Edge Lap Joint — What It Is, Where the Strength Comes From, and the Numbers

An edge-lap joint moves the notching from a board's face to its edge — the pieces stand on edge rather than lying flat, and the half-depth notches let them cross or meet while both boards stay upright. It's the joint behind most interlocking wine racks, shelf dividers, and standing lattice partitions.

Why standing the boards on edge changes what the joint has to resist

A face-notched half-lap or cross-lap mostly resists forces trying to pull the joint apart along a flat glue line. An edge-lap, by contrast, is almost always resisting bending — the boards are standing on their strongest axis (edge-up, the same orientation that makes a joist far stiffer on edge than flat), so an edge-lap divider carrying shelf load is leaning on the section's high resistance to bending in that orientation, with the notch's mechanical registration keeping the crossing pieces from sliding out of alignment under that load.

The same half-depth accuracy problem, in a less forgiving orientation

Because the boards stand on edge, any depth error in an edge-lap notch is immediately visible as a divider that leans or a grid that won't sit flat against a shelf or wall — there's no "which way is up" ambiguity to hide a small mismatch the way a horizontal lattice sometimes can. Both crossing notches need the same careful matched-depth layout a cross-lap needs, referenced from a consistent edge across the whole assembly.

Where this shows up, and why it's chosen over screws or dowels

An interlocking wine rack or divider grid built from edge-lapped pieces needs no fasteners at all — the notches alone hold the grid's shape once assembled, and many such grids are deliberately left unglued so the whole thing can be knocked down flat for storage or shipping. That's a real functional advantage over a glued or screwed grid: unlike almost every other entry here, an edge-lap divider system is routinely chosen specifically because it can skip permanent fastening entirely and still do its job.

Why glue is a genuine option here in a way it rarely is on a cross-lap

A cross-lap's notch runs across a board's face, glued in the same plane the wood most wants to move across seasonally, which is part of why some cross-laps are deliberately left dry. An edge-lap's notch runs across a board's edge instead — the direction that same board barely moves in at all — so gluing an edge-lap grid doesn't fight seasonal movement the way gluing certain face-notched joints can. That's a genuine structural argument for gluing an edge-lap grid where a knock-down design isn't wanted, on top of the purely aesthetic reasons a maker might choose one over the other.

Why the notch has to be cut dead square to the edge, not just to depth

Because the boards stand on edge, a notch that's slightly out of square to the board's edge (rather than just slightly off in depth) tips the crossing piece a few degrees out of vertical — invisible in a single joint, but compounding visibly across a multi-crossing grid where every piece is supposed to read as perfectly upright. This is a different accuracy demand than a face-lap ever makes, since a face-lap's crossing pieces lie flat regardless of small angular errors in the notch walls.

Cut sequence

1. Lay out every crossing point across the full grid before cutting anything, since edge-lap grids typically have many repeated identical notches and an error in the layout pattern compounds across every piece.
2. Mark notch width and half-depth on each board's edge, referencing from a consistent face.
3. Cut the shoulders with a saw or on a table saw with a stop block set for repeatable notch width across identical pieces.
4. Remove waste to the half-depth line, checking with a straightedge that the notch bottom is flat and consistent across the grid.
5. Dry-fit the entire grid assembled before any glue (if gluing at all) — an edge-lap grid with even one tight or loose notch usually won't sit flat until that one joint is corrected.

Load capacity is set by the notch depth removing exactly half the board's height

Because each notch removes half the board's height at the crossing point, an edge-lap divider carrying real weight (a loaded bookshelf's internal dividers, for instance) is weakest exactly where the load usually concentrates. This is manageable with adequately sized stock but is worth checking against actual expected load rather than assuming an interlocking grid is automatically as strong as the same boards uninterrupted.

Why a full grid takes longer than any single joint suggests

Similar per-joint time to a cross-lap, with the same layout-coordination cost multiplied across however many crossing points the full grid needs — a small wine rack with a dozen notches is a modest afternoon project; a full room-height divider grid with dozens of crossings is a considerably longer layout exercise.