Conveyor Drive Shaft — 4340 Steel, Heat Treated

Why Drive Shafts Crack at the Keyway

A keyway is a rectangular slot cut into a shaft that mates with a key in the hub of a sprocket, gear, or pulley. It’s how you transmit torque from the shaft to the driven component. The problem is geometry. A standard keyway has sharp 90° corners at the bottom of the slot where it meets the shaft’s cylindrical surface. Under torsional load, the stress at those corners is 3.5–4.0 times higher than the nominal stress in the shaft body — that’s the stress concentration factor, Kt.

For a shaft that sees constant, steady torque, this isn’t necessarily a problem — you just size the shaft for the amplified stress. But a conveyor drive shaft doesn’t see steady torque. It sees cyclic loading from start/stop cycles, load variations as product moves onto and off the belt, and vibratory loads from the conveyor chain or belt. At millions of cycles, even a small stress concentration can nucleate a fatigue crack. Once the crack starts, it propagates a tiny bit with each load cycle until the shaft fractures.

The Fix: 0.030″ Keyway Corner Radius

The textbook solution for keyway fatigue is simple — add a radius at the internal corners. A sharp corner has a theoretical Kt approaching infinity (limited only by the material’s grain structure). A 0.030″ radius drops the Kt to approximately 1.8, which is less than half the stress concentration of a sharp corner. The customer’s original shafts had essentially zero radius — the corners were as sharp as the keyway milling cutter left them.

We used a form endmill ground with a 0.030″ corner radius to cut the keyways. This is a more expensive tool than a standard keyway cutter, but it eliminates the need for a secondary radiusing operation and produces a more consistent radius than trying to manually blend the corners with a rotary file. The key still fits — the radius is at the bottom of the keyway where the key doesn’t contact, so the torque transmission geometry is unchanged.

Why 4340 Instead of 1045

The original shafts were 1045 steel — a medium-carbon steel that’s easy to machine and inexpensive. It’s adequate for many shaft applications, but its fatigue endurance limit (the stress level below which it can theoretically run forever) is relatively low, especially in the as-rolled condition the original shafts were supplied in.

4340 is a nickel-chromium-molybdenum alloy steel. Heat treated to 38–42 HRC, its fatigue endurance limit is roughly 2.5× that of as-rolled 1045. Combined with the keyway radius change (which cut the effective stress concentration in half), the new shafts operate at a stress level well below the infinite-life threshold. The customer’s engineering team ran an FEA analysis that confirmed an infinite-life safety factor of 2.8 at the worst-case keyway corner — meaning the shaft would need to see nearly three times its operating load before fatigue becomes a concern.

Machining 4340 at 38–42 HRC

We machined these shafts in the heat-treated condition rather than machining soft and then heat treating. This eliminates any distortion from the heat treat process — quenching a 36″ long shaft can introduce 0.010–0.020″ of bow, which would require straightening and potentially compromise the metallurgical properties. Machining at 38–42 HRC is harder on tooling, but it gives us direct control over the final dimensions.

The bearing journals were turned to ±0.0005″ and finished to 16 Ra using CBN (cubic boron nitride) inserts, which handle the hardened material without the surface finish problems that carbide inserts develop at this hardness range. The journals were then hard chrome plated to provide the wear surface that the bearings actually ride on.

Root Cause Analysis as a Service

The customer originally asked us to “just make the same shaft but stronger.” We could have quoted 4340 in the same design and called it done. Instead, we asked them to send us one of the failed shafts. The fracture surface told the story immediately — classic fatigue beach marks radiating from the keyway corner. The fix wasn’t just a harder material; it was a design change to eliminate the stress riser. We provided the DFM recommendation at no extra charge because it’s the right thing to do, and because a shaft that doesn’t fail is a customer that comes back.