Wednesday, October 30, 2013

BCM Door Repair

The past few weeks have been hectic, and I seem to have a few things on the go - whew!

I've been out to Oregon to look at material for the MFA project, and will be updating that thread soon enough. I've been deep into gate design these past 10 days.

I also work with the Boston Children's Museum, which has a complete 100+ year-old Japanese merchant's house, or machi-ya, installed within. Regular blog readers may recall a series from a few years back where I constructed a replacement hinged bench, battari-shōgi (⇐ link) on a sawhorse in my kitchen. Ah the good old days!

Just as with that hinged bench, the front door of the house takes some extraordinary wear and tear given the conditions of the installation, which sees hundreds of young and excited children descend upon the house daily. Recently the door lock completely broke and the Museum asked me to effect repairs. As with the battari-shōgi, I am choosing to replace the door's stile, which was originally made in a softwood, and stained dark brown, with wenge. Wenge, after rubbing down with finishing oil, looks pretty close to the stained pine, however it is many times tougher. It's a pragmatic solution to an unusual situation.

The Museum delivered the door to my shop last week, and this week I was able to give it some attention, commencing work by jointing and planing the wenge to yield a replacement stile which was 30mm x 54mm. Then I transferred marks over from the old stile, along with the exact center of the dor lock catch, and then cut off the old stile. What was revealed were multiple tenons on the door rails and nuki:

The replacement wenge stile lays atop the door, and has already been mortised for the multitude of tenons in various sizes.

The bottom of the door employs un-haunched quadruple tenons:

The top rail uses haunched double tenons:

Both top and bottom rails have a mitered return on the front face only.

The nuki have slightly shouldered bare tenons:

Here are the corresponding mortises for the upper rail, the cut out involving 0.25" hollow mortise chisel with a 0.180" web in between the mortises:

 The lower portion of the stile mortised for the quad tenons:

After trimming an arris on the stile to fit the mitered returns on the rails, I tried a fit:

With careful measurement and cut out, the frame assembled up without issue:

With the stile looking like it would fit on to the rest of the door, I turned my attention to the door lock. The Museum's door had been fitted with a door lock which is actually intended for use within a hollow core aluminum sliding door, which happens to be the most common sort of sliding door you will come across in Japan. Curiously, I wasn't even able to track down a replacement lock which was intended for solid wood doors. Not even a store specializing in temple building accoutrement and minka restoration hardware could be of help. In the end, I had to source another door lock intended to fit an aluminum door. Seems to be the only game in town.

I was a little apprehensive about fitting such a lock, as the one which had been fitted originally to the door had been put in fairly crudely. They had cut out most of the stile's cross section, leaving just a 15mm strip of material left, and then had glued in patch pieces around the lock. Not the most workmanlike job I have come across, and I worried I might have no choice but to do something similar.

After I received the new lock, which was a bit smaller than the one originally fitted, I made a test stile piece out of teak to see if I could fit the lock without hacking to door stile to bits. That test went well, so I felt I could move ahead with some optimism.

First, the edge of the door is mortised and countered for fitting the lock catch mechanism:

This mortise required some fiddly modifications with little slots and kerfs to allow the catch to be wiggled into place:

Here's the trial fitting of the catch following cut out:

I had to make some tricky internal cuts to allow the machnism to slide on in there - the mortise for the other portion of the lock mechanism provides a convenient window into the internals as they come into place:

Down in place and working smoothly:

Then out comes an appropriate tiny VIX bit to pre-drill the screw holes:

I fitted Robertson (square drive) screws, as I prefer these (and Torx-head screws, though these are less conveniently found at the local hardware store!).

With the catch installed, the remaining portions of the lock could be connected up, the stile already mortised for those pieces:

A final trial fit of the stile to the door ensued, to confirm there was clearance for one of the nuki tenons. Here's the lock assembly installed:

All that is left cut-out wise is to cut a couple of scallops on the stile to act as pulls. I should be able to jig up my groover to do this tomorrow. Then a finish planing and chamfering and some oiling, and I'm planning to put on teak oil, which doesn't build up a sheen as it has little to no varnish mixed in.  Then I can glue the new stile up, and trim the ends of the stile to the line. Should be done tomorrow.

One last pic:

All for today. Should have more to post tomorrow on this topic. Thanks for your visit!

Friday, October 4, 2013

Gateway (II)

Since my last post in this thread, where I mentioned an upcoming project with a well-known and long-established Museum on the east coast, things have been moving along gradually.  I'm patient with the process and re-adjusting expectations when dealing with a larger project and one with many involved parties. So far so good in that regard.

The process I have been involved with over the past year or so has comprised a number of phases. There were initial discussions and meetings, then I wrote up a written 12 page preliminary proposal, followed by a revised and extended 'proposal II'. In 'II', I mentioned that the 25 year lifespan of the current gate - now in a poor state - was really far short of ideal, especially in light of the precious material involved, namely Japanese Hinoki, a relative of Port Orford Cedar and Atlantic White Cedar in the US. The material specified for replacing the Hinoki is in fact to be Port Orford Cedar, however POC is a species with a very limited habitat (a small portion of southern Oregon and a few patches in northern California), and is currently under attack from a root-borne fungus which is killing off the trees. With climate changes in that portion of the country likely to be in the direction of increased aridity, and given the seemingly inexorable march of the root rot problem, the future supply of Port Orford Cedar is, well, uncertain at best.

Given those factors, I argued in my proposal, both part I and II, for changing the gate design, moving away from the simple kabukimon arrangement towards a roofed gate form. Any wooden structure outdoors will last far longer with a roof on it than without. Without a roof, in most species of wood, you are looking at 10~25 years durability, tops. Clearly, the lifespan of Hinoki, left out in the weather, is but 25 years in the Boston climate.

In proposal 'II', I laid out some options for improving the weatherproofness of the gate, incremental steps ranging from foundation changes, to additional copper cladding/flashing and capping, to added-on mini-roofs, to a fully-roofed gate form, of which there are various options. As one moves along through the steps, the appearance of the gate is modified further and further from the original configuration. While these weatherproofing options would certainly increase the up-front build cost of the gate, the reward, as such, is some increase in durability: the fully-roofed option would gain, I would think, at the minimum a 5- to 6-fold extension of gate lifespan. The steps short of a full roof would gain a certain increase in durability, but it was more difficult to anticipate how much bang for the buck that would actually mean. The question as ever: spend money up front to do it right, or, as is more commonly the case, spend far more money later to make it right?

One important point in this equation is that Port Orford Cedar is getting scarce and expensive now - what's the availability and price picture for this species going to look like in 25 years? From what I have seen with other species, I would not be optimistic in that regard. Materials in general are going to be more expensive in 25 years than they are now, and one might expect labor costs, insurance, taxes, etc., to have also risen some amount.

Moreover, the elephant in the room, so to speak: is it ethical to take a 300~400 year old tree, convert it into timber and configure it into a structure which, by its very design, sees that wood rot away in 25 years? I don't need to spend much time thinking on that issue to reach a conclusion, personally. In years past, this might have been something that people once barely batted an eye at, in the time when trees seemed a limitless resource and there was always another hillside to be found and more trees to chop, 'as far as the eye could see'. Those days are over, and how we use natural materials must move forward into a sustainable paradigm or else we simply repeat the fate of the Easter Islanders, chopping down that last tree, comfortable perhaps in the thought that 'it was the way things have always been done'.

That sustainability piece is really important to me, and it was clear to me from discussions that simply from an economic ROI (return on investment) perspective, the Museum was looking for a better outcome for their dollar than a 25-year gate lifespan. They knew as well that they could not simply count on funding to materialize for another gate if it were required 25 years from now - - make hay while the sun shines in other words.

In the end the Museum received a funding grant which allowed for the construction of a roofed gate, as per the proposal recommendations I had made. While the course may have seemed straightforward at that point, it turned out that there were peripheral parties involved with the process who were uncomfortable with any change away from the original garden designer's specifications in regards to the gate. While I understand and greatly respect those motives, for me it comes down to the simple fact that the previously installed gate only survived 25 years, and that short lifespan is due almost entirely to the design of the gate. For me what counts is how can best use a precious and perishable material to create a structure of beauty and durability. The roof is the best solution to that puzzle.

The Museum was trying to steer a smooth path through these competing visions of what should be done with the garden, and so when I met with staff there after funding had come through, they indicated that there were looking to replace the gate much as is, with the addition of some mini-roofs to the top of the cross-beam and post caps, etc.. This is the sort of thing they were contemplating - a picture of one of the roofing options I had supplied in the proposal 'II':

If you look closely, you can see through the gates opening that there are a pair of smaller and lower roofs heading out to the rear of the gate, cross-wise to the main beam. These cover over the rear support posts, hikae-bashira, and associated ties. The form of this gate, having three small roofs, is fairly similar to a kind of gate most often associated with Japanese castles, a kōrai-mon, with the exception of the relationship between the main posts and beam. Notice above that there are portions of the main posts which protrude above the main roof. The projections are shingled in copper completely (which makes good sense to do). With a kōrai-mon, however, the roof sits atop the posts and covers them completely, which makes better sense for other reasons. I can't say the arrangement seen in the above photo is entirely attractive, and all in all it would make more sense, at least to me, to have a gate in which the roof completely covered the posts, instead of inviting future trouble by having the posts pierce the roof fabric. A golden rule of sound roof design is to keep the folds and penetrations of the roof membrane as low in number as possible.

Key point: It's always better to depend upon geometry to shed water as opposed to depending upon materials to keep water out.

The Museum asked me to undertake some preliminary sketches of a gate similar to the above-pictured example so as to have something with which they could move forward in internal discussion. While they were sold on the idea of protecting their investment by adding some roofing, they were also seeking a balanced solution to the various viewpoints on the project, and so they were looking at a middle path. Sometimes this middle path seems like a fair compromise to solve tension between competing visions and expectations, however it is not always the case that a compromise in direction is in fact a wise direction. Sometimes a compromise means everyone 'wins' to some extent and sometime it means everyone 'loses' to some extent. But great designs are rarely produced by trying to balance a weigh scale....

I was very excited from the get-go at the prospect of a project such as this. For me, to have the chance to build a traditional Japanese structure, and one that would be publicly accessible, was an incredible opportunity, and one which I have been hoping would come along. I was slightly deflated the museum was indicating a design direction that fell short of what I thought would be the wisest course, however I returned home intending to draw up a version of the gate with the mini-roofs and additional flashing. I felt I had made, in my proposal, a strong pitch for a roofed gate and they had chosen not to go with that idea, and thought that maybe it was best if I simply accepted the situation as it was and get on with designing what they asked me to do (?)

It has often been the lot of craftspeople to be merely the 'hands', executing the plans and designs of others, and this is something I've brought up many times before on this blog - and railed against. What it boils down to, it seems to me, is to use one's skill and knowledge to benefit the client, as it is not always the case that the client is intimately familiar with the technical details that can make or break a design, regardless of how attractive the design may appear to them. And some design ideas, regardless of aesthetics, are simply bad ideas, no two ways about it. One task of a craftsperson is to endeavor to educate their clients about the things which matter in terms of craftsmanship. It's a case that needs to be made, though in practice it is not always an effective selling proposition. The hierarchy of needs with each client may mean that crafstmanship, sound building practice, wise use of materials, etc., can take a place several steps down their list of priorities. It's not that those things aren't important, it's just that they may be invisible to the client and other things play larger in their minds. For me, those things are on the top of the list of what is important, and some projects must be turned down even if certain minimums cannot be met in a project, though saying 'no' can be hard to do in a lot of cases. Fortunately though, there is often some room for getting the critical details funded in most projects, if the matter is explained succinctly and persuasively. A craftsman-builder does no good service putting up a structure they know from the outset will have defects in its design that reduce the integrity and lifespan of a structure.

As I explored the gate design in more detail, I soon came to the conclusion that the mini-roof idea was simply not a very good direction to take regardless. I also asked myself if I this was simply a case of confirmation bias, and I don't think so - I saw clear reasons why the mini-roof idea fell short (excuse the play on words). While the small roofs offered an improvement over having no roofs, after taking time to consider the matter in detail, the improvement was, in my estimation, a marginal one at best.

For one thing, if you look at how the weather deteriorates a wooden structure, presuming normal functioning of the structure (in other words, no leaks at the flashing, etc.), the portion of the structure which takes the brunt of the damage is the lower 1/3rd of the structure closest to the soil. The damage is most severe just above the ground, where you get the unfortunate tag team of dripping water and mud splash, with the associated wood-devouring fungi which eventually take hold. Kind of like a  Muay Thai fight, where one kick to the thigh too many and the opponent can no longer stand up, the weather effects give a beating on the wood elements most acutely on the lowest parts of the structure. The lowest portions of the structure get wet the most quickly, and dry the most slowly. Any moisture which may have accumulated higher up on the structure, drains downward, keeping the lower parts wet for longer. Eventually rot sets in. And when the lowest parts of the structure become wobbly, the entire edifice becomes vulnerable to collapse, while upper areas of the structure may be in reasonable condition.

Looking at the current gate, there is ample evidence of this weathering process:

Having a mini-roof on the gate is the same thing as having a house with a small eave projection. New England is a great place to see such houses, and when it rains, as one would expect, where the rain drips is where the bottom 1/4~1/3 of the house walls can be readily seen to be soaked, and these are the portions of wall which are invariably in the worst condition. Here's one local example, a fairly modern small barn on a  concrete foundation:

A closer look:

The trend is clear to see on a building about 15 years old.

I realized that, like short eaves, the mini-roofs tacked onto a gate would do very little to stop the process of the lowest portions of the gate taking the brunt of the weather.  In fact, leaving the roofs off altogether would confer but one advantage: the entire gate would rot out more or less evenly. A consolation prize, perhaps.

Another problem relates to the presence of the cross-wise roofs heading out to the rear - here's an example of a kōrai-mon for illustrative purposes:

Now imagine it's raining and you are standing between the gate's doors - would that likely be a nice spot to hang out? I didn't think so either. The only way that arrangement could work for people who did wish to stand in that space would be to put gutters on all the roofs (which some kōrai-mon have). That's a route which then percolates on down, design-wise, to other fundamentals of the site, like drainage. It didn't seem like a good direction to go. One of the advantages to a roofed gate - and in fact a significant reason for larger temple gates having roofs in the first place - is that it provides shelter from the weather, which allows someone who might want to take in the garden, but not stand out in the rain, to linger a while. In the first few centuries after Buddhism and Buddhist temples appeared in Japan, it was not possible for commoners to actually enter these sacred compounds - they could approach to the gate only, and pray towards the inner recesses. Even priests were only able to enter the most sacred spaces but once a year. A roofed gate allows people a comfortable place to stop and contemplate what lies within, and thus the roofed gate takes a place in the cultures of both Japan and China.

On another front, I realized that cost-wise, the added mini-roofs were every bit as complex and time-consuming to build as having a single main roof covering the entire structure. There was essentially no cost-saving to be realized by going to the mini-roofs - in other words, for the same investment one could have a single covering roof, which was a vastly superior solution in terms of protecting the structure over time.

It then became quite clear to me, after some pacing and cogitating, that I had to make a second push for the fully roofed gate. I feel, in retrospect, that this was in interesting junction in my professional career. I believed firmly in the virtues of the fully roofed gate, and yet after proposing that initially, the client had chosen a different direction. I had to make a decision about whether to make another push, to make the case again for the roofed gate, in the hope that they would listen and would see the logic in it, and would see how that represented the wisest investment in terms of money and materials. But to make another push was also to invite the possibility that they might react negatively - some people, after their 'minds are made up that's it', and they may resent any further entreaties to reconsider their decision. I worried that I risked the possibility of damaging my relationship with my contacts at the Museum if I brought the topic up again. And yet, when I reflected on my duty as a designer and carpenter, and considered the sort of work I wanted to leave behind in this world, I could only conclude that the risk was one very much worth taking. I was confident in my beliefs, and was hopeful that the client would be open to my input. I decided to tread carefully, but determinedly.

I took a breath and made the phone call, explained my conclusions about the mini-roof idea, made my argument, and discussions ensued within the Museum. They came back to me a week later and asked me to draw what I was proposing, and that they would "take that into consideration". A crack in the door thus appeared. I then spent about a week drawing the gate, taking into account some of the aesthetic direction that the Museum staff had indicated. They were looking for a gate which was not tending towards 'monumentality', a gate without a lot of ornament and elaboration. Not being intimately familiar with Japanese architecture, they found a lot of Japanese wooded buildings had roofs which appeared (overly) massive to them. They wanted something toned down from that. Hah! - I was now the one trying to steer a middle path....

Considering the issue of a full roof, it was not long before I came to see the wisdom in placing the roof over the structure so that the roof ridge was set back from the main front posts, centered over the entire structure. This pattern of gate is the physician's gate, or yakuimon. Based on some examples I had seen in Japan, I decided to modify the typical yakuimon framing so that the structural core of the gate was in fact virtually identical to the kabukimon pattern, save for the posts sticking out through the roof. I did this as I thought it would mean the gate was changing less from its predecessor than otherwise. I then dropped the roof slope down a notch or two from what is otherwise typical, and scaled the mass of the gate to match the feel of its predecessor. This gate not only fronts the garden, which suggests a certain scale, but also is placed within a wall attached to the massive museum building itself, so the proportioning had to work for both contexts. While most Japanese gates have a more massive and steeply pitched roof, there were certainly classic and pleasing examples to be seen which had shallower roof lines.

Here's my design for the new MFA Tenshin-en gate as it stands currently - front elevation:

Perspective view from the front:

The flanking section on the right side contains a small door, while the one on the left is paneled.

Side elevation, showing the non-centered roof:

I have not rendered the roof's small minoko completely cleanly or accurately in SketchUp due to time constraints and the hassle of doing such double-curvature drawing work in SketchUp. This roof employs the traditional hidden roof/decorative roof system. The roof will be shingled in copper, and the above drawing does not show that shingling in detail, again. The barge boards will be profiled with mayu, also not shown on the drawings. Some details may yet change, design-wise, and I am open to massaging out further refinements and adjustments. The above represents but a week of exploration, and I'm sure further consideration may well result in some tweaks at least.

The design was submitted more than a month ago and has now made its way through several levels of approval at the Museum. They decided that the final arbiter of whether the design was appropriate to the garden would be the son of the original garden designer, a fellow who lives in Japan. In the past few days we have received his blessing (whew!), and I am in fact going to be meeting with him later this month, along with other people involved in the garden rebuild project. I'll be interested to hear his impressions of my design and see if he has any desires for stylistic changes.

The Museum's Conservator has requested that I build a scale model of the gate as part of the project, so I am likely to be starting in on a 1/10 scale model sometime later this month. That will be a lot of fun to construct, and I'll be interested to see how the scale model might affect perceptions of the design.

I had been waiting for some paperwork to complete so that a purchase order comes through, and that happened yesterday so the project is officially underway! I will make my way out to Oregon next week to look at some POC logs. I'm looking forward to that, and of course you'll read and see the developments here on the Carpentry Way. Thanks for coming by. On to post III.