Monday, May 27, 2013

Joinery


Joinery
There are numerous potential types of joints that may be used in a timber framing project. They vary in complexity, size, strength and utility for a specific purpose. I used the types of joints that Ed used and taught me how to make.
Each of these joints also may be housed or flush. A housed joint may be stronger because more of the timber end is encased within its accompanying mortise. This additional timber end wood also more gradually reduces the size of the end of the timber which provides more support of the tenon in its mortise and reduces the likelihood of the timber cracking at the abrupt change in dimensions of a tenon that that does not have a housing. I used housed joinery for the beams and girts and flush joinery for the braces and joists. When the wood eventually shrinks and a space between the two joined timbers appears, what can be seen in that space of a housed joint is the housing but not the joint. When the space appears without a housing, the tenon is visible.
The entire frame employs 4 different types of joints, a mortise and tenon, a dovetail, a scarf, or a spline. Each of these joints are designed for the particular timbers that are to be joined. The large mortise and tenon joints are housed while the smaller ones are not. When the timbers meet perpendicularly, the mortises and tenons are simple rectangles. All the braces are joined at a 45 degree angle so the mortises and tenons are also cut at a  45 degree angle. The rafter bases sit on the top of posts so the rafter bases are cut at the angle of the roof. This is done with a joint called a birds beak.
“Timbers and Joinery” is a spreadsheet listing of all the 379 timbers (“Timber List") that are in the frame. Each timber has at least a mortise or tenon at each end (like joists and purlins) and several timbers have many joinery sites. 20 of the timbers have 10 or more joint sites and two girts have16 joint sites. There are a total of 648 mortises and 648 tenons in the frame. Dovetail mortises and tenons account for 270 of these joints. Brace mortises and tenons account for another 168 of these joints (84 braces). The 66 spline joints are counted as 3 tenons, one for the spline itself and two for the mortises to which they attach, so there are a total of 22 splines. There are 12 scarf mortise and tenon joints with a mortise and tenon for each half of the joint, so there are 6 scarf joints. The other infrequently used type joint are the 16 rafter birds beaks. The other 124 mortises and tenons are beam and girt end joints. In all, there are 1,296 individual joints to be cut.
All of the deck joinery follows traditional techniques. The deck corners are tongue and fork joints with a large center peg that extends up into the base of the corner posts. The 40 foot long sill beams are two 20 foot long timbers joined with a three feet long scarf joint with a center mortise for the perpendicular beam that locks the lap joint lengthwise. The beam tenon is pegged through the scarf joint, again, with a large center peg that extends up into the base of the post that rests on the intersection.

 


These are photos of one off the sill beams under construction. On the left on the sawhorses is the center scarf joint. The end of the beam below it is a tongue and fork joint. The photo below is of the other ends of these beams.

The photo to the left shows the process of cutting three of the six dovetail mortises that are in one of these sill timbers along with the tools employed. On the far right next to the drill, the mortise sides and back have been cut with a circular saw. Three holes were drilled to remove much of the wood. With the mallet and chisel in the center, the excess wood has been removed. On the left, the mortise has been fine tuned for the correct depth and straight back and sides of the mortise.

 

 In the photo below, these two timbers are joined end to end on top of the concrete foundation wall. The engine crane is placing one half of the connecting beam that will span to the opposite side of the foundation and connect to a similar set of sill beams. The photo above is a close-up of the mortise for the beam tenon that the crane is holding. I chiseled my initials and "1-1-96" when I made this mortise - Happy New Years day! The lower date, "6-18-96", is the date that this joint was put together.
 














The previous photos are of the assembly of the east center set of basement spanning beams. Above is the joining of the west center set of basement spanning beams. The post is free-standing (it has a copper plate underneath it to protect the wood from the concrete floor) and when the right hand beam tenon slides into its mortise, the connecting beams will rest on top of the post while the two scarf joints also fit together.

 

The photo on the left is of the last of the eight sill beams being raised. The scarf joint hanging in mid-air will connect to its mating scarf joint in the upper left of the photo. Meanwhile, the mortise at the far end of the beam will fit with its matching tenon on top of the concrete wall. Out of the photo on the left side is a post which supports this beam - the brace from this post is visible.

Below is a photo of the daylight basement wall, ready for the installation of the three connecting girts and their 18 joists.
The finished deck
All of the beams have housings that are one inch less than the beam size on the bottom and each side. All of the tenons are one and a half inches by four inches including the three quarter inch deep housing. Thus, the tenon is three and one quarter inches long and the peg center is placed one and a half inches from the outside of the beam. When I created mortises, I made sure that they were at least one-quarter inch deeper than the four-inch tenon. This puts the one inch peg edge one quarter inch inside of the housing and two and three quarter inches (the relish) from the end of the tenon. This housing, mortise and tenon design is based upon my assumptions of potential wood shrinkage over time. I assumed that the length of a beam would shrink at most about an eighth of an inch, or about one sixteenth at each end. I also assumed that the width of a post would shrink at most about a quarter of an inch on each side. Therefore the eventual shrinkage at a post to beam joint would be about five sixteenths of an inch. Thus, the three quarter inch housing, after the post and beam shrank would still be at least half engaged. As I inspect these joints that are now between five and seven years old, the average gap between the post and beam is about one quarter of an inch so one half inch of housing is still engaged.


 There are a number of different post connection problems in my frame design. The simplest joint is when two beams meet a post at an outside corner. I used 8x10 beams in the bent and 6x10 girts for the outside of the bays. This smaller bay girt allows space for the 2 inches needed for the bent beam peg, a one-inch peg centered one and a half inches from the post edge. The bay girt peg has to be drilled and installed from the outside of the post since the peg ends at the housing of the bent beam end. Care must be taken to avoid pounding this peg in so far that it might push the bent beam joint apart. To allow for future shrinkage, I made sure that the peg end would be at least one quarter inch from the bent beam housing.



Note the model "birds beak" test joint
 In the case where three beams meet at a post, I resorted to using a single spline for the bent beams and pegged the bay girt. The trick here is to install the bay girt before both of the bent beams are attached to the spline. Otherwise, the two bent beam-ends cover the peg site for the bay girt since the peg for the bay girt tenon is inside of the bent beam housing. This bay girt peg can not be driven so far that it contacts the inside of the opposite housing. I made sure that the peg end remained one quarter inch behind the back of the beam housing and was driven one quarter inch deeper than the back of the other beam housing. If the post is eight inches wide, the bent beam housings are both three quarter inches deep, and the peg ends are one quarter inch inside the back of the bent beam housings, then the bay girt peg has to be six inches long.

 
 The joining of the bent beams and bay girts at the central box corners involved the intersection of four beams all at one point. At these four points, two splines were needed. The bent beam splines run through the posts to support the beams from underneath. The bay girt splines run the opposite direction, divide the top of the posts and are set in long mortises along the top of the girts.




 Above is a photo of the spline joinery at the top of the posts used in the central box of the frame. At the right is one of these posts at the base of the loft stairway.
     






The roof framing, like the deck, also follows traditional joinery techniques. Since I felt no need to create an attic and wanted the cathedral ceiling look, the beams and girts did not have to meet at the same point on the posts to create the ability to install a flat floor. The ten-foot tall queen posts (photo below) left plenty of space for the bent beam to be at the eight-foot height of the first floor beams. The bay girts are joined to the posts below the bent beam in the more traditional fashion. The lower rafter runs from the top of the first floor post to the top of the queen post.



The upper rafter continues from the top of the queen post to the top of the king post (photo above). The king posts are 8x10s that narrow to 8x8s at their bases. This provides plenty of joinery space for the two rafter tenons that are joined on each side of the king post, as well as the ridge beam tenon that is joined to the opposite face of the king post. Connecting the rafters are dovetailed purlins on four-foot centers. The entire frame did actually get finished in November of 1998.



For some reason, lost in the mists of time, I wanted a 9-12 roof pitch. It could be that the 3-4-5-triangle that describes this pitch intrigued me. Given the immense number of forty-five degree angles created by the multitude of braces and from a practice standpoint, making 6-12 pitched roof might have been easier. Another reason for this particular roof pitch is that the roof angle closely corresponds to the appropriate angle for solar panels in the winter. I suspect that the 9-12 roof pitch simply provided a more pleasing appearance and geometry. The reason could not have been the need for loft space since a forty-foot span, starting two feet above the floor level, would create a huge loft at any reasonable roof pitch.


 
I spent the last days of my apprenticeship with Ed, going over my house plan drawings. This was the occasion when he told me that I would need a crane. Not wanting to believe this, I changed the subject. The only kind of joint that I had no shop experience making by then was a birds-beak, the crucial joint between the posts and rafters. I don’t know if it was his confidence in me, or total skepticism about this whole project, but his response was, “you’ll figure it out when you get there.” Curiously enough, I began this project starting with the decking through to the erecting of the wall framing having never made a birds-beak joint. Why I did not believe his pronouncement that I would need a crane, but did believe this proclamation that I would know how to cut a birds-beak joint when I got there still baffles me! It turns out that Ed was wrong about the crane but right about the birds-beak.

The day I met Ed - he sent me to the top of the left hand dormer!

Ironically enough, my first hands-on experience with timber framing involved a dormer. I included a south-facing dormer in my plan because I had visions of a loft full of light, utility and space - a whole living space unto itself. I took photos of the dormers that I had helped to construct but the joinery at the valley rafter peak eluded me. This point was the only place in my plan that was a vague sketch rather than an exact drawing. I had asked Ed about valley rafters and their connection to a purlin several times. On this last day before leaving for Colorado to begin my adventure, he revealed to me that he was still searching for the perfect solution to this joinery problem! I was thunderstruck! The last house that we had built included five dormers in a single room! Rather than talk about mundane birds-beak joints, we went out to the shop for a quick lesson on Ed’s current notion on how dormers should be framed in timbers. While I made a sample joint to take with me to Colorado, Ed fired up his portable band saw and made a ten-foot long 8x14 to serve as the purlin to house the two valley rafters and dormer ridge beam. I was touched and thrilled. Ed really did think that I could build my own timber frame, complete with a dormer. I thought of my apprenticeship with Ed as Timber Framing 101. Obviously, Ed had confidence that I would pass Timber Framing 102, birds-beaks and valley rafters, all on my own.
As it turned out, a dormer did not get installed. The massive roof timbering that I had committed myself to building without a crane was more than enough of a challenge at the time. I figured that if I really needed a dormer, I could, at some point in the future, simply rebuild that section of the roof. I still have that ten-foot long 8 by 14 and the lack of a dormer still disappoints me somewhat. On the other hand, the roof is finished and does not leak. A ten-foot long 8x14 will make a dandy mantel for the fireplace to be. Sorry about that Ed.
As fate would have it, my neighbor in this wilderness development, two lots south of our lot and about a quarter mile distant, had the audacity to build his home in the direct line of sight of the view from this un-built dormer. Just recently, I planted five Douglas fir seedlings at the edge of my lot. Hopefully, someday these trees will screen the view of that house. Then, some enterprising owner of this house may wonder why there is no dormer and install one!
The braces within the bent are 36 inches long and the braces within the bays are 30 inches long. Since the beams all meet at the same level the different sized braces insure that the brace mortises are at different level on the posts.
Although I had ordered extra timbers to cover the possibility of cutting disasters or flawed timbers, virtually every timber in the pile had been carefully evaluated for its size and shape and marked ahead of time. Once a section was raised, I went to the timber pile to find the timbers that I would need to complete the next section. It never failed that the timbers that I wanted to use next were not at the top of the pile. I stored timbers on rails that were supported off the ground by the short blocks that were the waste end of other timbers. These rails were usually the ten-foot timbers that I had rejected. On these rails, I tried not to stack timbers more than two high with three-inch stickers between them. I also left the ends of the rails unused. That way, if the timber that I wanted was on the bottom layer, I could slide the upper layer onto the unused part of the rails, lift the timber that I wanted up on top of the stack, slide it to the edge of the pile and onto my cart and wheel the timber to the basement workshop. I repeated this process to get the particular timbers that I wanted to work with into the basement. Then I went back to the pile and re-stacked the remaining timbers and recovered them with the tarp.
I started this project using plastic tarps. Not only were they hard to keep from blowing away in the wind, when I held them down with blocks and tied their edges with ropes, the wind just flapped them to death. In due course, the sun also rotted these tarps and I ended up with blue plastic threads all over everywhere. The building inspector at the time sold me his large (50 feet square), used, army-green canvas tarps. These tarps take much longer to rot in the sun and weigh a ton, which at least made them easier to tie and weigh down and harder for the wind to blow them away.
Once I had the timbers that I needed in the workshop, I stored them on blocks as well. This not only kept them dryer (rain and snow did blow into a daylight basement) but also made them easier to get to the tops of the sawhorses. I made an “in” pile, and an “out” pile with the sawhorses in between. The trick is to make the piles close enough to the sawhorses so that timbers could be easily moved from place to place (the one end at a time trick) and still leave enough room around the sawhorses to work on one.

Once a timber finally made it to the sawhorses, I planed it with a three-inch electric planer and sanded it with a three-inch belt sander until it as smooth, straight and square as I could make it. This often resulted in the timber being smaller than the planned size. Though this finishing work may seem like overkill, Linda thanked me five years later when she was oiling the finished house timbers. The timbers are so smooth that she did not get even one splinter.
The difference between the design post size and the planed and sanded post size results in the need to make minor beam and girt length corrections. For instance, the designed beam between two posts is 176 inches long and the two posts are designed to be 8 inches wide. If one of the posts is 7 3/4 inches wide, then the the beam needs to be cut 1/4 inch longer (176 1/4 inches long). Therefore, in order for final house dimensions to be correct, before any beam or girt timber is cut to length, all of the actual post dimensions must be established. The goal is to insure that all of the outside timber surfaces are flush with each other and the outside frame dimensions are correct. Minor timber dimension differences are left on the inside of the frame where flush surfaces are not necessary. If a beam that is 7 3/4 inch wide meets a post that is 8 inches wide, the joinery is cut so that the outside surface is flush and the missing 1/4 inch of width is on the inside of the frame.
Before any joinery is cut into any timber, several important preparations need to be made. The work space should be large enough to easily maneuver the timbers, well lit, at a comfortable temperature and have adequate ventilation. The sawhorses used to support the timber must be sturdy enough to support several hundred pounds without wiggling, at a height that is appropriate to the comfort of the timber framer and sit firmly on the floor. It is handy when using electric tools to have several outlets of adequate amperage available so that all the tools that will be needed can stay plugged in. One could argue however, that having only one extension cord to trip over and one tool plugged in at a time is safer. It is also important to have a place out of the way to dispose of cut-off pieces of wood. I usually kicked them under the sawhorses and cleaned up the mess at the end of the day. This is not the safest technique since these pieces under the sawhorses still present a tripping hazard.
Although it might appear obvious, it is easy to forget how heavy a timber can be. Never allow a body part to be downhill from a timber. Once gravity unexpectedly decides to move the timber, you cannot move your precious body part fast enough to get out of the way. One should always roll a timber away, rather than toward you. Although a timber falling off of the sawhorses and crashing to the floor is not desirable, having to try and jump out of the way of a falling timber is worse.
I often had two timbers on the sawhorses at one time. This trick is handy when you are creating identical timbers since it is easy to see that they match each other. The other timber also serves as a handy table on which to set tools and measuring devices. However, it is also easy to knock the other timber off of the sawhorses when rolling the timber that you are working on. It is also easy to find your fingers between the timber you are rolling and the other timber. It is also probably obvious that you should not roll a timber such that your fingers will end up between the timber and the top of the sawhorse.
I found that the best sawhorse arrangement is to separate them by a bit less than half of the length of the timber. Keeping the timber approximately centered on the two sawhorses places the center of gravity of the timber between the sawhorses and distributes the weight of the timber evenly on each sawhorse. This arrangement also provides plenty of room to work at the ends of the timber. If you should slide the timber along the sawhorses too far, the end that slides off of the sawhorse will fall between the two sawhorses where you are not likely to be standing. If for some strange reason you want to spin the timber around, with a great deal of caution, one can slide the timber onto one sawhorse and it will be balanced. One can then walk the end of the timber in a circle while it is balanced on this sawhorse. The need to perform this trick is probably a result of poor planning.
Dropping a chisel on concrete is also a really bad thing to happen and causes hours of sharpening time. My chisels were made in England and the steel is extremely hard. The good thing about this is that they can be made to be extremely sharp and keep that sharp edge for a long time. The bad thing is that sharpening them takes forever. I once had my corner chisel sharpened in a machine shop. The owner happily agreed to undertake this project but when I returned to pick up my chisel and asked if he would sharpen my other chisels, he said that the price would be considerably higher. It seems that he did not realize the hardness of the steel of my chisels and vastly underestimated how long it would take to sharpen them!
Another chisel edge hazard that one probably doesn’t think about is the stuff that people put in trees, like nails and bullets. Bullets are usually made of lead or copper and cutting through one is merely a surprise. An iron nail, on the other hand, can make a dandy nick in a chisel edge. Surprisingly enough, a nick in the chisel edge has a significant effect on its sharpness and grinding out a good size nick can take quite a while. Pounding your chisel through a nail buried in a timber is extremely annoying and to be avoided at all costs!

No comments:

Post a Comment