Grain direction decides where wood breaks

Orienting a wooden part so that it fails somewhere useful, or not at all.

Wood is built from long fibres bundled together like a sheaf of drinking straws running along the length of the tree, strong when pulled or bent along that fibre direction and far weaker when forced to split apart across it, so a part cut and loaded with its grain running the right way resists loads well, while the same part cut across the grain, or loaded in a direction that tries to split the fibres apart, can fail with startlingly little force.

What is actually happening

Bending a strip of wood along its grain puts individual fibres into tension and compression along their own length, which is exactly the direction wood fibres are strongest in, since each fibre is essentially a thin, tough tube reinforced along its length, and bending or pulling along that length has to actually break the fibre material itself to fail. Splitting wood along the grain works completely differently, since a wedge driven in along the fibre direction is not trying to break any fibre at all, it is only trying to separate one fibre from its neighbour, and the natural bonding holding parallel fibres to each other side by side is far weaker than the fibres themselves. Load the same piece of wood across the grain instead, bending it so that the fibres are being asked to resist being pulled apart sideways rather than stretched along their length, and the wood behaves almost as weakly as it does when being deliberately split, because sideways loading puts stress directly onto that same weak bond between neighbouring fibres rather than onto the strong fibres themselves. A knot or an awkward twist in the grain makes this worse locally, since wherever the fibres are forced to run diagonally across a piece rather than straight along it, any load applied along the piece's intended length is partly resolved across that diagonal grain instead, exposing the weak bond between fibres to stress it would never see if the grain ran straight.

The firewood comparison

Splitting a log for firewood shows exactly how little force is needed once a cut runs the right way relative to the grain. A wedge driven in along the length of a log, aligned with the grain running down its length, parts the wood with a single solid strike or two, the crack running ahead of the wedge with surprisingly little resistance, because the wedge is only separating fibres from their neighbours rather than cutting through any of them. Try to sever the same log crosswise instead, cutting straight across the grain to shorten it, and the same wedge and the same effort achieve almost nothing, since now every fibre running across the cut has to actually be severed individually rather than simply parted from its neighbour, which is why cutting a log to length takes a saw working steadily through the fibres rather than a single well-aimed strike.

The number that matters here

Wood can take many times less force to split along the grain, separating fibres from each other, than it takes to break the same piece cleanly across the grain, severing the fibres themselves, an imbalance large enough that a part loaded the wrong way relative to its grain can fail at a small fraction of the force the same piece would resist if the grain ran the other way.

What follows from this

Choosing how to cut a wooden part from a sheet or a plank, and how to orient it once installed, matters as much as choosing the wood itself, since a thin part cut so its grain runs across a point where it will be bent, rather than along it, is set up to fail exactly there under a fraction of the load it could otherwise resist. The same weakness can be used deliberately rather than avoided, since a part meant to break away safely under overload, rather than transmit damage further into a structure, can be given a grain direction that makes it the weakest link on purpose, failing cleanly at a predictable point instead of splintering unpredictably somewhere else. Inspecting a broken wooden part afterward usually tells this story clearly enough on its own, since a clean, fibrous split running along the length of the break points to a load that caught the grain at its weak angle, while a short, splintered break straight across the piece points to genuine overload of the fibres themselves rather than a grain-direction mistake.