber Beams 1

[KutEng]

I have a residential floor framing that is being converted into a shop and thus an increase in loads.

As a result, the existing floor system cannot take the increase in loads.

I need to nail and glue laminate a new timber member to the existing beams to increase the strength capacity (deflections are not an issue here). The new member will be the same dimensions (240x70) however will have different section properties as the existing timber. (The existing timber is 40-year-old Oregon, the new timber is Hyspan LVL).

My existing beam fails in bending and has a characteristic bending value of fb'=31 MPa. My new beam has a characteristic bending value of fb'=46.5 MPa.
My question is when recalculating my strength in bending - do I have to use the lowest fb' value, or due to composite action can I take an average of the two values?

Alternatively, am I able to calculate the relative stiffness between the two members and assume relative load distributions between the members and just design the existing and new members separately using their relative loads?

 

[struct_eeyore]

Composite action will result in a lower stress per a given loading, but it will not have any effect on the material properties itself. Other questions - Is the existing beam carrying any load - make sure to take the pres-tress into account if so, or full unload the beam. Also, glue and nails have different elastic response when loaded, and you should ideally only design the load to be carried by one or the other.

 

[KutEng]

Is the existing beam carrying any load - make sure to take the pres-tress into account if so, or full unload the beam

I believe the existing beam will be propped up when the new beam is installed - the existing beam will only be carrying the timber floorboards above during the installation.

glue and nails have different elastic response when loaded

Could you please elaborate on this

If the composite beam was loaded to the existing beams capacity, then loaded some more - would the extra load go into the new beam or would my existing beam fail and the entire load be taken by the new beam?

 

[Ceinostuv]

If the joists are properly fastened together, divvying the load based on stiffnesses is what I would do/have done. Verifying that the existing members will be propped before installation of the new joists will be critical for this to be accurate.

Judgement-In-Training

 

[KutEng]

What if the new members are only required to take the extra live loads and the dead loads can be carried by the existing beam? Will propping still be required?

 

[struct_eeyore]

the way I see it, you will still have spillover of liveloads into the old beam once they are in place. You can prorate the stress based on the existing dead load, but you composite beam still has to pick up the entire live load. to clarify my previous comment re. glue and nails - it's the same principle with welds and bolts (where welds are much stiffer and will pick up most all load before it reaches the bolts) - except here glue would be stiffer than the nails, if you use both, make sure you design the glue to carry the entire load.

 

[Lomarandil]

Without propping, the existing dead loads will be carried entirely by the existing beam, and the entirety of live loads (old and new) will be divided between the beams according to stiffness.

Without running the numbers ourselves, it sounds like this will likely overload your existing beam, causing it to fail, and then overloading your new beam when it stands alone.

As for glue and nails having different stiffnesses, nails tend to slip a lot before engaging fully. They're not particularly for building up composite beams for that reason. So similarly to the beams, the glue would take the load at first, and if not adequate to carry the load by itself, would fail in shear before the nails carry much if any load.

That said, for beams of the same dimension side-by-side (I presume) and both being loaded by the subfloor above (or whatnot), you don't see any major benefit by connecting the two.

Edit: Structee beat me to it!

----
just call me Lo.

 

[JoshPlumSE]

I went through a similar situation in a previous project. We reinforced existing roof joists by sister-ing in some LVL's. My thoughts based on this experience:
1) First thing I did was split the loading between the two based on stiffness similar to what Ceinostuv suggested. This resulted in the LVL taking a lot of load and being "overstressed" by my calculation.
2) I talked about this with the SEOR of the project. He suggested calculating the capacity of the two beams independently, and adding them together regardless of stiffness.
3) The rational, which I agree with, was that this wasn't a brittle, catastrophic failure mode. So, as the one beam (the LVL) gets overtressed in tension, the additional moment / stress that would have gone into that beam bleeds over into the original one....
4) Also, there was the fact that a portion of the dead load probably would stay with the original beam as well since we weren't removing the roof, or shoring the beam.
5) The key to me is that we did enough calculations to justify why the SEOR felt comfortable stamping the drawings / design, and that we felt the design was perfectly safe.

 

[KutEng]

Thank you all for the input,

Another question I have which Structee alluded to above - since the existing members are quite old they would have suffered from creep deflection and would have in essence pre-stressed to a certain point. If the existing members are propped, the new members installed and then the props removed, the existing beam will want to deflect back to that point without any loads on the members due to creep correct? Will this impose a load on the new member which also needs to be considered?

 

[Lomarandil]

Josh, I'm not certain that timber bending overstress leads to a ductile failure. I don't have anything to back that statement up, but tension/shear in timber feels brittle to me.

----
just call me Lo.

 

[KootK]

I don't have anything to back that statement up, but tension/shear in timber feels brittle to me.

It is. We pretty much settled the bejeezus out of that question here with videos and everything: Link

 

[KootK]

Will this impose a load on the new member which also needs to be considered?

All this fancy stuff is true but I'd not get lost in the weeds. If it's me:

1) I'm splitting the load based on stiffness and keeping the stresses in each piece at 85% max stress on this basis.

2) I'll assume that in shoving the new, same depth piece in there, some jacking de-stressing has occurred naturally.

3) To hell with creep, you'll never estimate that with any accuracy anyhow.

4) With these being side by side beasties, I see no reason at all to glue them. Just nail/screw to bring maybe 2/3 of the total load from the existing into the new. Maybe screws to reduce squeak potential.