Now we can take a look at the two previous joint mechanisms (half dovetailed and cogged), except this time secure the stretchers using two wedges instead of one.
There are a couple of different ways to go about it, and both involve inserting a wedge from each side. Here is the first type, where the wedges are gapped apart inside the post:
- Nuki to post connections double wedge a.jpg (585.3 KiB) Viewed 7751 times
In the background we see an example where the opposed wedges enter a mortise with a straight-across upper mortise wall, and in the foreground we have a version in which the upper end grain wall of the mortise is sloped to match the upper surfaces of the wedges.
While the version in the foreground has a bit better mechanical connection between the parts, as a result of the fact that the middle of the mortise dips downward to the midpoint, we must make the mortise taller overall and the wedges considerably thicker to allow enough space for the stretcher to enter upon assembly. Trade offs....
Connections in which the wedges are kept from touching one another especially suit situations where the post is of comparatively deep section and the mortise for the wedge to slip into, and have good bearing with, is long. The longer the better.
Consider for the moment the flip side, where the post is less deep than in the illustrated example - like a 2x6 timber for example. In order to keep the wedges apart from one another they would have to be extremely short, and that mean less security in the connection all around.
The other form of connection is to have the wedges slide by one another. Now, if the work is going to be concealed in a wall, under a floor, or something similar, we can pay less mind to appearances. Here is something appropriate to that scenario:
- DSC05414.jpg (183.61 KiB) Viewed 7751 times
For finish work, furniture or the like, the above work would be considered ugly. Too much stuff sticking out, even if the joint intersections were cut more cleanly, the pressure on the skinny end of the wedge as it emerges is likely to cause that skinny end to not lay flat upon the wedge below.
So, the clean solution is to make the wedges slightly different from one another in form, and bring each one just up to a hair shy of the wedge tip making a mortise exit:
- Nuki to post connections double wedge b.jpg (360.6 KiB) Viewed 7751 times
Note the wedges carefully, as the configurations do vary from one to the other.
To judge this connection properly, so that when the wedges are driven in tight, and do not evidence any tips poking out to view, is a matter of some finesse.
Now, in the article mentioned about a Van Arsdale sawhorse design, I do recall that the leg on that piece was made rectangular, and tapered on its narrow sides from base to top. Presumably this was done in interest of making the joint between the leg and the sill piece, or foot, as wide and strong as possible. That makes sense on account of the fact the the joint between post and foot is the weak spot in that type of sawhorse. Sometimes though, solving one thing with one aspect leads to challenges elsewhere in a structure.
In the next sketch, I have taken the former post section of 120mm, and made it half as deep, at 60mm:
- Nuki to post connections double wedge c.jpg (407.48 KiB) Viewed 7751 times
Another view, without the post sectioned shows the completed connection:
- Nuki to post connections double wedge d.jpg (279.05 KiB) Viewed 7750 times
I'm illustrating the cogged joint version, though the half dovetail connection is much the same in terms of the ramifications of marking the post section shallower than the stretcher is tall. Simply put, the double wedges have to ramp up at a much more extreme angle than before - halving the post depth makes the wedge slope double. Such steep wedge ramps make the wedges themselves more liable to come loose as a result of frame movement - - which can easily happen on a sawhorse when loaded with a heavy shiftable mass like a large timber. So, in order to keep the wedge angle more reasonable, one has the recourse of making the wedges longer, sticking out bother entry and exit faces, and having the ugly look, or going to a single wedge lock solution. I'd be inclined in that latter direction myself if faced with such a choice. It's a sawhorse and is utilitarian, but clean joinery is worth doing I think regardless.
The question is then: are either form of stretcher joint suited, really, to connecting to posts which are shallow in section? No. Does making the post shallower in section benefit these types of joint in any way? No.
There are other connection options better suited to the situation of a stretcher meeting a post of shallow section, however we have some more variants to look at with the standard square post yet, so that will form the subject of the next installment in this thread.