Why Two Diamond Blades With the Same Drawing Cut Differently

A buyer sends us a drawing and a complaint. The blades match the drawing exactly: 350 mm diameter, 25.4 mm bore, 10 mm segment height, 3.2 mm segment width, laser-welded. And yet the last two batches cut noticeably slower than the first, and they feel “slippery” on the granite.

Nothing on the drawing is wrong. Everything that matters is missing from it.

What the drawing does not contain

A diamond blade is a steel core with a composite cutting layer. The core geometry is fully specified by the drawing. The cutting layer is specified by four numbers, usually not on it:

Diamond grade and concentration
Metal bond composition
Sintering parameters (temperature, pressure, time)
Diamond coating, if any

Two factories can produce segments that measure identically and behave completely differently, because the difference is in the metallurgy of the bond — and the bond is what decides whether the diamonds are held long enough to cut, or released too early, or held so long that they glaze.

The glazing problem

The failure mode in the buyer’s complaint sounds like glazing. The segment develops a polished, shiny surface where the diamonds have worn flat instead of fracturing and exposing fresh cutting edges.

Glazing happens when the bond is too hard for the material being cut. The bond does not wear back at the rate the diamonds dull, so dull diamonds stay in place and rub rather than cut.

On granite, this is the most common specification error. Granite is hard and abrasive; the bond needs to wear at a rate matched to the diamond’s fracture rate. Too hard, and the blade glazes and stops cutting. Too soft, and the segment is gone before the diamond is spent.

The “slippery” feel is the symptom. The blade is not cutting; it is burnishing.

Why the first batch was fine

Three possibilities, in order of likelihood:

The factory changed bond supplier. Metal bond powder is bought in batches. If a supplier changed, or the factory reformulated to hold a price, the hardness shifts. Nobody tells the buyer, because the drawing still matches.

The sintering parameters drifted. Slightly higher temperature or longer hold time increases bond hardness. This is often invisible in the factory’s own records because they record setpoints, not measured outcomes.

The granite changed. Different quarry, different mineralogy. A bond that was correctly matched to one granite can be too hard for another. Buyers usually rule this out early; it should be ruled out last.

How to specify bond hardness

You cannot put a Rockwell number on a metal bond working layer meaningfully — it is a composite, not a homogeneous metal. What you can do:

Specify by performance, with a reference. “Must achieve at least X linear metres on Y granite with Z machine” is enforceable. “Bond hardness HRB 95” is not.

Hold a reference sample. A sealed segment from the approved batch, retained by both parties. New batches are compared against it, not against the drawing.

Test on arrival, not after installation. A simple field test: cut a known material for a fixed number of metres and measure segment wear. If wear is below the expected band, the bond is too hard — the blade will glaze in service.

What we do about it

For any blade programme that runs beyond a single order, we hold a reference segment and a reference cut report. Every subsequent batch is checked against both. When a factory cannot explain a deviation, we stop the shipment rather than let the buyer discover it three months later.

The drawing is the starting point. It is not the specification.

Sourcing cutting tools and hitting this problem? Send us the application details — we will tell you what to specify beyond the drawing.