Saw Kerf Compensation in Stone Cut Lists

A cut list is a list of finished parts. A slab layout is a plan for turning one piece of stone into those parts. The gap between the two is kerf — the width of material the blade turns into dust on every pass. Shops that treat kerf as a rounding error get away with it on a two-piece vanity and get burned on a kitchen with a long run of pieces nested down one slab. The last part comes up short, and nobody can explain where the material went.

It went into the slurry, one cut at a time. Kerf is not an abstraction; it is a physical quantity you can measure in your own shop in about ten minutes, and it is different for every blade on your rack. The catalog figure is a starting point, not your number. This article covers how to measure real kerf, where to put the allowance in a layout, how it drifts as a blade wears, and how the behaviour changes when the same job moves between a bridge saw, a CNC and a waterjet.

What Kerf Is and What It Costs

Kerf is the total width of the slot the tool cuts, not the thickness of the steel core. A diamond blade removes more than its core because the segments are set slightly wider than the plate, giving side clearance so the core does not bind or drag in the cut. That extra width is deliberate and necessary. It also means the cut you get is always wider than the plate you measured with a caliper at the arbor, which is one reason shop estimates of kerf tend to run low.

Bridge saw diamond blades for stone commonly run roughly 0.090 to 0.110 inch of kerf, and segment heights for granite production are commonly in the 20 to 26 mm range, though both vary by blade and manufacturer. Granite sits around 6 to 7 on the Mohs scale, hard enough that segment bond and geometry are chosen for the material rather than for convenience. Engineered quartz is a separate case and requires diamond tooling rated for engineered stone, not a general-purpose granite blade pressed into service.

A tenth of an inch sounds trivial until you count cuts. Nest eight parts down a slab and you have seven internal cuts plus a trim at each end. At roughly a tenth of an inch apiece that is close to an inch of material consumed before anyone has made a mistake. If your layout software assumed zero kerf, that inch has to come out of something, and it comes out of the last part in the sequence — the one already cut to size in everybody’s head but not yet cut in the stone.

The failure is worse than a single short part, because it usually shows up at the end of a job when the remaining material is committed. You either recut from a second slab, which destroys the yield you planned for, or you shave a seam location and hope the layout still lands where the template said. Neither outcome is cheap. Measuring the kerf once and building it into the layout costs a scrap offcut and a set of calipers.

Measuring and Applying the Allowance

There are two separate skills here, and shops usually have one without the other. The first is establishing the real kerf of the specific blade on the specific saw. The second is knowing where in the layout that number belongs, which is not simply “add it everywhere.” Getting the first right and the second wrong still produces short parts, so treat them as a pair.

Measure the Kerf You Actually Have

Take a piece of scrap in the material you are running, mark a line, and make a single straight cut all the way through at your normal feed and depth. Keep both halves. Butt them back together on a flat surface with the cut faces aligned as they were before the pass, and measure the gap that remains against the original mark with a caliper or feeler set. That gap is your kerf. It includes side clearance, any blade wobble, and whatever the saw itself contributes.

Do it more than once. Take a reading near the entry of the cut and another near the exit, and repeat the whole test on a second piece. If the numbers disagree by a meaningful amount you have found something worth knowing: a worn arbor, a flange that is not seating, a blade with runout, or a feed rate that is deflecting the plate. A blade that cuts a wider slot at the end of the pass than at the start is telling you about deflection, not about kerf.

Put the Allowance Where It Belongs

Inside a slab, every line between two parts consumes one full kerf. If two parts share a cut line, the allowance sits between them and both parts lose material to the same pass. At the slab edge, the situation is different: a trim cut removes kerf plus whatever irregular edge you are cutting away, and only one side of that cut becomes a part. Layouts that apply the same allowance to edge trims and internal cuts overestimate the edge and underestimate nothing, which at least fails safe.

The dangerous case is the opposite: a layout drawn as though parts butt directly against one another with no gap at all. On paper the slab holds everything. In stone it does not, and the shortfall accumulates. Stacking error is the term for it — each cut is fine individually, but the sum of the allowances you did not book shifts every subsequent part along the slab until the last one runs off the end.

Record It So It Survives the Week

A measured kerf that lives in one person’s head is worth very little. Number your blades, write the measured value on a card at the saw, and enter it in whatever field your layout or nesting software provides. Note the date and the material it was measured in. When a blade is swapped, the card gets updated before the next job is laid out. This is a two-minute habit that eliminates an entire class of expensive surprises.

Kerf Across Bridge Saw, CNC and Waterjet

On a bridge saw the operator is the compensation system. The blade follows a line, and whether the part ends up on-size depends on which side of that line the blade runs. Cutting to the line, on the line, or leaving the line are three different finished dimensions separated by roughly a full kerf. Shops that have run the same saw for years do this without thinking, which is exactly why it breaks when a new operator or a new blade arrives.

A CNC saw applies the offset in software, which removes the operator error and introduces a data error in its place. The control offsets the toolpath by the value stored in the tool table, and it will do that faithfully whether or not the stored value matches the blade currently on the spindle. Swap a blade without updating the entry and every part on that program is wrong by the difference, consistently and invisibly, across the whole job.

Router bits on a CNC are simpler in principle: the kerf equals the bit diameter, and CAM offsets the path by half of it to leave the part on-size. The complication is that a bit measured at nominal diameter and a bit that has been running for weeks are not the same tool. Profiling and cutout bits wear, and a worn bit leaves stock the finishing pass was never told about, which shows up as a part that is fractionally oversize on the outside dimension.

Waterjet kerf is narrower than a diamond blade and it is not perfectly parallel through the thickness — the stream is slightly tapered, so the entry side and exit side of the cut are not identical widths. For most fabrication that taper is small enough to ignore, but it matters where two waterjet-cut edges have to mate tightly, and it matters on thicker material. Nesting on a waterjet can be tighter than on a saw precisely because the kerf is narrower.

Problems multiply when one job crosses machines. A layout nested for waterjet kerf and then cut on a bridge saw will not fit the slab. Parts rough-cut on the saw and finished on the CNC need the saw allowance to leave enough stock for the router pass, not just enough to hit the finished dimension. Decide at layout time which machine cuts which line, and apply that machine’s allowance to that line.

Where the wrong allowance really shows is at a seam. Two mitred or butted pieces that were each cut a half-kerf off in the same direction leave a gap that has to be filled rather than closed, and on a book-matched or veined slab a shifted cut line also shifts the pattern match. Finished dimensions against a wall behave the same way: a countertop cut a kerf short leaves a gap the backsplash has to hide, and a kerf long does not go in at all.

Wear, Dressing and Keeping the Number Honest

Kerf is not a constant across a blade’s life. Segments wear on their faces and their sides, and the side clearance that produced your measured slot on day one is not what you have after months of production. The direction and size of the drift depend on the blade, the bond, the material and how hard the saw is being pushed, which is another way of saying you should measure rather than predict it.

Dressing changes things again. Running a blade through a dressing stick re-exposes diamond and restores cutting action, and the blade that comes out of that operation is not cutting the same way it was ten minutes earlier. If you are about to lay out a tight nest with very little spare slab, take a fresh test cut after dressing rather than assuming last month’s card is still valid.

Different blades on the same saw need different allowances, and this catches shops out constantly. A thin-kerf blade kept for delicate work, a heavy production blade for granite, and a blade rated for engineered stone will each cut their own slot on the same machine. One number written on the wall by the saw is one number too few. The card should list every blade in rotation with its own measured value.

Build re-measurement into events rather than the calendar, because events are what change the number. New blade, dressed blade, new material type, a saw that has just been serviced, or a job where the yield is tight enough that a tenth of an inch decides whether it fits. Any one of those is worth a scrap cut and two minutes with a caliper before the first slab goes on the bed.

The payoff is yield. A shop that books kerf accurately can nest closer to the real limit of the slab, because the layout it draws is the layout the saw will actually produce. A shop that guesses has to leave margin everywhere to protect against being wrong, and that margin is material bought and thrown away. Over a year of kitchens the difference between those two habits is measured in slabs, not in inches.

Blade choice drives both the kerf you get and how stable it stays, so it is worth reviewing the blades and cutting tools we carry against the materials you actually run rather than defaulting to whatever the last operator ordered. Our shop guides and technical articles cover the related setup work — flange condition, feed rates and blade care — that keeps a measured kerf figure meaningful from one job to the next.