On the current series of This Old House, they're (among other things)
putting a two-story, one-room-wide-ish addition on a house. The
addition has a gambrel roof, at right angles to the main roof.
For the ridge on the addition, they're using a double 2x12. This seems
kind of massive, to me. For instance, our (old) barn (not large, but
way bigger than their addition), also has a gambrel roof, and just a
1-x-whatever for the ridge board.
I'd be interested in any thoughts on this.
I would guess that your barn didn't have a cathedral ceiling, and that
their addition did. Their ridge is a structural ridge, holding up the
roof, and needs to be pretty big, especially if a snow load is
involved. A double 2x12 is on the small side. We have a 20' long
addition on the back of our house that has a double 14" Microlam as
If that's a structural ridge, a double 2x12 seems kind of small to
me. There may have been a steel plate sandwiched in between, or the
2xs were actually LVLs.
A structural ridge, such as the ridge over a cathedral ceiling,
requires a much bigger design load than a standard ridge. The ceiling
joists in a standard ceiling situation tie the rafters together and
keep them from spreading. This can also be accomplished by collar
ties at a higher level, but may require larger rafters and/or stronger
connections to the rafters.
There frequently isn't a ridge board in timber frame construction, and
the ridge board in a stick built house is mainly there to assist in
building the roof. In most typical stick-framed roof situations you
could omit the ridge board entirely and it wouldn't appreciably affect
the roof strength. A 2x4 truss roof doesn't typically have a ridge
It's all about the loads.
OK, thanks. Just to clarify: with collar ties (and suitably-sized
rafters), the ridge load would be much lower? When you say the ridge is
holding up the roof, does that mean their rafters are in tension?
The ridge load is really nothing. It is being held up by the
rafters. And you are correct that a structural ridge puts the
rafters are in tension (I think).
I am not a structural engineer, but I did point out a math error to
the one that did the calcs for my addition, causing him to change the
footing diameter from 6" to 18"!
Consider a frame consisting of two walls studs and two rafters, with
hinges at all the connections. That configuration is not stable: it
can freely deform via a mechanism in which the ridge falls and the
tops of the walls spread apart. The two usual ways to deal with this
are either to provide a tension element (ceiling joist or collar tie)
to prevent the walls from spreading, or to provide a structural ridge
element which is properly supported.
I think it is typical to do just one or the other, so if you are using
collar ties, there would be no ridge load at all. But you could
design a hybrid solution that uses both collar ties and a structural
ridge. In that case, yes, the collar ties would reduce the load on
Keep in mind that the gravity load of the roof provides a distributed
axial force on the rafter, so whether the rafter is in compression or
tension can change along its length. When using a structural ridge,
the reaction from the ridge puts the top end of the rafter in
tensions, but the other loads are compressive, so the bottom end of
the rafter is in compression.