Why solid timber tables split in Adelaide homes

Solid timber tabletops split in Adelaide homes because the house gets drier than the timber was built for, and because the top was fixed in a way that would not let it shrink. Both halves of that sentence matter. Dry air alone does not break a table. Dry air plus restraint does, and restraint is entirely preventable at the bench.

This is the most common avoidable failure in the Adelaide market, and the question that separates makers fastest. The wider species context sits in choosing timber for custom furniture.

The one number that explains it: equilibrium moisture content

Timber is not inert. It exchanges moisture with the air around it until the 2 reach a balance, and that balance point is called equilibrium moisture content, or EMC. It is expressed as a percentage of the timber's dry weight.

Put a board in humid air and it takes moisture on and swells. Put the same board in dry air and it gives moisture up and shrinks. This never stops. A 90-year-old table in an Adelaide villa is still doing it every season, and always will be. A maker is therefore not building to a fixed dimension. They are building to a dimension that will change, and the job is to control where the change goes.

Timber moves across the grain, not along it

This is the single most useful thing to understand, and it is why tables split lengthwise rather than crosswise.

Across the width of a board, seasonal movement is significant: measured in millimetres, and easily visible. Along the length of the same board it is negligible, well under 1mm over 2,400mm. The cells run lengthwise, and they change diameter far more than they change length.

Grain orientation changes the magnitude again:

• Flatsawn (growth rings running roughly parallel to the face) moves the most across its width, and it cups, because the 2 faces move differently.
• Quartersawn (rings running roughly perpendicular to the face) moves about half as much and stays flat. It costs 10% to 20% more and it is the best-value upgrade available on a wide top.

How much an Adelaide summer moves a tabletop

These figures are for a 1,000mm wide solid Tasmanian oak top of mixed grain, which is what most tables actually are. Take them as the right order of magnitude rather than a guarantee for a specific board.

| Indoor relative humidity | Approximate EMC | Width of a 1,000mm top | | --- | --- | --- | | 65% (damp July, unheated room) | 12% | 1,005mm | | 55% (mild spring) | 10% | 1,001mm | | 45% (Adelaide indoor average) | 8.5% | 999mm | | 35% (summer, cooling running) | 7% | 997mm | | 25% (February heatwave, refrigerated cooling) | 5.5% | 995mm |

Total swing: 10mm across the year, on a top most people assume is a fixed object. A flatsawn top in the same conditions moves closer to 15mm. A 1,200mm top moves proportionally more again.

If that top is held at both ends by something that will not move, 10mm has to go somewhere. It goes into the weakest glue line, and it opens.

Why Adelaide is harder on furniture than Melbourne or Sydney

Adelaide is the driest of the mainland capitals. Long-term temperature and humidity records are published by the Bureau of Meteorology, and the pattern is consistent: hot, dry summers with long runs of days above 35 degrees and very little atmospheric moisture.

Indoors it gets more extreme, not less. A house on ducted refrigerated cooling strips moisture out of the air as a side effect of cooling it, and a closed-up house through a February heatwave can hold under 30% relative humidity for weeks at a time. Evaporative cooling does the opposite, pushing humidity up sharply while it runs and letting it collapse when it stops.

Both are hard on restrained timber. A piece glued up at 12% moisture content in a workshop with no climate control, then delivered into a house running at 6%, will move 6mm to 10mm in its first summer. That is the year most failures appear.

The construction that prevents it

None of this is exotic. It is standard practice, it costs hours rather than materials, and it is the difference between a table that lasts 40 years and one that opens in year 2.

Buttons, z-clips and slotted holes

The top must be attached to the base in a way that holds it down firmly while letting it slide sideways. There are 3 accepted methods:

• Buttons. Small timber blocks with a tongue that sits in a groove cut inside the rail. The button screws to the top and slides in the groove.
• Z-clips or figure-8 fasteners. Steel equivalents, faster to fit, equally effective.
• Slotted or elongated screw holes. A screw through a hole widened across the grain direction, with a washer, tightened firmly but not crushed.

What must never happen is a screw driven straight through a fixed rail into the top with no slot. That single shortcut causes more Adelaide table failures than every other cause combined.

Breadboard ends

A breadboard end is the cross-grain strip on the end of a tabletop. It keeps the top flat and hides end grain, and it is the second most common failure point.

Built correctly, it is joined with a tenon running into a mortise, glued only at the centre 100mm to 150mm, and pinned through elongated holes toward each end so the pins can travel as the top shrinks. Built incorrectly, it is glued across its full length. When the top then loses 8mm and the breadboard end does not, the top splits or the end snaps off. Both are visible from across the room.

Finish both faces

A finish slows moisture exchange. If the top face is sealed with 4 coats and the underside is bare, the 2 faces reach equilibrium at different rates and the top cups. Finishing the underside costs perhaps 40 minutes and 1 extra litre of product, and skipping it is a reliable marker of a rushed job. Check under the table before you accept delivery.

Acclimatise before machining

Timber should sit in a conditioned space, stickered so air moves around every board, until its moisture content is stable and measured with a meter rather than judged by feel. For local hardwood that is typically 2 to 6 weeks. This is dead time a workshop pays for and cannot bill as a line item, which is precisely why the cheapest quote often skips it.

What to ask before you order

There are 4 questions, and they take a maker under 2 minutes to answer well.

1. How is the top fixed to the base? The answer must contain the word movement, and should name buttons, z-clips or slotted holes.
2. What moisture content is the timber at now, and how was it measured? The right answer is a number between 8% and 12% and the word meter.
3. Is the underside finished to the same number of coats as the top?
4. If there is a breadboard end, where is it glued? The centre only.

A maker who answers all 4 quickly is telling you how they run every job. Vagueness on questions 1 and 4 is the reliable warning sign. Reference data on the movement behaviour of Australian species is published by WoodSolutions.

What a warranty should actually say

Most furniture warranties exclude movement entirely, which is a reasonable position taken badly. Movement itself is normal and cannot be warranted against. Failure caused by construction that prevented movement is a defect and should be covered.

A good warranty distinguishes the 2. It will exclude seasonal width change, minor gapping at a breadboard end and surface checking from direct sun, and it will cover a glue line that opens, a top that cups, and a joint that fails. If the document simply says "solid timber moves and is not covered", the maker has written themselves out of responsibility for their own construction.

The same rules govern the species you pick, and 2 of them behave quite differently here: the comparison is in blackwood versus Tasmanian oak. Reclaimed stock carries an extra layer of this problem, covered in recycled timber furniture in Adelaide. If a top has already opened, it is repairable: furniture restoration covers what that involves. For material selection across piece types, start at custom timber furniture.