Fig. 498.
Fig. 499.
Fig. 500.
All such arrangements are, however, suited for the rougher class of work. When we come to nicer work we must have something more scientific, that will swell and shrink as little as possible and that will look better. So, instead of using a broad flat surface with the boards all running one way, we try to overcome the faults of the flat door by framing the pieces together. Suppose, for a theoretical case, that you make a door like Fig. 498. It will not warp or curl because of the cleats at the top and bottom, but it will swell and shrink in width because there is such a wide surface of board to be affected by the atmosphere, etc., and it may become winding. To lessen these objections the middle part of the board can be removed—all but a strip at each edge (Fig. 499). It will not now swell and shrink much in width because most of the board has been removed. This frame will hold its shape quite well, but it is only a frame, not a door. How can you fill up this open frame to make a door, so as to avoid the trouble about warping, winding, swelling, and shrinking? First, however, as this frame is considerably taller than it is wide, you will readily see that it will be a better arrangement to make it as shown in Fig. 500, with the cross-pieces between the uprights, according to the usual custom in such cases. Suppose, now, that you fill up the open space with a thin board, fastened on one side (Fig. 501), instead of the thick wood which occupied the space at first. The thin board will tend to warp and twist, but, being thin, it will not exert force enough to change the shape of the thick frame. That will prevent the warping and winding from doing much harm. If the screw-holes in this thin piece are reasonably loose, they will allow play enough for the board to expand and contract without putting any strain on the frame. This arrangement does not, however, look very nice on the side to which the board is screwed, though it can be used in some situations. Suppose, finally, that you cut a groove around the inside edge of the frame of the door (Fig. 502) into which this thin board can be fitted loosely, making the groove deep enough to give the board room to shrink and swell in width without dropping out or pushing against the frame. You now have a complete door (Fig. 503), and the warping, winding, swelling, and shrinking will do as little harm as possible. That is all there is to the theory of framing doors, panels, and the like.
Fig. 501.
Fig. 502.