Accuracy, sawing variation and oversizing

Why variation costs money

Sawing variation is the scatter in thickness of boards cut with the same setting. It is caused by saw wander, vibration, feed problems and worn guides. Because the planer needs a minimum green size to reach the final dimension, mills add an oversizing allowance proportional to that variation. Every millimetre of oversize is wood that is sawn, dried and then planed away.

The portal's yield article lists improving this as the first improvement step for a mill:

• Steele (1984, North America) cites an estimate that improving sawing accuracy from 0.2 to 0.1 inch on a machine with a 0.125 inch kerf would raise board-foot yield by 6.0%.
• Kilborn (2002, Alaska) found oversizing at 18 of 23 studies and variation above 0.030 inch at 17, and estimated about 2.5% more recovery per 0.030 inch removed.
• Steele also points out that planing allowances are often excessive and can usually be reduced.

These are North American, older figures; use them for the logic, not as a promise.

Worked example: a Ghanaian band saw

A thesis from KNUST (Ghana, 2002) studied a band saw cutting wawa (Triplochiton scleroxylon) at a mill in Kumasi. Key numbers:

• Target thickness 41.00 mm, mean actual thickness 42.42 mm.
• Oversize: 42.42 - 41.00 = 1.42 mm.
• Oversize share: 1.42 / 41.00 = 0.0346, about 3.5%.
• The thesis translated this into 3.46 m3 lost per 100 m3 of lumber.

The main driver of thickness variation was excessive kerf width: stellite-tipped saws averaged 4.20 mm kerf, swage-set saws 3.06 mm (a difference of 1.14 mm), and variation was lowest at a mean side clearance of 0.69 mm.

The lesson is not that band saws are inaccurate as such, but how much precision depends on setting and maintenance. It is a single mill and species, so do not copy the numbers.

What the Polish comparison says

In the Polish test (2023), thickness standard deviation was about 0.39 to 0.61 mm for the band saw and 0.28 to 0.71 mm for the frame saw. The authors concluded that the frame saw was slightly more accurate because of rigid guidance. The ranges overlap heavily, so this is a weak statement, and the paper has internal inconsistencies. Treat it as an illustration of what to measure, not a ranking.

A simple variation audit for your own mill

1. Choose one machine, one product and one setting.
2. Measure thickness of at least 20 consecutive boards at the same position (a caliper is enough).
3. Record the mean, the target, and the spread (highest minus lowest).
4. Compare the mean with the target: this is your oversize.
5. Repeat after a blade change, after realigning guides, and after adjusting side clearance.
6. Only then discuss reducing the planing allowance.

Quick calculation

If the target is 38.0 mm and the mean is 39.6 mm, oversize is 1.6 mm, or 1.6 / 38.0 = 4.2%. If the spread is 2.4 mm, a large part of that mean offset is probably buying safety margin against variation. Reducing variation lets you lower the mean without producing undersize boards.

Where accuracy problems come from

From the sources in this course, check these in order:

• Kerf and set: excessive set widens the kerf and increases variation (Ghana).
• Side clearance and guides: variation was lowest at one measured side clearance; worn guides allow wander.
• Blade stiffness and tension: thin or poorly tensioned blades deflect (lesson 2 and lesson 4).
• Feed: stability depends on thickness, width, feeding force and guide spacing (OSU Extension).
• Tooth condition: a dull tooth raises force and deflection (lesson 5).

Key points

• Variation forces oversizing; oversizing is lost yield.
• Reported estimates (6.0% for halving variation, about 2.5% per 0.030 inch) are North American and old; the logic travels, the numbers may not.
• In the Ghanaian example, 1.42 mm oversize on 41 mm is about 3.5%.
• Measure mean, target and spread on your own machine before changing allowances.
• Setting, side clearance, kerf and maintenance often matter more than the saw type.