I really like the look of the spline tenons in a couple of Benson's recent books. I'd like to use some in a cottage I'm designing. Here's a specific question about them: The rafters are white oak, 7"x7"x13ft. Can I cut the through mortises (1" thick, 7" wide) for splines for the purlins (which are 4"x5") so that the purlins are 4' on center? there would be three such purlins between bents, and then a differently joined ridge purlin. Or would such spline mortises weaken the rafter too much? Thanks for your thoughts. Tim

in my neck of the woods, you can only remove 25% of your timber when notching. If all you are taking at the purlin joint location is 1x7(7sq in) and the beam is 7x7(49sq in) then you are well under the 25% rule. Is there also a housing to consider?

mark-- thanks for the reply. Yes, there is a housing to consider: 1 inch deep, therefore 1"x4"x5" total volume of wood removed. By your 25% rule, would that mean 20sq.inches in addition to the 7sq. inches for the spline, for a total of 27sq. inches, just over 50% (yikes!) of the rafter's dimension?? Maybe this idea won't work. . . I haven't done similar math to the traditional scheme of using dovetails on the purlins (with or without housings), but it seems to me intuitively that such joints remove as much if not more wood. Your thoughts?

easy now... we are looking in two dimensions only
Looking at the end view of the 7x7
so the housing is either 1x4 or 1x5 in the end view
worst case is the 1x5
so now we have 49 sq minus 7 for the spline and 5 for the housing leaving 37 sq in
this gives me 75.510%... should be ok.

Sorry but my view is that the purlin could be severely weakened. The traditional joint of choice would be a tusk tenon with a diminished haunch or to drop the plane of the principal rafter and run the purlin over the rafter. Nontraditional solution would be drop-in housing to receive square cut purlins with toe nails or log screws with simpson straps linking purlins across the rafter.

The tusk tenon and your spline method would require a more challenging assembly than a drop in or an over/under method.

I would also avoid the drop-in dovetail.

Roger,
I would be very interested to hear your reasons for not using the drop-in dovetail in this particular instance. As you are, I am big believer in traditional, historic joinery. Being a rookie and depending on my books for accurate info, one book in particular on traditional joinery advocates the use of drop-in dovetails. What are the drawbacks to its use in a cp/pr joint? Thanks. Ron

The drop-in dovetail removes critical compression fiber in the upper portion of the receiving beam. So I look for alternatives, a good commentary can be found in Cecil Hewett's "Historic English Carpentry"

If you choose to use dovetail drop-ins, I believe you should increase the beam size well beyond the calculated size of the undamaged beam.

raising the purlins sounds good to me.
you could still use splines if you wanted, to join the purlins together above the principal rafter. If the splines were long enough, they would show from below. Actually, it should be possible in this case to orient the splines perpendicular to the top face of the principal rafter, then they would show even more...

roger, in your first post, you say that the "purlin could be severly weakened" Did you mean the principal rafter?

In my three bent office frame with 8x8" principle rafters and flush purlins I've used drop in dovetails at the gable ends and tusk tennons w/ diminished haunch in the middle which gives me an unbroken plane of wood in compression on top of the rafter pair that takes the most weight.
This seemed to be a good balance between maximum strength and ease of assembly. The tusks are just barely pegable with a full 8" rafter, I think raising the purlins is the best approach for narrower rafters. I'm not sure how the ridge purlin was going to go in but you might want to think about a king post.

A solution I have seen a photo of and liked, but never used looked like this:

Prin rafters moved inside roof plane, purlins (in this case prin purlins with common rafters on top, but could work as well for common purlins) interrupted at each prin rafter sitting in a small notch about 1" or so into the rafter and 2" or so down the face of the rafter. The purlin was kept from rolling with a shaped sprocket spiked or screwed to the principal rafter. The purlins were joined across the top of the rafter like Mark was saying with a spline (a blind top spline in this case). Wind braces in the plane of the purlin running from the purlin down to the prin rafter (sitting on top, I assume lagged).

This was from an English timber frame company.

It seemed to me to be a nice design because it has the ease of raising of interrupted purlins and a lot of the strength of continuous purlins. Also, minimal notching of the principal rafter, especially compared to the dreaded dovetails. Much easier to assemble than tenoned purlins.

Anybody seen or done anything like this over here?

cheers,
GH