prism

The first time I looked at this particular plan of Johan van Lengen for a bamboo house, I thought it was boring, the second time I saw that it was very strong, and the third made me realize that it was based on a simple polyhedra - a hexagonal prism, and the gable truss was simply a portion of a hexagon and hexagram combined.  I then realized that all longhouses are essentially geometrical prisms.  So here is one way of generating them.  While the pentagonal prism is the simplest, the vertical posts on the hexagonal prism engenders a more intuitive design, and the steeper slope of the roof on Lengen's drawing is more aesthetically pleasing while allowing for the physical construction of the building using poles.

Below is a clearer image of how such a building could be constructed, with end and side view.  The sheathing and roofing material is left out showing only the bare structural members.  On the side view, the diagonal bracing members run from the bottom left to the upper right; an additional set running perpendicular to these may be added to improve rigidity. The sheathing, once added, will also contribute to the strength of the building. 

In this layout, the downward force exerted by the ridge beam will tend to push apart the posts it rests upon, and so a better design would involve placing these two posts in the same plane, cutting their tops to allow them to rest against one another while cradling the ridge beam above. Fastened to the two ridge beam bearing posts (that form a more or less equilateral triangle shape) are the outside vertical posts and beam in a second plane. If the rafters locked together supporting each other in pairs, as they do in a Norwegain trestle frame building, then the ridge beam and its supporting posts would be unnecessary. But having a central beam eliminates much of the outward force the rafters would otherwise exert, thereby allowing a simpler wall structure. Also, the posts supporting the ridge beam help triangulate the structure, and thus serve a dual purpose. Despite these advantages, because the ridge beam rests on a joint (in buildings this tends to be a weak spot) and not a single post, the utility of this design is necessarily limited to small structures that do not need to bear too much weight. As this is only one of many plans Johan van Lengen provides, and he is writing to a particular target audience, I am sure he is more aware of this than I.

As all buildings, no matter how carefully conceived and crafted, are subject to eventual failure and dissolution, is there a way around this particular problem?  As van Lengen describes the building, it is composed of lashed poles.  In such buildings the lashings must be periodically tightened or replaced with new lashings.  If built this way, then even if the ridge beam forces apart the joint it rests upon, periodical re-tightening will bring it back into its proper position.  It is important to recognize that as builders, we often co-opt objects designed for one purpose and use them to fulfill another purpose for which they are only incidentally suited.  A tree must only stand tall and bear vertical forces, but a building is composed of trees placed at many angles, and so joints are created that eventually will fail before the trees themselves do.  If van Lengen's house were built of concrete, then the central equilateral triangle could be made without joints, or even turned into a catenary arch.  If it were, there would be no problematic structural issues at all.

One solution arrived at hundreds of years ago is evident in the construction of the English "cruck house".  The scale here is generally larger than the bamboo pole buildings described up till now.  I had disliked the appearance of cruck frame houses, but now that I see them in a different light, they appear very interesting. Perhaps I could make a small one with poles?

Above is a simple illustration I made of a cruck house. The "ridge beam" (top red circle) rests on the "yoke" (short green bar) which binds the top of the "cruck blades" (blue diagonals). There is also a "tie beam" (longer green bar) and two "wall beams" (lower red areas), that support the lower end of the rafters (yellow diagonals). Also visible are two vertical posts (brown bars). Variations on this basic theme are visible in many cruck buildings, the photos of which are very instructive.

It is notable to mention that Ben Law built a house exactly like the one van Lengen drew and wrote a book documenting the entire process (a photo of one of his buildings is above). It was also featured on TV.  Due to the careful attention to joinery details such a house would require, I would first build one like the cruck diagram I created, which I think would be simpler since it includes the "yoke", a detail excluded from Ben Law's design (though increasing its contemporary appeal).

---

3/20/10 I just got an inter-library loan book. 

An interesting point and comparison of cruck buildings is made by N W Alcock, in his book Cruck Construction: An Introduction and Catalogue on the first page:

"Cruck trusses have only one essential joint, that at their apex holding the blades together.  All their attached timbers can be removed or replaced, and even the feet (whether ground-fast or set on sill beams) can be reset without destroying their integrity.  Furthermore, their apexes are well protected from weathering and decay.  Contrast a box-framed building: its vital joint, between post, tie-beam, and wall-plate, is complicated, weakes the timbers, is exposed to the weather, and is unrepairable if it decays."  

I would add that this arrangement of two cruck blades is also a minimalist design, as any fewer, that is to say just one central post, would not provide sufficient stability for a building.