Tuesday, September 28, 2010


This past weekend I visited Hill-Stead, a 152-acre, 10-building museum and a National Historic Landmark in Farmington, Connecticut. It was primarily designed by the architect Theodore Pope Riddle, the very same woman who designed the Avon Old Farms School I wrote about several months back (Part I, II):

The builder was a local named Richard Jones.

Hillstead is the place she designed for her parents to live in, and was constructed in the period spanning 1989~1901, in the Colonial Revival Style:

Some additions were made to the house over the next few years after its completion. The history of Hillstead has been covered well by other writers, so I will not delve into that here. What struck me, funny enough, was the parallel between this place and sukiya teahouse architecture. Not any parallel in visual or constructional terms obviously, but in the fact that the common concept for the two forms is that of wealthy people trying to capture something rustic, yet quite refined, in their built environment. Also, like a Japanese house, at Hillstead it was the floor plan that determined the overall form; the rooms were not stuffed into a preconceived format like most Victorian-era structures. Roofs were drawn deliberately at an assortment of angles, the layout of the place was on the 'big house - little house - back house - barn' archetype of connected farm buildings seen in other parts of New England. The barn had to not only look old fashioned, but be timber framed with exposed pegging and uneven width siding. The apex of the barn roof has holes to simulate a dovecote. Hand forged nails were exposed prominently, etc. The object for this residence was to convey a 'charming sense of rambling informality', and the goal was attained, it is clear, in a high class way. Yes, it's a little artificial in that it's trying to simulate something else, but it's a well designed and well made place, and perfectly preserved.

The Popes collected many paintings and inside you can see works by Monet, Degas, and other French Impressionists of the period. The Popes clearly had good taste, collecting these works before the artists gained their fame. There is an extensive library inside with some 3300 volumes, some of which are quite unusual, though some apparently are in very fragile condition. There is a sweet set of Ming Dynasty Ox-Blood glazed vases and Chinese chessmen, lots of neat clocks, etc., etc. A lot of neat stuff to gawk at. Every bedroom seemed to have a 4-post bed, each one different. I managed to locate a few pictures of these beds elsewhere, which I'll share. This first one is my favorite in form, though the columns are veneered it would appear:

Another one:

And another:

There were lots of nice furniture pieces in the house.

To be honest, I don't really notice the accoutrement other than the furniture - mostly I stare the architecture, the room arrangements, the staircase work, the hardware. All the doors, for instance, are nice and thick at 2.5" - even the louvered interior doors are a full 2" thick. Many of the interior walls appear to be about 12~14" thick, which I find very pleasing. I like thick walls in a house.

Photography was not permitted inside the house, and though there was much I would have loved to record, I contented myself with taking snaps of various exterior features, pictures which I would like to share with you today.

One of the first things I always notice, and have written about in the past (see "A Bracing Situation", II, III, and IV), are braced doors and gates. setting aside the roadway gates, the first board and batten doors I saw at Hillstead were well-designed:

A close up shows that the joinery details are cleanly executed with the let-in braces:

Those doors weren't a fluke in their design either, as these nearby doors attest:

I like little garden structures, though it is a bit rare to come across anything all that nice. At Hillstead, there's a beautifully configured sunken garden near the house, on a stretched octagonal plan and with a summerhouse:

Another view:

Though it's not really a double roof, the skylight framing did remind me of that with the double rafters visible:

Another view:

It's a good-looking little building, ah, but the devil is in the details, now ain't it? The long openings on the sides, for example, feature scant 4x6" headers over some 16" span, which have sagged noticeably over time:

Those headers might be more appropriately sized at 4x 12 or so I imagine.

While sitting in the summerhouse, I noticed some non-native conifers in the garden:

So I went over to take a look, and to my surprise these are trees of Japanese origin:

Sawara(椹) reaches a maximum height of about 30m, and is a close relative of Japanese Hinoki (Chamaecyparis obtusa), along with Alaskan Yellow 'Cedar' (Chamaecyparis nootkatensis), Port Orford 'Cedar' (Chamaecyparis lawsonia), and here on the East Coast of the US, Atlantic White 'Cedar' (Chamaecyparis thyoides). I'm using the term 'cedar' loosely, as none of those trees are true cedars.

Check that! - I just learned that sands have shifted in the plant classifications for 'Alaskan Yellow Cedar' - chamaecyparis nootkatensis: it has now been transferred on the basis of strong genetic and morphological evidence to the separate genus, callitropsis, as calliptropsis nootkatensis, or, possibly, back to cupressus nootkatensis (the name it was originally described under in 1824 apparently). The terminological details are to be hammered out in the upcoming 2011 International Botanical Congress. It's no longer classified by botanists as chamaecyparis. One key difference relates to the amount of time it take the cones to mature, which is 2 years for the Yellow cedar, and 1 year for the members of chamaecyparis.

In Japan, sawara is a premium choice for wood shingles and wooden bathtubs, rice bowls, cookpot lids, among other applications. Here's a sawara bathtub as one example:

The botanical name for Sawara, chamaecyparis pisifera, tells you much. The first word, chamaecyparis, in Greek breaks down into khamai, meaning 'ground', and kuparissos, for 'cypress'. The second word, pisifera, means 'little peas', a reference to the small cones produced by this variety:

I haven't come across too many of these in North America, and it was fun to discover these specimens here in Connecticut. There is another Sawara in Elizabeth Park, in Hartford CT, though is is not in good health.

Well, all for today - I recommend a visit to Hillstead, one of the finest examples of Colonial Revival architecture in the US, to anyone who happens to be in the area. A tour is $10, and the docents are very enthusiastic.

Monday, September 27, 2010

A Way Forward

Having built in and lived in a cob house (see "Mud and Sticks" I and II <-- those underlined are links) for a couple of years, I am sold on the benefits of clay-based thick walls of high thermal mass. I've been thinking for a long while that the next house I build will be made with double walls and the German light clay straw, but hemp/lime, originally developed in France as IsoChanvre™ might be an equally good choice because of the benefits which accrue to the lime, along with use of a waste product (the hemp hurd). Here's a general overview, though a little dry as videos go:

Another project in Britain:

Part 2:

Here's the first permitted house built in the US, for the mayor of Ashville, NC no less, at about $133/sq.ft.:

A follow up piece on that project provides some more information.

The company which did the above project in the US has done a bunch of the groundwork to give the hemp/lime product the materials testing data and other 'hard data' attributes which help make it more easily salable to building departments (at least in forward-thinking jurisdictions). The lime and hemp materials come premixed and bagged, which is very convenient, and the central distributor for these materials is in Chicago. Good for them! I hope it catches on.

*Sorry about the video formatting being slightly too wide - I tried making various adjustments, but nothing I did seemed to solve the problem. If it is annoying to have the clipped pictures, simply click on through to youTube to see them.

Sunday, September 26, 2010

This One Rings a Bell (15)

The design phase is now principally complete. The client has been sent the drawings for review, and I now present them here.

The minoko was a royal pain to draw, and ended up having to be put together square cm by cm like a giant quilt of triangles. It took me hours!! It was the only way to produce the complex curvilinear shape in SketchUp. At this point, it doesn't look exactly as the one I intend to build looks, but it's close enough for the time being. I'll be doing a blog soon on minoko in more detail.

Without further ado then...

Top view:

Bottom view:

Side elevation:

Front elevation:

3/4 view:

Cutaway, with cantilevers removed:

Cutaway, with rafters removed, showing decorative ridgecap internals, nuki, etc.::

Cutaway, rafters and cantilevers removed, to show upper roof framing and bracing:

There will be a few scabbed-on braces to be added to the upper framing which have been omitted from the drawing.

Another view of same, purlins now removed:

As it might look looking up from the ground:

A few details are omitted at this stage, such as the gable end grill and pendant, and the carving details which may or may not be added to the various lower structural members are not established either.

Now I can take a breather and the client can give me feedback, go and visit his local building department with the plans, and so forth. I have some preparations to do for the Fine furnishings Show in Providence, RI coming up in October, among other things. I also plan to finish off my 'Fan of the Fan' series here along with a few other topics which may be of interest to readers. Stay tuned, and thanks for coming by today. --> on to post 16

Wednesday, September 22, 2010

This One Rings a Bell (14)

I wonder how many readers out there, in the segment of the woodworking population who tend to not spend much time in the design phase of a project, who may well be getting restless with this thread? After all, I haven't cut a stick of wood yet. I'm starting to get to that maxed-out point with the drawing myself, after another 20 hours or so of work since the last post. It's getting really close now though, with 95% of the structural parts now in place. A lot of the drawing time at this point is more perspiration than inspiration, I'll say that. SketchUp doesn't always deliver on some of its functions so well when the surfaces which intersect are of complex curvature. I'm doing the best I can.

I thought that this time I would show the structural details in a manner similar to that of the parts being added to the roof. Whaddya say?

Starting with the blank slate, so to speak, of the lower exposed eave, the first move is a set of interior beams, along with a ring of hanegi osae (which serve as fulcrums for the various cantilevers):

Since the previous post, those interior beams have been re-sized and reconfigured a fair bit. They bear more or less directly down upon the principal structural timbers carrying the bell and will be through-bolted to those support timbers.

On now with the ha-ne-gi (桔木), the cantilevers, and a few stub posts sprinkled in here and there:

Configuring the cantilevers is challenging in a small building with a square plan, as the cantilevers all are essentially needing to meet more or less in the middle of the roof and be as long as possible. The solution in this case meant that some of the hanegi had to arrive in the middle at different heights.

Here's a closer view showing how the hanegi come together:

The result of the positioning and re-postioning required to get them into a good arrangement is that four of the pieces are of curved form and will need to come from bowed logs. I've got a source for some excellent material in that regard.

Next layer is the purlins:

With the field hips being irregular, some creative solutions were required to support those hips properly in the roof.

Next, the field rafters:

Here's a look at the asymmetrical positioning of a cluster of purlins around one of those hips- some of the purlins serve partially as cantilevers themselves:

And a view from the other side of the hip:

The purlins I have shaped so as to curve up with the field rafters, and then for the last 50~100 cm they straighten out to run under the hip and lap over one another. That simplifies the connections between the hip and the purlins. Additionally, I strengthened the area generally with the addition of corbels on one side at purlin level 2, and the purlin entering from the opposite side of the hip (which is the 3rd one up on the high pitched side) is at that location sandwiched in between the corbel and the purlin it carries.

Another view of the same corner:

Those stub posts in the second picture of this post support the decorative purlins which carry the verge board assembly:

In front of the central stub post is roof for the assembly which will infill the gable and conceal the posts. As this gable infill is primarily a decorative element and the client has not had an opportunity to view some of the options in that regard, I have left off designing it for the time being.

Another view of the gable:

Last one for today - the planking is in place and the roof form is shown:

All that remains is the minoko, along with the decorative ridge, and a few more interior posts and beams. Those posts and beams will serve to support those purlins, in case you might have been thinking that I was intending to run them across the roof without support. Uh-uh. it will hopefully all make sense when all the parts are in place, so please stay tuned.

All the structural components should be complete and presented in the following post in this thread. Thanks for coming by today. Oh yes, we're nearly there...

--> on to post 15

Monday, September 20, 2010

This One Rings a Bell (13)

More progress has been realized in the past few days. Getting pretty close to completion now with the drawing work, and I thought it was high time to share a few pictures of the bell tower as it stands.

The first view is of the structure with most of the layers in place:

The exposed planks at each end of the gable are the interior of the fukiji, and are where the minoko will be placed, and I haven't drawn them yet. I will probably draw them in a similar piece-by-piece manner to which they are constructed.

Pulling off a few layers now, it can be seen that a few parts have been added to the interior;

First off, of course, the irregular hip rafters are completed, down to all their jack rafter mortises. Here's a look at the terminal end of the hip as it meets the fukiji miter:

Looking in the gable end, the cantilevers - ha-neh-gi - are now in place:

Also visible in the above drawing are the stub posts atop the ridge to carry the decorative ridge. This building has three different ridges!

A bit more pulled off the model now gives a clearer look at those hanegi:

The hanegi are all compound splayed, parellelogram in cross-section, and the outer pair are slightly curved as well. I may use logs in the round if I can locate something appropriate.

Another view of the hanegi and one of the irregular hips, with jacks attached:

A bird's eye view:

There's getting to be a fairly tidy pile up of chunky timbers in the middle of the roof, and there are more to come. First I need to draw out the cantilevers for the other two sides of the roof, and then I can develop the various purlins, which rest upon the cantilevers in the lower end of the roof. The purlins carry the common and hip rafters of course.

All the hanegi are mortise and tenoned into the kaya-oi:

A slight space is left between the end of the hanegi and the inside face of the kaya-oi to accommodate settling.

Last one today is a shot from directly above:

Along the uphill edge of the eave ceiling boards are the hanegi-osae-ita, which are the fulcrum for the cantilevers. They bear directly down, through the hiro-komai and fan rafters, right onto the wall plate. They will be notched to receive the hanegi.

All for today. I envision just one or two more posts on this, the principal design phase, and it will be all done. Certain details I will deal with later on mind you, such as the gable wall treatment, the drop ceiling, and the gegyo (the carved pendant hung at the apex of the verge boards). Also, the carving details, if any, for the beam ends in various places has yet to be specified.

It's been a long slog with many challenges to get to this point in the process, and I am feeling a curious mix of exhaustion and elation. Tomorrow I will tackle the remaining cantilevers and the so-called 'drooping verge', or minoko. It should be fun.

Thanks for your visit today. Comments always welcome. --> on to post 14

Saturday, September 18, 2010

This One Rings a Bell (12)

Post 12 in the series, so far covering the design of a Japanese bell tower or shōrō (鐘楼). Construction phase to follow, once the design is fully realized. I am at last getting closer now in that regard.

I thought I'd post up a few pictures of where things stand now that I've worked out the gable end treatment. I hope you'll agree that it is really starting to look like a building now:

I'm really pleased with the improvements realized by re-sizing and re-profiling the fukiji.

The gable eave edge build-up consists of 3 layers:

The lower layer, forming the wide board, is the hafū - what in English would be termed a verge or barge board. Above that is a piece which shows it's edge to the front, and climbs on top of the edge of the hafū - this is termed a nobori-uragō. On top of that, just like with the main roof eave edge, is a fukiji, which will be shingled on its face and exposed underside with copper. The slightly down-tilted board on the top of the assembly is termed the koma-bitai, or 'horse head'. In a copper shingled roof, the slope of the koma bitai is typically fairly slight, while on a tiled roof it is more pronounced.

A bird's eye view, showing how the slight lean given the verge board assembly gives it a slight prow forward from this perspective:

The slight lean of the verge board assembly forward is done to counter the effect of visual foreshortening, when viewing the structure from ground level. If the hafū assembly were dead vertical, to the eye it looks slightly as if it is tilting back into the roof.

Next is a view one might have from the ground:

Last one for today:

At this point virtually all of the internal structural components are not illustrated. A decision about whether to go with a copper or tiled decorative ridge needs to be settled before those structural elements can be configured, as the weight difference between the copper and the tile treatments is considerable. At this point the upper ridgepole is provisional, being sized for the case of using copper for the decorative ridge.

Thanks for visiting by the Carpentry Way on your travels today. --> on to post 13