Step-Over Guide

a_e and a_p

Two independent quantities describe how much of the tool is in the cut:

• a_e — radial engagement (width of cut, step-over): how far the tool is stepped into the material sideways, measured perpendicular to the feed direction.
• a_p — axial depth of cut: how deep the tool is engaged along its axis.

The pair defines the strategy. A large a_p with a small a_e is the high-efficiency approach: a narrow, deep cut using the full flute length. A small a_p with a large a_e is the traditional approach: a wide, shallow facing cut. Both remove material; they load the tool in entirely different ways and fail in entirely different ways.

A full slot is the special case where a_e equals the tool diameter. It is the most demanding operation an end mill performs — 180° of arc engagement, no free side for the chips, and heat with nowhere to go — and it is where most breakages happen.

Engagement strategies compared

Percentages are conventional industry starting points, not limits from a specific tool. The permissible depth for any particular tool depends on its flute length, core diameter, geometry and the rigidity of the setup.

The high-efficiency row is the one that changes shop economics. Cutting at 10% radial engagement and full flute depth spreads wear over the whole flute length instead of concentrating it in the bottom few millimetres, keeps arc-of-contact time short so the edge cools between engagements, and — with the chip-thinning correction applied — removes material at a competitive rate at much lower load. It requires a controller and a CAM system that can maintain constant engagement through corners; without that, corner spikes remove the benefit.

Every step-over decision is also a feed decision

Below 50% radial engagement, chip thickness falls below the programmed feed per tooth and the feed must be corrected upwards to keep the edge cutting rather than rubbing. This is not optional refinement — at 10% engagement the uncorrected chip is roughly 60% of the intended thickness, and at 5% roughly 44%.

The Feed Rate Guide carries the full derivation and a correction-factor table. The practical rule: whenever you reduce the step-over, recalculate the feed in the same edit.

Step-over and surface finish — the scallop calculation

On a ball-nose finishing pass, the step-over sets the height of the cusp left between adjacent passes. That cusp — the scallop — is the surface finish, and it is fully predictable from geometry.

Halving the step-over quarters the scallop but doubles the cycle time. Beyond the point where the scallop is smaller than the deflection, runout or thermal movement of the setup, further reduction buys nothing measurable.

The corollary is the useful part: if the drawing calls for a surface the scallop calculation already satisfies at 0.5 mm step-over, running 0.25 mm doubles the finishing time for a difference no instrument in the shop will resolve.

Corners, entries and the moments engagement spikes

Programmed engagement is an average. Several moments in a normal toolpath exceed it substantially:

• Inside corners. Engagement rises as the tool sweeps a corner whose radius is close to its own. A corner radius at least 1.3× the tool radius, or a CAM corner-smoothing strategy, keeps the load steady.
• Straight plunge entries. An end mill entering vertically at full diameter is cutting with the least capable part of its geometry. Ramp or helical entry spreads the load along the flute; helical is generally the gentler of the two.
• Exit burr formation. As the edge leaves the material the supporting wall thins and the chip rolls over rather than shearing. Reducing the feed on exit, or leaving a small finishing allowance, controls it.
• Re-entry into a partially cut region. Any toolpath that returns into an area already roughed can meet a step of unpredictable width. Constant-engagement strategies exist precisely to prevent this.

What limits the depth you can actually take

The axial depth a tool tolerates is set by rigidity, not by the flute length printed in the catalogue. Deflection scales with the cube of unsupported length, so the practical order of interventions is fixed:

1. Shorten the stickout to the minimum the geometry allows — this is worth more than every other change combined.
2. Use the shortest tool that reaches, rather than a long tool held short.
3. Improve the holder: a shrink-fit or high-precision hydraulic holder reduces runout and raises effective stiffness against a worn end mill holder.
4. Reduce radial engagement and correct the feed — this lowers deflecting force while maintaining removal rate.
5. Only then reduce the axial depth.

Reaching for step 5 first is the common error: it lowers productivity without addressing the deflection that caused the problem.