By Nicholas Bogosian
In the Materials Science of Wood class at BPR this quarter, we’ve been assigned six projects: Bracket reconstruction, wood epoxy repair, dutchman repair, lathe turning, wood carving, and parquetry design. The focus of the course is to get us fully acquainted with the character of wood, the tools by which we manipulate it, its common deterioration mechanisms, and basic methods by which to conserve, preserve, and restore it. The nature of the more significant projects (bracket & parquetry) lead us into aspects of fine wood working, whereas the separate Building Carpentry class focuses on wood as a framing material in a historic context. The Building Pathology component of the program, in turn, reinforces the study of deterioration and stabilization of materials such as wood.
This month I documented the process of my bracket reconstruction. “Case by case basis” is a phrase we hear all the time in our classes. The goal of the program is to equip us with an index of options. Much like a doctor upon hearing her patient’s symptoms, she must catalogue in her brain potential causes and possible remedies. If she is a good doctor, the cause of the symptoms will be considered the first priority to solve. In the field of preservation we also have other variables dictating our actions: time, the vision of the owner of the object/property (are we restoring to mid-18th century or are we leaving “as is” and conserving what we have only?), and the budget of the owner.
In the context of my bracket reconstruction I pretty much assumed the vision of the project as a restoration of sorts. I also assumed that if any problem exists that was a direct contributor to the bracket’s complete failure/disappearance, that it has been investigated and fixed. Whereas dutchman and wood epoxy repairs are repairing a wooden object and retaining as much original fabric as possible, a reconstruction effort is dealing with recreating an object based on documentation of what used to be. Perhaps only a couple of the brackets remain. Perhaps none exist at all. If it fits the parameters of the project’s vision, the reconstruction process may begin once all proper documentation and research has been accomplished.
All documentation and research aside, I began at the drafting table rendering the bracket in detail. Generally, all profiles need to be explored. I learned very quickly in the construction process, that this time spent at the drafting table is the most difficult and most important part of the entire process. Every dentil, every depth, every component of the design must be understood in your mind and explained on the paper. If you can see its multiple layers coming together accurately, then the construction process will run much more smoothly.
A bracket’s width is determined by the height of the individual boards that compose it. A process of glue lamination will give us our bulk. Once the height of these individual boards is determined, they are planed down to the correct size. In our case we’re dealing with rough-cut Poplar. Rough-cut boards are not necessarily the dimension we need and may show signs of crooking, cupping, and bowing.
A note on dimensional lumber…
The most cost-effective and resourceful method of dimensioning lumber in a lumber mill is the plain sawing method.
The downside to plain sawn planks is the nature of the growth rings in relation to moisture evaporation processes. They are more prone to warping. The quarter sawn method produces a more durable cut of wood that is less prone to this warping.
In our case, the boards are roughly plain sawn. Each face grain is planed down to the correct level in the planer which also provides a finer finish. The purpose of the planer is to give plumb dimensions on these face grains as well.
After the face grains have been planed, one edge grain per board must be joined in the joiner to remove any imperfections such as crooking. Once a single edge grain side has been joined, the other side must be trimmed off on a table saw setting the recently joined side against the fence. End grain sides may be simply trimmed on a chop saw. Now the board should be square on all sides.
After all individual boards have their proper height, the edges are glued together with a Poly Vinyl Acetate adhesive (i.e. white glue and wood(yellow) glue). These adhesives are water based and work best on porous materials. F-clamps keep the boards in place in the drying process.
It is best to arrange the boards in alternating end grain patterns. Should further warping occur, ideally the warpings will oppose each other and cancel themselves out.
After the boards have dried, the process of tracing the side profile of the bracket onto these begins. I used a simple carbon paper. I needed to trace seven profiles, as seven profiles would create the width of my bracket once placed side by side.
Once the individual profiles have been cut using a scroll saw, they are aligned together and once again glued in the final lamination process.
There might be irregular edges along this profile after the lamination process. Using a bobbin sander, the bulk of the bracket may be sanded down to a smooth and regular shape.
One component of my bracket was a turned rosette. After a block is attached to the end of the lathe, using various turning speeds and different turning chisels, my contoured shape was created. These discs were then glued to both sides of the bracket.
In creating the decorative scrolls which flank the bracket, a 3-D carved depth illusion is given by joining two pieces: one creating the elevated portion and the other providing the backing.
Using a scroll saw once again, the piece is “carved out.” Once the two pieces are glued together, a simple dremel tool helped to establish even more depths in the scrolls. These too were glued to each side of the bracket.
The last decorative element of the bracket was creating the partial architrave on the top and base consisting of a simple cornice and dentil run. It is worth noting that options for replicating historic and even rare molding profiles must be indexed as well for future “case by case” assignments. Options can run the gamut from locating rare router bits, creating custom router bits, or even doing a combination of routing with existing bits in one’s collection and hand planing/shaping. All decorative trim and molding must be carefully tagged, photographed, and organized if detached from a structure in a preservation endeavor.
Once a matching router bit was found, the cornice was shaped using the router. Various miter joints must be cut with miter saws to create the corners of the cornice.
Dentil blocks can be created with a few different methods. The most time-efficient method is using a dado blade on a common table saw. The dado blade is intended to carve out the wood. The width of this uniform shape is determined by placing spacers in between two saw blades and based on the height of the saw blade. A jig is created for the assignment if not already in your jig collection. By simply passing the dentil plank inside a jig over the dado blade, the spacing in between the dentils is created accurately.