Cutting Granite Without Chipping: Three Chip Types, Three Fixes
Cutting granite without chipping starts with naming the chip. Exit chipping is a slab with nothing behind it, edge chipping is a dull blade, and a corner bite is the geometry of a round blade. The fixes: a cuttable backer board with the depth set to thickness plus 2mm, dressing or changing the blade, and cutting past the corner before squaring it.
Cutting granite without chipping — which step goes wrong?
Name the chip first, because there are three kinds and each has its own cause and its own fix. A, exit chipping: the blade cuts downward, and the moment it breaks through, the last layer under the slab has nothing behind it, gets pushed outward and snaps off. It reads like poor technique, but the slab had nothing behind it. B, edge chipping along the cut: the blade is dull, the diamonds are flattened and the metal bond is not wearing away, so the blade skids, friction heats it up and grit sheds along both sides of the kerf; feeding too hard does the same, one diamond taking too deep a bite. C, corner bite: pure geometry of a round blade, nothing to do with technique. Telling them apart is a matter of location: the slab underside at the exit, the two sides of the kerf, or the internal corner. Name the type first, then choose the fix.
What goes under the slab so the exit does not chip?
The exit side needs something behind it all the way, and slowing down does not replace it. Put a board that can be cut under the slab — plywood, XPS foam or a rubber mat — and set the cut depth to slab thickness plus 2mm so the blade runs 2mm into it. The board has to be in full contact. A few timber blocks dotted underneath leave air between them, and the chip appears exactly over the air. The backer is consumable: move it after each cut, because sawing repeatedly in the same spot means running into a groove that is already open, which is the same as having no backer. Steel and concrete tables cannot be cut into, the blade is ruined on contact, so the cut can only stop 1~2mm short, and that last 1~2mm is snapped off, which means the exit always chips. On those tables, lay a cuttable board first, then talk about technique.
| What is under the slab | Support at the exit | Can the blade cut in | Result |
|---|---|---|---|
| Steel or concrete table | Yes, but the blade cannot touch it | No | Depth stops 1~2mm short, that 1~2mm is snapped off → the exit always chips |
| Plywood or XPS in full contact | Yes | Yes, 2mm into it | Cut face stays flush |
| A few timber blocks | Only along the block lines | Yes | Air between the blocks → the chip lands over the air |
The blade drags and the cut edge sheds grit — time to change it?
Do not reach for a slower feed first; check whether the blade is still sharp. Four field signs, all things you watch rather than numbers you read: it pushes harder than usual; the slurry runs clear because less stone powder is coming off; the kerf is wider than with a new blade and the cut faces show bright friction marks; dry cutting throws denser sparks. Any one of them means dress the blade — run a few passes in a scrap slab or a firebrick to wear off a layer of bond and expose fresh diamond. Dressing is not sharpening the blade; it is removing the worn layer of bond. If it still drags after dressing, or cuts off line, or a segment has a chipped corner, change it. Do not fit a larger diameter: the diameter must not exceed what the machine and its guard allow, and a larger blade raises rim speed in proportion to diameter — 125÷110≈1.14, 14% faster — while the guard no longer covers it. That is a safety rule, not a way to cut quicker.
Does water poured on the slab face count?
No. Water has to reach the kerf, not the surface. One test decides it: milky white slurry must flow out of the kerf entrance continuously; if it does not, the water is not getting in. Too little water sets off three chains. Segment temperature climbs, the bond wears early and diamonds drop before they are spent. The steel core heats unevenly, the blade wobbles, the kerf widens and the edge chips with it. Stone powder cannot escape, the segment loads up, its face is welded shut with powder and the blade skids. So aim the water where the blade enters the cut and judge by continuous slurry. A thin film across the slab face never enters the kerf and does nothing. The deeper the cut and the faster the push, the more powder is ground off per minute, and the water has to rise with it; a blocked water nozzle is a common fault, so check the flow before you start.
Why is a corner always bitten off when cutting an internal right angle?
Because the blade is round. The blade centre has to travel directly above the kerf, and the blade is a disc far larger than the slab is thick; when the cut stops at the corner the centre is still above the corner, so the disc sweeps past it within the slab thickness, and that sweep is the piece that goes missing. That is geometry, not technique. The single-side bite width at the slab face is w=√(2rT−T²), where r is blade radius (diameter÷2) and T is slab thickness. With φ110, r=55, T=30: √(2×55×30−30²)=√2400≈49mm. The 30mm column is the thickness we run; 50mm is a substitution only. Change T and the number changes. The bigger the blade, the wider the bite, which is why a small-diameter blade is easier on corners. Two ways out: cut past the corner point and square it with an angle grinder, or bring in a waterjet if the corner has to come out square in one pass, since a circular blade cannot do it. Internal corners on moulding profiles follow the same geometry, and designing them as a radius or chamfer saves the rework.
| Blade diameter | Blade radius r | Bite on 30mm slab | Bite on 50mm slab |
|---|---|---|---|
| φ110 | 55 | 49 | 55 |
| φ125 | 62.5 | 53 | 61 |
| φ180 | 90 | 67 | 81 |
| φ350 | 175 | 98 | 122 |
Drilling holes and cutting narrow strips — where does the chipping happen?
Core drilling chips on the back face. Use a diamond core bit, enter at an angle to start a locating dimple, then straighten up so the bit does not wander; stop when the tip breaks through the back, turn the slab over and carry on from that hole — skip the flip and the back blows out in a ring. Leave at least one slab thickness of material between the hole edge and the slab edge; on 30mm that means 30mm or more, and that is an empirical starting point, not a code value. The thinner the slab, the smaller that margin works out, but a thin slab is far less stiff, so the same strip snaps more easily and the packing and clamping have to make up for it. Narrow strips chip at the end with nothing under them: whatever is going to fall must be supported, so pack and prop the strip you are cutting off before the blade goes in. Support its full length, not just one end: pack only one end and the far end is still snapped off. On the Maple Red G562 30mm slabs we run, 100 m² minimum and 5 days to ship, holes and narrow strips are trial-cut in the plant first.
Related reading