This is the fourth in a series of articles on the development of our homestead project. In it I look at some practical measures we took in our early building projects as well as the considerations that led us in one direction or another.
A timber frame structure is a fairly complicated affair. Those who are familiar with conventional residential building methods, known as western platform framing, know that the technique includes just a few discrete classes of dimension lumber typically all a nominal 2 inches in thickness and varying only in length and width, being from a nominal 4 inches to 12 inches wide and of whatever length between 8 feet and 20 feet or so that is necessary for the application. Just a few different fastening devices from 6 penny to 16 penny nails satisfy all the requirements for connecting framing members together. The worker at the sawmill does his or her job without the least necessity of knowing the nature of the wood’s ultimate destination. Trees are turned into 2 x 4’s or 2 x 10’s, planks or veneer based solely on what the tree will yield without reference to the needs of any particular building or plan. Likewise in the nail factory, one, two, or a million 3 inch long nails are produced, all identical, all perfectly suited to joining together two pieces of wood whose actual thickness is a consistent 1 1/2 inches. No further reference to application is necessary and though there may seem to be a bewildering array of fasteners at the local hardware store, they all function within the context of a single convention.
The success of this system of building rests on a single central assumption, which is that a standardized set of lumber sizes and fastener types can be adapted at the point of use to serve a theoretically limitless array of applications. (Theoretical limits to one side, there are practical limits to what can be done with dimension framing lumber. I remember a friend telling me about a time he site-built gambrel trusses for the roof of a 60 foot clear-span dairy barn. Normally trusses for such a building are factory built to engineered specifications and transported to the site, but the site was inaccessible. The engineering firm wanted an exorbitant amount of money to bring their crew in so my friend decided to do it himself. He told me how he awoke in the middle of the night during a major snowstorm, sweating bullets, convinced that the barn was coming down. Fortunately it did not collapse that night or since but in my mind as well as his it tests the limits of what can be done with ordinary framing lumber.) The system is convenient for the lumber producer, the designer and ultimately for the builder, too, however there is a down side to all that convenience and that is waste. Estimators assume that 20% of all the wood that shows up on a building site will be waste. As an example, 9 foot ceilings are increasingly popular in the larger McMansions that are being built these days. A 9 foot ceiling requires a stud height of 104 5/8 inches or 8 feet 8 5/8 inches. The closest length to this in dimensions lumber is 120 inches or 10 feet which means that every stud cut discards 15 3/8 inches of 2x4 or 2x6. Even if 2 studs are cut form a single 18 foot length, about 7 inches is wasted from each board (In contrast, for 8 foot walls the industry provides a stud 92 5/8 inches long, known as a “precut.” No on-site waste occurs with these). I don’t have information on waste at the production end but I know form personal experience that trees do not grow to convenient even numbered whole foot dimensions. It is certain at least that the odd extras from logs do not make it to the dimension lumber racks of your local lumber yard. Ted Benson, author of Building the Timber Frame House, joked that the loggers in New Hampshire wouldn’t own a tape measure longer than 12 feet.
Compared to western platform framing, timber framing starts from completely opposite assumptions. A timber frame starts from a plan and that plan consists of a detailed rendering of each structural member. In a typical stick framed building all that is necessary is a line drawing showing the positions of walls and openings. One or two framing details will suffice to show the technique to be repeated over and over throughout the structure. In a timber frame repetition is the exception rather than the rule. In some cases (and our house plan proved to be one of them) nearly every major structural member is unique. There is nevertheless some room for factory work in a timber frame. Typically floor joists, ceiling joists or rafters will repeat and braces will be confined to 2 or 3 different sizes with several of each size required. But even in these cases, the actual dimensions of these repeating structural members will be unique to the plan. Whether a joists is a 4x4 or a 5x5 or a 5x7 or some other dimension is determined not by the availability of lumber at the yard but by its function in the timber frame structure. Whereas in common stick framing the plan might call for nailing four 2x10’s together to make a load bearing header, a timber frame plan assumes that you never cut them apart in the first place and calls it an 8x10. Once the plan for a modern McMansion is in the builder’s hand, the lumber to build it is already at the lumber yard and a simple phone call will bring it to the site. In contrast, in timber framing, it is not until the plan is complete and the required dimensions of every piece are specified that the first tree is felled and the harvesting commenced.
Finally the connections between timbers that guarantee the integrity of the structure differ materially form the off-the-shelf systems that characterize western platform framing. In a timber frame it is the inter-penetration of wooden members (in the form of mortise and tenon joints, dovetails, tongues and forks, or various scarf joints) and the use of hand-made wood pegs or trennels which unite the frame. Each joint is unique and satisfies the unique requirements of that point in the frame.
I bring all this up by way of saying that my years of experience in conventional building did little to prepare me for the experience of building a timber frame structure.
As the snow flew in the fall of 1994 we were back-filling the dirt around the foundation of the shop and preparing to start in earnest on the cutting of timbers for the frame. Laurie’s job as a teaching assistant at Skidmore College afforded her a substantial break between the fall and spring semesters and during that time we concentrated intensively on our task. January of 1995 was cold but it was mercifully snowless in Meco. We were able to get into the woods everyday and by the end of the month we had nearly half of our timbers on the ground ready to move to the building site. I should note that with our small, hand-held sawmill we developed a technique different form normal logging operations. On a typical day we would fell 2 or 3 trees and clean them of their limbs. Once we had determined what timbers they would yield we cut them to appropriate lengths and set them up more-or-less ready for the mill. All this was done right where the tree fell. The next day the sawmill and its equipment were carried to the tree and the milling done. Typically we would spend two days on milling for every day we felled trees.
There were over 250 separate pieces in the shop frame ranging from 3”x5” braces 3 or 4 feet long to sill timbers 7”x9” and 24 feet in length. Most timbers also had a species associated with them (sills were ash, bearing beams were birch, rafters were maple, etc.) and we developed an inventory system to keep track of what we cut. We first made a master timber list. This was our main tracking document. Each timber was listed with its location in the frame, cross-sectional dimensions, length, species and check-offs for when it was cut and when the joinery was complete. In addition we had index cards arranged by species listing all the pieces of each species that we needed, The cards went into the forest with us so that when a tree come down we could compare its potential yield to the appropriate index card and decide how best to cut it up. Each day’s production was recorded on the index card and each night we could sit around the wood stove and update the master timber list.
Once a timber was on the ground it still had to be delivered. This was done generally by simply dragging it out of the woods. We would use a short piece of rope with a loop at each end. One loop went around the end of the timber and the other loop attached to a stout round pole called a “draggin’ stick.” This was about 2 inches in diameter and 4 feet long and was held, Laurie on one side and me on the other, about waist high. Like a team of oxen, cursing and grunting, we would drag the timbers out of the woods and to the building site. The largest timbers required an extra pair of legs and snow was essential to lubricate the process, but the job got done. (I should note two factors which made this dragging operation easier than it might sound. First, we tried to harvest trees as close to the building site as we could. We estimate now that most of the wood came from an area of about 5 acres surrounding the building. Also, the land slopes sharply upward form the road so, for the most part, we were dragging down hill. In the one case where we harvested a number of trees a distance from the building site we were able to drag them to the main road and then truck them to the site from there.)
With Laurie back to work our timbering operations were restricted to the weekends. During the week I began working on the joinery. This work was unfamiliar to me and fairly exacting. A timber frame is like a giant jigsaw puzzle. The pieces are fashioned to fit together to very close tolerances. Even the peg holes into which the hardwood pegs are driven are predrilled to precision alignment as each joint is fashioned. An error in any one place can throw the entire frame out of alignment and mean that nothing will fit together right. To complicate matters it is impossible with timbers weighing sometimes several hundred pounds to test the fit of the joinery before the actual assembly. As I described in a previous article, we had made detailed drawings of each part of the frame. Now with the work in progress, those drawings were checked and rechecked for accuracy and each dimension transferred to the actual timber was checked twice before any cutting was done.
It surprises some people to hear that all the joinery was done with traditional hand tools without recourse to power tools, but there are good reasons for this. Going back to the discussion of dimension lumber from earlier in this article, when we buy a 2x4 or 2x6 from the lumber yard we can assume that the opposite sides of the board will be parallel to one another, that adjacent sides will be perpendicular, that it will be 1 ½ inches thick and, in the case of a 2x4 3 1/2 inches wide. If you’re careful with your selection, the board will also be straight. It is, in part, the precision-manufactured character of dimension lumber that makes possible the use of power tools in the final stage of processing. In the case of the large timbers used in a timber frame none of those assumptions can be safely made. Timbers are frequently out of square, rarely straight and dimensions can vary significantly even over the length of a single timber. Framers long ago developed techniques to overcome the difficulties associated with these inconsistencies. Collectively the techniques are called the Square Rule method (a complete description of this method can be found in Jack Sobon’s books on timber framing.) This method has its own central assumption which is that inside every timber regardless of its actual size and shape there is a smaller, perfectly straight, perfectly true timber. Using the Square Rule, we do all our work on that imagined perfect timber. The technique is no different from (and considerably easier than) that of the sculptor or carver who finds the beautiful figure within the undifferentiated lump of raw material. In short, hand saws, chisels and the indispensable barn beam boring machine are perfectly suited to the rigorous precision of the Square Rule Method.
By the end of the spring semester of 1995 when Laurie could look forward to an extended time of concentrating on the homestead, the die was cast. Invitations had been sent out to all we know inviting them to our old-fashioned barn raisin’ on August 14. Our one task then was to somehow prepare the myriad pieces of the frame so that on that day they could be assembled and raised into position with nothing more than the collective strength of a community of friends.
Deadlines aside, the rhythm of the seasons dictate the pace of life for Jim and his partner, Laurie Freeman, on their homestead in Meco, NY