Hi All,
I'm getting ready to cut a small frame for an outdoor kitchen out of white oak and am wondering if anyone has pictures or anecdotes to convey just how much movement and opening to expect around joints? Because its just a naked frame the joinery will be quite on display, so I'd like to be prepared for what might happen visually. My experience to date has been more with softwood.

Does it help the case that the frame will sit air drying under its own roof, with no sudden drying caused by a heating system as in a house?

Whats a good timeframe between cutting the joints and raising?

Finally, is there any problem or benefit to be had from oiling the frame once its up?

Appreciate any advice, thanks for your time.
jon santiago

In my practice, I tend to fully house the joint abutments as much as possible to help conceal the shrinkage margins, also I tend to graduate widths of timbers. In oak the shrinkage is troublesome with braces so housing is essential to hide the change of angle and drift from checking in boxed heart braces, so free of heart in braces is helpful. You can oil the timber on the horses, right after cutting. End coat the joints with Anchorseal. Keep the timbers in shade. Erect frame as soon as you can. Drawbore the pins.

Roger is right on. In my experience oiling and sealing end grain help slow the drying process and that has to be good.
No direct sun on timbers if at all possible and raise it as soon as you can or the raising is twice as much work as you are trying to fix problems. It doesn't take long for the stuff to start moving.

Good design is key -- don't stack deep beams, avoid posted ridges where the post sits on a deep beam, etc.

If the timbers are going to sit more than a week or two before you cut them, you should seal the end grain of the raw timbers to cut down on end splits.


Also, use the absolute best oak you can get as it moves less (shrinks the same but twists and bows less).


It is common to see joinery gaps of 3/8" and checks up to 3/4".

Green oak can be a bit of a heart breaker, but to me, nothing else is as pretty and as solid looking as oiled oak timbers.

hmm...i do have a posted ridge in the design on top of a 9" deep beam. the tie beam will shrink away, and the ridge will want to settle, rafters will thrust? how big will that movement be and can i mitigate it somehow?

Jon, in your scenario, I would look at shrinkage across the frame and look for differences in shrinkage from the total accumulated at the ridge vs at the plate. Plus deflection under load of the tie can effect the position of the ridge.

If you have a building in service that you concerned about, set up a monitoring plan and track changes. If it appears that joint failure is ongoing, get the building surveyed by someone with relevant experience and is licensed to do so.

thanks roger. to clarify, this is just a design in process not yet built. its a small outdoor structure (12x12). perhaps im overthinking it

What is the depth of the ridge? Is it the same depth as the plate? If so, you can disregard that since they will drop the same amount. If the ridge is deeper than the plate, (say 12" vs 8") you have the shrinkage of the difference compounding the problem. In this example, that's 13" total of oak -- could be as much as 5/8" or 3/4".

If they are the same and you are just looking at 9" of oak shrinkage that would be more like 1/2".


I try not to do posted ridges in green wood. You can leave a gap between rafter plumb cuts and don't screw opposing rafters to each other - just to the ridge. That way when the ridge settles, the rafters don't thrust.

But with 1/2" of ridge drop, you can't really leave gaps that big. You would need a gap of 2x(tan roof pitch x 1/2". That would be 1" if it is 12 in 12 and 3/4" if it is 9 in 12.

Where posted ridges are used, this concern is well known

The solution has always been to let the rafters rest unattached on top of the plates and purlins, pegged or nailed only to the ridge. Then as the roof structure settles, the rafter ends will simply slide over the purlins/plates and not thrust on them

I do not believe a ridge supported roof system is a bad concept. I hold that the relevant shrinkage in the ridge is the portion under the bearing not the entire beam, so in a simple example, if a ledger is applied to ridge four inches up from the bottom of the ridge, then the shrinkage that effects the the rafters support is that four inch portion. Also keep in mind the ridge will shrink in width, so the likelihood of severe thrust at the rafter foot will be limited. The same applies to a mortised and/or housed rafter, it is the height of the bearing not the total height of the beam that would concern me.

I cannot think of a reason to affirm Davids no connection at rafter and plate.

Roger,

This is the way it is done in the cultures that traditionally use a ridge support style of roof framing. There are a lot of differences between a ridge roof and a truss style roof with plate support. We have to be careful to keep these differences in mind.

I focus on ridge roofs, it's my own personal favorite and it is the most versatile possible way of framing a roof. It's the only practical method of framing a shallow roof (thrust is too great otherwise)

In the life of the roof, 'floating' rafters help to ensure the overall integrity of the frame. Stiffness is attained by bridging or decking from rafter to rafter, if uplift is a major concern (It shouldn't be, the gusts that strike hung roofs in the Bernese Alps are matched only by hurricane or tornado force winds) then you can simply strap the rafters to the plate.