Every piece of timber you work with is either shrinking or swelling. Right now, wherever it’s stored, it’s adjusting to the air around it. Understanding shrinkage and swelling isn’t optional — it’s the difference between a project that lasts and one that tears itself apart.
In the previous guides, we covered the three axes of wood movement — tangential, radial, and longitudinal — and the reasons behind each. Now it’s time to bring them together.
This guide explains how total shrinkage values are measured, how to read the numbers you’ll find in reference tables, and how to use them to predict what your timber will actually do in the real world.
What Shrinkage and Swelling Actually Mean
<aside> 🖼️
Image placeholder: Shrinkage vs swelling (same process, opposite direction)
</aside>
- Simple before/after diagram of a board changing width with arrows.
- Label: “shrinkage (MC down)” vs “swelling (MC up)”.
Shrinkage is the reduction in size when wood loses moisture. Swelling is the increase in size when wood gains moisture. They are the same process in opposite directions.
Both happen only when moisture enters or leaves the cell walls — the bound water range. Free water sitting in cell cavities has no effect on dimensions.
This means:
- Above the fibre saturation point (roughly 28–30% MC), dimensional change stops — the cell walls are fully saturated and any additional water simply fills empty space
- Below the fibre saturation point, every percentage point of MC change produces a proportional dimensional change
- The relationship is approximately linear between the fibre saturation point and oven-dry (0% MC)
<aside> 📊
Diagram placeholder: A graph showing the relationship between moisture content (x-axis) and dimensional change (y-axis). A flat line above ~28% MC, then a straight diagonal line from ~28% MC down to 0% MC. Label the fibre saturation point clearly.
</aside>
How Shrinkage Is Measured and Reported
<aside> 🖼️
Image placeholder: Green → oven-dry measurement concept
</aside>
- Simple schematic showing a sample measured at green vs measured after oven-dry.
- Label: tangential, radial measurement directions.
When you look up shrinkage data for a species, you’ll typically find values reported as percentages from green to oven-dry. This is the total possible shrinkage — the full range the wood could experience if it went from fully saturated cell walls to completely dry.
Standard reference values include:
- Tangential shrinkage (%) — the percentage change measured along the growth rings
- Radial shrinkage (%) — the percentage change measured from pith to bark
- Volumetric shrinkage (%) — the total volume change (roughly the sum of tangential + radial + longitudinal, though not exactly additive)
- T/R ratio — tangential divided by radial, indicating how uneven the movement is
Longitudinal shrinkage is usually omitted from tables because it’s negligible in normal wood (as we covered in Guide 5).
Example: European Oak
| Property | Value | | — | — | | Tangential shrinkage | ~8.5% | | Radial shrinkage | ~4.5% | | Volumetric shrinkage | ~13.5% | | T/R ratio | ~1.9 |
These numbers tell you that oak moves almost twice as much tangentially as radially, and that it’s a moderate-to-high shrinkage species overall.
Reading the Reference Tables
Shrinkage values vary between sources, sometimes significantly. That’s normal. Here’s why:
- Different test specimens (heartwood vs sapwood, mature vs juvenile wood)
- Different measurement methods (from green, from a specific MC, at different temperatures)
- Natural variability within a species — two trees of the same species grown in different conditions will produce wood with different shrinkage characteristics
What to do with this variability
- Treat published values as indicative, not exact
- Use them to compare species and estimate behaviour, not to predict movement to the nearest tenth of a millimetre
- When multiple sources give different numbers, the range is more useful than any single value
- For critical applications, test your actual stock if possible
<aside> ⚠️
Don’t over-rely on decimal precision. A table might say oak’s tangential shrinkage is 8.5%, but your particular boards might be 7.8% or 9.2%. The important thing is the order of magnitude and the relative behaviour between directions and between species.
</aside>
Shrinkage Coefficients: The Practical Tool
<aside> 🖼️
Image placeholder: From total shrinkage → coefficient
</aside>
- Mini visual showing: Total shrinkage % ÷ FSP ≈ coefficient (% per 1% MC).
Total green-to-oven-dry shrinkage gives you the full picture, but most timber doesn’t swing through that entire range. In practice, you need to know how much a board will move for a given change in moisture content.
This is where shrinkage coefficients come in.
A shrinkage coefficient expresses the dimensional change per 1% change in MC, as a percentage of the dimension.
To calculate it from total shrinkage data:
$$ S = \frac{\text{Total shrinkage %}}{\text{FSP (assume ~28–30)}} $$
Example: European Oak, tangential
- Total tangential shrinkage: 8.5%
- Fibre saturation point: ~28%
- Shrinkage coefficient: 8.5 ÷ 28 = ~0.30% per 1% MC change
This means for every 1% change in moisture content, a flat-sawn oak board changes approximately 0.30% in width.
Using the coefficient to estimate movement
The formula:
$$ \Delta D = D \times \Delta MC \times S $$
Where:
- $\Delta D$ = change in dimension
- $D$ = current dimension
- $\Delta MC$ = change in moisture content (as a decimal, e.g. 4% = 0.04)
- $S$ = shrinkage coefficient (as a decimal, e.g. 0.30% = 0.003)
Worked example
A flat-sawn European oak board, 250mm wide.
Seasonal MC swing: 10% to 14% (a 4% change).
Tangential shrinkage coefficient: 0.003.
$$ \Delta D = 250 \times 0.04 \times 0.003 \times \frac{250}{0.25} $$
Let’s simplify:
- 4% MC change × 0.30% per 1% MC = 1.2% total change
- 1.2% of 250mm = 3.0mm
That board will change by about 3mm across its width over the course of a year. If it’s a tabletop panel made from several boards totalling 800mm, the total cross-grain movement is roughly 9.6mm — almost a centimetre.
Now the same board quarter-sawn (radial coefficient ~0.16% per 1% MC):
- 4% × 0.16% = 0.64%
- 0.64% of 250mm = 1.6mm
Half the movement, same board width, same environment.
Shrinkage Data for Common Species
<aside> 🖼️
Image placeholder: Species stability comparison
</aside>
- Simple chart grouping species into low / medium / high movement.
- Call out beech (high), cedar/teak (low), oak/walnut/pine (moderate).
Here’s a comparison table for species frequently encountered in woodworking:
| Species | Tangential % | Radial % | Volumetric % | T/R ratio | | — | — | — | — | — | | European Oak | 8.5 | 4.5 | 13.5 | 1.9 | | European Beech | 11.8 | 5.8 | 17.9 | 2.0 | | Scots Pine | 7.7 | 4.0 | 12.4 | 1.9 | | Douglas Fir | 7.8 | 4.8 | 12.4 | 1.6 | | Western Red Cedar | 5.0 | 2.4 | 6.8 | 2.1 | | Black Walnut | 7.8 | 5.5 | 12.8 | 1.4 | | Hard Maple | 9.9 | 4.8 | 14.7 | 2.1 | | Cherry | 7.1 | 3.7 | 11.5 | 1.9 | | Ash | 7.8 | 4.9 | 13.2 | 1.6 | | Teak | 4.0 | 2.2 | 7.0 | 1.8 |
What to notice
- Beech is one of the highest-movement species — a flat-sawn beech panel will move substantially more than the same panel in oak or walnut
- Western Red Cedar and Teak are among the most stable — low total shrinkage and good stability, which is partly why they’re favoured for outdoor use
- Walnut has a low T/R ratio (1.4), meaning the difference between flat-sawn and quarter-sawn behaviour is smaller than in most hardwoods — walnut is relatively forgiving
- Hard Maple has a high T/R ratio (2.1), meaning cut orientation matters more
Swelling: The Other Direction
Everything we’ve said about shrinkage applies in reverse to swelling. When timber gains moisture, it expands.
The coefficients are the same. The directions are the same. The ratios are the same.
But there’s an important practical difference: shrinkage opens gaps, swelling closes them.
- A tabletop that shrinks in winter might show gaps between boards
- The same tabletop in a humid summer might swell and push against its frame or fixings
- If there’s no room to expand, something has to give — boards buckle, frames crack, finishes fail
This is why designing for movement means allowing for both directions. A gap that’s right in summer might be too wide in winter. A tight fit in winter might be dangerously tight in summer.
What the T/R Ratio Tells You
<aside> 🖼️
Image placeholder: T/R ratio intuition
</aside>
- Visual showing tangential bar roughly twice radial bar.
- Caption: “Higher ratio = more cupping risk in flat-sawn boards.”
The tangential-to-radial ratio is one of the most practically useful numbers in any species reference.
Low T/R ratio (below 1.6)
- Movement is relatively even in both cross-grain directions
- Less tendency to cup
- More forgiving of ring orientation
- Examples: walnut (1.4), Douglas fir (1.6), ash (1.6)
High T/R ratio (above 2.0)
- Movement is very uneven — tangential much greater than radial
- Strong tendency to cup in flat-sawn boards
- Ring orientation matters a lot
- Quarter-sawing has a bigger payoff
- Examples: Western Red Cedar (2.1), hard maple (2.1), beech (2.0)
Moderate T/R ratio (1.6–2.0)
- Most species fall here
- Normal cupping behaviour in flat-sawn boards
- Quarter-sawing helps but isn’t always essential
- Examples: oak (1.9), Scots pine (1.9), cherry (1.9)
Volumetric Shrinkage: The Whole Picture
Volumetric shrinkage is the total volume change from green to oven-dry. It’s roughly the sum of tangential, radial, and longitudinal shrinkage, though the actual calculation is slightly more complex.
Volumetric shrinkage is useful for:
- Comparing overall stability between species
- Estimating density changes with moisture content (denser species tend to have higher volumetric shrinkage because there’s more cell wall material to swell and shrink)
- Understanding drying behaviour — high-volumetric-shrinkage species are more prone to drying defects
General categories
- Low volumetric shrinkage (<10%): Very stable species like cedar, teak
- Moderate volumetric shrinkage (10–14%): Most common commercial species — oak, walnut, pine, Douglas fir
- High volumetric shrinkage (>14%): Species like beech, hard maple — require more careful handling
Putting It All Together: A Decision Framework
<aside> 🖼️
Image placeholder: Decision flow
</aside>
- Simple flow diagram: species → cut (ring orientation) → environment (EMC range) → estimate ΔMC → calculate movement → choose joinery/fixings.
When planning a project, here’s how to use shrinkage data:
1. Know your species
Look up the tangential, radial, and volumetric shrinkage for your timber. Note the T/R ratio.
2. Know your cut
Check the end grain of your boards. Are they flat-sawn (tangential face), quarter-sawn (radial face), or somewhere in between?
3. Know your environment
Estimate the likely MC range your project will experience. A centrally heated UK home might swing from 8% MC in winter to 14% MC in summer. An outdoor structure might see a wider range.
4. Estimate the movement
Use the shrinkage coefficient and your MC swing to calculate the expected dimensional change across the grain.
5. Design for it
Allow for the movement you’ve calculated:
- Floating panels
- Slotted fixings
- Expansion gaps
- Appropriate joint design
<aside> 📐
The Timber Logic movement calculator (coming soon) will automate steps 1–4. But understanding the principles behind the numbers is what separates good design from guesswork.
</aside>
Common Mistakes This Knowledge Prevents
- Gluing a solid wood panel to a rigid substrate — the panel needs to move, and if it can’t, it will crack or the glue joint will fail
- Not accounting for seasonal change when fitting drawers or doors — a drawer front that fits perfectly in summer may be too tight in winter (or vice versa)
- Choosing beech for a wide, unsupported surface without realising it’s one of the highest-movement species available
- Ignoring the T/R ratio when choosing between flat-sawn and quarter-sawn stock — for species with a high ratio, the difference is dramatic
- Using green-to-oven-dry values as real-world predictions — your timber won’t experience that full range in service; use shrinkage coefficients with your actual expected MC swing
The Simple Rule
<aside> 💡
Look up the shrinkage values for your species. Calculate the movement for your board width and expected moisture swing. Then design to accommodate it — not fight it.
</aside>
What’s Next
We now understand how much timber shrinks and swells, and how to calculate it. But what drives those moisture changes in the first place? In Guide 7 — How Humidity Affects Wood, we’ll explore the relationship between the air around your timber and the moisture inside it — and why a centrally heated room in winter is one of the harshest environments a piece of wood can face.
🔗 Knowledge Network
Species Pages
- European Oak — moderate shrinkage, T/R ratio ~1.9
- European Beech — high shrinkage, T/R ratio ~2.0
- Scots Pine — moderate shrinkage, T/R ratio ~1.9
- Douglas Fir — moderate shrinkage, low T/R ratio ~1.6
- Western Red Cedar — very low shrinkage, highly stable
- Black Walnut — low T/R ratio ~1.4, forgiving movement
- Hard Maple — high T/R ratio ~2.1
- Cherry — moderate shrinkage, T/R ratio ~1.9
- Ash — moderate shrinkage, low T/R ratio ~1.6
- Teak — very low shrinkage, excellent stability
Glossary Terms
- Shrinkage
- Swelling
- Shrinkage Coefficient
- Volumetric Shrinkage
- T/R Ratio
- Fibre Saturation Point (FSP)
- Tangential
- Radial
Calculators
- Movement Calculator — explicitly referenced as “coming soon” in this guide; uses shrinkage coefficients covered here
Categories
- Shrinkage and swelling
- Shrinkage coefficients
- Fibre saturation point
- T/R ratio
- Species stability
- Wood movement basics
Related Guides
- Track 2 – Guide 4 – Tangential vs Radial Movement — the directional movement this guide quantifies
- Track 2 – Guide 5 – Longitudinal Movement (and Why It’s Small) — the third axis (negligible but important to understand)
- Track 2 – Guide 7 – How Humidity Affects Wood — what drives the MC changes behind shrinkage and swelling
- Track 2 – Guide 3 – Why Wood Moves — the cellular mechanism behind dimensional change
- Track 3 – Guide 9 – Stability Differences Between Species — species-level stability comparisons