These posts are of my own design concept (and are copyrighted), and are intended to show the general arrangement of pieces in the structure. So, certain joinery and geometrical details have been omitted from the illustrations. I take the liberty of noting that if you seek to emulate this design in some manner, be forewarned that it is somewhat difficult to execute. Pentagonal roof framing with a reciprocal support ring composed of parallelogram shaped beams, and incorporating a double roof with curved eaves is not a basic roof form. Ideally, you will have want to have mastered regular hip roof framing with joined timbers, then polygonal roof framing with timbers, and then standard hip roof, double form, with curved eave before venturing into this territory.
At the conclusion of the last post, the reciprocal beam assembly was in place and the lower hip rafter were fitted:
You can see how the reciprocal beams have been extended beyond the wall plate to pick up the outrigger beam at midpoint. Also, note that the upper ends of the reciprocal beams have been trimmed so as to form a leveled surface.
Next, an inner purlin ring is fitted:
Then the main lower hip rafters are fitted:
Next, the common rafters are shown, which, in this example are of parallel type not fanned:
Atop the common rafters are two layers of perimeter eave fascia, the hirokomai and yodo respectively:
A perspective view of the half-completed roof structure, which shows the curved eave line a bit better:
A view from directly above:
And the view which is a prime consideration in the design - inside the building looking up (click the image to enlarge):
A bit like a snowflake, no?
In the next post in the series I will add more parts as the roof structure builds. I hope you've enjoyed the tour so far. Thanks for dropping by the Carpentry Way. On to part 8
Chris, this is awesome. I hope you get commissioned to build one!
ReplyDeleteTico,
ReplyDeleteglad you like the design so far and I also hope for that commission on day. The design process has been rewarding in and of itself.
~C
Can you share what load this roof is designed to carry?
ReplyDeleteSince I'm not used to designing in wood (I'm used to working with composites) it kind of looks very heavy to me.
Roland,
ReplyDeletethanks for your comment and question. I might suggest you take a look back at post 4 of this thread;
(http://thecarpentryway.blogspot.com/2012/02/orphan-story-of-gazebo-iv.html);
especially the Japanese structures pictured therein, and then consider again if my structure looks so 'heavy' by comparison.
I have been working this as a design exercise primarily and the roof is capable, I would expect, of carrying a heavy tile roof. However I would tend to prefer a lighter copper shingle roof on the same frame - that is the intended covering. I haven't taken it to the stage of calculating for loads, as that would require knowing something about the environment in which the building is to be placed (ie., typical snow and wind loads, seismic conditions, etc.) and something about which woods are to be employed for the framing, whether they are dried, etc..
I am designing more upon a 'rule of thumb' basis at this point, and upon common Japanese proportioning methods. In that regard, personally I tend to like a slightly 'chunky' look to the framing, as opposed to the 'bird cage' or 'minimalist' aesthetic.
Your question is an interesting one in light of your deign work in composites. Wood is different in that its fibers can not be as optimally arranged for the various loads imparted on the structure, and it is weakened where there are joints and connections. therefore you will always need more of it to deal with the imposed loads and a certain redundancy is inevitable.
There are certainly methods of designing a roof structure with lighter wooden components, but there are always trade-offs. In wood, it is often the case that the more 'efficient' a structure is, the more pared-down in elements, the less resilient and durable it tends to be.
For instance, I love trusses, and they certainly can be made out of some slender pieces of wood - they usually are these days - and work quite well for their design loads, but in a fire they burn pretty quick and they are not really intended to last long. Aesthetically too they don't always look so nice. With timber trusses too the designs are such that the loads on the material can become quite high and I've seen plenty of cases where distortion/cracking results. Really, combining long spans, high loads and wooden trusses is really a recipe for METAL.
I tend to like a certain amount of redundancy and slight over-sizing in timber structures, especially one like this where most of the loads are simple bending in compression, as it is in this case.
A good case in point for redundancy I suppose, would be a Town lattice bridge truss as compared to a conventional trussed bridge frame.
As to the visual 'heaviness', please keep in mind that many of the structural components currently visible in the drawings will be largely concealed by the eave boards, ceiling, etc. The rafters for instance, are fairly svelte and they make up the largest visual component of the eaves. The hip rafters are heavy and that is a lesson I have brought forward by looking at a lot of old Japanese hip roof frames and how they deform over time. Again, check out post 4 and look both at the hip rafter used in Japanese pavilion structures and the size of reciprocal beams employed in the Bunraku Puppet Theater. All of these structures are engineered for seismically active environments, and with both snow and wind loads in mind.
~C
Compared to the other pictures it is indeed not that heavy. It struck me because of the relative size of the rafters when compared to the pillars, because those are subject to compression loads which could lead to buckling; easily the worst load to subject a beam to.
ReplyDeleteYour remarks comparing wood to composites are interesting. I would say more or less the opposite! Wooden beams are in my view principally very similar to unidirectional composites, which I tend to use the most. It is possible to make a quasi-isotropic material with as little as 3 or 4 layers of composite material, but doing that drastically lowers the performance of the material. When one uses unidirectionally reinforced composites is when the material really shines.
I've been looking at wood as a more renewable alternative compared to high-performance carbon fibers, but the relatively large range in material properties does indeed make for relatively low design allowables.
A Town lattice truss is visually quite appealing, I think. Simple elements combined in a sturdy yet open structure. It has a kind of elegance. Although personally I find a variant like the geodetic frame of the Vickers Wellington even more interesting, because of the continuous load paths and huge tolerance for damage.
But I wonder why sandwich construction in wood has never really taken off? (with the exception of some De Haviland planes, most notably the Mosquito). I'm guessing that a sandwich roof made from 1/2" end-grain Balsa clad on both sides in two layers of say 1/16" wooden strips in suitable directions would make a relatively light yet strong structure.
Roland,
ReplyDeletethanks for your follow up. My comment about composites was in terms of the easy arrangement of fibers, which is not really possible in solid wood except by glue lamination.
I like the Town lattice truss as well. It's a common bridge form and there are plenty around here to view - most no longer in service however.
Sandwich construction in wood - the description you provide of a balsa clad roof might be light and relatively strong, but I'm not so sure about the durability.
I prefer to avoid glues myself - most of my use of them comes in doing patches and repairs.
An about the rafters: please keep in mind that the drawings in the post are showing the decorative rafters, and those rafters do not bear a significant portion of the roof load.
Thanks again for your comment. I'm always interested in learning from people who work with different materials than I do.
~C
Why a pentagon?
ReplyDeleteKathy,
Deletecheck out earlier posts in this series for a thorough description of the various polygons, and the relationship between reciprocal framing and different polygons, and the reason for a pentagonal plan will become clear. See part Vi in particular:
http://thecarpentryway.blogspot.com/2012/02/story-of-gazebo-vi.html
~C