Tuesday, June 30, 2009

French Connection 4

This is my second post today, so if you are just checking in you may wish to take a look at the archive for my previous post.

It was about 6 months ago that I first started intensively tackling some projects in a 19th century French carpentry drawing book I have by Louis Mazerolle. One of those projects, the tréteau, was first shown in this blog back in late January, in a post entitled The French Layout Challenge. This post, though of a slightly different title is a continuation of that series beginning in January. Subsequent posts have been French Connection, Part Deux, and French Connection 3. New readers, or those, like me who find their memory occasionally betrays them, may wish to read those posts first before proceeding.

When I last left off in this series, I was somewhat impatiently waiting to hear back from a campagnon who had said he would help me with the areas of the drawing that were giving me trouble. Well, eventually I'd gotten to a point where I had pretty much given up on that fellow, when just last week I got a message from him apologizing for the delay, and that he hadn't been able to make much of the drawings, text and photos I had sent him and let me know that he was going to France for the summer and would be picking up a copy of the Mazerolle book and would be better able to help me from that point.

That was great news. Then it occurred to me that in the meantime, now that I had gained some facility with SketchUp, I might be able to render the parts of the drawing that were puzzling me into 3D, which would be much clearer a means of showing the campagnon exactly what the issues were. So, over the last few days I have been devoting a fair amount of time to re-drawing (this must be the 6th time overall now) the tréteau. My intention was to try and use SketchUp as a 2D platform, where I could use the development techniques to locate the positions and shapes of the parts, then replace the central plan view with a 3D view.

Here's a look at how the overall drawing looks, not quite complete but mostly there, with all the traces still in view:

While it might look to some as some kind of spaghetti junction gone mad, it's actually pretty clear once you know which line is going where and more importantly, why.

Here's a closer view of the tréteau, not quite complete, showing how I have developed the 3D from the 2D plan view:

Well I've gotta say, this SketchUp has really helped me out a lot! The reason being, is that once I have completed a 2D development, I can construct it directly (that is, virtually) by raising the 3D view from the plan. If the points from the 2D plan and its developments connect to form planes (which is what you want) in 3D, SketchUp shows that by automatically filling in the planes. It's not that I necessarily want the planes filled in, as in many cases I don't - it's the fact that it confirms that my 2D development is correct. And once I know that a given part is correct, that is one less thing to be uncertain about and the process can roll onwards.

It goes further though, the 3D, in helping me out. There were concerns in my mind that I wasn't seeing the 2D correctly in some views, and that the legs were surely going to interfere with one another in a couple of spots under the top beam of the horse. Developing the 3D has shown me that I was correct in these suspicions.

I have also been able to clearly establish that the original drawing of the sawhorse in the book is in fact incorrect in several places. Whether these errors are as a result of original error or miscopying, or are deliberately there to keep the student on their toes, I can't say. The main thing though is that I'm not hung up on these little issues any longer - the book, wonderful as it is, has a few minor glitches is all.

Another aspect of the original drawing is that a fair number of lines are in fact omitted, and I think this makes sense given how many lines are eventually required. I think that L. Mazerolle left out lines where it was presumed that the development solution was 'obvious' (perhaps to him and his students, but not always to me!), or where, as I have discovered, he is using the same method elsewhere on the drawing and only chooses to illustrate it once. Further, there are parts of the drawing where even after looking very closely with a magnifying glass, I have been unable to make out exactly what is happening at certain places. At other ares of the drawing, I have been unsure whether certain locations where a lines meet are in fact intersections or are simply lines that end up close to one another. There are a lot of places where the drawing is quite bewildering, and given my poor command of 19th century French, apparently very tough even for French speakers today, the text is not of much help even though I have translated all of it that pertains.

In the classroom, were Mazerolle to be teaching the layout of something like this, he would more than likely have an example already made, and having the actual object there while working on drawing it, is very helpful indeed. Now with the capacity to develop 3D views closer at hand, in effect I do have that model to refer to. In some cases I can draw the 2D, develop from that a 3D assembly, and then actually remove the part in question from the assembly, rotate it around and down to be flat on the floor, so to speak, and then I can superimpose it over the 2D and check things out. That technique has helped me resolve several areas of confusion and bewilderment on the drawing, and I'm happy to say that I have managed to now complete the drawing of the tréteau! I won't actually be needing the campagnons help after all it looks like (though we'll see about that - I haven't made it yet!).

Here's the 99% complete tréteau drawing:

The above drawing, as well as the one to follow, are in the large size format, so if you click on them you will get a better view.

As a final bonus from this application of SketchUp, I have been able to readily re-scale the tréteau to be more useful to my needs. The example in the book scales out to 90cm high (@35.4" tall). I prefer sawhorses that are lower myself, and my other current sawhorse, the irregular-splayed Japanese style one I detailed in the series of earlier posts entitled "Irregular Situation", is 24" tall. I was able to scale the 90 cm tall tréteau down to 60.94cm, and now I have been able to re-measure the parts, and thus it will be easier to re-draw the horse (yup, I'll need to draw it at least one more time) knowing how big to make all the bits so they work in a more compressed space. That means, for example, that the legs of the tréteau are going to shrink from 8cm (@3.1") to 5.8cm (@2.3") in section. Here's a look at the original, along with the scaled-down tréteau:

I'm feeling quite elated to have cracked the code of this drawing at long last. I must have hundred of hours in now on the drawing work, and it's good to not have to pull any hair for the next while at least. This realization of an accurate model also saves me at least one round of trial and error in the construction process, as previously anticipated, since I am now very confident that the layout will be correct and I can cut pieces hopefully without wasting any wood. The assembly of this horse is likely to be pretty tricky, so at this point the unresolved issues concern the tenons on various pieces, and how their shape will affect the assembly process. In many cases the tenons appear to diminish in size by about half from root to tip. The above drawings do not show all the through tenons yet, nor have I performed the diminishing-down to some of the exposed tenons.

I posted twice today as I am going to be away from my desk for about a week. When I get back to the keyboard, I can return to the lantern project, and of course the next installment in the kō-ko-gen-hō series. I hope my readers in the US have a great 4th of July, and those in Canada a happy Canada Day on July 1st (and my apologies to readers from other countries as to my ignorance of your holidays).

I will have net access, so any comments posted will be receive a response.

à bientôt

Comment Reply

I received an interesting comment to yesterday's post, and discovered that my reply was too long to be posted, so I have decided to create a new post. Here is the comment by Derek Cox:

G'Day Chris,
I am thoroughly enjoying this lamp build, probably the most of all your posts so far...

I noted earlier your use of the half laps with mitered abutments and I had thought that they must be undesirable in one aspect in that they reduce the cross section of the joining members and therefore decrease overall strength. They look nice though and I was pondering the main reason for their use. Are they mainly used here so that finish planing will not create gaps in the joint like it would with normal half laps; is there some mechanical advantage such as resisting twisting in the plane of the joining pieces or is it largely aesthetic?


Derek Cox

Here's my reply:

Hi Derek,

glad to read your greeting from the land of Oz and that you are enjoying the thread so far. Also, great question! I've been wondering if someone was going to be curious about that.

You are quite right that the mitered half laps - any housed half lap for that matter - reduce the cross section of the pieces at the joint and thus the strength. I strive to compensate for this at the outset, if I know that I will be using these joints, by designing the pieces for finished size in view of their reduced cross section and resultant lowered strength after cut-out. I would observe that the lap itself is really a bigger culprit than the housing in terms of reducing the joint strength.

I am doing all I can to retain as much strength at these joints as possible, and thus my earlier comment a few posts ago that the initial half lap with the hijiki was not ideal with a 0.25" housing reduction on each face, and my recent decision to add a cap to the top of the post which also served to stiffen up the connection for the hijiki. The lantern is a prototype after all, so I am learning, revising, and fine-tuning as I go. The more important piece, for structural strength to carry/resist the loads is the second tier level of beams, and I have made the lap in the beams a fair bit stronger, using only a 0.125" housing, and that's despite the fact that it is a slightly smaller section than the pieces used for hijiki.

In any case, the purpose of housing the joint, let it be said, is twofold: depending upon the case it allows more of one timber's grain to be carried (not just the portion in the half-lap), and (a little more to the point here) is that it allows for shrinkage and swelling, through seasonal moisture cycling, to be concealed in the joint. That said, mahogany is renowned for having very slight in-service movement, so that would not appear to be as much of a concern in this case.

So, why use these kind of joints then? While they do, I agree, look nice, in this case virtually none of these joints will be visible once the lantern is assembled. Another aesthetic aspect which favors the use of mitered abutments, and is definitely a factor in my decision to use them, is that they allow for clean flowing chamfers of the arrises of the involved pieces. I will be making a slight chamfer on the pieces, about 1/16' or so, and the miters allow that to happen harmoniously where they intersect. This is not the main reason I choose the mitered abutment however.

The finish planing aspect IS one of the reasons, yes, though still NOT the main one for me. I have left myself just a pass to plane on each face, so the potential for gaps after finish planing is but slight (hah -I'm hoping!).

The MAIN reason I like to use the mitered abutments in a half lap relates to assembly, as I can fit a tighter joint with the mitered abutments. How so? Well, with a non-mitered abutment, you will have un-yielding end grain on one piece pressing/rubbing against the comparatively soft face grain of the other piece. If you are trying for a tight fit, it is easy for the face grain to get torn or marred by the end grain of the other as the two surfaces slide past one another, especially when several trial fittings may be involved. While the vulnerable surface can be 'killed' by pre-compressing with a hammer, and re-dampened later to swell the grain back, sometimes the results aren't quite as satisfactory. With the mitered abutments, a similar surface density of is found on each abutment (about 45˚ to end grain), and the joint can be fitted relatively tightly without concern for damage to the abutting surfaces.

I've had experiences, especially with softwoods, with the face grain portion getting a bit torn up by the end grain in housed half laps and that certainly makes a mess of a joint in a hurry. This is a lesson I bring forward from timber framing. The twisted half lap (neji-gumi) with mitered abutments is thus my preferred connection for the wall plate underneath the hip rafter.

A final reason one might choose the mitered abutments is in relation to timber framing and the behavior of such a joint under horizontal-plane shear loading. The miters, I believe, would provide a decrease in a potential stress riser (as opposed to a 90˚ abutment) and I think the joint would better resist those sorts of loads, and that extends to temporary loads as might occur during raising, where a timber might need to be wiggled from side to side - the mitered abutment is less likely to show up any damage (wood denting from the end grain pressing into the face grain when the timber presses one side to the other - or worse damage).

As an aside, I would add that in timber framing connections at the corners of the wall plate, there aren't a lot of choices, fundamentally, in how to join the pieces. I would argue that the half lap (or it's more refined version of twisted half lap) leaves more meat from each piece, and gives better support to each piece in the connection than the other two common choices, which are M&T (no projection of nose beyond the joint, at least not on both pieces), or sliding dovetails with a detached nose on one so as to give each a projecting nose (the detached nose with dovetail is so weak that it ultimately is little but decoration). And of course, these joinery choices become more critical when a hip rafter is introduced on top of the joint and must take some material out of it to be interfaced.

I hope that explanation made good sense to you. To reiterate: the main reason I use the mitered abutments in a housed half lap is to be able to fit the joints up tight, through several trial fits, without fear of damaging the surfaces. Secondary, and still important reasons why one might choose this type of joint relate to the fact that chamfering or other treatments of the arrises of the intersecting pieces can be seamlessly accommodated, and that finish planing will not spoil the fit.



Note to readers: I will have a second post later today, and then I will be away from my desk and unable to post for about a week.

Monday, June 29, 2009

First Light XI

The previous 10 parts of this series may be found archived to the right side of this page - those readers who are new to this blog may wish to back track in this thread a little before proceeding.

After completing work on the post cap, which also serves as a support for the upper tier of beams, I commenced the work on the dodai, or sill pieces for the lantern housing. These are to be partially housed in the upper tier of beams, which is a more stable and interlocked arrangement in general. The dodai themselves are joined with the same half-laps with mitered abutments I have been using elsewhere in this project. Here are the 4 sill pieces after most of the joinery cut out has been concluded:

You might note the presence of some little diagonal mortises in the piece on the left of the picture - these are involved with the lantern housing posts.

I checked the fit of the dodai pieces as I fittend them to one another:

Here's a close up of one of the corners of the sill:

Next it was time to begin fitting the sill to the support beams, which required they be fitted both one by one, and checked simultanueously for fit. I've pulled one piece out to fit:

Here I'm checking them simultaneously:

Also note in the above picture (and in some of the ones to follow) that the mortises for the lantern housing posts, which I elected to orient diagonally, have been processed at the dodai joint intersections.

After much fiddling, the first piece of sill was fitted. Here it is about 1/32" short of fully down in the housings:

Then two sill pieces were in place and it was time to lap the third piece, where it fits both on top of the first two and in among the supporting beams:

This shot gives a good view of the number of simultaneous engagements that the dodai must engage with, both with other sills and in the tier of support beams:

Finally it's down to the last piece of sill, here being slid down into position, then scribed to the housings, then removed, adjusted, and refit (and repeat as necessary!):

The dodai ring is now in place and ready for the next layer (well, almost!):

A view from the top:

And a close up of one of the intersections:

The nose atop of the diagonal support beam has been planed to a slope so as to shed and water that lands on it. Fitting the housing of the dodai to the sloped abutment takes a little extra time. The end grain of the sill pieces has yet to be finish planed.

The next installment in this thread will deal with the fitting of the lantern housing posts, which are splayed in two directions and thus involve the same sort of cross-section modification and compound geometry as we see in Japanese splayed sawhorses. See you next time at post 12.

Saturday, June 27, 2009


I came across a book at the local library, a recent publication that looked like it might be of interest to read: "Green From the Ground Up: Sustainable, Healthy, and Energy-Efficient Home Construction" (2008). The Taunton Press is the publisher, the same place the magazines Fine Homebuilding and Fine Woodworking emanate from. The authors are Scott Gibson and David Johnston.

I was curious to see what the 'latest' in green building might be, and how a mainstream publisher would address such a topic. Well, the biggest hint as to the orientation of the book comes on the cataloging data page, immediately following the inner title page. Here one finds the following paragraph at the bottom:

The following manufacturers/names appearing in Green from the Ground Up are trademarks:

air-krete®, Andersen®, Cedar Breather®, Cor-A-Vent®, Corian®, Cracker Jack®, Duette®, Dumpster®, Durisol Wall Forms®, EnergyStar®, Enerjoy®, FSC®, Gortex®, Grace Ice and Water Shield®, Gravely®, GreenSure®, Heat Mirror®, Home Slicker®, Hunter Douglas®, Hylar®, IceStone®, InterfaceFlor®, Kynar®, Lyptus®, Medite®, Metlund®, Micronal®, PCM Smartboard®, Milgard®, Mylar®, NuCrete®, PaperStone®, Polyureseal BP®, Rainscreen®, Richlite®, Silestone®, Skyblend®, SolaHart®, Superglass®, Thermafiber®, Trex®, Typar®, Tyvek®, TyvekDrainwrap®, Ultra Touch®, Wal-Mart®, Zodiaq®

Nothing against any of those manufacturers and their products, or the propriety of their trademarks (I didn't even realize that Dumpster® and Rainscreen® were in fact trademarks, but then so is Heroin® a trademark, and where would, um, civilization be without that?). The point is that one can tell at a glance that the thrust of the book is to provide 'green solutions' that are industry-driven solutions. This is confirmed rather early in the text, page 9, when a sidebar makes mention of
Building with Straw: here we learn that while it is a "low tech construction method that can be managed without a lot of building background", on the downside, "the availability of straw probably has something to do with where you live" (really?!), that it must be kept from getting wet during construction, and building code officials unfamiliar with the method might not like it. Well, big news there folks. Though numerous books have been written on the subject of straw bale, from folksy-hippy in content to highly technical, all it merits in this book is a couple of paragraphs, and with a certain tinge of disapproval or that is it a risky proposition. You're insurance company might not like it (enter horror movie theme music). And they are right in what they state, technically- straw is not local to many places (but then, neither is Tyvek®, etcetera), and water getting at the straw either during shipment, construction, or after the building is complete are all distinct negatives. I suppose the same could be said for finish-grade wooden building materials too. Anyway, it's apparent that this book isn't going to be devoting much time to such methods as straw-bale, which as they state at the outset of the sidebar, will "probably never be mainstream".

On the immediately following page we come across another sidebar, this time on
Rammed Earth and Adobe. Here we learn that while "both techniques are appealing for their use of natural materials", unfortunately a "variety of factors is likely to keep them confined to a limited geographic region". We find that building a rammed earth structure is "not a beginner's game", taking "specialized equipment as well as know-how", and that the process "isn't inexpensive". I wonder how the millions of Chinese rammed earth structures, built without any specialized equipment, and looking no more complex a process than that of erecting forms with a few planks and tamping the soil in place by hand, serves as a counterpoint to that perspective? As for adobe, while "there's probably no reason that adobe homes couldn't be built outside their traditional geographic stronghold", on a practical level, they say, "you'll also need hot, sunny weather and the right kind of soil to make the bricks". Well, that sounds like a waste of time doesn't it? I guess it wouldn't be possible to simply dry bricks under a roof.... Funny though how with cob, which is after all monolithic adobe (same mix of materials, clay, sand, straw), there is a fairly wide range of mix possibilities that will work - it's quite a forgiving material in fact, especially in comparison to say, mixing concrete, and I would be puzzled if adobe bricks were much different a proposition....

After those brief two sidebars, no further mention is made of any sort of alternative building methods, until a later short section on living roofs (where again they are marginalized due to 'high cost' an the fact that the plantings must be appropriate to local climatic conditions (well,
duh!)), and a small sidebar on clay plasters (again, 'difficult' to do). Building techniques using timber crucks, hemp hurd walls, earthbag foundations, cob, light-clay woodchip infill systems and so forth, receive no consideration at all. The owner builder process is not mentioned. Too bad.

Well, no time to ponder such matters in a book like this - there's a bright future of exciting new products to be consumed to build your very own 'green' building! You see, 'green' is nothing more than a
different set of consumer choices, it's not, heaven forbid, about avoiding products of industrial manufacture, say seeking to find ways to build using materials of low embedded energy and high resiliency and durability - the type of buildings that learn as Stewart Brand talks about. No! we can't have that. That's not glamorous enough I guess. And is small beautiful, as E.F. Schumacher so eloquently put it? Not in this book - no mention is made whatsoever about building smaller homes, more conserving of materials and energy inherently. Again, too bad.

Well, to the books credit they do spend some time talking about the benefits of passive solar orientation, and how to site a building optimally for that benefit. They do discuss briefly the issue of on-site construction waste, which, by golly, is one of the arguments that the manufactured home industry puts forward too. They talk about saving and conserving water, and all that is very good, but I can't get away from the fact that the solutions they put forward in nearly every case, are
technological solutions of the sort made possible by big industry. As usual they advocate for the 'tight' house, the 'living in a plastic bag' approach to construction, and once again, the solution to any possibility of stale air in such a system is to be provided by the HVAC people, and a whole-house mechanical ventilator, preferably with computerized controls.

They devote a paragraph in the book to recycling, however then make little mention of the post-use recycle-ability of many of the materials they advocate, like finger-jointed studs, OSB sheet goods, glued-on closed-cell insulation, gypsum board, vinyl siding, insulated form concrete foundations, engineered lumber products, and so forth. The book strongly advocates for SIPs (
Structural Insulated Panels), which they claim to be "a green building product on every level". Huh? Well, how about the embedded energy level at the manufacturing, shipping and crane-erecting end? How about the recyclability of those SIPs when the lifespan of the osb and foam board sandwich comes to an end? How about the inherent flexibility of a product like that, where inevitable building modifications, either to the structure or the building systems will be mandated- how easily modified are SIPs? Not so much methinks. Curiously too, while they mildly deride straw bale due to the issue of water getting at the bales during construction, with SIPs the same problem is noted, that the panels and water "don't make good bedfellows", but never fear, the solution is at hand: "careful flashing is key". The next page reiterates this: "The oriented strand board faces of structural panels would be damaged by water. A continuous waterproof membrane and careful flashing details are essential". In other words, there's a problem, but it's all solvable so long as you pay attention to a few details. Again, comparing to their comments on straw bales, where one of the negatives is that "straw isn't local to every area", how local are SIP panel factories to 'every area'? Does the 'straw industry' take out color adverts in Taunton Publications? I'll let the reader guess.

We learn in the 'Framing' section of the text that the timber frame, while "beautiful in it's own right", has a drawback: due to the quantity of wood used, is "inherently a lesser choice than a house built with advanced framing techniques". By
advanced they mean using the absolute minimum of wood, with a high prevalence of osb, finger jointed studs, engineered lumber, lvls, and so forth. Again the question of material consumption and use seems to be from the ever so narrow lens of the short term - the fact that larger timbers can be, and are, readily recycled into useful material many years down the line, while all that other modern junk simply ends up in the burn pile or landfill when its life comes to an end (or building fashions change, or the new owner scrapes it off the lot to build something bigger) is not acknowledged. Neither is the significance of the emotional-connective realm of people living within a timbered structure, a place where one can see and relate to its natural elements, as opposed to the sheet-rocked, fundamentally alienating box most of us inhabit and work within. If that is the sort of advanced that is proffered as the solution, I want no part of it - low tech, for me, is the way to go. Low-embedded energy, locally-procured, owner-builder driven, based in traditional craftsmanship and the hand the artisan, with high recycle-ability, materials that are safe to work with, easy to make changes to if the need arises (resilient)-- those are virtues to be sought in truly building green, in my view.

I guess it's not much a surprise that such a book would come from a magazine, especially from Taunton. Most magazines exist for the purpose of selling advertising, end of story. They may start out with ideals, but eventually the needs of ad revenue dictate that their sponsors have some say in the content and slant of the publication. I have read virtually every issue of both Fine Woodworking and Fine Homebuilding, and it is most noticeable how much better they were, each of them, through the first 50 issues of publication or so. More varied, eclectic, open to alternatives, more committed to craftsmanship, even if it was quirky or non-standard. Nowadays, the magazine's focus seems to be primarily in new tool and product reviews and any form of construction that is pushing the latest industrial product. The 'solutions' are ever more uniform and standardized. Articles on such things as the house that the artist Wharton Escherick built (June/July 1984), Charles Greens's James House (January 1985), or the Bow-Cot Cottages in New York of the architect Ernest Flagg (November 1981) are now but quaint memories in Fine Homebuilding magazine. How can
Fine homebuilding, for instance, be reconciled with vinyl siding and vinyl handrails, as have been featured cover stories in the past year? Ya got me. So, for me, these magazines no longer merit purchase, except on a very rare occasion when there a particularly good article appears. And that happens, in my view, about once in two years or so, and that itself is a dwindling rate. If you come across old back issues in a used book shop, they're worth a look.

Anyhow, the bigger issue is the version of 'sustainable building' that this proffered by this book, and others like it, and I don't see that message, of industrialized 'green' building, as part of the solution, just more of the same, only with a new color: green. The t-shirts are on sale as we speak - look for the ® mark and you're all set.

Friday, June 26, 2009

First Light X

This is part 10 of a series - please look to the archive at the right sidebar to link to previous installments if you are new to this page or haven't visited in a while.

Continuing on with the work on the post cap. This is where things stood after it was fitted to the top:

The next step was to deck the top surface which had a scallop in it from the planer snipe, as I mentioned yesterday, and was far from level with the surrounding wood. I made up a couple of pieces of mdf (Medium Density Fiberboard) a material that I like for making jigs and nothing else:

With the top surface of the cap clean and flat, I then laid out the mortises for the tie rods:

As usual, as I don't don't have a hollow chisel mortiser handy, I resort to drilling and chopping to produce mortises:

Mortises complete:

Then I could check the tie rod mortises for fit:

I'll explain more about the tie rods in a future post when I have the post apart again and can show how it works.

The next step was to double-check the height relation between the top of the cap and the surrounding makitō:

Of course, I checked both diagonals too, and things were looking fine.

Then I plowed out the trench for the lowest piece in the upper tier of support beams. You might recall that I showed a picture in a previous post where the bottom of this beam cleared the top by 0.125". Well, now with the cap fitted, I realized that to plow the cap out so as to leave only 0.125" of material above the post surface would make for a weak cap. I solved this by cogging the joint and reducing the drop down of the lower beam by another 0.125", that way I believe I found a happy medium between weakening the cap too much and weakening the beam. Here's the trench:

Next step was to cut the cog seats:

As you can see, the stepped cog allows the middle of the cap to retain 0.25" of wood, giving plenty of strenght in such a hard and dense wood as Goncalo Alves, while the exit points reveal but a 0.125" step from the top of the post surface.

Following that I prepared the lower beam cog joint:

And here they are, getting to know each other for the first (uh, okay, second) time:

Was the cap done now? Not yet! Holy crow, a lot of work goes into this little cap, a piece largely hidden from view.

The hole for the electrical cable needs to pass through the cap of course. I wanted to shift the wire chase over so that it would come through the upper tier of beams away from the center lap joint, which I wanted to keep as intact as possible for strength reasons. I located a spot to the side of the upper beam, drilled a hole which came through and marked a point on the cap. From that point, I then drilled through the cap, freehand, towards the middle and taking care not to put the hole into the trench:

Then it was time to remove the cap again, using a Bessey clamp, and make a trench in the top to allow the wiring to bend to the side:

However, I was still not there yet. I needed to find a way to make the cap removable since my plan was to bevel the sides inwards. I solved that by drilling and tapping for 8 x 1.25 mm bolts. These bolts are turned into the cap, and when they bottom out on the wood below, will drive the cap up and out. Here's a shot showing the test to see if they work:

Of course, these bolts are not part of the lantern, they are simply used when I need to get the cap out during dis-assembly. When you are building something to last for the long term, you have to consider re-buildability and serviceability. I'll mark the cap so that someone in the future who is (hopefully) trying to repair the lantern will know what size bolt to use and (hopefully) will understand the intention of the holes. That's a fair bit of wishful thinking, but all I know that I sure am grateful, in any piece of equipment I work on, if it becomes evident during repair work that the maker spent a little time considering what it would be like to service their product years down the line and made the piece capable of repair. I so dislike the modern manufacturing rationale of pre-planned obsolescence.

Finally, it was time to bevel the cap. Lacking a tablesaw at the moment, I sawed the bevel waste off...

... then planed to the line, here with my 54mm Mosaku:

Lastly I pared the end grain with a paring guide and a chisel:

The finshed cap, installed in the top of the beam:

Too bad in a way that 95% of it will be covered up- the Gancalo Alves is a pleasant wood to look at, takes a nice polish, and I must say that I enjoyed the brief amount of work I did with it. Although hard and dense, it is nicely workable by hand tools, though you have to be sharp.

In the upcoming installment, post 11 in the series, I turn my attention back to the process of building this lantern ever upwards, layer by layer- the lantern housing sill, or dodai is the next item on the agenda.

Thursday, June 25, 2009

First Light IX

This is the ninth part in a thread of posts describing the construction of a freestanding Japanese garden lantern. In previous posts I have described the construction of the central post, the initial tier of support arms, or hijiki, and then went on to show the construction and fitting of the second tier of support beams.

Today I will continue on with the build-up, though later on in fact it will seem like a bit of a strip down as I need to take some of the lantern apart to work on some details I had omitted earlier to tackle.

First though, the second tier of arms had been fitted together at a central 4-way lap joint, half-lap for the diagonal pieces, and 1/4 lap for the other two, which are oriented normal to the post faces. The lower piece then had to be re-shaped so it would fit on the pillow blocks. This necessitated removing about 0.5" from each end. I proceeded by ripping the pieces with the ryoba nokogiri:

After cross-cutting, I pared the surfaces down to dimension:

Then I started working the returns into a serpentine shape:

Continuing to shape the lower surface, this work done entirely by chisel:

The finished result:

There's a little more to be done yet on this lower beam, but I moved on to work the noses on the uppermost beam, on which I trimmed a slope:

At this point, with the cluster of beams assembled atop the hijiki, the scene looked like this:

You can see that the lower beam comes within an 1/8" of the top of the post. You might wonder why I didn't run it all the way down to the surface (?) - well, I have some plans for this, which I will explain shortly.

I had noticed that with the second tier of beams in place there were a few discrepancies in fit, and decided to to some checking and adjusting to the lower layer:

A couple of small adjustments were made - one of the hijiki wasn't quite sitting fully down in its trench.

Now that I had all the upper tier parts removed, I could return to dealing with a few tasks needed to be done in the lower layer. One of those concerned the routing of the electrical cord, which was to come up the center of the post. Though I hadn't cut the passage in the center of the post yet, I could now make the entry through the middle of the hijiki:

The completed hole:

Given that this is a prototype, I have made a design revision or two as the build has progressed. Another such revision concerned the connection at the top of the post. I originally designed the lantern to have a single row of hijiki and then the makitō (pillow blocks) as support for the lantern housing. Part and parcel of this design was that the central space atop the post and up to the floor of the lantern housing would be largely occupied by a hopper, as you can see in yellow in this early drawing:

With the revision in design to a two-tier support system, the lower beam of the upper tier now intersected the space in which the hopper would have been situated. I wanted the top of the post, where there was some end grain and a large number of component intersections to be protected from the elements. Wind could blow rain into that space, where it could wick its way in-between the pieces. With the hopper no longer a factor, I came up with a solution that not only protected the top of the post, but helped support the beam work and improved the strength of the entire post-hijiki connection. This solution was in the form of a cap which would be attached to the post top with a square tongue and groove joint.

First I needed a piece of material for the cap. I love the workability of the mahogany, but it didn't strike me as being tough enough for this particular role. I remembered that I had a couple of offcuts of a wood that would be perfect however: Goncalo Alves (astronium fraxinifolium). A had bought a couple of huge wide planks of it from a dealer in North Carolina several months back, a couple of planks that the dealer had imported 15 years ago and set aside for his personal stock and from which he had intended to make a dining table. That plan changed, and I was in the right place at the right time, as they say. After they arrived, I had them planed at the window and door shop just down the road. Their planer was in poor condition at the time, so the planing had resulted in some severe snipe at the ends of each board, necessitating that I trim about 4~5" off each end. That was a bummer, but for this project a blessing as I still had the offcuts:

The lighter colored sections of the piece are very close to the mahogany in color. The black streaks are clearly what set it apart, plus the fact that Goncalo Alves is about twice as dense as mahogany (the planks I have weigh about 100 lbs (45kg.) each!). It is described in one of my wood identification books as "noted for its durability", and is a wood commonly used for boat-building in its native area. This was a perfect choice of material for the cap I thought.

I made up a jig of mdf and proceeded to cut a 0.25" groove on the post top with a router:

I cut the required piece of cap, and then used the router to cut a recess in the center so that a perimeter tongue was left:

Here it goes:

It's a tight fit, by design, and need to be tapped down into place with a hammer. Once down, it helps lock the hijiki into place - the diagonal pieces in particular are noticeably stiffer after the cap is fully down on top. Now, removing the cap requires the use of a Bessey clamp to grab it - I can't do it with my grip alone. This posed an issue as my intention was to make the profile of the cap beveled on all four sides, which would make it un-grabable, un-removable by such means - this problem I also solved, to be detailed, among other things, in the next post in this thread. See you then.