Roof Thrust load

[MEBob]

What is equation for rafter tie load ( thrust) for uniformally loaded roof? If the load were applied at the ridge the thrust load should be the vertical load divided by the pitch i.e. 3 times vertical load for a 4:12 roof but would seem to be less ( like half) for a uniformally distributed load. Tables seem to show some odd multiple less than half.

 

[Lion06]

This depends on whether the ridge has a ridge beam or not and what the stiffness of that beam is.

For a load directly on the ridge, the lateral thrust is only as you have noted if that ridge beam is not doing any work and has no stiffness whatsoever.

If the ridge isn't doing any work, then the thrust should be (wl^2)/(2h); where h is the vertical distance between the ridge and the eave, w is the uniformly distributed load, and l is the horizontal distance between the ridge and the eave.

 

[msquared48]

If you put a metal tie strap over the top of the rafters at the ridge beam, assuming the ridge beam is designed to take the vertical load, then there is no lateral kick to the wall, except due to the vertical deflection of the ridge beam and associated spreading of the walls.

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

 

[msquared48]

But even that lateral spreading will be restrained by the stiffness of the roof sheathing, if present.

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

 

[Ron]

Mike,
How will the sheathing restrain lateral thrust due to deflection of the ridge beam?

Ron

 

[Stillerz]

Yea, how does that work ?

 

[mmillerpe]

It would help to resist lateral thrust if it was tied into a shear wall at the end. The horizontal component (thrust load) would be resolved into the roof diaphragm and then into the shear wall(s)

 

[Stillerz]

One would have to consider a global shear lag effect in that case in order to consider points far from the shear wall.
Anyway, if near a shear wall, thrust loads would be inherently mitigated....by the shear wall.

 

[msquared48]

Ron:

The inherent stiffness of the roof diaphragm would act like a deep beam, not only:

1. slightly reducing the live load deflection of the beam, not the dead load (I did say slightly), but also...

2. reducing the associated lateral deflection of the wall.

The greater the roof pitch, the greater the effect, assuming you do have shear walls at the end of the diaphragm to develop the stiffness as previously mewntioned. Wew never really consider this effect, but it is there.

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

 

[dhengr]

Mike:
We know that roof sheathing diaphragm action is there, assuming it is well connected at the end walls. But, isn’t it true that a greater pitch will tend to reduce the ridge deflection, while a flatter pitch would be more effective in counteracting trust at the exterior walls?

 

[BAretired]

If the roof diaphragm is capable of taking the sloping load, then there is no need for rafter ties. The roof simply acts as a folded plate, transferring the load to the end walls.

BA

 

[msquared48]

Looking at the pitch of the roof, the greater the pitch, the greater the vertical component imparted to the ridge beam, the less deflection of the beam, and the less lateral spreading of the walls.

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

 

[cntw1953]

ms48 is correct on the pitch issue, and the notion that a stiffer roof diaphragm can minimize the lateral thrust on the wall, in the mid portion of wall span, by deform less. However, the effect of the latter phenomenon is difficult to measure, and it would not eliminate lateral thrust, thus is ignored in the design.

I have not dealt with wood design, correct me if there is mistake/mis-understanding.

 

[BAretired]

The reason we don't consider wood roofs as rigid diaphragms is that we don't trust them to carry the required membrane stresses. The main concern is the adequacy of nailed connections. If we were designing a concrete folded plate, we would not be concerned about lateral thrust at the side walls. The force at the roof/wall junction would be resolved into two components, one parallel to the roof diaphragm and the other parallel to the wall.

In the present example, we have not yet determined whether or not there is a ridge beam. If there is, the ridge beam carries half of the uniform load between side walls. The side walls each carry one quarter of that load plus load on eave projection if any. Rafter ties are not required as each rafter is a simple span.

If there is no beam at the ridge, then the rafter tie tension is WL/8h where W is the total load per unit of length on the roof, L is the span between side walls and h is the vertical distance from the middle of tie to the middle of rafters at the ridge line.

In the absence of both a ridge beam and rafter ties, the roof must act as a folded plate.



BA