This spring in preparation for my workshop rebuild I needed to build a window sash. I started by reviewing an array of sources to understand traditional window sash design and construction (a list of my references are included below). It became clear that the basic design of sash joinery is pretty similar among sources, but there is important variation in building techniques.
The fundamental joint in a window sash is the mortise and tenon. Nearly all of the sources suggest that the joints should be prepared before the moulding and glazing rabbet are completed. The method of “sticking” the moulding profile (typically an ovolo) and rabbet can be accomplished in a variety of ways. The British approach would be to stick the ovolo with a moulding plane and the rabbet with a sash fillister. The resulting material between what is cut by these planes would be the fillet. The American approach would be to employ a “stick and rabbet” plane which has two irons and cuts the ovolo and rabbet simultaneously. A similar and more “modern” (not necessary better) method would be to use a Stanley 45 with a sash iron. Because of the Stanley 45’s square iron and minimal supporting surfaces, it can be harder to use with difficult grain. For my window build, I utilized my Stanley 45 and the sash iron fresh from rehab and honing. Really not a matter of preference as I do not own dedicated sash planes, so the 45 was my only option. Hope to change that in the future.
Window sashes are traditionally built from stable, straight grain material and softwoods were typical in the US. Without a source for 6/4 eastern white pine, I built my window sash from a home center 2×6 (it was ponderosa pine). It took much selection to find a board where I could rip material so that the stock had the growth rings perpendicular to the face much like quarter sawn lumber. My next step was to create a “story stick” based upon the open size for the window. A “story stick” is a traditional, shop-made template in sash making that outlines the overall dimensions of the window as well as location and size of the joints and moulding. I followed the process described in Doormaking and Window-Making to mark-out the joints and moulding dimensions onto the “story sticks” (one for rails and one for stiles).
I dimensioned the rails and stiles for the sash based upon the opening height and width from the “story stick”. For this small casement window, the stiles and upper rail were 1 3/4″ wide, while the bottom rail was 2 1/2″ wide. The rails and stiles were a bit thinner (~ 1 1/4″) than the traditional 1 1/2″ because of some twist that had to be planed out of the board. Fortunately the sash iron of the Stanley allows for a small range of thickness; stock that is too thin would yield a glazing rabbet that is too narrow. Next I clamped the rails (and then stiles) to the “story sticks” and transferred the joinery layout lines using a knife. Layout lines were transferred around the pieces and a marking gauge was used to define the tenon, mortise, and moulding depth (i.e., 3/8″).
Once the pieces were marked, the next step was to chop the mortises and saw the tenon cheeks. My Stanley 45 was then used to work the glazing rabbet and stick the ovolo moulding. The mortise chisel should match the width of the fillet. Mine was slightly undersized (5/16″) and this discrepancy was corrected by some additional paring in several of the steps below.
The tenon shoulders were then sawn and the tenon width was cut to match the mortise. The haunch sockets on the rails that accept the fillet from the stiles were then chopped.
The final step before assembly was creating a scribed joint so that the ovolo moulding on the stile was overlapped by the ovolo on the rails. The scribing process was as follows.
Cut away the moulding (ovolo) on the stile at the mortise leaving about 1/2″ of ovolo near the inside mortise wall, but let the fillet remain to form the haunch.
Stile after paring away the portion of ovolo moulding
On the end of the rail, first pare the ovolo using a 45 degree miter block. Next, using an in-cannel gouge suitable for the size of the ovolo moulding, pare away the mitered material to form the excavated area where the ovolo on the stile will pass under the ovolo on the rail.
Having finished all joinery and scribing each joint should be test fitted. Once a satisfactory fit is achieved the sash should be assembled. Before assembly I sawed kerfs into the tenons to accept oak wedges. I used hide glue and held the joints in place using clamps then the oak wedges were driven to serve as a mechanical counterpart to the glue.
It was pleasing to measure the diagonals of the sash and to find them in congruence. This meant I was successful at creating square joints all around. After the glue had cured, I sawed off the horns and cleaned up the sash with my smoothing plane. This project was a great learning experience for me and I hope this post assists some others trying to tackle sash joinery.
Anonymous. 2013. Doormaking and Window-Making (reprint). Lost Art Press, Fort Mitchell, KY.
Anonymous. 2016. The Woodworker: The Charles Hayward Years, Volume II. Lost Art Press, Fort Mitchell, KY.
George Ellis. 1987. Modern practical joinery (reprint). Linden Pub. Co., Fresno, CA. 486 p.
Paul Hasluck. 1907. Cassells’ carpentry and joinery. David McKay, Publisher, Philadelphia, PA. 614 p.
Charles Hayward. 1979. Woodwork Joints. Sterling Pub. Co. Inc. 128 p.
Roy Underhill. 1983. The Woodwright’s Companion – Exploring Traditional Woodcraft. UNC Press, Chapel Hill, NC. 203 p.
Roy Underhill. 2008. The Woodwright’s Guide: Working Wood with Wedge and Edge. UNC Press, Chapel Hill, NC. 192 p.