Wooden Sleeper Load Distribution

[seungkunoh]

Hi everyone,

I have a rooftop cell site where (2) existing joists fail under appurtenance loading (gravity only). Initial design was to install a blocking between (2) joists which fail under added load.

I was proposing to sister joists underneath, but GC proposes 4x8 DFL No. 2 sleeper to land a steel post. The sleeper sits on (4) existing 2x joists below (please see attached). Statically, I do not think sleeper can spread loads rather equally onto all (4) joists, rather most loading (gravity) should be resisted by (2) joists - Joist 2 & 3.

I have seen this approach many times by other engineers, but is it correct?


Thank you,

 

[Flotsam7018]

If the sleeper is relatively much stiffer than the existing joists the idea can work to spread the load out - a 4x8 probably won't cut it. If you can go with a deeper SCL beam or steel, that could get you there.

 

[jjl317]

Or you could add multiple layers of cribbing to force the load to spread out over the (4) joists - lowest level would span from 1 to 2 and 3 to 4. Middle level centered between 1 & 2, and 3 & 4, then top level span between the middle levels. Would require much stronger/stiffer blocking, but would force the load the way you want.

 

[Ron247]

A really stiff sleeper can most likely spread some load to joists 1 and 4, but it most likely will be less than the amount routed to joists 2 & 3. You should be able to make a computer model or hand calc that considers the stiffnesses of the joists and sleeper acting integrally to determine the distribution. Your model also needs to consider what happens if some other area loads (including wind) are on 1 or more joists but not the others since that is a real possibility.

You are essentially letting the initial sag of the 2 interior joists allow load routing to the outer joists until you reach equilibrium. For this reason, I do not see where the loads to each joist will be truly equal; and the inner ones will always be higher than the outer ones. One way to better balance them is too difficult to apply in real life. Leave some space between the inner joists so the initial load goes to the outer ones first before sag makes the sleeper come into contact with the inner ones and they start taking load. In both cases, you are hoping deflection calcs are 100% accurate, lumber is perfectly straight and the Greek God of Optimism is smiling on you. Since we are dealing with wood, right off the bat there is will be variance. Theoretical does not equal practical.

 

[JAE]

I like what Ron247 suggests - you can model these joists and sleeper to get a handle on the load distribution across the four joists.

However...........The locations of these sleepers along the joists spans will affect the results dramatically.

As you approach the ends of the 2x joists there stiffnesses go way up (getting near the bearing and small deflections under any load) such that almost all the load would go to the two inner joists (think infinitely stiff boundary conditions).

As you approach midspan, the joists are, to use a highly technical term, "squishier" and the load will get distributed across all four joists more equally.

 

[Ron247]

As you approach midspan, the joists are, to use a highly technical term, "squishier" and the load will get distributed across all four joists more equally.

In my first post I was trying to think of way to explain this concept about location along the joist but my wording got so complex I punted it because I had to run somewhere. I forgot about chapter 4 of my structures book; "Squishy Joists Are as Squishy Joists Do". Much simpler explanation.

 

[gte447f]

OP, the Steel Joists Institute (SJI) has published information on a method to distribute loads to multiple joists using a load distribution beam or strongback. The method is based on the relative stiffness of the joists and the distribution beam. You can find this information in SJI Technical Digest 12 as well as some of their other publications. The method should be applicable to your situation.