A few years back I had the opportunity to work on a project a bit south of San Francisco for a very wealthy individual whom I'll refer to as Client E. This project had been ongoing for 5 years at the point which I joined it, and comprised a 35 acre property that had been transformed by the construction of an multi-acre artificial lake, ozonated no less, and some dozen buildings in the look of Japanese classic architectural tradition(s). This project was very much the Hearst Castle of our day, and unlike any job I had been on before, this one had no budget. By that I mean, there was no concern about how much something cost or how long it took to do, only that it be done to the highest standard. In fact, even if the work was done to perfection, there were cases where it would be ripped out after completion if the client, or a designer employed by the client decided that they didn't quite like the look or specification after all. That could be hard to deal with from the carpenter's perspective, trying to do your best and yet knowing that it might all be for naught, to be ripped out and then done again.
Sometimes this do it and re-do it mentality could get a little out of hand it seemed to me. A custom bathroom faucet in one of the buildings had been the subject of so many prototypes and design revisions, and designer turf battles that it ended up costing close to $40,000. Moving a toilet in the main residence building, after the G.C. had not taken sufficient notice of the client's request in the matter from a couple of years prior, ended up turning into a complicated affair and cost something like $125,000 to achieve by the time it was actually moved - and that's a move of some 3' or so.
It was a bit of a surreal work environment to say the least. Some days I was utterly repelled by the excess of it all, and other days I felt really lucky to have the chance to work on a project where the objective was to do my best work, not the usual 'get it done and get on to the next thing' attitude which is the norm elsewhere. I lived in a rented house a few miles from site, along with half a dozen other site carpenters. We were transported to the site each morning in a company van, had to pass through a security checkpoint and then get out of the van and walk through a foam pad soaked with anti-fungal chemicals, in an attempt to keep a fungus that affected oak trees from spreading to the site. It was not your normal work site by any stretch.
I think the final cost on this project came in around $200,000,000, though I'm not exactly sure. Last I heard the client had purchased the neighboring property and was having some construction work done there as well. I had the great fortune to work on several of the buildings on site and to work with some of the best carpenters in the Bay area, so it was really a tremendous opportunity in all respects, and I learned so much during that period. There were some colorful personalities among the crew, most of whom I hit it off with very well, and a few, ah, not so much.
The last project to be tackled on this site was the only building near to the road, the guardhouse. We called it the 'guard shack', since it was a tiny building, 14' x 21' or so, in comparison to the others on the site, and at a million dollars or so, relatively inexpensive. The head carpenter on this job was Walter, here studying one of the roof layout full scale renderings:
Walter had in fact laid out the roofs for most of the buildings on site, and is thus far the only other person I have met who shares my passion for the Japanese roof and who has studied it to the same extent I have. We often spent our lunch hours poring over a Japanese layout text or two, and I would help with the readings of certain sections, since I can read Japanese to a decent extent. Walter cannot read Japanese, so it is a real testament to his dedication that he has reached his pinnacle of expertise on the topic purely from studying the descriptive geometrical drawings in Japanese layout texts. He is a Harvard graduate, so he certainly has the smarts necessary to get into the material. He's truly a master carpenter.
You might notice from the previous picture that the building is inside a scaffolded structure. This is the normal way things are done in Japanese traditional architecture - a scaffolded and roofed building is erected over the foundation so that all woodwork can be carried out without worry of damage to the valuable material from the sun and rain. That way, when the project is complete, the scaffolded structure can be removed, kind of like unwrapping a gift, and the client is presented with a absolutely pristine building, no water stains or sun-bleached parts anywhere. The Japanese follow this practice even for extremely large projects, where the scaffolded structure cost can run into the millions:
The 'guard shack' had an upper hipped roof which was mildly convex (termed a mukuri roof), and a lower wrap-around roof, or hisashi, which was concave, and hipped on two corners. The roof structure was done in the manner of a Japanese roof, with both exposed decorative eave components and hidden structural rafters. The decorative rafters were round poles, and these in fact came from Japan, and are grown specifically to be rafters:
These rafters cost anywhere from $300~500 a piece - a lot of the traditional Japanese sukiya building materials tend to be similarly pricey like that - despite the rustic appearance, these parts aren't simply dragged out of the bush. The rafters are grown from a tree that is cut, coppice-fashion, down low, and the resulting shoots that grow up from the stump tend to be relatively un-tapered and are pruned so as to be branch-free. In this decorative exposed eave, things are further complicated by having the rafters fan, and sweep up to an upturned hip, not quite so visible however in this view:
We've wrapped the blue painters tape around the rafter tips because at that stage we were working on determining the best way to cut the rafter tails. With a fanning rafter arrangement there are three traditional Japanese methods of dealing with the rafter tip cuts. Here's a view of that hip corner from up high:
And a slightly better 'big-picture' view of the situation:
Walter and I both concluded that one method in particular of marking the end cuts gave the most pleasing result - it allowed for a different cut on each rafter so that as they fanned they presented a fairly similar look at their end grain when viewed from the ground. However, we did not have final say in this decision, and when the 'decider' (not the client) came around he chose one of the simpler methods after we presented the options to him. He and Walter did not get along so well, so I think the decision to do something different than what Walter and I had determined was the best way, was based in part upon their interpersonal rivalry, but hey, that's water (Walter?) under the bridge now.
We cut the rafter tips the way the 'decider' wanted:
In the above photo you can see the roof surfaces essentially complete except for the covering material, which was to be bark shingles. Unlike the bark shingles described in yesterday's post, these came not from a cedar or cypress tree, and not from Japan: these shingles came from Redwood tree bark.
The Redwood bark used came from the base of the tree, and this raw material was several inches thick. There was no way that the bark was simply peeled from the tree in the sustainable manner of practice followed in Japan - it's not possible to pull 4~6" thick bark from a tree, so undoubtedly this bark came from fairly large Redwood tree(s) that had been felled (hopefully not just for the bark). While I have no idea as to the source of this material, I did feel a bit glum about using it.
When we got the shingles, they had been cut by some mill into even-thickness, about 0.75", however we needed them tapered. I made a tapering jig for the planer, a little Makita jobbie, and one of the laborers spent some 2 weeks running shingles through that planer, a machine which, in the end, was destroyed by that work. The planing was really noisy, dusty and toxic work - again, contrast this to the sensible Japanese practice of working with bark detailed yesterday.
So we had the tapered shingles and spent several weeks laying them up along the roof edge. Once the lay-up was complete, it looked like this:
The next step was the critical one of trimming the edge of the shingle build up. Again, such matters are given serious consideration in Japanese roof work, as the way the edge of the roof is cut can have strikingly different consequences for the appearance of the roof edge when viewed from the ground. Since the roof curved upwards at each hip corner, the treatment of the face and lower edge of the shingled edge could flow seamlessly with that up-sweep, or if done in error could clash with the rafter and eave perimeter fascia and look mis-cut. Done incorrectly, the surface might present a twisted surface, or make the perimeter fascia below look as if it were not straight.
We spent a fair amount of time determining the best way to proceed. in previous buildings on site we had cut off the edge of the roof (conventionally shingled in wood) with a handsaw, and then hand-planed the shingle end grain, however bark did not lend itself to being cut with a handsaw or hand-planed. Barks got lots of grit in it and this would trash our valuable hand tools.
Remember again from yesterday how the Japanese work the bark with hatchet and adze. That wasn't our plan. I proposed that we could set up a long guide track for the big Makita circular saw to run along, and that's what we did - here's the top view of the track after set up:
You might call it 'early Festool' (inside joke for the woodworkers in the readership).
So, the moment came, and Walter was a bit anxious, because the final appearance of the roof, and thus judgment of its appearance, boiled down largely to this one cut, so we had to get it right:
However, after running the saw in a short distance, things weren't going so well. The saw blade had wandered off the cut line and the cut was spoiled. Walter was really stressing, and very worried that the whole show was all screwed up.
From my furniture work, I had some experience cutting difficult abrasive hardwoods on the table saw, and had learned how critical blade sharpness can be. Though the blade in our Makita seemed sharp enough to the touch, I knew that could be a little misleading and that very slight differences in blade sharpness could make all the difference in having a blade hold to a line or wander slightly from deflection. I proposed that I could go to the tool store and locate a new blade, explaining my reasoning to Walter. He was a bit 'doom and gloom' at that point, but realized it was worth a shot. While I was away, Walter would re-set the guide jig slightly so we could- hopefully - create a fresh cut line.
An hour or so later I was back from the tool store, and luckily I had managed to find a blade (it's not a particularly common model) for that saw. I fitted the new blade, on went the respirators again, for another kick at the can. It was 'do or die', there was no more room to play with in the cut line, and if we messed up this cut, the roof shingles would have to be torn off and re-done. The 'moment of truth':
Well, all went well. The new blade made all the difference and we were able to execute a smooth and clean cut. There was great relief on Walter's part:
And I was pretty chuffed at having solved the problem and getting a good result, and I think I make quite a fashion statement too:
It's the 'barking up the wrong tree' look, perhaps.
Here's a view of the edge of the fresh cut roof:
Form there, we worked our way around the rest of the roof edges, and then the remainder of the roof shingling could be completed:
Here's a view of the gabled returns on the lower hisashi after cutting and shingling had been completed, along with a view of the main roof:
You have to be really careful in working with such valuable materials, that's why we considered every decision carefully and kept the components protected with painter's tape and foam padding throughout the project. You can see also in the above photo how the ridge of the roof was treated, using a Japanese-sourced ceramic ridge cap, the ends of which were infilled with bark shingles. I left the site soon after our work was done, and I've never seen that building out in the daylight.
So, that's the story of my bark roof shingling experience. In retrospect, it certainly wasn't a material I would want to work again, at least not in the manner we worked it. The Japanese approach the use of a material like that much more sensibly it seems to me, and if things could be done over here in a similar manner, then bark might very well become more of an option. I'm not sure how durable that Redwood bark roof will prove to be. Bark is inert essentially, not absorbing moisture, so it doesn't move much in service, which is a plus. I think the look of a bark roof is quite pleasing, so we'll see how that balances against initial cost and service life.
It is often difficult to choose a roofing material for building, when looking at it from a myriad of perspectives - material cost, durability, environmental load, appearance, ease of install, interplay with structural and other roof components, etc. I feel fortunate to live in the North East now, where a great material for roofing is found in abundance and can even be readily re-claimed from old buildings: slate. I'm looking forward to opportunities to work with that material, something that I had no chance to use on the West coast, where it is rare and expensive (and most of that expense is in the trucking it seems). The French, I note, have some very sophisticated slating practices, including working with curvilinear slate, so I have much to study and learn in that regard.