Newsletter Article: Cross-Grain Construction
Cross-grain Construction
I am surprised at how many woodworkers ignore the perils of cross-grain construction in their building. One of the underlying principles of woodwork is to respect the ever expanding/contracting characteristic of wood. Like a sponge, wood is always seeking equilibrium with the surrounding atmosphere. As humidity increases, wood gets wider (across the grain) and thicker. However, it does not get longer in length, it's one stable direction. As humidity drops the opposite occurs, again, there is no change in length.
Prior to our advanced climate control in our homes, furniture makers had to pay strict attention to building techniques so as not to restrict wooden parts from moving seasonally. As far less air tight homes experienced much greater humidity swings with the changing seasons, the furniture inside the home did the same.
Frame and panel construction is a great example of the craftsman's genius. The traditional frame-and-panel door has a solid-wood panel in the middle, enclosed in a groove around the inside of the frame. It is very important not to put any glue in that groove; the panel must be able to expand and contract freely. There's also some extra room in the groove to allow for expansion and contraction of the solid-wood panel. During seasonal changes, that panel will get a bit wider as it absorbs moisture.
The frame is made of parts that are much longer compared to their width. The door height is determined by the length of the stiles, no change seasonally. The door width is determined mostly by the length of the rails. Although this dimension also includes the width of two stiles, because they tend to be narrow (usually around 2" each) there is very little movement.
The frame is considered a stable component and can be a door or the side of a desk or cabinet or chest of drawers. Inside the stable frame is an unstable solid panel that is designed to move freely (float) in the grooves cut in the frame pieces. Typically, the panel is pinned top and bottom at the mid-point so the seasonal movement will be equal on both sides of center. If not pinned, a wide panel could pull completely out of the stile on one side, exposing a gap. Frame and panel construction has been around a long time and is a major reason antique furniture has survived.
Another area where wood movement must be factored is in drawer work. Drawers expand and shrink in their height, so the appropriate clearance must be made. I like to keep my drawer sides thin to all but eliminate any movement in the width of a drawer and to be nice and light. On a side note, drawer dovetails front and back are so strong they enable very thin sides without compromising strength. Other corner joints require more mass in the side material to make the joint strong enough to survive.
With thin sides, drawer side clearance can be made undetectable and feel like the drawer is floating on a cushion of air. Fitting the drawer height is somewhat of a guessing game, leave too much and in the dry period drawers get "sloppy ", not enough room and in the humid season they get welded shut!
Solid drawer bottoms also need special attention, grain always runs parallel to the drawer front. Expansion of the bottom is designed to go out the back of the drawer, under the drawer back. Typically, the bottom is housed on three sides, a long grain edge is glued into the groove in the drawer front, the two end grain ends are dry fit into the drawer side grooves where they should "bottom out". This further strengthens the drawer and prevents any racking. The fourth edge (long grain), sits under the drawer back where it can shrink and expand.
Depending on the width, I use one or more screws each in a long slot cut in the long grain edge of the drawer bottom. They need to be tight enough to keep the drawer bottom in place and free from rattling yet allow the seasonal movement of the solid bottom . In case it doesn't go without saying, should the drawer bottom be installed with the grain parallel to the side, seasonal expansion would bind the drawer in the case!
Another problem area is the fitting of the drawer front to the solid wood case. drawer length doesn't change seasonally but the carcass will get wider and narrower. Don't use the back of the carcass as a drawer stop, doing so will change the drawer front reveal as the carcass either shrinks or expands. I like inset drawers that use the drawer front reveal as part of the design, best way to keep that reveal constant is to use a stop that references off the inside bottom of the drawer front.
Perhaps the biggest area of concern for cross grain construction is in attaching a solid wood table top to a base. Like the frame portion of a frame and panel door, the base of a table tends to be stable in both length and width. The solid top of a dining table will want to move as much as a 1/4" and possibly more seasonally. Ignoring this will result in cracks and stressed joints that will eventually fail. Allowance must be made for seasonal changes while at the same time the top needs to be held tight to the base.
Buttons are the traditional method to address this problem. More modern methods include pocket hole screws with over-sized holes to allow for movement or figure “8” hardware with over-sized holes.
My favorite method, one I came up with 30 years ago, uses short dowels with a narrow flat spot along their length so they can be glued and screwed to the underside of the table top. The dowels fit in holes drilled through the stretchers at the top of the base. The holes are partially exposed at the top edge of the stretchers so instead of 360 degrees there is about 300 degrees of the hole. When in place, the top of the dowels are flush with the top of the stretchers.
To attach the solid top to the base there are several dowels along both long stretchers but only a few on the short stretchers at either end of the frame. These are in the middle and best if only a few inches either side of center. When the table top is attached, the dowels are glued and screwed to the underside of the table. They are "dry fit only" in the holes in the stretchers. As the table swells the dowels move in the holes while maintaining a tight fit between the top and the base. The two end dowels keep the table top centered on the base, equal movement either side of center.
I often use contrasting wood dowels to make a feature of this method when the design allows it, otherwise the holes in the stretchers are not drilled all the way through and can only be seen from the underside. I this latter case be sure to leave enough room in the bottom of the hole for the dowel to expand into.
I find the challenge of working with this "live" material the thrill of the sport! Designing around these challenges keeps the creative engineering juices flowing! Now after all that, why does plywood not explode?
Till next time, enjoy your time in the shop!
--
Rob Cosman, Enjoy a free 30 days of learning, Visit freemonthonrobsworkshop.com