Timber Frame Shrinkage and What Moves After Completion

Timber changes dimension with moisture content, and most of that change happens across the grain rather than along it. In a platform frame, the horizontal members are the ones laid across the grain, and they are stacked at every floor band, so the change accumulates in bands rather than being spread up the height of the wall.

That single fact explains most of what follows. The studs are not what move; the plates and rim members at each level are, and everything that runs past those bands without moving with them becomes a place where the difference has to appear.

The entries describe the expected condition as the frame shortening slightly while anything rigid and continuous past that band does not. The named consequences are a distorted opening, a cracked junction, a tall window that no longer sits square in its surround, or a stair or riser fixed as though both sides moved together.

So the expected behaviour is the frame changing, and the reported symptom is whatever was attached across it. That is why the useful question is never whether the frame moved but what was allowed to bridge the band it moved at.

Where a masonry face is carried on a framed backup, the anchor is where two different movement regimes meet. The veneer changes dimension as a fired or cast product responding to moisture and temperature; the frame behind moves with its own moisture content, its loads and time.

An anchor stiff in every direction simply transfers that difference into the masonry as cracking, so the connection is asked to restrain the leaf outward and inward while yielding along the face of the wall. The entry notes the effects tend to appear at corners and beneath long horizontal supports rather than evenly.

Large sections dry slowly and go on changing shape long after they are in place, mostly across the grain, and the post and beam entry frames the design question as not how to stop that but what is allowed to be rigid against it. As the sections dry the joints themselves change: shoulders close up, pegs bear differently, and the frame settles into a slightly different shape from the one it was raised in.

Mass timber plates move less than sawn sections but still respond to moisture, and where a structure stacks many horizontal plates the accumulated effect at the top of a building is described as worth designing for explicitly.

The post and beam entry names another misreading directly: splitting along the grain of a large section is treated as failure, when it is a normal consequence of drying that is assessed rather than assumed. The word assessed is doing the work there. It means somebody qualified looks at it, not that it can be dismissed.

The same distinction applies to gaps opening at joints and to shoulders closing. What separates normal drying behaviour from a condition needing attention is a judgement about the member and the building, and it is not available from a photograph or a description.

Where a non-loadbearing partition meets a floor above, one side is still moving and the other has to be finished. The head has to hold the wall sideways, which is a positive connection, and let the deck descend, which is the absence of one, along the whole length of the wall.

The published reading is unusually specific and worth knowing: cracking in a long straight horizontal run along the head line that is filled and comes back, doors within the partition binding at the top and needing easing more than once, studs found bowed when the lining is opened. It is the crack returning after a second making-good that separates a propped partition from a finish that simply moved.

The rule that emerges across all these records is simple to state and easy to break: anything crossing a band where movement was expected has to be able to follow that movement or has to stop short of it. A rigid fixing, a continuous screed, a bonded finish or a tight-packed frame reinstates exactly the restraint that a detail existed to remove.

Movement joints obey the same rule from the other side. A joint bridged by a rigid finish applied over it has been eliminated, and the crack will simply appear elsewhere, usually a short distance away and usually in a position nobody would have chosen.

The difference between an expected change and a defect is established by somebody qualified looking at the building, and what helps them most is a record rather than an opinion. Where the cracking runs, whether it follows a consistent line, whether it has been made good before and whether it came back are all part of the evidence.

Photographs dated and taken from the same position over time are worth more than any single observation, because they show whether something is still moving or has settled.